I U

1035-

BULLETIN OF

THE BRITISH MUSEUM

(NATURAL HISTORY)

-1 OCTW74 ]

ZOOLOGY

Vol. 19
1970

BRITISH MUSEUM (NATURAL HISTORY)
LONDON: 1974

DATES OF PUBLICATION OF THE PARTS

No. I . . . . 30 January 1970

No. 2 . . . . . 27 February 1970

No. 3 ..... 18 March 1970

No. 4 . . . -23 April 1970

No. 5 . . . . . -23 April 1970

No. 6 . . . . . -19 June 1970

No. 7 . . . . . .21 July 1970

No. 8 . . . . . .21 July 1970

No. 9 . . . . . .21 July 1970

Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 5NU

CONTENTS

ZOOLOGY VOLUME 19

No. i. Notes on the family Amphiuridae (Ophiuroidea). By AILSA M.

CLARK i

No. 2. A further systematic study of the genus Acarus L., 1758 (Acaridae,
Acarina), with a key to species. By DONALD ALISTER GRIFFITHS
(Pis. 1-4) 83

No. 3. Skull and swimbladder connections in fishes of the family Megalopidae.

By PETER HUMPHRY GREENWOOD (Pis. 1-3) 119

No. 4. Fourteen new species of freeliving marine nematodes from the Exe

estuary. By RICHARD MARTYN WARWICK 137

No. 5. The anatomy and taxonomy of Indostomus paradoxus Prashad &

Mukerji. By KEITH EDWARD BANISTER 179

No. 6. New records, and one new species, of teleost fishes from the Gilbert

Islands. By GEOFFREY PALMER 211

No. 7. A review of the South American iguanid lizard genus Plica. By

RICHARD ETHERIDGE 235

No. 8. On the genus Ly copter a and its relationship with the family Hiodonti-

dae (Pisces, Osteoglossomorpha) . By PETER HUMPHRY GREENWOOD 257

No. 9. A redescription of the species of Eupolybothrus Verhoeff S. Str.
preserved in the British Museum (Natural History) and the Hope
Department of Zoology, Oxford (Chilopoda, Lithobiomorpha) . By
EDWARD HOLT EASON 287

Index to Volume 19

NOTES ON THE FAMILY
AMPHIURIDAE (OPHIUROIDEA)

AILSA M. CLARK

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. i

LONDON: 1970

NOTES ON THE FAMILY AMPHIURIDAE
(OPHIUROIDEA)

BY

AILSA M. CLARK

Pp. 1-8 1 ; ii text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. i

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is issued
in five series corresponding to the Departments of the
Museum, and an Historical series.

Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be com-
pleted within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19, No. i of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 30 January, 1970 Price 2

NOTES ON THE FAMILY
AMPHIURIDAE (OPHIUROIDEA)

By AILSA M. CLARK

THIS paper deals primarily with the family Amphiuridae in Mortensen's sense, that
is excluding Ophiactis and its allies, although it was necessary to include Amphiactis,
mainly because of a nomenclatorial problem. The relationship between the Amphi-
uridae and the Amphilepididae is reviewed with the result that the Amphilepididae
is reduced to the rank of a subfamily. The taxonomic weight of various characters
which have been used in grouping the species of amphiurids is assessed and a modified
formula for expressing oral armament concisely is introduced.

My thanks are extended to the International Federation of University Women for
a grant which enabled me to study important collections of amphiurids in the United
States, providing much of the background to the present work.

The family Amphiuridae now consists of over four hundred nominal species of
ophiuroids, most of them living on or burrowing in more or less muddy sand or
gravel and some of them very delicate, easily losing the disc, so that a few species
are known only from incomplete specimens. The family has recently (1962) been
the subject of a revision by Fell, who has divided the larger genera Amphiura,
Amphiodia, Amphipholis and Amphioplus into groups of more manageable size which
are unfortunately rather artificial in my opinion.

The family was established in 1867 by Ljungman, though with the now extraneous
subfamilies Ophiacanthinae and Ophionereidinae included (since raised to family
rank), while his subfamily Amphiurinae comprised five divisions, of which the present
Ophiactidae made two (Hemipholis being separated from Ophiactis and Ophiopholis] ,
Amphilepis a third and the amphiurids proper the other two : Amphipholis, Ophio-
phragmus and Ophiostigma in one and Ophiocnida, Amphiura, Ophiopeltis and
Ophiocentrus in the other.

Lyman (1882) in his " Challenger " report merged Amphipholis and Ophiopeltis
with Amphiura but recognized Ophiocnida, Ophiophragmus, Ophiostigma and
Ophiocentrus, as well as Ophionema and Ophionephthys, the last two having been set
up by Liitken (1869).

In 1899 Verrill subdivided Amphiura on the basis of differences in the number
and arrangement of the oral papillae, reviving Amphipholis Ljungman and estab-
lishing two new genera, Amphiodia with three (rarely four) oral papillae on each side
of each jaw, all of them arising from the dental and oral plates, with no papilla on
the adoral shield and Amphioplus with four or five oral papillae, the outermost one
at least arising from the adoral shield.

4 AILSA M. CLARK

Although he recognized the distalmost oral papilla (sometimes also the third one)
as representing the scale or scales of the distal or second oral tentacle, Verrill makes
no special mention in his keys to the occurrence of the first oral tentacle scale in his
two new genera. When present, this scale arises at a higher level in the mouth slit
above the oral papillae. 1

FIG. i, a-d. Amphilepis norvegica (Ljungman), B.M. reg. no. 98.5.3.688, Trondhjem Fjord,
Norway ; a. Oral plate in adradial view (above) and abradial view (below), b. dental
plate, c. dorsal (internal) view of a whole jaw and d. adradial view of projecting part of
intact jaw (compare with usual ventral view, fig. 8f). e. Amphiura chiajei Forbes,
98.5.3.700, same locality, dorsal view of jaw for comparison with c. f-m. Adradial
views of projecting part of intact jaws of : f. Amphiura filiformis (O. F. Miiller), 72.2.3.156,
Vigo Bay, Ireland ; g. A mphipholis squamata (Delle Chiaje), 1937.12.31.1, Beachy Head,
Sussex ; h. Amphioplus abditus (Verrill), 90.8.23.43, Martha's Vineyard, two jaws of same
specimen (see fig. gf for ventral view) ; i. A. archeri A.M.C., 1956.5.23.1, paratype,
Ghana, two jaws of same specimen, the left-hand one with the only oral tentacle scale ;
k. A. depressus (Ljungman), 82.12.23.375, Fiji ; 1. A. integer (Ljungman), Morrumbene
Estuary, Mozambique ; m. Amphiodia occidentalis (Lyman), 98.11.28.10, Queen Charlotte
Is., B.C. [The oral tentacle scale(s), where present, is stippled and the infradental
papillae cross-hatched.]

1 This scale or papilla has been given a variety of names by ophiuroid specialists and even by the same
person at different times. The following people have called it the " first oral tentacle scale (or papilla) " :
Lyman, Verrill, Koehler in 1905, H. L. Clark, Mortensen and Thomas. Others, including Matsumoto,

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDE A) 5

Subsequently other specialists, including Matsumoto (1915 and 1917), Djakonov
(1954) and Fell (1960) in their keys to the genera of Amphiurids, have included the
presence of the first oral tentacle scale as characteristic of Amphioplus as opposed
to its absence in Amphiodia. In fact a number of species which do not comply with
this requirement have been referred to these genera, particularly as regards Amphio-
plus, which is subdivided accordingly further on in this paper.

After VerriU's the next important contribution to the study of the family was that
of Matsumoto (1917) who devised a formula to express the occurrence of the oral
papillae on each side of a jaw. In his scheme the roman number " I " represents
the infradental papilla (the pairing of which at each apex is characteristic of the
family), " II " is the superficial papilla or papillae on the side of the oral plate and
" III " the papilla or papillae on the adoral shield. By this formula the oral papillae
of Amphioplus are +1+11+111 and those of Amphiodia +I+II III. However,
Matsumoto made the proviso that " when a papilla arises partly from the oral plate
and partly from the adoral shield it is referred to the second group " (i.e. II). Since
such a dual origin is shown by many of the species of Amphiura, particularly those
with the distal papilla spiniform unlike that of A. chiajei the type-species, then
Matsumoto's designation of Amphiura as +1 11+ III is not always correct,
although it does express the discontinuity of the series of oral papillae so character-
istic of the genus. It seems to me that the proviso should be ignored and group III
expanded to include species with the distal papilla arising from the distal end of the
oral plate. However, a more elaborate formula is really needed to help pinpoint
the position and occurrence of the individual oral papillae more precisely, as well
as to cover the oral tentacle scale. Unfortunately Matsumoto's formula does not
lend itself to subdivision and qualification. Accordingly I put forward the following
alternative devised by Dr. J. P. Harding, Keeper of Zoology. The examples given
after each paragraph show the evolution of the formula stage by stage.

1. The individual papillae are represented by the letter m, so that most Amphiura
species can be expressed as mm and Amphioplus as mmmm (see figs. 2, 3 and 9).

2. The three main areas involved, corresponding to Matsumoto's I, II and III,
namely the apex of the jaw below the dental plate, the side of the oral plate and the
edge of the adoral shield, are delimited from each other by two commas, thus most
species of Amphiura become m,,m and Amphioplus can be m,mm,m.

3. When a papilla arises jointly from two adjacent loci, for instance at the
junction of oral plate and adoral shield, then the m is split up into two ns with the
comma between and for emphasis a linking bar above, thus m.ri^n for certain species
of Amphiura which cannot be shown as m,,m and m,mn^nm for most species of
Amphioplus. (There are also a few species of Amphioplus in which the fourth and

Djakonov, Chang and Fell, have used " internal ", " lateral ", " additional " or " supplementary
papilla ", while Koehler in 1914 used " intermediate " but in 1922 (possibly altered in translation?)
called it the " second " papilla. I propose to use the term " oral tentacle scale " even though it is
normally more papilliform than squamiform. It seems unnecessary to specify that it is the first such
scale since the one or two corresponding scales of the second oral tentacle on the edge of the adoral shield
and/or at its junction with the oral plate are universally regarded as oral papillae. However, it should
be pointed out that in a few species of Amphioplus it becomes more or less superficial in position, in some
individuals fitting in to the series of oral papillae so as to become almost indistinguishable from them in
ventral view.

6 AILSA M. CLARK

distalmost oral papilla is partly based on the first ventral arm plate ; for these it is
necessary to add a third comma and to split the last m as well, thus m,mnjnnji but
the great majority of species can be dealt with without this refinement.)

4. If there is a significant gap in the series of papillae this is indicated by the
insertion of an o, so that for Amphiura the formula becomes m,o,mo (or m,on^no).

5. The presence or absence of the oral tentacle scale can be shown by -f- or t,
which can be added at the end of the formula, thus for Amphioplus it is usually
m,mnjim +t and for Amphiodia m,mm,o t. Since the oral tentacle scale is rarely
in sequence with the oral papillae but normally arises at a higher level it seems better
to separate it from the main formula in this way. However, there are a few species,
notably of Amphioplus, and the type-species of Ophionephthys, in which some
individuals have a more or less superficial second papilla lateral to the infradental
and clearly homologous with the oral tentacle scale, since on some jaws it is set at a
higher level. This is indicated by equating t with the first m in group II, which can
be linked within brackets ; thus in Amphioplus abditus (figs. I h and 9 f-i) the formula
may be m,(m = t)m,onji for some series but m,om,onji +t or m,mm,ori^n +t for
others.

6. Finally, if one papilla is enlarged, this can be shown by the use of capital
letters, thus m,o,M +t for Amphiura chiajei or mediterranea or m,mN,Nm + or t
for Amphioplus hastatus (see figs. 2] and gp,q).

As a parallel to this formula, it may be of use in discussing the genera of Amphiurids
if the four main arrangements of oral papillae, those of Amphiura, Amphiodia,
Amphipholis and Amphioplus, are designated as A to D respectively, as follows:

A Amphiura (usual formula m,o,mo +t)

B Amphiodia (usual formula m,mnji t)

C Amphipholis (usual formula m,mN,N t)

D Amphioplus (usual formula m,mnjim +t)

Most of the genera of the family can then be simply qualified by one letter or the
other, for instance Amphiacantha (D), having Amphioplus-type oral armament,
although a few such as Ophionephthys cannot be fitted into any one category.

Although this particular sequence of genera with two, three and four oral papillae
is the logical one to adopt, it must be emphasized that Amphiura and Amphioplus
are probably more closely related to each other than either is to Amphiodia or
Amphipholis. This is discussed further under the heading of Amphioplus.

Until 1962 the most important characters used for separating many amphiurid
genera were those furnished by modifications of the disc. For instance Ophiocentrus
(A), Ophiocnida (B), Ophiostigma (C) and Amphiacantha (D) have spinelets on some
of the disc scales, while Acrocnida (A), Amphichondrius (C) and Paracrocnida (D)
have the ventral disc scales either thickened and themselves almost granuliform
(Acrocnida and Paracrocnida} or else covered with superficial granules (Amphichon-
drius). However, Thomas (1962) has cast some doubt on the taxonomic value of
some disc modifications, notably the development of the marginal " fence " of
papillae in Ophiophragmus (B), since he has observed some West Indian individuals
in which this is completely lacking, so that they are superficially indistinguishable

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 7

from Amphiodia. Since the West Indian species concerned are very liable to lose
and regenerate their discs, like a number of other Amphiurids, I think it important
that a proper study be made of the resultant disc modifications. Almost certainly
the primary rosette with enlarged central and five radial plates is lost in regeneration
and the proportions of the radial shields may be altered, though possibly the latter is
only transitory; just how much any armament is affected remains to be seen.

In his revision of the four largest genera of the family, Fell (ig62a) attempted to
subdivide them firstly on the basis of another character provided by the disc namely
the extent of the disc scaling, which in some species is reduced so that the skin is
naked, semitransparent and appears dark in preserved specimens and secondly by
the number of tentacle scales.

More or less extensive lack of scaling on the disc is not an entirely new character
for distinguishing genera in this family, extreme reduction of the scaling having been
used as the main distinction of Ophionema and Ophionephthys by Liitken (1869) as
well as of Ophiopeltis Diiben and Koren (1845), though the last-named was for eighty
years regarded as a synonym of Amphiura until revived by Fell.

It seems to me best to consider first the use of these two characters as they bear
on the subdivision of Amphiura and then to go on to deal with the other nominal
genera in the light of this study.

AMPHIURA Forbes

Amphiura Forbes, 1843 : 149-150 ; Liitken, 18593 : 54-55 ; Ljungman, 1867 : 318 ; Lyman,
1882 : 122-126 ; Verrill, 18993 : 24, 25-26 ; 18990 : 306-308 ; Matsumoto, 1917 : 194 ;
Fell, 19623 : 4-5, U-I2 ; 19620 : 81. Type-species : A. chiajei Forbes, 1843, designated by
Verrill, 1899, Lyman's earlier (1865) designation of A . filiformis (O. F. Miiller) being invalid,
this not being among the three species included by Forbes, namely Amphiura florif era Forbes
(a synonym of A. chiajei according to Liitken (1869 : 75) followed by Ljungman), A, neglecta
(a synonym of A mphipholis squamata according to Liitken) and A . chiajei itself.

Fell's subdivision of Amphiura on the basis of the extent of disc scaling and the
number of tentacle scales can be represented in a grid as shown below.

Tentacle scales

o i 2

{fully scaled Nullamphiura Monamphiura Amphiura

partly scaled Icalia Pandelia Hemilepis

more than half naked Ophiopeltis 1 Amphinephthys

Apart from Amphiura itself, all these names were new except for Ophiopeltis and
Hemilepis. Ophiopeltis was described with the rank of a genus by Diiben and Koren
(1845) but Hemilepis was a name first used by Ljungman (1871) together with
Ophiopeltis (spelled Ophiopelte] for two groups of Atlantic species of Amphiura with
partial or almost complete lack of disc scaling respectively ; these names might

1 Ophionema Liitken was not taken into consideration by Fell but falls into the same section of the
grid as Ophiopeltis.

AILSA M. CLARK

FIG. 2. Dorsal and ventral views of parts oiAmphiura spp. a and b. A . semiermis Lyman,
M.C.Z. no. 1343, " Blake " st. 44, West Indies ; c and d. A. brachyactis H.L.C., M.C.Z.
4933. holotype, Broome, NW Australia ; e and f. A. abbreviata Koehler, Copenhagen
Museum collection, Amboina ; g and h. A. microsoma H.L.C., M.C.Z. 3736, holotype,
Torres Strait area (there are exceedingly tenuous disc scales in the stippled area, which
are impossible to outline without over-emphasizing their distinctness, as well as those
shown around the radial shields) ; i and j. A. mediterranea Lyman, M.C.Z. 1330, syntype,
Nice, S. France ; k and 1. A. stepanovi Djakonov, M.C.Z. 1346 (labelled as syntype of
A. stepanovii Tscherniawsky, a nom. nud ; d.d. only 3 mm.), Sevastopol, Black Sea.
[In e and g the separation of the radial shields is probably unnatural and due to shrinkage
in preservation.]

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 9

therefore be interpreted as of subgeneric rank as used by Ljungman. Lyman (1882)
ignored them when dealing with Amphiura in his " Challenger " report but Verrill
(1899 a and b) included both names to cover small groups of species of Amphiura
(again without precisely designating them as subgenera) ; also Koehler (1914)
mentioned Hemilepis only as a " group " name, while Matsumoto (1917) inferred
that it is synonymous with Amphiura, together with Ophiopeltis and Ophionephthys.

Because of the great number of species of Amphiura, the usual practise when
describing new ones has been to restrict comparisons to other species with the same
combination of tentacle scale number and disc covering, some specialists having
gone so far as to call these similarly-endowed species a " group ". In keys also these
two characters have been used for convenience to split up the large number of
species involved. However, from the time of Lyman's " Challenger " monograph
(1882) until Fell's paper of 1962, no-one has regarded the groups so formed as
sufficiently distinct to warrant even a subgeneric distinction.

The taxa of Fell's with reduced disc scaling may be considered first.

Of the four species included in Hemilepis by Ljungman, Fell designated Amphi-
ura semiermis Lyman, 1869 (fig. 2a, b) as type-species and referred twenty other
species to the genus. A . semiermis has rather indistinct scales on the dorsal side of
the disc and although in some specimens these cover the whole dorsal side stopping
at the margin, in others they are lacking interradially and reduced centrally too.
Koehler's photograph (1914, pi. iv, fig. 6) shows that the holotype of Amphiura
latispina Ljungman, i867b (Hemilepis according to Fell) also has the dorsal scaling
reduced, the interradii being bare (but for a few isolated scales) and the central scales
also inconspicuous so that the scaling appears to be limited to the five radial areas
around the pairs of radial shields. This condition of the disc scaling is at variance
with Fell's diagnosis of Hemilepis as having the disc " scaled above but wholly or
partly naked below " but approaches his Amphinephthys in which the disc is " naked
above and below save for the radial shields and a narrow zone of scales bordering the
shields ".

Amphinephthys itself is inconsistent among Fell's other nominal genera, as he
himself admits, since it spans two spaces in the grid, including species such as the
type, Amphiura crossota Murakami, 1943, with two tentacle scales and also A.
microsoma H. L. Clark, 1915 (fig. 2g, h) with only one scale. Amphinephthys is
otherwise distinguished by having the disc covered with skin except for scales around
the radial shields. I have not seen any material of A. crossota but have been able
to examine the holotype of Amphiura microsoma, which does have scales in the
skin of the disc, though these are so fine and nearly transparent that when dried the
skin turns dark and appears to lack calcification. Also H. L. Clark (1938) has
referred small specimens with fully-scaled disc to the species. The type specimen
of the closely related (if not conspecific) nominal species Amphiura brachyactis
H. L. Clark, 1938 (also from Australia) (fig. 2C, d) has slightly more opaque disc
scales, though they are still obscured in the skin. Amphiura diacritica H. L.
Clark, 1938, which has two tentacle scales like A. crossota, was also referred to
Amphinephthys by Fell, although the type specimen was described as having obscured
scales in the ventral skin of the disc, though not in the dorsal side.

io AILSA M. CLARK

Another species which Fell referred to Amphinephthys (though doubtfully) was
Amphiura polyacantha Liitken and Mortensen, 1899, from the East Pacific off
Panama, his doubt being aroused by the reversal of the names polyacantha and
gymnogastra on the caption to Liitken and Mortensen's plate ix, though the correct
figures 3-5 are quoted both in their text and on the plate itself. These figures show
that A. polyacantha does have the scale reduction characteristic of Amphinephthys
and also that it is very similar to the type specimen of Amphiura arcystata H. L.
Clark, 1911 from California, though in that the disc scaling is even more limited in
extent. As I noted in 1965, A. arcystata was referred not to Amphinephthys but to
Hemilepis by Fell, apparently as a compromise measure between the three specimens
figured by H. L. Clark, only one of which could be described as having the disc scaled
above only, a condition supposed to be diagnostic of Hemilepis. One of the other
specimens figured by H. L. Clark (1911, fig. 58d-f, p. 146) has the scales around the
radial shields more extensive on the dorsal side than those of the type but still not
quite continuous interradially or centrally and not extending on to the ventral side
at all, while the other (fig. 58g-i) has scales all over both sides of the disc, relatively
shorter radial shields and fewer and sharper arm spines (judging from the figure). It
might seem hard to believe that these two specimens are conspecific with the type or
possibly even with each other) since their proximal dorsal arm plates are quite a
different shape, markedly rhombic and only narrowly contiguous as opposed to
rounded with only the slightest lateral angles in the type. However, the Japanese
specimens which I studied in 1965 do have more nearly ovate plates. Matsumoto
(1917) recorded some additional Japanese specimens as arcystata, noting that the
larger ones have discs of the " Ophionephthys-type " (i.e. with scales only around the
radial shields) ; presumably the smaller ones have more extensive scaling. Dj akonov
(1954) also found that some small specimens from the north-west Pacific have fine
granuliform scaling on both sides of the disc in contrast to the usual condition.

Further records of A . arcystata from the eastern Pacific have been made by May
(1924), Nielsen (1932) and Ziesenhenne (1937). May noted that specimens from
Monterey Bay usually have the disc naked at the centre and interradially but some-
times the upper side is completely scaled. Nielsen found that all his nine specimens
from La Jolla had the discs naked except around the radial shields and the same was
probably true of Ziesenhenne's seven lower Californian ones since he noted that on
the dried specimens the discs were brown but the radial shields white. The only
specimens of A . arcystata in the British Museum collections, for which I am indebted
to Captain Ziesenhenne, are three from Santa Cruz Island, southern California, all
of which have the discs mainly covered with dark brown skin in which there is not
a trace of imbedded scales. The shape of the dorsal arm plates in all of them is
rounded, not at all rhombic.

It remains to be seen from direct comparison of abundant material from both
sides of the North Pacific whether Japanese specimens can be distinguished in any
way from Californian ones. The extent of the disc scaling does not appear sufficiently
constant to warrant a specific distinction (let alone a generic one). A comparison
between Amphiura arcystata and the type-species of Amphinephthys, Amphiura
crossota Murakami, from the Caroline Islands is also called for.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) n

Yet another name which should be considered in dealing with Amphiura arcystata,
polyacantha and gastracantha is Amphiura verticillata Ljungman, 1867, known only
from a damaged specimen from the Galapagos Islands. This was described by
Ljungman as having minute scales covering the disc and accordingly Fell retained
it in Amphiura but Koehler's photographs (ig27a, pi. ii) show scales apparently only
remaining in oval areas around the radial shields, as if the interradial and central
scales were much less substantial and so were easily lost or obscured. The distal
oral papilla of A. verticillata arises from the oral plate as in the Californian type of
arcystata, not from the edge of the adoral shield as in H. L. Clark's figures of the other
(Japanese) specimens which he named arcystata. The shapes of the radial shields,
oral and adoral shields and of the ventral arm plates in verticillata are not unlike
those of the type of arcystata and of polyacantha but the almost rectangular dorsal
arm plates, thick distal oral papilla and somewhat square-tipped arm spines of
verticillata may provide significant differences.

Thus we have three closely related species, Amphiura arcystata, polyacantha and
verticillata referred to three different nominal genera, Hemilepis, Amphinephthys and
Amphiura, by Fell's system.

There are several examples of differences in the extent of the disc scaling in
individuals which have been attributed to separate species as well as others such as
those of arcystata already cited which are considered conspecific. For instance,
H. L. Clark (1938) described Amphiura brachyactis, which Fell placed in Pandelia
on account of its single tentacle scale and the occurrence of fine scales (by implication
in the description) on the upper side of the disc only. Clark commented that
brachyactis is very closely related to A . microsoma (see above p. 9) which he had
described in 1915 and which Fell refers to Amphinephthys since the type was
described as having the disc almost naked above as well as below. However, in
1938 H. L. Clark also recorded some specimens which he attributed to microsoma
even though their discs were distinctly covered with scales on both sides. I have
examined the holotypes of both nominal species and find that even in microsoma
there is an almost continuous coat of scales although these are so fine and transparent
as to be almost obscured by the skin, which appears more or less dark in colour except
around the radial shields where the scaling is more nearly opaque. The two species
appear to be distinguished by the relative arm length but they are certainly con-
generic. Among the species with no tentacle scales, Amphiura syntaracha H. L.
Clark, 1915, which Fell refers to Icalia, has the disc scaled except in the proximal
parts of the ventral interradii but is otherwise very similar to four other Japanese
nominal species with much more reduced scaling, namely A. ecnomiotata H. L. Clark,
1911, aestuarii Matsumoto, 1915, vadicola Matsumoto, 1915 and sinicola Matsumoto,
1941, all of which Fell places in Ophiopeltis. Further, Djakonov (1954) has described
two forms, incompleta and profundi of Amphiura psilopora H. L. Clark, 1911
(lacking tentacle scales so that Fell refers it to Nullamphiura), these forms being
distinguished from the fully-scaled form psilopora by having the disc skin-covered
ventrally with either coarse or fine soft skin. Mortensen (1936) has noted specimens
of Amphiura deficiens Koehler, I922b (Icalia according to Fell) in which the disc
is almost completely scaled ventrally, unlike the condition in the type-material.

12 AILSA M. CLARK

Another species of Djakonov's, Amphiura inepta (1954) is referred by Fell to
Monamphiura (with one tentacle scale and the disc fully-scaled), presumably on
the basis of Djakonov's figures. However, the description notes that the disc
appears bare ventrally in specimens preserved in spirit although dried specimens
show very small, thin, rounded scales arranged loosely in the ventral interradii.

Until much more work has been done on individual and seasonal variation in the
disc scaling of these burrowing amphiurids it is difficult to decide to what extent it
can be used in taxonomic distinctions, even at specific level, unless supported by other
characters.

With regard to the use of the number of tentacle scales as a generic character,
there are several species of Amphiura which have been shown to be variable in this
respect. An extreme example is Amphiura belgicae Koehler, 1900, which Mortensen
(1936) notes may have the tentacle scales numbering two, one or none, some speci-
mens having one number throughout while in others it varies even on different pores
of the same arm. Other species may also have mixed scale counts, the proximal or
at least the basal pores having a different number to the distal ones, or odd numbers
may occur at random, one number may predominate in certain individuals or
immature specimens may have fewer scales than adults of the same species. Ex-
amples of all these conditions include Amphiura angularis Lyman, 1879, anomala
Lyman, 1875, aster Farquhar, 1901, concolor Lyman, 1879, dacunhae Mortensen,
1936, fibulata Koehler, 1913, rosea Farquhar, 1893, simonsi A. M. Clark, 1952,
stimpsoni Liitken, 1859, sundevalli (Miiller & Troschel, 1842) and verticillata Ljung-
man, 1867. Accordingly I very much doubt the wisdom of giving generic weight to
the number of tentacle scales in this family.

As a result of these observations, I consider that the following should be referred
to the synonymy of Amphiura Forbes, 1843 :

Ophiopeltis Diiben & Koren, 1845 (type-species 0. securigera Duben & Koren,
1845) [but see below].

Hemilepis Ljungman, 1871 (type-species Amphiura semiermis Lyman, 1869).

Amphinephthys Fell, 1962 (type-species Amphiura crossota Murakami, 1943).

Icalia Fell, 1962 (type-species Amphiura denticulata Koehler, 1896, a synonym
of A, fragilis Verrill, 1885).

Pandelia Fell, 1962 (type-species Amphiura hinemoae Mortensen, 1924).

Nullamphiura Fell, 1962 (type-species Amphiura psilopora H. L. Clark, 1911).

Monamphiura Fell, 1962 (type-species Amphiura alba Mortensen, 1924).
The range of form within the assemblage of species so reunited is extensive and
some natural groups distinguished by combinations of characters may prove to be
recognizable but I very much doubt whether these should be ranked higher than
subgenera.

Of these seven names, there is only one which I think merits recognition as a sub-
genus of Amphiura, this is Ophiopeltis as noted by me in 1966 when describing
a new Australian species.

Among the many amphiurids with no disc armament and reduced tentacle scales
coupled with Amphiura (A) -type mouth parts, the most obviously different from the
type-species of the genus, A. chiajei, are those which have no tentacle scales at all

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 13

combined with extremely reduced disc scaling. Some of these have been referred
in the recent past to the genus Ophionephthys but, as Fell has pointed out, are clearly
not congeneric with the type-species, 0. limicola Lutken, 1869, from the West Indies,
since that species has two (rarely three) small oral papillae distal to the infradental
one, besides an oral tentacle scale, all arising from the oral plate (its formula being
normally m,omm,o +t), one small tentacle scale on most of the arm pores, a con-
spicuous row of scales around the periphery of the disc dorsally and the adoral
shields meeting widely at their proximal (interradial) ends. In Fell's revision a
generic distinction is rightly made between 0. limicola and these other species,
namely 0. ecnomiotata (H. L. Clark, 1911), 0. africana Balinsky, 1957, 0. decacantha
H. L. Clark, 1938, 0. heptacantha Mortensen, 1940 and 0. octacantha H. L. Clark,
1915, all of which he regards as congeneric with Ophiopeltis securigera Diiben and
Koren. Ophiopeltis was a monotypic genus as originally proposed, including only
Ophiopeltis securigera, characterized by its very long arms, 12-15 times the disc
diameter, completely naked disc, absence of tentacle scales and modification of the
middle of the three arm spines with an axe-headed tip (fig. 3b, d).

In 1861 Michael Sars redescribed 0. securigera and noted that drying renders
visible some microscopical imbricating scales in the skin covering the disc. No
mention of this had been made by Diiben and Koren, who described the skin of the
disc as so very thin and soft that it often ruptures in preserved specimens over the
arm bases, as it has done in the single (dried) specimen from the Shetland Islands in
the British Museum collections, which is the only example of securigera available to
me and which shows no trace of the microscopic scales noticed by Sars. 1

It was presumably on the strength of this observation by Sars that Lutken (1869)
rediagnosed Ophiopeltis as having extremely minute scales in the skin of the disc and
accordingly separated off a new genus Ophionema characterized by complete absence
of any such scales. However, as a supporting character he cited the unspecialized
arm spines of the type-species, Ophionema intricata.

Two years later in 1871, G. O. Sars described a new species of Ophiopeltis, 0.
borealis, but Ljungman treated Ophiopeltis (spelled Ophiopelte) as a subgenus of
Amphiura and expanded it to include filiformis (O. F. Miiller, 1776), atlantica Ljung-
man, 1867 and sarsi Ljungman, 1871, as well as securigera and borealis.

However, Lyman in his " Challenger " report (1882) and subsequent authors until
1962 have considered Ophiopeltis as a synonym of Amphiura, not even a subgenus,
because of the existence of intermediate species like Amphiura borealis (G. O.
Sars) with shorter arms and partially-scaled discs but other characters approximating
to those of Ophiopeltis securigera. The wisdom of this is supported by Mortensen's
figure of these two species side by side (1927, fig. 122), showing that they are very
similar, differing only in the presence of disc scales over most of the dorsal side
coupled with inward divergence of the radial shields and by the lesser modification

1 This specimen of securigera also shows two interesting anomalies. Firstly it has a small poorly-
calcified tentacle scale on most of the arm pores, which is quite distinct in the dried condition (though
could well be inconspicuous in spirit ones). Secondly there are two distal oral papillae on each side of
all the jaws, the inner of the two arising from the outer part of the oral plate so that the formula is
m,om,m+t. (Fell (ig62a) has also observed this phenomenon.) On two of the jaws there is a small
median apical papilla between the infradental pair, as may also occur in a number of amphiurids,
including Amphiura borealis.

I 4 AILSA M. CLARK

of the second arm spines in borealis. Other species are known in which the disc
scaling is intermediate in extent between the complete or almost complete absence
of scales in securigera and the partial covering in borealis but the issue has been
clouded by the fact that, with the exception of some Japanese species, these others
have been referred to Ophionephthys, which Matsumoto (1915 and 1917) maintained
in Amphiura, being unfamiliar with the type-species, 0. limicola, the mouth parts of
which had not then been figured. The Japanese species described by Matsumoto
are Amphiura aestuarii, A. vadicola and, in 1941, A. sinicola, which are more or less
intermediate, particularly in the rather wider dorsal and ventral arm plates, between
other Japanese species such as A. syntaracha H. L. Clark, 1915, which has scales all
over the disc except on the proximal parts of the ventral side, and A. ecnomiotata
H. L. Clark, 1911, subsequently referred to Ophionephthys by H. L. Clark in 1915
since it has scaling around the radial shields. Following Matsumoto, Koehler (i922a)
also considered Ophionephthys as a synonym of Amphiura. However, Mortensen in
1924 and 1936 had a different conception of the nature of Ophionephthys (identical
with that adopted by Fell in 1962) ; he referred to the genus two species (stewartensis
and magellanica} with tentacle scales present and Amphioplus (D)-type oral papillae,
of which the discs were assumed to be deficient in scaling just because they were
missing in the specimens collected. The fallacy in this has been revealed by the
recent discovery of Castillo (1968) that intact magellanica has a scale-covered disc
and is clearly referable to Amphioplus ; probably the same will prove to be true of
Ophionephthys stewartensis, which should be referred to Amphioplus meantime.
By 1940 Mortensen had changed his mind and was again describing species of
Ophionephthys (namely heptacantha and iranica} with Amphiura (A) -type oral
papillae, as H. L. Clark had been doing since 1915.

The species that Fell has referred to Ophiopeltis from Ophionephthys are Amphiura
aestuarii Matsumoto, 1915, Ophionephthys africana Balinsky, 1957, 0. decacantha
H. L. Clark, 1938, Amphiura ecnomiotata H. L. Clark, 1911, Ophionephthys hepta-
cantha Mortensen, 1940, 0. iranica Mortensen, 1940, 0. octacantha H. L. Clark, 1915
and Amphiura vadicola Matsumoto, 1915 (A. sinicola Matsumoto, 1941, which
clearly belongs to the same group having been overlooked by Fell). In addition
there are three other species which Fell has included in Ophionema but which, in my
opinion, are close to some of these listed above, namely Ophionephthys phalerata
Lyman, 1874, 0. tenuis H. L. Clark, 1938 and Ophionema hexactis Mortensen, 1940.
Both phalerata and hexactis have the second from lowest of the arm spines strongly
spinose at the tip, while tenuis has the corresponding spine on the more distal arm
segments modified into an axe-headed shape (a feature not mentioned by H. L. Clark
in his description but observable on a paratype in the British Museum collections
and also noted by Mortensen in 1940 on one sent to him for comparison with Ophio-
nephthys iranica, which species he emphasized is closely related to tenuis}. On the
contrary, the two American species of Ophionema, 0. intricata Liitken, 1869 and
0. hexacantha Nielsen, 1932, have the arm spines relatively unspecialized, tapering in
shape and only slightly blunted and rugose at the tips. In addition these two
evidently lack even the narrow area of disc scales round the proximal end of each
radial shield which is visible in hexactis, iranica and tenuis. In securigera also there

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 15

is little or no scaling here. I have seen no examples of Ophionema myself and can
rely only on Nielsen's remarks and figures and Liitken's earlier description (in Danish)
of 0. intricata. Fell did not include the genus in the key given in his main revision,
nor did he diagnose it, though this was done in his Australian amphiurid paper, with
particular reference to Ophionephthys tennis which he was referring to Ophionema
(while simultaneously placing the related iranica in Ophiopeltis}.

The type-species of Ophiopeltis , 0. securigera, has only four arm spines proximally
and three for most of the length of the arm, of which the middle one is strongly
modified into a flattened, axe-headed shape (Fig. 3c, d) ; the dorsal arm plates are fan-
shaped, wider than long and the disc scaling is more or less completely absent so that
the bar-like radial shields are parallel. An additional species from southern Aust-
ralia, Amphiura (Ophiopeltis} parviscutata, was described by me in 1966 and is very
closely related to securigera, which it resembles in these characters.

The twelve other affiliated species mentioned above fall into two fairly natural
groups on the basis of arm structure. In the first come those having not more than
five arm spines and only the second from lowest one more or less strongly modified at
the tip, with at the same time the arm plates not particularly narrowed so that the
dorsal ones are all wider than long. In this group come aestuarii, iranica, tenuis,
phalerata and hexactis. Owing to Lyman's very poor figures of phalerata it is difficult
to make a proper comparison of it and his statement that its adoral shields are
" curved and united at their ends so as to form a continuous ring " a condition
supported by the diagrammatic figure is completely at variance with the condition
found in any of the other species under consideration, all of which have the adorals
spaced interradially with a pit between their ends. Only the holotype of hexactis is
known and this has six arms and three madreporites ; it remains to be seen whether
other specimens from the Persian Gulf share this irregularity. The arm spines
number four and the second from lowest has a very spinose tip. The four other
species all have five arm spines proximally. In this group of five species some have
the disc scaling less reduced than others, so that in aestuarii and phalerata the radial
shields are separated proximally on the surface by a wedge of scales, whereas in
iranica, tenuis and hexactis the shields are parallel with scales only at their proximal
and abradial ends.

The second group of species, comprising africana, decacantha, ecnomiotata, hepta-
cantha, octacantha, sinicola and vadicola, is marked by the possession of seven to ten
arm spines proximally and several of these, not just the second one, are flattened and
markedly square-tipped or spinose at the ends ; also the arm plates are more or less
narrowed so that, at least at the base of the arm, the dorsal ones are longer than wide.
All these species (with the possible exception of africana, the photograph of which
does not show the condition clearly) have the radial shields inwardly divergent.

It follows from this that Ophiopeltis securigera is more closely related to the first
group than to the second and indeed it seems to me that if Ophionema is to be main-
tained as separate from Ophiopeltis, then this second group should also be distin-
guished with the same taxonomic rank.

The question is, what should this rank be? Fell would make it a generic one
on a par with his other taxa but I cannot bring myself to do this because the different

16

AILSA M. CLARK

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 17

groups of species involved are not sharply defined. For instance in European waters
Amphiura borealis (as already quoted) links Amphiura with the type-species of
Ophiopeltis, while in Japan there is a series of species from A. syntaracha H. L. Clark
(Fig. 3m, n), still with no tentacle scales but with the disc completely (though
loosely) scaled except in the proximal parts of the ventral interradii and with arms at
least ten times the disc diameter, through aestuarii with more restricted disc scaling
but still only five arm spines and relatively wide dorsal arm plates, then to sinicola,
vadicola and finally ecnomiotata, with seven to ten spines, progressively narrower arm
plates and longer arms. Accordingly I propose to regard these three taxa as sub-
genera of Amphiura. Unfortunately since one of them is split off from Ophiopeltis
sensu Fell it is necessary to add yet another name to the literature.

AMPHIURA subgenus OPHIOPELTIS Diiben & Keren, 1845, restricted

Ophiopeltis Diiben and Koren, 1845 : 437 ; 1846 : 236-238, pi. vi, figs. 3-6 ; M. Sars, 1861 : 14 ;
Ljungman, i86yb : 321 ; Liitken, 1869 : 94 (75) ; G. O. Sars, 1871 : 15-17 ; Fell, ig62a :
8-9 ; ig62b : 82. Type-species : Ophiopeltis securigera Diiben and Koren, 1845.

Amphiura (Ophiopelte] : Ljungman, 1871 : 643.

Amphiura (part), Lyman, 1882 and following authors until 1962.

DIAGNOSIS. A subgenus of Amphiura in which the disc scaling is absent or
reduced to a small area proximal to each of the bar-like parallel radial shields or to a
slightly larger patch with a wedge between the proximal ends of the radial shields,
which consequently appear divergent ; the arms are extremely long, more than ten
times the disc diameter ; the oral shields have only a slight proximal lobe, if any, and
the adorals are widely separated interradially (? in phalerata) ; the distal oral papilla
is never very large and sometimes (hexactis) rudimentary ; the dorsal arm plates are
wider than long ; the arm spines proximally number four or five, the second from
lowest one becoming more or less modified with an axe-headed or very spinose tip,
especially on the outer part of the arm, tentacle scales are normally lacking though
a poorly-calcified small scale may occur on some pores of certain individuals.

INCLUDED SPECIES. Ophiopeltis securigera Diiben & Koren, 1845, from northern
Europe, Ophionephthys phalerata Lyman, 1874, from the Philippines, 0. tenuis H. L.
Clark, 1938, from north-west Australia, Amphiura (Ophiopeltis} parviscutata A. M.

FIG. 3. Dorsal and ventral views of parts of Amphiura spp. (except for i). a-d. A . (Ophio-
peltis) securigera Diiben and Koren, 1953.4.8.138, Shetland ; a. radial shields and a small
area of surrounding skin (stippled), b. two jaws (the madreporite on the right) showing
the abnormal occurrence of two distal oral papillae each side and incipient single tentacle
scales, c. fiftieth arm segment from above and d. from below ; e-g. A . (Fellaria) octacantha
(H.L.C.), M.C.Z. 3744, holotype, Torres Strait area (in f the fifteenth free arm segment is
shown and the spines are slightly foreshortened ; in g one distal oral papilla is displaced
and another, extreme left, lost) ; h. A. (Ophionema) intricata (Liitken) (after Nielsen,
1932) ; i. Ophionephthys limicola Liitken (modified from Nielsen and from Thomas,
1962) ; j and k. A. fragilis Verrill, U.S.N.M. 32601, holotype, "Albatross" st. 2025,
E coast of U.S.A. ; 1 and m. A, sarsi Ljungman, M.C.Z. 1342, syntype, Azores ; nand o.
A. syntaracha H.L.C., M.C.Z. 1361, holotype, Bay of Yeddo, Japan.

i8 AILSA M. CLARK

Clark, 1966, from Victoria, Australia, Ophionephthys iranica and Ophionema hexactis
Mortensen, 1940 from the Persian Gulf (the former also taken recently in the vicinity
of Karachi, West Pakistan) and Amphiura aestuarii Matsumoto, 1915, from Japan.
Geographically therefore these are a mixed lot ; A. (Ophiopeltis) parviscutata is the
one most closely related to the similarly temperate A . (Ophiopeltis) securigera, while
the rest are from the tropical Indo-West Pacific and are morphologically intermediate
to some extent with the species of the following subgenus from the same region.

AMPHIURA subgenus FELLARIA nov.

Amphiura (part) H. L. Clark, 1911 : 148-149 ; Matsumoto, 1915 : 71 ; 1917 : 211-213 > I 94 I :

339-342.
Ophionephthys (part) H. L. Clark, 1915 : 239-240 ; 1938 : 240-241 ; Mortensen, 1940 : 79-81 ;

Balinsky, 1957 : 7-9. [Non Ophionephthys Liitken, 1869].
Ophiopeltis (part) Fell, ig62a : 8.

DIAGNOSIS. A subgenus of Amphiura in which the disc scaling is reduced to a
small area lateral and proximal to each pair of radial shields ; the arms are extremely
long, usually fifteen to twenty times the disc diameter but sometimes even longer ;
the oral shields have only a slight proximal lobe, if any, and the adorals are widely
separated with a pit between their interradial ends ; the distal oral papilla is not
rudimentary but may be large and clavate ; the dorsal arm plates are narrow, the
basal ones at least, if not all of them, longer than wide ; the arm spines number seven
to ten proximally and all or most of them are flattened with their tips rugose and
squared-off or slightly flared ; tentacle scales are lacking.

TYPE-SPECIES. Ophionephthys octacantha H. L. Clark, 1915, from northern
Australia.

INCLUDED SPECIES. Ophionephthys africana Balinsky, 1957, from Mozambique,
Amphiura ecnomiotata H. L. Clark, 1911, from Japan and Ophionephthys heptacantha
Mortensen, 1940, from the Persian Gulf ; also Amphiura vadicola Matsumoto, 1915
and A. sinicola Matsumoto, 1941, from Japan, which are to some extent intermediate
with Amphiura sensu stricto. Ophionephthys decacantha H. L. Clark, 1938, I
consider is a synonym of A. (Fellaria) octacantha, having studied the holotypes of
both. Despite the name decacantha, I could not find more than eight arm spines on
any part of the arms of the former and the only apparent morphological difference
consists of the barely contiguous proximal dorsal arm plates in the holotype of
decacantha, those of octacantha tending to overlap slightly for most of their width, at
least on the proximal segments. D.d. in both is about 5 mm. but the disc of the
holotype of octacantha is more drastically shrunken and I think that the original size
was greater than in the other specimen, which might account for the small difference
in arm structure. The type-localities are Friday Island, Torres Strait (octacantha)
and Broome, north-west Australia (decacantha) .

It may be noted here that H. L. Clark's figure of the holotype of Amphiura
ecnomiotata is incorrect in showing the distal oral papillae as long and narrow ; in
fact they are very thick and not particularly elongated ; also the oral shields are

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDE A) 19

in reality as wide as long or a little wider. In addition, in the skin of the disc between
the radial shields interradially are scattered some very fine scales, not mentioned by
Dr. Clark.

AMPHIURA subgenus OPHIONEMA Liitken, 1869, restricted

Ophionema Liitken, 1869 : 27-28 (9-10), 94 (76), 98 (80) ; Lyman, 1882 : 151-152 ; 1883 : 253 ;
H. L. Clark, 1918 : 282-284, pi. 2, figs. 1-3 ; Nielsen, 1932 : 265-267 ; Thomas, 1964 :
159-160. [Non Ophionema : Mortensen, 1940 : 79-81 (O. hexactis, now referred to Ophio-
peltis)]. Type-species : Ophionema intricata Liitken, 1869.

DIAGNOSIS. A subgenus of Amphiura in which the disc scaling is absent and the
radial shields are bar-like and parallel ; the arms are extremely long, fifteen to twenty
times the disc diameter ; the oral shields have no proximal lobe and the adorals are
widely separated interradially ; the distal oral papilla is small and often more or less
rudimentary ; the dorsal arm plates are fairly narrow, about as wide as long or a
little longer ; the arm spines proximally number five or six and all taper to a rounded
and only slightly rugose tip ; tentacle scales are lacking.

INCLUDED SPECIES. Ophionema intricata Liitken, 1869, from the West Indies and
0. hexacantha Nielsen, 1932, from Panama. 1

It seems advisable at this point to add references and a diagnosis to Ophionephthys
for comparison with the superficially rather similar subgenera of Amphiura just
dealt with.

OPHIONEPHTHYS Liitken, 1869

Ophionephthys Liitken, 1869 : 25-27 (7-9), 93 (75), 98 (80) ; H. L. Clark, 1918 (part) : 278-282 ;
Nielsen, 1932 : 266, fig. 8 ; Fell, ig62a (part) : 3-4, 15 ; Thomas, 1962 : 680-681. Type-
species : Ophionephthys limicola Liitken, 1869. [The many references to species other than
the type-species apply to Amphiura (Ophiopeltis) , Amphiura (Fellaria), Amphioplus and (in
the case of OP sesquipedalis Bell) to Amphiodia or Ophiophragmus.]

DIAGNOSIS. A genus of the family Amphiuridae in which the disc scaling is
reduced to a small patch of scales at the proximal end of each pair of radial shields
and a line of scales near the edge of the disc extending across the interradii from the
distal end of each radial shield ; the arms are extremely long, up to twenty times the
disc diameter ; the oral papillae are normally limited to three in each series, the
two distal ones on the side of the oral plate as well as the oral tentacle scale (occasion-
ally odd series are found with a fourth papilla but this is nearly always situated also
on the oral plate and rarely on the adoral shield) so that the oral formula is normally

1 Since completion of this MS, a paper by Domantay and Domantay (Studies on the classification
and distribution of Philippine littoral Ophiuroidea (brittle stars). Philipp. J. Sci. 95: 1-77; 1967) has
come to my notice. This includes a description of a new species referred to Ophionema, O. philippinensis
(p. 25), the affinities of which are difficult to assess from the information given. It is clear from the
drawing of the upper side (fig. 4) that the disc has been lost, the central structure being the oral frame.
The excessively high ratio of arm length to d.d. of 25-30:1 (d.d. in fact the diameter of the oral
frame being given as only 2 mm.) is almost certainly incorrect, the true ratio being probably much
lower, rendering unlikely a close relationship with Ophionema intricata.

20 AILSA M. CLARK

m,mm,o +t (occasionally m,mmm,o +t and only rarely m,mm,om +t). In the type
species the adoral shields meet widely interradially, the oral shields have a large
rounded proximal lobe, the dorsal arm plates are as wide as or wider than long and
there is a single tentacle scale.

REMARKS. Having started as a monotypic genus, I consider that Ophionephthys
should revert to that condition. The treatment of some of the nominal species
which have been referred to it since 1869 is dealt with in the discussion leading up to
the subgenera of Amphiura above but an extended account of the history is needed.

The species which have been referred to Ophionephthys since 1869 fall into two
categories. The first includes those with Amphiura (A) -type oral papillae, namely
phalerata Lyman, ecnomiotata (H. L. Clark), octacantha H. L. Clark, aestuarii (Mat-
sumoto), vadicola (Matsumoto), decacantha H. L. Clark, tennis H. L. Clark, heptacantha
Mortensen, iranica Mortensen and africana Balinsky. The second category consists
of species with Amphioplus (D)-type oral papillae, namely stewartensis Mortensen,
magellanica Mortensen, Amphioplus cyrtacanthus H. L. Clark, 1915, A. lucidus
Koehler, I922a, A. seminudus Mortensen, 1940 and doubtfully A. luctator Koehler,
I922a, the last four having been referred to it by Fell (i962a) (and not having been
previously mentioned in this paper merit the fuller details). In 1955 I myself named
some discless West African specimens " POphionephthys sp.", noting, however, that
although they are probably congeneric with 0. magellanica and stewartensis, the
same could not be said of their relationship with 0. limicola.

Of the species in the first category, Fell removed all except phalerata and tenuis
(which he included in Ophionema) to Ophiopeltis ; I am now transferring those two
also to Ophiopeltis but simultaneously abstracting to the new subgenus Fellaria six
of the others ecnomiotata, octacantha (with synonym decacantha}, vadicola, hepta-
cantha and africana.

In his revised diagnosis of Ophionephthys (ig62a : 15), Fell gives the number of
oral papillae as four or five (i.e. of the Amphioplus (D)-type) and used the genus as a
repository for several species of Amphioplus the type-specimens of which were
lacking the disc, arousing the supposition that their disc scaling was reduced.
However, a number of amphiurids with fully-scaled discs also frequently shed or
otherwise lose them. Anyway the figures of Nielsen (1932) and Thomas (1962) as
well as some specimens kindly sent to me by Dr. Thomas show that Fell's description
of the mouth parts is incorrect and that Ophionephthys limicola has oral papillae
arranged unlike those of any of the major genera of Amphiurids. However, it would
be remiss not to point out that occasional specimens, particularly some immature
ones, of a few species of Amphioplus like A. strongyloplax (fig. gb) and macraspis
(both originally described as species of Amphiodia by H. L. Clark, 1911) may lack
the fourth oral papilla from the adoral shield ; but then the third papilla, as often in
Amphioplus, arises from the point of junction of the oral plate and the adoral shield,
not from the oral plate alone as in 0. limicola, the oral formula in these two species
of Amphioplus being m,mn^nm +t or m.mrfjio +t. The presence of the oral tentacle
scale of course distinguishes both Ophionephthys and Amphioplus from Amphiodia.

Of the second category of species, i.e. those referred to Ophionephthys by Fell, only

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 21

the types of Amphioplus seminudus were in possession of their discs ; these have
scales present only around the proximal end of each pair of radial shields, the rest
of the disc being naked. However, Castillo (1968) has recently discovered additional
material of Ophionephthys magellanica from southern Chile, which he says show
fully-scaled discs so that the species is undoubtedly referable to Amphioplus. It
may be noted that Chang (1948) has figured a specimen with the disc fully-scaled
above but naked below under the name of Amphioplus cyrtacanthus but this does not
appear to have the arm spines hooked, unlike the holotype, and I think the identifica-
tion needs confirmation. The three other species, lucidus, luctator and stewartensis
also await the discovery of intact specimens ; meanwhile I am sure their affinities lie
with Amphioplus.

There is, however, one species normally included in Amphioplus which appears to
have some affinity with Ophionephthys limicola ; this is Amphioplus coniortodes
H. L. Clark, 1918, also from the West Indian area. The holotype (fig gd, e) has the
oral formula m,(m = t)mm,o in most series, with three papillae on the side of the
oral plate, the proximal-most of these being very slightly higher and clearly homo-
logous with the oral tentacle scale of other species. The outermost papilla does not
seem to me to make contact with the adoral shield although H. L. Clark describes it
otherwise. In some series there is a very small papilla on the side of the first ventral
arm plate in series with the rest giving a formula of m,(m = t)mm,,m. This
arrangement shows some difference from the specimen figured by Thomas (1962,
fig. IOB) which has the more usual Amphioplus (D)-type formula m,(m = t)mm,m,
while one of the four series drawn even has a second papilla on the edge of the adoral
shield. The disc of A. coniortodes also tends to approximate to that of 0. limicola,
being covered with naked skin below while the dorsal scaling is extremely fine (so
much so in the holotype as to be barely distinguishable). In addition there is an
enlarged series of scales extending interradially from each radial shield, something
like the line of scales in 0. limicola. Finally, A. coniortodes also has extremely long
arms, equal in length to about twenty times the d.d. Thomas has pointed out
that the two species are commonly associated in the same soft mud in Florida, so
some of the resemblances between them may be convergent. Fell (i_962a) has
referred A. coniortodes to his new nominal genus Ailsaria, together with several other
species of Amphioplus with more or less reduced disc scaling. This is dealt with
below under the heading of Amphioplus.

AMPHIURA (cont.)

Amphiura correct a Koehler

fig. 4a, b

Amphiura correcta Koehler, 1907 : 300-301, pi. 12, figs. 30, 31.
Amphiura novae Benham, 1909 : 104, figs. 1-3 ; Fell, 1958 : 26-27.
Amphiura abernethyi Fell, 1951 : 1-3, figs. 1-4.

22 AILSA M. CLARK

Among the specimens studied by me at the Museum of Comparative Zoology,
Harvard in 1954 was a syntype of Amphiitra correcta, the locality of which was
unknown although the types were collected by Quoy and Gaimard on the voyage
of Dumont d'Urville (on the " Astrolabe " primarily to Australia, New Zealand and

FIG. 4. a and b. Amphiura correcta Koehler, M.C.Z. 1303, syntype, locality ? (presumably
New Zealand) ; c. A. lorioli (Koehler), 1948.5.26.69 (pt.), paratype of A. ceramis
H.L.C., Zanzibar area, one jaw.

the South Seas). This specimen (fig. 4a, b) proved to be conspecific with Amphiura
norae Benham from New Zealand. The same is true of a syntype of A. abernethyi
sent to the British Museum by Dr. Fell in 1952, which indicates that Fell's drawings
over-exaggerated the peculiarities such as the shape of the radial shields and the
scales between them ; the latter are simply rounded in the syntype sent to me ; also
there is a suture separating oral plates and adoral shields each side of the apex of the
oral shield so that the two adorals are not contiguous, the oral shields themselves are
more nearly hexagonal than rhombic, the distal oral papilla is broad rectangular or
fan-shaped and the dorsal arm plates have the distal side with a median convex angle
and are not simply curved. The fifth arm spine found in abernethyi compared with
the four counted by Benham can be attributed to the much larger size of the types
of the former. In a letter of November, 1952, Dr. Fell agreed with me that norae
and abernethyi could be " lumped together ", having studied a further range of
specimens which indicated to him that a single variable species is present. However,
in his paper of 1958 he mentioned this synonymy as only a possibility.

Amphiura lorioli (Koehler)
fig. 4c

Ophiactis lorioli Koehler, 1897 : 328 ; 1899 : 41-42, pi. 6, figs. 46, 47.
Amphioplus lorioli H. L. Clark, 1915 : 258 ; 1918 : 295 ; Fell, ig62a : 17.
Amphiura ceramis H. L. Clark, 1939 : 58-60, figs. 14, 15.

The holotypes of both Ophiactis lorioli and A . ceramis are in the British Museum
collections. Comparison reveals no significant difference. Both have extremely

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 23

large and few disc scales for an amphiurid and unusually small oral shields with the
constricted though elongated distal lobe markedly depressed and a close pair of
distal oral papillae each side. The oral formula is m,om,,m -ft (or m,onjinji +t),
the distal papillae, though close together, hardly abutting at all on the edge of the
adoral shield (as H. L. Clark's figure incorrectly shows them, though his description
is better ; the teeth are also broader than his figure shows). There is some resem-
blance in the oral structure to Ophiomonas though the oral tentacle scale is rather
deeper in the slit basally in A . lorioli. The paratypes of A . ceramis are remarkably
variable in the number and arrangement of their oral papillae ; one even has a
formula of m,omm,mn^n +t in five series but m,omn^nfT/h +t in the other five, thus
illustrating the close relationship of Amphiura with Amphioplus.

AMPHIODIA Verrill

Amphiodia Verrill, iSgga, : 306, 313 ; iSggb : 25 ; H. L. Clark, 1915 : 245 ; Nielsen, 1932 : 267 ;
Djakonov, 1954 : 61 ; Fell, ig62a : 5, 15 ; Thomas, 1962 : 640 ; 1964 : 164-165. Type-
species : Amphiura pulchella (Lyman, 1869).

Amphiodia (Amphispina) Nielsen, 1932 : 276.

[See also p. 25 for synonyms.]

The Amphiurid species with Amphiodia (B)-type jaw armament and unarmed discs
exhibit a much narrower range of form with regard to other characters than those of
the (A)-type. In none of the species so far described does the disc lack scales on the
upper side, though they may be naked below, also tentacle scales are never lacking
altogether and the number of arm spines rarely exceeds five. Accordingly Fell
(ig62a) could only split the genus into three using his criteria of disc scaling and
tentacle scale number, dividing off Gymnodia with the disc skin-covered below
from Amphiodia and Diamphiodia with fully-scaled discs and one or two tentacle
scales respectively. (Evidently he suffered a change of mind about the new nomen-
clature involved, since the name Ailsaria appears in his key in the proper place for
Gymnodia, judging from the diagnoses, while the nomen nudum Gymnoplus takes the
place of Ailsaria.}

The type-species of Amphiodia, Amphiodia pulchella (Lyman, 1869) has been
fully described and figured by Thomas (1962) who points out that the ventral scales
" may be difficult to see unless the specimen is dried ". Accordingly Thomas refers
Amphiura repens Lyman, 1875 (included by Fell in Gymnodia}, to the synonymy of
A. pulchella. The type-species of Gymnodia, Amphiodia tabogae Nielsen, 1932,
is evidently very variable. The second specimen figured by Nielsen (1932, fig. roa
and c, p. 273) appears very similar to Amphiodia pulchella and I fully agree with
Thomas (1964) that there cannot be a generic distinction between them. Of the
other two nominal species referred to Gymnodia, Amphiodia platyspina Nielsen,
1932 is very similar to A. tabogae, differing mainly in the modification of some of the
middle arm spines, while Amphiodia psilochora H. L. Clark, 1911 (fig. 5g) has the
oral tentacle scale present and so does not belong to the Amphiodia (B) -group at all ;
it resembles an Amphioplus with the distalmost oral papilla suppressed as may occur
in A. coniortodes already mentioned (p. 21) ; I have also observed some scattered
scales in the ventral interradii of the holotype of A. psilochora. Nielsen has also

AILSA M. CLARK

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 25

noted that in larger specimens (d.d. up to n mm. in his material) of Amphiodia
occidentalis (Diamphiodia according to Fell) " the scales in the proximal part of
the ventral interradii are resorbed, so that the skin of the disc is transparent on
specimens in alcohol ". It should also be mentioned that Amphiodia obtecta
Mortensen, 1940, which Fell left in Amphiodia, has the ventral side of the disc naked ;
Mortensen also commented that specimens in alcohol appear to be naked above as
well but drying reveals the scales on the dorsal side. In face of these observations
it seems impossible to accept variations in disc scaling as providing characters of
generic or even subgeneric weight in this context.

As for tentacle scale number, supposed to characterize Diamphiodia, at least one
species of Amphiodia, Amphiodia euryaspis H. L. Clark, 1911, has either one or
two scales, as I found on studying the type-material in the Smithsonian Institution.
Here again I cannot accept this as a generic character. Accordingly the following
are here referred to the synonymy of Amphiodia Verrill, 1899:

Gymnodia Fell, I962a (type-species Amphiodia tabogae Nielsen, 1932).

Diamphiodia Fell, I962a (type-species Amphiura violacea Liitken, 1856).

Nielsen's subgenus Amphispina (1932), for which no type-species was designated
by Nielsen, is strictly speaking a nomen nudum on that account under Article I3b
of the Code of Nomenclature. I cannot trace a subsequent type-designation and
since I consider that Amphiodia (Amphispina} digitata Nielsen, 1932, together with
Ophiophragmus duplicatus Koehler, 1930 and Amphiodia microplax Burfield, 1924,
form a natural group distinguished by the spiniform prolongations of the uppermost
row of ventral disc scales, it is necessary to re-establish the name as follows :

AMPHIODIA subgenus AMPHISPINA nov.
Amphiodia (Amphispina) Nielsen 1932: 276-277.

TYPE-SPECIES. Amphiodia (Amphispina] digitata Nielsen, 1932.

DIAGNOSIS. A subgenus of Amphiodia distinguished by the modification of the
uppermost row of ventral disc scales with the free edges prolonged into single or
forked spines.

FIG. 5. a and b. Amphiodia psara H.L.C., M.C.Z. 965, holotype, Corona del Mar, California;
c and d. Amphiodia crassa Koehler, U.S.N.M. 40953, " Albatross " st. 5282, Philippine
Is. (d.d. only 2-5 mm ; in d the two giant arm spines of the first segment of the arm drawn
are both lost but those of the adjacent arms are shown) ; e and f. Amphiodia (Amphispina)
duplicata Koehler, 1949.8.15.24, paratype, Amboina, e showing spinose projections of
marginal scales, f relatively more enlarged to show details of oral structure ; g. " Amphi-
odia " psilochora H.L.C., U.S.N.M. 25544, holotype, Sagami Bay, Japan, showing the
oral tentacle scale in contrast with f ; h. Ophiophragmus wurdemanni Lyman, M.C.Z.
1119, holotype, Charlotte Harbour, Florida ; i. O. brachyactis H.L.C., M.C.Z. 1120,
holotype, off Sombrero Key, Florida, side view of base of arm showing marginal papillae
fringing the distal edge of the large radial shields with a few additional papillae just above
the arm base each side ; j. " Amphiodia " affinis (Studer), 1936.12.30.769, South Georgia.

26 AILSA M. CLARK

I do not agree with Nielsen that Amphiodia urtica (Lyman, 1860) should be
referred to the subgenus Amphispina ; it has many or all of the ventral disc scales
angular or with spinose more or less projecting points on the free edges and possibly
merits a separate supraspecific distinction comparable with Acrocnida of the Amphi-
ra-group. Fell (i962a) mistakenly refers it to Ophiophragmus although it lacks
articulated marginal papillae.

AMPHIODIA sensu extenso (cont.)

Of the forty-seven nominal species included in Amphiodia, Diamphiodia and
Gymnodia by Fell (19623.), a large number are from east or west America (excluding
Alaska and Chile), namely Amphiodia air a (Stimpson, 1852) (referred to Micropholis
nov. by Thomas, 1966 and with synonyms Ophiolepis limbata Grube, 1857 and
Amphiodia gyraspis H. L. Clark, 1915, according to Thomas, 1964), A. grisea
(Ljungman, i867b), A. occidentalis (Lyman, 1860), A. oerstedi (Liitken, 1856), A.
peloria Bush, 1921, A. periercta H. L. Clark, 1911, A. planispina von Martens, 1867,
A. platyspina Nielsen, 1932, A. psara H. L. Clark, 1935, A. pulchdla (Lyman, 1869)
(with synonym Amphiura repens Lyman, 1875, according to Thomas, 1962), A.
riisei (Liitken in Lyman, 1860), A. sculptilis Ziesenhenne, 1940, A. tabogae Nielsen,
1932, A. trychna H. L. Clark, 1918 (with synonym A. tymbara H. L. Clark, 1918,
according to Thomas, 1962), A. vicina H. L. Clark, 1940 and A. violacea (Liitken,
1856). To these can be added A. urtica (Lyman, 1860) with synonym Amphiura
barbarae Lyman, 1875, according to Nielsen (1932).

More than half of these species, namely A . grisea, oerstedi, planispina, platyspina,
psara (fig. 5a, b), riisei, sculptilis, tabogae, trychna, vicina and violacea, show a more
or less marked tendency for broadening of the third oral papilla, which may approxi-
mate in proportions to the operculiform third papilla of Amphipholis sensu extenso,
witness the uncertain position of Ophiolepis air a Stimpson, transferred to Amphipholis
by Ljungman, 1867, to Amphiodia by Verrill, 1899 and back to a newly-formed
division of Amphipholis by Thomas, 1966. In addition, a number of the American
species of Ophiophragmus, such as 0. brachyactis H. L. Clark, 1915, 0. luetkeni
(Ljungman, 1871), 0. marginatus (Liitken, 1859), 0- Paucispinus Nielsen, 1932, and
0. pulcher H. L. Clark, 1918 (but not 0. wurdemanni (Lyman, 1860), the type-species),
also share this tendency. Since Thomas (1962) has found the holotype of Amphiodia
rhabdota H. L. Clark, 1918 to be a specimen of Ophiophragmus pulcher deficient in
the marginal papillae supposedly characteristic of the genus Ophiophragmus, I think
that similar synonymies will also be discovered among some of the other nominal
species, for instance Amphiodia violacea, A. psara (fig. 5a, b) and 0. marginatus
would bear comparison i,nd indeed the supra-specific limits between the species of
Amphiodia and Ophiophragmus named above, together with most of the American
species of Amphipholis sensu extenso, especially those which Thomas (1962) has
referred to his new nominal genus Micropholis (namely Amphipholis atra (Stimpson,
1852), gracillima (Stimpson, 1852), subtilis Ljungman, i867b, platydisca Nielsen,
1932, geminata (Le Conte, 1851) and puntarenae (Liitken, 1856), are badly in need
of revision. It is to be hoped that Thomas will go fully into this matter in his

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 27

monograph of west Atlantic amphiurids (now in course of preparation), though it is
unfortunate that the compass of this work apparently excludes the many closely-
related species described from the western coasts of America.

The abstraction of the species of Amphiodia with the third papilla enlarged leaves
only four American species with three approximately equal oral papillae on each half
of the jaw, namely A. pulchella, A. occidentalis, A. periercta and A. peloria. Out of
the 25 nominal species remaining which Fell referred to Gymnodia, Diamphiodia and
Amphiodia, only eight more appear to me to be valid species of Amphiodia. These
are A. craterodmeta H. L. Clark, 1911, A. crassa (Koehler, 1904) (fig. 5c, d), A. dividua
Mortensen, I933C, A. euryaspis H. L. Clark, 1911, A.fissa (Liitken, 1869), A.frigida
(Koehler, 1897), A. minuta H. L. Clark, 1939 and A. obtecta Mortensen, 1940, to which
may be added A. debita Koehler, i922a and acutispina Koehler, 1914, of the three
which Fell mentioned as incertae sedis. The remainder are disposed of as follows:
Amphiodia psilochora H. L. Clark, 1911 (fig. 5g) has an oral tentacle scale (as
mentioned above, p. 23) and resembles an Amphioplus with the fourth (distalmost)
oral papilla suppressed ; its oral formula is m,mm,o +t. It is here provisionally
referred to Amphioplus and discussed further on p. 57. Amphiodia affinis
(Studer, 1885), ochroleuca (Brock, 1888), grata (Koehler, 1904), reposita (Koehler,
1905) and servata (Koehler, 1904) (see figs 5J and 8 (part)) also have the first oral
tentacle scale present but usually have a slight diastema between the infradental and
middle papillae (very small in the case of some specimens of ochroleuca see fig. 8t)
and the distalmost papilla is based at least partly on the edge of the adoral shield
(except in affinis where both the distal papillae are usually inset somewhat on to the
oral plates though there is a great deal of variation in their number and arrangement) ;
the formula for most of these species can be expressed as m,om,m -ft and they are
referred to subdivisions of Amphioplus as stated on pp. 48 and 49. Amphiodia
assimilis (Liitken & Mortensen, 1899) and caulleryi (Koehler, 1897), judging from
the respective figures, resemble species of Amphiura with two distal oral papillae,
having a large diastema at the side of the oral plate and the two distal papillae
forming a symmetrical pair each side based on the adoral shield ; their formula may
be given as m,o,mm -ft, possibly m,on3m -ft, much as in Amphiura filiformis, with
which I believe they are congeneric. (See also p. 43). Amphiodia digital a
H. L. Clark, 1911 is another such species and I agree with Djakonov (1954) in referring
it to Amphiura. Amphiodia olivacea (Brock, 1888) has short blunt marginal
spines and should be referred to Ophiophragmus, while Amphiodia fuscoalba
(Brock, 1888) (the third of Fell's species incertae sedis) appears from the description
to have Amphiura (A) -type oral papillae and granuliform scales on the underside of
the disc ; it may therefore be referred to Acrocnida. Ctenamphiura sinensis
A. H. Clark, 1917, which Fell rather incomprehensibly placed in Diamphiodia
despite its four oral papillae more like Amphioplus, I have recently (1967) referred
to Paracrocnida. Similarly Amphioplus brachiostictus Tortonese, 1949 has four
oral papillae and cannot therefore be included in Amphiodia. However, Mortensen
(1952) mentions it as being similar to Amphiodia chilensis (Miiller and Troschel),
which he maintains is better-placed in Amphiodia than in Ophiophragmus although
some specimens of it have marginal papillae. Three of the four remaining names are

28 AILSA M. CLARK

synonyms, Amphiodia antarctica (Ljungman, 1867!}) of A. chilensis (Miiller &
Troschel, 1842), according to Mortensen (1936), A. rhabdota H. L. Clark, 1918
of Ophiophragmus pulcher H. L. Clark, 1918, according to Thomas (1962) and A.
rossica Djakonov, 1935 of A.fissa (Liitken, 1869), according to me (1965). Finally,
Amphiodia cyclaspis (originally cyclospis), also of Djakonov, 1935, was clearly
based on a specimen with incompletely-regenerated disc, which accounts for the
supposedly diagnostic characters of the diminutive separated radial shields and the
mosaic of small plates on the upper side of the basal arm segments ; Djakonov com-
pared it otherwise with the sympatric A. crater odmeta, with which I suggest it is
synonymous. Djakonov (1954) gives a figure of the dorsal side but the ventral has
never been illustrated.

OPHIOCNIDA Lyman

Ophiocnida Lyman, 1865 : 133 ; 1882 : 152-153 ; Verrill, iSggb : 3i5~3 l6 (sensu Lyman),
p. 317 (restricted) ; H. L. Clark, 1915 : 250 ; Thomas, 1962 : 683-684. Type-species :
Ophiolepis hispida Le Conte, 1851 (designated by Verrill, iSggb : 317).

Ophiocnidella Ljungman, 1871 : 649. Type-species : Amphiura scabriuscula Liitken, 1859.

Ophiocnida hispida, the type-species, has oral papillae of the Amphiodia (B)-type
(fig. yc). The same is not strictly true of one of the other species which is currently
still included in Ophiocnida, having been referred to it by Lyman (1874), namely
Ophiophragmus echinatus Ljungman, i867b. This species has Amphioplus (D)-type
oral papillae (fig. 7k), as least when mature, and accordingly with its spinose disc is
now referred to the genus hitherto known as Amphiacantha Matsumoto, 1917 for
which a new name is proposed on p. 33.

Another species is a doubtful member of the genus Ophiocnida, namely 0. loveni
(Ljungman, i867a), for further mention of which see p. 31,

AMPHIPHOLIS Ljungman

Amphipholis Ljungman, i867a : 165 ; 1867!) : 311-312 ; Verrill, iSgga : 24 ; iSggb : 306, 311-
312 ; H. L. Clark, 1915 : 240 ; Djakonov, 1954 : 64 ; Fell, ig62a : 15 ; Thomas, 1962 : 657 ;
1966 : 827-833. Type-species : Amphipholis januarii Ljungman, i867a.

[See below, p. 31 for synonyms.]

As with Amphiodia, the Amphiurid species with Amphipholis (C)-type jaw
armament and unarmed discs exhibit a fairly narrow range of form in comparison
with Amphiura with regard to disc scaling, though their tentacle scales may be
reduced to one or (in one case) totally absent. Fell (i962a) was thus able to divide
off only Monopholis with one tentacle scale and Nullopholis with none from Amphi-
pholis restricted. As Thomas (1966) has pointed out, the second species which Fell
refers to Nullopholis, Amphipholis pentacantha H. L. Clark, 1915, has tentacle
scales present on the basal segments (in fact there are twc on several pores of the
very first segments though only one on the next two segments, the remaining pores
having none, see fig. 6e, f). (Similarly A . murmanica Djakonov, 1929 loses its single
scale distally.) Thomas also notes that A. pentacantha is not at all closely related

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA)

29

either to Amphipholis nudipora Koehler, 1914 the type-species of Nullopholis, or to
A. vitax (Koehler, 1904), the type-species of Monopholis, being isolated from them
both by its very coarse disc scaling and from nudipora also by its relatively numerous
and elongated arm spines. Djakonov (1954) notes that A. murmanica is closely
related to A . squamata except for having only a single tentacle scale, while the one
remaining species referred to Monopholis by Fell, A. loripes Koehler, I922a (fig. 6a),
is known only from a single discless specimen and appears to have some affinities
with Amphiodia. Certainly the three species do not form a natural group and I can
see no justification for separating them off from Amphipholis. Nor can I find other
characters to support the absence of tentacle scales in A. nudipora to warrant its
separation and I fully agree with the rejection of these two nominal genera. However,
Thomas himself proceeds to divide up the species of Amphipholis on different criteria,
most notably the shape of the oral plate in side view and the occurrence of perfora-
tions through the arm ossicles. The observation of both these characters requires
dissection and special preparation ; also their use as taxonomic characters is of recent
introduction, the study of oral plates being initiated by Murakami (1963) (discounting
his preliminary paper of 1947 in Japanese) and of arm ossicles by Buchanan &
Woodley (1963). Although Murakami covered a fair number of Japanese species,
these included only a few from each of the larger genera, four of Amphipholis for
instance, while only two species of Amphiura, filiformis and chiajei have had their
arm ossicles described with regard to this character, apart from Thomas's summarized
remarks for his subdivisions of Amphipholis. The perforation of the arm ossicles
appears to be correlated with the degree of extension of which the tube feet are cap-
able and this appears to depend on the feeding habits of the animals. Evidently in
Amphipholis januarii the distal (more immature) arm ossicles may be perforated but
not the proximal ones ; so there is also a chance that this is an ontogenetic rather
than a functional modification. Much more work is clearly needed on the functional
morphology of a wide range of species of Amphiurids before it can be determined to
what extent the sharing of perforated ossicles is attributable to convergence of habit.
Since Thomas is now engaged on such a study, it is obviously premature to take
decisive action here with regard to his new taxa. More details are needed to render
the preliminary distinction of these convincing.

The diagnoses Thomas gives for these three nominal genera can be summed up
in the form of a table :

Oral papillae able to close slit

Tentacle scales

Arm spines

2nd spine modified

Successive dorsal arm plates

contiguous

Arm ossicles perforated
Oral plates with strong ab-

radial wings
Disc scales
Radial shields

A mphipholis
almost

just
only distally

" prominent "
not stated but

Axiognathus

+ or almost

i or 2

3-5

Micropholis
almost

2
3-5

" large "
" joined their

" small "
not stated but

joined for c. 2/3 whole length " +/ contiguous

AILSA M. CLARK

In the case of Amphipholis restricted, the diagnosis is derived from only a single
species and in Axiognathus from four. Certainly most of the characters involved
are only of specific weight, apart from the two already mentioned, the importance
of which remains to be seen, while it is just possible that modification of the second
arm spine will prove to allow a supra-specific grouping, though below the generic

a

FIG 6. a. Amphipholis loripes Koehler, U.S.N.M. 41190, holotype, " Albatross" st. 5375,
Philippine Is. (only a few disc scales remain attached to the oral shields but enough to
show that the disc was not naked below) ; b-d. A. januarii Ljungman, M.C.Z. 4140,
holotype of A. pachybactra H.L.C., Buccoo Bay, Tobago, c shows the thirtieth arm seg-
ment from above (all the ventral interradii are ruptured, hence the gaps in d) ; e and f.
A. pentacantha H.L.C., M.C.Z. 1437, holotype, off St. Lucia, West Indies.

level unless supported by other characters. Superficially the differences between
Amphipholis januarii (fig. 6b-d) and A. squamata, the type-species of Amphipholis
and Axiognathus respectively, are very slight and considerable evidence will be
needed to convince me that they are generically distinct.

Inevitably my reluctance to accept this schism is increased by desire to retain the
familiar combination Amphipholis squamata (Delle Chiaje, 1829) f r tn ^ s tne

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 31

most nearly cosmopolitan echinoderm of all and one which must feature on a great
number of fauna lists, so that a change of generic name would cause extensive
confusion. However, after scrutinizing the literature I cannot find any explanation
why the specific name squamata has been used since 1900 in face of the obvious
priority of Ophiura elegans Leach, 1815. Leach's short description and more
particularly his notes that the disc scales are " of nearly one size and the margin has
a distinct row of scales running round it ", together with the drab colour and the
locality on the rocks in South Devon, convince me that his material was conspecific
with A . squamata. If the latter name is to be retained, then some approach is needed
to the International Commission on Zoological Nomenclature to request that the
plenary powers be exerted to reject 0. elegans Leach formally and validate 0.
squamata Delle Chiaje. Simultaneously I would have liked to be able to recommend
that the usage of Amphipholis in Ljungman's preliminary paper (i867a) be sup-
pressed in order to validate that in his main paper (i867b), where the first-named
and best-known species included is A. elegans (Leach) with squamata given as a
synonym. [A type-designation of A. squamata has already been given by Verrill
(iSgga : 24 and iSggb : 306, but not p. 311 which names januarii) .] However, it is
difficult to justify squamata as type-species in face of Ljungman's treatment of the
genus.

Accordingly I propose to suspend consideration of the adoption of Axiognathus
and Micropholis pending publication of Thomas's monograph ; meanwhile I hope
for some response by interested parties to my proposals for stabilizing the specific
name squamata. At present therefore, Amphipholis is treated in the wide sense with
only the following synonyms :

Nullopholis Fell, 1962. Type-species : Amphipholis nudipora Koehler, 1914.
Monopholis Fell, 1962. Type-species : Amphiura vitax Koehler, 1904.

In addition to the species listed by Fell, there are two others which have been
referred to Amphipholis. One of these is A. tetracantha Matsumoto, 1941, from
Japan and the other Amphiura murex Koehler, 1908, from Brazil. The holotype
of the latter is relatively small (d.d. only 2-5 mm.) and has the under side of the disc
covered with conical granules while the upper side has relatively large naked scales.
This unusual combination is shared by Ophiocnida loveni (Ljungman, I967a), also
known from Brazil, of which I believe A. murex is a synonym. I am uncertain
whether 0. loveni is congeneric with the type-species of Ophiocnida, 0. hispida (Le
Conte, 1851), the latter having the disc scales distinct on both sides and with small
spinelets scattered all over. The affinities of 0. loveni appear to be more with
Amphichondrius granulosus Nielsen, 1932 or Ophiophragmus filograneus (Lyman,
1875), once also included in Ophiocnida.

AILSA M. CLARK

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 33

OPHIOSTIGMA Liitken

Ophiostigma Liitken, 1856 : 13 ; i85gb : 233 ; Lyman, 1865 : 103 ; Ljungman, iS6jb : 317 ;
Lyman, 1882 : 164-165 ; H. L. Clark, 1915 : 243 ; Fell, 1960 : 22 ; Thomas, 1962 : 689.
Type-species : O. tenue Liitken, 1856, by monotypy.

In 1960 Fell referred Amphistigma H. L. Clark, 1938, to the synonymy of Ophio-
stigma, being evidently misled by H. L. Clark's figure of the disc of A. minuta, the
type-species, which is not very clear. Another figure of the holotype of A . minuta
is given here (fig. yf, g) which I hope shows up more conspicuously the marked
difference in the disc in comparison with Ophiostigma, though certainly both have
oral armament of the Amphipholis (C)-type. Ophiostigma tenue (fig. yd, e) has
papilliform spinelets scattered all over both sides of the disc, whereas in Amphistigma
minuta there are -only a few relatively very large papillae restricted to the margin
of the disc while in the centre the five approximately radial scales of the rosette are
extremely thickened and projecting but do not bear articulated papillae.

Amphistigma minuta H. L. Clark
% 7f> g

Amphistigma minuta H. L. Clark, 1938 : 245-247, fig. 16.

DOUGALOPLUS* nom. nov.

Amphiacantha Matsumoto, 1917 : 177-179 ; Nielsen, 1932 : 296 ; Djakonov, 1954 : 57- Type-
species : Amphioplus acanthinus H. L. Clark, 1911. Non Amphiacantha Popofsky, 1904,
Protozoa.

Unfortunately the generic name Amphiacantha used by Matsumoto is preoccupied,
as discovered recently by Miss M. Downey of the U.S. National Museum and com-
municated to me by Dr. D. L. Pawson, to both of whom I am indebted. The
replacement name Dougaloplus is therefore proposed for the genus.

Apart from the type-species, the genus also includes Ophiostigma formosa Liitken,
1872, Amphiura gastracantha and notacantha both of Liitken and Mortensen, 1899,
Ophiocnida libera Koehler, 1907 and 0. amphacantha McClendon, 1909, all referred
to Amphiacantha by Matsumoto, who also included a new nominal species Amphia-
cantha dividua and Ophiocnida sexradia Duncan, 1887, which I consider are
synonymous and referable to Ophiocomella (sexradia having been transferred in 1963

* Dougal a shaggy dog of television fame. Gender: masculine.

FIG. 7. a-c. Ophiocnida hispida (Le Conte), M.C.Z. 1496, Hassler Expedition, Panama,
b shows the twentieth free segment from above ; d and e. Ophiostigma tenue Liitken,
M.C.Z. 5712, " Velero " st. 870, Isabel Is., W Mexico (the texturing of the surface is
indicated on a few plates) ; f and g. Amphistigma minuta H.L.C., M.C.Z. 4983, holotype,
Lord Howe I. (the dorsal view slightly oblique) ; h-k. Dougaloplus echinatus (Ljungman),
h and i. Stockholm Museum collection, holotype, N. Caspar Strait, E of Sumatra, j and k.
B. M. 98.6.30.1, Batavia (Jakarta) ; 1. Amphioplus (?} luctator Koehler, U.S.N.M., Kei Is.
Expedition st. 77.

34 AILSA M. CLARK

Clark in Parslow and Clark). Matsumoto also placed Amphilimna pentacantha H. L.
Clark, 1911 in Amphiacantha ; this species has since been referred to the synonymy of
Amphiacantha amphacantha by Nielsen, 1932. Subsequent additions to the genus
have been A. transacta Koehler, 1930 and A. derjugini Djakonov, 1949. Amphia-
cantha transacta appears to have multiple apical oral (or dental) papillae and the
genital slits bordered with genital papillae unlike other Amphiurids ; it must there-
fore be a member of the Chilophiurida and should be referred to the Ophiocomidae
or possibly Ophiochitonidae. Amphiacantha derjugini, however, is certainly con-
generic with Dougaloplus amphacantha, resembling it in the spiniform two middle
oral papillae but differing in the single tentacle scale.

One further species is now referred to the genus, namely Ophiocnida (originally
Ophiophragmus] echinata (Ljungman, i867b), to which Lyman (1874) doubtfully
referred a specimen from the Philippines although he thought it differed in two
important respects from the holotype. There are two specimens in the British
Museum collections from Batavia (Jakarta), Java which agree closely with Lyman's
description and figures but again have the same two differences from Ljungman's
description. To settle their identity, I requested Dr. P. A. Andersson of the Stock-
holm Museum to lend me the holotype of Ophiophragmus echinatus, which he very
kindly did. A description of it is given below but the two supposed differences can
be dealt with here. Firstly Ljungman describes the radial shields as almost con-
cealed by the disc spines. This is true to some extent in the holotype, only the
distal ends of the radial shields being naked and exposed but it looks to me as if the
specimen may have been dried at some time (it is now in spirit) since the disc is rather
shrunken and the scaling has distorted considerably, in most cases so as to form a
fold coinciding with the radial shield and concealing it for most of its length. In
one radius this does not apply but here the scaling around the shields appears to
have been scraped free of spinelets so it is not absolutely certain that the naked
condition of the shields is natural, though I believe it is. These two shields are very
long and thin, equal in length to about half r of the disc. The second character
concerns the number of oral papillae ; Ljungman describes only three whereas Lyman
found four in his specimen with d.d. 10 mm. (at least 2 mm. larger than the holo-
type), though in a small specimen, d.d. 4 mm., Lyman notes that there are often only
three. In the holotype the first impression is that there are only three papillae but
on most of the adoral shields there is a small calcification on the edge distal to the
third papilla. At best this " papilla " is inconspicuous and it may be quite lacking ;
usually it is angled slightly down into the oral slit and partly concealed by the large
third papilla so it is not difficult to see how Ljungman overlooked it.

Besides the two specimens from Batavia there are in the British Museum three
from Macclesfield Bank, South China Sea and one from the Great Barrier Reef
(recorded by H. L. Clark), all of them I believe conspecific with the holotype of
echinatus. One Batavia specimen is discless but is similar in size to the other where
d.d is 12-5 mm. (fig. 7], k). Both of them have four oral papillae in each series but
the outermost one is distinctly the smallest and sometimes partly overlain by the
third. Two of the Macclesfield Bank specimens are also discless, though their d.d.
must have been about 7 mm. ; their fourth papilla is developed in most series but

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 35

often partly overlain. The same is true of the Barrier Reef specimen but this is
dried and the fourth papilla has a much rougher texture than the other three. The
third Macclesfield Bank specimen has d.d. only 4 mm. and the fourth papilla is just
distinguishable in a few series ; clearly its development is progressive with growth
and seems to be retarded in comparison with most species of Amphioplus, though in
1955 I noted something similar in Amphioplus congensis. All three intact specimens
have long, narrow, separated and quite bare radial shields, as in Lyman's specimen.
I think therefore that echinatus is better referred to Dougaloplus than left in Ophioc-
nida. Its affinities are probably with D. formosus from further north in the China
Sea, which resembles it in jaw structure with the two innermost papillae widely
spaced to the sides of the lowest tooth, the third papilla the largest and the fourth
the smallest ; also both have very broadly contiguous adoral shields. D. formosus
differs from echinatus in having the radial shields short, broad and contiguous.

Dougaloplus echinatus (Ljungman)
fig. 7h-k, ?1

Ophiophragmus echinatus Ljungman, 18670 : 316.

POphiocnida echinata (O. longipeda MS) : Lyman, 1874 : 230-231, pi. 4, figs. 22, 23.

Ophiocnida echinata : Koehler, 1898 : 68, pi. 5, fig. 47 ; 1905 : 32 ; H. L. Clark, 1915 : 250 ;

Koehler, 1930 : 113.
PAmphioplus luctator Koehler, ig22a : 178-179, pi. 68, figs. 1-3 ; 1930 : 106.

DESCRIPTION. I estimate that the holotype (fig. yh, i) has the d.d. 8 mm.,
although Ljungman puts the measurement at 7 mm. The arms are all broken
distally ; their length is 70+ mm., probably c. 100 mm. The disc is shrunken and
partially detached from some of the arm bases. The scaling is obscured on the upper
side by the numerous small, slightly tapering, spinelets, many of which have become
displaced and directed almost horizontally, especially across the radial shields where
the disc is distorted as described above. On the under side the spinelets are pro-
gressively fewer proximally and the scaling is distinct.

The oral shields are widest at about the middle of their length or just proximal to
this, the proximal angle is acute and the distal lobe broadly rounded. The length
just exceeds the breadth. The adoral shields are very broadly contiguous inter-
radially but have no exposed part extending distal to the lateral angle of the oral
shield. Radially they are separated by the very small first ventral arm plate. The
innermost oral papilla of each series is offset to the side of the lowest tooth and is not
strictly infradental. The second papilla is slightly smaller but the third much
larger, though it cannot really be called operculiform as in Ophiostigma and Amphi-
pholis. As described above the fourth papilla is rudimentary and sometimes appears
to be absent ; usually it is partially concealed by the third.

The arms are distinctly convex above and below with the result that the dorsal
arm plates appear almost carinate and the ventral arm plates have a broad flattened
median area but incline inwards at the sides. The dorsal plates are broad fan-shaped
with the widest part towards the distal end and the distal side becomes more or less

36 AILSA M. CLARK

flattened in the middle beyond the basal segments. The ventral arm plates are
approximately quadrangular and about as wide as long, proximally slightly wider
than long. The arm spines are slightly tapering and blunt-tipped ; they number
four just beyond the disc but this soon falls to three.

The type-locality is 2 30' S : 107 10' E, Gaspar Strait (between Sumatra and the
island of Billiton to the east), 33 metres.

SYNONYMY. I believe that Amphioplus luctator Koehler, i922a, (fig. 7!) known
only from specimens from the Philippine area and Kei Islands all lacking the disc,
will prove to be a synonym of D. echinatus. Although Koehler (1930) recorded both
nominal species simultaneously from the Kei Islands, both the (presumably com-
plete) specimens which he referred to Amphiacantha echinata were small and the
distinctive disc would have served to identify them whereas the discless ones which
he named luctator would all have been perforce identified from the mouth parts.

AMPHIOPLUS Verrill

Amphioplus Verrill, iSgga. : 25 ; 18990 : 306, 314 ; H. L. Clark, 1915 : 251 ; Matsumoto, 1917 :
169-170 ; Djakonov, 1954 : 59-6o ; Fell, igfaa. : 6, 16-17 ; Thomas, 1962 : 651. Type-
species : Amphiura tumida Lyman, 1878.

[See also p. 50 for synonym and p. 56 for diagnosis.]

The majority of species included in Amphioplus up to 1962 have two tentacle
scales and fully-scaled discs, like the type-species, A. tumidus (Lyman, 1878), so
Fell's subdivision on the basis of these characters still leaves as many as 59 species
in Amphioplus as restricted. The remainder Fell distributes between three new
nominal genera, Silax, Unioplus and Ailsaria. (The last-named is misplaced in
Fell's key as explained under the heading of Amphiodia.} Silax has the disc fully-
scaled but no tentacle scales ; Unioplus also has full scaling but one tentacle scale ;
while Ailsaria has the disc at least partially naked below but two tentacle scales.
As with Amphiodia and Amphipholis sensu extenso, the range of combinations of
these two characters is much less than in the species of Amphiura sensu extenso.
The number of species Fell refers to Silax, Unioplus and Ailsaria is respectively two,
twelve and four.

The type-species of Silax is Amphioplus verrilli (Lyman, 1879) from deep water
(down to nearly 5,000 metres) in the West Indies. In the holotype of A. verrilli the
lowest tooth on three of the jaws is distinctly tricuspid (fig. 8a) and the innermost
oral papilla of each series cannot strictly be called infradental since it is offset to the
side of the tooth ; however, on the two remaining jaws the lowest tooth is straight-
edged and there is a symmetrical, infradentally-placed pair of papillae at the apex of
the jaw. The oral plates have a slight horizontal flange level with the base of the
first oral tentacle scale and dorsal to the second oral tentacle, rendering the scale
almost superficial in comparison with its position deeper in the oral slit in most
species of Amphioplus including A. tumidus. Also, apart from the adapical papilla
of each half -jaw, there are only two distal oral papillae (or scales of the second oral
tentacle) making a total of three oral papillae, as in Amphichilus, not four as in

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 37

Amphioplus sensu stricto, the fourth one mentioned by Lyman being the first oral
tentacle scale. The oral formula is m,om,m -ft (or m,omn^n -ft), though on the
half -jaws which have the first oral tentacle scale most nearly superficial it could be
written as m,(m = t)m,m.

This kind of jaw structure is an incipient modification from that of Amphioplus
sensu stricto in the direction of Amphilepis, since Matsumoto's revision of 1915
isolated in the family Amphilepididae, the relationship of which to the Amphiuridae
is discussed further below (p. 70). An even more important modification in my
view is that of the arms of Silax verrilli. Lyman described the ventral arm plates as
swollen but I would call them keeled, though the apex of the keel is blunted. The
consecutive dorsal arm plates are separated, even basally, but this condition is
shared by a number of other deep-water amphiurids (as well as by Amphilepis).

The combination of the keeled arms with the oral structure and the complete
absence of tentacle scales even basally in my opinion justifies a generic distinction of
verrilli from Amphioplus tumidus ; I therefore support Fell in this instance in
recognizing Silax as a valid genus.

Fell (1962) tentatively referred a second species to Silax, namely Amphioplus
trepidus (Koehler, 1904). However, as he comments, this species is not at all
Amphiurid-like, not only does the disc merge with the arms dorsally but the radial
shields are small and very widely spaced, especially at their distal ends. I have
examined the holotype and find that it has a group of tooth papillae at the apex of
each jaw, numbering six or seven but with two of the more superficial ones arranged
as a symmetrical pair, simulating the infradental pair of oral papillae of amphiurids ;
usually there is also a single smaller median papilla superficial even to these two.
In addition the dorsal-most arm spine is often enlarged on alternate sides of the arm
and there are remnants of about eight irregularly-placed spines on the disc, one as
much as 0-35 mm. long, the d.d. being only 4 mm. Despite the considerable depth
(289 metres) I am certain that the species is referable to the family Ophiocomidae
and probably to the genus Ophiomastix, which also has disc spines and enlarged
uppermost arm spines.

There are, however, a number of other species which show some degree of affinity
with Silax verrilli. These include the following :

Amphilepis patens Lyman, 1879 W South America ; m,tt,om (fig. 8c)
Ophiomonas protecta (Koehler, 1904) E Indies ; m,oo,m -ft (fig. 8d)
Ophiomonas bathybia Djakonov, 1952 W Pacific ; m,oo,m -ft
Amphioplus daleus (Lyman, 1879) E Pacific-N and S Atlantic ; m,(m = tjmnji

(fig. 81)

Amphioplus patulus (Lyman, 1879) Southern Ocean ; m,(m = t) mnji (fig. 8j)
Amphioplus cernuus (Lyman, 1879) N Pacific ; m,(m = t)mn^h
Amphiodia grata (Koehler, 1904) E Indies ; m,(m = t)m,m (fig. 8k)
Amphioplus falcatus Mortensen, I933a SE Africa ; m,(m = t)m,m
Amphioplus confinis (Koehler, 1904) E Indies ; m,(m = t)mnji
Amphiura koreae Duncan, 1879 Japan ; m,om,m -ft (fig. 8h)
Amphioplus gentilis (Koehler, 1904) E Indies ; m,oM,m -ft (fig. 8r)
Amphiodia servata (Koehler, 1904) E Indies ; m,oM,m -ft (fig. 8p)

AILSA M. CLARK

s

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDE A) 39

Amphioplus incisus (Lyman, 1883) W Indies ; m,om,m -ft (fig. 8q)
Amphiura concolor Lyman, 1879 Japan ; m,om,m -ft (fig. 8 g)
Amphiodia caulleryi (Koehler, 1897) Indian Ocean ; m,onjim -ft?
Amphiura diomediae Ltk. & Mtsn., 1899 Indo-Pacific ; m,onjim -ft (fig. 8n)
Amphiura lunaris Lyman, 1878 W Indies ; m,oo,mm -ft (fig. 8v)

Amphichilus cesareus (Koehler, 1905) E Indies ; m,(m = t?)mn,n
Amphiodia reposita (Koehler, 1905) E Indies ; m,oN,Nm -ft (fig. 8m)
Amphioplus conditus (Koehler, 1905) E Indies ; m,om,m -ft (fig, g s )
Amphiodia ochroleuca (Brock, 1888) Indo-Pacific ; m,m,m -ft (fig. 8t)
Amphichilus trichoides Matsumoto, 1917 Japan ; m,n/nm -ft
Amphiodia affinis (Studer, 1885) Southern Ocean ; m,omm,o -ft (fig. 5])
Amphiura bidentata H. L. Clark, 1938 Australia ; m,omm, -ft (fig. 8u)
The generic names given in the list are those currently in use before 1962 and
mainly correspond to the ones given in H. L. Clark's catalogue (1915). The oral
formula in some of the species is subject to variation but the version given is believed
to be the most common one. The seven species below the dividing line are from
shallow water, those above from deep ; only the southern Amphiodia affinis ranges
from shallow depths to several hundred metres.

A tricuspid lowest tooth similar to that found in the holotype of 5. verrilli, though
a little smaller, occurs on two jaws of the holotype of Amphilepis patens (fig. 8c),
the two innermost papillae again being offset and not truly infradental, though on

FIG. 8. a and b. Silax verrilli (Lyman), 82.12.23.17, holotype, " Challenger " st. 54,
N of Bermuda ; c. Amphilepis patens Lyman, 82.12.23.14, holotype, " Challenger " st.
299, off Chile (the tip of one first oral tentacle just shows at top left) ; d. Ophiomonas
protecta (Koehler), U.S.N.M. 41013, " Albatross " st. 5359, Philippine Is. ; e. Amphioplus
(Unioplus) sp. (Amphiura concolor according to Lyman), 82.12.23.455, " Challenger "
st. 191, Kei Is. ; f. Amphilepis norvegica (Ljungman), 98.5.3.688, Trondhjem Fjord ;
g. Amphiura concolor Lyman, 82.12.23.107, holotype, " Challenger " st. 195, Banda Sea ;
h and i. Amphiura koreae Duncan, 80.1.3.4, holotype, Japan Sea (i shows the large tentacle
scales on a proximal arm segment) ; j. Amphioplus (Unioplus) patulus (Lyman),
82.12.23.369, holotype, " Challenger " st. 156, Southern Ocean, SW from Australia ;
k. Amphioplus (Unioplus) gratus (Koehler), Amsterdam Museum, syntype, " Siboga "
st. 211, E Indies ; 1. Amphioplus (Unioplus) daleus (Lyman), 82.12.23.370, holotype,
" Challenger " st. 325, SW Atlantic, oblique ventral view of superficial part of one side
of a jaw ; m. Amphioplus (Unioplus) repositus (Koehler), Amsterdam Museum, syntype,
" Siboga " st. 89, E Indies ; n and o. Amphiura diomediae Liitken and Mortensen, n.
1948.5.26.56 (pt.), John Murray Expedition st. 54, S Arabian coast, o. 1966.1.20.9,
" Albatross " st. 5053, Suruga Gulf, Japan (tentacle scales for comparison with i) ;
p. Amphioplus (Unioplus) servatus (Koehler), Amsterdam Museum, holotype, " Siboga "
st. 262, (showing the single series of four oral papillae on the left) ; q. Amphioplus (Unio-
plus) incisus (Lyman), M.C.Z. 1468, holotype, " Blake " st. 161, off Guadeloupe, W Indies;
r. Amphioplus (Unioplus) gentilis (Koehler), Amsterdam Museum, syntype, " Siboga "
st. 178, E Indies (see also fig. nf and g) ; s. Amphioplus (Unioplus) conditus (Koehler),
Amsterdam Museum, holotype, " Siboga " st. 89, E Indies ; t. Amphioplus (Amphichilus)
ochroleucus (Brock), M.C.Z. 4986, Broome, NW Australia ; u. Amphiura bidentata H.L.C.,
M.C.Z. 4947, holotype, Broome ; v and w. A. lunaris Lyman, M.C.Z. 1324, holotype,
" Blake " st. 29, Gulf of Mexico.

4 o AILSA M. CLARK

the three other jaws the teeth are all broadly rounded and the two papillae lie at
least partially below the lowest one. As is usual in the genus Amphilepis, the apex
of the pair of oral plates is less truncated than in most amphiurids, though less so in
patens than in the type-species, Amphilepis norvegica (fig. 8f), and these two inner-
most papillae are relatively broad and short with their bases at an angle to each other,
not in the same plane. Also the oral plates have a superficial flange bearing the first
oral tentacle scale. Although in A. norvegica this oral tentacle scale is operculiform
and extends right to the first ventral arm plate, in A . patens it is broken up into two,
sometimes three, separate papillae while distally it curves outwards as a rim almost
encircling the second oral tentacle, which is superficial in Amphilepis ; ventral to
the second tentacle, on the edge of the adoral shield, this rim is again rather irregu-
larly divided abradially, usually with a single papilla marked off from it, though
Liitken & Mortensen (1899) fig ure two papillae in this position in their " Albatross "
specimen. Lyman failed to record the distal papillae of the holotype.

A very similar oral structure occurs in Ophiomonas bathybia with which Djakonov
justifiably believes Amphilepis protecta to be congeneric. In Ophiomonas protecta
(fig. 8d) the operculiform first oral tentacle scale is not quite superficial, hence the
addition of t at the end of the formula. A similar condition appears to occur in
Amphilepis diastata Murakami, 1942, which may have the second tentacle even
more inset into the oral slit, judging from Murakami's pi. 4, fig. 34 (1963). In 1965
I provisionally referred diastata to the synonymy of Amphiura koreae (that is sensu
Matsumoto, which I now think may be better referred to diomediae, see pp. 41-42) .
It should certainly be included in Amphiura, the jaw structure is quite unlike that of
Amphilepis. Before discussing Amphiura koreae there are at least seven species in
which the first oral tentacle scale is papilliform and almost superficial, coming more
or less into series with the oral papillae ; these include Amphioplus daleus, patulus,
cernuus, cesareus, confinis andfalcatus, also Amphiodia grata, all of which have a basic
oral formula of 3m +t like Silax verrilli. The first three also approximate to 5.
verrilli in reduction of the tentacle scales, though this was not remarked on by
Lyman. In fact the largest of the three holotypes, that of Amphioplus patulus
(d.d. 14-5 mm.) is the only one to have any tentacle scales beyond the eighth arm
segment ; they extend to about the twenty-fifth. The types of A. daleus and
cernuus have d.d. respectively 9 mm. and just under 6 mm. ; the former has the
single scales on only the first two segments (with one exception on the third). Some
descriptive notes are given below (pp. 44, 45, 58-59 and 62-63) on type-material of
the other four species.

A similar 3m -ft formula occurs in all the other species in the list but in these the
first oral tentacle scale is normally deeper in the oral slit and not in the same super-
ficial plane as the oral papillae. Descriptive remarks for Amphioplus gentilis,
Amphiodia ochroleuca, reposita and servata are given on pp. 46, 62 and 65-66 and they
are also illustrated in fig. 8, as are Amphioplus incisus and conditus. It can be seen
that all of them have a semblance of continuity in the series of oral papillae, the
small diastema between the infradental one and the two distal papillae being partially
filled by the oral tentacle scale above. This contrasts with species such as Amphiura
lunaris (fig. 8v, w) and digitula, in which the two distal papillae are much more

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDE A) 41

sharply dissociated from the infradental by the long diastema.

A common feature among the deep-water species in the list is irregularity in the
symmetry of the infradental papillae. These may be unequal in size (e.g. in the
holotype of Amphiura concolor, fig. 8 g), or subdivided, usually with three papillae
in place of the two (as in the specimen of Ophiomonas protecta illustrated here, fig.
8d). Matsumoto (1917) figured this latter condition in what he calls Amphiura
koreae and it is also found in the holotype of Amphioplus patulus (though not in the
paratype), causing Matsumoto to refer the latter species to Amphiactis which is in-
correct, the teeth and the apices of the jaws being broad. The positions of the distal
papillae and their relations with the oral tentacle scale are also subject to some
variation. Again this has been noted particularly by Matsumoto in the species he
calls Amphiura koreae, in which the inner of the two distal papillae may be based
entirely on the oral plate rather than partially on the adoral shield and sometimes
there may be an additional papilla making a total of four or conversely one of the
two distal papillae may be lost. Having studied Duncan's holotype of Amphiura
koreae (fig. 8h, i) I am not altogether convinced that Matsumoto was right in
synonymizing with it Amphiura diomediae Liitken & Mortensen (1899) (type-
locality in the East Pacific) despite the considerable variation also reported by Liitken
and Mortensen. Not only is the oral structure of the holotype of A . koreae inconsistent
with its inclusion in Amphiura, the second oral papilla being entirely on the side of
the oral plate, but also the oral plates themselves are relatively short, leaving only a
very small diastema between the infradental and second papillae, the arm spines are
tapering but slightly blunted at the tips and the two tentacle scales (fig. 8i) are
distinctly larger than in any of the specimens which I have seen with oral structure
like the holotype of diomediae (see Liitken and Mortensen's pi. 12 figs, i, 2), including
two paratypes. It also seems to me that in " typical " diomediae not only are the
arm spines sharper, the tentacle scales smaller and the oral diastemas longer, but also
the disc is smoother with flatter scales showing little change of size at the marginal
junction, while the other surface plates are also thinner. At least some of the differ-
ential characters shown by the holotype of A . koreae are shared by Liitken and Mor-
tensen's unnamed variety of diomediae (1899, pi- I2 > fig 5 - 5~?)> namely the shorter
oral plates and the more marked size change between dorsal and ventral disc scales.
Koehler's specimen (1922, pi. 67, fig. 9) from the Philippines shares at least the
second of these characters besides having thicker, and incidently less imbricating,
disc scales in the centre. The specimens from the Maldive area (John Murray
Expedition station 143) recorded as diomediae by H. L. Clark in 1939 also have
thickish disc scales, coarser slightly blunted arm spines and the second oral papilla
based entirely on the oral plate though the plate itself is relatively longer and the
diastema larger than in the holotype of A. koreae. The other Murray Expedition
specimens, from off the Arabian coast, are more like the holotype of diomediae with
longer oral plates, sharper spines and smoother discs. I am inclined to think that
there are two forms intergrading to some extent but it will need some much larger
samples (particularly from Japan) than are available to me at present to investigate
this. Unfortunately these two forms together bridge the borderline between Amphi-
ura and Amphioplus, the holotypes of diomediae and koreae respectively agreeing in

42 AILSA M. CLARK

their oral structures with the accepted diagnoses of the two genera Amphiura with
a large diastema and one or sometimes two distal oral papillae besides the infra-
dental one, as opposed to Amphioplus with a more or less continuous series of four or
sometimes three papillae, in each case with the addition of the oral tentacle scale.
This emphasizes the close relationship between the two genera. Since none of the
other specimens seen by me or figured by others shows either such a small diastema
or such large tentacle scales as the holotype of A . koreae, I propose to term all of
them Amphiura diomediae. [As remarked opposite there is also a possibility that A .
diomediae is conspecific with Amphiura caulleryi Koehler, 1897.] The best generic
position for specimens with the oral structure of the holotype of A . koreae is discussed
further on pp. 43-44.

The great geographical range of Amphiura diomediae suggests that some of the
other deep-sea nominal species in the list above may prove to be synonymous, for
instance Amphioplus daleus, patulus and cernuus, despite the wide separation of their
type-localities.

As I mentioned in 1965, the " Challenger " specimens from station 191 (fig. 8e)
named Amphiura concolor by Lyman are not conspecific with the holotype but are
more closely related to Amphioplus confinis (Koehler, 1904), differing in having much
thinner scales on the disc though the mouth parts are rather similar ; two jaws out of
five of the specimen figured have irregular and multiple papillae at the apex but the
other three are undeniably amphiurid in form. A similar irregularity often occurs in
Amphiodia affinis, not only in the apical papillae but also in the more distal ones.
The usual arrangement of these is shown in fig. 5] , with the two more distal papillae in
series with, though slightly spaced from, the infradental and the third one not quite
contacting the adoral shield. However, this is not always the case and there may be a
fourth papilla present, also the apical arrangement varies widely, the lowest tooth is
often distinctly conical, which provides some justification for Hertz's inclusion
(1927) of the species in the genus Amphiadis, though the blunt apex of the jaw itself
precludes this, while the infradental papillae vary in their spacing and size. An un-
usual feature is that the oral tentacle scale may be duplicated with two papillae
arranged obliquely. [This contrasts with the duplication I have also seen in occa-
sional specimens of species such as Amphiura diomediae in which the oral tentacle
scale is considerably broadened horizontally and can be split into two parts, only the
innermost being at all papilliform.] In fact Amphiodia affinis was included in
Amphioplus by Koehler (1917) but transferred to Amphiodia by Mortensen (1936)
followed by Madsen (1955). It is a very anomalous species with its very thick disc
scales, convex disc, widely-separated radial shields and very short arms tapering
abruptly at the tip to a large conical terminal plate and I can understand Studer's
reference of a small specimen to the genus Ophioceramis (family Ophiuridae) ; only
the jaw armament resembles that of amphiurids, especially those from deep water in
which the apical structure is similarly somewhat variable. I suspect that a more
detailed study of the species will lead to its inclusion in a genus distinct from
Amphioplus but it certainly cannot be left in Amphiodia.

Notes on the type-material of Amphioplus confinis and some other interesting
'Siboga' species of Koehler are given on pp. 58 to 66 at the end of the section on

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 43

Amphioplus together with a description of a new genus and species based on a
specimen in the Museum of Comparative Zoology, Harvard which Liitken & Morten-
sen (1899) misidentified as Amphiura (i.e. Amphioplus) dalea.

There are two species in the list which are fairly remote from Silax verrilli and
stand apart from the others by the symmetrical pairing of their two distal oral
papillae and their isolation from the infradental one. Firstly Amphiura lunaris
(fig. 8v, w), of which the holotype (d.d. only 3-5 mm.) has two, sometimes one, distal oral
papillae close together on the edge of the adoral shield and an almost superficial
flange on each oral plate, the edge of which is indistinctly subdivided in rather the
same way as that of Amphilepis patens (fig. 8c) ; however, the apex of the jaw is
broadly truncated with wide rounded teeth and the infradental papillae are not at all
divergent, unlike those of Amphilepis. The consecutive dorsal arm plates are not
contiguous but this is common to a number of deep-water amphiurids. A better
range of material is needed to show the true affinities of the species. Secondly
Amphiura bidentata (fig. 8u), a shallow- water species, has a rather smaller
diastema between the infradental and the two distal oral papillae, both of which are
based on the more superficial part of the oral plate, not on the edge of the adoral.
These two species are not closely related to each other but it seems best to leave both
of them in Amphiura. The same may be true of Amphiura caulleryi (referred to
Amphiodia by H. L. Clark in 1915), judging from Koehler's figure which shows a
large diastema between the infradental and the pair of distal papillae situated on the
adoral " dans Tangle buccal " " recouvrent de chaque cote 1'orifice du pore tentacu-
laire buccal " according to Koehler. In spite of the presence of the first oral tentacle
scale on the side of each oral plate, Koehler (1930) follows H. L. Clark in including the
species in Amphiodia. Until the type-material can be re-examined, the true position
of the species cannot be determined though it probably belongs to Amphiura,
possibly to Amphioplus, but certainly not to Amphiodia. Koehler's early figures
were very diagrammatic and it may be noted that his illustration of Amphiura
reposita (1905, pi. 4, fig. 13) similarly shows the inner of the two distal papillae as
based entirely on the adoral shield whereas in fact it is more often entirely on the
oral plate in the holotype and never more than partially in contact with the adoral
(see my fig 8m). However, it should be noted that Koehler in 1904 referred Amphi-
ura caulleryi to the Amphioplus section and simultaneously identified a specimen as
Amphiura diomediae, referring that species to the Amphiodia section. Were it not
for this, I would have thought that caulleryi and diomediae could be synonymous,
judging from Koehler's description of 1897 ; the type-locality of caulleryi is in the
region of Ceylon and diomediae has been taken in the Maldive area nearby. A
solution of the problem must await a re-examination of the holotype of A . caulleryi.

Although the diastema in the papilla series of the holotype of Amphiura concolor
is hardly more conspicuous than that of species such as Amphioplus gentilis and
incisus and the second papilla arises wholly from the oral plate, I am inclined to
leave concolor in the genus Amphiura, unlike the second specimen referred to the
species by Lyman (fig. 8e), which is certainly congeneric with Amphioplus falcatus.
The problem is even more questionable when it comes to the holotype of Amphiura
koreae. Strictly speaking, this specimen cannot be regarded as an Amphiura, the

44 AILSA M. CLARK

diastema being so small ; its affinities appear to be all with species such as Amphio-
plus daleus, as far as oral structure goes. However, in view of its other close relation-
ship and even possible identity with Amphiura diomediae, it would be irresponsible to
refer it to a different genus than Amphiura, at least until a more detailed study can be
made of large samples from Japanese waters.

It remains to determine the best generic position for the rest of the species in the
list above, excluding those just named as better retained in Amphiura and the three
first species, Amphilepis patens, Ophiomonas bathybia and 0. protecta at the top.
Amphiodia can be ruled out to start with because of the presence of the first oral tentacle
scale. This leaves Unioplus Fell, Amphichilus Matsumoto, Amphioplus or Amphiura.

The type-species of Unioplus is Amphioplus falcatus Mortensen, I933a,
from moderately deep water (c. 400 metres) off SE Africa. It has the oral tentacle
scale more or less superficial and in series with the true oral papillae, which number
three more often than four. A paratype of A. falcatus in the British Museum
collections shows the considerable variation described by Mortensen in the number
and arrangement of the papillae and also has a third infradental papilla between the
usual two on three of the five j aws. About half of the papilla series have four papillae
plus the oral tentacle scale, the formula being m, (m = t)m, mm but in the rest it is
m, (m = t)m,m with minor variations in the precise position of the distal papillae
and the vertical alignment of the scale. The dorsal arm plates are barely contiguous
but not so markedly separated as in the deep-water species such as A . daleus ; also
the single tentacle scale extends for almost the entire length of the arm and is not
restricted to a few basal segments. The last character distinguishes falcatus from
cernuus and patulus which Fell refers to Unioplus since both have been described as
having a single tentacle scale. Amphioplus daleus he omitted since it had been
referred by Matsumoto to Amphichilus, an interesting taxon intermediate between
Amphiura and Amphioplus (at least as regards oral armament), which was un-
fortunately not taken into consideration in Fell's revision.

There is some affinity between these three nominal species of Lyman (daleus,
cernuus and patulus} and A . falcatus but the same cannot be said of most of the other
species which Fell has referred to Unioplus. At least four of them, Amphioplus
strongyloplax (H. L. Clark, 1911), A. thrombodes (H. L. Clark, 1918) A. philohel-
minthius Ziesenhenne, 1940 and A. glaucus (Lyman, 1879) differ from falcatus in
having bare skin with only scattered scales over part or all of the ventral side of the
disc (according to Fell's system thereby justifying further subdivision) and the first
two, together with A. macraspis (H. L. Clark, 1911), also disagree with the diagnosis
of Unioplus in having two tentacle scales on some of the proximal pores rather than
one. They differ too in having the oral tentacle scale higher in the slit and usually
four oral papillae (though in strongyloplax and macraspis the fourth or outermost
rather than the second papilla may be lacking.) Of the other species referred to Unio-
plus, Amphioplus formatus (Koehler, 1905) (fig. na, and see pp. 61-62) is probably
not an Amphiurid at all ; the holotype and only recorded specimen has very narrow
elongated teeth and the two innermost papillae of each jaw are offset to the sides of
the lowest tooth, though they each form a straight series of four with the other oral
papillae, the distalmost of which is the broadest ; the formula is njimm,M t. The

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 45

oral shields are relatively very small and almost circular and the adorals are com-
paratively large and very broadly contiguous inwardly. Tentacle scales are lacking
altogether on the first four to seven segments. The arms are only about three times
the disc diameter and the arm spines are extremely long, even exceeding the width of
the arm. The holotype of Amphioplus capax (Koehler, 1905) has d.d. only 1-5
mm. and was very decalcified when I saw it recently ; most of its characters are
juvenile ones but Koehler figures four oral papillae in a straight row and unusually
elongated oral shields. The type of Amphioplus aciculatus Mortensen, 1936 is
also small, d.d. 2 mm. ; its oral formula is m,om,mm+t and the inner one of the two
distal papillae is spiniform. The holotype of Amphioplus dispar (Koehler, 1897)
also has the second from outermost papilla spiniform but otherwise shows no par-
ticular resemblance to aciculatus.

With the possible exception of A. daleus and its two allies and A. falcatus, there
appears to me to be no particular affinity between any of these species and I do not
consider that Unioplus can be maintained on the basis of the single tentacle scale
alone. However, it may be worth redefining in terms of the oral structure for some of
the species in the list on pp. 37-39. Even so, the oral structure is liable to confusion
with that of Amphichilus, which also has only three oral papillae, though in the type-
species, Amphichilus trichoides Matsumoto, 1917, the tentacle scale is distinctly
higher in the oral slit and the formula is better expressed as m,onjim,+t than as
m,(m = t)ivim, as in Amphioplus falcatus. The radial shields of Amphichilus
trichoides are almost fully contiguous, unlike those of most of the other species under
consideration ; Matsumoto included this in his generic diagnosis. He also referred
Amphioplus daleus (Lyman, 1879) as we ^ as A. intermedius (Koehler, 1905) to
Amphichilus. I do not consider A . daleus is closely related to Amphichilus trichoides ;
apart from being a deep-sea species it has separated or barely contiguous radial
shields, non-contiguous dorsal arm plates, no tentacle scales (except a single one on a
few of the basal pores) and irregular apical oral papillae on some jaws. In the
holotype of Amphioplus intermedius I found the second papilla to be perfectly in
series with the other three, the four oral papillae closing the slit so well that it is
impossible to tell without mutilation if the oral tentacle scale is also present behind
the papillae ; the general appearance of the jaws agrees well with that of Amphio-
plus impressus (Ljungman, i867b), unlike Amphichilus trichoides.

Koehler (1930) has referred Amphioplus cesareus (Koehler, 1905) to Amphichilus
with some justification ; it has almost fully contiguous radial shields and comes from
relatively shallow water, usually 5-100 metres ; in one of the two syntypes in the
Amsterdam Museum the second oral papilla is slightly inset in some series and with its
conical rather than square shape is clearly homologous with the oral tentacle scale
which is otherwise missing. However, after comparison of specimens named both
Amphiura cesarea (or Amphichilus cesareus) and Amphioplus impressus by Koehler at
different times (see pp. 63-64) with the same author's description and figures of the
type-material of A . impressus, I have come to the conclusion that cesarea is a synonym
of impressus . The normal position of the second oral papilla in the species is superficial
(fig. gm) and, like A. intermedius, I do not think that it can be referred to Amphi-
chilus. These two species differ most obviously from one another in the much finer

46 AILSA M. CLARK

disc scaling of intermedium ; together they form a link between Amphichilus and
Amphioplus sensu stricto as well as between the latter and Amphipholis, from which
only the presence of a fourth oral papilla distinguishes them, the outermost papilla
approximating to an opercular form. However, I believe that Amphiodia ochro-
leuca (Brock, 1888) (fig. 8t) can be referred to Amphichilus ; it is another shallow-
water Indo-West Pacific species and cannot be left in Amphiodia since it has the first
oral tentacle scale present ; this scale is distinctly higher than the oral papillae, of
which the two outer ones are fairly broad and rectangular. Like A. trichoides the
radial shields are contiguous, though in some specimens only for their distal halves.
In contrast, all the other species under consideration, like Amphioplus falcatus,
have the radial shields only contiguous by their distal adradial corners or else
separated altogether. Apart from the single antiboreal species Amphiodia affinis
(Studer, 1885), which ranges in depth from shallow-water to several hundred metres,
and with the exception of Amphiodia reposita (Koehler, 1905) and Amphioplus
conditus (Koehler, 1905), all these are from deep-water (400+ metres, many of them
1000+ metres). [By coincidence, the type- and only recorded specimens of the
two latter species were labelled as coming from " Siboga " station 89, the depth of
which was only n metres ; it seems suspicious that they have not since been collected

FIG. 9. Amphioplus spp. (except t). a. A. tumidus (Lyman), M.C.Z. 1476, holotype,
" Blake " st. 47, Gulf of Mexico ; b. A. strongyloplax (H. L. Clark), U.S.N.M. 25615,
holotype, off Washington ; c. A. cyrtacanthus H. L. Clark, M.C.Z. 1489, holotype,
Philippine Is. ; d and e. A . coniortodes H. L. Clark, M.C.Z. 4217, holotype, Key West, Florida,
the disc shrunk away from the arm base ; f-i. A . abditus (Verrill), showing the exceptional
variation in the oral papillae and oral tentacle scale, f. B.M. 90.8.23.43, Martha's Vineyard
(the oral shield shown is the madreporite), an oral tentacle scale occurs on one side in
addition to the second oral papilla (the two oral tentacles are also shown on this side),
g. U.S.N.M. 18007, holotype of A. macilentus (Verrill), a synonym, off New England, with
five superficial papillae one side and four the other, h. M.C.Z. 1456, Noank, Connecticut,
with no second papilla but the oral tentacle scale more or less superficial, i. M.C.Z. 4842,
Charleston, South Carolina, with superficial second papilla presumably representing the
oral tentacle scale though rounded in shape and with an interstitial papilla between third
and fourth on one side (and in five other series) ; j and k. A . stenaspis H. L. Clark, M.C.Z.
4990, holotype, near Darwin ; 1 and m. A. impressus (Ljungman), U.S.N.M. 41171, "Alba-
tross " st. 5371, Philippine Is. ; n. A. archeri A.M. Clark, B.M. 1956.5.23.1, paratype,
Ghana ; o. A. cinctus (Koehler), B.M. 1956.5.23.21, off Accra, Ghana, the two middle
papillae of the left-hand series reflexed ; p and q. A. (Lymanella) hastatus (Ljungman),
p. Amsterdam Museum, " Siboga " st. 167, W of New Guinea, " absolument conforme a
1'exemplaire type de Ljungman " (Koehler), q. B.M. 1948.5.26.92, John Murray Expedi-
tion st. 89, northern Arabian Sea (d.d. only 3-5 mm.), the three distal oral papillae
removed on one side to show the rudimentary oral tentacle scale ; r. A. (Lymanella)
megapomus H. L. Clark, U.S.N.M. 25633, holotype, " Albatross " st. 4964, Japan ;
s. A. (Lymanella) laevis (Lyman), M.C.Z. 5590, syntype of A. praestans (Koehler),
"Siboga " st. ii, Java Sea ; t. Paracrocnida persica Mortensen, B.M. 1964.7.17.26, Eilat,
Gulf of Akaba ; u. Amphioplus (Lymanella) depressus (Ljungman), Amsterdam Museum,
" Siboga " st. 51, Molo Strait, E Indies (named Amphiura velicta by Koehler, 1905),
edge of disc from above ; v. A, (Lymanella) spinulosus (Koehler), Amsterdam Museum,
paratype, " Siboga " st. 319 (not in published report but added MS to Amsterdam
Museum copy), Java Sea, edge of disc from below ; w. A. (Lymanella) laevis (Lyman,
M.C.Z., holotype, third dorsal arm plate (from drawing by Miss A. Schoener) ; x. A.
(Lymanella) megapomus H. L. Clark, holotype (as above).

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 47

48 AILSA M. CLARK

if this was true.] It is possible therefore to redefine Uniophts in terms of the
formula (often with t = m) coupled with the divergent or separated radial shields
contrasting with those of Amphichilus. Also most of the species to be included in
Unioplus show some irregularity in the armament of the apex of each jaw and the
consecutive dorsal arm plates are barely contiguous or even separated.

Neither Unioplus nor Amphichilus seems to me to be sufficiently distinct for
recognition at the generic level. There is clearly considerable variation in the oral
structure, not only in the species already discussed but also in others such as Amphio-
plus abditus (Verrill, 1871) (see fig. gf-i) and Amphioplus timsae Mortensen, 1926. In
addition, as I remarked in 1965, the relationship between Amphiura and Amphio-
plus is very close, as exemplified by the Amphiura koreae-diomediae problem detailed
above, as well as by some geographically coincident species such as the Japanese
Amphiura digitula (H. L. Clark, 1911) an Amphiura with two distal oral papillae
and Amphioplus ancistrotus (H. L. Clark, 1911). Accordingly I think it best to
reduce both Unioplus and Amphichilus to the rank of subgenera of Amphioplus.
This accords with Verrill's original diagnosis of Amphioplus as having four or five oral
papillae in each series since he included the oral tentacle scale in his papilla count.

AMPHIOPLUS subgenus UNIOPLUS Fell, re-defined

Amphioplus (part) : Mortensen, iQ33a, and others.

Unioplus Fell, ig62a : 16. Type-species : Amphioplus falcatus Mortensen, iQ33a.

DIAGNOSIS. A subgenus of Amphioplus including species with the second oral
papilla normally suppressed but the first oral tentacle scale often more or less super-
ficial and taking a place in the papilla series after the infradental papilla, so that the
oral formula usually approximates to m, (m = t)m,m or m,om,m + t ; the infradental
papillae and lowest tooth usually rather variable in size, shape and arrangement,
sometimes asymmetrical ; the two radial shields of each pair either just touching
by their distal adradial corners or else completely separated by the scales ; successive
dorsal arm plates barely if at all contiguous ; tentacle scales numbering one or two,
sometimes none except on the basal pores.

The following nominal species are referable to Unioplus 1 :

Amphiura cernua Lyman, 1879 : 28. N Pacific, E from Japan.

Amphiura dalea Lyman, 1879 : 27. SW Atlantic.

Amphiura patula Lyman, 1879 : 31. Southern Ocean, SW from Australia.

Amphiura incisa Lyman, 1883 : 250. West Indies.

Amphiura affinis Studer, 1885 : 162. South Georgia.

? Amphiura caulleryi Koehler, 1897 : 330. Ceylon area.

Amphiura confinis Koehler, 1904 : 89. East Indies.

Amphiura gentilis Koehler, 1904 : 86. East Indies.

Amphiura grata Koehler, 1904 : 85. East Indies.

Amphiura servata Koehler, 1904 : 84. East Indies.

1 In original combination, chronological order and with type-localities.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 49

Amphiura condita Koehler, 1905 : 46. East Indies.
Amphiura reposita Koehler, 1905 : 40. East Indies.
Amphiura falcatus Mortensen, I933a : 365. SE Africa. Type-species.

Of all these, only Amphioplus cernuus and patulus were included in Unioplus by
Fell, besides the type-species.

AMPHIOPLUS subgenus AMPHICHILUS Matsumoto
Amphichilus Matsumoto, 1917 : 175. Type-species : Amphichilus trichoides Matsumoto, 1917.

DIAGNOSIS. A subgenus of Amphioplus including species with the second oral
papilla suppressed and the first oral tentacle scale inset, so that the oral formula is
m,m,m + t or m,njim + t ; armament of jaws normally symmetrical ; the two radial
shields of each pair contiguous for at least the distal half of their length, often more ;
dorsal arm plates contiguous, at least medially ; two tentacle scales in the two
species now included.

The following are referable to Amphichilus :

Amphiura ochroleuca Brock, 1888 : 484. East Indies.

Amphichilus trichoides Matsumoto, 1917 : 175. Japan. Type-species.

AMPHIOPLUS sensu extenso (cont.)

As for Ailsaria, apart from the type-species, Amphioplus echinulatus Morten-
sen, 1940, ~F ellreterstoitAmphioplusarcheri A. M. Clark, 1955, A. coniortodes (H. L.
Clark, 1918) and A. personatus (Koehler, 1905). Of these, Amphioplus archeri is
not at all closely related to echinulatus ; its radial shields are short, wide and fully
contiguous, rather than elongated and more than half separated proximally and the
oral shields too are quite a different shape ; the oral formula (from Mortensen's
figure) is m,onjimni+t in A. echinulatus but normally m,mm,onji t in A. archeri
(figs. li, gn), there being no oral tentacle scale on most jaws and a distal rather than
proximal gap in the sequence of papillae ; the arm structure is also very different in A .
archeri. In Amphioplus coniortodes(fig. gd, e) the dorsal disc scaling is fine and
indistinct, the oral shields are triangular with no distal lobe, the oral formula is
m,omm,om+t or m,(m = t)mm,om, the oral tentacle scale being more or less in
series with the oral papillae while the arm spines are simple at the tip, numbering only
three rather than five with hooked tips, as in echinulatus, or eight proximally as in A .
archeri. In the holotype of Amphioplus personatus, (see also pp. 64-65) the oral
formula is m,omn^nnTn+t, the distalmost papilla being based partly on the first
ventral arm plate ; with its rather similar oral formula, divergent radial shields, five
arm spines and spearhead-shaped oral shields with a distal lobe (though this is a little
wider than it appears in Koehler 's figure) A. personatus may prove to be closely
related to A. echinulatus (both of them are from the Indo-West Pacific) but it does not
agree with the two Atlantic species, archeri and coniortodes. Because of its oral

D

50 AILSA M. CLARK

structure, I think A. archeri is more closely related to Amphioplus cinctus (Koehler,
1914) (fig. 90) which has the disc fully scaled below, while Amphioplus echinulatus
shows a parallel affinity for the fully-scaled A. timsae Mortensen, 1926.

These observations lead me to the conclusion that the following is not a valid
taxon of the genus-group and should be referred to the synonymy of Amphioplus
Verrill, 1899 :

Ailsaria Fell, (type-species Amphioplus echinulatus Mortensen, 1940).

As I noted in 1965, among the many species of Amphioplus with two tentacle
scales, a natural group exists which differs from the type-species, A. tumidus (Lyman,
1878) (fig. ga) in having the four oral papillae in a continuous row, their free edges
forming a straight line so that they are capable of closing the oral slit almost com-
pletely, the third papilla being distinctly enlarged, often truncated conical in shape,
while the oral tentacle scale is rudimentary and concealed behind the papillae, if
present at all ; at the same time the radial shields are contiguous for more than half
their length, there are only three arm spines and the two tentacle scales are large or
very large, also there is often some specialization of the disc scales in the marginal
region at the line of junction of the dorsal and ventral scales.

The most frequently recorded species of this group are probably Amphioplus
depressus, integer and hastatus (fig. gp, q), all described (as Amphipholis) by Ljung-
man, i867b, also Amphioplus laevis (Lyman, 1874), relictus (Koehler, 1898) and
japonicus (Matsumoto, 1915).

H. L. Clark (1918) first referred japonicus to the genus Amphioplus, protesting
that it has no true articulated marginal papillae and four rather than three oral
papillae, unlike Ophiophragmus wurdemanni Lyman, the West Indian type-species of
Ophiophragmus, which up to that time had never been properly figured. However
Matsumoto (1941) persisted in including japonicus in Ophiophragmus, Koehler (1930)
and Murakami (1943 and 1944) following the same practice, though in 1963 Mura-
kami uses the generic name Amphioplus ; Matsumoto's grounds for this were that
japonicus is distinguished from Amphioplus by the absence of the first oral tentacle
scale. I have seen a number of specimens of japonicus but have not observed this
scale in any of them; however, I have found it, at least in rudimentary form, on some
oral plates of certain individuals of Amphioplus hastatus, A. laevis and particularly A .
integer, all of which but especially the first are closely related to japonicus. The
marginal " fence " on the disc of A. japonicus consists of the enlarged scales of the
outermost dorsal row bordered by the up-turned scales of the uppermost ventral
row which in young specimens Matsumoto found to be sufficiently prominent to form
a " hem-like row of denticles " so that the disc edge is serrated. Comparable
modifications of the marginal scales have also been recorded in A . furcatus (Morten-
sen, i-933a), in A. hastatus by Mortensen, 1940, in A. gibbosus (Ljungman, i867b) and
in A. affinis (Duncan, 1887) ; also in A. relictus (by Koehler in I922(a) though I
believe that he confused relictus and japonicus in this work) , while in A . spinulosus
(Koehler, 1904) (fig. gv) even greater modification occurs. Not all individuals of
japonicus show the marginal " fence " or flange and Mortensen (1940) notes that
marginal thorns may be present or absent in hastatus in the Persian Gulf, so their
occurrence is not diagnostic.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 51

In amphiurids such as Amphioplus with numerous oral papillae there is often
considerable variation in the arrangement of the papillae and their relationship with
the first oral tentacle scale. It should not be forgotten that the two distalmost oral
papillae of Amphioplus are homologous with the scales of the second oral tentacle, as
indeed Verrill (1899) termed than in his diagnosis of Amphioplus. It is then more
understandable that the first oral tentacle scale can be transformed into an oral
papilla by shifting to a superficial position, as already described in species such as A .
falcatus or alternatively more or less suppressed at the expense of the oral papillae as
in the group of species now under discussion.

In fact a straight row of four oral papillae with no visible oral tentacle scale is not
restricted to the species of the hastatus-group, being shared notably by Amphioplus
impressus (Ljungman, iSCyb) (fig. 9m), intermedius (Koehler, 1905) and aurensis A.
M. Clark, 1955. These also have contiguous radial shields and only three arm spines
but are distinguished from the hastatus-group by the smaller size of the two tentacle
scales and the enlarged fourth rather than third oral papilla.

In A. stenaspis H. L. Clark, 1938 (fig. 9], k) and A. iuxtus Murakami, 1943, also
the oral papillae are in a straight row and at the same time the tentacle scales are
very large, but both species have more than three arm spines proximally and their
radial shields are separated or only contiguous for less than half their lengths, also in
A. stenaspis at least the oral tentacle scale is present and visible above the oral
papillae.

Amphioplus causatus and exsecratus (Koehler, 1905), as well as A. cinctus (Koehler,
1914) and ailsaclarki Cherbonnier, 1957 (if the last two are distinct) resemble
hastatus in the oral structure even to the extent of having the third oral papilla the
largest ; they also have fairly large tentacle scales and, with the possible exception of
exsecratus, the radial shields largely contiguous (though for only about half their
length in causatus). However, causatus has four arm spines proximally and the
others often have five ; even so it might be better to ally cinctus (and ailsaclarki) with
the hastatus-group by setting aside the distinction of arm spine number ; although
most specimens studied appear to have the oral papillae in a concave row, this is
largely an illusion caused by the extension of the second oral tentacle and the erection
of the third and fourth papillae to the vertical position.

There are of course several other species of Amphioplus with only three arm spines
or with two more or less enlarged tentacle scales, sometimes both together.

Clearly none of these characters is restricted to the species of the hastatus-group,
although they combine to distinguish it. Accordingly the limits are not sharply
defined and I consider that the group merits only the rank of a subgenus.

AMPHIOPLUS subgenus LYMANELLA nov.

TYPE-SPECIES. Amphipholis hastata Ljungman, i867b. [See Koehler,
pi. 3, figs. 2 and 3 for photographs of the holotype.]

DIAGNOSIS. A subgenus of Amphioplus in which the four oral papillae are arranged
in a continuous row with their free edges (when aligned horizontally) forming a

52 AILSA M. CLARK

straight line, so that they can almost completely close the oral slit, the third papilla
being more or less distinctly enlarged so that the formula is m,mN,Nm +/ t ; l the
radial shields are contiguous for over half their lengths ; the two tentacle scales are
large or very large, more or less completely covering the corresponding pores and
the arm spines number not more than three. In some of the species the uppermost
ventral row of disc scales (which usually comes just above the margin in preserved
specimens) may project in a flange or else have small thorny prolongations of the
scales ; rarely (A . spinulosus) the thorns are larger, rather like those of the subgenus
Amphispina of Amphiodia but are still extensions of the scales and not articulated.
The following nominal species are referable to Lymanella 2 :

Amphipholis hastata Ljungman, i867b : 313. Mozambique.
Amphipholis depressa Ljungman, i867b : 312. Java-Singapore.
Amphipholis integra Ljungman, i867b : 313. Port Natal.
Ophiophragmus gibbosus Ljungman, i867b : 316. Port Natal.
Amphipholis andreae Ltitken, 1872 : 106. Java.
Amphiura laevis Lyman, 1874 : 229. Philippine Is.
Amphiura congensis Studer, 1882 : 19. Congo, West Africa.
Ophiophragmus affinis Duncan, 1887 : 89. Mergui Archipelago, Burma.
Amphiura relicta Koehler, 1898 : 69. Bay of Bengal.
Amphiura (Amphioplus) spinulosa Koehler, 1904 : 92. Lesser Sunda Is.
Amphiura (Amphioplus} praestans Koehler, 1905 : 52. Lesser Sunda Is.
Amphioplus megapomus H. L. Clark, 1911 : 170. Southern Japan.
Ophiophragmus japonicus Matsumoto, 1915 : 70, with var. parvus Matsumoto,

1941 : 334. Japan.

Amphioplus bocki Koehler, ig27b : 6. Fiji.
Amphioplus potens Koehler, 1930 : 107. Kei Is.
Amphioplus furcatus Mortensen, I933a : 370. Natal.
Amphioplus caelatus Ely, 1942 : 39. Hawaiian Is.
Amphioplus miyadii Murakami, 1943 : 227. Southern Japan.

Of these eighteen names, several have already been relegated to the synonymy of
others and I think that when a proper appreciation is reached of the extent of varia-
tion, especially of the disc in relation to growth and regeneration, the total number of
species recognized will be less than half that number.

Mortensen (1940) considers that A. furcatus is a synonym of A. hastatus, which he
found in abundance in the Persian Gulf, although the type-material of furcatus
appears to me to have relatively finer and more numerous disc scales than is usual in
hastatus of the same size.

At the same time Mortensen notes that Amphioplus bocki could be a synonym
of A. laevis. However, since Lyman published only very poor figures of the

1 In preserved specimens the third oral papilla particularly may be found projecting outwards so that
the small, V-shaped gap between its tip and that of the fourth papilla, through which the second oral
tentacle may be seen, is considerably widened. This is the case in the holotype of A. bocki Koehler,
igayb for instance.

2 With original combinations, in chronological order and with type-localities.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 53

holotype of A. laevis, this is uncertain. Unfortunately I omitted to study the
holotype when visiting the Museum of Comparative Zoology, Harvard, but Miss Amy
Schoener tells me that it is in poor condition with only a few damaged basal segments
of the arms left attached to the disc. Her drawing of an arm base (fig. gw) shows
rather irregular dorsal arm plates with some suggestion of a median angle in their
distal sides but with the broadest part of the plate towards the proximal end which is
unusual. The specimen figures by Mortensen (1940) under the name of laevis has
distinct median distal angles to the dorsal arm plates but the broadest part of each
plate is near the distal end. This is just the shape figured by Koehler from a syn-
type of Amphioplus praestans and found also in another syntype in the Harvard
Museum (fig. 95), although some of the smaller " Siboga " specimens in the Amster-
dam Museum have little sign of a median distal angle to these plates. In some ways
it might be best to reject the name laevis on the grounds of the almost unrecognizable
condition of the holotype but the name has been widely used in recent years for the
species figured by Mortensen. Koehler (1905) referred two small " Siboga " speci-
mens (d.d. c. 5 mm.) to laevis at the same time that he described praestans and these
two have particularly narrow radial shields, length : breadth = 3-5-3-75 : i (4 : i in
the holotype according to Lyman), while their oral shields are relatively broad,
though still longer than broad. Nevertheless I do not think that either is beyond the
range of variation of a single species, nor is the slight difference in the shape of the
dorsal arm plates compared with the holotype significant. There are specimens in
the British Museum collections from the Red Sea, New Guinea and the Solomon
Islands which agree with the type-material of praestans and at the same time with
Lyman's description of laevis in having the disc flat and usually thin and more or less
sharp at the edge, the scales and plates thin and the spines slender. The distinctive
light line along the arms mentioned by Lyman is an illusion caused by the thinness
and semitransparency of the arm plates, allowing the mid-line of the vertebrae to
show through. Unfortunately most of the specimens studied have the arms more or
less badly broken, except for some from the Solomon Islands in which the arm length
is about ten times the d.d. This compares with a proportion of 13-14 : i for the
type-material of A . bocki. However, I am inclined to think that Mortensen may be
right in referring bocki to the synonymy of laevis. Even so, it should be noted that
Koehler describes the arm spines of bocki as blunt-tipped whereas in all the specimens
I regard as referable to A . laevis the slender spines are more or less sharp.

Matsumoto (1941) thinks that Amphioplus megapomus is allied to his variety
parvus of japonicus but, as noted in 1965, I doubt this since the discless holotype of
megapomus (fig. gx) is another specimen with a median distal angle on each dorsal arm
plate, differing in this from japonicus. H. L. Clark in 1915 recorded four smaller
Japanese specimens as megapomus, noting that their radial shields are relatively short.
However, he also notes that their arm plates show some differences from those of the
types and it is possible that they were not conspecific in view of the number of other
nominal species of Lymanella recorded from Japan. If the identification was correct
and the radial shields are short in megapomus then it is very close to A . andreae, but
if long then the affinity is with laevis again, or with Amphioplus miyadii, if that is
itself distinct from laevis.

54 AILSA M. CLARK

H. L. Clark (1918) and Koehler (i922a) both note that Amphioplus affinis
(Duncan) is a homonym of A. affinis (Studer, 1885) (though Mortensen in 1936 and
Madsen in 1955 put affinis Studer in Amphiodia) and that Duncan's material was
probably conspecific with A . relictus. However, the Kagoshima specimens figured by
Koehler in ig22(a) (pi. 71, figs. 7 and 8) named by him Amphioplus relictus
appear to be conspecific with A . japonicus of which Kagoshima is the type-locality.
Koehler's earlier specimens from the Indian Ocean which he named relictus at least
may be distinguishable from japonicus by the longer distal lobes of their oral shields
(though the shape of these is variable in many of the species of Lymanella) but his
figures are so diagrammatic that this is not certain. In 1930 he recorded Ophio-
phragmus japonicus from the Gulf of Siam and Amboina, thus extending the range
well beyond Japan, though simultaneously he recorded other specimens from these
same areas as Amphioplus relictus. H. L. Clark (1946) refers relictus to the synonymy
of depressus, which seems to me well justified. I am even doubtful whether depressus
and japonicus can be distinguished from hastatus. All of them have relatively broad
radial shields, opaque scales and plates, no marked median distal angle to the dorsal
arm plates and often more or less specialized marginal scales (fig. gu). The distinc-
tion of Amphioplus hastatus from the others by its prominent primary rosette is
not reliable because of the likelihood of loss and regeneration, in which case the
rosette is probably not reformed. Mortensen (1940) records a number of discless
specimens and assumes that the loss was incurred during capture since there is no
sign of regeneration. However, I suggest that natural loss may occur during the
breeding season when the disc is liable to be distended but this remains to be investi-
gated. The considerable length of the distal lobe of the oral shields (fig. gp) is also
supposed to distinguish hastatus. Certainly all 21 Japanese specimens which I
referred to japonicus in 1965, have a very short distal lobe, more of a convex side,
however, a very similar form of shield occurs in a specimen of hastatus (fig. gq) from
the northern Arabian Sea (John Murray Expedition) as well as in Koehler's 1905
figure of hastatus. Unfortunately Koehler did not include Amphioplus depressus
among those of Ljungman's types of which he gave photographs in 1927 (a), though in
1930 he illustrates a specimen with d.d. 5 mm. which he says is very like the holo-
type. The scales of this specimen are thick and stand out in relief and the central
disc scales though large do not form a proper rosette ; it also has very long distal
lobes to the oral shields, although Ljungman described those of the type as having
" angulo externo parum producto ". It does not seem markedly different from
some of the specimens which I have seen attributed to hastatus.

Lyman (1882) referred two lots of " Challenger " specimens to A. depressus. The
single discless specimen from Japan is almost certainly conspecific with A . megapomus
since it has median distal angles to the dorsal arm plates. The other two are from
Fiji, one having the disc intact and measuring 9-5 mm. in diameter. The primary
rosette is just distinct though the radial plates are very little larger than the scales
around them ; on a line between the radial shields across each interradius there are
seven to nine scales. The radial shields are fairly wide and almost triangular in
shape with length : breadth 2-3 : i and length : disc radius i : 2-7. The uppermost
row of ventral disc scales form a projecting flange. The dorsal arm plates are

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 55

extremely wide (as might be expected at this large size) and the dorsal side of the
arms is very convex, though when viewed immediately from above the distal edges of
the plates appear straight. The oral shields are narrow spearhead-shaped with the
distal lobe only a little shorter than the proximal angle and not very constricted.
There is no sign of an oral tentacle scale in either specimen. Three other specimens in
the British Museum collections can also be referred to A . depressus ; these are from
Madras in 9 metres depth and all have d.d. c. 5 mm. The primary rosette can just
be distinguished in two of them and two have a slight marginal flange. The radial
shields are relatively larger than in the Fiji specimen (in conjunction with the
smaller total size), the ratio of length : disc radius being i : 2-5 in one specimen and
i : 2-2 in the other two. The dorsal arm plates are convex distally but this curve is
truncated medially to some extent.

The largest Murray specimen of A. hastatus has d.d. 5 mm. It also has relatively
large radial shields with length : disc radius 1:2-1 and length : breadth c. 1-8 : i.
The primary rosette is quite distinct, there are about seven scales between the radial
shields interradially and the disc has a sharp margin but no modification of the
scales into a fringe of spiniform processes as Mortensen describes for some specimens
from the Persian Gulf (1940). A rudimentary oral tentacle scale is present though
normally concealed behind the oral papillae and the oral shields have only a short
distal lobe. [A specimen from the Red Sea recorded as hastatus by Burfield (1924)
with d.d. only 3-2 mm. has relatively longer radial shields, length : breadth 2-5-2-8 :
i and the scales are much smaller with 11-13 m a l me between neighbouring radial
shields interradially. Also the dorsal arm plates have a slight median distal angle.
Accordingly I have re-identified Burfield's specimens as A. (Lymanella) laevis.]

Some of the other nominal species listed are also very closely related, for instance
Amphioplus integer and caelatus, and need further investigation of their ranges of
variation. Although it has often been noted that most of the species are liable to
lose their discs very easily, little account has been taken of the consequences of re-
generation on the characters shown by the disc, such as the distinctness of the primary
rosette, the proportions of the radial shields, the size of the scales and the modification,
if any, of the marginal scales. While regeneration is in process the newly formed
radial shields undergo a marked change, starting as short wide separate plates but
becoming longer and more contiguous as the disc grows to the original size. Morten-
sen (i933a) illustrates a regenerating specimen of A. integer showing an early stage in
the regrowth of the radial shields and Koehler's figure 3, pi. 18 (1930) of the type and
only specimen of Amphioplus potens, which is clearly in process of regeneration but
at a rather later stage, shows the shields in contact but barely longer than wide and
the whole disc still diminutive. I believe that potens will prove to be a synonym of
either A. andreae (if its radial shields were originally short with a length : breadth
ratio of c. 2 : i) or A. laevis (if the ratio was about 3:1). There is of course some
variation in the proportions of the fully-grown radial shields ; as noted in 1965, those
of 21 specimens of A. japonicus have the length : breadth ratio ranging from 1-75 to
2-55 : i, averaging 2-1 : i. In 15 specimens of A. integer (the only other sample of
any size available) the variation is greater, from 2-3 to 3-7 : i, averaging 2-8 : i.

5 6 AILSA M. CLARK

It remains to give a diagnosis of the genus Amphioplus as reconstituted, as well as
of the subgenus Amphioplus, excluding the species now referred to Unioplus,
Amphichilus and Lymanella ; also to construct a diagram illustrating the possible
relationships between these taxa and some other amphiurids.

Genus AMPHIOPLUS Verrill

DIAGNOSIS. A genus of Amphiuridae with three or, more usually, four oral papillae
forming a fairly continuous series on each half of the jaw, the consecutive ones either
contiguous or else separated by relatively small spaces much less conspicuous than
the large diastema of Amphiura which is often equal in length to the entire oral plate ;
a first oral tentacle scale usually present on the side of the oral plate, sometimes
suppressed (in which case the oral papillae do not number less than four, unlike
Amphiodia), or displaced so as to come more or less into series with the superficial
oral papillae (which then often number only three) ; disc scaling usually complete but
sometimes reduced on the lower side leaving more or less extensive areas of naked
skin, the scales not armed with spinelets though some of the marginal scales may have
inconspicuous prolongations ; the two radial shields of each pair variously related,
most often contiguous only by their distal adradial corners but sometimes almost
completely contiguous or conversely completely separate ; tentacle scales usually
numbering one or two, sometimes suppressed beyond the basal segments which have
a single scale.

AMPHIOPLUS subgenus AMPHIOPLUS Verrill

TYPE-SPECIES. Amphiura tumida Lyman, 1878.

DIAGNOSIS. A subgenus of Amphioplus with the oral papillae numbering four and
arranged in a concave, sometimes discontinuous, series not capable of closing the
oral slit completely ; a first oral tentacle scale present, usually visible between and
behind the first and second oral papillae so that the oral formula usually approxi-
mates to m,mn^nm + t ; the radial shields usually only contiguous distally, sometimes
more fully contiguous or completely separate ; one or two tentacle scales.

The following nominal species are referable to the subgenus Amphioplus :

abditus (Verrill, 1871)
aciculatus Mortensen, 1936
acutus Mortensen, 1936
ancistrotus (H. L. Clark, 1911)
asterictus H. L. Clark, 1915
basilicus (Koehler, 1907)
conductus Koehler, 1922
coniortodes H. L. Clark, 1918
cuneatus (Lyman, 1878)
cyrtacanthus H. L. Clark, 1915
debilis (Koehler, 1904)

didymus H. L. Clark, 1938
echinulatus Mortensen, 1940
exsecratus (Koehler, 1905)
firmus (Koehler, 1904)
hexacanthus H. L. Clark, 1911
legatus Koehler, 1922
lobatus (Ljungman, 1867)
lucidus Koehler, 1922
macraspis (H. L. Clark, 1911)
magellanicus (Mortensen, 1936)
pectinatus Mortensen, 1933

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA)

57

personatus (Koehler, 1905)
platy acanthus Murakami, 1943
rhadinobrachius H. L. Clark, 1911
seminudus Mortensen, 1940
stenaspis H. L. Clark, 1938
stewartensis (Mortensen, 1924)

strongyloplax (H. L. Clark, 1911)
stratus Cherbonnier, 1963
tesselatus (Koehler, 1904)
thrombodes H. L. Clark, 1918
timsae Mortensen, 1926
tumidus (Lyman, 1878) Type-species

In addition there are a number of species incertae sedis, either because of in-
sufficiently precise description of their jaw armament, or else showing some characters
more or less intermediate between the subgenera as defined above. These are as
follows :

ailsaclarki Cherbonnier, 1957
archeri A. M. Clark, 1955
aurensis A. M. Clark, 1955
brachiostictus Tortonese, 1948
capax (Koehler, 1905)
causatus (Koehler, 1905)
cinctus (Koehler, 1914)
consors (Koehler, 1907)
difficilis (Duncan, 1887)
dispar (Koehler, 1897)
famula (Koehler, 1910)
glaucus (Lyman, 1879)
impressus (Ljungman, 1867)

intermedium (Koehler, 1905)
iuxtus Murakami, 1943
longirima Fell, 1952
magnificus (Koehler, 1907)
occidentalis Koehler, 1914
parviclypeus H. L. Clark, 1915
peregrinator (Koehler, 1912)
philohelminthius Ziesenhenne, 1940
psilochora (H. L. Clark, 1911)
refectus (Koehler, 1905)
signalis Koehler, 1930
textilis (Koehler, 1907)

Of these last nominal species, A. archeri at least may warrant distinction in a
special taxon, either a subgenus, or possibly even a genus, since it diverges from
Amphioplus sensu stricto in the direction of Ophiophragmus, having some develop-
ment of marginal papillae, though the true extent of these remains to be seen from
better samples with original as well as regenerating discs. Only the presence of the
fourth oral papilla, which in this case arises from the side of the first ventral arm
plate, distinguishes it from Ophiophragmus, of which the oral structure is of the
Amphiodia (B)-type. Also in A. archeri the oral tentacle scale, or at least an addi-
tional papilla probably corresponding to this, is only rarely present (fig. li), on one
out of ten oral plates of the specimen figured.

Amphioplus psilochora is newly referred to the genus from Amphiodia and seems to
occupy an isolated position. Its oral formula is m,mm,o +t, there being no fourth
papilla at the distal end of the series, though the presence of the oral tentacle scale
prevents its inclusion in Amphiodia ; in Amphichilus and Unioplus, also with only
three papillae, it is the second one which is usually undeveloped. The affinities of
psilochora may lie with A . strongyloplax, in which the fourth papilla can be reduced or
lacking in some series or even entire specimens.

In addition to the species listed under the headings of the various subgenera of
Amphioplus (pp. 48, 49, 52, 56-57) and those given as incertae sedis above, there are

58 AILSA M. CLARK

several other names which have been included in the genus which are now disposed of
as follows :

agassizi Verrill, iSggb, referable to Ophiochitonidae (fig. nb).

cyihera A. H. Clark, 1949, referable to Amphiactis (fig. ud,e, see p. 74).

diacritus Murakami, 1943, ? a synonym of A. ancistrotus (see A.M.C., 1965).

formatus (Koehler, 1905), not an amphiurid (fig. na, see pp. 61-62).

lorioli (Koehler, 1897), referable to Amphiura, with synonym A mp h iuraceramis

H. L. Clark, 1939 (see pp. 22-23).

luctator Koehler, 1922, ? a synonym of Dougaloplus echinatus (see p. 36).
modestus (Koehler, 1897), incomprehensibly referred from Ophiochiton (with

which its affinities clearly lie) by H. L. Clark (1915).
nereis (Lyman, 1883), referable to Ophiochitonidae (fig. nc).
papittatus (Ltitken & Mortensen, 1899), referable to Ophiochitonidae.
trepidus (Koehler, 1904), referable to Ophiocomidae (see p. 37).

The presence of genital papillae in A . agassizi, nereis and papillata prohibits their
inclusion in the Amphiuridae ; the narrow teeth with no infradental papillae exclude
A . cyihera and formatus ; the diminutive widely-separated radial shields distinguish
A. modestus and trepidus, while the latter also has a group of apical tooth papillae
rather than an infradental pair of oral papillae.

My studies to date lead me to the conclusion that Amphioplus occupies a central
position among amphiurids, having affinities with other taxa as indicated in the
diagram on p. 60.

Some notes on type-material in the Amsterdam Museum of some relevant " Si-
boga " species of Koehler follow, arranged in alphabetical order of species.

Amphioplus (Unioplus) conditus (Koehler)

Amphiura (Amphioplus) condita Koehler, 1905 : 46-47, pi. 5, figs. 9, 10.
Amphioplus conditus : H. L. Clark, 1915 : 257.

LOCALITY. " Siboga " st. 89 ; Pulu Kaniungan, East Indies, n metres ; Amster-
dam Museum, the holotype.

DESCRIPTIVE NOTES. D.d. is 4 mm. The radial shields are c. 0-9x0-3 mm. =
3:1. The dorsal arm plates are more fan-shaped than shown in Koehler's figure and
barely contiguous, if at all. There are only three oral papillae, the supposed second
papilla being distinctly higher than the two distal papillae, the inner one of which is
the largest ; the formula is m,om,m +t.

Amphioplus (Unioplus) confinis (Koehler)

Amphiura (Amphioplus} confinis Koehler, 1904 : 89, pi. 14, figs. 7, 8.
Amphioplus confinis : H. L. Clark, 1915 : 257.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 59

LOCALITY. " Siboga " st. 211, S of Celebes, 1158 metres ; Amsterdam Museum,
four syntypes. St. 221, Flores Sea, 2798 metres ; two specimens. (The latter were
not included in the " Siboga " report.)

DESCRIPTIVE NOTES. D.d. is 3 - 5-4'0 mm. The rosette is conspicuous ; the radial
shields are long, divergent and completely separated, even distally. The disc scales
are so thick as to produce an uneven profile. The dorsal arm plates are fan-shaped
and not contiguous. The arm spines are conical and very acute, evenly tapering.
The oral structure of the four syntypes shows some slight differences. One specimen
has two jaws with the innermost oral papillae not infradental but offset to the sides
of the lowest tooth, a third jaw has a close infradental pair, a fourth has three
asymmetrical apical papillae and the fifth has three symmetrical papillae ; the first
oral tentacle scale is superficial and in series with the oral papillae so that the formula
is m,(m = t)mffji ; the adoral shields are widely contiguous and the oral shields are
triangular with a slightly convex distal edge but the distal part distinctly hollowed.
A second specimen has a close infradental pair of papillae on each jaw and the lowest
tooth above may have a conical apex ; the oral tentacle scale is superficial for its
whole length, even its base not being overlapped by the inner of the two distal
papillae, while the outermost oral papilla barely makes contact with the adoral shield
so that the formula is m,(m = t)mm, ; the triangular oral shields have the distal
side more convex but are less hollowed than in the first specimen. The third
specimen has the infradental papillae more or less spaced on three jaws with a
pointed tooth between, though the other teeth are broad as usual ; the oral tentacle
scale is in continuous series with the oral papillae ; the inner of the two distal papillae
is the largest one and the outer is small and seems to arise from the side of the first
ventral arm plate rather than the edge of the adoral or the oral plate so that the
formula is best expressed as m,(m = t)m,,m. The fourth syntype (in fact only three
specimens were mentioned in the report but there is no way of telling whether one
was added) is very like the second one with the infradental papillae regularly paired
but the oral tentacle scale is less superficial and is overlapped distally by the middle
oral papilla ; the oral shields are very wide and short and triangular in shape, as well
as markedly hollowed. The smaller of the two specimens from station 221 has three
infradental papillae on one jaw ; the larger specimen is poorly preserved ventrally.

Amphioplus (Amphioplus) exsecratus (Koehler)

Amphiura (Amphioplus) exsecrata Koehler, 1905 : 48-49, pi. 3, figs. 7, 8.
Amphioplus exsecratus : H. L. Clark, 1915 : 254.

LOCALITY. " Siboga " ; Amboina reefs ; Amsterdam Museum, the holotype.

DESCRIPTIVE NOTES. The oral structure is very like that of Amphioplus (Lyman-
ella] hastatus with the four papillae in a straight series closing the oral slit and the
third one the largest and blunt-conical in shape. However, the oral tentacle scale
is present although concealed behind the papillae so that the formula is m.mnjmi -ft.

6o

AILSA M. CLARK

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 61

The oral shields are shaped like the sole of a shoe, being elongated with a wide distal
" heel " and a very short and obtuse proximal angle or " toe ". The two tentacle
scales are very large, especially the ventral one. There are five arm spines proximally,
the number falling to three. The upper surface is somewhat contorted ; the radial
shields are probably narrow.

Amphioplus (Amphioplus) fir mus (Koehler)

Amphiura (Amphioplus) firma Koehler, 1904 : 87-88, pi. 12, figs. 7, 8.
Amphioplus firmus : H. L. Clark, 1915 : 257.

LOCALITY. " Siboga " st. 159, Molucca Is., 411 metres ; Amsterdam Museum,
two syntypes.

DESCRIPTIVE NOTES. D.d. is 4-5 mm. Both specimens have a very large central
disc plate but no distinct primary radial scales though one appears to have distinctly
enlarged primary interradials. In one the radial shields measure 1-2 : 0-4-0-5 mm. but
in the other they are relatively a little broader ; the shields are completely separate,
the scales between them not tapering outwardly ; interradially there are three to five
scales between adjacent radial shields ; the dorsal arm plates are elliptical but slightly
angled laterally ; two jaws have the pointed infradental papillae slightly separated
and the tooth between them with a cusp ; the three distal oral papillae each side all
arise from the side of the oral plate or else the narrow fourth one may abut on the
first ventral arm plate so that the formula is m,(o)mmm(o),o +t or m,(o)mm,,m ;
the first ventral arm plate is large and its proximal angle is quite superficial ; the
tip of the first oral tentacle scale shows beyond the tip of the second oral papilla ;
the third oral papilla is the largest. Both specimens have dark brown areas on the
ventral side of the disc as if the scales are transparent.

" Amphioplus " formatus (Koehler)
fig. na

Amphiura (Amphioplus) formula Koehler, 1905 : 51-52, pi. 5, figs. 11-13, pi. 16, fig. 6.
Amphioplus formatus : H. L. Clark, 1915 : 257.

LOCALITY. " Siboga " st. 310 or 312, N of Sumbawa, 73 or 274 metres ; Amster-
dam Museum, the holotype.

DESCRIPTIVE NOTES. D.d. is 12-5 mm. (Koehler estimated it as 12 mm.). The
disc scaling is very fine, the rosette distinct but its plates very widely separated as in
Koehler's figure ; there are some enlarged scales proximal to the radial shields, not
shown by Koehler. The radial shields are contiguous for only about the distal fifth
of their length ; their proportions are c. 2-2 : 0-8 mm. = 2-75 : i ; the proximal arm
spines are 1-0-1-15 mm. long compared with an arm breadth of 1-07 mm. No oral
tentacle scale could be found but behind the oral papillae there is a horizontal flange
on the oral plate.

62 AILSA M. CLARK

The relatively short arms with length only about three times the d.d., the large
size, the narrow teeth with the two innermost oral papillae offset, the diminutive
oral shields and the relatively long arm spines leave little doubt that this specimen
is not an amphiurid at all ; however, at present I am not prepared to hazard a guess
as to its true taxonomic position.

Amphioplus (Unioplus) gentilis (Koehler)
figs. 8r, ng

Amphiura (Amphioplus) gentilis Koehler : 1904 : 86-87, pi- I 6, figs. 8, 9.
Amphioplus gentilis : H. L. Clark, 1915 : 254.

LOCALITY. " Siboga " st. 178, Molucca Is., 835 metres; Amsterdam Museum,
several syntypes.

DESCRIPTIVE NOTES. All the specimens are about the same size with d.d. 2-3 mm.
The scales and plates are notably thickened and shiny, standing out in relief. In
one specimen the radial shields are separated as shown in Koehler's figure but the
six plates of the primary rosette may be closer than drawn, with only interstitial
scales between their angles. The dorsal arm plates are all separate. The lowest
tooth of this specimen has an abrupt median cusp on some jaws with the two inner-
most oral papillae offset rather than infradental ; their shape is conical. The two
distal oral papillae each side are both partly or completely based on the edge of the
adoral shield so that the formula is m,o,mm -f-t or m,onjim -f-t. The adoral shields
are separated interradially in this specimen but in another they are broadly contigu-
ous and the oral shields are much shorter. The latter also has the radial shields
contiguous at their distal tips, the disc scales are flatter, also the outermost oral
papilla is transparent or absent. Another has the oral shields intermediate between
the two extremes. The infradental papillae may be almost contiguous and more
rectangular in shape in some specimens, the first oral tentacle scale may be deeper
in the oral slit and more rounded at the tip, the primary disc scales may be relatively
larger and the radial shields contiguous distally for up to a quarter of their length.
It is possible that not all these specimens are conspecific. Indeed a syntype in the
collections of the Harvard Museum, no. M.C.Z. 3501 (fig. nf) has been labelled
(probably by H. L. Clark) " Amphiactis prob. canescens " and I agree that it is an
Amphiactis, the teeth are acute.

Amphioplus (Unioplus) gratus (Koehler)
fig. 8k

Amphiura (Amphiodia) grata Koehler, 1904 : 85-86, pi. 18, fig. 6, pi. 19, fig. 7.
A mphiodia grata : H. L. Clark, 1915 : 249.

LOCALITY. " Siboga " st. 211, S of Celebes, 1158 metres ; Amsterdam Museum,
two syntypes.

DESCRIPTIVE NOTES. The larger one has d.d. 10 mm. (Koehler gives 9 mm.).

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 63

The radial shields are 2-0 x i-o mm. and rather more angular than shown in Koehler's
figure. There is a large central disc plate but the five widely spaced primary radials
are barely larger than the other scales. The oral shields are rounded spearhead-
shaped and hollowed in the middle ; the adorals are very convex. There are three
oral papillae with which the oral tentacle scale is more or less in series, being super-
ficial, so that the formula is m,(m = t)m,m. The outermost oral papilla is the
widest and rectangular in shape. The dorsal arm plates have a slight median distal
angle. The smaller syntype has d.d. 8 mm. It has the oral tentacle scale superficial
on only three half -jaws so that the usual formula is m,om,m -j-t.

Amphioplus impressus (Ljungman)
fig. 9!, m

Amphipholis impressa Ljungman, i86yb : 314.

Amphiura (Amphioplus) cesarea Koehler, 1905 : 44-45, pi. 5, figs, i, 2.

Amphioplus cesareus : H. L. Clark, 1915 : 257.

Amphioplus impressus : H. L. Clark, 1915 : 258 ; Koehler, 1922 : 174-176, pi. 64, figs. 3, 4 ;

i927a : 16-19, pi. 2, fig. 10, pi. 3, fig. i.
Amphichilus cesareus : Koehler, 1930 : 119.

LOCALITIES. " Siboga " st. 294 ; off Timor, 73 metres ; Amsterdam Museum,
two syntypes of Amphiura cesarea. Mortensen, 1922 ; Kei Is., 20 metres ; B.M.
1949.8.15.16, one specimen. " Albatross " st. 5371 ; Philippine Is., 152 metres ;
U.S.N.M. 41171, one specimen.

DESCRIPTIVE NOTES. The larger of the syntypes of Amphiura cesarea has d.d.
6-5-7-5 mm - the disc being distorted ; its radial shields are c. 1-15 x 0-5 mm. =
2-3:1 and contiguous for about three-quarters of their length. The smaller speci-
men has d.d. 5-5 mm. and its radial shields are also 1-15 mm. long but slightly
narrower with the ratio 2-5:1. The rosette is distinct in both, with one row of
scales between the primary plates. The disc scales are all thin and flat so that the
disc appears fairly smooth ; there are six or seven scales between the radial shields
across the interradii in the larger specimen but only three to five in the smaller. The
dorsal arm plates of the smaller specimen particularly show the median distal angle
which is rather exaggerated in Koehler's figure. In the larger one the oral slits are
completely closed by the straight rows of four papillae but the second papilla is
always small and conical and more often than not lies slightly above the level of the
other papillae and clearly represents the oral tentacle scale ; the formula is therefore
m,(m = t)mN,N, the distalmost papilla being very broad and almost opercular, with
only its distal end contacting the adoral shield. Koehler's figure shows the oral
shields as too pear-shaped ; they are relatively small and rhombic with the distal
lobe only a little more blunt than the proximal. The arm spines are unusually
diminutive on the first three segments. The tentacle scales are of moderate to small
size. In the smaller specimen the second oral papilla is almost perfectly in series
with the other three.

REMARKS. The Kei Islands specimen was identified as Amphichilus cesareus by

64 AILSA M. CLARK

Koehler (1930), whereas the " Albatross " specimens in the Smithsonian Institution
he determined as Amphioplus impressus (1922) ; in the meantime (i927a) he re-
described and figured the two syntypes of A. impressus. Comparison of all these
reveals no significant difference between the two nominal species, as far as I can see.
The disc scaling appears to stand out a little more in relief in Ljungman's types than
in those of cesarea but this could easily be due to the shading effect of angled lighting
for the photographs. The border of narrow rectangular scales around the disc
shown in Koehler's pi. 3, fig. i (i_927a) of one syntype of impressus and in his drawing
of one of cesarea (1905, pi. 5, fig. i) is an unusual and distinctive feature. Another
significant resemblance is the form of the first ventral arm plate ; in nearly all
amphiurids this inclines more or less gradually up into the oral slit, especially at its
proximal end. In the types of both impressus and cesareus and the other specimens
seen by me its superficial part is horizontal, ending proximally in an abrupt angle
from which it passes vertically into the oral slit. I am convinced therefore that only
a single species exists and that Amphiura cesarea is a synonym of Amphioplus im-
pressus.

As for the precise position of the species within Amphioplus, without a larger sample
I find this hard to determine. Apart from the one syntype of A. cesarea none of the
specimens I have seen have the second oral papilla inset and it is difficult to ally
them with Amphichilus trichoides, although the second papilla must represent the
first oral tentacle scale. For the present at least I propose to leave the species in
Amphioplus sensu stricto.

Amphioplus (? Amphioplus) inter medius (Koehler)

Amphiura (Amphioplus) intermedia Koehler, 1905 : 42, pi. 4, figs. 14, 15.
Amphioplus intermedius : H. L. Clark, 1915 : 258.

LOCALITY. " Siboga " st. 77, Borneo Bank, 59 metres ; Amsterdam Museum,
the holotype.

DESCRIPTIVE NOTES. D.d. is 4 mm. The radial shields are much more fully
contiguous than is suggested by Koehler's figure, only the proximal tips being
separated by a narrow V-shaped scale ; their measurements are i-o X 0-25 mm. The
fourth and outermost oral papilla is twice as wide as the third, while the second
papilla is conical but perfectly in series with the other three. Together they form a
straight series completely closing the oral slit so that it is impossible to see if there
is an additional oral tentacle scale present above the series.

Amphioplus (Amphioplus) per sonatus (Koehler)

Amphiura (Amphioplus) personata Koehler, 1905 : 47-48, pi. 5, figs. 7, 8.
Amphioplus personatus : H. L. Clark, 1915 : 258.

LOCALITY. " Siboga " st. 50, E of Flores, 40 metres ; Amsterdam Museum,
the holotype.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 65

DESCRIPTIVE NOTES. D.d. is 6 mm. The radial shields are 1-6 x 0-4 mm. The
distal lobe of the oral shields is wider than is shown in Koehler's figure. The infra-
dental papillae are huge, while the fourth and outermost papilla is very small and
inconspicuous, apparently partly joined to the first ventral arm plate. The oral
formula can be expressed as M,omn7hivn -f-t, since the inconspicuous oral tentacle
scale is placed well above the first and second papillae between which there is a space.
The adoral shields almost meet radially as well as meeting interradially.

Amphioplus (? Amphioplus) refectus (Koehler)

Amphiura (Amphioplus) refecta Koehler, 1905 : 52-53, pi. 6, figs. 3, 4.
Amphioplus refectus : H. L. Clark, 1915 : 258.

LOCALITY. " Siboga " st. 193, Molucca Is., 22 metres ; Amsterdam Museum,
the holotype.

DESCRIPTIVE NOTES. The holotype is somewhat decalcified. The edge of the
disc is armed with a series of acute or slightly thorny spinelets. The radial shields
are divergent for the proximal half to two-thirds of their length. The condition of
the specimen makes it difficult to tell if the oral tentacle scale is present ; the oral
formula is otherwise m,mnjim (or m,mm,m). The tentacle scales are moderate in
size.

Amphioplus (Unioplus) repositus (Koehler)
fig. 8m

Amphiura (Amphiodia) reposita Koehler, 1905 : 40-41, pi. 4, figs. 12, 13.
Amphiodia reposita : H. L. Clark, 1915 : 250.

LOCALITY. " Siboga " st. 89, n metres ; Amsterdam Museum, three syntypes.

DESCRIPTIVE NOTES. All three have d.d. c. 6-5 mm. In one specimen the second
oral papilla is the largest and may have a double peak ; it arises mainly from the side
of the oral plate so that the formula is m,on7nm +t (fig. 8m). However, in another
specimen the second papilla arises from the edge of the adoral shield so that the
formula is m,o,mm +t, more like the condition shown in Koehler's figure. The radial
shields are almost contiguous at their distal ends ; their adradial sides are straight
and each measures c. 1-3 x 0-43 mm. The ventral disc scaling is very obscure.

Amphioplus (Unioplus) servatus (Koehler)
fig. 8p

Amphiura (Amphiodia) servata Koehler, 1904 : 84-85, pi. 15, figs. 6-8.
Amphiodia servata : H. L. Clark, 1915 : 250.

LOCALITY. " Siboga " st. 262, W of Aru Is., 560 metres ; Amsterdam Museum,
the holotype.

66 AILSA M. CLARK

DESCRIPTIVE NOTES. D.d. is 5 mm. The radial shields are c. 1-25 x 0-43 mm.
and are neither so long nor so much separated as Koehler's figure shows. Nine of
the series of oral papillae consist of three, with a formula of m,om,m -f-t but on the
tenth there is an additional papilla in the space between the first two and the formula
is m,mm,m +t. The first ventral arm plate has a markedly concave inner side.

Amphioplus (Lymanella) spinulosus (Koehler)
fig. QV

Amphiura (Amphioplus) spinulosus Koehler, 1904 : 92-93, pi. 17, figs. 7, 8.
Amphioplus spinulosus : H. L. Clark, 1915 : 258.

LOCALITY. " Siboga " st. 306, off Flores, 247 metres ; Amsterdam Museum, the
holotype. St. 319 (omitted from report but added MS in Amsterdam Museum
copy) ; Amsterdam Museum, one paratype.

DESCRIPTIVE NOTES. The holotype has d.d. 5 mm. The primary rosette is
distinct and there is a V-shaped scale between the proximal ends of each pair of
radial shields. The scales of the uppermost ventral row, which lie around the dorsal
disc margin as preserved, have their edges prolonged into usually single spine-like
processes. The paratype has d.d. 4 mm. The radial shields are oval, 0-8 x 0-46
mm., and hardly separated even at their proximal ends. There are six or seven
scales between the adjacent radial shields interradially. The primary rosette is not
distinct. The series of marginal spinose processes (fig. gv) is better preserved than in
the holotype and is situated towards the ventral side of the margin. Some of the
processes have a forked tip. The arms are distinctly rounded carinate above and
the dorsal arm plates have the distal edge curved with a suggestion of a median
angle. The arm spines are very acute. The tentacle scales are fairly large. The
series of four oral papillae forms a straight row ; the third papilla is massive, trun-
cated conical, while the others are rounded. They completely close the oral slits so
that the occurrence of the oral tentacle scale cannot be ascertained. The oral
shields are rather small and sunken in the middle, with a long distal lobe ; the adorals
are very widely contiguous interradially.

Amphioplus capax (Koehler, 1905) and debilis (1904)

Unfortunately the type-material of both species is in poor condition orally, the
holotype of capax being somewhat decalcified while the syntypes of debilis are either
obscured with mud or else not conspecific with the specimen described and figured,
having contiguous radial shields or only three oral papillae.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 67

TRIODIA gen. nov.

TYPE-SPECIES. Triodia abdita 1 nov.

DIAGNOSIS. A genus of Amphiuridae with three oral papillae, the second one
running the length of the short oral plate and the third bordering the adoral shield ;
with no oral tentacle scale developed, the oral formula is m,m,m t ; a triangular
supplementary adoral shield (possibly to be interpreted as one of a pair of supple-
mentary oral shields) is present between each adoral and the genital slit ; the disc
scales are unarmed ; the dorsal arm plates are not contiguous, even basally ; the
ventral arm plates are somewhat swollen ; single crescentiform tentacle scales are
present on a varying number of proximal arm segments, each scale bordering the
lateral edge of the ventral arm plate and curving round on to the proximal part of the
following lateral arm plate, the remaining tentacle pores lack scales.

Triodia abdita sp. nov.
fig. 10

Amphiura dalea (part) : Liitken & Mortensen, 1899 : 154-155, pi. 12, fig. n only. [Non A.
dalea Lyman, 1879 : 27.]

LOCALITY. " Albatross " st. 3361 ; 6io'N : 836'W (Gulf of Panama), 2790
metres ; Museum of Comparative Zoology, Harvard, no. 1488, one specimen.

DESCRIPTION. D.d. is 14 mm. ; all the arms are broken within three joints of the
base but there are some associated arm fragments, at least one of which belongs to
this species.

The disc is covered with fine, well-defined scales, slightly larger and smoother
around the radial shields and on the lower side towards the oral shields. The six
plates of the primary rosette are enlarged but widely separated. In preservation
the disc has tended to shrink into a crease in each interradius. In length the radial
shields are just over a third of the disc radius, with length : breadth just under 3:1.
The two shields of each pair are contiguous for about a quarter of their length distally,
otherwise being separated by about four enlarged scales.

The oral papillae number three in each series. The infradental pairs of papillae
are compact ; on one jaw three papillae take the place of two. The two other
papillae each side also form contiguous series. The oral plates are unusually short
and the second papilla is broad and occupies the whole length of the superficial edge
of the oral plate. In shape it has some resemblance to an oral tentacle scale but it
is completely superficial and perfectly in series with the two other papillae. The
third and outermost papilla is equally broad but is based entirely on the edge of the

1 The generic name is derived from the triangular tentacle scales, the series of three oral papillae and
the triads of plates formed by the oral shields and adjoining supplementary adoral shields. The specific
name signifies " hidden ", the holotype having for so long been submerged under another name in the
collections of the Harvard museum. Gender: feminine.

68

AILSA M. CLARK

adoral shield ; in one out of the ten series, this papilla is split into two. The second
oral tentacle is visible behind the third papilla but there is no sign of a first oral
tentacle scale on the vertical face of the oral plate. The oral formula can therefore
be expressed as m,m,m t. The oral shields are almost heptagonal, the distal part
having three straight sides separated by distinct angles, while the four more proximal
sides tend to merge into one another ; the length is greater than the breadth. The
adoral shields just meet interradially ; distally they are very broad, separating the
oral shield widely from the first lateral arm plate, also they are divided by an oblique
suture forming a supplementary adoral shield adjacent to the genital slit.

The dorsal arm plates are very broad and short, the basal ones more than 2-5 times
as broad as long and elliptical in shape ; those on a detached arm fragment have
become more pentagonal with the distal side flattened. All the plates are distinctly
separated, the lateral arm plates meeting midradially for over a quarter of the seg-
ment length proximally. The ventral arm plates are relatively narrow, mostly

FIG. 10. Triodia abdita gen. & sp. nov., holotype, M.C.Z. 1488, " Albatross " st. 3361,
Gulf of Panama, a. dorsal and b. ventral views of part of disc and an arm base.

slightly longer than broad. Their shape changes from rectangular on the first two
segments to hexagonal by tapering of the proximal and distal ends. The large
lateral arm plates bear eight or nine short tapering spines on the first free segments ;
judging from the arm fragments the number rapidly falls to three. Even the longest
spines are distinctly shorter than the segment. The middle spines are somewhat
flattened.

On about the first five segments a crescent-shaped tentacle scale corresponds to
each pore, running along the side edge of the ventral arm plate and curving around
distally on the proximal side of the following lateral arm plate. The extent of the
segments with scaled pores is clearly variable since one arm fragment has them at
its proximal end although the number of arm spines has dropped to as little as four.
On the other arm segments the large pores are naked.

REMARKS. This specimen was labelled Amphioplus dalea when I saw it in 1953,
with a double query of the name added by H. L. Clark. Although in many ways it
is similar to the other specimens which Liitken and Mortensen named Amphiura
dalea, notably in the disc scaling, radial shields, the absence of tentacle scales from

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 69

most arm pores (though this seems not unusual in deep-water amphiurids) and the
non-contiguous dorsal arm plates (again a feature of many deep-water species), the
coincidence of the modified oral structure, including the papillae and the adoral
shields, and the unusual shape of the tentacle scales make it necessary to distinguish
the specimen.

Of the other " Albatross " specimens, most have the oral formula approximating
to m,(m = t)m,m. Three of the four from station 4670 have no tentacle scales at
all and the fourth only has three or four proximal segments with single shorter scales
restricted to the sides of the ventral arm plates. Those from station 4493 have more
triangular oral shields, much wider distally ; also they have only three or four arm
spines, even at d.d. c. 14 mm. Their identification is probably correct.

[AMPHILEPIDIDAE]

Amphilepididae Matsumoto, 1915 : 66 ; 1917 : 144-146 ; Fell, 1960 : 23 ; Spencer & Wright,

1966 : Uioo.

Amphilepididae : Koehler, 1922 : 202 ; Mortensen, 1927 : 222 ; 19330 : 372.
Amphilepide'es : Koehler, 1930 : 132.

When established by Matsumoto this family included Amphilepis Ljungman,
i867b with the second oral tentacle pore opening outside the mouth, also Amphiactis
Matsumoto, 1915 (non Verrill, 1867) and Ophiochytra Lyman, 1880 with this pore
concealed inside the slit. 1 Most of Matsumoto's diagnosis appears to have been
derived from Amphiactis umbonata, probably the only species personally studied by
him. The main point of resemblance between Amphilepis and Amphiactis appears
to be in the oral structure as seen in internal (dorsal) view, shown by a comparison
of fig. ic or Lyman's pi. XL, fig. ig 2 and Matsumoto's pi. Ill, fig. 7 ; these show that
both have unusually reduced lateral wings to the oral frames. Regardless of this,
H. L. Clark (1918), Mortensen (1927) and A. H. Clark, (1949) remove Amphiactis
from the Amphilepididae and ally it with Ophiactis, while Fell (1960), followed by
Spencer & Wright (1966), even adopt H. L. Clark's short-lived idea of Amphiactis
as a synonym of Ophiactis, expressed in 1915 before due consideration. Only
Koehler (1922 and 1930) seems to have retained the genus in the family Amphi-
lepididae, mistakenly in my view since the arm structure particularly is very similar
to that of Ophiactis, though I am sure the differing jaw structure warrants a generic
distinction.

The removal of Amphiactis drastically undermines the validity of the Amphi-
lepididae by implying that the reduction of the lateral wings of the oral frame is a

1 In fact Matsumoto was mistaken about this character as far as the type-species of Ophiochytra,
0. epigrus Lyman, 1880, is concerned, the second oral pore being superficial in the holotype and only
recorded specimen. Ophiochytra is not closely related to either Amphilepis or Amphiactis and I do not
know why Lyman in his " Challenger " report placed it between Ophiostigma and Ophiocentrus (both
Amphiurids according to Ljungman (iSGyb) and subsequent authors). With its short arms covered
largely by the lateral arm plates, few peglike arm spines, very firm convex disc and strong oral skeleton
and diminutive radial shields, I consider that it is a member of the Ophiurinae.

2 It should be noted that Mortensen (iQ33b) has re-identified Lyman's " Challenger " specimens as
Amphilepis ingolfiana, though this is very closely related to the type-species, A. norvegica.

yo AILSA M. CLARK

character of lesser taxonomic weight, probably of generic value only, though Matsu-
moto's figures show marked variation from one species to another even within the
same genus.

The affinities of AmphiUpis are clearly with the Amphiuridae, as mentioned above
(p. 37) under the heading of Amphioplus. It is quite easy to derive the jaw form
of AmphiUpis norvegica from that of Amphioplus via Ophiomonas (formerly Amphi-
Upis ) protecta and AmphiUpis patens by progressive flattening of the jaws, bringing
the second oral tentacle pore into a superficial position while the apex of the jaw and
the teeth themselves become narrower, the two infradental papillae becoming
slightly offset laterally with their bases inclined with respect to one another and at
the same time come into series with the now superficial extended and subdivided
first oral tentacle scales which have come to assume the function of the more or less
reduced (ultimately absent) distal oral papillae (or second oral tentacle scales). I
find Lyman's description of the oral structure of AmphiUpis as stouter than that of
most amphiurids very surprising. On the contrary it appears to me to be much
more delicate, though this is especially noticeable in lateral view of the jaws (fig. la,
d), which Lyman did not illustrate. [It should be noted that Murakami's remarks
about the jaw structure of the Amphilepididae (1963), being based on AmphiUpis
diastata, are also misleading since the species is more closely related to Amphiura
diomediae than to AmphiUpis norvegica ; see p. 40.]

In view of the series of species bridging the gap between Amphioplus and Amphi-
Upis, as far as jaw structure goes, I find it difficult to maintain the latter as repre-
sentative of a distinct family. Accordingly I propose to reduce the rank of the
Amphilepididae to that of a subfamily of the Amphiuridae.

AMPHILEPIS Ljungman

Amphilepis Ljungman, i86yb : 322 ; Lyman, 1882 : 149 ; H. L. Clark, 1915 : 244. Type-
species : Amphiura norvegica Ljungman, 1864.

FIG. 11. a. " Amphioplus " formatus (Koehler), Amsterdam Museum, holotype, " Siboga "
st. 310 or 312, off Sumbawa, E Indies, showing narrow teeth and very long arm spines ;

b. " Amphioplus " agassizii Verrill, M.C.Z. 1261, holotype, off St. Lucia, W Indies ;

c. " Amphioplus " nereis (Lyman), M.C.Z. 1260, holotype, off Montserrat, W Indies
(b and c showing chilophiurid-type distal oral tentacle scale and genital papillae) ;
d and e. Amphiactis cythera (A. H. Clark), U.S.N.M. E.6gi6, syntype, Hawaiian Is., e. the
second free arm segment showing serrated arm spines ; f. Amphiactis sp. Icanescens
(Lyman), M.C.Z. 3501, syntype of Amphioplus gentilis (Koehler) but not conspecific with
most other types, " Siboga " st. 178, N. of Ceram, E Indies ; g. Amphioplus (Unioplus)
gentilis (Koehler), Copenhagen Museum, Kei Is. Expedition st. 23 (see also fig. 8r) ;
h and i. " Amphiactis " astarte A. H. Clark, U.S.N.M. E.yo2i, holotype, Hawaiian Is. ;
showing chilophiurid-type distal oral tentacle scale ; the oval at the interradial apex of
the adoral shields is an erect papilla, there is also a small nodule at the distal abradial
corner of each shield not mentioned by A.H.C. ; j and k. Amphiactis lycidas A. H. Clark,
U.S.N.M. .7019, holotype, Hawaiian Is., the disc rather distorted.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 71

a

72 AILSA M. CLARK

Thirteen species have been referred to Amphilepis, namely :
antarctica Koehler, 1907 patens Lyman, 1879

diastata Murakami, 1942 platytata H. L. Clark, 1911

gymnopora Hertz, 1927 protecta Koehler, 1904

ingolfiana Mortensen, 1933!) remittens Koehler, 1922

mobilis Koehler, 1904 scutata Mortensen, ig33a

norvegica (Ljungman, 1864) tennis Lyman, 1879

papyracea Lyman, 1879

Of these, A. protecta has been referred to Ophiomonas by Djakonov (1952) and
diastata to Amphiura by me (1965), while Madsen (1967) suspects that gymnopora
will prove to be a synonym of Amphiura belgicae. Among the remainder, Amphi-
lepis infolfiana, mobilis, platytata and scutata are closely related to the type-
species, A. norvegica, having relatively narrow teeth, the more superficial ones at
least pointed. The affinities of Amphilepis antarctica are uncertain, the holotype
and only recorded specimen having d.d. only 2-5 mm. It is unusual in having the
radial shields no bigger than the disc scales. Koehler's figures are so diagrammatic
that it is impossible to determine its relationships. The holotype and only known
specimen of Amphilepis tenuis is very badly decalcified and few of its characters
are still distinguishable ; nor were Lyman's figures of it very good. However, it
appears to have relatively broad squared-off teeth, in which it resembles Amphilepis
gymnopora, 1 papyracea, patens and remittens, though in all these the ventral-
most one may be more or less rounded or with a small median cusp and the apex of the
jaw is also broad, the proportions and positions of the associated papillae and lowest
tooth being somewhat variable. In some of these species there may be a superficial
median apical papilla (e.g. in the two syntypes of A . gymnopora on some jaws) or the
first oral papilla of each side may approximate to its partner and occupy an infra-
dental position (e.g. in the left hand jaw shown of the holotype of A . patens in fig. 8c).
In the case of A. remittens at least this blunting of the apex of the jaw is carried so
far that the bases of the two innermost papillae, which are truly infradental, are
tangential rather than oblique and I think it may be better referred to Ophiomonas
like A . protecta. Unfortunately the holotype, which I have seen, is in poor condition
and the details of the oral structure are difficult to make out. One other species,
Amphiura pycnostoma H. L. Clark, 1911, should be added to this group of species
of Amphilepis. The holotype in the Smithsonian Institution has the elongated
oral tentacle scales superficial, as indeed H. L. Clark's figure suggests.

With only the holotypes of Amphilepis patens and papyracea to work on I am
unable to compare the internal structure of any of these four species with that of
Amphilepis norvegica. There is however, one obvious difference in A . patens, namely
that it has long genital slits, though I do not think this is true of papyracea. A
proper assessment of the relationships of these four species to those of Amphilepis
sensu stricto (i.e. including norvegica, infolfiana, mobilis, platytata and scutata) must
await further material.

1 Hertz describes the teeth of A. gymnopora as pointed but this is not supported by her photograph,
which shows the apex of the jaw as relatively broad and blunt ; possibly the most superficial tooth is
pointed.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 73

Family OPHIAGTIDAE
HISTAMPICA 1 nom. nov.

Amphiactis Matsumoto, 1915 : 66-67 > J 9 J 7 : 146-147 ; H. L. Clark, 1918 : 294 (discussed
under heading of Ophiactis) ; Koehler, 1922 : 204 ; A. H. Clark, 1949 : 29. Type-species :
Amphiactis umbonata Matsumoto, 1915. Non Amphiactis Verrill, 1869, Coelenterata.

Unfortunately the generic name Amphiactis as used by Matsumoto is preoccupied
and must be replaced.

As noted in the discussion above of Matsumoto 's family Amphilepididae, Fell
(1960) followed H. L. Clark's short-lived idea (1915, revoked in 1918) that Amphiactis
is a synonym of Ophiactis. I consider that, on the contrary, it is more widely marked-
off from Ophiactis than Amphioplus is from Amphiura, indeed almost as remote as
Amphilepis is from Amphiura ; not only are there four or five oral papillae in each
series rather than one or sometimes two, as found in Ophiactis, but also the teeth of
Amphiactis are narrowed in a comparable fashion to those of Amphilepis, with which
Matsumoto related it. I regard this as a convergent modification, not a sign of
affinity. The internal structure of Amphiactis and Ophiactis is very different,
judging from Matsumoto's studies.

The following nominal species have been referred to Amphiactis :
Amphiura canescens Lyman, 1879 Amphiactis umbonata Matsumoto, 1915

Ophiactis dissidens Koehler, 1904 type-species

Amphiura duplicata Lyman, 1875 Ophiactis pectorals Lyman, 1880

Ophiactis parata Koehler, 1904 Ophiocnida picteti de Loriol, 1893

Amphiura partita Koehler, 1897 Amphiactis astarte A. H. Clark, 1949

Amphiura patula Lyman, 1879 Amphiactis lycidas A. H. Clark, 1949.

The first seven were included by Matsumoto, followed by H. L. Clark who added
pectoralis and picteti (the latter conditionally) and the last two by A. H. Clark.

As mentioned on p. 41 Amphiura patula Lyman is an Amphioplus, the apex of
the jaw being broad, the additional median third infradental papilla being a peculiar-
ity of the holotype and not shared b}^ the paratype.

I have examined the holotype of Ophiactis pectorale and find that its jaw
structure and interradially-folded disc accord not with the Ophiactidae but with
certain members of the family Ophiacanthidae ; also it possesses the very distinctive
diminutive hooks on the upper parts of the lateral arm plates characteristic of the
ophiacanthid Ophiodictys uncinatus Koehler, 1922, though these were not noticed
by Lyman. The two are certainly conspecific so pectorale is now referred to Ophio-
dictys with uncinatus becoming a synonym.

I think that H. L. Clark's misgivings about Ophiocnida picteti were justified ;
since it has only two distal oral papillae each side, broad teeth with only a single
infradental papilla and the disc with scattered spinelets, it is undoubtedly referable
to Ophiactis.

Unfortunately I did not study the types of Ophiactis parata and dissidens when
visiting Amsterdam. Koehler 's figures are not sufficiently precise to fix the

1 Histampica, an anagram of Amphiactis. Gender: feminine.

74 AILSA M. CLARK

relationships of the species with any certainty. The three tentacle scales on the
proximal arm segments of 0. dissidens and the rugose arm spines are features other-
wise unknown in the Ophiactidae but which do occur in the Ophiacanthidae, to
which the species must be referred ; it has some affinity with Ophiophrura H. L.
Clark, 1911, except that the radial shields are distinct. Ophiactis parata, while not
sharing these two characters, has an even closer resemblance to another ophiacanthid,
namely Ophiocopa singularis Koehler, 1922, particularly the very short fully-
contiguous radial shields with granular armament along their distal edges, unlike any
ophiactid. I suspect that a comparison of specimens will prove the two to be
conspecific, certainly they are congeneric.

Ophiactis partita, described earlier by Koehler, is evidently an Amphiactis in
Matsumoto's sense, having been referred to the synonymy of Amphiactis duplicata
by Koehler (1922).

I have seen the holotypes of both Amphiactis lycidas (fig. nj, k) and astarte
(fig. uh, i) in the Smithsonian Institution. A. lycidas superficially recalls some
ophiacanthids, especially in the proximal constriction of each arm segment, however,
there seems no good reason for removing it from Amphiactis. Amphiactis astarte
on the other hand resembles some of the Ophiolepidinae, notably Ophiozonella.
Matsumoto has pointed out that Amphiactis umbonata, the type-species, has some
resemblance in the disc scaling to Ophiozona (from which he split off Ophiozonella).
However, in A. astarte an additional feature is that there is a fifth oral papilla
at the distal end of the series (not mentioned by A. H. Clark) which is inclined up
into the oral slit, a characteristic of the order Chilophiurida, which agrees with
inclusion of the species in the Ophiolepidinae not the Ophiactidae, the latter being
a family of the Gnathophiurida. I think that astarte should be referred to Ophio-
zonella.

Finally, the holotype of Amphioplus cythera A. H. Clark, 1949 (fig. nd, e) is in
my opinion congeneric with Amphiactis umbonata. It has rounded or tapering teeth,
narrower than is usual in Amphioplus and there are no infradental papillae ; the
structure of the arms is very like that of ophiactids.

The following species can therefore be included in Histampica : H. canescens
(Lyman), cythera (A. H. Clark), duplicata (Lyman), lycidas (A. H. Clark) and
umbonata (Matsumoto).

SUMMARIES OF PROPOSED TAXONOMIC CHANGES
I. Changes in Rank, New Taxa and Synonyms

A. GENUS-GROUP TAXA page

Amphinephthys Fell a synonym of Amphiura Forbes . . . . . 12

Icalia Fell a synonym of Amphiura Forbes . . . . . . .12

Pandelia Fell a synonym of Amphiura Forbes . . . . . . 12

Hemilepis Ljungman a synonym of Amphiura Forbes . . . . . 12

Nullamphiura Fell a synonym of Amphiura Forbes . . . . . 12

Monamphiura Fell a synonym of Amphiura Forbes . . . . . 12

Ophiopeltis Diiben & Koren reduced to a subgenus of Amphiura Forbes . . 12

NOTES ON THE FAMILY AMPHIURID AE (OPHIUROIDE A) 75

Fellaria subgen. nov. of Amphiura for type Ophionephthys octacantha H. L. Clark 18

Ophionema Liitken reduced to a subgenus of Amphiura Forbes ... 19

Ophionephthys Liitken restricted to the type-species ..... 20

Gymnodia Fell a synonym of Amphiodia Verrill ...... 25

Diamphiodia Fell a synonym of Amphiodia Verrill ..... 25

Amphiodia (Amphispina) validated by type-designation .... 25

Nullopholis Fell a synonym of Amphipholis Ljungman . . . . .31

Monopholis Fell a synonym of Amphipholis Ljungman . . . . .31

Amphistigma H. L. Clark a valid genus distinct from Ophiostigma ... 33

Amphiacantha Matsumoto pre-occupied, replaced by

Dougaloplus nom. nov. .......... 33

Amphichilus Matsumoto reduced to subgenus of Amphioplus Verrill . . 48

Unioplus Fell reduced to a subgenus of Amphioplus Verrill .... 48

Ailsaria Fell a synonym of Amphioplus Verrill ...... 50

Lymanella subgen. nov. of Amphioplus for type A. hastatus Ljungman . . 51

Triodia gen. nov. for type T. abdita sp. nov. ...... 67

Amphiactis Matsumoto preoccupied, replaced by

Histampica nom. nov. .. \ ....... 73

B. SPECIES-GROUP TAXA

Ophionephthys octacantha H. L. Clark designated type-species of Amphiura

subgen. Fellaria nov. . . . . . . . . . .18

Ophionephthys decacantha H. L. Clark a synonym of A. (Fellaria) octacantha

(H. L. Clark) 18

Amphiura novae Benham and abernethyi Fell both synonyms of A. correcta

Koehler ............ 22

Amphiura ceramis H. L. Clark a synonym of A. lorioli (Koehler) ... 23

Amphiodia cyclaspis Djakonov a synonym of A. craterodmeta H. L. Clark . 28

Amphipholis murex (Koehler) a synonym of Ophiocnida loveni (Ljungman) . 31

A mphiacantha dividua Matsumoto a synonym of Ophiocomella sexradia (Duncan) 33

Amphioplus luctator Koehler ?a synonym of Dougaloplus echinatus (Ljungman) . 36

Amphichilus cesareus (Koehler) a synonym of Amphioplus impressus (Ljungman) 45

Amphioplus praestans (Koehler) a synonym of A. (Lymanella) laevis (Lyman) . 53

Amphioplus bocki Koehler ? a synonym of A. (Lymanella) laevis (Lyman) . 53
Amphioplus megapomus H. L. Clark a synonym of either A. (Lymanella) laevis

(Lyman) or of A. (L.) andreae (Liitken) ....... 53

Amphioplus miyadii Murakami ? a synonym of A. (Lymanella) laevis (Lyman) . 53
Amphioplus potens Koehler a synonym of either A. (Lymanella) laevis (Lyman)

or of A. (L.) andreae (Liitken) ........ 55

Ophiodictys uncinatus Koehler (Ophiacanthidae) a synonym of O. pectoralis

(Lyman), referred from Ophiactis ........ 73

Ophiactis parata Koehler ? a synonym of Ophiocopa singularis Koehler (Ophia-
canthidae) ............ 74

Pre-1962 combination

Genus according
to Fell

Genus according
to A.M.C.

Ophionephthys africana Balinsky
Ophionephthys heptacantha Mtsn.
Ophionephthys iranica Mtsn.

1 Not exhaustive for Fell's larger subdivisions.

Ophiopeltis
Ophiopeltis
Ophiopeltis

Amphiura (Fellaria)
Amphiura (Fellaria)
A mphiura (Ophiopeltis)

7 6

AILSA M. CLARK
TRANSFERS OF SPECIES (cont.)

Pre-1962 combination

Genus according
to Fell

Genus according
to A.M.C.

Ophionephthys magellanica Mtsn. Ophionephthys

Ophionephthys octacantha H.L.C. Ophiopeltis

Ophionephthys phalerata Lyman Ophionema

Ophionephthys stewartensis Mtsn. Ophionephthys

Ophionephthys tenuis H.L.C. Ophionema
Ophionema hexactis Mtsn.

Amphiura aestuarii Matsumoto Ophiopeltis

Amphiura alba Mtsn. & other spp. ref. to Monamphiura
Amphiura crossota Murakami & other spp.

ref. to Amphinephthys

Amphiura ecnomiotata H.L.C. Ophiopeltis
Amphiura fragilis Verrill & other spp. ref. to Icalia
Amphiura hinemoae Mtsn. & other spp. ref. to Pandelia
Amphiura psilopora H.L.C. & other spp. ref.

to Nullamphiura

Amphiura pycnostoma H.L.C. Monamphiura
Amphiura semiermis Lyman & other spp. ref.

to Hemilepis
Amphiura sinicola Matsumoto

Amphiura vadicola Matsumoto Ophiopeltis

Amphipholis loripes Koehler Monopholis

Amphipholis murmanica Djakonov Monopholis

Amphipholis nudipora A.M.C. Nullopholis

Amphipholis pentacantha H.L.C. Nullopholis

Amphipholis vitax Koehler Monopholis

Amphistigma minuta H.L.C. Ophiostigma

Amphiodia affinis (Studer) Amphiodia

Amphiodia assimilis (Ltk. & Mtsn.) Diamphiodia

Amphiodia caulleryi (Koehler) Diamphiodia

Amphiodia fuscoalba (Brock)

Amphiodia grata (Koehler) Diamphiodia

Amphiodia ochroleuca (Brock) Diamphiodia

Amphiodia olivacea (Brock) Amphiodia

A mphiodia platyspina Nielsen Gymnodia

Amphiodia psilochora H.L.C. Gymnodia

Amphiodia reposita (Koehler) Diamphiodia

Amphiodia servata (Koehler) Diamphiodia

Amphiodia tabogae Nielsen Gymnodia

Amphiodia urtica (Lyman) Ophiophragmus
Amphiodia violacea (Lutken) & other spp. ref.

to Diamphiodia
Ophiocnida echinata (Ljungman)
Ophiocnida picteti de Loriol

Ctenamphiura sinensis A.H.C. Diamphiodia
Amphiacantha acanthina (H.L.C.)

Amphiacantha amphacantha (McCl.)

A mphioplus (A mphioplus) 1
Amphiura (Fellaria)
Amphiura (Ophiopeltis)
A mphioplus (A mphioplus)
A mphiura (Ophiopeltis}
Amphiura (Ophiopeltis)
Amphiura (Ophiopeltis)
Amphiura (Amphiura)

Amphiura (Amphiura)
Amphiura (Fellaria)
Amphiura (Amphiura)
Amphiura (Amphiura)

Amphiura (Amphiura)
Amphilepis

Amphiura (Amphiura)
Amphiura (Fellaria)
Amphiura (Fellaria)
Amphipholis or Amphiodia

A mphipholis

Amphipholis

Amphipholis

Amphipholis

Amphistigma

A mphioplus

Amphiura (Amphiura)
? Amphiura or Amphioplus

(A mphichilus)

Acrocnida

Amphioplus (Unioplus)
Amphioplus (Amphichilus)

Ophiophragmus

A mphiodia

A mphioplus
Amphioplus (Unioplus)
Amphioplus (Unioplus)

A mphiodia

Amphiodia

A mphiodia

Dougaloplus

Ophiactis

Paracrocnida

Dougaloplus

Dougaloplus

1 Referred to Amphioplus by Castillo (1968).

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA)

TRANSFERS OF SPECIES (cont.)
combination

77

Genus according
to Fell

Genus according
to A.M.C.

A mphiacantha formosa (Ltk.)
A mphiacantha gastracantha (Ltk. & Mtsn.)
A mphiacantha liber a (Koehler)
Amphiacantha notacantha (Ltk. & Mtsn.)
Amphiacantha transacta Koehler
Amphichilus daleus (Lyman)
Amphichilus intermedius (Koehler)
Amphichilus trichoides Matsumoto

Amphioplus aciculatus Mtsn. Unioplus

Amphioplus agassizi Verrill Amphioplus

Amphioplus andreae (Liitken) Amphioplus

Amphioplus brachiostictus Tort. Amphiodia

Amphioplus caelatus Ely Amphioplus

Amphioplus capax Koehler Unioplus

Amphioplus cernuus (Lyman) Unioplus

Amphioplus conditus (Koehler) Amphioplus

Amphioplus confinis (Koehler) Amphioplus

Amphioplus congensis (Studer) Amphioplus

Amphioplus coniortodes H.L.C. Ailsaria

Amphioplus cyrtacanthus H.L.C. Ophionephthys

Amphioplus cythera A.H.C. Amphioplus

Amphioplus depressus (Ljungman) Amphioplus

Amphioplus dispar (Koehler) Unioplus

Amphioplus echinulatus Mtsn. Ailsaria

A mphioplus falcatus Mtsn. Unioplus

Amphioplus formatus (Koehler) Unioplus

Amphioplus gentilis (Koehler) Amphioplus
Amphioplus gibbosus (Ljungman)

Amphioplus glaucus (Lyman) Unioplus

Amphioplus hastatus (Ljungman) Amphioplus

Amphioplus incisus (Lyman) Amphioplus

Amphioplus integer (Ljungman) Amphioplus

Amphioplus japonicus (Mats.) Amphioplus

Amphioplus laevis (Lyman) Amphioplus

Amphioplus lorioli( Koehler) Amphioplus

Amphioplus lucidus Koehler Ophionephthys

Amphioplus macraspis H.L.C. Unioplus
Amphioplus modestus (Koehler)

Amphioplus nereis (Lyman) Amphioplus

Amphioplus patulus (Lyman) Unioplus

Amphioplus personatus (Koehler) Ailsaria

Amphioplus philohelminthius Zies. Unioplus

Amphioplus seminudus Mtsn. Ophionephthys
Amphioplus spinulosus (Koehler)

Amphioplus strongyloplax (H.L.C.) Unioplus

Amphioplus thrombodes (H.L.C.) Unioplus

A mphioplus trepidus (Koehler) ? Silax
Amphilepis diastata Murakami
Amphilepis remittens Koehler

Dougaloplus

Dougaloplus

Dougaloplus

Dougaloplus
order Chilophiurida
Amphioplus (Unioplus)

Amphioplus

Amphioplus (Amphichilus)
A mphioplus (A mphioplus)
family Ophiochitonidae
Amphioplus (Lymanella)

A mphioplus
Amphioplus (Lymanella)

Amphioplus
Amphioplus (Unioplus)
Amphioplus (Unioplus)
Amphioplus (Unioplus)
Amphioplus (Lymanella)
Amphioplus (Amphioplus)
Amphioplus (Amphioplus)

Histampica
A mphioplus (Lymanella)

Amphioplus

Amphioplus (Amphioplus)
Amphioplus (Unioplus)
? not amphiurid
Amphioplus (Unioplus)
Amphioplus (Lymanella)

Amphioplus

Amphioplus (Lymanella)
Amphioplus (Unioplus)
Amphioplus (Lymanella)
Amphioplus (Lymanella)
Amphioplus (Lymanella)

Amphiura

Amphioplus
Amphioplus (Amphioplus)

Ophiochiton
Ophiochitonidae
Amphioplus (Unioplus)
Amphioplus (Amphioplus)

Amphioplus

Amphioplus (Amphioplus)
Amphioplus (Lymanella)
Amphioplus (Amphioplus)
Amphioplus (Amphioplus)

Ophiocomidae

Amphiura

? Ophiomonas

78 AILSA M. CLARK

TRANSFERS OF SPECIES (cont.)

Pre-1962 combination Genus according Genus according

to Fell to A.M.C.

Amphiactis astarte A.H.C. Ophiozonella (Ophiuridae)

Amphiactis canescens (Lyman) Histampica

Amphiactis dissidens (Koehler) Ophiacanthidae

Amphiactis duplicata (Lyman) Histampica

Amphiactis lycidas A.H.C. Histampica

Amphiactis pectorale (Lyman) Ophiodictys (Ophiacanthidae)

Amphiactis umbonata Matsumoto Histampica

Ophiochytra epigrus Lyman Ophiuridae

REFERENCES

BALINSKY, J. B. 1957. The Ophiuroidea of Inhaca Island. Ann. Natal Mus. 14 : 1-33, 7

figs., 4 pis.
BENHAM, W. B. 1909. Scientific results of the New Zealand Government Trawling Expedition,

1907. Rec. Canterbury Mus. 1(2) : 1-34, pis. 7-11.

BROCK, J. 1888. Die Ophiuridenfauna des indischen Archipels. Z. wiss. Zool. 47 : 465-539.
BUCHANAN, J. B. & WOODLEY, J. D. 1963. Extension and retraction of the tube-feet of

Ophiuroids. Nature, Lond. 197 : 616-617, 2 figs.
BURFIELD, S. T. 1924. A new species and a new variety of ophiuroid with notes on a collection

of Ophiuroidea from the Sudanese Red Sea. Ann. Mag. nat. Hist. (9)13 : 144-154, 2 figs.
CASTILLO ALARCON, J. 1968. Ophiuroideos colectados por la XIX Expedicion Antartica

Chilena. Publnes Inst. antart. Chile No. 13 : 1-34, 6 pis.
CHANG, F.-Y. 1948. Echinoderms of Tsingtao. Contr. Inst. Zool. Acad. Peiping 4 : 33-104,

24 figs., ii pis.
CHERBONNIER, G. 1957. Ophiures rares ou nouvelles des cotes de Sierra-Leone. Bull. Mus.

Hist. nat. Paris 29 : 163-171, 3 figs.
CLARK, A. H. 1949. Ophiuroidea of the Hawaiian Islands. Bull. Bernice P. Bishop Mus.

No. 195 : 1-133, 22 fig 8 -
CLARK, A. M. 1955. Echinodermata of the Gold Coast. // West Afr. Sci. Ass. 1(2) : 16-56,

23 figs., pi. 2.

J 965- Japanese and other ophiuroids from the collections of the Munich Museum. Bull.
BY. Mus. nat. Hist. Zool. 13(2) : 1-71, 6 figs., i pi.

1966. Port Phillip Survey, 1957-63. Echinodermata. Mem. natn. Mus. Viet. 27

289-355. 4 P ls -
1967. Echinoderms from the Red Sea, part 2. (Crinoids, Ophiuroids, Echinoids and

more Asteroids). Bull. Sea Fish. Res. Sta. Haifa 41 : 26-58, 5 figs.
CLARK, H. L. 1911. North Pacific Ophiurans in the collection of the United States National

Museum. Bull. U.S. natn. Mus. 75 : 1-302, 144 figs.
1915. Catalogue of recent Ophiurans. Mem. Mus. comp. Zool. Harv. 25(4) : J ^5-3

20 pis.

1917. Reports on the scientific results of the expedition to the tropical Pacific , on

the U.S. Fish Commission Steamer " Albatross ", 1899-1900. XVIII. Reports on the

scientific results of the expedition to the eastern tropical Pacific , by the U.S. Fish

Commission steamer " Albatross ", 1904-1905. XXX. Ophiuroidea. Bull. Mus. oomp.
Zool. Harv. 61 : 429-453, 5 pis.

1918. Brittle-stars new and old. Bull. Mus. comp. Zool. Harv. 62 : 265-338, 8 pis.
1938. Echinoderms from Australia. Mem. Mus. comp. Zool. Harv. 55 : viii + 596, 63 figs.,

28 pis.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 79

CLARK, H. L. 1939. Ophiuroidea. Scient. Rep. John Murray Exped. 4 : 29-136, 62 figs.

1946. The Echinoderm fauna of Australia. Publs Carnegie Instn No. 566 : 1-567.

DJAKONOV, A. M. 1954. [Ophiuroidea of the Soviet Union.] Opred. Faune SSSR 55 : 1-136,

47 figs., i pi. [In Russian ; translation, 1967 ; Israel Program for Scientific Translations,

123 pp.]

DUBEN, M. W. VON & KOREN, J. 1845. Arch, skand. Beitr. Naturgesch. Part 1 : 436-440.
1846. Ofversigt af Skandinaviens Echinodermer. K. svenska VetenskAkad. Handl. 1844 :

229-328, pis. 6-1 1.
DUNCAN, P. M. 1879. On some Ophiuroidea from the Korean Seas. /. Linn. Soc., Zool. 14 :

445-482, pis. 9-1 1.

1887. On the Ophiuridae of the Mergui Archipelago, collected for the Trustees of the
Indian Museum by Dr. John Anderson. /. Linn. Soc., Zool. 21 : 85-106, figs. 28-40, pis. 8,
9 and n.

ELY, C. A. 1942. Shallow-water Asteroidea and Ophiuroidea of Hawaii. Bull. Bernice P.

Bishop Mus. No. 176 : 1-63, 18 figs., 13 pis.
FELL, H. B. 1951. Some off-shore and deep-sea ophiuroids from New Zealand waters. Zool.

Publs Viet. Univ. Wellington No. 13 : 1-4, 4 figs.

1958. Deep-sea Echinoderms of New Zealand. Zool. Publs Viet. Univ. Wellington
No. 24 : 1-40, 5 pis.

1960. Synoptic keys to the genera of Ophiuroidea. Zool. Publs Viet. Univ. Wellington

No. 26 : 1-44, 6 figs.
i962a. A revision of the major genera of amphiurid Ophiuroidea. Trans. R. Soc. N.Z.,

Zool. 2 : 1-26, i pi.

19620. A revised classification of the Australian Amphiuridae (Ophiuroidea). Proc.
Linn. Soc. N.S.W. 87 : 79-83.

FORBES, E. 1843. On the Radiata of the eastern Mediterranean. Pt. I. Ophiuridae. Trans.

Linn. Soc. Lond. 19 : 143-153, pis. 13, 14.
KOEHLER, R. 1897. Echinodermes recueillis par 1'Investigator dans 1'Ocean Indien. i. Les

Ophiures de Mer Profonde. Ann. Sci. Nat. (8)4 : 277-372.
1899. An account of the deep-sea Ophiuroidea collected by the Royal Indian Marine

Survey Ship Investigator. Echinoderma of the Indian Museum, viii + 76 + 11 pp., 14 pis.

Calcutta.
1904. Ophiures de 1'Expedition du Siboga. i. Ophiures de mer profonde. Siboga

Exped. 45a : 1-167, 3^ pis.

1905. Ophiures de 1'Expedition du Siboga. 2. Ophiures littorales. Siboga Exped.
45b : 1-142, 18 pis.

1907. Revision de la collection des Ophiures du Museum d'Histoire naturelle de Paris.

Bull, sclent. Fr. Belg. 41 : 279-351, 5 pis.
1914. Asteroidea, Ophiuroidea et Echinoidea. In Michaelsen & Hartmeyer. Beitrdge

zur Kenntnis der Meeresfauna Westafrikas. 2 : 129-303, pis. 4-15. Hamburg.
1917. Echinodermes (Asteries, Ophiures et Echinides) recueillis par M. Rallier du Baty,

aux lies de Kerguelen, en 1913-14. Annls Inst. oceanogr. Monaco 7(8) : 1-82, 10 pis.
i922a. Ophiurans of the Philippine Seas and adjacent waters. Bull. U.S. natn. Mus.

100(3) : 1-486, 103 pis.
i922b. Ophiuroidea. Scient. Rep. Austral. Ant. Exped. C8(2) : 1-98, pis. 76-90.

I927a. Revision de quelques ophiures de Ljungman, appartenant au Musee d'Histoire
Naturelle de Stockholm. Ark. Zool. 19A(2) : 1-29, 4 pis.

I927b. Ophiures recueillis aux iles Gilbert, Marshall et Fiji. Goteborgs K. Vetensk. o.
VitterhSamh. Handl. 33(3) : I - I 3. I pi-

1930. Papers from Dr. Th. Mortensen's Pacific Expedition, 1914-16. LIV. Ophiures
recueillis par le Docteur Th. Mortensen dans 1'Archipel Malais. Vidensk. Meddr dansk.
naturh. Foren. 89 : 1-295, 2O pis-

LEACH, W. E. 1815. Zoological Miscellany. 2 : 1-154, I2 P^ 3 -

LJUNGMAN, A. i867a. Om nagra nya arter af Ophiurider. Ofvers. K. VetenskAkad. Fork.
Stockh. 1866 : 163-166.

So AILSA M. CLARK

LJUNGMAN, A. i867b. Ophiuroidea viventia hue usque cognita enumerat. Ofvers. K.

VetenskAkad. Fork. Stockh. 1866 : 303-336.
-- 1871. Forteckning ofver uti Vestindien af Dr. A. Goes samt under Korvetten Josefinas

Expedition i Atlantiska Oceanen samlade Ophiurider. Ofvers. K. VetenskAkad. Fork.

Stockh. 1871 : 615-658.
LUTKEN, C. F. 1856. Bidrag til kundskab om Slangestjernerne. II. Oversigt over de

vestindiske Ophiurer. Vidensk. Meddr dansk. naturh. Foren. 1856 : 1-19.

and b. Additamenta ad Historiam Ophiuridarum. i and 2. Beskrivelser af ny

eller hidtil kun ufuldstaendigt hjendte Arter af Slangestjerner. K. dansk. Vidensk. Selsk.
Skr. 6(5) : 1-74, 2 pis. ; 179-271, 5 pis.
1869. Additamenta ad Historiam Ophiuridarum. 3. Beskrivende og kritiske Bidrag

til Kundskab om Slangestjernerne. K. dansk. Vidensk. Selsk. Skr. 5(8) : 24-109, 3 figs.
LUTKEN, C. F. & MORTENSEN, T. 1899. Reports on an exploration off the west coasts of

Mexico, Central and South America, and off the Galapagos Islands. XXV. The Ophiuri-

dae. Mem. Mus. comp. Zool. Haw. 23(2) : 97-208, 23 pis.
LYMAN, T. 1865. Ophiuridae and Astrophytidae. Illust. Cat. Mus. comp. Zool. Harv.

No. 1 : 1-200, 19 figs., 2 pis.
1869. Preliminary report on the Ophiuridae and Astrophytidae dredged in deep water

between Cuba and Florida reef. Bull. Mus. comp. Zool. Harv. 1 : 309-354.
1874. Ophiuridae and Astrophytidae new and old. Bull. Mus. comp. Zool. Harv. 3 :

221-272, 7 pis.
1879. Ophiuridae and Astrophytidae of the " Challenger " Expedition. II. Bull. Mus.

comp. Zool. Harv. 6(2) : 17-83, 8 pis.

1880. A structural feature hitherto unknown among Echinodermata found in deep-sea
Ophiurans. Anniv. Mem. Boston Soc. nat. Hist. : 1-12, 2 pis.

1882. Ophiuroidea. Rep. scient. Res. Challenger, Zool. 5 : 1-386, 46 pis.

1883. Reports on the results of dredging, , in the Caribbean Sea (1878-79) and on

the east coast of the United States, during the summer of 1880, by the U.S. Coast Survey
steamer " Blake ". XX. Report on the Ophiuroidea. Bull. Mus. comp. Zool. Harv. 10:
227-287, 8 pis.

McCLENDON, J. F. 1909. The Ophiurans of the San Diego region. Univ. Calif. Publs Zool.

6(3) : 33-64. 6 pis.
MADSEN, F. J. 1955. Echinoderms other than Holothurians collected in Sub-Antarctic and

Antarctic seas, mainly by the Norvegia Expeditions, 1928-30. Scient. Res. Norweg.

Antarct. Exped. No. 37 : 1-17, 2 figs.
1967. Ophiuroidea. Rep. B. A. N.Z.Antarctic Res. Exped., 1929-31. B9 : 123-145, 8 figs.,

i pi.
MATSUMOTO, H. 1915. A new classification of Ophiuroidea : with descriptions of new genera

and species. Proc. Acad. nat. Sci. Philad. 67 : 43-92.
1917. A monograph of Japanese Ophiuroidea, arranged according to a new classification.

/. Coll. Sci. Tokyo 38(2) : 1-408, 100 figs., 7 pis.
1941. Report of the biological survey of Mutsu Bay and vicinities. Sci. Rep. Tohuku

Univ. (4)16 : 331-344, 10 figs., pis. 21-23.
MAY, R. M. 1924. The Ophiurans of Monterey Bay. Proc. Calif. Acad. Sci. (4)13 : 261-303,

17 figs.
MORTENSEN, T. 1924. Echinoderms of New Zealand and the Auckland-Campbell Islands.

II. Ophiuroidea. Vidensk. Meddr dansk. naturh. Foren. 77 : 91-177, 36 figs., 2 pis.

1927. Handbook of the Echinoderms of the British Isles, ix + 471 pp., 269 figs. London.
I933a. Echinoderms of South Africa (Asteroidea and Ophiuroidea). Vidensk. Meddr

dansk. naturh. Foren. 93 : 215-400, 91 figs., 12 pis.
I933b. Ophiuroidea. Dan. Ingolf-Exped. 4(8) : 1-121, 52 figs., 3 pis.

I 933 C - Biological observations on Ophiurids with descriptions of two new genera and

four new species. Vidensk. Meddr dansk. naturh. Foren. 93 : 1-121, 7 figs., i pi.

NOTES ON THE FAMILY AMPHIURIDAE (OPHIUROIDEA) 81

MORTENSEN, T. 1936. Echinoidea and Ophiuroidea. " Discovery " Rep. 12 : 199-348, 53 figs.,
9 pis.

1940. Echinoderms from the Iranian Gulf. Dan. sclent. Invest. Iran. Part 2 : 55112,
24 figs., 2 pis.

1952. Reports of the Lund University Chile Expedition, 1948-49. 3. Echinoidea and
Ophiuroidea. Ada Univ. Lund N.F. (2)47(8) : 1-22, 3 figs., i pi.

MURAKAMI, S. 1942. Ophiurans of Izu, Japan. /. Dep. Agric. Kyushu Univ. 7 : 1-36, 12
figs.

1943. Reports on the Ophiurans of Palao, Caroline Islands. Reports on the Ophiurans
of Yaeyama, Ryukyu. Ophiurans from some Gulfs and Bays of Nippon. /. Dep. Agric.
Kyushu Univ. 7 : 159-204, 17 figs. ; 205-222, 2 figs. ; 223-234, 2 figs.

1944. Report on the Ophiurans from off Ogasawara Islands and from off the Yaeyama
group, Nippon. Note on the Ophiurans of Amakusa, Kyusyu. /. Dep. Agric. Kyushu
Univ. 7 : 235-257, 14 figs. ; 259-280, 5 figs., i pi.

1963. The dental and oral plates of Ophiuroidea. Trans. R. Soc. N.Z., Zool. 4 : 1-48,
i fig., 7 pis.

NIELSEN, E. 1932. Ophiurans from the Gulf of Panama, California and the Strait of Georgia.

Vidensk. Meddr dansk. naturh. Foren. 91 : 241-346, 42 figs.
PARSLOW, R. E. & CLARK, A. M. 1963. Ophiuroidea of the Lesser Antilles. Stud. Fauna

Curacao 15 : 24-50, 4 figs.
SARS, G. O. 1871. Nye Echinodermer fra den norske Kyst. Fork. VidenskSelsk. Krist.

1871 : 1-31.

SARS, M. 1861. Over sigt af norges Echinodermer. 160 pp., 16 pis. Christiania.
SPENCER, W. K. & WRIGHT, C. W. 1966. Asterozoans. In Moore, R. C. Treatise on

Invertebrate Paleontology. Part U. Echinodermata 3, vol. 1 : 4-107, 89 figs. (Geo-

logical Society of America Inc. ; University of Kansas Press.)
STUDER, T. 1885. Die Seesterne Siid-Georgiens nach der Ausbeute der deutschen Polarstation

in 1882 und 1883. Jahrb. Hamb. Wiss. Anst. 11 : 143-166, 2 pis.
THOMAS, L. P. 1962. The shallow water amphiurid brittle stars (Echinodermata, Ophiuroidea)

of Florida. Bull. mar. Sci. Gulf Caribb. 12 : 623-694, 23 figs.

- 1963. A redescription of the amphiurid brittle star Ophiocnida cubana A. H. Clark, 1917.
Proc. biol. Soc. Wash. 76 : 217-222, 3 figs.

1964. Amphiodia atra (Stimpson) and Ophionema intricata Liitken, additions to the
shallow water amphiurid brittlestar fauna of Florida (Echinodermata : Ophiuroidea).
Bull. mar. Sci. Gulf Caribb. 14 : 158-167, 3 figs.

- 1966. A revision of the tropical American species of Amphipholis (Echinodermata :
Ophiuroidea). Bull. mar. Sci. Gulf Caribb. 16 : 827-833.

VERRILL, A. E. iSgga. Report on the Ophiuroidea collected by the Bahama Expedition from

the University of Iowa in 1893. Bull. Univ. Iowa 5(i) : 1-86, 8 pis.
iSggb. North American Ophiuroidea. i. Revision of certain families and genera of

West Indian Ophiurans. 2. A faunal catalogue of the known species of West Indian

Ophiurans. Trans. Conn. Acad. Arts Sci. 10 : 301-386, 2 pis.
ZIESENHENNE, F. C. 1937- Echinoderms from the west coast of Lower California, the Gulf

of California and Clarion Island. Zoologica, N.Y. 22 : 209-239, 2 figs.

AILSA M. CLARK, M.A.

Department of Zoology

BRITISH MUSEUM (NATURAL HISTORY)

LONDON, S.W. 7

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

A FURTHER SYSTEMATIC STUDY

OF THE GENUS ACARUS L., 1758

(ACARIDAE, ACARINA), WITH A

KEY TO SPECIES

D. A. GRIFFITHS

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 2

LONDON : 1970

A FURTHER SYSTEMATIC STUDY OF THE

GENUS ACARUS L., 1758 (AC ARID AE, ACARINA),

WITH A KEY TO SPECIES

BY

DONALD ALISTER GRIFFITHS

Pp. 83-118; 4 Plates, 40 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

ZOOLOGY Vol. 19 No. 2

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 2 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. not. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 27 February, 1970 Price i 8s.

KM)

A FURTHER SYSTEMATIC STUDY OF THE
GENUS ACARUS L, 1758 (ACARINA)

By D. A. GRIFFITHS

CONTENTS

page

1. INTRODUCTION .......... 85

2. GENERIC CHARACTERS ......... 86

3. MORPHOLOGICAL VARIATION ........ 87

4. PREPARATION OF MATERIAL FOR MICROSCOPIC EXAMINATION . . 92

5. KEY TO ADULTS AND DESCRIPTIONS OF SPECIES .... Q2

(i) Key to Adults ........ 92

(ii) Acarus nidicolous sp.n. ....... 93

(iii) Acarus macrocoryne sp. n. . . . . . . 102

(iv) Acarus chaetoxysilus sp.n. ...... 105

(v) Acarus calcarabellus sp.n. ...... 107

6. SPECIMENS INCERTAE SEDIS ........ 109

(i) Acarus specimens ex Formosan citrus fruit . . . 109
(ii) Acarus specimens ex Brazilian cocoanuts . . . . no
(iii) A carus specimens ex Chinese water-chestnuts . . . 112

7. RE-ATTRIBUTION OF MATERIAL LISTED AS SPECIMENS INCERTAE SEDIS

IN GRIFFITHS 1964 . . . . . . . . .113

8. AN APPRAISAL OF THE ACARUS SPECIMENS CONTAINED IN THE OUDE-

MANS' COLLECTION RIJKSMUSEUM VAN NATUURLIJKE HISTORIE,
LEIDEN . . . . . . . . . . .113

9. SUMMARY . . . . . . . . . . .117

10. ACKNOWLEDGEMENTS . . . . . . . . . 117

11. REFERENCES . . . . . . . . . .117

i. INTRODUCTION

VIRTUALLY on the eve of the publication of my revision of the genus Acarus L., 1758
(Griffiths, 1964) I found a thriving population of the species here described as Acarus
nidicolous sp.n. in the grounds of the Pest Infestation Laboratory. The chaetotaxy
of the dorsum, in particular the length of setae la and d\_ to d^, and the results of
crossing it with the species of the Acarus siro complex was such that had this taxon
been discovered earlier, some of the taxa which in the revision were placed as speci-
mens incertae sedis would have been given specific status. More recently, Dr. van
Eyndhoven, Dr. D. Furman and Dr. D. J. Johnston kindly presented me with Acarus
material which they considered did not correspond to any of the forms included in
the revision.

Thus, within a few months of completing a long study, it became necessary to
prepare a second paper, as a consequence of which the number of species in the genus
is almost doubled, an accurate reflection of the ever-changing situation in acarine
taxonomy at this present time.

The descriptions which appear within follow the pattern used to re-describe Acarus
farris Oudemans, 1905 in Griffiths, 1964. They were made with a new microscope

86

D. A. GRIFFITHS

and, as a result, I have discovered some omissions and some mistakes in my previous
descriptions of the A earns hypopi which I will amend here.

Firstly, a pair of supra-coxal setae are present in the motile and inert hypopal
forms. They are situated above coxae I, respectively. In the motile hypopi they
are similar in size and shape to the supra-coxal setae of Acarus nidicolous (Text-fig.
16). Those of Acarus immobilis Griffiths, 1964 are much shorter.

Secondly, in the motile Acarus hypopi and, so far this appears to be true for motile
hypopi of many other genera within the family, the structure which I termed seta ba
on tarsus I (see Text-figs. 10 and 21 of Griffiths 1964) is actually solenidion oj 3 , seta
ba is absent. This rule also applies to the inert hypopus of A . immobilis, the seta
labelled ba in Text-fig. 28 of Griffiths 1964 is a minute solenidion o> 3

2. GENERIC CHARACTERS

The generic characters of the adult stage as given in Griffiths 1964 still stand. In
addition, I am now satisfied that the morphology of the female reproductive system
is sufficiently different from that of closely related genera for it to be used as a
diagnostic character. It is described in Text-fig, i.

FIG. i. Acarus siro L., $ reproductive system; p. posterior margin of body; c. chitinised,
saucer-like depression in body wall; e. opening of expansible tube (d); b. chitinised, bell-
shaped structure attached to which is a thin walled, dilatable sac (s) t. chitinised opening
to thin walled oviduct (o).

SYSTEMATIC STUDY OF THE GENUS ACARUS 87

Zatchvatkin 1941 (see p. 115 in the English Translation by Ratcliffe & Hughes
1959) stated that " the setae of the dorsal surface [in Acarus hypopi] are longer than
in the hypopi of all other representatives of the family sc [scapular setae] being
particularly long. The fact that the hypopus of A. nidicolous sp.n., described
herein, has short scapulars which barely reach the posterior margin of the propodo-
somal shield (Text-fig. 16) means that the relative length of the scapular setae can no
longer be used as a diagnostic character at generic level. As more species are described,
it is becoming increasingly difficult to find characters which can serve to separate
motile hypopi of the family Acaridae into their respective genera. The practice of
describing new species and especially new genera based solely on the hypopal form
must be considered of doubtful benefit to acarine taxonomy.

3. MORPHOLOGICAL VARIATION

In my revision of the genus Acarus (Griffiths, 1964) material representing eight
taxa was designated as specimens incertae sedis because, at the time, there was no
way of determining whether the morphological differences exhibited by these taxa
represented intra-specific or inter-specific variation. Material collected since 1964
and the results of nutritional experiments performed on sibling females of A. siro
has now helped to clarify the position.

In Griffiths 1966 I showed that the tarsal length of adults varied according to the
nutritional quality of the food they received during the pre-adult period. An example
of this type of variation in a wild population is represented by Text-figures 23 and 24.
Recently, I have examined the effect which high and low nutrient diets have upon the
growth of the dorsal hysterosomal setae. Examples of the results obtained for
groups of sibling females are illustrated in Text-figs. 2a and b. A poor diet (Fig. 2b)
produces short, thin setae whereas a rich diet favours longer, thicker ones (Fig. 2a).
Their length expressed as a percentage of the respective idiosomal length is given
below.

Setal length as percentage of idiosomal length

Setae ? reared on ? reared on

poor food rich food

(Text-fig. 2b) (Text-fig, aa)

di 4-6 8-6

d z 7-7 J 5-6

d* 9.4 I4-4

d* 9-2 I5-9

la 6-2 10-8

lp 8-3 13-3

sa e 7-8 11-4

Seta d z shows the greatest variation. It is an interesting point that in the eco-
logically isolated species Acarus gracilis seta d% is much longer than d\ or d$ (Text-
fig. 8) and serves to distinguish this species from all Acarus species.

88

D. A. GRIFFITHS

sae

FIG. 2. Acarus siro L., idiosoma of sibling $'s illustrating variation in length of dorsal
setae (d\ to di), brought about by exposing pre-adult stages to (a) a rich diet and (b) a
poor one.

FIG. 3. Acarus siro L., idiosoma of $$'s taken from a natural population, showing
extreme range of variation in length of dorsal setae.

SYSTEMATIC STUDY OF THE GENUS ACARUS 89

I have found one population of A. siro in which the length of the hysterosomal
setae is extremely variable. This was taken from a bag of oatmeal which had been
left in a garden shed for more than five years. The range of variation in setal length
exhibited by the females of this population is illustrated in Text-fig. 3. It should
be noted that in individuals with the longest setae, the tips of the d% dorsals overlap
the bases of the d^ pair. However, in the vast majority of the many populations of
A . siro which I have examined, the length of the dorsal setae d\ to d is never as great
as the distance between its base and the base of the dorsal seta immediately posterior
to it.

FIG. 4. Acarus immobilis Griffiths. Idiosoma of $$'s taken from a seven year old stock
laboratory culture, illustrating extremes of variation in length of dorsal setae (di to di).

Some populations of A. immobilis and A. f arris which I have obtained since 1964
also exhibit variation in dorsal setal length. But, so far, I have not found an
individual of either species in which the length of seta dz or d% exceeds the distance
between the base of dz and d% or d% and d^. Text-figure 4 gives examples of the
extremities of the range of variation found in a population of A . immobilis, cultured
in the laboratory for seven years.

After studying many populations representing the A. siro species complex, I
believe the tendency in most populations is towards short dorsal setae. But, at the
extreme range of variation so far observed (i.e. in A. siro), the tips of the dz setae,
when projected backwards in a straight line, often overlap the bases of pair d%.

There is another group of species, described herein for the first time, in which the

go

D. A. GRIFFITHS

dorsal setae are usually long so that the tips of d\, d% and d% distinctly overlap the
base of the seta next in line (e.g. Text-fig. 9). But, in two species (A. nidicolous,
which I have in culture and A . macrocoryne, known from two preserved populations)
individuals can be found in which some dorsal seta are shorter than the distance
separating the relevant bases (Text-fig. 5).

Previously, the shape of solenidion on on tarsus II was the only diagnostic charac-
ter which, by itself, could be used to distinguish all the species in the genus. Plates
I, II and III illustrate that this is now no longer possible.

FIG. 5. Acarus macrocoryne sp.n. Idiosoma of $$'s taken from a natural population
showing range of variation in length of dorsal setae (d\ to d$).

Prior to this publication, genu I of Acarus males was said to bear two large and one
small process and, ventrally, femur I extended into a strong conical spur (Text-fig,
n). The new species described here include one in which the genu bears a single
large process sometimes subtended by a much smaller one and another species in
which the genual processes are completely lost with the femoral spur considerably
reduced in size. I have also seen a single A . immobilis specimen in which there is
only one large genual process, similar to the condition occurring in A . macrocoryne.

It is hoped that the above discussion on morphological variation will provide an
insight into why certain taxa (some previously described as specimens incertae sedis)
have been given specific rank whilst others have not. It has been dealt with in some
detail because despite the problem of discontinuous variation, I have had to use the
characters which exhibit this phenomenon in order to provide a key to species.
Therefore, identifications based on single specimen samples, or samples not properly
prepared for microscopic examination, may not be reliable. It is perhaps heartening
that the five known hypopal forms are easily distinguishable.

SYSTEMATIC STUDY OF THE GENUS ACARUS

FIGS. 6, 7, 8. 6, Acarus mirabilis Volgin. $. Dorsum of idiosoma; external scapular
setae sc e ; internal scapular setae sc i ; dorsal setae d\ to d*. Re-drawn after Volgin (1965) .
7, Acarus tyrophagoides (Zachvatkin) $. Dorso-lateral view idiosoma. 8, Acarus
gracilis Hughes $. Dorsum of idiosoma.

92 D. A. GRIFFITHS

4. PREPARATION OF MATERIAL FOR MICROSCOPIC EXAMINATION

Before making a permanent mount, the specimens require to be cleared. Add two
or three specimens to a drop of 70 per cent lactic acid and cover with a number o, one
cm diam. cover-slip. Place on a slide-warmer or in an oven at 5o-6oC for one or
two days.

Before making a permanent preparation, the lactic acid should be neutralised by
placing the cleared specimens in a drop of Vitzthum's fluid* for up to one hour or
more. This can best be achieved by surrounding the cover-slip of the temporary
lactic acid mount with Vitzthum's fluid, using the fluid to float off the cover-slip.
The specimens can then be pushed to a clean spot of fluid with the tip of a very fine
needle. Finally, remove to a clean slide and mount in Berlese's fluid.

If more than one specimen is available, some should be mounted in a dorso- ventral
position, others should be manoeuvred so that solenidion o>i on tarsus II can be viewed
laterally. One specimen per slide under a small cover-slip is preferable. The key to
adults is based on material prepared as above and examined by means of phase-
contrast microscopy. An oil-immersion objective of high power is required in order
to assess the degree of pectination of hysterosomal setae.

5. KEY TO ADULTS AND DESCRIPTION OF SPECIES
(i) KEY TO ADULTS

1 External scapular setae (sc e) more than twice as long as internals (sc i), e.g. Text-

fig. 6. ..... 2

External scapulars about equal in length to internal pair, e.g. Text-fig. 10. . . 3

2 Seta ds long, equal or almost equal in length to d^ and at least three times as long as

di or d z which are short (Text-fig. 7). Solenidion o>i with a stout shaft and
distinct terminal head. . . . . . . .A. tyrophagoides.

Seta ds short, about equal in length to di and dz, so that all three are much shorter
than ^4 (Text-fig. 6). . . . . . . . .A. tnirabilis.

3 Dorsal setae dz four to five times longer than di and twice length of da (Text-fig. 8) .
Solenidion wi usually slender, always tapering towards the apex (PI. I, fig. 3)

A. gracilis.
Length of dz intermediate between length of di and da, never more than three times

longer than di and at most a little longer than da ...... 4

4 Ventral distal spine s of tarsus I and II (not tarsus I of $) large, about equal in length

to length of tarsal claw ; ventro-posterior margin concave, claw tip directed back-
wards (PI. IV, fig. i). Solenidion <a\ recumbent with distinct 'goose-neck' before
terminal expansion (PL I, fig. 4). Usually, dorsal setae are short (see p. 89).

A. siro s. sir.
- Ventral distal spine s slender, about half length of tarsal claw; ventro-posterior

margin of claw convex, claw tip directed forwards (PI. IV, fig. 2) . . . . 5

5 Hysterosomal setae hi, la, Ip and dorsals di to d\ short. In particular, the length of

dorsal seta dz or ds does not exceed the distance between its base and base of the
dorsal seta immediately posterior to it. (e.g. Text-figs. 2a and b) . . . . . 6

Hysterosomal setae hi, la, Ip and dorsals longer; generally, in the majority of
individuals of a given population, the length of dz and ds is greater than distance
between its base and base of the seta immediately posterior to it (e.g. Text-fig. 15).
(See p. 90 for details of setal length variation) ....... 8

6 Solenidion o>i with sides almost parallel, expanding into a distinct egg-shaped terminal

head which at widest part is wider than widest portion of stem (PL I, figs, i & 2).

* Chloral hydrate 5og, Phenol 45g, Distilled water 500.

SYSTEMATIC STUDY OF THE GENUS ACARUS 93

Femoral spur and genual processes on leg i of ^ normal (e.g. Text-fig, n).

. A. immobilis.

- Solenidion wi without distinct egg-shaped terminal head: <$ leg I normal or with

reduced spur and genual processes absent ....... 7

7 Male leg I with normal spur and two tooth-like genual processes (Text-fig, n)*.

Solenidion o>i short, compact, with sides expanding gradually from the base then
narrowing to an indistinct neck before expanding into a terminal head. Width of
widest part of head equal to width of widest portion of stem (PI. II, fig. i). A. f arris
Male leg I with femoral spur much reduced so that seta vF arises almost at spur
apex; genual processes absent. (Text-figs. 33 & 35). Solenidion wi much more
slender, with a distinct neck in distal half (PL II, fig. 2). . A. calcarabellus sp.n., p. 107

8 Scapular setae with one or at most two pectinations, hysterosomal setae smoothf,

long (especially those of opisthosoma) terminating in fine whip-like endings (Text-
figs. 27 & 28). Solenidion coi short with parallel sides (PI. Ill, fig. 2).

A. chaetoxysilos sp.n., p. 105
Scapular setae strongly pectinate; hysterosomal setae with varying degrees of

pectination but some opisthosomal setae always pectinate .... 9

9 Hysterosomal setae dz, ds, sa i and sa e strongly pectinate ; also pectinations usually

but not always present on hi, d\, la and Ip. ....... 10

Setae dz and da smooth as are hi, di and la ; setae Ip and sa i may also be smooth . 1 1

10 Solenidion wi large, expanding gradually from the base into a large swollen tip; the

whole solenidion slightly curved (PL II, fig. 4). Male with normal femoral spur and a
single large tooth-like process on genu I (Text-figs. 22-24). ....

A. macrocoryne sp.n., p. 102

- Solenidion on short and thick, as thick as or thicker than basal portion of solenidion

psi of tibia; sides parallel with an indistinct terminal head (PL II, fig. 3). Male
genual processes normal in British populations (Text-fig, n); known from one
specimen from U.S.A. in which genu I bears a single tooth (Text-fig. 12).

A. nidicolous sp.n., p. 93

1 1 Hysterosomal setae of $ including sacrals are smooth (Text-fig. 39) . $ sacrals sparsely

pectinate but remainder of hysterosomal setae are smooth. Supra-coxal seta
long, tenuous and sparsely pectinate (Text-fig. 39). Solenidion wi short, strong,
terminating in an egg-shaped head, with convexity of one side more distinct (PL
III, fig. 3). . . . . Specimen incertae sedis ex cocoanuts, Brazil, p. no
Both sexes with opisthosomal setae Ip, d\ and sacrals pectinate. Supra-coxal seta

long and strongly pectinate ........ 12

12 Solenidion on with a slender shaft terminating in an obvious egg-shaped head.

(Text-fig. 38 and PL III, fig. 4)

Specimen incertae sedis ex citrus fruit, Formosa, p. 109
Solenidion on short, with a thick shaft terminating in an egg-shaped head (PL III,

fig. i). . . . Specimen incertae sedis ex water-chestnuts, China, p. 112

(ii) Acarus nidicolous sp.n

The gross morphology resembles that of the adults of the Acarus siro complex
(Griffiths, 1964) and, in the main, only the differences are discussed below.

MALE. Length of idiosoma of five males, reared on a wheat germ and yeast
mixture at 90 per cent relative humidity and 20C, = 0-365 mm average. Idiosomal

* The two species in couplet 7 are easily separated on secondary sexual characters on leg I of the $.
In the absence of cj<?'s, identification is based on shape of solenidion Wi on tarsus II, viewed laterally and
is more difficult. A . calcarabellus is very rare.

f Should be examined under oil-immersion phase-contrast microscopy.

94

D. A. GRIFFITHS

measurements were taken from the centre of an imaginary line connecting anterior
tips of epimeron fork to the centre of the posterior margin of the body.

(i) Dorsum : (Text-fig. 9) . Number and distribution of setae on dorsum and venter

10

FIGS. 9, 10, 9, Acarus nidicolous sp.n. $. Idiosoma, dorsum. Specimen collected in
Britain. 10, Acarus nidicolous sp.n. <$. Idiosoma, dorso-lateral view. Unique speci-
men from U.S.A.

same as for A . siro complex, but degree of pectination different in that all hysterc
somal setae are pectinate or, occasionally d\, hi and sometimes la may be smooth.
Length of dorsal idiosomal setae expressed as a percentage of idiosomal lengtl
(average of 5 individuals) are as follows : vi 17-2%, ve 4-9% sc i 29%, sc e 26%,

SYSTEMATIC STUDY OF THE GENUS ACARUS

95

hi 10-9%, he 23-3%, laio-1%, dt8-6%, ^2237%, ^324-6%, d 4 19-2%,
Ip 16-0%, sa e 13-5%, sa i 31-4%. Thus, with the exception of sa i, the dorsal
hysterosomal setae are longer than corresponding setae in the A cams siro complex.
In particular, dz and d% are over 20 per cent of the idiosomal length, so that when
directed backwards in a straight line, the tips of dz extend beyond bases of d% and
tips of dz extend just beyond or level with bases of d^.

11

C0 2

v+u

FIGS, u, 12. n, Acarus nidicolous sp.n. $. Leg I, post-axial face; structures labelled
co, a and <p = solenidia; e = famulus; remainder (e.g. aa, e, gT) = setae. 12, Acarus
nidicolous sp.n. <$. Leg I, post-axial face. Unique specimen from U.S.A.

Propodosomal shield similar in shape to that of A. f arris. Grandjeans' organs,
supra-coxal setae, not significantly different in shape from corresponding organs in A .
siro complex. Electron scanning micrographs showing the structure of Grandjeans'
organ and the supra-coxal seta of A. siro appear in PL IV, figs. 3 and 4, respectively.

(ii) Venter: Closely resembles ventral surface of males of the A. siro complex.
Same complement of setae, smooth or at most with one or two pectinations. Position
of post-anal setae relative to anal opening as for A . siro complex ; their lengths relative

9 6

D. A. GRIFFITHS

to each other also similar. No significant difference in structure of genitalia or
apodemes.

(iii) Legs : I & II. Five segmented with well developed pre-tarsus terminating in a
stalked claw. Genu and femur of leg I of $ much enlarged, femoral apophysis and
genual projections as found in A. siro complex (Text-fig, n). The chaetotaxy of
leg II and the shape of solenidion oi is described in Text-fig. 12. A micrograph of
o>i on leg II is also shown in Plate II, fig. 3. In shape, it is relatively short, thick,
and " cigar-shaped ". Throughout its whole length it is thicker than the basal
section of solenidion psi of the tibia.

cG

13

FIGS. 13, 14. 13, Acarus nidicolous sp.n. $. Leg II. 14, Acarus nidicolous sp.n. <J. Leg
II. Unique specimen from U.S.A.

(iv) Morphological variation :

Dr. Dean Furman presented me with a single $ specimen taken from the nest of
Neotomafuscipes, Wildcat Canyon, Alamada Co., California, U.S.A., June 1964. The
length of the idiosomal setae of this specimen appears to be the same as for British
males of A . nidicolous (compare Text-fig. 9 with 10) . However, genu I bears only one
large process instead of the usual two large and one small process (compare Text-fig,
ii with 12). These same figures illustrate that seta hT of leg I is smooth in British
males but pectinate on the North American specimen. Also on genu II solenidion
sigma of the North American specimen is much longer (compare Text-fig. 13 with 14).

Until further material can be obtained from the south-western areas of the United
States, this specimen must be considered a variant of A. nidicolous.

FEMALE (Text-fig. 15). Body larger and more roundly oval than male.

SYSTEMATIC STUDY OF THE GENUS ACARUS

97

(i) Dorsum : As for male, except that propodosomal setae, expressed as percent-
age of idiosomal length are slightly shorter, whereas the hysterosomal setae are
longer, except sa i which is about the same in both sexes. The average lengths of
idiosomal setae (from 5 females) are as follows :
w 14-9%, ve4'0%, sc i 24-5%, sc e 23-3%, hi 13-3%, he 24%, la 15-2%,

FIG. 15. Acarus nidicolous sp.n. $. Dorsum.

i n-6%, ^225-2%, ^327-4%, ^423-3%, lp22-3%, sa 622-3%, sa i
3i-o%.

(ii) Venter : Morphology same as for A. siro complex females. Length of anal
and post-anal setae, relative to each other, is very similar to that in A. siro sensu
stricto.

(iii) Legs : Leg chaetotaxy and shape of solenidion on as for male.

HYPOPUS (Text-figs. 16 to 21). Colour of live specimens pinkish-beige. Gross

98 D. A. GRIFFITHS

morphology the same as that of motile hypopi of the A. siro complex. Only the
differences are given below.

(i) Dorsum (Text-figs. 16) : Surface punctations smaller and not so dense.
Setae ve almost smooth compared with dense pectinations of the same setae in the
Acarus siro complex. All idiosomal setae are short, narrow and finely pointed. In
particular, the four scapular setae (sc i and sc e) do not extend beyond posterior margin
of the propodosomal shield. On extreme posterior margin a pair of setae, possibly
homologous with pa are the longest pair on the hysterosoma.

17

FIGS. 16, 17. Acarus nidicolous sp.n. Hypopus. 16, Dorsum. 17, Venter.

(ii) Venter (Text-fig. 17) : On gnathosoma, at base of each lobe which bears an
arista, there is a small clear circular area which is not found in other motile Acarus
hypopi.

Coxal skeleton much more defined. In freshly mounted specimens coxal fields

III and IV can be seen to be enclosed. However, in older, permanent mounts those
parts of the coxal skeleton which join apodemes III and IV and divide coxal fields

IV fade, so that coxal fields III and IV may appear to be open. I have recently dis-
covered that this phenomenon is true for other Acarus motile hypopi. Thus, the
character " closed or open coxal fields " is not a reliable one to use in a taxonomic
key.

SYSTEMATIC STUDY OF THE GENUS ACARUS

99

The pair of genital setae (g) arise on a level with the anterior margin of the pairs of
suckers which outflank them.

(iii) Legs : The four apical segments of legs I to IV are described in Text-figures
18 to 21.

LEG CHAETOTAXY OF ALL STAGES. Nomenclature follows Grandjean (1939).

The numbers of setae and solenidia on each segment of each leg of all stages is
given below as a formula. The famulus (e), which is present immediately in front
of o>i on tarsus I in all stages is not included in the formula. The five groups of
figures inside each parentheses represent, from left to right, tarsus, tibia, genu, femur
and trochanter.

Setae

Larva

Leg I
(13.2.2.1.0.)

II

(12.2.2.1.0.)

Ill

(lO.I.I.O.O.)

IV

Protonymph.

(13 2210)

(l2 22IO)

(lO.I.I O O.)

( 8. o.o. o.o.)

Hypopus

( 02210)

( Q22IO)

( 8. 1. 1. o.i.)

( 8.1.0.1.0.)

Tri tony mph

(13 2211)

(I2.2.2.I.I.)

(lO.I. I.O.I.)

(lO.I.O.I.O.)

Adult

(13.2.2.1.1.)

(I2.2.2.I.I.)

do. 1. 1. o.i.)

(lO.I.O.I.O.)

Solenidia

Larva

( T

1.2 O

)

(I

.i.i

.O.O.)

(o,

T T

.O.O.)

(

)

Protonymph

(2

,1.2 O

)

(l

T T

.O.O.)

(o.

T T

.O.O.)

(o.o

.0.0.0.)

Hypopus

(2

I I O

)

(l

I i

O O )

(o

I O

O O.)

(o.i

.0.0.0.)

Tritonymph

h

I 2 O

)

(l

I I

O O )

(o

I.I

.O.O.)

(O.I

.0.0.0.)

Adult

h

,I.2.O

0)

(T

T T

.O.O.)

(0

T T

,O.O.)

(O.I

0.0.0.)

The chaetotactic formula for A. f arris given in Griffiths (1964) should be amended
to agree with the above. This involves three alterations, namely:

(a) Protonymph setae, leg IV, tarsus I replace figure 7 with an 8. I am now
satisfied that spine s is present as a minute structure.

(b) Hypopal seta, leg I, tarsus I replace figure 10 with a 9. Seta ba is NOT present.

(c) Hypopal solenidia, leg I, tarsus I replace figure i with a 2. Solenidion oj 3 is
present. In Griffiths (1964) it was mistakenly termed seta ba.

When the above corrections are taken into account, the chaetotactic formula for
this species is identical with the formula for those species of the A. siro complex
which have a motile hypopus. Also, the position of the respective setae and solenidia
and, in general, their length and the degree of pectination of the setae are remarkably
similar. However, as already mentioned, in adult specimens the shape of a>i is
distinct.

BIOLOGY AND HYBRIDIZATION EXPERIMENTS. This species has been reared suc-
cessfully in the laboratory for a period of four years. Single pair matings of virgin
adults have been made to determine egg output and hatch. The pairs were reared
in individual cells under the same physical conditions as the stock cultures namely,
20C and 90 per cent relative humidity, with dried yeast powder as food.

Twenty-five pairs laid 2044 eggs over a 15 day period. Maximum for any one pair

D. A. GRIFFITHS

18

19

20 1

FIGS. 1 8-2 1. Acarus nidicolous sp.n. Hypopal leg chaetotaxy (excluding trochanter
segment). 18, left leg I. 19, left leg II. 20, left leg III. 21, left leg IV.

SYSTEMATIC STUDY OF THE GENUS ACARUS 101

was 137 and minimum 40, with an average per pair of 82. Percentage hatch was

85-9%.

The same technique as described above has been employed to cross A. nidicolous
with the three species of the A . siro complex. Twenty initial crosses, in each direc-
tion were completed. Viable eggs were obtained only from the cross A. immobilis
. nidicolous $. The results (Table i) show that only 4 pairs produced eggs and

TABLE i

Details of the four viable interspecific crosses

A . immobilis $ + A nidicolous $

No. of offspring
reaching maturity*

-L CbU

-fl ULLLLJ^L V^J. ^tiCiO

/o

number

per cross

hatch

??'s

(J(J'S

i

76

47-3

4

I

2

58

29.9

ii

5

3

45

13-3

2

2

4

85

21. 1

O

8

* Fj crosses made with these adults failed to produce an F 2 generation.

indicate the presence of a reproductive isolating mechanism at the gamete level
since each fertile pair produced very few offspring. With these offspring, 10 FI
sibling and non-sibling crosses were performed. Seven crosses produced small
numbers of eggs but none hatched.

DISTRIBUTION, HABITAT AND LOCALITY. Known from animal nests in Great
Britain, namely; in unidentified small rodent nest, grounds of Pest Infestation Lab.,
Slough, Bucks., March 1964; breeding nest of hedgehog (Erinaceus europaeus L.),
Pest Infestation Lab., Slough, Bucks., April 1964; squirrel drey, (Sciurus carolinensis) ,
Wytham, near Oxford, April 1964; nest of hooded crow (Corvus comix), Eileen mor,
Inverpolly, Scotland, June 1964; mole (Talpa europaea L.) breeding nest, Foxfield,
Wilts., January 1964 ; mole nest, Sittard. December 1911. Also from cheddar cheese,
Slough, Bucks., 1967.

One c from nest of Neotoma fuscipes, California, U.S.A., has also been assigned
to this species (see p. 96).

MATERIAL DEPOSITED IN NATIONAL MUSEUMS

(i) British Museum (Natural History).

HOLOTYPE : $. England, Bucks., Slough; nest of unidentified small mammal
in grass bank; March, 1964. Coll. D. A. Griffiths. BM(NH) registration number
1969.38

PARATYPES : Three <$<$s, 6$$ and 3 hypopi, on n slides. BM(NM) registration
numbers 1969.39 to 1969.41 inclusive. Data as for holotype.

OTHER MATERIAL : Three $<$, i $, 16 hypopi, on 5 slides. England, Oxfordshire,
Wytham woods; Sciurus carolinensis nest (Grey squirrel's drey); April, 1964. Coll.

B*

102 D. A. GRIFFITHS

D. A. Griffiths. BM(NH) reg. numbers 1969.44 and 45 inclusive. One $, Scotland.
Eileen mor, Inverpolly; Corvus cornix nest; June, 1964. Coll. G. E. Woodroffe.
BM(NH) reg. number 1969.46.

One <J, 2 hypopi, on 2 slides. England, Wilts., Foxfield; Talpa europaea breeding
nest; Jan., 1964. BM(NH) reg. numbers 1969.47 and 48.

One $, i hypopus, on 2 slides. England, Bucks., Slough; Erinaceus europaens breed-
ing nest, April 1964. Coll. D. A. Griffiths. BM(NH) reg. numbers 1969.42 and 43.

(ii) United States National Museum (Washington).
PARATYPES : Four <$<$, 6 $$ and 3 hypopi, on 8 slides. Data as for holotype.

OTHER MATERIAL : USA, California, Alameda Co., Wildcat Canyon; Neotoma
fuscipes nest, June ist, 1964. Collrs. Chaniotis & Stumpf.

(iii) Institute of Acarology, Wooster, Ohio, U.S.A.

PARATYPES : Four <&'s, 3 $$ and 2 hypopi, on 7 slides. Data as for holotype.
(iv) Rijksmuseum van Naturlijke Historic, Leiden, Holland.

OTHER MATERIAL : One slide bearing eleven hypopi. For data see item 2, page 114.

(iii) Acarus macrocoryne sp.n.

Previously, (Griffiths, 1964, p. 456), I referred this taxon to the category 'speci-
mens incertae sedis' '. The species is represented by two populations collected near
Santa Fe", New Mexico, U.S.A. in 1952 and 1953. They were taken from two separ-
ate but similar habitats namely, the nests of the small mammals Perognathus flavus
and Dipodomys spectabilis.

The mites from the P. flavus nest (36 specimens on one slide) are all very small and
have short tarsi. The material from the D. spectabilis nest consists of specimens with
similar short cylindrical tarsi together with others in which the tarsi are much longer,
tapering towards the extremity (Text-figures 22, 23 and 24). Hysterosomal setae,
sc e, dz, dz and ip of long tarsi specimens also tend to be longer than the corresponding
setae in the respective sexes of short tarsi specimens. The extremes of variation in
length of hysterosomal setae shown by this population is illustrated in Text-fig. 5.

Since it has been shown (Griffiths, 1966) that tarsal length and shape can be in-
fluenced by varying the nutritional value of the food given to pre-adult stages, and
because Text-figures 2 and 3 show that similar variations in setal lengths occur
in Acarus siro sensu stricto, it is considered that the forms within the D. spectabilis
nest population represent intraspecific variation.

MALE AND FEMALE (Text-figures 22 to 26 inclusive). Distribution and number of
setae on the idiosoma is normal. Setae he, d\ and la appear to be smooth in some
" short tarsi " specimens, remainder are pectinate. Setae of " long tarsi " speci-
mens with slightly more pectinations than " short tarsi " specimens. Because of the
variation in length exhibited by the dorsal setae of the idiosoma, the lengths of these
setae relative to body length is not given, but are adequately illustrated in the
accompanying text-figures. In general, the lengths of the dorsal setae dz, dz and d$

SYSTEMATIC STUDY OF THE GENUS ACARUS

103

may be used with discretion to distinguish this species from members of the A . siro
complex. Where populations are represented by single specimens or a few individ-
uals this character may not in itself be conclusive.

22

25 M

FIGS. 22-24. Acarus macrocoryne sp.n. Leg I chaetotaxy and tarsal length variation.
22, a " long tarsus " specimen ex D. spectabilis nest. 23, a " short tarsus " specimen
from the same population. 24, leg I of specimen from a population found in the nest of
Perognathus flavus. The whole population has very short tarsi.

Remaining features of the idiosoma not significantly different from those of the
A. siro complex.

Genu and femur of leg I of the male much enlarged as usual but the projections

io 4

D. A. GRIFFITHS

on the genu are reduced to one large tooth and sometimes a minute sub-tending
tooth, compared with the two large and one small projection found on genu I of the
males of Acarus species described previously.

LEG CHAETOTAXY (Text-figures 22 to 24). Position, length and, where applicable,
the degree of pectination of setae not significantly different from that described for
the A. siro complex.

25

FIGS. 25, 26. Acarus macrocoryne sp.n. dorsum. 25, $. 26,

Solenidion on on tarsus I and II is long, slightly curved, expanding towards the
apex into a distinct head. The internal margin of the head is more expanded than
the external margin (PI. II, Fig. 4).

The famulus is minute and difficult to discern even under oil-immersion phase-
contrast microscopy.

MATERIAL DEPOSITED IN NATIONAL MUSEUMS.
(i) United States, Washington.

HOLOTYPE : on slide with data as given below I select the male specimen
which bears long setae. The position of the holotype specimen is indicated by a
labelled diagram attached to the slide. It is the second specimen in from the right-
hand edge of the cover-slip as viewed through a stereoscopic binocular microscope and

SYSTEMATIC STUDY OF THE GENUS ACARUS 105

the second from the left as viewed through a compound microscope. There are
another 3 << and a nymph on the slide, these 4 specimens constitute paratypes.

Data : USA, New Mexico, Santa Fe; Dipodomys spectabilis nest 63A ; Feb. 2nd,
1953. Coll. H. B. Morlan.

PARATYPES : Three $$ on one slide, with same data as for holotype.
Twenty ^s and 16 $$s, on one slide. USA, New Mexico, Santa Fe ; Perognathus
flavus nest ; Feb. 29, 1952. Coll. H. B. Morlan. USNM reg. number Lot 52-11181.

(ii) British Museum (Nat. Hist.)

PARATYPES : Single <$ and 2 $$, on 2 slides. Data as for holotype. BM(NH) reg.
numbers 1969.51 and 52.

(iv) Acarus chaetoxysilos sp.n.

In Griffiths, 1964 this taxon was not given specific rank because at the time it was
known from two specimens only. These were imported into the United States of
America from Japan on Citrus sp. More recently, in a collection of Acarus material
kindly loaned by Dr. D. E. Johnston, Institute of Acarology, Ohio, U.S.A., I found
another male which was also introduced from Japan into the U.S.A. But, on this
occasion, the cargo was " Irish " potatoes and the time interval separating the two
collections is twelve years.

MALE AND FEMALE

(i) Idiosoma (Text-figures 27, 28) . General body conformation same as that des-
cribed for other Acarus species. Genu I of male bears two strong teeth and one
smaller projection as in A. nidicolous. Spine s angled forward and about half length
of tarsal claw. (See pi. IV, fig. 2 for example.)

On the idiosoma the vertical internal setae (vi) are pectinate but, except for a single
pectination on some of the scapular setae, the remaining setae are smooth, long and
whip-like, tapering so finely towards their tips that only when viewed by oil-immer-
sion phase-contrast equipment can their full length be seen. Setae hi and all setae
of the dorsum posterior to dz are exceptionally long, resembling the posterior train of
setae of A . tyrophagoides but different from this species because the internal scapulars
(sa i) are long. The terminal portion of the supra-coxal seta is also extended into a
fine tapering thread.

(ii) Legs (Text-figures 29, 30). The glabrous setae of the tarsi are also long with
fine whip-like endings. Setae of the tibia and genu are pectinate as for other species.

On tarsus I and II o>i is fairly short, evenly cylindrical from the base until just
before the apex, where it expands almost imperceptibly into a poorly defined head
(PI. Ill, fig. 2). It is similar in shape to on of A. nidicolous. The trivial name has
been derived from chaeta, hair; xy silos, shaven.

HOLOTYPE : I select as holotype the $ specimen on United States National
Museum slide number 40.21938. There is one $ paratype of the same taxon on the
same slide. Data : On Citrus sp., Japan, at Seattle, Washington, October loth,
1940. Seattle No. 9081.

PARATYPES : In addition to the $ mentioned above, there is one # (together on

io6

D. A. GRIFFITHS

same slide with a number of $$ which are not of this taxon) with the following data :
" Irish " potatoes, Japan : at Boston, February 28th, 1956; Collr. Lantz & Averill,
Lot 56 ; 3105. In the collection of the Institute of Acarology, Columbus, Ohio,
U.S.A.

27

28

100,

FIGS. 27, 28. Acarus chaetoxysilos sp.n. 27, $ dorso-lateral view. 28, $ holotype, dorsum.

SYSTEMATIC STUDY OF THE GENUS ACARUS

107

30

FIGS. 29, 30. Acarus chaetoxy silos sp.n. Leg I chaetotaxy (excluding trochanter segment).

29, <J. 3. ?

(v) Acarus calcarabellus sp.n.

MALE AND FEMALE (Text-figs. 31 to 36).

(i) Idiosoma. The general body conformation is the same as that described for
other Acarus species. However, it should be noted that in all specimens the idio-
somal setae are very short and sparsely pectinate (Text-figs. 31 and 32).

(ii) Legs. Genu I of <$ is devoid of all protruberances and the spur on femur I is
reduced so that seta vF arises very close to the tip of the spur. When viewed from
some angles it appears to arise directly from the tip. The structure of genu I and
femur I serves to distinguish the male from males of all other known species in the
genus (Text-figs. 33 and 35).

The shape of solenidion on on tarsus I and II in both sexes is distinct (PI. II, fig. 2).
It is relatively long, especially on tarsus II, and a little more slender than coi of
previously described species. Well before the apex, it expands slightly to form a
long lozenge-shaped club, best observed when the solenidion on tarsus II is viewed
laterally.

A population consisting of 3 females and a single male was given to me by Dr.
van Eyndhoven, who commented upon the unusual shape of the male spur. I took

io8

31

32

100i

FIGS. 31, 32. Acarus calcarabellus sp.n. <$. 31, $ ex cheese, dorso-lateral view. 32,

cj ex Erithracus rubecula nest.

another small population from the nest of a robin (Erithacus rubecula) where it was
in association with A. f arris.

HOLOTYPE : ,. England, Berks., Ascot ; Erithacus rubecula nest, box no. DAG/yo ;
Feb., 1966. Coll. D. A. Griffiths. BM(NH) reg. number 1969.49.

PARATYPES : the following are in my collection ; one slide bearing single $, data as
for holotype, two $$ and one $ on three slides, data Holland, 1964; ex kaas [cheese].

SYSTEMATIC STUDY OF THE GENUS ACARUS 109

6. SPECIMENS INCERTAE SEDIS

(i) Acarus population ex Citrus grandis fruit, Formosa,
imported into U.S.A., Oct. 8th 1956.

This taxon is known from two females and nine nymphs on one slide. Data ;
United States National Museum number 56-12672, on Citrus grandis fruit, Phila. Pa.
October 8th 1956, J. Freedland Coir.

The females are distinguishable from A. immobilis only in as much as firstly, the
dorsal setae d^ to d\ inclusive are longer than the corresponding setae representing the

FIGS. 33, 34. Acarus calcarabellus sp.n. Leg I chaetotaxy. 33, $; note absence of genual
processes and small femoral spur. 34, o.

extremes of variation observed in A . immobilis (Text-fig. 37) . Secondly, solenidion
oji, although similar in shape to on of A. immobilis, has a much more slender shaft.
(PI. Ill, fig. 4).

Plate I, fig. i and 2 illustrates that wi of A. immobilis can vary in thickness. It is
also true that setal length is variable. Therefore, the Formosan specimens may
represent an extension of the variability already observed in populations from Europe
and the Near East. Until such time as further populations become available from
the Far East, these specimens must remain incertae sedis.

no

D. A. GRIFFITHS

(ii) Acarus population ex cocoanuts, Brazil, imported into U.S.A., 1935.
One <$, one and a tritonymph were taken from cocoanuts imported into San
Francisco, from Brazil on July 7th, 1935. Coll. F. J. Phelan. U.S. National
Museum collection, San Francisco number 5368.

35

FIGS. 35, 36. Acarus calcarabellus <$ holotype. Leg chaetotaxy (excluding trochanter

segment). 35, leg I. 36, leg II.

FEMALE (Text-fig. 39) : Dorsal hysterosomal setae relatively long, with d% and
reaching or overlapping bases of next distal pair of setae. Scapular setae each bears
five or more sets of pectinations but all setae of the hysterosoma are glabrous with
fine endings. The supra-coxal setae are long and slender, bearing only seven
pectinations (Text-fig. 39). Solenidion o>i has a shape which closely resembles that
found in A. immobilis (PI. Ill, fig. 3).

MALE : Represented by one poor specimen. Genual processes and tibial spur

SYSTEMATIC STUDY OF THE GENUS ACARUS

38

FIGS. 37, 38. Acarus specimen, incertae sedis ex Formosan Citrus grandis fruit. 37,

lateral view. 38, leg I of same $.

100 ji
FIG. 39. Acarus specimen incertae sedis ex Brazilian cocoanut. $, dorsum.

112 D. A. GRIFFITHS

normal. Idiosomal setae as for female except d and sc e each bear two pectinations.
Only one supra-coxal seta on specimen, this bears only four, long, slender pectina-
tions. Solenidion ai\ as for female.

TAXONOMIC POSITION : This taxon combines the distinguishing characters of a
number of the species described above. Solenidion o>i is similar in shape to o>i of
A. immobilis (compare PI. I, fig. 2 with PL III, fig. 3). However, the hysterosomal
setae are too long and devoid of pectinations for it to be considered conspecific with
A. immobilis.

The glabrous nature of the hysterosomal setae puts it close to A . chaetoxysilos and
to the specimens incertae sedis ex China (see below). The long, slender, sparsely

FIG. 40. Acarus specimen incertae sedis ex Chinese water-chestnuts. $, dorso-lateral

view.

pectinated supra-coxal seta serves to distinguish it from both of these taxa. Since
all the characters cited above are known to exhibit intra-specific variation, it is con-
sidered that the taxon be placed as specimens incertae sedis until further material
becomes available.

(iii) Acarus population ex Chinese water chestnuts,
imported into U.S.A., Jan. 22nd, 1937.

The taxon is represented by one $ and $ on U.S. National Museum slide number
37-2183. One other female, not an Acarus, is also mounted on the slide. The
material was imported from China into Chicago.

The specimens are close to A. immobilis but the dorsal setae are longer and less
pectinate than those of A. immobilis populations examined to date (Compare
Text-fig. 4a with 40).

SYSTEMATIC STUDY OF THE GENUS ACARUS 113

7. RE-ATTRIBUTION OF MATERIAL LISTED AS SPECIMENS INCERTAE
SEDIS IN GRIFFITHS 1964 (pages 453 to 457).

These were catalogued under the country from which they probably originated
although some were collected at quarantine centres in the U.S.A. I shall re-list them
as first published.

(i) BRAZIL (on orange navel end). Since I am now satisfied that in A. immobilis
the dorsal idiosomal setae dz, ^3 and d may be longer than given in my original
description, I place this specimen in A . immobilis.

(ii) NEW ZEALAND. Now considered to represent hybrids of the cross A. siroxA.
immobilis.

(iii) FORMOSA. Remains as specimens incertae sedis (see p. 109).

(iv) JAPAN. Given specific rank in this paper (see p. 105).

(v) AZORES $ AND KENYA -. These were originally thought to represent intra-
specific variation in A. siro. On the evidence illustrated in Text-figs. 2 and 3, I am
now convinced that this is correct.

(vi) CHINA. Still placed as incertae sedis (see p. 112).

(vii) U.S.A., NEW MEXICO. Raised to specific level (see p. 102).

8. AN APPRAISAL OF THE ACARUS SPECIMENS CONTAINED IN THE
OUDEMANS COLLECTION, RIJKSMUSEUM VAN NATUURLIJKE HISTORIE,

LEIDEN.

I am most grateful to Dr. Van der Hammen of the Rijksmuseum van Natuurlijke
Historic, Leiden, who kindly loaned the material in the Oudemans collection apper-
taining to the genus Acarus. Namely; slides labelled Tyroglyphus f arris, Tyrogly-
phus farinae and Tyroglyphus africana. The mounting medium on many of the
slides has deteriorated but the important morphological characters of most specimens
can be resolved with phase-contrast equipment. This most interesting collection
has been examined and specimens have been determined to species level according
to the information obtained from my investigations into the genus Acarus.
(i) Material labelled Tyroglyphus f arris Ouds.

It consists of seven slides representing material collected from three localities.
The label on the left side of each of the slides has been crossed through by two pencil
lines. Buitendijk (1945), in her catalogue to Oudemans' collection, states that the
word 'fout' (error) is written on the drawings of farris, but the reason for this is
unknown. I believe that Oudemans crossed through the slide labels and wrote
'fout' on his drawings when he decided, in 1925, that farris was a race of farinae and
not a discrete species.

The left label of each slide bears the word 'type' (in ink) in Oudemans' handwriting.
Since his original description of farris is based on material which he collected at
Arnhem in 1902, I have previously selected as lectotype the single male specimen
mounted on the slide numbered 45 (Griffiths, 1962). Details of the slides are as fol-
lows: Catalogue numbers 43 and 44 :- Utrecht Sept., 1885. One chelicera and part
of a gnathosoma is mounted on slide 43, which is in poor condition. Slide 44 con-

114 D. A. GRIFFITHS

sists of seven females, all of which are identifiable as A car us J 'arris. The habitat is
not recorded. Catalogue numbers 45, 46 and 47 :- Arnhem February 1902; taken
from cheese. The cover-slip of slide 45 is crazed but the diagnostic characters of
tarsus II of the single male are plainly visible under a phase-contrast microscope
(mag. X 800). Slides numbered 46, 47 and 48 are not in very good condition. They
bear, respectively, single specimens of the tritonymph, protonymph and larva.
Catalogue number 49 :- Arnhem March 1909, Mus decumanus. The cover-slip
is crazed and the seven tritonymphs mounted on the slide have shrivelled,
(ii) Material labelled Tyroglyphus farinae L. 1758.

It consists of twenty-seven slides (catalogue numbers i to 27), representing twenty-
one collections made between 1898 and 1932 by various people. Details of these
slides are listed below in the following manner :
Slide number :

L = Inscription on left label ; since the words 'Tyroglyphus farinae L. 1758'
appears on each label, they have not been repeated each time in the
following notes.
R = Inscription on right label ; generally a record of date of collection, habitat,

location and collectors' name.

This information is followed by my determination of the specimens on the slide,
where this is possible, and remarks when necessary.

1. L. Female ds. ve.

R. Zuid Willems-varrt Brabant 9.viii.i9i8. Dr. G. Romyn prep. no. 196.
Two females both of which are A . siro.

2. L. Nph II. hypop. Kortharig ras.

R. Nest van Talpa europaea L. Sittard 12. 1911 F. Heselhaus. Eleven
motile hypopi of A. nidicolous sp.n. Oudemans' comment 'Kortharig
ras' [short-haired race] refers to the short scapular and dorsal setae
peculiar to this hypopus.

3. L. Nph II hypopialis vent.

R. Cucumis melo en sativa. Zwyndrecht 20.5.1922.

Institut v. Plantensiekten Wageningen.

The morphology of dorsal and ventral surface of this single motile hypopus is
extremely similar to that of an FI hybrid motile hypopus resulting from the cross
f arris X immobilis.

4. R. op Kass, Rykszuivel station Leiden; Dr. van Sillers Edt.

Eight specimens, comprised of a mixed population of siro and f arris.
Taken from cheese, State Dairy Produce Station, Leiden, Holland.
5 and 6.

R. Hyacinthebol, Sassenheim Febr. 1924.
F. Schoevers Wageningen don.

Slide 5 is of nine males and slide 6 bears twenty-four females. This
material represents a mixed population off arris and immobilis. Twelve
of the specimens are immobilis, six are f arris, nine appear to be hybrids
and six I cannot identify because it is impossible to obtain a lateral
view of solenidion omega i.

SYSTEMATIC STUDY OF THE GENUS ACARUS 115

7. R. Insection collectie Amsterdam, 1916. MacGillavry.

Sixteen hypopi (possibly taken from an insect specimen by Mac-
Gillavry). This slide is in a very poor condition. I can identify eight
specimens as f arris but the remainder cannot be identified.

8. L. Nph I dors. vent. lat.

R. in grutterswaren, Keuringsdienst Amsterdam. Oct., 1923.

Three protonymphs which judged on the size of spine S, are siro.
Translation of details on right label is : 'in corn-chandlers wares,
Inspection Service, Amsterdam'.

9. L. Male dors. vent, normal gebouwd ras.
R. Arnhem 27 Mei, 1898 Oudemans.

A very poor slide bearing two males which I believe are A . siro, but
the specimens are too obscured to be absolutely certain. These speci-
mens represent Oudemans' 'normal born race' in his discussion of races
(Oudemans Feb., 1924).

10. L. Protonympha dors. vent.

R. op veegsel meelmager Arnhem 2.2.1902 Oudemans.

Five protonymphs, all siro; collected by Oudemans from old flour
sweepings.

11. L. Nyph II hypop. dors, mooi exemplar 258.

R. Mus rattus L. Arnhem 8.10.1904. Oudemans.
This is a hypopus of A. f arris.

12. L. male, zeer sterk ontwikkeld ras.

R. Cavia cobdya, Arnhem Feb. 1909 Oudemans.

One male A. f arris. Translation of note on left label reads :- 'very
strongly developed race'. Oudemans (Feb., 1924) gave this male as an
example of a strong race having dark coloured legs. But, differing
from other strong races in that tarsi of legs II, III and IV were long
and slender; as long as the genu and tibia together. It is interesting
to note that Oudemans did not recognise this specimen as f arris.

13. L. nph. female, male.

R. in een gat in een appal. Arnhem 17-111-1918 Oudemans.

Three specimens too poor to be identified, together with one male
which is f arris. They were taken from a hole in an apple.

14. L. groot ras.

R. auf Kase, Bremen 16.9.1904 F. Koenike coll.

One male, one female, both f arris. Oudemans (Feb., 1924) gave
these as an example of a strong race which had shorter tarsi than the
male on slide 12.
15 and 16.

L. 15- hypopi uit Nph I gekweekt

16- Nph II uit hypopi gekweekt.
R. Berlyn 14.5.1923, Hanna Schulze.

Slide 15 bears 8 motile hypopi of A . f arris which emerged from proto-
nymphs and slide 1 6 is of 3 tritonymphs which emerged from hypopi.

n6 D. A. GRIFFITHS

But, since I am unable to distinguish between nymphs of farris and
immobilis, I cannot be certain to which species the tritonymphs belong.
17, 18 and 19.

R. Shag-tabak, Hamburg, Feb., 1928. Dr. L. Reh, visit.

Three slides, totalling fifteen females and one male; all are A. siro.
20. R. op vorkens-dormen Berlyn Aug. 1927. Dr. F. Zacher.
Twelve specimens all identifiable as A. Siro.

21 and 22.

L. Lv, Nph I, III, adulti.

R. Evotomys glareolus. Schreber Hemmelsdorper See. 10.4.1926 Erna
Mohr.

There are over one hundred and fifty specimens on slide 21, which is
not very clear. All the specimens which can be identified are A. siro.
The single hypopus on slide 22 is definitely an Acarus siro. It is the
only one in the whole collection. Presumably, these specimens were
taken from the nest of a bank vole.

23. L. Printed label ACAROTHECA I-g Aleurobius farinae (D.G.)

R. Printed label ITALICA. In farina, caseo, etc., Italia septentre. A.
Berlese.

One male and one female. The male is an Acarus farris and the
female is an Acarus siro.

Berlese used to make slides which he then sold to institutions and
private collectors. Slide 23 is undoubtedly an example of this work.
To each slide of any one species Berlese attached printed labels. The
right-hand label of each was inscribed with identical information con-
cerning the habitat of the species. Usually, this information was of a
general nature, giving the habitat in which Berlese believed the mite
was most commonly found. In this case he gave 'in farina, caseo, etc.,'.
Since it is possible the mites may have been collected from some other
habitat, this data is not necessarily true for these particular specimens.

24 and 25.

L. Nympha.

R. Steelgrove Citrus aurantium Zind Europa. Arnhem, Oudemans II,
1932.

Both slides have been re-mounted. I failed to find any specimens on
either slide!

26. L. Nph II hyp. dors. vent.

R. Osmoderma eremita Scop. Weenen Aug., 1924.
H. E. Wickmann, legit Vitzthum donavit.

One hypopus of A. farris which presumably was taken from 0.
eremita (a species of the flower-beetles).

27. R. in een buisje met Hevea parasieten; Medan, Deli Proepstation AVROS

Mei, 1918.

SYSTEMATIC STUDY OF THE GENUS ACARUS 117

Two males, one is siro, the other specimen is crushed and cannot be
identified. Translation of words on right label reads 'in a tube with
Hevea parasite, Medan, Deli Agricultural Research Station',
(iii) Material labelled Tyroglyphus africana

This consists of one slide labelled as follows :
Catalogue no. i,

L. Tyroglyphns africana (Oudms) female, dors type.
R. Platygenia barbata (Cetonide) Togo (W. Afr.)
H.Voigts lo.ix.igoi.

A single female which is undoubtedly an A earns siro, placed in
synonymy in Griffiths 1964.

9. SUMMARY

Ten species taxa are recognised ; four are described for the first time, two of which
were previously considered as specimens incertae sedis. A key to adults is provided.

Generic characters together with interspecific and intraspecific variation are
reconsidered. An appraisal is made of the Acarus specimens contained in the A. C.
Oudemans collection, Rijksmuseum van Natuurlijke Historic, Leiden.

10. ACKNOWLEDGEMENTS

I am most grateful to the many acarologists who were sufficiently interested in my
first revision of A earns to send me material which they considered to be unusual, and
which I have subsequently used as the basis of this new study. I am particularly
indebted to Dr. E. W. Baker, United States National Museum, Washington, Dr.
E. van Eyndhoven, Zoological Museum, Amsterdam, Dr. D. Furman, Division of
Parasitology, University of California, Dr. D. Johnston, Institute of Acarology, Ohio
State University, and Dr. W. Knulle, Acarology Laboratory, Department of Zoology
and Entomology, Ohio Agricultural Research and Development Centre. I am most
grateful to Dr. L. van der Hammen, Rijksmuseum van Natuurlijke Historic, Leiden,
for the loan of Oudemans' material.

ii. REFERENCES

BUITENDIJK, A. M. 1945. Vorloopige catalogus van de acari in de collectie Oudemans. Zoo/.
Meded. 24, p. 69.

GRANDJEAN, F. 1939. La chaetotaxie des pattes chez les Acaridae. Bull. Soc. zool. Fr. 64,
p. 50.

GRIFFITHS, D. A. 1962. The flour mite Acarus siro L., 1758 as a species complex. Nature, 196,

No. 4857, p. 908.
- 1964. A revision of the genus Acarus L., 1759 (Acaridae, Acarina). Bull. Br. Mus. Nat.

Hist. (Zool.) 11, (6), p. 415.

1966. Nutrition as a factor influencing hypopus formation in the Acarus siro species com-
plex (Acarina, Acaridae). /. Stored Prod. Res. 7, p. 325.

HUGHES, A. M. 1957. A new species of Acarus (A. gracilis sp.n.) belonging to the family Tyro-
glyphidae Donnadieu 1868 (Acarina). Ann. Mag. nat. Hist. 10, p. 753.

OUDEMANS, A. C. 1924. On races of Tyroglyphus farinae L., 1758. Tijdschr. Ent. 67.

il8 D. A. GRIFFITHS

VOLGIN, V. I. 1965. Two new species of Acaroid mites (Acarina, Acaroidea) of the fauna of the

U.S.S.R. Ent. Obozr. 44 (i), p. 203.
ZACHVATKIN, A. A. 1941. [Tyroglyphoidea (Acari)]. Fauna SSSR No. 28, 475 pp. (In Russian,

English translation by RATCLIFFE, A. and HUGHES, A. M., Washington, D.C. : American

Institute of Biological Sciences).

Dr. D. A. GRIFFITHS,

A.R.C. PEST INFESTATION LABORATORY,

LONDON ROAD,

SLOUGH, BUCKS.

PLATE i

Lateral view of solenidion omegai (wi) situated on tarsus II.
FIGS, i & 2. Acarus immobilis Griffiths.
FIG. 3. Acarus gracilis Hughes.
FIG. 4. Acarus siro L.

X2250

Bull. BY. Mus. nat. Hist. (Zool.) 19, 2

PLATE i

C*

PLATE 2
Lateral view of solenidion omegai (o>i) situated on tarsus II. x 2250

FIG. i. A earns f arris (Oudemans)

FIG. 2. Acarus calcarabellus sp.n.

FIG. 3. Acarus nidicolous sp.n.

FIG. 4. Acarus macrocoryne sp.n.

Bull. Br. Mus. nat. Hist. (Zool.) 19, 2

PLATE 2

PLATE 3

PLATE 3
Lateral view of solenidion omegai (on) situated on tarsus II. x 2250

FIG. i. Acarus specimen incertae sedis ex Chinese water-chestnuts.

FIG. 2. Acarus chaetoxysilos sp.n.

FIG. 3. Acarus specimen incertae sedis ex Brazilian cocoanuts.

FIG. 4. Acarus specimen incertae sedis ex Formosan Citrus grandis fruit.

Bull. Br. Mus. nat. Hist. (Zool.) 19, 2

PLATE 3

PLATE 4

FIG. i. Acarus siro L. $. Apex of tarsus I showing large spine 5 curving away from pulvilus.

FIG. 2. Acarus immobilis Griffiths. $. Apex of tarsus I showing small spine s curving towards
pulvilus.

FIG. 3. Acarus siro L. A scanning electron micrograph of Grandjean's organ showing its
structure and position. x 3200

FIG. 4. Acarus siro L. A scanning electron micrograph of the supra-coxal seta. x 3200

Bull. Br. Mus. nat. Hist. (Zool.) 19, 2

PLATE 4

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

<4L

SKULL AND SWIMBLADDEfe
CONNECTIONS IN FISHES OF
THE FAMILY MEGALOPIDAE

P. H. GREENWOOD

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 3

LONDON : 1970

SKULL AND SWIMBLADDER CONNECTIONS
IN FISHES OF THE FAMILY MEGALOPIDAE

BY

PETER HUMPHRY GREENWOOD

Pp. 119-135; 3 plates, 2 text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 3

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is issued
in Jive series corresponding to the Departments of the
Museum, and an Historical series.

Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 3 of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 1 8 March, 1970 Price

SKULL AND SWIMBLADDER CONNECTIONS IN
FISHES OF THE FAMILY MEGALOPIDAE

By P. H. GREENWOOD

CONTENTS

Page
INTRODUCTION .......... 121

THE LINKAGE BETWEEN SKULL AND SWIMBLADDER IN Megdldps Cyprinoides 122

The soft tissues ......... 122

Otic region of the skull . . . . . . . .124

SKULL AND SWIMBLADDER LINKAGE IN Tarpon atlanticus . . . 125

The soft tissues ......... 125

Otic region of the skull . . . . . . . . 127

THE SACCULAR OTOLITH IN Megalops AND Tarpon . . . . 128

SWIMBLADDER-SKULL CONNECTIONS IN OTHER ELOPIFORM FISHES . . 130

DISCUSSION ........... 131

The status of the genera Megalops and Tarpon . . . . 131

Relationships between the Elopomorpha and Clupeomorpha . . 132

SUMMARY ........... 133

ACKNOWLEDGEMENTS . . . . . . . . .133

APPENDIX: Notes on the caudal fin skeleton in Megalops and Tarpon . 133

REFERENCES . . . . . . . . . . . 134

LIST OF ABBREVIATIONS USED IN THE FIGURES ..... 135

INTRODUCTION

THE only detailed account of connections between the swimbladder and the otic
region in megalopid fishes is that given by de Beaufort (1909), who described the
Indo-Pacific Oxeye, Megalops cyprinoides (Brouss.). The closely related Atlantic
species Tarpon atlanticus (C. and V.) is mentioned in de Beaufort's paper solely
by the quotation of Hyrtl's (1855) brief description of the swimbladder. No
comparative study, or comment, was made by de Beaufort.

Hyrtl's remarks are rather vague and would not lead one to suspect the existence
of any marked interspecific differences in swimbladder anatomy. Later authors,
as did de Beaufort, have tacitly assumed that the swimbladder and its otic connec-
tions are alike in both species.

Externally, the two species are very similar, and there is still a great deal of
uncertainty about the generic status of the Atlantic Tarpon. Some authors treat
it as a monotypic genus, Tarpon atlanticus, while others refer it to the genus Megalops
(see Hollister, 1939, and Hildebrand, 1963, for further discussion). This problem
will be considered anew in the light of the marked interspecific differences which I
have found in many details of the skull-swimbladder linkage, and in the otic region
of the skull.

122 P. H. GREENWOOD

Other taxonomic problems, but particularly the relationship between the Elopo-
morpha and the Clupeomorpha, will also be discussed (see Greenwood et al., 1966).

THE LINKAGE BETWEEN SKULL AND SWIMBLADDER IN
MEGALOPS CYPRINOIDES

In general, my investigation of the swimbladder in this species confirms the
description given by de Beaufort (1909). The situation is redescribed here in some
detail mainly to serve as a basis for comparison with Tarpon atlanticus (see pp. 125-8).

The soft tissues. There is one aspect where de Beaufort and I differ, and that is
in our interpretation of the outer thickened tunic which covers the precoelomic
extension of the swimbladder.

The tunic is identified by de Beaufort as swimbladder wall which has become
greatly thickened and differentiated into an anterior, transversely aligned, " U "-
shaped portion linking the periotic bullae with the longer egg-shaped part running
forward from the pneumatic duct (see de Beaufort, 1909, p. 529 ; fig. 2, and plate 29,
fig. 3) . de Beaufort does not identify precisely which of the swimbladder tunics is
thickened, but his figures and description would seem to imply that it is the tunica
externa. Certainly his figure 2 shows the egg-shaped portion as continuous with
the outer wall of the main bladder. He is quite explicit in identifying the tissue
which enters the periotic bullae (the " U "-shaped part remaining outside) as tunica
interna.

I find, however, that the thickened tissue of the egg-shaped portion is actually
continuous with the much thinner peritoneal membrane covering the ventral and
lateral aspects of the main swimbladder. Although the thick and thin parts of the
peritoneum are continuous, there is no obvious gradient in thickness between them.
Instead, there is a clear-cut line of demarcation running obliquely upwards and
posteriorly from a point just anterior to the pneumatic duct. Without special care
in dissection, the peritoneum parts cleanly along this line. The thin tissue then
collapses and adheres closely to the swimbladder, while the thick walled egg-shaped
portion retains its shape. In this condition the egg-shaped part certainly does not
appear to be part of the peritoneal lining, but nor does it seem to be part of the swim-
bladder, which can be seen lying within it.

From a careful dissection of the tissues involved, and considering that the egg-
shaped portion is continuous with the peritoneum, I find it difficult to interpret the
precoelomic, thick-walled tunic as being derived from intrinsic swimbladder tissue.
Instead, I would interpret it as being derived from the peritoneum; in effect, an
evagination of the transverse septum. Thus, the precoelomic portion of the swim-
bladder, like the coelomic part, lies in a peritoneal cover.

As far as can be determined from gross dissection and microscopic studies of
sections through a juvenile fish (ca. 5 cms standard length), the tunica externa (sensu
Fahlen 1967) of the swimbladder is present throughout its precoelomic extension,
apparently including the vesicles which lie within the periotic bullae. Further
histological study on better preserved material is, however, necessary to confirm
these points.

The beginning of the precoelomic peritoneal conduit for the swimbladder is taken

FISHES OF THE FAMILY MEGALOPIDAE 123

to be the line of demarcation between thick and thin tissue (see above); dorsally,
the thick tissue is closely associated with the 6th and 7th vertebrae. As de Beaufort
noted, the main part of the conduit is approximately ovoid in form. Anteriorly
(below the ist or second vertebra) it bifurcates to form the " U "-shaped arms
associated with the base of the skull, and extending upwards to the openings of the
periotic bullae. At the point of bifurcation, the lumen of the ovoid body is deeply
constricted, but it immediately expands in the base of the " U "-shaped arms.

Superficially, the egg-shaped body has, in preserved material, a silvery-white
colour. Under moderate magnification the tissue appears to be made up principally
of horizontally aligned, elongate, parallel and slightly sinuous fibres. Dorsally, this
part of the conduit is firmly attached to the ventral and ventrolateral face of the
vertebral column. The tissue is thickest along this line of attachment, and it
surrounds the basal portion of several ribs which, in consequence, seem to emerge
through the wall of the conduit. The point at which the egg-shaped body bifurcates
to form the " U "-shaped part is clearly defined externally as well as internally.
The " U "-shaped part has very thick walls, and in preserved specimens is yellowish-
white in colour. Each arm of the " U " is only half a hollow cylinder but the
transversely aligned basal part is a complete cylinder. The arms lie somewhat
obliquely forward; the edges of each hemicylinder are closely applied to the skull.
Anteriorly and dorso-anteriorly each arm is attached to the intercalar of its side,
anteroventrally to the basioccipital, and posteriorly to the exoccipital. The cylindrical
transverse part of the " U " is firmly attached to the base of the centrum fused with
the basioccipital.

The precoelomic extension of the swimbladder is variously modified. Posteriorly,
it differs but little from the region immediately behind the pneumatic duct, and like
that region has a non-alveolate inner surface. (For a description of the alveolate
main bladder, see de Beaufort, op. cit.}. Further forward, the cross-section narrows
rather rapidly until, at a point about half-way between the anterior bifurcation and
the pneumatic duct, the bore is reduced to a narrow duct. As the bore narrows
the wall thickens so that over its anterior third the swimbladder has a characteristic
appearance, rather like a short section of plastic tubing. Throughout its passage
in the peritoneal conduit, the swimbladder is loosely attached to the conduit walls
by thin strands of connective tissue.

The thick-walled tube fills the narrow lumen which connects the " U "- and egg-
shaped portions of the conduit. It penetrates for a short distance into the base of
the " U ", and then bifurcates into two, thin-walled tubes, each of which completely
fills the space between the skull and peritoneal wall in its arm of the " U ". On
entering the periotic cavity each tube again expands to form an even thinner-walled
vesicle, the periotic vesicle. The outer wall of the periotic vesicle is closely applied
to the inner walls of its bony periotic bulla, with diverticula extending into the various
recesses of the bulla's inner wall (see below) .

Unfortunately, my material is not suitably preserved for detailed histological
examination. Thus I cannot determine definitely if both swimbladder tunics
contribute to the thick-walled tubular part; nor is it certain whether the periotic
vesicle is formed from both tunica interna and externa or from the interna alone.

124 P. H. GREENWOOD

de Beaufort believed that only the tunica interna lined the periotic region, but then
he apparently thought the tunica externa formed the " U "-shaped part of the conduit
and therefore remained outside the bulla.

Otic region of the skull. Although the periotic swimbladder vesicles are completely
surrounded by bone, they are nevertheless strictly extracranial in position, and lie
alongside the lateral skull wall.

Each vesicle lies partly in a horizontally aligned, almond-shaped depression
formed in the outer wall of the otic region (i.e. in the prootic, basioccipital and exocci-
pital bones). The greater volume of the chamber is, however, provided by the
concave inner face of a shield-like extension from the intercalar. The shield reaches
outwards and lateral to the depression in the otic wall, and forms the lateral and
ventral walls of the periotic bulla. In other words, this hypertrophied portion of
the intercalar forms a false lateral wall to the otic region (see Plate, I, Fig. i). The
space between it and the true lateral skull wall (formed by prootic, basi- and exocci-
pital bones) being occupied by the swimbladder vesicle. Ridewood's statement
(1904, p. 62) that " Megalops has a cavity in its opisthotic ..." (italics mine) is
inaccurate.

There are three, inwardly directed diverticula from the medial (i.e. skull) wall
of each chamber. From the anterior part there is, ventrally, a broad, deep pit into
the floor of the prootic bridge over the myodome. The opening into this diverticu-
lum lies immediately below the anterior tip of the saccular recess in the prootic; the
diverticulum of each side is separated in the midline by a thin bony wall. From the
posterior part of the chamber there is a dorsomedially directed pit in the exoccipital,
and a ventromedially orientated one into the basioccipital. In a dry skull, the
ventromedial pits from each side meet one another in the midline of the basioccipital,
and are continuous, forming a narrow transverse tunnel underlying the posterior
part of the saccular recess. In an alizarin preparation, however, there is a thin bony
partition between the medial tips of the diverticula. A foramen (situated immediately
below the opening for the hyomandibular branch of the facial nerve) opens into the
anterior part of the periotic bulla.

de Beaufort's description of the periotic chamber implies that it is divided into
anterior and posterior parts, but none of the specimens that I have examined clearly
shows any such division.

The bones forming the inner wall of the chamber (i.e. the skull wall) are thick and
without foramina or fenestrae. Consequently the periotic swimbladder vesicles are
nowhere in direct contact with the sacculus or utriculus (cf. the condition in Clupeo-
morpha where there is a direct contact with the utriculus; see Wohlfahrt, 1936).

The myodome in M. cyprinoides extends below almost the entire length of the
saccular cavity, and the periotic bulla is entirely above and lateral to the myodome.

The intercalar (opisthotic of Ridewood, op. cit.} is a large bone of complex shape.
For descriptive purposes three parts can be recognized; a relatively small dorso-
posteriorly situated basal portion, an expanded lateral shield-like part, and an
anterodorsally placed saddle-like wing (see Plate i, Figs. 2 A & B).

The basal part is roughly ovoid in shape, and of rather spongey texture ; it has firm
sutural unions with the exoccipital medially and the pterotic dorsally.

FISHES OF THE FAMILY MEGALOPIDAE 125

From the ventrolateral aspect of the basal part there is a short, broad, pedicel
connecting it with the expansive, medially concave lateral shield. Posteriorly, the
shield is suturally joined with the exoccipital (where it forms the floor to the foramen
for the head vein and glossopharyngeal nerve) ; ventrally there is a suture with the
basioccipital, and anteriorly one with the prootic.

The relatively deep dorsoposterior margin of the lateral shield curves medially
and somewhat ventrally before expanding to form the broad, saddle-shaped wing.
This portion provides the posterior part of the dorsal roof to the periotic bulla.
Medially, the hind part of this roofing wing extends upwards and surrounds the
basal portion (to which it is very closely applied) ; it also forms a backwardly directed,
almost completely closed groove lying lateral to the basal ovoid (a narrow diver-
ticulum from the periotic vesicle occupies at least part of this groove). The main
body of the saddle expands ventrolaterally, and overlaps the upper margin of the
lateral shield. In lateral view, the saddle-like wing appears to be an independent
crescentic bone closely applied to both the basal part of the intercalar and to its
lateral shield-like expansion ; the impression is enhanced by the fact that, posteriorly,
the crescentic element is suturally joined to a dorsal projection from the shield.
However, when the intercalar is isolated from the surrounding skull bones its integrity
is at once apparent.

Ridewood (op. cit., fig. I2B, and p. 44) misinterprets certain relationships between
the intercalar and its neighbouring bones. He describes the posterosuperior part
as . . . small and wedged in between the main part of the exoccipital and the
part of the bone that forms the posterior border of the subtemporal fossa". The
bone which he identifies as part of the exoccipital is, in fact, the posterior, upwardly
directed groove formed by the saddle-like wing of the intercalar (cf. Plate 2, with
Ridewood's fig. I2B).

There are no obvious ontogenetic changes in the morphology of the skull wall in the
otic region, such as are seen in Tarpon atlanticus (see page 128).

SKULL AND SWIMBLADDER LINKAGE IN TARPON ATLANTICUS

In its basic features, the swimbladder-skull connection in this species resembles
that of Megalops cyprinoides. The chief differences lie in the absence of any contribu-
tion from the intercalar to the wall of the periotic chamber (which is formed entirely
from the peritoneal tissue), a simpler form of precoelomic peritoneal conduit, and the
absence of a thick-walled section in the precoelomic swimbladder extension.

The soft tissues. The precoelomic prolongation of the swimbladder originates
immediately anterior to the pneumatic duct. It is of slightly narrower diameter
than the main bladder at that point, and is composed of very thin tissue. It almost
fills the peritoneal conduit (see below), but, except for a short distance posteriorly
in the midline, is free from the conduit walls. The lumen of the swimbladder
decreases but slightly as it runs forward. At a point below the first vertebra the
bladder divides into two, each arm then passing forward within the periotic chamber.
These periotic vesicles are, like those of Megalops, very thin-walled; unlike that
species there is, in Tarpon, no thick-walled tube immediately preceding the point of
bifurcation.

126

P. H. GREENWOOD

The main (ie. coelomic) part of the swimbladder, like that of Megalops cyprinoides,
is alveolate; the description given by de Beaufort (op. cit.) for that species could
well apply to Tarpon atlanticus.

The peritoneal conduit surrounding the precoelomic swimbladder extends from
about the I3th vertebra to the ist vertebra. It is closely applied to the ventrolateral
face of the vertebral column and is perforated by the I2th and J_3th ribs posteriorly
and the first five or six ribs anteriorly. Throughout its length the conduit is of
almost uniform diameter, and its walls are of uniform thickness. Unlike Megalops,
there is no deep ventral groove between the main body of the conduit and the otic
arms, which in Tarpon atlanticus are just hemicylindrical lateral extensions of the
median tube.

par

epi

fr

1st v

psph-

FIG. i. Tarpon atlanticus. Otic region of skull (right) in a specimen 91 - o mm S.L. to show the

periotic vesicle (p.w.) Drawn by Sharon Lesure

The otic arms arise at about the level of the first vertebra and run forwards and
upwards at a slight angle (see text fig. i). The ventral and median part of the main
tube is, at the point of otic arm divergence, closely applied to the basioccipital.

Each hemicylindrical otic arm extends forward almost to the anterior margin
of the prootic. Dorsally it is attached to the lower margin of the exoccipital, to the
lower limb of the intercalar (see page 127), along the ventral part of the upper inter-
calar limb, and along the prootic (following a line level with the intercalar-prootic
bridge). Ventrally it is attached to the ascending limb of the parasphenoid (its
posterior margin), and to the basioccipital along the parasphenoid-basioccipital
suture; the margin of the tube skirts the base of the first infrapharyngobranchial
where the latter articulates with the ascending parasphenoid limb.

FISHES OF THE FAMILY MEGALOPIDAE 127

Otic region of the skull. From the description given above it will be seen that
Tarpon atlanticus, unlike Megalops cyprinoides, has the entire wall of the periotic
cavity formed from peritoneal tissue. The intercalar in Tarpon serves only as a
point of attachment and, to a certain extent, suspension for the peritoneal tissue.
In Megalops it will be recalled, the peritoneal wall is replaced by a bony plate
developed from the lower limb of the intercalar.

The medial wall of the periotic chamber (which is, of course, the lateral skull wall)
is formed chiefly by the basioccipital and prootic with a small contribution pos-
teriorly from the exoccipital (Plate 3, Fig. i). The line of contact between the
prootic and basioccipital runs obliquely upwards and backwards; near its antero-
dorsal tip, the basioccipital margin diverges slightly from that of the prootic, and
a small triangular fenestra is formed. It opens directly into the saccular recess.
The fenestra is most obvious in small specimens but it is still present in the largest
skull examined (21 cms long). In all specimens throughout the size range available
(skull lengths 4-3-21-0 cms) certain areas of the bony inner wall to the periotic
cavity are thinner and more flexible than are other areas. The flexible zone is
usually centred in the basioccipital, but may include the posterior part of the prootic
as well. This condition contrasts markedly with that in Megalops where there is
neither a fenestra nor thinning of the otic bones. In fact, the otic bones in Megalops
are noticeably thicker and denser than those of Tarpon.

Pronounced ontogenetic changes take place in the morphology of the otic region,
particularly with regard to the relative position of the saccular cavity. These changes
are discussed below.

The intercalar is a large bone of somewhat complex form (Plate 3, Fig. 2A & B),
but relatively simpler than that of Megalops (see page 124) . Again, three regions can
be recognized. The ovoid basal portion fits deeply into the exoccipital; the pit in
which it is accommodated lies immediately above the foramen for the vagus nerve.
Both the intercalar base and the walls of the pit are deeply ridged, the ridges from
the two bones interdigitating.

Two limbs arise from the basal portion. One (the broader and more substantial)
is crescentic in lateral outline. Its flattened horizontal portion extends forward to
meet the prootic ; the vertical portion curves dorsally and articulates with the pterotic.
Medially the stout pedicel connecting this limb with the base is firmly articulated
with the exoccipital.

The other limb lies below the crescentic one, is less substantial and, apart from its
curved posterior margin (which serves as a pedicel) is almost horizontally aligned;
it is slightly longer than the crescentic limb (see Plate 3, Fig. i).

In fishes with a skull length of < 20 mm, the lower arm is well-separated laterally
from the upper crescent, but in larger individuals (skull lengths 55 mm and above)
the two arms are in contact for most of their lengths (cf. text fig. i, and Plate 3,
Fig. i).

Four levator muscles from the 2nd-4th gill arches originate on the ventral and
posterolateral aspects of the lower intercalar limb; the suprapharyngobranchial of
the first gill arch articulates near the anterior tip.

Comparing the intercalar of Tarpon with that of Megalops it is clear that the

128 P. H. GREENWOOD

expanded, shield-like part in Megalops corresponds to the lower limb in Tarpon,
and that the dorsal, crescentic limb in that species is the homologue of the saddle-
shaped wing in Megalops. Because the upper crescent in Tarpon has a more definite
pedicel from the basal part, the base itself has no contact with the pterotic (that
contact being effected through the pedicel).

Ontogenetic changes in the shape of the lateral skull wall of the otic region were
mentioned earlier. The bones involved are the basioccipital and prootic, especially
the former. In all specimens less than 23 cms standard length (skull length 4-3 cms)
the wall of the otic region (which is the inner wall of the periotic chamber) is slightly
convex, bulging a little into the periotic chamber. The large saccular otolith is
visible through the thin bone; its ventral margin lies slightly above the level of the
basioccipital-parasphenoid suture .

In two larger skulls (one 13 cms long from a fish 78-2 cms S.L., the other 21 cms
long from a fish of unknown length) the lower part of the otic wall (particularly the
basioccipital) is deeply concave (see Plate 3, Fig. i). This once convex area now
has the form of an almond-shaped depression. The depressed area lies principally
in the basioccipital, below the ventral margin of the sacculus, which now lies in the
same horizontal as the lowest point on the ventral intercalar arm. In the 13 cm
skull there are two depressions (one in the basioccipital, the other in the prootic)
separated by a narrow ridge (see Plate 3, Fig. i). The posterior, almond-shaped
depression in the basioccipital is the larger of the two. In the 21 cm skull, the
depressions are confluent.

I have been able to examine only one skull of an intermediate size (5-5 cms, from
a fish 31 cms S.L.). Here there is a noticeable concavity in the wall of the basi-
occipital, posterior and a little ventral to the saccular recess. There is also a shallow
indentation of the prootic (just anterior to its suture with the basioccipital) below
and in front of the saccular recess. Neither depression is as deep as those in the
13 cm skull, and the lower margin of the sacculus is still ventral in position.

Without more material from the upper part of the "intermediate" size range it is
impossible to tell what growth changes are involved in producing the very charac-
teristic form of the otic wall in " large " fishes. The dorsal migration of the saccular
recess is particularly intriguing. Measurements made on the otic region in fishes
of all available sizes do not indicate a marked differential growth anteroposteriorly
or dorsoventrally in any part of the bones involved. There is, however, a suggestion
that certain areas in these bones may grow inwards, thereby forming concavities
laterally and, as the inner surfaces of the bones extend medially towards one another,
the saccular recesses are forced upwards.

THE SACCULAR OTOLITH IN MEGALOPS AND TARPON

The sagitta in Megalops cyprinoides and Tarpon atlanticus differs in a number of
features (see text fig. 2).

In M. cyprinoides the anterior outline is acute, with the ventral margin sloping
gently upwards; in T. atlanticus it is relatively obtuse, with the ventral margin
curved abruptly upwards near the tip. There is a distinct antirostrum and excisura
in M. cyprinoides but neither is present in T. atlanticus. In that species, the dorsal

FISHES OF THE FAMILY MEGA.LOPIDAE

129

outline posterior to the ostium is domed, whereas in M. cyprinoides the outline is but
slightly curved.

The shape of the sulcus (which is somewhat constricted near its midpoint, has a
broad and rounded cauda, and an expansive ostium) is alike in both species.

2mm

FIG. 2. Saccular otolith (sagitta) of A. Megalops cyprinoides (from a fish of unknown
length, B. Tarpon atlanticus (from a fish 31 cms S.L.). c. Tarpon atlanticus (from a
large specimen of unknown length). Drawn by Sharon Lesure

No details of possible growth changes could be determined from the M . cyprinoides
sagittae available since both pairs are the same size (5-5 mm long). The two pairs of
T. atlanticus sagittae show that in larger otoliths the dorsal outline is more produced,
and the margin of the sulcus is raised and slightly thickened. The smaller sagitta

130 P. H. GREENWOOD

(10-0 mm long) is from a fish 31-0 cm S.L.; the larger otolith is 21-0 mm long but
no record is available of the fish's size.

SWIMBLADDER-SKULL CONNECTIONS IN OTHER ELOPIFORM FISHES

An intimate and complex connection between the otic region and swimbladder
exists only in the Megalopidae although in at least some Elopidae a simpler linkage
is developed.

Elops lacerta C. and V. (two specimens, 190 and 270 mm S.L. examined) has two
narrow, finger-like projections from the anterior part of the swimbladder. These
extend forward from about the level of the first vertebra to a point just behind the
saccular swelling on the basioccipital. Each projection is closely applied to the basi-
occipital and lies in a faint groove on that bone. A peritoneal sheath covers each
diverticulum, but the tissue is not thicker or in any other way macroscopically
distinguishable from the peritoneal layer covering the swimbladder itself.

de Beaufort (1909) describes paired anterior prolongations of the swimbladder
in Elops saurus Linn., and notes that these reach only as far forward as the middle
of the first vertebra. Cuvier and Valenciennes (1846, p. 373) also describe the paired
diverticula but do not indicate how far forward they extend (except to note that
there is no connection between the swimbladder and the interior of the skull).
Finally it may be noted that Hyrtl (1855) makes no mention whatsoever of anterior
swimbladder projections in this species.

In contradistinction to these earlier observations I find that the diverticula
do extend forward onto the basioccipital (at least in a fish 280 mm S.L.), the situation
being identical in all respects with that in E. lacerta. These discrepancies are difficult
to reconcile; the size of the various fishes examined may, however, be partly
responsible (see below).

In Elops machnata (Forsk.) (three specimens, two of 175 mm S.L., one 108 mm
S.L.) there are two peritoneum covered anterior projections. These extend only
to the level of the first vertebra in the larger specimens, and do not even reach that
point in the smaller fish. Thus in this species it seems likely that the extent of
diverticulum development is positively correlated with the size of the individual.

If in E. saurus there is a similar relationship, then the discrepancies between my
observations and those of de Beaufort, Hyrtl, Cuvier and Valenciennes might be
explained by their having dissected smaller fishes.

No linkage between skull and swimbladder is developed in the two species of
Albuloidei I have examined.

In Albula vulpes (Linn.) de Beaufort (op. cit.} describes a pair of short blind-sacs
from the anterior wall of the swimbladder at the level of the fourth vertebra. This
is the condition I found in a specimen of 97 mm S.L. (BMNH 1862. 4. 24 : 31) ; the
diverticula extend to the level of the anterior rim of the first vertebra. In contrast
I could find no trace of diverticula in the other specimens (160 mm and 165 mm S.L.
BMNH 1949. ii. 29 : i and 1855. 9. 19 : 875) I dissected ; in these fishes the for-
ward end of the swimbladder is smoothly rounded and lies below the first vertebra.
Likewise, I find no trace of the anterior diverticula in the swimbladder of Ptero-
thrissus gissu Hilgen. (a single specimen 394 mm S.L., the type of Bathythrissus

FISHES OF THE FAMILY MEGALOPIDAE 131

dorsalis Giinther). Giinther (1887), however, using the same specimen, described
two short horns from the anterior wall of the swimbladder, which lies posteriorly
in the visceral cavity and thus well behind the strongly developed transverse septum.
The viscera of this fish are now in a rather poor state of preservation, and this could
well account for my failure to find the horns.

Clearly the situation in elopoid and albuloid fishes requires further investigation
on more species and over a wide size-range of individuals. Nevertheless, it does
seem certain that no otophysic connection comparable in complexity with that of the
Megalopidae, exists in the Elopidae and Albulidae.

DISCUSSION

The status of the genera MEG A LOPS and TARPON. There is a great deal of
uncertainty about the generic status of the Atlantic Tarpon (see Hildebrand, 1963) ;
current opinion seems to favour it being congeneric with the Indo-Pacific species
Megalops cyprinoides.

Jordan and Evermann (1896) erected the genus Tarpon for the Atlantic species,
principally on the basis of it having a more posterior insertion of the dorsal fin than
Megalops cyprinoides. Hollister (1939) showed, however, that the dorsal fin position
is the same in both species, but that the ventral fins in Tarpon are nearer the snout.
In addition she emphasized that Tarpon has fewer vertebrae (57, comprising 33
trunk and 24 caudal, compared with 68 [38 trunk and 30 caudal] in Megalops}, and
pointed out differences in the fin ray counts (12-15 dorsal, and 19-22 anal rays in
Tarpon, cf. 19-21 dorsal and 24-27 anal rays in Megalops). Finally, Hollister drew
attention to certain differences in the caudal fin skeleton, particularly the longer
first uroneural in Megalops (see appendix, page 133).

A number of other osteological and soft anatomical differences associated with
the swimbladder-skull connection can now be added to those already known. In
particular may be noted the entirely peritoneal lateral wall of the periotic cavity in
Tarpon, the more differentiated precoelomic extension of the swimbladder and its
peritoneal sheath in Megalops, and the differences in the degree of development and
complexity shown by the intercalar in the two species.

At the same time, one should not lose sight of the fact that basic similarity in the
" bauplan " of the swimbladder-skull connection seems to stress the phyletic
integrity of the two taxa. In this character it seems reasonable to consider the
Atlantic species as the plesiomorph partner of the Indo-Pacific species. In other
features it is impossible to obtain a clear-cut picture of relative plesiomorphy or
apomorphy.

The nature of the interspecific morphological divergence in the swimbladder-skull
connection would, nevertheless, seem to warrant the separation of the two species
at generic level. Hence I propose to redefine the genera as follows :

Genus Megalops Lacepede 1803 : megalopid fishes with an expanded arm of the
intercalar forming the entire lateral wall of the large periotic bulla, which is thus
completely bounded by bony walls ; precoelomic extension of the swimbladder
with a short, thick-walled section, of narrow bore, immediately posterior to its
division into two periotic vesicles ; peritoneal cover to the precoelomic swimbladder

132 P. H. GREENWOOD

with a " U "-shaped otic portion leading to the periotic bullae anteriorly, and a
longer, egg-shaped main body posteriorly ; 67 or 68 vertebrae; anterior tip of the
first uroneural reaching the anterior part of the second preural centrum.

Genus Tarpon Jordan and Everman 1896 : megalopid fishes in which the inter-
calar does not contribute to the lateral wall of the periotic cavity, the latter bounded
laterally by precoelomic peritoneal tissue only ; precoelomic peritoneal cover of the
swimbladder extension passing smoothly into the otic arms forming the lateral
walls of the periotic cavity ; precoelomic swimbladder extension without a thick-
walled tubular portion ; 55-57 vertebrae ; anterior tip of the first uroneural reaching
the anterior part of the first preural centrum.

Relationships between the ELOPOMORPHA and CLUPEOMORPHA. Although
Greenwood et al. (1966) included the superorders Elopomorpha and Clupeomorpha
in a single division (later named the cohort Taeniopaedia, see Greenwood et al.
1967) they did so with little conviction, pointing out that if any relationship did
exist it was a very distant one. Subsequent research has certainly not provided
any new evidence to support any close phyletic grouping of the two taxa. In fact
the new evidence indicates quite the contrary situation. For example, a detailed
study of the clupeomorph family Denticipitidae (Greenwood, 1968) shows that its
supposedly elopomorph-like characters (Greenwood et al., 1966) are not of that
nature, but are in fact, clupeomorph characteristics. The earlier confusion was
due to insufficient knowledge of denticipitid anatomy. Other research on clupeo-
morph fishes (mostly unpublished, but see Patterson, 1970) has served to emphasize
the uniqueness of this group in a number of specialized characters, characters which
in all living and fossil elopomorphs show either no specialization at all or are
developed along very different lines.

The swimbladder-otic linkage provides a good example of the latter condition.
In the Clupeomorpha the swimbladder diverticula enter the skull, and the vesicles
into which they expand are lodged within the bullae developed from the prootic
and pterotic bones (or the prootic alone).

In megalopid Elopomorpha (the only members of the superorder with what could
reasonably be considered a specialized otophysic connection) the linkage is entirely
extracranial and involves different cranial bones. Such a condition could not be
considered basic for the otophysic connection characterizing the Clupeomorpha.
The otophysic connection found in the Elopidae (see p. 130), on the other hand, is of
such a generalized type that it could be basic to either a megalopid or a clupeo-
morph type. As an indicator of phyletic relationship in this particular context it is
valueless. Similar difficulties are encountered when one considers other characters ;
indeed, the more one analyzes the Clupeomorpha and Elopomorpha, the more
difficult it becomes to suggest a close relative for either taxon. In particular,
any attempt to find a sister group (sensu Hennig, 1966) for the Elopomorpha is
hampered by the extreme plesiomorphy of its basal, that is non-anguilliform,
members.

Thus, since no positive evidence can be found to support the hypothesis that the
Elopomorpha and the Clupeomorpha are each other's closest relatives, I propose
that the Clupeomorpha be removed from the cohort Taeniopaedia.

FISHES OF THE FAMILY MEGALOPIDAE 133

This action, of course, raises the question of where the Clupeomorpha should now
be classified. For the moment I can find no satisfactory answer. The clupeo-
morphs are not the sister group of the other primitive teleost cohort, the Archaeo-
phylaces (osteoglossomorph fishes), and there is, at present, insufficient evidence to
include them in the cohort Euteleostei (although this relationship is suggested by
some characters). As a temporary expedient, then, the Clupeomorpha will have to
be given cohort status and recognized as being a monophyletic group phylogenetically
incertae sedis. Hopefully they will not remain too long in limbo.

SUMMARY

The skull-swimbladder linkage in Megalops cyprinoides and Tarpon atlanticus is
described and compared. Although both species have the same fundamental
bauplan, the linkage in Megalops shows greater differentiation of the tissues, in
particular the peritoneal cover to the precoelomic swimbladder extension, and the
involvement of the intercalar bone in forming a periotic bulla.

In Tarpon the peritoneal cover is a simple tube, and the intercalar does not
contribute to the formation of a periotic bulla ; instead there is a periotic chamber
with a non-bony lateral wall (formed from the peritoneal swimbladder sheath).
Size correlated changes in the otic region are described for T. atlanticus; no such
changes occur in Megalops cyprinoides.

Other members of the Elopiformes were also investigated, and a simple otophysic
connection found in the three Elops species examined. Albula and Pterothrissus
lack an otophysic connection.

The disputed generic status of the Atlantic Tarpon (T. atlanticus) is reviewed,
and it is concluded that this species should be recognized as constituting a monotypic
genus, albeit one closely related to Megalops. The two genera are redefined.

Finally, the phyletic relationships of the Elopomorpha is considered, particularly
with regard to the supposed relationship of these fishes with the Clupeomorpha.
There is apparently no positive evidence to support the idea of close affinity, and the
Clupeomorpha are given cohort status.

ACKNOWLEDGEMENTS

Many people have contributed to this paper by presenting material or by dis-
cussing various parts of the problem. To all I express my gratitude. Especial
thanks in this respect are due to Drs. Colin Patterson, Donn E. Rosen, Gareth J.
Nelson and J-P. Gosse, and to Mr. Peter Forey who has freely made available to me
his unpublished data on fossil and living elopiforms.

APPENDIX

NOTES ON THE CAUDAL FIN SKELETON IN MEGALOPS AND TARPON
I have examined the caudal fin skeleton in three specimens of Tarpon atlanticus,
and four of Megalops cyprinoides. This structure is somewhat more variable in M.
cyprinoides than was indicated by Hollister (1939). In contrast the characters of

134 P. H. GREENWOOD

the caudal skeleton that are variable in M. cyprinoides are constant in Tarpon
atlanticus.

Variability in Megalops involves the neural arches and spines associated with the
first and second preural centra. Two of the four specimens have only a neural
arch on the first preural centrum ; in the other two specimens this centrum carries
a neural arch posteriorly, and an arch with short spine anteriorly. Three specimens
(including the second pair noted above) have a neural arch with a fuUy-developed
spine on the second preural centrum ; the fourth fish has an arch with very reduced
spine, together with an arch and fully-developed spine articulating between this
vertebra and the preceding one.

In all three Tarpon specimens there is a neural arch (but no spine) on the first
preural centrum, and an arch with fully developed spine on the second preural
centrum.

REFERENCES

BEAUFORT, L. F. DE. 1909. Die Schwimmblase der Malacopterygii. Morph. Jb., 39 : 526

644.

CUVIER, G. & VALENCIENNES, A. 1846. Histoire naturelle de poissons, 19, Paris.
FAHLEN, G. 1967. Morphological aspects on the hydrostatic function of the gas bladder of

Clupea harengus. Acta Univ. lund., Sect. 2, No. I : 1-49.
GREENWOOD, P. H. 1968. The osteology and relationships of the Denticipitidae, a family of

clupeomorph fishes. Bull, Br. Mus. nat. Hist. (Zool.), 16 : 213-273.
GREENWOOD, P. H., ROSEN, D. E., WEITZMAN, S. H., & MEYERS, G. S. 1966. Phyletic studies

of teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat.

Hist., 131 : 339-456.

GREENWOOD, P. H., MEYERS, G. S., ROSEN, D. E., & WEITZMAN, S. H. 1967. Named divi-
sions of teleostean fishes. Proc. biol. Soc. Wash., 80 : 227-228.
GUNTHER, A. 1887. Report on the deep-sea fishes collected by H.M.S. Challenger during

the years 1873-1876. Rep. Sci. Res. Challenger Exp. Zool., 22 : 1-335.
HENNIG, W. 1966. Phylogenetic systematics (Davis, D. D. and Zangerl, P. transl.). University

of Illinois Press, Chicago.
HILDEBRAND, S. F. 1963. Family Elopidae, in: Fishes of the western north Atlantic, Pt. 3.

Sears Foundation, New Haven.

HOLLISTER, G. 1939. Young Megalops cyprinoides from Batavia, Dutch East Indies, includ-
ing a study of the caudal skeleton and a comparison with the Atlantic species Tarpon

atlanticus. Zoologica, N.Y., 24, 449-475.
HYRTL, J. 1855. Ober die accessorischen Kiemenorgane und dem darmcanal der Clupeaceen.

Sber. Akad Wiss. Wien, 10 : 47-57.
JORDAN, D. S., & EVERMANN, B. W. 1896. The fishes of Middle and North America, part I.

Bull. U.S. natn. Mus., No. 47.

PATTERSON, C. 1970. A clupeomorph fish from the Gault (Lower Cretaceous). In the press.
RIDEWOOD, W. G. 1904. On the cranial osteology of the fishes of the families Elopidae and

Albulidae, with remarks on the morphology of the skull in the lower teleostean fishes

generally. Proc. zool. Soc. Lond., 2 : 35-81.
WOHLFAHRT, T. A. 1936. Das Ohrlabyrinth der Sardine (Clupea pilchardus Walb.) und seine

Beziehungen zur Schwimmblase und Seitenlinie. Zeitchr. Morph. Okol. Tiere, 31 : 371-410.

FISHES OF THE FAMILY MEGALOPIDAE

135

LIST OF ABBREVIATIONS USED IN FIGURES

basph basisphenoid ophv

bic basal part of intercalar par

boc basioccipital prot

boc dep depression in basioccipital prot dep

epi epiotic psph

exoc exoccipital pter

fica foramen for internal carotid ptf

artery ptsph

fn foramen for unidentified branch pw

of trigeminal nerve

fr frontal soc

fu V fused vertebral centrum sph

gr groove stf

ic intercalar X

ica saddle-like wing of intercalar

icagr posterolateral wall of groove ist v

formed by saddle-like wing of i.

intercalar 2.
icb lower limb of intercalar

icsh shield-like expansion of lower 3.

intercalar limb 4.

icu upper limb of intercalar 5.

io orbitosphenoid

oaf opening to foramen for orbital

artery

opening for head vein
parietal
prootic

depression in prootic
parasphenoid
pterotic

posttemporal fossa
pterosphenoid

peritoneal wall to periotic cham-
ber

supraoccipital
sphenotic
subtemporal fossa
foramen for vagus nerve

first vertebra

Foramen for head vein

Foramen for hyomandibular

branch of VII

Foramen for orbital artery

Foramen

Foramen opening into periotic

bulla

P. H. GREENWOOD, B.Sc.(Rand)., D.Sc.(Rand).
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON S.W.y

B

PLATE i
FIG. i. Megalops cyprinoides. Otic region of skull in lateral view.

FIG. 2. Megalops cyprinoides. Intercalar, A, lateral aspect, viewed somewhat obliquely
from above and anteriorly. B, dorsal aspect, the long axis of the bone aligned horizontally.

Drawn by Sharon Lesure

Bull. Br. Mus. nat. Hist. (Zool.) 19, 3

sph par epi

fr

PLATE i

pter

exoc

fu v

A bic

FIG. 2

icagr

ica

icsh

B*

PLATE 2

Megalops cyprinoides. Posterior aspect of skull, viewed obliquely from the left side and some-
what ventrally. Drawn by Sharon Lesure

Bull. BY. Mus. nat. Hist. (Zool.) 19, 3

PLATE 2

epi soc

pter

sph

exoc

ophv

boc

psph

PLATE 3

FIG. i. Tarpon atlanticus. Otic region of skull (from a specimen 78-2 cms S.L.). The foramen
numbered 4 is not present in the left prootic, and was absent in other specimens examined.

FIG. 2. Tarpon atlanticus. Intercalar and exoccipital. A, Bones orientated in their natural
position but viewed obliquely from above and anteriorly. B, Bones orientated in their natural
position, but viewed obliquely from above and behind. Drawn by Sharon Lesure

Bull. Br. Mus. nat. Hist. (Zool.) 19, 3

soc

epi

PLATE 3

IO

fr

boc

boc dep

prot dep prot fica

psph

1cm

FIG. i

ICU

ICU

FIG. 2

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

24

FOURTEEN NEW SPECIES OF

FREELIVING MARINE NEMATODES

FROM THE EXE ESTUARY

R. M. WARWICK

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 4

LONDON: 1970

FOURTEEN NEW SPECIES OF FREELIVING
MARINE NEMATODES FROM THE EXE ESTUARY

BY

RICHARD MARTYN WARWICK

- - . -- X-
Department of Zoology, University of Exeter

and
Dove Marine Laboratory, Cullercoats, Northumberland

Pp. 137-177; 16 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

ZOOLOGY Vol. 19 No. 4

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 4 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 23 April, 1970 Price i 45.

(1-2)

FOURTEEN NEW SPECIES OF FREELIVING
MARINE NEMATODES FROM THE EXE ESTUARY

By R. M. WARWICK

SYNOPSIS

Fourteen new species of freeliving interstitial marine nematodes are described from sandy
habitats in the Exe estuary area, South Devon. They are all referable to existing genera. Eno-
plolaimus denticulatus sp. nov. is similar to E. connexus Wieser, 1953, but differs from it princi-
pally in the shorter cervical setae of the female, and in the form of the spicules in the male;
Mesacanthion africanthiforme sp. nov. is characterized by the absence of a pre-cloacal supple-
ment, by the presence of stout post-cloacal setae in the male, and by the unique structure of the
spicules and gubernaculum ; Epacanthion gorgonocephalum sp. nov. is related to E. enoploidiformis
(Gerlach, 1952) comb. nov. and E. oliffi Inglis, 1966, differing from the former in the possession
of a hirsute anterior end, and from the latter in the structure of the male copulatory
apparatus and in the setal arrangement on the head; Axonolaimus orcombensis sp. nov. is closest
to A. steineri Timm, 1954, but differs from it in the number of subcephalic setae and in the
shape of the amphids, gubernaculum and spicules; Leptolaimus ampullaceus sp. nov. is charac-
terized by the number of pre-cloacal supplements, by the structure of the spicules and guber-
naculum, and by the presence of longitudinal files of papillae down the body-length ; Camacolaimus
barbatus sp. nov. is closest to C. prytherchi Chitwood, 1935, but has shorter cephalic setae, no alae
on either side of the cloaca, a prominent post-cloacal supplement, lateral papillae in the oeso-
phageal region, and a different setal arrangement on the tail; Paralinhomoeus uniovarium sp.
nov. is characterised by its large amphids, long cephalic setae, and by the possession of a single
anterior ovary; Theristus (Theristus) denticulatus sp. nov. is characterized by the typically
complex structure of the gubernaculum and by the presence of stout ventral spines on the
male tail; Theristus (Theristus) inter stitialis sp. nov. is characterized by its long cephalic setae,
by the position and size of the amphids, and by the structure of the male copulatory apparatus ;
Theristus (Trichotheristus) psammoides sp. nov. is closest to T. (T.) vicinus Riemann, 1966, but
has larger amphids and differently shaped spicules and gubernaculum; Chromaspirina inglisi sp.
nov. ( C. pontica sensu Gerlach, 1951, nee. Filipjev, 1918) is characterized by a uniform
covering of fine hairs over the entire body-surface ; Microlaimus spirifer sp. nov. is characterized
by the position and form of the amphids, by the form of the copulatory apparatus in the male,
and by the long tail; Pomponema reducta sp. nov. is characterized by the possession of only six
cephalic setae, and by the arrangement of denticles in the buccal cavity; Paracanthonchus
opheliae sp. nov. is characterized by the distinctive structure of the gubernaculum and of the
spicules.

INTRODUCTION

DURING the course of an ecological survey of the distribution of freeliving nematodes
in the Exe estuary, several species were discovered which proved to be new. This
paper deals with the species which were fairly common in the estuary, and their
ecological distribution, together with that of other species, will subsequently be
discussed elsewhere. Suffice it to say that all the species described here are inter-
stitial, and were found in sandy habitats at various tide levels at Shelly Bank, just
inside the mouth of the estuary at Exmouth, and on Exmouth beach near Orcombe
Point.

B 139

140

R. M. WARWICK

Measurements have been made from glycerine mounts under the camera lucida with
an opisometer. Curved structures, e.g., spicules, are measured as the curve and not
the chord, the oesophagus length is measured from the anterior end of the body, and
the head diameters are measured at the level of the first circle of cephalic setae unless
otherwise stated. The type-material has been deposited at the British Museum
(Natural History). The classification used is that of De Coninck (1965), but following
Inglis (1966) the only groups recognized are families and genera. Where a large
amount of material is available only three males and three females have been
measured, since relative and absolute measurements have only a limited taxonomic
value. Attention has rather been given to the morphology of the male copulatory
apparatus, in response to recent pleas by Wieser and Hopper (1967) and Inglis
(1967). All the species are described from a syntypic series.

Family ENOPLIDAE

Enoplolaimus denticulatus sp. nov.

(Fig. i)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 262.

Shelly Bank: M.L.W.S.T. Orcombe Point:

a

b

Males

5 2> 93
56-00

3'54

59-22

4-12

Females

5775
51-63
61-49

4-02
4-00
4-48

c

J 9'95
14-30

23-11

17-08
21-14
16-25

v%

60-24
62-16

67-25

Body
length (mm.)

3'97

379
4-10
4-44
4-55

MEASUREMENTS (in mm. in order of body lengths). Males: Body breadth: 0-075;
0-060; 0-064. Oesophagus length : 0-98 ; 0-95 ; 0-92. Distance of nerve ring from
anterior : 0-225 > ' 22 '> * 22 - Head diameter : 0-053 J '45 1 0-049. Length of
labial setae : 0-016 ; 0-016 ; 0-017. Length of longer cephalic setae : 0-082 ; 0-082 ;
0-071. Length of shorter cephalic setae : 0-040 ; 0-040 ; 0-032. Length of longest
cervical setae : 0-043 ; 0-043 ; 0-025. Length of mandibles : 0-032 ; 0-026 : 0-027.
Length of onchia : 0-019 > ' OI 6 ; 0-014. Tail length : 0-199 ' 2 35 > 0-164. Cloacal
diameter : 0-050 ; 0-044 '> 0-042. Spicule length : 0-053 ; 0-060 ; 0-055. Guber-
naculum length : 0-022 ; 0-020 ; 0-023. Supplement length : 0-012 ; 0-008 ; o-on.
Distance of supplement anterior to cloaca : 0-139 > 0-144 ; 0-132.

Females : Body breadth : 0-071 ; 0-086 ; 0-074. Oesophagus length : 1-02 ;
I'll ; 1-015. Distance of nerve ring from anterior : 0-235 > ' 2 5 '> ' 2 35- Head
diameter : 0-055 '> 0-056 ; 0-050. Length of labial setae : 0-017 ; 0-018 ; 0-017.
Length of longer cephalic setae : 0-092 ; 0-070 ; 0-099. Length of shorter cephalic

FOURTEEN NEW SPECIES OF MARINE NEMATODES

141

setae : 0-037 > 0*029 > '34- Length of longest cervical setae : 0-008 ; o-oio
o-oio. Length of mandibles : 0-029 0-032 ; 0-029. Length of onchia : 0-019
0-019 0-020. Tail length : 0-24 ; 0-21 ; 0-28. Anal diameter : 0-049 '> O'49
0-051. Distance of vulva from anterior : 2-47 ; 2-76 ; 3-06.

FIG. i. Enoplolaimus denticulatus sp. nov. A, Details of cephalic structure of male. B,
Lateral view of female tail, c, Lateral view of male tail. D, Male head. E, Lateral view
of spicules and gubernaculum. F, Female head.

This is a relatively large member of the genus. The cuticle is smooth. The
mouth is surrounded by three relatively high lips which have pronounced subsidiary
lobes and bear semi-lunar striations on their inner surfaces (Fig. I A) . The six stout
conical labial setae arise at the level of the tips of the mandibles. The six long
cephalic setae are fairly stout, and the four shorter ones more slender. The posterior
border of the short cephalic capsule has small incisions with fenestrae to accommo-
date the bases of the cephalic setae. In the male there is a series of long cervical

142 R. M. WARWICK

setae just posterior to the cephalic capsule (Fig. iD), but the female only has a
circle of six very short setae in this position (Fig. iF) . The mandibles have the usual
appearance of two lateral rods united by an anterior curved bar, with the pointed
tips projecting into the buccal cavity. The onchia are large, and distinct from the
mandibular plate. The terminology used in describing this and subsequent members
of the Enoplidae follows that of Inglis (1964).

The oesophagus is cylindrical with three files of glands down its length. Short
setae are fairly numerous in the oesophageal region of both sexes, but are scarce
posterior to this, becoming more numerous on the tail. The proximal third of the
tail is conical and the remainder cylindrical, the tip being rounded and slightly
swollen (Figs. iB and iC)

MALE. The spicules are more or less straight when viewed laterally. The distal
tips are strongly hooked and bear a series of fine denticles (Fig. lE) . The proximal
ends are open and funnel shaped, and there are dorsal swellings just posterior to
them. When viewed from the ventral aspect the spicules appear to be far apart
proximally, bent in the middle, and closely adherent to the narrow gubernaculum in
their distal half, as illustrated for E. connexus by Wieser (1953, p. 77, Fig. 386).
The gubernaculum is also hooked at its tip, but bears no denticles. The supplement
.is small, simple and tubular. The testes are paired, opposed and outstretched.

FEMALE. The ovaries are paired, opposed and doubly reflexed. They are about
equal in size.

DISCUSSION. This species is closest to E. connexus Wieser, 1953, but differs from
it in having much shorter cervical setae in the female, spicules of a different shape
with denticulate tips, and the supplement in a more anterior position. The present
species is much larger than E. connexus, and the cephalic setae are accommodated in
fenestrae of the cephalic capsule and do not lie wholly posterior to it as in Wieser 's
species.

Mesacanthion africanthiforme sp. nov.
(Fig. 2)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 264.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.
Orcombe Point : M.H.W.N.T., M.L.W.N.T.

Body
a b c V% length (mm.)

Males 81-84 5-41 19-61 4-49

77-27 4-32 16-14 2-55

65-83 4-09 16-01 2-37

Females 56-06 4-48 15 -10 58-65 3-70

57-97 4-72 21-17 61-56 3-07

6375 4-25 I7' 58-82 2-55

FOURTEEN NEW SPECIES OF MARINE NEMATODES 143

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth : 0-055 '
0*033 i 0-036. Oesophagus length : 0-83 ; 0-59 ; 0-58. Distance of nerve ring from
anterior : 0-24 ; 0-18 ; 0-18. Head diameter : 0.047 > 0*032 ; 0-028. Length of
labial setae : 0-008 ; 0-007 ; 0-006. Length of longer cephalic setae : 0-041 ; 0-026 ;
0-024. Length of shorter cephalic setae : 0-020 ; 0-013 ; o-oio. Length of longest
cervical setae : 0-014 '< 0-016 ; 0-007. Length of mandibles : 0-029 '< 0*020 ; 0-019.
Length of onchia : 0-017 ; 0-013 ; 0-012. Tail length : 0-229 ' T 5^ '> 0-148. Cloacal
diameter : 0-047 '> '33 1 '33- Spicule length : 0-033 ; 0-025 ; 0-020. Guber-
naculum length : 0-013 ; o-on ; o-oio. Distance of first pair of supplementary
setae posterior to cloaca : 0-026 ; 0-014 ; 0-015. Distance of second pair of sup-
plementary setae posterior to cloaca : 0-085 J 0-050 ; 0-052.

Females : Body breadth : 0-066 ; 0-053 ; 0-040. Oesophagus length : 0-825 i
0-65 ; 0-60. Distance of nerve ring from anterior : 0-26 ; 0-20 ; 0-19. Head dia-
meter: 0-050 ; 0-036 ; 0-036. Length of labial setae : 0-007 J 0-006 ; 0-006. Length
of longer cephalic setae : 0-030; 0-023; 0-023. Length of shorter cephalic setae :
0-013 ; o-oio ; o-oii. Length of longest cervical setae : 0-012 ; o-oio ; 0-008.
Length of mandibles : 0-032 ; 0-024 ' 0-023. Length of onchia : 0-020 ; 0-016 ;
0-014. Tail length : 0-245 i 0>I 45 '> o>1 5 - Anal diameter : 0-045 > o '37 ', 0-033.
Distance of vulva from anterior : 2-17 ; 1-89 ; 1-50. Egg length : ; 0-215 ; .
Egg breadth : ; 0-039 '

This is a long slender species, and has a characteristically dome-shaped head
(Fig. 2 A). There are some irregular areas of pigmentation just posterior to
the cephalic capsule. The cuticle is smooth. The three lips are relatively low, and
have no internal striation. The six labial setae are stout and conical, and situated
at the level of the tips of the mandibles. The cephalic setae originate from about the
middle of the cephalic capsule, the posterior border of the capsule being deeply
incised and the bases of the setae accommodated in fenestrae. The four shorter
cephalic setae are about half the length of the six longer ones. There are six files
of long setae extending one third to two thirds of the way down the oesophagus
length in both sexes, but they tend to be more numerous in the male. Setae are
scarce further posteriorly and become more numerous again in the tail region. The
three mandibles have the usual appearance and bear prominent lateral processes.
The onchia are large and equal in size.

The oesophagus is cylindrical, with the usual three files of glands down its length.
The tail is conical in its distal third and the remainder is cylindrical (Figs. 26 and
2C). The tip is slightly swollen in some specimens.

MALE. The spicules are small and fairly straight. They are pointed and open
ended distally with a thickened cuticular ring subterminally. The gubernaculum
consists of a pair of tubes surrounding the distal ends of the spicules, and these tubes
are united by a median bar which has a roughly triangular appearance in lateral view
(Figs. 2D and 2E). There is no cuticularized pre-cloacal supplement. The tail,
however, bears two pairs of stout conical setae posterior to the cloaca, and these
probably have a copulatory function. This feature is parallelled in Africanthion
nudus Inglis, 1964, which has a similar series of setae anterior to the cloaca. The
testes are paired, opposed and outstretched.

144

R. M. WARWICK

FOURTEEN NEW SPECIES OF MARINE NEMATODES 145

FEMALE. The ovaries are paired, symmetrical, and doubly reflexed. The eggs
are large and elongate.

DISCUSSION. In this species the male is distinct from all others in the combination
of the absence of a pre-cloacal supplement and the presence of stout post-cloacal
setae. Also the structure of the gubernaculum and spicules is unique. It is close to
the genus Africanthion Inglis, 1964, in that the supplement is replaced by stout
setae, but cannot be referred to this genus because in Africanthion the onchia are
unequal in size, the lips have semi-lunar striations, and the copulatory setae are pre-
cloacal and not post-cloacal as in the present species. Since one other species
Mesacanthion infantile (Ditlevsen, 1930), is also reported to have no supplement,
and the head structure of the present species is typical of the genus, it will be referred
to Mesacanthion.

Epacanthion gorgonocephalum sp. nov.
(Figs. 3 and 4)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 263.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.L.W.N.T., M.L.W.S.T.

a

b

c

v%

Body
length (mm.)

48-52

47-44
52-32

4-36
4-38
4-38

19-32

20-53
18-90

4-27
4-27
4-29

34-86
44-10
37-32

4-58
4-45
4-56

21-38
20-13
20-70

58-84

60-04

60-34

4-8l
4-63

474

Males

Females

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth : 0-088 ;
0-090 ; 0-082. Oesophagus length : 0-98 ; 0-975 ; 0-98. Distance of nerve ring
from anterior : 0-185 ; 0-192 10-190. Head diameter: 0-059 >'55 > 0-052. Length
of labial setae : 0-021 ; 0-019 '> 0-022. Length of longer cephalic setae :
0-074 ; 0-063 I 0-073. Length of shorter cephalic setae : 0-043 ; 0-042 ; 0-039.
Length of longest subcephalic setae : 0-036 ; 0-037 > '35- Length of shortest
subcephalic setae : 0-012 ; 0-017 i o-on. Length of longest cervical setae : 0-048 ;
0-044 i 0'35- Length of longest body setae : 0-029 ; 0-027 ; 0-024. Distance of
dense ring of setae from anterior : 0-31 ; 0-32 ; 0-30. Mandible length : 0-020 ;

FIG. 2. Mesacanthion africanthiforme sp. nov. A, Male head. B, Lateral view of male tail,
c, Lateral view of female tail. D, Lateral view of spicules and gubernaculum. E, Ventral
view of spicules and gubernaculum.

146 R. M. WARWICK

0-019 ; 0-020. Tail length : 0-221 ; 0-208 ; 0-227. Cloacal diameter : 0-060
0-057 J 0-052. Spicule length : 0-053 '> 0-051 ', 0-050. Gubernaculum length
0-020 : 0-018 : 0-018.

FIG. 3. Epacanthion gorgonocephalum sp. nov. A, Male head. B, Lateral view of spicules
and gubernaculum. c, Ventral view of spicules and gubernaculum.

Females : Body breadth : 0-138 ; 0-105 ; 0-128. Oesophagus length : 1-05 ; 1-04 ;
1-04. Distance of nerve ring from anterior : 0-221 ; 0-198 ; 0-215. Head diameter :
0-059 > 0-062 ; 0-062. Length of labial setae : 0-023 '> 0-019 ; 0-019. Length of
longer cephalic setae : 0-082 ; 0-071 ; 0-073. Length of shorter cephalic setae :
0-035 I 0-032 ; 0-034. Length of longest subcephalic setae : 0-012 ; ; o-on.
Length of longest cervical setae : 0-028 ; 0-029 '> 0-026. Length of longest body

FOURTEEN NEW SPECIES OF MARINE NEMATODES

setae : 0-023 i 0-024 ; 0-019. Mandible length : 0-022
0-225 i 0-230 ; 0-229. Anal diameter : 0-072 ; 0-068
from anterior : 2-83 ; 2-78 ; 2-86.
0-113 ; 0-092 ; 0-091.

0-019 > ' OI 9- Tail length :
; 0-072. Distance of vulva
Egg length : 0-264 ; 0-363 ; 0-312. Egg breadth :

FIG. 4. Epacanthion gorgonocephalum sp. nov. A, Anterior end of female. B, Lateral
view of female tail, c, Lateral view of male tail. D, Anterior end of male.

The body is relatively long and slender, and the cuticle smooth. The three lips
have a marked subsidiary lobe, and internally are marked by semi-lunar striations
posterior to a line running backwards from the tips of the mandibles. The outer
parts of the lip are marked by wider spaced striations, and the lip margin is scalloped.
The subsidiary lobe has no internal striation or scalloped edge (Fig. 3 A) . The labial
setae are stout, and originate just posterior to the tips of the mandibles. There are

148 R. M. WARWICK

six long and four shorter cephalic setae situated at the level of the bases of the
onchia. In the male there are twelve groups of subcephalic setae, three per group,
one being long, one of medium length and one short. In the female only six single
short subcephalic setae are present, or these may be absent altogether (Fig. 4A). The
bases of the larger cephalic setae are supported by small areas of thickened cuticle in
the head. The mandibles are typical of the genus. They appear in optical section
as two longitudinal rods joined by a thin sheet of cuticle, and are relatively long and
slender. The three onchia are equal in size and are situated at the bases of the
mandibles. The onchial plate is rounded posteriorly, and has well-developed radial
processes. Each of the three plates curve inwards posteriorly, and their posterior
edges almost meet at the base of the buccal cavity. The cephalic slits are prominent
and crescentic, and are situated just anterior to the level of the tips of the onchia.
There are numerous long dense cervical setae in the male, and there is a region of
very dense, but rather shorter setae about one-third of the way down the oesophagus
length (Fig. 40). Posterior to this setae become scarcer, increasing in density again
on the tail. In the female cervical setae are sparse (Fig. 4A), and there are very few
setae on the body-surface until just before the anus.

The oesophagus is swollen anteriorly round the base of the buccal cavity, and
is otherwise cylindrical with the usual three files of glands down its length.
The anterior half of the tail is conical and the posterior half cylindrical (Figs. 46
and 46).

MALE. The gubernaculum consists of a short double tube surrounding the ends
of the spicules, and bears two small lateral projections distally (Figs. 36 and 36).
The spicules are short and slightly curved, and each bears distally a pair of laterally
curving knobs. There is no pre-cloacal supplement. The testes are not visible.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed. There is
usually only one egg per uterus, but one specimen was found with two in the anterior
uterus.

DISCUSSION. This species is similar to Enoplolaimus enoploidiformis Gerlach,
1952, which should be transferred to Epacanthion, and also to Epacanthion oliffi
Inglis, 1966, which both have short spicules and no pre-cloacal supplement. E.
enoploidiformis does not have a hirsute anterior end as do E. oliffi and E. gorgono-
cephalum, while E. oliffi differs from the present species in having no gubernaculum,
spicules with denticulate tips and a different setal arrangement on the head.

Family AXONOLAIMIDAE

Axonolaimus orcombensis sp. nov.

(Fig- 5)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 265.

DISTRIBUTION IN EXE ESTUARY. Orcombe Point : M.L.W.N.T., M.L.W.S.T.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 149

Body
a b c V% length (mm.)

Males 111-07 14-81 20-73 3-11

93-13 14-90 20-55 2-98

113-13 15-74 21-55 3-62

Females 83-32 14-18 20-53 62-50 3-12

91-43 14-55 21-92 56-25 3-20

92-00 15-33 21-61 59-63 3-22

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-028 ; 0-032 ; 0-032. Oesophagus length : 0-21 ; 0-20 ; 0-23. Distance of nerve
ring from anterior : 0-113 0-120 ; 0-134. Distance of excretory pore from anterior :
0-034 > 0-029 ; 0-036. Head diameter : 0-012 ; 0-013 ; 0-013. Length of cephalic
setae : 0-023 0-024 '< 0-026. Length of subcephalic setae : 0-018 ; 0-018 ; 0-017.
Amphid length : 0-007 ' 0-007 ; 0-007. Amphid breadth : 0-006 ; 0-006 ; 0-006.
Body diameter at level of amphids : 0-014 ; 0-015 ; 0-014. Length of buccal cavity:
0-021 ; 0-022 ; 0-024. Tail length : 0-150 ; 0-145 ; 0-168. Cloacal diameter :
0-028 ; 0-030 ; 0-031. Spicule length : 0-043 ; 0-045 '> 0-048. Gubernaculum
length : 0-021 ; 0-022 ; 0-023.

Females : Body breadth : 0-037 > 0' 35 ', 0-035. Oesophagus length : 0-22 ;
0-22 ; 0-21. Distance of nerve ring from anterior : 0-132 ; 0-122 ; 0-124. Distance
of excretory pore from anterior : 0-035 ; 0-032 ; 0-035. Head diameter : 0-013 J
0-014 '> 0-014. Length of cephalic setae : 0-023 i 0-027 '> 0-024. Length of sub-
cephalic setae : 0-013 ; 0-018 ; 0-015. Amphid length : 0-006 ; 0-007 ; 0-006.
Amphid breadth : 0-006 ; 0-006 ; 0-006. Body diameter at level of amphids :
0-016 ; 0-016 ; 0-016. Length of buccal cavity : 0-023 0-023 i 0-022. Tail length:
0-152 ; 0-146 ; 0-149. Anal diameter : 0-025 ; 0-029 > 0-028. Distance of vulva
from anterior : 1-95 ; 1-80 ; 1-92. Egg length : 0-198 ; ; . Egg breadth :
0-027 ; ;

The body is long and slender, and the cuticle marked by fine transverse striations,
which commence just behind the cephalic setae and extend almost to the tail- tip.
The mouth is surrounded by six rounded lips each bearing a small conical labial
papilla. There are four long cephalic setae, and in the male a circle of eight long
subcephalic setae is situated just anterior to the base of the buccal cavity, while in
the female the sublateral subcephalic setae are paired, giving twelve in all (Figs. 5A
and 56). The posterior portion of the buccal cavity is long and tapering, and there
are no odontia in the anterior portion. The ampids are small, describing a single
loop, and are almost circular in outline.

The oesophagus is cylindrical throughout its length. A few short scattered setae
are present on the body-surface, particularly in the cervical region and on the tail.
The tail tapers throughout its length, and in the male bears two ventrolateral files
of short setae and a pair of terminal setae about 0-009 mm. long, but these are absent
in the female (Figs. 5D and 5E).

MALE. The gubernaculum has a strong, irregularly-shaped, dorsally-pointing

150

R. M. WARWICK

FIG. 5. Axonolaimus orcombensis sp. nov. A, Lateral view of male head. B, Lateral view
of female head, c, Lateral view of cloacal region in male. D, Lateral view of female tail.
E, Lateral view of male tail.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 151

apophysis (Fig. 56). The spicules are arcuate, strongly cephalate proximally and
pointed distally. There are 20-30 pre-cloacal supplements, more widely spaced
posteriorly than anteriorly. Each consists of a small pore, and between them the
ventral body-cuticle is markedly thickened. The testes are paired, opposed and
outstretched.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed. The eggs
are very large and elongate.

DISCUSSION. The only other species with cephalic setae approaching two head
diameters in length and with amphids which do not reach the level of the base of
the buccal cavity is A. steineri Timm, 1954. A, orcombensis differs from this species
in the number of subcephalic setae, in the shape of the amphids, in the form of the
gubernaculum and spicules and in the general body proportions.

Family LEPTOLAIMIDAE

Leptolaimus ampullaceus sp. nov.
(Fig. 6)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 266.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.
Orcombe Point : M.L.W.S.T.

Body
a b c V% length (mm.)

Males 61-43 7-05 7-17 0-86

58-67 6-93 6-52 0-88

67-87 8-48 7-04 0-95

Females 45-52 7-17 7-76 48-49 1-32

61-18 7-70 6-62 46-15 1-04

51-76 6-98 6-29 48-86 0-88

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth:
0-014 ; 0-015 ; 0-014. Oesophagus length : 0-122; 0-127 > 0-112. Distance of nerve
ring from anterior : 0-075 ; 0-080 ; 0-178. Head diameter : 0-004 '> 0-005 ;o 0-05.
Length of cephalic setae : 0-005 '> 0-006 ; 0-005. Amphid breadth : 0-005 ; 0-005 '>
0-005. Amphid length : 0-006 ; 0-007 ; 0-006. Body diameter at level of amphids :
0-007 '> 0-008 ; 0-008. Tail length : 0-120 ; 0-135 '> ' I 35- Cloacal diameter :
0-014 i 0-015 ; 0-014. Spicule length : 0-017 ; 0-018 ; 0-019. Gubernaculum
length : 0-009 ' 0-009 0-009. Supplement length : o-oio ; o-on ; o-oio.

Females : Body breadth : 0-029 > 0-017 ; 0-017. Oesophagus length : 0-184 ;
o^SS i 0-126. Distance of nerve ring from anterior : o-in ; 0-094 ; 0-083. Head
diameter : 0-007 > 0-007 > 0-005. Length of cephalic setae : 0-008 ; 0-007 ; 0-005.

D

152

R. M. WARWICK

FIG. 6. Leptolaimus ampullaceus sp. nov. A, Lateral view of male head. B, Lateral view of
cloacal region in male, c, Lateral view of posterior end of male. D, Region of oesophageal
bulb. E, Lateral view of female tail.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 153

Amphid breadth : 0-005 '> 0*004 '> 0-004. Amphid length : 0-006 ; 0-005 '> 0-004.
Body diameter at level of amphids : 0-009 ; 0-008 ; 0-007. Tail length : 0-170 ;
0-157 ; 0-140. Anal diameter : 0-015 ; 0-013 ; o-oio. Distance of vulva from
anterior : 0-64 ; 0-48 ; 0-43. Egg length : 0-066 ; ; . Egg breadth : 0-015
>

Transverse striation of the cuticle extends from the level of the cephalic setae to
the tip of the tail. There is no lateral differentiation, but a single file of prominent
conical papillae is present down each side of the body. The lips are indistinct and
no labial papillae can be seen. Four long cephalic setae are present. The buccal
cavity appears to be small and conical (Fig. 6A), but merges imperceptibly with the
lumen of the oesophagus and may possibly be long and tubular as in other members
of the genus. The amphids are large, and oval in outline. The amphidial nerve is
prominent, and exits posteriorly. The amphids are slightly larger in the male than
in the female. The oesophagus is narrow at its anterior end and broadens just pos-
terior to the amphids. It has three small swellings along its length, and terminates
in a small bulb (Fig. 6D). The long tail is more or less cylindrical in the female, but
in the male the first third is conical and the posterior two-thirds filiform (Figs. 6C
& 6E). There are a few scattered setae on the male tail.

MALE. The gubernaculum is small, and bears a pair of dorsal apophyses which
are relatively slender (Fig. 6B). The spicules are dilated distally, with pointed tips,
and become narrow proximally with cephalate ends. There are 7-9 tubular well-
cuticularized pre-cloacal supplements, each being cephalate proximally and attached
to a pear-shaped ampulla. Distally they pass through a small densely-thickened
area of body-cuticle. The testes are paired, opposed and outstretched.

FEMALE. The ovaries are paired, symmetrical, opposed and outstretched. The
eggs are large and elongate. In the third female measured above there is a single
ventral supplement identical in structure to those found in the male. It is situated
0-170 mm. posterior to the vulva.

DISCUSSION. This species is probably closest to L. setiger Schuurmans Stekhoven
& De Coninck, 1933, the males of which are described by Gerlach (1952). It differs
from this species in the number of pre-cloacal supplements (16 in L. setiger), in the
structure of the gubernaculum and spicules, and in the presence of lateral files of
papillae down the body-length. There may also be a difference in the structure of the
buccal cavity but, as mentioned above, this feature is difficult to interpret in the
present specimens.

Family CAMAGOLAIMIDAE

Camacolaimus barbatus sp. nov.

(Fig. 7)

MATERIAL STUDIED. Two males and one female. B.M. (N.H.), Reg. No. 1968.
267.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.H.W.N.T., M.T.L., M.L.W.S.T.

154 R- M. WARWICK

Body
a b c V% length (mm.)

Males 51*84 7-30 21-18 1-97

50-25 7-44 2I-85 2-01

Female 48-11 6-85 21-71 52-25 1-78

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-038 ; 0-040. Oesophagus length : 0-27 ; 0-27. Distance of nerve ring from an-
terior : 0-122 ; 0-148. Head diameter : o-on ; o-on. Length of cephalic setae :
0-003 0-003. Amphid width : 0-005 '> 0-004. Length of buccal tooth : 0-029 '
0-029. Tail length : 0-093 ; 0-092. Cloacal diameter : 0-031 ; 0-032. Spicule
length : 0-058 ; 0-053. Gubernaculum length : 0-016 ; 0-015.

Female : Body breadth : 0-037. Oesophagus length : 0-26. Distance of nerve
ring from anterior : 0-153. Head diameter : 0-012. Length of cephalic setae :
0-003. Amphid width : 0-005. Length of buccal tooth : 0-032. Tail length :
0-082. Anal diameter : 0-028. Distance of vulva from anterior : 0-93.

Coarse transverse striations of the cuticle commence just posterior to the cephalic
setae and extend to a level just anterior to the spinerette on the tail-tip. Two
lateral longitudinal lines indicate narrow fields of unstriated cuticle, and can first
be detected a short distance behind the base of the oesophagus. They fade out just
posterior to the anus or cloaca. There is a lateral file of small conical papillae between
the head and the base of the oesophagus, but otherwise the body is naked.

The lips are indistinct, and there are six small conical labial papillae. The four
cephalic setae are very short. The dorsal buccal tooth has a prominent shoulder
near its distal tip where it leaves the oesophageal musculature (Fig. 76). The distal
tip of the tooth is rounded, and proximally it extends far down into the lumen of the
oesophagus. The amphids are positioned just anterior to the cephalic setae. They
are more or less circular, with the amphidial nerve extremely prominent and attached
to the dorsal side of the amphid. The anterior four-fifths of the oesophagus is
narrow and cylindrical, with three small swellings down its length. The posterior
fifth is conical, giving the oesophagus a broad base. The tail is conical, with a pointed
unstriated spinerette (Figs. 7 A and 76).

MALE. The gubernaculum is small, with pointed apophyses projecting dorsally
on both the dorsal and ventral sides of the spicule tips (Fig. 76). The spicules are
bent in the middle, with the proximal half swollen and the distal half narrow.
Proximally they terminate in sharp right-angle bends so that the tips point ventrally.
There is a pair of stout conical setae just posterior to the cloaca. A large ventral
post-cloacal supplement is present one quarter of the tail-length from the tail tip. It
consists of a circular area of thin cuticle by which the epidermis communicates by a
protoplasmic thread with the exterior. On either side of the supplement is a stout
conical seta, and there is another pair of setae positioned more laterally between
the supplement and tail-tip. The testes are paired, opposed and outstretched.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 155

DISCUSSION. This species is closest to C. prytherchi Chitwood, 1935, but differs
from it in having much shorter cephalic setae (3/1 as compared with 8-io/x), no alae
on either side of the cloaca, and in having a prominent post-cloacal supplement. C.
prytherchi has only one pair of setae on the male tail (the pair immediately posterior
to the cloaca), and has no lateral papillae in the oesophageal region. C. prytherchi
has been redescribed by Wieser and Hopper (1967), and it is with this description that

FIG. 7. Camacolaimus barbatus sp. nov. A, Lateral view of female tail. B, Lateral
view of male head, c, Lateral view of male tail.

156 R. M. WARWICK

the present species is compared since these authors, in re-examining the type-
specimen, have found structures not originally described by Chitwood.

Family LINHOMOEIDAE

Paralinhomoeus uniovarium sp. nov.
(Fig. 8)

MATERIAL STUDIED. One male and two females. B.M. (N.H.), Reg. No. 1968.
268.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank: M.L.W.S.T.

Body
a b c V% length (mm.)

Male 64-38 1675 14-11 2-06

Females 61-30 16-79 *5'&4 59-22 2-82

53-58 1775 13-33 59-5I 2-84

MEASUREMENTS (in mm. in order of body lengths). Male : Body breadth :
0-032. Oesophagus length : 0-123. Distance of nerve ring from anterior : 0-084.
Distance of excretory pore from anterior : 0-072. Head diameter : 0-018. Length
of longer cephalic setae : 0-018. Length of shorter cephalic setae : 0-013. Length
of cervical setae : 0-007. Amphid length : 0-016. Amphid breadth : 0-013. Body
diameter at level of amphids : 0-018. Distance of amphids from anterior : o-on.
Tail length : 0-146. Cloacal diameter : 0-021. Spicule length : 0-023. Guber-
naculum length : 0-012.

Females : Body breadth : 0-046 ; 0-053. Oesophagus length : 0-168 ; 0-160.
Distance of nerve ring from anterior : 0-097 ; 0-098. Distance of excretory pore from
anterior : 0-088 ; 0-084. Head diameter : 0-027 '< 0< 2O. Length of longer cephalic
setae : 0-019 i 0-018. Length of shorter cephalic setae : 0-012 ; 0-012. Length of
cervical setae : 0-005 '> 0-007. Amphid length : 0-017 i 0-017. Amphid breadth :
0-013 ; 0-014. Body diameter at level of amphids : 0-025 i 0-024. Distance of
amphids from anterior : 0-013 '> 0-013. Tail length : 0-178 ; 0-213. Anal diameter :
0-023 J 0-023. Distance of vulva from anterior : 1-67 ; 1-69. Egg length : ;
0-142. Egg breadth : ; 0-041.

The body is long and slender. The cuticle is marked by fine transverse striations
commencing at the level of the amphids and extending almost to the tail-tip. The
mouth is surrounded by six low lips each bearing a small conical labial papilla. There
are four long cephalic setae and six shorter ones. In the male the cervical setae
commence well anterior on the body, and what appears to be a pair of lateral cephalic
setae in Fig. 8A is in fact one cephalic and one cervical seta. The buccal cavity is
cyathiform and rather weakly cuticularized (Figs. 8A and 8B). The amphids are
slightly subcircular, elongate longitudinally, and are broken at the base for the exit
of the amphidial nerve. They are concave, but have a circular raised region in the
centre.

FOURTEEN NEW SPECIES OF MARINE NEMATODES

157

FIG. 8. Paralinhomoeus uniovarium sp. nov. A, Lateral view of male head. B, Lateral
view of female head, c, Lateral view of spicules and gubernaculum. D, Lateral view of
female tail. E, Anterior end of male. F, Lateral view of male tail.

158 R. M. WARWICK

The oesophagus is slightly swollen at its base (Fig. 8E), and the ventral gland
extends just beyond this swelling. There are a few scattered setae in the cervical
region, and on the male tail, but otherwise the body is naked. The tail is long, and
in the male the proximal three-quarters is tapering and the distal quarter cylindrical
(Fig. 8F). The female tail is more evenly tapered down its entire length (Fig. 8D).

MALE. The gubernaculum is roughly triangular in lateral view, with a pair of
weakly-cuticularized postero-dorsal apophyses (Fig. 8C). The spicules are short and
strongly arcuate, very slightly cephalate proximally and pointed distally. The
testes are not visible.

FEMALE. There is a single long anterior outstretched ovary, and the posterior
one is reduced to a small post-vulvular sac.

DISCUSSION. Species are known from this genus both with paired and single
ovaries. However, a division of the genus on this basis cannot be made at the
present time since the form of the female reproductive system has not been noted in
all species descriptions, and cannot always be inferred from the position of the
vulva alone. Further, some species are described from males only. The present
species is in many respects intermediate between Paralinhomoeus de Man, 1907, and
Halinema Cobb, 1920. It has large amphids with a central raised portion, as in
Halinema, but the cephalic setae are shorter and there are neither subventral rows
of conoid appendages nor long terminal setae on the male tail. Paralinhomoeus
uniovarium is distinguished from other members of the genus by the form and size
of the amphids, by the length of the cephalic setae and by the shape of the tail.

Family MONHYSTERIDAE

Theristus (Theristus) denticulatus sp. nov.

(Fig- 9)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 269.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.

Body
a b c V% length (mm.)

Males 23-14 4-05 7-17 0-81

28-00 4-41 6-91 0-56

25-91 4-22 6-71 0-57

Females 26-67 4-28 6-78 57-50 0-80

22-92 4-20 7-05 60-00 0-55

21-85 4-31 7-n 57'6i o-59

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-035 ' 0-020 ; 0-022. Oesophagus length : 0-200 ; 0-127 ; 0-135. Distance of nerve

FOURTEEN NEW SPECIES OF MARINE NEMATODES

159

FIG. 9. Theristus (Theristus) denticulatus sp. nov. A, Lateral view of male head. B,
Lateral view of female tail, c, Lateral view of female head. D, Lateral view of cloacal
region in male. E, Lateral view of tail of another male. F, Ventral view of male cloacal
region.

160 R. M. WARWICK

ring from anterior : 0-123 J 0-073 ; 0-079. Head diameter : 0-009 > 0-007 '> 0-007.
Length of cephalic setae : 0-007 '> 0-004 i 0-004. Length of longest cervical setae :
0-023 ; 0-012 ; 0-015. Amphid width : 0-008 ; 0-005 > 0-005. Body diameter at
level of amphids : 0-013 ; 0-008 ; 0-008. Distance of amphids from anterior :
0-042 ; 0-017 > 0-021. Tail length : 0-113 ; 0-081 ; 0-085. Cloacal diameter :
0-026 ; 0-020 ; 0-022. Length of right spicule : 0-026 ; 0-025 > 0-026. Length of
left spicule : 0-032 ; 0-032; 0-032. Gubernaculum length : 0-017 > ' OI 5 i 0-017.

Females : Body breadth : 0-030 ; 0-024 0-027. Oesophagus length : 0-187 '>
0-131 ; 0-137. Distance of nerve ring from anterior : 0-107 ; 0-080 ; 0-079.
Head diameter : o-on ; 0-007 '> 0-008. Length of cephalic setae : 0-006 ; 0-003 i
0-004. Length of longest cervical setae : 0-008 ; 0-013 ', 0-012. Amphid width :
0-005 ' 0-004 ; 0-005 ' Body diameter at level of amphids : 0-012 ; 0-009 ' 0-008.
Distance of amphids from anterior : 0-022 ; 0-022 ; 0-021. Tail length : 0-118 ;
0-078 ; 0-083. Anal diameter : 0-022 ; 0-019 '> 0-017. Distance of vulva from
anterior : 0-46 ; 0-33 ; 0-34. Egg length : ; ; 0-042. Egg breadth : ; ;
0-021.

The head end is attenuated considerably from a point about half-way down the
oesophagus-length. There are ten cephalic setae, and the submedian pairs are only
slightly subequal in length. The labial sense organs are not detectable. The
structure of the buccal cavity is typical of the genus. The amphids are slightly
subcircular and elongated longitudinally. They are situated well posterior on the
body, although their position is somewhat variable (Figs. 9 A & gC). The oesophagus
broadens considerably towards its posterior end in association with the attenuation
of the neck region. Longish setae, up to about three-quarters of the body-width, are
scattered over the entire body-surface. The tail is conical (Figs. 96 and gE).

MALE. At the distal end of the gubernaculum is a complex array of structures.
There are two flat lateral plates which are triangular when viewed from the ventral
aspect, and covered ventrally with minute denticles. In a more median position
there is a pair of laterally-curving points, and also three more slender points, one
passing anteriorly between the distal tips of the spicules and the other two pointing
posterolaterally (Figs. gD, gE & gF). Proximally the gubernaculum bears a short
triangular dorsally, or slightly posterodorsally, pointing apophysis. The spicules
are unequal in size and structure. The right one is the shorter and is sharply bent
in the middle with a narrow rounded slightly cephalate proximal tip and a pointed
distal tip. The left spicule is less sharply bent. There is a file of stout conical spines
arranged ventrally down the length of the tail. The structure of the gonads is not
clear.

FEMALE. There is a single pre-vulvular outstretched ovary and a small post-
vulvular sac.

DISCUSSION. The size of the body, position of amphids, etc., are so variable in
this species that at first more than one species was thought to be present. However,
the measurements and fine structure of the various parts of the male copulatory
apparatus proved to be so constant that the specimens are considered monospecific.
It is probable, therefore, that the position and size of the amphids are of little value

FOURTEEN NEW SPECIES OF MARINE NEMATODES 161

in delimiting species among the homogeneous group of species to which T. denticulatus
belongs. This group contains species of the subgenus Theristus with unequal spicules
and comprises T. heterospiculum Allgen, 1932, T. heterospiculoides Gerlach, 1952, T.
diversispiculum Gerlach, 1953, T. asymetricus (Wieser, 1956), T. wimmeri Wieser,
1959, T. rhynchonemoides Hopper, 1961, and T, roscoffiensis Vitiello, 1967. T.
problematical (Allgen, 1927) is a doubtful species, and Wieser's (1956) redescription has
been named by Vitiello (1967) as T. asymetricus. The present specimens differ from
all these species in the complex structure of the gubernaculum and the ventral spines
on the male tail.

Theristus (Theristus) inter stitialis sp. nov.
(Fig. 10)

MATERIAL STUDIED. Three males. B.M. (N.H.), Reg. No. 1968. 270.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.H.W.S.T., M.L.W.N.T.
Orcombe Point : M.L.W.S.T.

Body
a b c length (mm.)

Males 55-00 6-53 979 1-43

53-33 6-32 9-66 1-44

51-11 6-19 9-65 1-38

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-026 ; 0-027 '> 0*027. Oesophagus length : 0-219 ; 0-228 ; 0-223. Distance of nerve
ring from anterior : 0-112 ; 0-102 ; 0-120. Head diameter : 0-021 ; 0-018; 0-019.
Length of long cephalic setae : 0-021 ; 0-019 ; 0-019. Length of medium cephalic
setae : 0-017 ; 0-017 0-017. Length of short cephalic setae : 0-008 ; 0-007 ' 0-009.
Amphid diameter : 0-009 ' o-oio ; 0-009. Body diameter at level of amphids :
0-024 > 0-022 ; 0-023. Distance of amphids from anterior : 0-016 ; 0-012 ; 0-017.
Tail length : 0-146 ; 0-149 I 0<I 43- Cloacal diameter : 0-026 ; 0-027 ; 0-027. Spicule
length : 0-042 ; 0-039 '< 0' 39- Gubernaculum length : 0-081 ; 0-017 ; 0-016.

The cuticle is fairly coarsely striated. The labial setae are prominent and conical,
5^ in length. There are 14 cephalic setae, the submedian pairs being only slightly
subequal and the lateral triplets having an additional much shorter seta (Fig. loA).
Body setae are sparse, short and scattered, becoming more numerous on the tail.
The buccal cavity and amphids are typical of the genus in structure. The oesophagus
is more or less cylindrical throughout its length. The tail is conical (Fig. loC).

MALE. The gubernaculum has a pair of short rounded caudo-dorsal apophyses
proximally, and a pair of lateral projections distally which are each tipped with two
teeth (Figs. loB and zoD). The spicules are paired and equal, strongly cephalate
proximally, but distally obscured by the gubernaculum.

1 62

R. M. WARWICK

DISCUSSION. This species belongs to group B.2. in the key of Wieser (1956), and
is distinguished from other members of the group by the long cephalic setae, by
the position and size of the amphids and by the structure of the male copulatory
apparatus.

FIG. 10. Theristus (Theristus) interstitialis sp. nov. A, Lateral view of male head. B,
Lateral view of spicules and gubernaculum. c, Lateral view of male tail. D, Ventral
view of spicules and gubernaculum.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 163

Theristus (Trichotheristus) psammoides sp. nov.
(Fig. ii)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 271.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.

Body
a b c V% length (mm.)

Males 35-63 4-38 870 1-14

36-06 4-58 7-99 1-19

32-29 3-90 8-50 1-13

Females 30-00 4-31 8-63 63-04 1-38

29-74 4-04 7-96 60-18 1-13

28-72 3-86 10-00 64-29 1-12

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth : 0-032 ;
0*033 J '35- Oesophagus length : 0-26 ; 0-26 ; 0-29. Distance of nerve ring from
anterior : 0-104 '> 0-109 '> 0>II 5- Head diameter : 0-017 ) ' OI 5 ', 0-016. Length of
long cephalic setae : o-oii ; o-oii ; 0-012. Amphid diameter : o-oio ; o-oii ; 0-012.
Body diameter at level of amphids : 0-022 ; 0-021 ; 0-022. Distance of amphids from
anterior : 0-018 ; 0-017 ; 0-021. Length of longest body setae : 0-024 0-023 ; 0-029.
Tail length : 0-131 ; 0-149 '> ' I 33- Cloacal diameter : 0-028 ; 0-028 ; 0-027. Spicule
length : 0-041 ; 0-041 ; 0-043. Gubernaculum length : 0-016 ; 0-014 ; 0-017.

Females : Body breadth : 0-046 ; 0-038 ; 0-039. Oesophagus length : 0-32 ; 0-28 ;
0-29. Distance of nerve ring from anterior : 0-126 ; ? ; ?. Head diameter : 0-017 '>
0-016 ; 0-017. Length of long cephalic setae : 0-013 ; 0-013 ; 0-014. Amphid dia-
meter : 0-007 I 0-008 ; 0-009. Body diameter at level of amphids : 0-025 '> 0-024 ;
0-023. Distance of amphids from anterior : 0-017 ; 0-017 ; 0-020. Length of longest
body setae : 0-037 ; 0-036 ; 0-037. Tail length : 0-160 ; 0-142 ; 0-122. Anal
diameter : 0-031 ; 0-026 ; 0-029. Distance of vulva from anterior: 0-87 ; 0-68 ; 0-72.

There are six small slender labial setae and ten cephalic setae, of which four are
very slightly shorter than the remaining six. In some specimens four additional
very short setae (about 4/1) can be detected in association with the submedian pairs
(Fig. 1 1 A). The amphids are perfectly circular, and are larger in the males than
in the females. Longish setae are scattered generally over the body-surface, but are
most numerous in the oesophageal region and on the tail. The tail is filiform for its
distal two-thirds (Figs. nB & iiC).

MALE. The gubernaculum has a small rounded dorsally-directed apophysis and
a pair of rounded lateral projections at the distal end (Figs. nD & nE). The spi-
cules are equal, arcuate, cephalate proximally and strongly hooked distally. The
structure of the gonads is not clear.

FEMALE. There is a single anterior outstretched ovary and a small spherical
post-vulvular sac.

16 4

R. M. WARWICK

FOURTEEN NEW SPECIES OF MARINE NEMATODES 165

DISCUSSION. This species is closest to T. vicinus Riemann, 1966, but differs from
it in having larger amphids and a different structure of the male copulatory apparatus.
It is also close to T. erectus Wieser & Hopper, 1967, from which it differs in having
larger amphids, no lateral alae and in a differently-shaped gubernaculum.

Family SPIRINIDAE

Chromaspirina inglisi sp. nov.
(Fig. 12)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 272.

DISTRIBUTION IN EXE ESTUARY. Orcombe Point : M.L.W.S.T.

Body
a b c V% length (mm.)

Males 46-00 9-94 20-64 1-61

43.78 10-19 19-52 1-62

48-82 10-00 !9*53 1*66

Females 38-72 10-34 18-64 5&'94 i'5i

46-15 ii-ii 22.78 62-78 1-80

42-56 11-09 22-32 63-39 I>8 3

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0*035 I 0-037 '34- Oesophagus length : 0-162 ; 0-159 ' ' I 66. Distance of
nerve ring from anterior : 0-095 ; 0-099 '097. Head diameter : 0-020 ; 0-014 1
0-016. Length of shorter cephalic setae : 0-006 ; 0-006 ; 0-006. Length of longer
cephalic setae : 0-012 ; o-on ; o-on. Amphid diameter : 0-009 ' o-oio ; o-on.
Body diameter at level of amphids : 0-023 '> 0-022 ; 0-023. Length of oesophageal
bulb : 0-048 ; 0-045 ; 0-043. Breadth of oesophageal bulb : 0-028 ; 0-028 ; 0-026.
Tail length 10-078 ; 0-083 '> 0-085. Cloacal diameter : 0-031 ; 0-033 '> 0-032. Spicule
length : 0-048 ; 0-054 i 0-052. Gubernaculum length : 0-025 '> 0-022 ; 0-023.

Females : Body breadth : 0-039 '39 '> '43- Oesophagus length : 0-146 ;
0-162 ; 0-165. Distance of nerve ring from anterior : 0-082 ; ? ; 0-102. Head dia-
meter : 0-014 ; 0-019 ) 0-020. Length of shorter cephalic setae : 0-005 i 0-006 ;
0-006. Length of longer cephalic setae : 0-013 J o-on ; 0-012. Amphid diameter :
o-oii ; 0-012 ; o-oio. Body diameter at level of amphids : 0-022 ; 0-026 ; 0-027.
Length of oesophageal bulb : 0-043 ; 0-043 ; 0-051. Breadth of oesophageal bulb :
0-028 ; 0-030 ; 0-026. Tail length : 0-081 ; 0-079 5 0-082. Anal diameter : 0-029 '>
0-028 ; 0-032. Distance of vulva from anterior : 0-89 ; 1-13 ; 1-16. Egg length :
0-123 J Egg breadth : 0-024 '< '>

FIG. ii. Theristus (Trichotheristus) psammoides sp. nov. A, Lateral view of male head.
B, Lateral view of male tail, c, Lateral view of female tail. D, Ventral view of male
cloacal region. E, Lateral view of spicules and gubernaculum.

1 66

R. M. WARWICK

The head is invaginated in most specimens. Those with a head diameter of 0-019-
0-020 mm. are the least invaginated and this should be taken as the correct head-
diameter. The cuticle is marked by fine transverse striations commencing at the

FIG. 12. Chromaspirina inglisi sp. nov. A, Lateral view of male tail. B, Lateral
view of male head, c, Lateral view of female tail. D, Oesophageal bulb.

anterior border of the amphids and extending to the spinerette on the tail. The body-
surface is covered uniformly with very fine hairs, which commence a short distance
posterior to the head. These are probably cuticular in origin, but the possibility
that they are epigrowths of unicellular cyanophyecous algae cannot be excluded,
since these are known to form associations with other species of nematode (e.g.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 167

Leptonemella species). The lips and labial sense-organs are not visible, because of
the invagination of the head. There is an anterior circle of six short cephalic setae,
and four longer setae just posterior to them (Fig. 126). There is a ring of indistinct
odontia present in the anterior portion of the buccal cavity, and the posterior portion
contains a prominent heavily-cuticularized dorsal tooth opposed by a pair of very
small sub ventral projections. The amphids describe a single rounded loop.

The oesophagus has a pyriform posterior bulb, the cuticular lining of which is not
markedly thickened (Fig. I2D). Scattered setae between 0-005 an( i ' OI 3 mm -
in length extend from the head to about 30% of the way down the oesophagus.
The middle region of the body is virtually devoid of setae, but they become fairly
numerous on the tail, especially in the male. The tail is conoid, and the spinerette
unstriated (Figs. 12 A & I2C).

MALE. The gubernaculum is slender and crescentic. It is closely applied to the
spicules. The spicules are arcuate, strongly cephalate proximally, and have a
prominent ventral ala (Fig. 12 A). A small pore is present in the cuticle a short
distance in front of the cloaca, but otherwise supplements are absent. The testis
appears to be single and outstretched.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed. The eggs
are very elongate.

DISCUSSION. This species is very similar to C. pontica Filipjev, 1918, sensu Ger-
lach, 1951, but as Wieser & Hopper (1967) point out the conspecificity of Gerlach's
and Filipev's species is doubtful. In the type the amphids are smaller, the cephalic
setae only 5/1 in length and there are some differences in the structure of the guber-
naculum. The present specimens agree with Gerlach's in all respects, except for the
uniform covering of fine hairs, which could easily have been overlooked. Otherwise
they would appear to belong to the same species.

Family MICROLAIMIDAE

Microlaimus spirifer sp. nov.

(Fig. 13)

MATERIAL STUDIED. Three males and one female. B.M. (N.H.), Reg. No.
1968. 273.

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.

Orcombe Point : M.H.W.N.T., M.T.L.

Body
a b c V% length (mm.)

Males 36-55 7-31 11-16 1-06

38-00 6-69 9-79 0-95

36-67 6-24 8-54 0-88

Female 37-67 6-48 9-68 53'26 0-92

MEASUREMENTS (in mm. in order of body lengths). Males: Body breadth :

1 68

R. M. WARWICK

B

FOURTEEN NEW SPECIES OF MARINE NEMATODES 169

0-029 i 0-025 I 0-024. Oesophagus length : 0-145 ; 0-142 ; 0-141. Distance of
nerve ring from anterior : 0-066 ; 0-077 ; 0-087. Distance of excretory pore from
anterior : 0-040 ; ? ; ?. Length of oesophageal bulb : 0-032 ; 0-032 ; 0-032. Breadth
of oesophageal bulb : 0-021 ; 0-018 ; 0-018. Head diameter : 0-009 '> 0*009 0-008.
Length of longer cephalic setae : 0-006 ; 0-005 > 0-005. Length of shorter cephalic
setae : 0-003 > 0-003 i 0-003. Length of buccal cavity : o-oio ; 0-009 o-on.
Amphid diameter : 0-009 '> 0-009 ' 0-008. Body diameter at level of amphids :
0-014 > 0' 01 3 ; 0-012. Distance of amphids from anterior : 0-027 '> 0*024 ' 0-024.
Tail length : 0-095 ; 0-097 ; 0-103. Cloacal diameter : 0-017 '> 0-018 ; 0-018.
Spicule length : 0-025 ', 0-023 i 0-022. Gubernaculum length : 0-008 ; 0-007 ;
0-006.

Female : Body breadth : 0-030. Oesophagus length : 0-142. Distance of nerve
ring from anterior : 0-083. Length of oesophageal bulb : 0-031. Breadth of oeso-
phageal bulb : 0-020. Head diameter : 0-008. Length of longer cephalic setae :
0-006. Length of shorter cephalic setae : 0-003. Length of buccal cavity : o-on.
Amphid diameter : 0-008. Body diameter at level of amphids : 0-013. Distance of
amphids from anterior : 0-024. Tail length : 0-095. Anal diameter : 0-016.
Distance of vulva from anterior : 0-49.

The cuticle is transversely striated. The head has the usual arrangement of sense
organs ; a ring of six small conical labial papillae, an anterior ring of six short cephalic
setae and a posterior ring of four longer setae (Fig. I3A). The buccal cavity is
fairly well cuticularized. It contains a dorsal and ventral tooth of about equal size,
the dorsal one being the more anterior. There is a very small posterior cavity set off
from the main buccal cavity by a constriction. The amphids are spiral, describing
about 6-5 turns, the outer margin of the amphid being more prominent than the inner
spiral. There are four longish setae situated at about the level of the excretory pore,
but otherwise the body is naked. The oesophagus terminates in a pyriform bulb.
The tail is long and conoid (Figs. 136 and 130).

MALE. The gubernaculum is very small, and the spicules are only slightly bent,
rounded proximally and pointed distally. There are no supplements.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed.

DISCUSSION. This species is characterized by the posterior position and spiral
form of the amphids, by the form of the copulatory apparatus in the male, and by
the long tail.

Family GYATHOLAIMIDAE

Pomponema reducta sp. nov.
(Fig. 14)

MATERIAL STUDIED. Three males and three females. B.M. (N.H.), Reg. No.
1968. 274.

FIG. 13. Microlaimus spirifer sp. nov. A, Lateral view of male head. B, Lateral
view of male tail, c, Lateral view of female tail.

1 7 o

R. M. WARWICK

DISTRIBUTION IN EXE ESTUARY. Shelly Bank : M.L.W.S.T. Orcombe Point :
M.H.W.N.T., M.L.W.S.T.

v%

Males

Females

46-25

43-44
46-25

33-33

42-65

40-79

7-05
6-18

6-95

6-22

6-33

6-22

10-96

10-67

9-67

10-94
10-51

IO-2O

61-79
61-38

59-35

Body
length (mm.

1-48

i-39

1-48

1-40
i-45
i-55

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-032 ; 0-032 ; 0-032. Oesophagus length : 0-210 ; 0-225 ' " 2I 3- Distance of nerve
ring from anterior : ? ; 0-093 ; ?. Distance of excretory pore from anterior : 0-117 '>
? ; ?. Head diameter : 0-023 '< 0-022 ; 0-024. Length of labial setae : 0-004 ;
0-004 > 0-004. Length of cephalic setae : o-oio ; 0-009 ' o-oio. Amphid diameter :
o-oio ; 0-009 > o-oio. Body diameter at level of amphids : 0-024 ' 0-024 > 0-026.
Tail length : 0-135 \ 0-138 ; 0-153. Cloacal diameter : 0-030 ; 0-025 > 0-028. Spi-
cule length : 0-043 ; 0-046 ; 0-046. Gubernaculum length : 0-023 I 0-026 ; 0-027.
Distance of anterior supplement from cloaca : 0-330 ; 0-315 ; 0-295.

Females : Body breadth : 0-042 ; 0-034 '> 0-038. Oesophagus length : 0-225 I
0-229 ' 0-249. Distance of excretory pore from anterior : ? ; 0-127 J ? Head dia-
meter : 0-023 > 0-024 J 0-024. Length of labial setae : 0-004 0-004 ; 0-004. Length
of cephalic setae : 0-008 ; 0-009 ; o-oio. Amphid diameter : 0-008 ; 0-007 0-007.
Body diameter at level of amphids : 0-026 ; 0-025 ; 0-026. Tail length : 0-128 ;
0-138 ; 0-152. Anal diameter : 0-023 '> 0-024 ; 0-025. Distance of vulva from
anterior : 0-865 ; 0-89 ; 0-92.

The cuticle is marked by transverse rows of rounded punctuations, and there is a
marked lateral differentiation. Punctuations on the head are small, sparse and
irregularly arranged, but just posterior to the amphids they become larger and closer
together, although still not arranged in definite rows (Figs. I4A & 146). Along the
length of the oesophagus the markings begin to sort themselves into distinct trans-
verse rows, and at the base of the oesophagus the lateral differentiation begins. This
comprises four files of larger dots, which gradually increase in size down the length of
the body. They are very large and rather irregular in shape in the anal or cloacal
region, and extend about two-thirds of the way down the tail-length.

The mouth is surrounded by six small rounded lips each bearing a stout conical
labial seta. There are also twelve slender and much shorter setae in positions cor-
responding with the tips of the buccal rugae (Figs. I4A & 146). There appear to be
only six cephalic setae, each consisting of a stout basal half and a filiform tip. The
four shorter cephalic setae reported from all other species of Pomponema appear to
be absent, unless they are very closely adherent to the larger ones. This reflects the
trend begun in P. stomachor Wieser, 1954 and P. polydonta Murphy, 1963, in which

FOURTEEN NEW SPECIES OF MARINE NEMATODES

171

FIG. 14. Pomponema reducta sp. nov. A, Lateral view of female head. B, Ventral view of
female head, c, Posterior end of male. D, Lateral view of female tail. E, Ventral view
of spicules and gubernaculum. F, Lateral view of cloacal region in male.

172 R. M. WARWICK

these four setae are very reduced. The buccal cavity is of the typical form. An-
teriorly its walls are supported by twelve buccal rugae. More posteriorly there is a
large dorsal tooth opposed by a pair of much smaller subventral teeth and additional
groups of denticles arranged as in Figs. I4A and 146. The amphids describe a spiral
of 4-4-5 turns in the male and 3-5 turns in the female.

The oesophagus has a distinct anterior pharyngeal bulb dorsally, and broadens
gradually towards its posterior end, with no definite posterior bulb. There are no
somatic setae. The tail is long and slender, its proximal half tapering and distal
half filiform (Figs. 146 & 140).

MALE. The gubernaculum consists of two identical halves each having an out-
wardly-curving point distally and a rounded lateral projection near the middle
(Figs. I4E & I4F) . The spicules are arcuate and have small ventral alae. There are
20-24 equally spaced pre-cloacal supplements, the cuticle between them being
lamellated and giving them the complex appearance which is characteristic of the
genus. The testes are not visible.

FEMALE. The ovaries are paired, more or less symmetrical, opposed and reflexed.

DISCUSSION. This species is closest to P. polydonta Murphy, 1963, but differs
from it in several respects. The absence of the four shorter cephalic setae has
already been noted. In P. polydonta the longer cephalic setae are 23/1 long as com-
pared with 8-io/x in P. reducta. These setae are similar in structure, since Murphy
states for P. polydonta that 'there is indication of segmentation and articulation of
larger setae on some specimens at a point one-third to one-half of the length from the
base'. The buccal denticles in P. polydonta are arranged in two rows, and not in
discrete groups as in the present species. The spicules in P. polydonta appear to be
non-alate.

Paracanthonchus opheliae sp. nov.
(Figs. 15 & 16)

MATERIAL STUDIED. Three males and two females. B.M. (N.H.), Reg. No.
1968. 275.

DISTRIBUTION IN EXE ESTUARY. Shelley Bank : M.T.L., M.L.W.N.T., M.L.W.S.T.
Orcombe Point : M.T.L., M.L.W.N.T.

Body
a b c V% length (mm.)

Males 32-92 9-03 10-09 2-14

37-12 9-13 10-95 2-19

34-33 9-62 10-70 2-30

Females 20-65 9-41 12-57 50-00 2-54

18-46 10-31 11-09 49-62 2-64

FOURTEEN NEW SPECIES OF MARINE NEMATODES 173

MEASUREMENTS (in mm. in order of body lengths). Males : Body breadth :
0-065 I 0' 59 ,* 0-067. Oesophagus length : 0-237 0-240 ; 0-239. Distance of
nerve ring from anterior : 0-140 ; 0-141 ; 0-137. Distance of excretory pore from
anterior : 0-054 > 0-040 ; ?. Length of longer cephalic setae : 0-018 ; 0-016 ; 0-018.
Length of shorter cephalic setae : o-oio ; 0-009 > 0-008. Head diameter : 0-031 ;
0-038 ; 0-035. Amphid diameter : 0-014 ; 0-014 ; 0-014. Body diameter at level
of amphids : 0-039 ; 0-041 ; 0-040. Tail length : 0-212 ; 0-200 ; 0-215. Cloacal
diameter : 0-056 ; 0-055 '> 0-063. Spicule length : 0-059 > O'56 ; 0-071. Guberna-
culum length : 0-067 ; 0-064 ; 0-065.

Females : Body breadth : 0-123 I 0>I 43- Oesophagus length : 0-270 ; 0-256 ;
Distance of nerve ring from anterior : 0-177 ? Distance of excretory pore from
anterior : 0-052 ; ?. Length of longer cephalic setae : 0-017 > 0-018. Length of
shorter cephalic setae : 0-008 ; o-oio. Head diameter : 0-038 ; 0-053. Amphid
diameter : 0-008 ; 0-009. Body diameter at level of amphids : 0-055 > '59- Tail
length : 0-202 ; 0-238. Anal diameter : 0-052 ; 0-074. Distance of vulva from
anterior : 1-27 ; i'3i.

This is a relatively large species. The cuticle is marked by transverse rows of small
rounded punctations which commence at the level of the bases of the cephalic
setae. There is no lateral differentiation. Type-i campaniform organs (see Inglis,
1963) are found in two lateral files down each side of the body, commencing just
posterior to the amphids and extending almost to the tip of the tail. In the middle
of the body these files are about i6ji apart. There are six files of setae down the
length of the body, and these are shorter and more widely spaced posterior to the
base of the oesophagus.

In all specimens the head is somewhat invaginated, so that the labial sense-organs
cannot be seen. The arrangement of cephalic setae is typical of the genus, with an
anterior ring of six short setae and more posteriorly four longer setae. The buccal
cavity is cyathiform and contains twelve buccal rugae. The dorsal tooth is rela-
tively small (Fig. 15 A). The amphids in the male describe a spiral of 4 turns and in
the female 3-5 turns. There are two large ocelli positioned dorsolaterally 1-1-5.
head diameters from the anterior. They each consist of an anterior lensatic unit
and a posterior chomatic unit containing dark blackish pigment granules. The
pigment can only be seen in living or freshly killed specimens, and is removed com-
pletely on clearing in glycerine.

The oesophagus is more or less cylindrical throughout its length. Numerous
large pennate diatoms were seen in the gut of several specimens. The tail is conical
(Figs. 150 & 16) and in the male bears two files of longish ventrolateral setae.

MALE. The spicules are curved and have broad ventral alae. They are pointed
distally and the proximal end has an undulating appearance (Figs. 156 & I5C). The
gubernaculum is a complex paired structure, the distal end of which is set off from
the long proximal portion as a pair of massive swellings which appear rectangular
in lateral view. Each swelling bears three large teeth, the arrangement of which is
shown in Figs. 156 & 156. Anteriorly two small protuberances project from the
distal portion of the gubernaculum on either side of the spicules. The proximal
part of the gubernaculum consists of two long club-shaped structures which lie very

174

R. M. WARWICK

B

FIG. 15. Paracanthonchus opheliae sp. nov. A, Lateral view of male head. B, Ventral
view of spicules and gubernaculum. c, Lateral view of spicules and gubernaculum. D,
Lateral view of female tail.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 175

FIG. 1 6. Paracanthonchus opheliae sp. nov., lateral view of male tail.

176 R. M. WARWICK

close to one another. They bear narrow alae along their dorsal edges. There are
seven tubular pre-cloacal supplements arranged as in Fig. 16. The testes are paired,
opposed and outstretched, the anterior one commencing about one-third of the way
down the length of the body.

FEMALE. The ovaries are paired, symmetrical, opposed and reflexed. The eggs
are roughly spherical, about 6^/j. in diameter.

DISCUSSION. This species belongs to group 'B' in the key of Wieser (1954). It
differs from all other species assigned to this group in the structure of the guber-
naculum and spicules.

ACKNOWLEDGEMENTS

The main part of this work was conducted at the University of Exeter under the
supervision of Professor L. A. Harvey, for whose advice and encouragement I am
very grateful. I should also like to thank Dr. W. Grant Inglis and Mr. John W.
Coles of the British Museum (Natural History) for my initial instruction in nematode
taxonomy and for the continued assistance and hospitality they have afforded me on
subsequent occasions. I am indebted to Mr. S. Prudhoe for critically reading and
correcting the manuscript. The work was supported by a grant from the Natural
Environment Research Council.

REFERENCES

ALLGEN, C. A. 1927. Freilebende marine Nematoden von den Campbell-und Staten-inseln.
Nyt. Mag. Naturvid. 66 : 249-309.

1932. Weitere Beitrage zur Kenntnis der marinen Nematodenfauna der Campbellinseln.

Ibid. 70 : 97-198.

CHITWOOD, B. G. 1935. A new nematode, Camacolaimus prytherchi, n. sp. (Camacolaimidae) .

Proc. helminth. Soc. Wash. 2 : 49-50.
COBB, N. A. 1920. One hundred new nemas. (Type species of 100 new genera). Contr. Sci.

Nematology 9 : 217-343.
DE CONINCK, L. A. P. 1965. Systematique des nematodes. Traite de zoologie. Anatomie,

systematique, biologique. 4(2) : 586-681.
DITLEVSEN, H. 1930. Papers from Dr. Mortensen's Pacific Expedition 1914-16. No. 3.

Marine freeliving nematodes from the Aukland and Campbell Islands. Vidensk. Meddr

dansk naturh. For en. 73 : 1-32.
FILIPJEV, I. N. 1918. Freeliving nematodes from the region of Sevastopol I. Trav. Lab.

Zool. Sebastopol. 2(4) : 1-350. (In Russian).
GERLACH, S. A. 1951. Revision der Metachromadoracea, einer Gruppe freilebender mariner

Nematoden. Kieler Meeresforsch. 8 : 59-75.

1952. Die Nematodenbesiedlung des Sandstrandes und des Kiistengrundwassers an der

italienischen Kiiste. Archo zool. ital. 37 : 519-640.

I953- Freilebende marine Nematoden aus dem Kustengrundwasser und dem Brackwasser

der chilenischen Kiiste. Acta Univ. lund. N.F. 49(io) : 1-37.
HOPPER, B. E. 1961. Marine nematodes from the coast line of the Gulf of Mexico. Can. J.

Zool. 39 : 183-199.
INGLIS, W. G. 1963. 'Campaniform-type' organs in nematodes. Nature, Lond. 197 : 618.

1964. The marine Enoplida (Nematoda) : a comparative study of the head. Bull. Br.

Mus. nat. Hist. (Zoology). 11 : 263-376.

1966. Marine nematodes from Durban, South Africa. Ibid. 14 : 81-106.

FOURTEEN NEW SPECIES OF MARINE NEMATODES 177

INGLIS, W. G. 1967. Interstitial nematodes from St. Vincent's Bay, New Caledonia. Expdn.

Franfaise Recifs Corall. N.-Caledonie 1960-1963. (Ed. Fondation Singer- Polignac, Paris).

2 : 29-74.
DE MAN, J. G. 1907. Sur quelques especes nouvelles ou peu connues de Nematodes libres

vivant sur les cotes de la Zelande. Tijdschr. ned. dierk. Ver. 10 : 227-243.
MURPHY, D. G. 1963. A new genus and two new species of nematodes from Newport, Oregon.

Proc. helminth. Soc. Wash. 30 : 73-78.
RIEMANN, F. 1966. Die interstitielle Fauna im Elbe-Aestuar, Verbreitung und Systematik.

Arch. Hydrobiol. Suppl. 31 : 1-279.
SCHUURMANS STEKHOVEN, J. H. & DE CONINCK, L. A. P. 1933- Diagnosis of new Belgian

marine nemas. Bull. Mus. r. Hist. nat. Belg. 9 : 1-15.
TIMM, R. W. 1954. A survey of the marine nematodes of Chesapeake Bay, Maryland. Biol.

Stud. Cath. Univ. Am. 23 : 1-70.
VITIELLO, P. 1967. Nematodes libres marins de Roscoff. I. Description de cinq especes

nouvelles. Cah. Biol. mar. 8 : 403-416.
WIESER, W. 1953. Freeliving marine nematodes I. Enoploidea. Reports of the Lund

University Chile Expedition 1948-49. Ada Univ. lund. N.F. 49(6) : 1-155.

- 1954. Freeliving marine nematodes II. Chromadoroidea. Reports of the Lund Uni-
versity Chile Expedition 1948-49. Ibid. 50(i6) : 1-148.

1956. Freeliving marine nematodes III. Axonalaimoidea and Monhysteroidea. Reports
of the Lund University Chile Expedition 1948-49. Ibid. 53(i3) : 1-115.

- 1959. Freeliving marine nematodes and other small invertebrates of Puget Sound
beaches. Univ. Wash. Publs Biol. 19 : 1-179.

- & HOPPER, B. E. 1967. Marine nematodes of the East Coast of North America I.
Florida. Bull. Mus. comp. Zool. Harv. 135 : 239-344.

Dr. R. M. WARWICK
DOVE MARINE LABORATORY
CULLERCOATS, NORTHUMBERLAND

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

THE ANATOMY AND TAXONOMY

OF INDOSTOMUS PARADOXUS

PRASHAD & MUKERJI

K. E. BANISTER

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 5

LONDON: 1970

THE ANATOMY AND TAXONOMY

OF INDOSTOMUS PARADOXUS

PRASHAD & MUKERJI

BY

KEITH EDWARD BANISTER

Pp. 179-209 ; 15 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

ZOOLOGY Vol. 19 No. 5

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is issued
in five series corresponding to the Departments of the
Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 5 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HIOSTRY)

Issued 23 April 1970 Price Sixteen Shillings

(0-80)

THE ANATOMY AND TAXONOMY OF

INDOSTOMUS PARADOXUS PRASHAD

AND MUKERJI

By K. E. BANISTER

CONTENTS

Page

ABSTRACT ........... 181

INTRODUCTION .......... 182

Materials and methods . . . . . . . .182

Terminology .......... 182

THE OSTEOLOGY OF Indostomus paradoxus . . . . . . 183

Body armour .......... 183

Vertebral Column . . . . . . . . .184

Caudal fin .......... 185

Median fins and their skeletal supports . . . . . 185

Pectoral fin skeleton . . . . . . . . .186

Pelvic fin and skeleton . . . . . . . . 187

SYNCRANIUM ........... 188

Olfactory region . . . . . . . . .188

Orbital region .......... 190

Otic and occipital regions . . . . . . . .192

Oromandibular region . . . . . . . .192

Palatoquadrate arch . . . . . . . . 193

Opercular series . . . . . . . . .194

Hyoid arch .......... 194

Branchial skeleton . . . . . . . . .195

NOTES ON THE SOFT ANATOMY . . . . . . . .196

Jaw mechanism of Indostomus . . . . . . .196

Swimbladder .......... 197

DISCUSSION ........... 197

SUMMARY ........... 204

ACKNOWLEDGEMENTS ......... 204

APPENDIX i ........... 205

KEY TO FIGURES .......... 207

REFERENCES ....... ... 208

ABSTRACT

The osteology of Indostomus paradoxus Prashad and Mukerji is described and the evidence for
its affinities discussed. The array of characters common to both Indostomus and the higher para-
canthopterygii suggests a relationship between the two groups which is expressed by proposing
a new paracanthopterygian order Indostomiformes of which the only known representative
is Indostomus paradoxus.

182 K. E. BANISTER

INTRODUCTION

Indostomus paradoxus was discovered by the Indian Museum Expedition of 1926 in
Lake Indawgyi, a large shallow lake in the Myitkyina district of Upper Burma.
The series of specimens collected (about 30) was named and briefly described by
Prashad and Mukerji (1929). Most of the specimens were housed in the Indian
Zoological Survey in Calcutta, except for five which were presented to the British
Museum (Natural History) and at least one was presented to the United States
National Museum.

Bolin (1936) briefly redescribed the U.S. National Museum specimen but the
unavailability of further specimens resulted in the only two descriptions of this
species being superficial.

Materials and Methods

One of the five specimens at the B.M. (N.H.) [No. 1937.9.25 : 1-5] had been
prepared by the alizarin method but was not suitable for detailed examination. I
was, therefore, delighted to receive another specimen from Dr. Earl Herald of the
Steinhart Aquarium, California Academy of Sciences (CAS-ACC i963-xii-i6). This
fish was one of six which arrived at an aquarium fish importers in America and was
described as having come from " Mountain streams remote from Bangkok ".
Doubt has since been cast on this locality (Herald pers. comm.). I prepared this fish
by the alizarin method and dissected it.

Since then I have obtained about 140 specimens from Dr. Alan Tubb of the
F.A.O. in Thailand. Dr. Tubb collected them from a stream at the west side of
Lake Indawgyi in 1956, kept them alive in Aquaria and made invaluable notes on
their behaviour (see Appendix i). It is with great pleasure that I acknowledge his
kindness in making both his notes and specimens available to me.

Some of the specimens have been stained with alizarin and dissected as before. I
have also had prepared longitudinal and transverse serial sections, stained in Mallory's
triple stain. Other specimens have been presented to various museums throughout
the world.

The size range of the specimens was 21-27 mm. (standard length). The line
drawings of the bones were made from different specimens but all were within the
24-27 mm. s.l. range.

Terminology

The bone terminology generally follows Harrington (1955) although with only
adult specimens available I cannot be too certain about the dual origin of some bones ;
in these cases the dermo- and auto-prefixes will not be used. The terms dentary,
angular and articular are retained for the three bones of the lower jaw. I fully
accept the inaccuracies implied in this but it is a convenient labelling system (Weitz-
man 1962), especially in the absence of embryological information.

Nelson (1967) has been followed for the naming of the pharyngeal bones.

Greenwood et alii (1966) have been followed for classification above the genus
level. The names given to the muscles are merely indicative of their position and

TAXONOMY OF INDOSTOMUS PARADOXUS

183

function ; identity with muscles of the same name in other fishes is not necessarily
intended.

THE OSTEOLOGY OF INDOSTOMUS PARADOXUS

The body of Indostomus is very slender, slightly depressed and covered in scutes
(fig. i). The presence of this dermal armour has had a modifying effect upon the
axial skeleton and musculature.

FIG. i. Indostomus paradoxus Prashad and Mukerji. Lateral view.

Body Armour

In the abdominal region there is a median row of dorsal plates and two rows of
ventrolateral plates. The latter irregularly overlap in the ventral midline and their
dorsal edges overlap the ventral margins of the dorsal plates. In front of the last
dorsal fin ray the dorsal plates are convex, behind it they are flat. The plates are
also flattened on the ventral surface of the abdomen.

There are twenty-one dorsal plates (rarely twenty) corresponding in number, but
not always in position, to the vertebrae. The first plate is loosely connected to the
head by a fold of skin which allows the fish to move its head (see Appendix i). Each
of the next five plates (Nos. 2 to 6 inc.) bears at its centre a small mobile spine which
supports a flap of tissue. The depression in which each spine articulates is perforated
by a pair of tendons attached to the muscles fastened on to the median keel of the
plates. The articulation mechanism of one of the spines is shown in fig. 2.

FIG. 2. Diagram of the articulation of one of the dorsal spines.

The first ventrolateral plate has been incorporated into the pectoral fin skeleton
(see below). The anteroventral margin of the second has been expanded forwards
and lies below the pectoral fin. A notch in its posterior margin marks the position of

184 K. E. BANISTER

the articulation of the pelvic fin, which is on the third plate (fig. i). The first five
dorsal and anal fin rays articulate between the plates ; the sixth, like the isolated
dorsal spines, is in the centre of the plate. Posterior to the median fins the plates
become fused to give the body a quadrangular cross section. A series of cristae run
along the length of the body ; towards the caudal region they coalesce and emphasize
the rectangular cross section of the caudal region.

Vertebral column

There are 21 (rarely 20) vertebrae, of which nine are abdominal. Transverse
processes, which originate at the base of the anterior part of the neural arch, are
present on vertebrae 2-9 inclusive (figs. 3 and 4). Ribs are absent.

0-5mm

FIG. 3. Dorsal view of the first three vertebrae. For explanation of the abbreviations

see key on page 207.

The first vertebra (figs. 3, 4, and 12) is modified to allow for the depression and
elevation of the head (see Appendix i). The transverse processes are broad and
arise from the centrum. A part of the leading edge of each transverse process is
extended forward as a shallow socket to receive the exoccipital process dorsally
(figs. 4 and 12). The neural arch of the first vertebra is fused to the centrum and has
a sloping anterior face to permit the elevation of the head. This elevation is effected
by the contraction of epaxial muscles which are connected to the epiotics by ossified
tendons. I am unable to explain the mechanism which causes the lateral flexures of
the head noticed by Dr. Tubb (see Appendix i).

The transverse processes on the abdominal vertebrae do not contact the dermal
plates. The neural arches of the abdominal vertebra (fig. 4) have a complex inter-
digitation with each other ; neural spines are not formed.

The last abdominal vertebra (the ninth) has short, thin transverse processes and
the posterolateral edges of the centrum develop as small wings which articulate with
the front edges of the haemal arch of the tenth vertebra. Vertebrae 11-21 (or 20)
have neural and haemal arches which become progressively shallower and after the
fourteenth vertebra fuse with the dermal armour so that the nature of the neural
spine of the last caudal vertebra cannot be determined.

TAXONOMY OF INDOSTOMUS PARADOXUS

185

A series of ossified tendons lies horizontally in the axial musculature. Some
originate on the ventral keels of the dorsal row of plates, others originate in the
musculature, but all run caudally to transmit the contractions of the anterior muscles
to the posterior part of the body. A cross section of one segment of the caudal

mm

FIG. 4. Lateral view of braincase and anterior vertebrae to show the occipital

articulation.

region reveals an external ring of armour divided vertically by the vertebra, with
four groups of tendons ; situated above and below the centrum on each side. Very
little musculature is present.

Caudal fin

The caudal fin is large and fan-shaped (fig. i). It is dorsoventrally symmetrical
and in each half there are two small procurrent rays, one principal unbranched ray
and four or five branched rays (figs, i and 5).

v2l

1 m m

hy

FIG. 5. Caudal fin skeleton.

The caudal skeleton consists of a preural centrum fused with the hypurals into a
single solid unit. It is extremely small and no evidence of any fusion of parts is
available.

Median fins and their skeletal supports

The isolated dorsal spines which may represent the first dorsal fin have been
mentioned above. The median keels of the dorsal plates (fig. 6) appear to fulfil the

i86

K. E. BANISTER

function of proximal radials for all the median fin rays. The dorsal and anal fins
both have six segmented rays, all but the last being branched. Differences in the
position of branched and unbranched rays relative to the scutes has been com-
mented on above (and see fig. i). A small nodule of cartilage (the distal radial?) is
found between the bases of the lateral halves of each fin ray.

FIG. 6.

Diagram of the last three dorsal spines and the first three dorsal fin rays to
show the fin supports. The fifth spine is out of alignment.

Pectoral fin skeleton

The pectoral fin has 23 rays, of which 15 are segmented. The supracleithrum
(fig. 7) is lodged ventrally in a niche on the first lateral scute which has been in-
corporated into the pectoral skeleton. Functionally, the cleithrum is represented by
this first scute and the median flange on its anterior edge, but it is probable that the
flange alone represents the cleithrum which has fused with the scute. The outer
part is ornamented in the same way as the other scutes and is clearly in series with

ptm

FIG. 7. Lateral view of the dermal elements of the pectoral fin skeleton.

them. Only the flange forming the hind wall of the gill chamber really distinguishes
it. I hesitate to refer to the scute-like part of this bone as an infracleithrum (Swin-
nerton 1905) which is a similar bone in Gasterosteus , because of the position of the
supracleithrum. Applying Occam's razor the most likely explanation is the fusion
of the cleithrum and the scute. The cleithra interdigitate in the midventral line.

TAXONOMY OF INDOSTOMUS PARADOXUS

187

The endochondral part of the pectoral skeleton consists of a coracoid, a scapula and
three pterygials (fig. 8). Baudelot's ligament is absent.

FIG. 8. Medial view of the pectoral fin skeleton. Only the upper and lower
pectoral fin rays have been drawn in.

The coracoid has a postero ventral process which rests on the cleithrum. The
scapular foramen is contained entirely within the scapula and only just fails to open
anteriorly. The dorsal margin of the scapula is expanded horizontally, and laterally
it fits closely on to the dorsal edge of the scute. The recess below the dorsal lip of
the scapula on its median face houses the origin of muscles which run to the pectoral
fin. The coracoid, scapula and pterygial ossifications are all contained within a
sheet of cartilage and are joined to each other more firmly than they are to the rest
of the pectoral skeleton. There is a possibility that the scapula has fused with the
uppermost pectoral pterygial. This has happened in other fishes eg. Scorpaenich-
thys (Starks 1930) and would account for the low number of pterygials.

Pelvic fin and skeleton

The pelvic fin consists of four segmented rays. The outermost ray is thickened,
and the middle two branched. The base of the fin articulates on a small boss in the
centre of a depression near the rear margin of the second ventrolateral scute (fig. i) .
Rays 2, 3 and 4 share either one elongated or two small rounded ossicles lying between
their divided bases. The first ray is not so markedly bifurcated at the base and
articulates directly with the boss.

188

K. E. BANISTER

The pelvic girdle is reduced to a small ridge of bone in the medial face of the
scute.

SYNCRANIUM

The skull is fairly flat on top and rather depressed. The dermal bones are pitted
and ridged, with most of their edges crenellated or spiny. A series of ridges on the
skull (figs. 9 andio) suggest the presence of an extensive cephalic lateral line system,
but close examination of sections has shown that in all probability the lateral line
system is confined to a pair of short tubes between and slightly in front of the orbits
(see below).

The eyes are large, the snout moderately elongated and the mouth is small and
terminal. The most remarkable feature of the mouth is the great discrepancy
between the lengths of the upper and lower jaws (fig. 9).

dsp dpt e R

ptm

pmx

FIG. 9. Head of an unprepared specimen in lateral view. The ridges are
illustrated diagramatically.

Olfactory region

The bones present are the ethmoid, lateral ethmoids, supraethmoid and vomer.
The nasal bone is absent.

The ethmoid and the supraethmoid are fused together to form one functional unit
with the supraethmoid as a dorsal shield over the wedge shaped and largely carti-
laginous ethmoid (figs. 10 and n).

On a few specimens up to four small pits are present near the anterolateral corners
of the dorsal surface of the supraethmoid.

The leading edge of the ethmoid block is formed entirely from the supraethmoid
and the maxillae and premaxillae are bound to it by fibrous tissue. The anterior
nostril (fig. 10 anf) has a flap of skin which directs the water over the olfactory organ
and into the canal between the lachrymal and the ethmoid. For most of its length
the canal is roofed by the dorsal edges of the supraethmoid and lachrymal which are
joined by collagen fibres. The anterior nostril is floored by a process from the
ethmoid block which has a ligament joining it to the lachrymal.

TAXONOMY OF INDOSTOMUS PARADOXUS

189

The lateral faces of the ethmoid are produced ventrally as sheets for the attach-
ment of the pterygoquadrate arch.

The vomer appears to have fused so firmly with the ventral face of the ethmoid
that prolonged maceration will not separate them. Serial transverse sections show
that there is no line of weakness nor any sign of a joint between the vomer and the
ethmoid. There are two possible explanations for this, but no conclusions can be

ptm

FIG. 10. Dorsal surface of the skull.

reached without an examination of much younger fishes than I have been able to
obtain. Either there is no vomer, and its place has been occupied by a posterior
projection from the ethmoid, or the ethmoid and vomer have co-ossified.

For convenience the process in question will be referred to as vomer. It is tooth-
less and has the usual relationships of a vomer bone, except at its anterior end (fig. n).

The lateral ethmoid (figs. 9, 10, n) forms the hind margin of the posterior nostril.

IQO

K. E. BANISTER

Little of this bone is visible in lateral view (fig. 9) . The ventral face is concave with
the median wall more or less contiguous with the expanded lateral edge of the eth-
moid bone. The parasphenoid separates the lateral ethmoids in the midline. The
portion of this bone which is visible externally is sculptured, so presumably a pre-
frontal bone is present as well.

O5mm

FIG. ii. Ventral surface of the snout. The left pterygoid bar and the right
maxilla and premaxilla have been removed.

Orbital Region

The small tubes containing the lateral line canals in the frontals vary in length in
different specimens. This is an indication that the cephalic lateral line system has
undergone reduction. Small descending frontal processes are present (fig. 12) which
form a groove with the lateral, post-orbital part of the frontal. This groove is
blocked posteriorly by the sphenotic, and houses what appears to be a part of the
muscle block associated with the operculum, although I hesitate to refer to them as
dilatator operculi muscles as they are highly modified and also seem to insert on to
the preoperculum. The frontals meet in the midline by a series of irregular overlaps.

The pterosphenoids are axehead-shaped bones which just meet each other in the

TAXONOMY OF INDOSTOMUS PARADOXUS

191

ventral midline above the parasphenoid (fig. 12). The posterior edge of each bone
contributes to the margin of the anterior foramen of the pars jugularis of the tri-
gemino-facialis chamber.

0-5 mm
FIG. 12. Ventral surface of the braincase and first vertebra.

The sphenotic is a bone of complex shape, the main part of which resembles a con-
vex cap forming the anteroventral corner of the neurocranium. It forms the
lateral margin of the anterior trigemino-facialis foramen (fig. 12). A thin, laterally
directed process is present which appears on the external surface of the skull, with the
frontal, just behind the orbit (figs. 9, 12). The extremity of this process is expanded
and sculptured and may well represent a dermosphenotic. It has been labelled as
such in fig. 9.

The parasphenoid penetrates deeply into the cartilage of the ethmoid, and is held
in place anteriorly by collagen fibres. Ascending processes are absent and the bone
is slightly flattened between the eyes. A thin sheet of muscles originates on the
orbital part of the parasphenoid and inserts on the " pterygoid ", this seems to be
the adductor arcus palatini.

Only one infraorbital bone is present and that is the lachrymal (figs. 9, 10, n).
The anterior extremity of the lachrymal is joined by a ligament to the ethmoid. The
lachrymal forms the outer wall of the olfactory cavity. An examination of the ol-
factory region reveals the presence of a few strands of tissue which Dr. L. Bannister
of Guys Hospital has suggested may well be the remains of a plunger mechanism.
The preservation of the fishes is unfortunately not good enough to comment further
on this. The plunger (if that is what is it) lies behind the nasal epithelium close to
the anterior nostril.

IQ2 K. E. BANISTER

Otic and Occipital Regions

Each prootic is separated from its partner in the midline by the parasphenoid
(fig. 12) . The distal wall of the trigemino-f acialis chamber is thickened and produced
anteriorly as a small spike below the anterior foramen for the chamber. From this
projection collagenous fibres insert on to the inner face of the hyomandibular. The
trigemino-facialis chamber is reduced, the medial wall is missing and only two
external foramina are present (fig. 12).

The bone labelled epiotic in fig. 10 appears to have a dermal component which
may represent a parietal ossification which has fused with the epiotic. The posterior
edge is smooth and is overlain by the first dorsal scute. Two pairs of ossified tendons
from the axial muscles insert on this region (see above). A shallow post-temporal
fossa is present, floored by the epiotic and pterotic.

Both the dermal and the endochondral parts of the pterotic are present. They
can be described as two flat plates meeting at an angle at the lateral edge of the skull
with the dermal plate overhanging the endochondral plate (fig. 12). At the widely
separated medial edges of the plates, two vertical struts of endochondral bone create
a chamber which has three openings. The posterior foramen opens into a cavity on
the inner surface of the prootic and the small anterior foramen opens close to the
hyomandibular socket. This cavity must largely house the inner ear but there is no
trace of otoliths in either whole specimens or in the sections.

The ventromedial corner of the exoccipital is produced backwards, slightly beyond
and level with the midpoint of the basioccipital condyle to articulate with the
modified transverse processes of the first vertebra (figs. 3, 4, 12). A small pit is
present on the posterior wall of the bone in which the descending process of the post-
temporal lodges. The foramen magnum is tube-like, and formed mostly from the
exoccipitals which just meet dorsally below the supraoccipital. The latter bone has
a sculptured surface ; there is no occiptal crest, just a flat plate narrowly separating the
epiotics (fig. 10).

The basioccipital has a convex condyle which does not protrude as far as the
exoccipital condyles.

The posttemporal overlies the rear margins of the epiotic and pterotic. The
serrated, sculptured dorsal plate is slightly convex (figs. 7 and 10) and from the under-
side of this there are two processes. The larger central process (fig. 7) abuts onto the
exoccipital and the small process has the supracleithrum adhering to its medial face.
The supracleithrum does not extend above the cleithrum.

Oromandibular region

The premaxillae are expanded at the symphysis and bear about four rows of fine
pointed teeth (fig. n). The symphysis is weak, the usual jaw ligaments are missing
and the upper jaw bones are contained within the fibrous tissue of the upper " lip ".
Ascending processes are absent (fig. 10).

The maxilla is a greatly reduced, weak bone (fig. n). Its posterior end is ex-
panded slightly and it is very firmly attached to the dorsal edge of the premaxilla.
The length of the maxilla varies from specimen to specimen, in some it almost

TAXONOMY OF INDOSTOMUS PARADOXUS

193

reaches to the midline. A discussion of the jaw mechanism will be found below,
(page 196).

The angular is a small bone, more easily seen in sections than in gross dissection,
which lies in a deep pit on the posterior margin of the articular (fig. 13). The articu-
lar is greatly elongated and occupies over three quarters of the length of the lower
jaw (fig. 13). The lateral face is sculptured and a ridge marking the position of
Meckel's cartilage is visible. The tendon from the adductor mandibulae muscles
inserts onto the dorsal margin in a gutter formed by the development of a medial
shelf below the eye. The anterior end of Meckel's cartilage is medially grooved and
receives the posterior projection of the dentary.

A small coronoid process is present near the posterior end of the dentary (fig. 13).
The symphysial region is expanded as a dentigerous plate which extends in front of
the premaxillary tooth plate so that occlusion only occurs on the posterior half of
the former plate.

The seamoid articular is a small, roughly cylindrical bone (fig. 13) in close proximity
to the ectosteal part of the articular. It lies in the tendon at the ventral end of the
adductor mandibulae muscles which are in an unusual position as a result of the
elongation of the mandible (see below).

mpt

pty

FIG. 13. Medial aspect of the suspensorium, opercular bones and mandible. Part of the
medial adductor muscle has been removed to show the sesamoid articular.

Palatoquadrate arch

The palatoquadrate arch lacks the usual complement of bones (fig. 13). One
long bone, the " pterygoid ", fills the places usually occupied by the ectopterygoid,
endopterygoid and palatine. There is no indication of separate centres of ossification
in the adult fish so the real identity of the bone must remain hidden until young stages
can be found. In Gasterosteus there is a large tri-radiate " pterygoid " bone formed
by the fusion of the ectopterygoid and endopterygoid. In the syngnathids the endo-

194 K. E. BANISTER

pterygoid is the largest element whilst in the triacanthid fishes neither the ecto-
pterygoid nor endopterygoid is greatly enlarged. Therefore, just because jaw or
snout elongation has occurred, it cannot be stated that certain pterygoid elements
necessarily enlarge, fuse or disappear.

If the vomer has co-ossified with the ethmoid block (see page 189) then there are
slight grounds for offering the suggestion that the descending sheets on the ventral
face of the ethmoid represent anterior suspensorial bones (probably the autopala-
tines) fused with the ethmoid as an added strengthening member. This must remain
conjectural however until the ontogeny is known.

The " pterygoid " is largely a thin lath of bone with the dorsal margin below the
eye bent over laterally to conform to the curvature of, and to support, the eye. In
front of this, a dorsal lamella develops at an angle to the rest of the bone to form the
groove which will anteriorly hold the descending sheets from the ethmoid (fig. n).
At the widest part of the pterygoid, just at the front of the orbit, a ligament joins
it to the lachrymal.

The quadrate has an elongated lower limb (fig. 13), with the articular condyle as
a transverse cylinder placed ventromedially dorsolaterally at its heel. The dorsal
limb is shorter than the ventral limb but extends further on the medial side where it
overlaps the metapterygoid. The posterior edge of the quadrate is channelled to
receive the symplectic.

The metapterygoid dorsally articulates with the hyomandibular, and the ventral
part of its posterior margin contacts the symplectic. Both the metapterygoid and
the dorsal blade of the quadrate are angled more towards the midline than is the
rest of the suspensorium. The symplectic is a thin conical bone which is housed in a
groove on the rear edge of the quadrate (fig. 13). It is poorly ossified and has a
synchondrosis with the hyomandibular.

Opercular series

The operculum is armed with six conspicuous spines (fig. 13). The hyomandibular
socket faces forwards at the anterodorsal edge of the bone. Medially, the posterior
margin bears a shelf underneath which the levator operculi muscles from the hyo-
mandibular insert. The dilatator operculi muscles insert on the anterior projection
of the operculum in front of the hyomandibular socket.

The suboperculum is greatly reduced to a thin rod inside the posterior margin of the
operculum (fig. 13).

The interoperculum is a flat sculptured bone connected to the mandible by a short
but ill-defined interoperculo-mandibulare ligament. A loose interdigitation exists
between the interoperculum and the operculum.

The preoperculum has a firm ligamentous connection to the quadrate along its
ventral edge. The greatest expansion of the adductor mandibulae muscles occurs
in the space between the preoperculum and the suspensorium.

Hyoid arch

The interhyal is a flat quadrangular bone with its dorsal margin fitting into a trans-
verse groove in the posterior face of the hyomandibular (fig. 13).

TAXONOMY OF INDOSTOMUS PARADOXUS

195

The epihyal and ceratohyal are ankylosed dorsally and both have a depressed
section (fig. 14). There is no foramen in the ceratohyal. Two slender branchio-
stegal rays articulate on the lateral face of the slender part of the ceratohyal and two

eh ch uh

1 mm

FIG. 14. Lateral aspect of the hyoid bar, branchiostegal rays
and urohyal.

bbl

FIG. 15. Dorsal view of the pharyngeal skeleton. Basibranchial 4 is cartilaginous.

196 K. E. BANISTER

thicker rays articulate on the same face of the expanded part. One thick and one
thin branchiostegal ray articulate on the epihyal (fig. 14).

The upper hypohyal is a splint-like bone lying along the dorsal face of the cerato-
hyal to meet the lower hypohyal (fig. 14) .

The basihyal is a flat bone about four times as long as it is wide. The urohyal (fig.
14) is keeled and the posterior half of the bone is protected ventrally by a triangular
median dermal plate which lies in front of the first ventrolateral body plates.

The hyomandibular (fig. 13) is a roughly cuboidal bone with the foramen for the
hyomandibular nerve below the anterior articular head. The sockets on the neuro-
cranium for the hyomandibular are shown in fig. 12.

Branchial skeleton

Ceratobranchials 4 and 5 have fused with tooth plates. Those on the fifth cera-
branchials are the larger but they do not meet in the mid-line (fig. 15).

The fourth epibranchial is expanded and lies above the fourth upper pharygeal
tooth plate. The larger anterior pharyngeal tooth plate is probably the result of the
fusion of the second and third toothplates with the second and third infrapharyngo-
branchials, the latter being represented by a thickening of poorly ossified endochon-
dral bone along the lateral edge of the plate. There is no trace of the fourth infra-
phary ngobranchial .

There are three irregularly cuboidal hypobranchials. The fourth basibranchial is
distinct but does not ossify.

NOTES ON THE SOFT ANATOMY

i. The jaw mechanism of Indostomus

The upper jaw cannot be protruded, the maxillae and premaxillae are bound to
each other and to the front of the ethmoid by fibrous tissue. Discrete jaw ligaments
are lacking. When the lower jaw is depressed the premaxillae and maxillae rotate
about the front of the ethmoid until the distal arm of the premaxilla is vertical. This
only allows the jaws to open to a little over 20. The feeding habits have become
modified and are described in the appendix.

The adductor muscles to the lower jaw are small, and are shown diagramatically
in fig. 13. They comprise a lateral and a medial muscle. At their insertion they are
approximately equal in cross sectional area but before the posterior end of the sesa-
moid articular the medial muscle has greatly increased in thickness and effectively
surrounds that bone. Still further back a band of muscles from the quadrate to
apparently the anterior of the operculum separates the medial and lateral adductor
muscles. The lateral adductor muscle originates on the hyomandibular. Just
anterior to the point of separation of the medial and lateral components the thick
medial muscles split into a dorsal and a ventral portion. The dorsal portion origin-
ates on the braincase and the ventral portion on the hyomandibular and possibly
on the metapterygoid.

I am not certain of the identity of these muscles, and the terms applied to them
are used purely descriptively.

TAXONOMY OF INDOSTOMUS PARADOXUS 197

2. Swimbladder

The swimbladder is simple and physoclistous. A gas gland and an oval are
present.

DISCUSSION

The systematic position of Indostomus paradoxus

Indostomus is a bizarre, highly modified fish. Two major factors have combined
to obscure its relationships; firstly the great elongation of the lower jaw and secondly
the presence of dermal body armour.

The fish fauna of Lake Indawgyi has been listed by Prashad and Mukerji (1929)
and does not contain any obvious relatives of this endemic species, so its peculiarities
may be due to long isolation in the one lake basin. Prashad and Mukerji put for-
ward the suggestion that the ancestors of Indostomus were not fresh water fishes, for
in the early Tertiary the Bay of Bengal extended as far as Lake Indawgyi, and some
marine fish could have been isolated and confined to the lake when the sea receded,
this seems to be the most likely suggestion.

Prashad and Mukerji thought that Indostomus was " closely allied to the family
Solenostomidae and to a certain extent the Syngnathidae of the order Solenichthyes
Regan ". However they pointed out that there were certain differences between the
three families. The structures described in this paper show that Indostomus is
distant from the Syngnathidae and the Solenostomidae. The anatomy of the
families comprising the order Gasterosteiformes (senu. Greenwood et alii 1966) will
be described in subsequent papers, but for the moment Jungersen (1910) gives a
good account.

Many of the similarities between Indostomus and the syngnathids are superficial
and due to the presence of the annular armour and an elongated snout. The Sy-
gnathidae and the Solenostomidae both show the following features of taxonomic
significance which are not possessed by Indostomus.

1 . Enlargement and/or complex articulation of the anterior 3 or 4 vertebrae with
each other.

2. Sutural contact between the basioccipital and the autopterotic.

3. Separate ecto- and endopterygoids.

4. A metapterygoid which does not ossify as a separate bone but forms the
metapterygoid process at the anterodorsal corner of the quadrate.

5. A dorsal lamina on the symplectic.

6. 4 pectoral pterygials.

7. No more than 3 branchiostegal rays.

These characters are enough to suggest that Indostomus and the syngnathids are
not closely related, or at least should not be placed in the same order.

Bolin (1936) thought that Indostomus was closer to the Aulorhynchidae and the
Aulostomidae than to the Syngnathidae and Solenostomidae. I hope to show in
subsequent papers that the Aulostomidae are closely related to the Syngnathidae and
Solenostomidae whilst the Aulorhynchidae is a family of moderately generalized
fishes from which the Gasterosteidae are specialized descendants.

igS K. E. BANISTER

The Aulorhynchidae is the most primitive family of the suborder gasterosteoidei
in which Greenwood et alii (1966) have placed Indostomus. Although Aulorhynchns
and Gasterosteus have several features in common and are clearly related there are
several reasons why Aulorhynchm and Indostomus are most unlikely to be related.
The trends in the aulorhynchid gasterosteid lineage have not developed in a direc-
tion likely to produce indostomid facies.

Bolin (1936) admitted that the similarities he saw between Indostomus and Aulo-
rhynchus were superficial, and this is so, but there is one surprising feature that they
both possess and that is an elongated mandible. It is relatively much shorter in
Aulorhynchus than in Indostomus but even so it is produced backwards relative to
the upper jaw further than in most fishes.

Certain other fishes eg. Exocoetus volitans (Halocypselus of Gregory fig. ioi),Para-
chaenichthys (Chaenichthyidae), Gregory 1933) show the same phenomenon so
perhaps one should not place too much weight on this character.

Indostomus and Aulorhynchus have few features in common, but because the
former is more specialized than the latter many characters of the more primitive fish
could have given rise to characters of the more specialized fish. But none of the
important features peculiar to Aulorhynchus and the gasterosteids are found in
Indostomus.

Aulorhynchus in particular and the gasterosteids generally have the following
characters not possessed by Indostomus.

1. Ascending processes on the premaxillae

2. Nasal bones

3. A well formed vomer bone

4. A triradiate " pterygoid " bone

5. Autopalatines

6. An infraorbital series of 5 bones in Aulorhynchus, 3 in the gasterosteids

7. Ascending processes on the parasphenoid which contact the f rentals

8. An intercalar bone

9. Parietal bones

10. Upper pharyngeal teeth plates consisting of a UP2 and a fused UP3 & 4. In
Indostomus the anterior element is the larger and may be a fused UP2 & 3, with a
separate UP4

11. 4 pectoral pterygials

12. Ribs

13. Pelvic fin skeleton

14. Dermal scutes along the lateral line, not full armour

15. Over thirty vertebrae.

Although many of these differences are only to be expected when comparing a
generalized fish with a specialized fish some have been retained by the specialized
descendants of aulorhynchids and are not present in Indostomus.

Indostomus must therefore be removed from the order Gasterosteiformes as it now
stands (Greenwood et alii 1966, p. 398). The status of the remaining families in that
order will be discussed in subsequent papers.

The next problem is to try and decide which of the higher taxa of Rosen and

TAXONOMY OF INDOSTOMUS PARADOXUS 199

Patterson's (1969) modification of Greenwood et alii (1966) can best accommodate
Indostomus. Only the superorders Acanthopterygii and Paracanthopterygii need be
considered.

Rosen and Patterson (1969) have included the Atherinomorpha of Greenwood et
alii (1966) as a series within the Acanthopterygii parallel to the berycoid percoid
lineage. There are certain features of the Atheriniformes which are shared to various
extents with Indostomus. This does not necessarily imply that Indostomus ought
to be placed in this order, but rather than it is of a similar level of complexity. This
is almost stating the obvious for Rosen (1964) concluded that the Atherinomorpha
arose from a group that stood somewhere in the ancestry of the Perciformes, and it is
in this zone above the Protacanthopterygii and below the Perciformes that Indosto-
mus must lie. The upper limit is set by Gosline (1966) who states that no percoid
fish nor percoid derivative has nodules between the pelvic fin rays and the pelvic
skeleton, which must therefore place Indostomus below the percoids.

Fishes in the series Atherinomorpha typically have a caudal fin skeleton which
originates on one supporting centrum (as is the case in the majority of the Acantho-
pterygii). The caudal skeleton of Indostomus conforms to this pattern although the
separate hypurals have fused up into one plate. This is possibly only to be expected
because of the minuteness of its caudal skeleton.

There are some interesting trends within the suspensoria of some of the Atherino-
morph families which may throw some light on the origin of the peculiar suspensorium
of Indostomus. In the family Isonidae (Rosen, 1964) the autopalatine is greatly
reduced and the endopterygoid is absent. This reduction in size and number of
suspensorial bones has been continued further in Indostomus. The suspensorium and
the jaws of the family Adrianichthyidae are of some interest. In many species of
atheriniform fishes the upper jaw is protractile, but not in the Adrianichthyidae.
There the maxilla is firmly joined to the upper edge of the premaxilla. The ethmoid
is very large and has lost the disc-like appearance found in most atheriniforms. In
these two respects it offers a parallel with Indostomus. The resemblance could go
even further. In the adrianichthyids the lower jaw is of normal length and only the
autopalatine links the upper jaw with the quadrate. If the proportions of the jaws
and snout were altered to the proportions of those of Indostomus, a suspensorium
very like that of Indostomus would result. This is pure hypothesis, but the adrianich-
thyids are a group of fish which satisfy the likely conditions for the pre-indostomid
suspensorium.

If the lower jaw were elongated to articulate behind the eye, the autopalatine is
already in position to elongate whilst keeping the upper jaw and quadrate tied
together. As the quadrate is now behind the orbit and the lower jaw articulation
almost vertical, there is no room for a large endopterygoid which could well dis-
appear leaving just the hyomandibular, quadrate and symplectic to form the vertical
column between the lower jaw and the braincase. The suspensorium of Xenopoecilus
does not fit exactly into the appropriate pattern because the metapterygoid is
now absent, yet is present in Indostomus, and although I have referred to the large
suspensorial bone in Indostomus as the " pterygoid ", it could possibly be the
autopalatine.

200 K. E. BANISTER

So, within the Atheriniformes there exist possible plans from which some of the
characters of Indostomus could have evolved. It must, though be borne in mind
that it is highly unlikely that the Adrianichthyidae have the basal suspensorial
pattern of the order, and also that convergence could play a large part in the produc-
tion of peculiarly modified suspensoria for somewhat similar jaws.

Parenthetically, it must be mentioned that the suspensorium of the ammodytoid
Hypoptychus dybowskii (Gosline, 1963) is not unlike that of Indostomus. It differs
mainly in having both the ectopterygoid and the palatine forming the lath in front
of the quadrate and in the presence of a normal palato-maxillary connection. The
jaws of Hypoptychus are approximately equal in length which suggests that this
unusual pattern of suspensorium is not obviously related to the disposition of the
snout and jaws and may well have occurred independently within the neoteleostei.

The Atheriniformes have the following features which differ from those in Indosto-
mus.

1. A much higher number of vertebrae

2. No dorsal spines, and rarely a spiny first dorsal fin

3. Pelvic girdles with a long lateral spur

4. No trace of bony armour

5. Four pectoral pterygials, recessed within the scapulo-coracoid margin

6. Nasal bones usually present

7. A different arrangement of the upper pharyngeal tooth plates

8. Unarmed opercular bones

This makes the inclusion of Indostomus within the atherinomorph series unlikely.
Little is known about atherinomorph ancestors, but the group evolved in the fresh-
waters of Asia, the area in which Indostomus now lives.

It is now necessary to try and fit Indostomus into the lower reaches of the Per-
comorpha series of the Acanthopterygii.

The order Lampridiformes contains the remarkable fish Stylophorous chordatus
whose anatomy was described by Regan (1924). The most significant feature is the
long mandible which in elongating backwards has carried the quadrate to a position
behind the hyomandibular so that the suspensorium is angled forwards. The
hyomandibular, quadrate and symplectic form a narrow column which is angled the
opposite way from that in Indostomus. In Stylophorous the ecto-, endo- and meta-
pterygoid bones are reduced to splinters between the quadrate and the hyomandibu-
lar. The autopalatine is a rod-like bone in a normal position but distant from the
pterygoids. The premaxillae have large ascending processes and the mouth is
highly protrusile. This elongated mandible is not universal amongst the Lampridi-
formes, but the modifications of the suspensorium form an interesting comparison
with the situation in Indostomus. The presence of a basisphenoid, an orbitosphenoid,
parietals, the dermopterotic-frontal contact which excludes the dermosphenotic from
the skull roof, the high vertebral number long based dorsal fins without true spines
and the absence of dermal armour are some of the many features which do not suggest
any close relationships between the Lampridiformes and Indostomus. It is however,
worth noting that three pectoral pterygials are common in that order.

Indostomus is far too specialized to be associated with the Beryciformes and their

TAXONOMY OF INDOSTOMUS PARADOXUS 201

early offshoot the Zeiformes (Patterson 1964). Any relationship with the Syn-
branchiformes and Channiformes can be dismissed.

The order Scorpaeniformes is suspected by Greenwood et alii (1966) of being poly-
phyletic, but typically these fishes have the second suborbital bone contacting the
preoperculum and a caudal skeleton consisting of two plate-like hypurals sutured to
the terminal half centrum. The family Agonidae contains armoured forms which
have retained the ribs and an unmodified caudal skeleton. Despite the precedent
in the dermal armour of the Agonidae, the Scorpaeniformes are a fairly well defined
order which, if any stock underwent reduction would be unlikely to produce the
indostomid facies.

This leaves two small orders to consider, the Dactylopteriformes and the Pegasi-
formes.

The only point of similarity between the Dactylopteriformes and Indostomus is
that both possess a low number of vertebrae, 22 in the former and usually 21 in the
latter. The osteology is otherwise substantially dissimilar (Allis 1909).

The Pegasiformes are as much a problem as is Indostomus. Pegasus has never been
described in detail, Jungersen (1914) has only given a brief account of the fish and
from this one may gather that Pegasus is, like Indostomus, in an apparently isolated
position.

The Paracanthopterygii is a superorder of fishes which, in the words of its definers,
Greenwood et alii (1966) "represents a spiny finned radiation more or less compar-
able morphologically with that of the superorder Acanthopterygii ". Rosen and
Patterson (1969) have shown that many of the acanthopterygian features of the
Paracanthopterygii are lost during the evolution of the superorder.

The majority of the species included in the five component orders are marine, but
the most primitive order the Percopsiformes is now confined to the fresh waters of the
new world. Rosen and Patterson (1969) have shown that the superorder is mono-
phyletic but contains two lineages : the percopsid-gadid lineage and the batrachoid-
lophiid-gobiesocid lineage. The most trenchant characters separating the Para-
canthopterygii from the Acanthopterygii are the presence of a levator maxillae
superioris muscle and a second free ural centrum in the caudal fin skeleton. During
the evolution of the batrachoid-gobiesocid lineage these two characters were modified
and are absent in the gobiesocids, nevertheless a multitude of less trenchant charac-
ters links the batrachoids to the gobiesocids.

The Gobiesociformes are highly specialized fishes, as is Indostomus, and comparison
of the characters of Indostomus with the trends within the batrachoid line that pro-
duced the gobiesocid facies reveals a number of similarities. Many of these can be
explained by convergence and parallelism, some characters are of much less sig-
nificance than others but whether convergence and/or parallelism can explain the
particular combinations of characters present in both is something that will have to be
argued when our knowledge increases.

Characters and trends present in the batrachoid lineage and in Indostomus are

1. Flattened skull roof

2. Fusion of parietals with epiotics

3. Autosphenotics flare forward

202 K. E. BANISTER

4. Absence of an intercalar

5. Horizontal alignment of the occipital condyles

6. Infraorbital series reduced to the lachrymal which lacks a suborbital shelf

7. Reduction of the pterygoid series

8. Hyomandibular and preopercular in intimate contact

9. Upper hypohyal elongated and lying over the cerataohyal

10. Six branchiostegal rays

11. Upper end of preopercular well below hyomandibular head

12. Posterior margin of opercular excavated

13. Caudal skeleton (in gobiesocids) fused up without a free parhypural

14. Pectoral radials enlarged

15. Coracoid with a postero ventral process

16. Spinous anterior dorsal fin rays

17. Rounded caudal fin

18. Scales absent

19. No orbito- or basisphenoids

20. Subopercular slender and directed towards the upper corner of the opercular

21. Adductor arcus palatini across the floor of the orbit

22. Pelvic fins subthoracic and with reduced number of rays

23. Upper and lower jaw teeth fail to occlude completely.

Some of these characters need further qualification, i, 5, 12 and probably 8, all
seem to be related to the flattening of the skull. Rosen and Patterson (1969) pointed
out that in paracanthopterygians the exoccipital condyles tend to lie alongside the
basioccipital condyle, whereas in the acanthopterygians they tend to lie above the
basioccipital condyle. This is mechanically necessary in fishes which raise their
heads, a practice found in both Indostomus and the paracanthopterygians. In the
Scopraenif ormes Hoplichthys langsdorfii the exocciptal flanges are level with the basi-
occipital condyle ; in Platycephalus insidiator and P. japonicus they are only a little
higher, and fail to meet medially above the basioccipital condyle. A similar situation
is found in the perciformes Gymnodraco acuticeps and Parachaenichthys georginaus.

Number 13, the fused caudal skeleton, is another character which needs some
comment. A caudal skeleton which has had all its components parts fused into a
single hypural plate supported on one centrum is found in advanced perciforms as
well as in the Paracanthopterygii and the Gasterosteif ormes. In the case of Indostomus
the evidence afforded by the second pre-ural neural spine (Patterson, 1968, Rosen and
Patterson, 1969) cannot be called upon, because the neural arches of all the caudal
vertebrae of Indostomus have fused with the dermal scutes. Most perciforms in
which the caudal skeleton is fused typically have an autogenous parhypural whilst
in the Gobiesocif ormes the parhypural is fused with the hypural plate. In the
scombrids, where the parhypural is in some species fused with the hypural plate, a
well developed parhypural process is evident. In the Gasterosteiformes (with the
exception of the highly reduced caudal skeletons of the Syngnathidae) the parhy-
pural, when fused, is usually obvious in alizarin preparations. The absence of an
autogenous parhypural in Indostomus suggests a real affinity with the Gobiesocids,
because although the caudal skeleton is as fused as it is in the syngnathids, the

TAXONOMY OF INDOSTOMUS PARADOXUS 203

caudal fin plays a much greater role in the locomotion of the fish. In other words the
consolidation of the urophore elements is more likely to be the result of a phyletic
trend (as in the batrachoid-gobesocid line) than of reduction resulting from a trend
to develop a prehensile tail and lose the caudal fin as in the Syngnathidae.

The excavated hind margin of the opercular (12), is a feature that is also present
in the Scorpaeniformes (a group that seems to have several paracanthopterygian
trends) but the very spiny type of opercular present in Indostomus and batrachoids
like Opsanus tau (with the consequent excavations of the rear margin) does not seem
to be repeated anywhere else in quite this form. None of the batrachoids have as
many pungent spines as does Indostomus.

Characters 2, 4, 6, 7, 10, 13, 14, 16, 17, 18, 19, 20 and 23 are far from rare in the
Acanthopterygians and their significance would seem to lie in their particular
combination in one group of fishes.

Characters 3,9, u, 15, 21 and 22 must indicate either that Indostomus has some
affinity with the batrachoid lineage of the Paracanthoptergyii or that it exhibits a
remarkable degree of convergence in these rather specialized features.

The jaws of Indostomus are highly aberrant but would seem to conform to the
" maxillary rotation " style present in the higher Paracanthopterygii. Within the
batrachoid lineage when the levator maxillae superioris muscles are absent, no
muscles insert on the maxillae and the rotation of the upper jaw in under the sole
control of the depression of the lower jaw.

One would expect that the peculiar ethmoideum of Indostomus would be too special-
ized to reveal any affinities with more orthodox fishes. The only points worth men-
tioning in this context are that in Opsanus beta the vomer is fused with the ethmoi-
deum (Rosen and Patterson, 1969), and that in Opsanus tau at least (see fig. 263 in
Gregory, 1933) the physical disposition of the ethmovomer block and the pterygo-
quadrate bars are much the same as they are in Indostomus.

Primitive Paracanthopterygian features present in Indostomus include the
following :

1. Absence of a supra-temporal fossa

2. Two openings to the pars jugularis

3. Parasphenoid-pterosphenoid contact

4. No pelvic spine.

Apart from the specializations of the snout and jaws, there are several important
differences between Indostomus and the paracanthopterygians. The following are
characters of Indostomus which are not found in the more highly evolved paracantho-
pterygians (although some are present in the Percopsidae) .

1. A strong ventral process of the posttemporal

2. Short hypobranchials

3. A distinct gap between the epibranchials 3 & 4

4. No reduction of the basibranchials

5. No ascending processes on the premaxillae

6. Considerable reduction of the lateral line system

7. Absence of ribs

8. Dermal armour.

204 K. E. BANISTER

It seems then that Indostomus contains a mosaic of a few features that are primi-
tive to fishes of the level of basal paracanthopterygians, several features also found in
members of the batrachoid-lophiid-gobiesocid lineage and a few characters sui
generis. The presence of the characters peculiar, in this combination, to the batra-
choid-gobiesocid lineage make it less likely (but not impossible) that a similar com-
bination could be found in acanthopterygian fishes.

The balance of evidence suggests that Indostomus could have evolved from one of
two species. It could have evolved from a species just on the paracanthopterygian
side of the paracanthopterygian-acanthopterygian dichotomy and then have evolved
along a line somewhat parallel to the batrachoid-gobiesocid line to produce a fish
with the short dorsal and anal fins and the few rayed caudal fins as in the Gobiesoci-
formes. The other possibility is that Indostomus evolved from a species somewhere
on the line leading to the gobiesocids. The fact that the most precise characters are
those shared with the more highly evolved paracanthopterygians is some evidence
for this. 1 The fact that the gobiesocids are marine littoral forms found in the Indian
Ocean does not contradict this when Prashad's and Mukerji's views on the origin of
Lake Indawgyi are taken into account. The presence of the primitive features
could then, in some cases be explained by secondary reduction and simplification.

Much of this discussion is, of necessity, speculative, but the balance of evidence
leads me to consider Indostomus paradoxus to be the only species of a new Paracan-
thopterygian order ; order Indostomiformes, which shows some affinities with the
Batrachoidiform-Gobiesociform lineage of that superorder.

The diagnosis of the order Indostomiformes is essentially that of the species viz.
small freshwater fishes from Upper Burma, the body has a complete dermal armour
covering ; the head is depressed and is mobile ; the pterygoid series of bones is
reduced in number ; the upper jaw is not protrusile ; the lower jaw is long ; a series
of isolated spines is present in front of the dorsal fin; the dorsal and anal fins are
short based and opposite each other ; the pelvic fin contains fin rays with cartilaginous
nodules at their bases ; there is no pelvic fin spine.

SUMMARY

1. The anatomy of Indostomus paradoxus is described.

2. A new order-INDOSTOMIFORMES is proposed to contain this single species.

3. It is suggested that the order Indostomiformes be placed within the superorder
Paracanthopterygii and probably near the Gobiesociformes.

ACKNOWLEDGEMENTS

I am very pleased to be able to thank Dr. Alan Tubb of the F.A.O. in Thailand for
providing the specimens of Indostomus and so generously letting me use his unique
observations on the living fish.

Drs. P. H. Greenwood, C. Patterson and E. Trewavas have helped immeasurably

1 Whilst in the realms of speculation I am tempted to compare the behaviour of captive Indostomus in
leaping out of the water and sticking onto the glass (presumably by a combination of surface tension and a
flattened belly see Appendix) with the trend to develop a sucking disc in the Gobiesociformes.

TAXONOMY OF INDOSTOMUS PARADOXUS 205

in putting this work on paper in an orderly fashion and eliminating many of my
mistakes. Needless to say any remaining errors are all my own work.
To all these people I gratefully offer my thanks.

APPENDIX i
Based on notes made on the behaviour of the living Indostomus by Dr. Alan Tubb

Notes on Indostomus

About 140 specimens collected from a small creek entering the west side of Lake
Indawgyi were brought alive to Rangoon and installed in two small (10 gallon)
aquaria, on 3 April 1956.

One aquarium tank contained a pure stock of Indostomus, about no individuals,
the other contained a variety of small Barbus spp., Platypoecilus and Xiphophorus
hybrids.

Both tanks had sand on the bottom and were planted with clumps of Cabomba
and dwarf Sagittaria.

In the community tank, the older inhabitants spent some time chasing and nipping
at the Indostomus, without however doing any obvious damage, and after two days
the enlarged community settled down and no further attacks occurred.

The Indostomus is quiet in behaviour, and generally slow-moving. Normal
movement is effected almost wholly by means of the pectoral fins which maintain
a constant rapid fluttering movement, even when the fin is apparently completely
at rest. When alarmed, these fish move with remarkable speed, apparently using
the relatively large fanshaped caudal fin. The dorsal and anal fins are usually
folded along the body, but may be fully spread when the fish is moving slowly, with
no apparent purpose, or when resting in an inclined position close to an ascending
strand of vegetation. The dorsal and anal fins, together with the tiny pelvics are
used as " brakes " and are widely spread to retard forward motion. They and the
caudal fin are often spread when the fish is "hanging " in the water, presumably to
provide an increased resisting surface.

The Indostomus appears to feed entirely on animal food, but will only take live
organisms. In the aquarium, they fed readily on small tubificid worms which were
scattered over the surface of the water and allowed to sink slowly to the bottom.
Only rarely were worms actually sinking taken by the fish. In almost all cases, the
worms fell to the bottom and were then taken.

When approaching a food item, the Indostomus moves very slowly, all fins depressed
except the pectorals, which maintain their constant fluttering motion. When close
to the food organism, the fish rolls gently over to one side, either right or left. In
the case of a worm protruding from the sand, the roll may be sufficient to place the
transverse axis of the body at 90 to normal. The final millimetre of movement
is accomplished by a sudden dart (the jaws are not protrusible) and the food item
snapped up. If the worm is deeply imbedded in the sand, the fish, having gripped
the exposed portion, wriggles violently backwards, and usually succeeds in with-
drawing the worm from its tube.

206 K. E. BANISTER

Colonial protozoans growing on the aquarium walls or on the vegetation are taken
in a similar manner, the degree of rolling depending on the situation of the colony.
In this latter case, the withdrawal by the fish is steady, and usually only a few
millimetres.

Normally there appears to be little belligerence between individuals, although
actual shoaling does not occur, each fish moving independently of other members of
the community. Occasionally, however, a fish seems to establish a kind of temporary
" territory ". These individuals, by their more slender form, and the absence of the
whitish or greenish mass visible through the translucent walls of the abdomen of
other less slender specimens, are probably males.

In the cases observed, the " territory ", which does not appear to exceed in
diameter twice the total length of the fish, is maintained only for brief periods,
rarely more than 15 minutes, but during this time, all other members of the species
which infringe the boundary are driven off, although no damage appears to happen
to the invader.

Despite its dermal armature, the fish is surprisingly flexible in life, and can bend
in an almost complete circle, the snout approaching within a few millimetres of the
caudal fin. The most remarkable feature, however, is the flexibility of the " neck ".
It is one of very few fishes known to the writer, which can actually " raise its head ".
The free anterior vertebrae, permit a vertical flexure of the head to an angle of about
10 from the main axis of the body. Lateral and ventral flexures of about the same
order of magnitude also occur.

In the tank containing only Indostomus the fish showed a strong tendency to
assemble in the darkened areas, particularly, a number would cluster, heads upward,
on the inner surface of the angles of the tank, where they were sheltered from the
direct front lighting. Not all followed this habit, many spending the great part of
the day lying flat on the bottom, or resting head upwards at an angle of 30-75
among the vegetation. These resting periods appeared to be prolonged, although it
was not possible to be sure that any one individual remained in the same place for an
extended period.

A peculiar reaction was noted when the fish were unusually disturbed, either when
the glass of the tank was being cleaned, or occasionally when the " mulm " was
being siphoned from the bottom. Most of the fish would dart wildly about, pro-
ducing a noticeable clicking sound as their bony snouts bumped against the glass
walls of the tank. A few, however, would leap clear of the surface and adhere to the
upper part of the glass, i or 2 centimetres above the water. They failed to respond
to a gentle touch, remaining quite immobile, and either had to be gently pushed
back into the water, or allowed to remain until they flipped back again of their own
accord. One specimen was observed to remain sticking to the glass for 4^ minutes
before wriggling into the water again. They apparently suffered no damage from
this exposure.

In the community tank, the Indostomus paid little attention to the other species,
simply moving slowly aside as another fish approached. The current induced by the
sudden sweep of the caudal fin of a larger fish often tumbled these extremely light
creatures over and over, but did not appear to cause them a great deal of alarm.

TAXONOMY OF INDOSTOMUS PARADOXUS 207

The respiratory movements are worthy of note. The mouth remains open con-
stantly, except when food is actually being ingested, and the gill covers are spread
outwards, forming a narrow frill one each side of the head. The gills are thus partly
exposed and are clearly visible from any position behind to the posterior edge of the
gill covers, or from below, the branchiostegals and their connecting membranes also
being widely spread. There is a steady rapid pulsation of the isthmus and connected
membranes, but as there does not appear to be any inner postlabial skin fold to serve
as an oral valve, such as that found in certain of the siluroids and other fishes, it is
difficult to avoid the impression that the respiratory current is not, in part at least,
induced by the continuous fluttering of the pectoral fins. When " stalking " its
food, the Indostomus depresses the otherwise extended opercles and branchiostegals
and the pulsation of the isthmus appears to cease entirely.

The breeding of this fish has not yet been observed, but several specimens, ap-
parently females, showed evidence of gonadal development during late June 1956.
These fish became markedly more robust in appearance and, through the translucent
body walls, the development of the ovaries was readily observable. The first obvious
stage was the appearance of a pale creamy mass occupying about half the length
and about one third of the volume of the abdominal cavity, extending from opposite
the first dorsal spine to the vent. In some of the fish under observation, this mass
later assumed a delicate jade green color, and although the length did not appear to
increase, the distension of the abdomen became more marked. In the noticeably
more slender specimens, and particularly in those showing the temporary " terri-
tory " habit mentioned above, no such evidence or gonadal development was
observed.

When captured, all specimens were generally dark to dusky brown the fins showing
the transverse banding mentioned by Prashad and Mukerji (loc. cit. p. 221). In the
aquarium tanks however, in the clear water and under the relatively bright light
conditions, most specimens rapidly became a very light, honeybrown color and the
fin patterns were reduced to extinction. Dorsally, on each side of and close to the
mid line were four bright golden squares, situated at the interspaces between the
dorsal spines, i.e. on the 4-7 rings. Some specimens occasionally become somewhat
darker in color, when indistinct transverse bands of cloudy black appear, apparently
coinciding with the rings and extending from the anterior (scapular of Prashad and
Mukerji) to the first caudal ring. On such individuals, the transverse banding of the
fins, particularly the caudal fin, becomes evident.

KEY TO FIGURES

an angular nipt metapterygoid

anf anterior nostril mx maxilla

ar articular na neural arch

bb basibranchial op operculum

bo basioccipital pmx premaxilla

bp body plate pnf posterior nostril

br branchiostegal ray pop preopercular

c centrum pro prootic

cb ceratobranchial ps parasphenoid

208

K. E. BANISTER

ch ceratohyal pt

cl cleithrum ptm

co coracoid pts

d dentary pty

dpt dermal part of pteroticum q

dsp dermal part of sphenoticum r

Dsp dorsal spine sa

e ethmoid sc

eb epibranchial scp

eh epihyal scl

epo epiotic se

ex exoccipital so

f frontal sop

hb hypobranchial sp

hm hyomandibular sy

hy hypural tp

ih interhyal tpf

iop interoperculum uh

la lachrymal up

le lateral ethmoid ur

Ih lower hypohyal v

lig ligament vo

lie lateral line canal

pterotic

posttemporal

pterosphenoid

pterygoid

quadrate

pectoral pterygial

sesamoid articular

scapular

scapular foramen

supracleithrum

supraethmoid

supraoccipital

suboperculum

sphenotic

symplectic

transverse process

facet on transverse process

upper hypohyal

upper pharyngeal

urohyal

vertebra

REFERENCES

ALLIS, E. P. 1909. The cranial anatomy of the Mail-Cheeked fishes. Zoologica, Stuttg. 22,

(57) : 1-129, 8 pis.
BOLIN, R. L. 1936. The systematic position of Indostomus paradoxus Prashad and Mukerji,

a freshwater fish from Burma. /. Wash. Acad. Sci. 26(io) : 420-423.
GOSLINE, W. A. 1963 Notes on the osteology and systematic position of Hypoptychus dybowskii

Steindachner, and other elongate perciform fishes. Pacif. Sci. 17(i) : 90-101.
GOSLINE, W. A. 1966. Comments of the classification of Percoid fishes. Ibid. 20(4) : 409-418.
GREENWOOD, P. H. ; ROSEN, D. E.; WEITZMAN, S. and Myers, G. 1966 Phyletic studies of

teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat. Hist.

131( 4 ) : 339-455-
GREGORY, W. K. 1933. Fish skulls : a study of the evolution of natural mechanisms. Trans.

Am. phil. Soc., New Ser. 23(2) : 75-481.
HARRINGTON, R. W. 1955. The osteocranium of the American cyprinid fish Notropis bifrenatus,

with an annotated synonomy of teleost skull bones. Copeia No. 4, 267-290.
JUNGERSEN, H. F. E. 1910. Ichthyotomal contributions. II. The structure of the Aulo-

stomidae, Syngnathidae and Solenostomidae. D. Kgl. Danske Vidensk. Selsk. Skrifter (7).

Raekke, Naturividensk. og Mattem. Afd VII 5.
JUNGERSEN, H. F. E. 1914. Some facts regarding the anatomy of the genus Pegasus. Rep.

Br. Ass. Advmt. Sci. 420-422.
NELSON, G. J. 1967. Gill arches of some teleostean fishes of the families Girellidae, Pomacen-

tridae, Embiotocidae, Labridae and Scaridae. /. Nat. Hist., 1 : 289-293.
PATTERSON, C. 1964. A review of the Mesozoic Acanthopterygian fishes, especially those of

the English chalk. Phil. Trans R. Soc. Ser B, 247(739) : 213-482.
PATTERSON, C. 1968. The Caudal skeleton in Lower Liassic Pholidophorid fishes. Bull. Br.

Mus. nat. Hist. (Geol.) 16(5) : 203-239.
PRASHAD, B. and MUKERJI, D. D. 1929. The fish of Indawgyi lake and the streams of the

Myitkyina district (Upper Burma). Rec. Indian Mus. 31 : 161-213.

TAXONOMY ON INDOSTOMUS PARADOXUS 209

REGAN, C. T. 1924. The morphology of a Rare Oceanic Fish, Stylophorus chordatus : based on

specimens collected in the Atlantic by the "Dana" Expeditions, 1920-1922. Proc. R. Soc.

Ser. B, 96 : 193-207.
ROSEN, D. E. 1964. The relationships and taxonomic position of the halfbeaks, Killifishes,

Silversides and their relatives. Bull. Am. Mus. nat. Hist. 127 (5) : 219-267.
ROSEN, D. E. and PATTERSON, C. 1969. The structure and relationships of the Paracanthop-

terygian fishes. Bull. Am. Mus. nat. Hist. 141 (3) : 357-474.
STARKS. E. C. 1930. The primary shoulder girdle in bony fishes. Stanf. Univ. Publs. Biol.

Sci. 6(2) : 1-92.
SWINNERTON, H. 1905. A contribution to the morphology and development of the pectoral

skeleton of teleosteans. Q. Jl. microsc. Sci. 194 (XLIX) : 363-382.
WEITZMAN, S. 1962. The osteology of Brycon meeki, a generalized characid fish with an

osteological definition of the family. Stanford ichthyol. Bull., 8 (i) : 1-77.

Dr. K. E. BANISTER

Department of Zoology

BRITISH MUSEUM (NATURAL HISTORY)

CROMWELL ROAD, LONDON S.W.7

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

NEW RECORDS, AND ONE NEW
SPECIES, OF TELEOST FISHES
FROM THE GILBERT ISLANDS

G. PALMER

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 6

LONDON : 1970

NEW RECORDS, AND ONE NEW SPECIES,

OF TELEOST FISHES FROM THE

GILBERT ISLANDS

BY

GEOFFREY PALMER

Pp. 211-234; 2 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 6

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19, No. 6 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. not. Hist. (Zool.)

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 19 June, 1970 Price Fifteen Shillings

(75 P.)

NEW RECORDS, AND ONE NEW SPECIES,
[. OF TELEOST FISHES FROM THE

f GILBERT ISLANDS

By G. PALMER

AN interesting collection of marine fishes, which has now been presented to the
museum, was made in 1962 by Mrs. M. J. Cooper, whilst she was living at Betio,
Tarawa in the Gilbert Islands.

Randall (1955) in his report on the fishes of the Gilbert Islands, stresses the fact
that there seem to be few species endemic to this group of islands, the majority of
those found in this area being common to many of the oceanic islands of the tropical
Pacific. In particular, the fauna is very similar to that of the Marshall Islands, some
of the icthyofauna of which has been dealt with by Schultz et al (1953 and 1960).
About 400 species were recorded by Randall, some of which are represented in the
present collection. Despite its comparatively small size of 297 specimens, the
present collection cosists of 75 species, 21 of which, including the new species, are
listed as new records for the Gilbert Islands. These are indicated by an asterisk.
In a number of instances live colour notes were made by Mrs. Cooper and these have
been included under the appropriate species.

APODES
MURAENIDAE

Echidna zebra (Shaw)

Gymnothorax zebra Shaw, 1797, Nat. Misc., 9: pi. 322
B.M. No. 1969.8.26.1-3.

3 specimens, 152-324 mm. in length. Taken off reefs at Bairiki, Tarawa.

The smallest example has 21 complete light bands encircling the body, the number
of these apparently increasing with size, as the other two larger specimens have 25
and 39 bands respectively. Between the complete bands are a varying number of
partial bands or spots.

Echidna nebulosa (Ahl)

Muraena nebulosa Ahl, 1789, Spec, ichthyologicum de Muraena et Ophichtho (Thunberg) : 7. pi.
right fig.

B.M. No. 1969.8.26.4-12.

9 specimens, 87-393 mm. in length, from bomb-crater tide-pool, Tarawa.

The teeth in the small examples are less molar-like than are those in the larger fish.

214 G. PALMER

The 315 mm. eel is anomalous in having the tubes of the anterior nostrils united,
although the nasal apertures are separate.

Echidna ? leucotaenia Schultz

Echidna leucotaenia Schultz, 1943, Butt. U.S. natn Mus. 180: 22.
B.M. No. 1969.8.26.13-14.

2 specimens, 39 and 40 mm. in length. Taken off Tarawa reefs.

Both specimens show the whitish snout and lower jaw, as well as the white margin
on the median fins, which are said by Schultz to be characteristic for this species.

* Pseudechidna brummeri (Bleeker)

Muraena brummeri Bleeker, 1858, Natuurk. Tijdschr. Ned. Indie 17: 137.
B.M. No. 1969.8.26.15.

I specimen, 350 mm. in length, from Tarawa lagoon.

Trunk 190 mm., tail 160 mm. Greatest depth n mm. Head 28 mm. Greatest
height of dorsal fin 5 mm. Median fins well-developed, the height of the dorsal in the
posterior part of the tail region being as high as the depth of the body. Dorsal,
caudal and anal fins with a marginal white band. Colour of body an overall light
brown with the anterior pores of the head and lower jaw a dark brown, as are the
tubes of the anterior nostrils. Teeth conical, depressible, a double row in the
maxillary series, those of the inner row being the larger. A single row in the inter-
maxillary series, with three mesial teeth. Vomer with a single row of 6 or 7 teeth.
Teeth of the lower jaw in a single series. This species was not recorded by Schultz
(1953) from the Marshall and Marianas Islands and appears to be a new record for the
Gilbert Islands.

Murenophis pardalis (Temminck & Schlegel)

Muraena par dalis Temminck & Schlegel, 1846, Fauna japonica Pisces Pt. 5: 268 pi. 119.
B.M. No. 1969.8.26.33.

i specimen, 132 mm. in length from Tarawa lagoon.

Colour brownish, with prominent white spots which tend to coalesce to form
vertical white lines.

Gymnothorax picta (Ahl)

Muraena picta Ahl, 1789, Spec, ichthyologicum de Muraena et Ophichtho (Thunberg) : 8.
B.M. No. 1969.8.26.16-26.

ii specimens, 72-474 mm. in length, taken from Tarawa lagoon.

Gymnothorax petelli (Bleeker)

>, Natuurk. Tijdschr. Ned. Indie 11
i specimen, 112 mm. in length, from Tarawa lagoon.

Muraena petelli Bleeker, 1856, Natuurk. Tijdschr. Ned. Indie 11 : 84.
B.M. No. 1969.8.26.27.

TELEOST FISHES FROM THE GILBERT ISLANDS 215

Gymnothorax riipelli (McClelland)

Dalophis riipelliae McClelland, 1845, Calcutta J . nat. Hist. 5: 213.
B.M. No. 1969.8.26.28-32.

5 specimens, 50-212 mm. in length, from Tarawa lagoon.

The three smallest examples have a line of solid, dark pigment running longitudin-
ally along the ventral surface, linking the dark rings.

Gymnothorax thyrsoideus (Richardson)

Muraena thyrsoidea Richardson, 1844, Zool. Voy. "Sulphur" Fishes: in.
B.M. No. 1969.8.26.35-38.

4 specimens, 202-270 mm. in length, from Tarawa lagoon.
Gymnothorax undulata (Lacepede)

Muraenophis undulatus Lacepede, 1803, Hist. nat. Poissons 5: 629, 642, 644 fig. 2.
B.M. No. 1969.8.26.39-40.

2 specimens, 114 and 302 mm. in length from Tarawa lagoon.
It is with some hesitation that the 114 mm. specimen is included in this species.
Colouration in life was stated to have been pale brown, mottled with darker brown.

Gymnothorax favaginea Schneider

Gymnothorax favagineus Schneider, 1801, Syst. Ichth. Block. : 525.
B.M. No. 1969.8.26.41-42.

2 specimens, 52 and 62 mm. in length from Tarawa lagoon.

These two specimens have the snout and the distal margins of the fins white. The
smaller of the two is very dark and shows only light interspaces on the dorsal half of
the body.

Gymnothorax fimbriata (Bennett)

Muraena fimbriata Bennett, 1831, Proc. zool. Soc. Lond. Pt. i : 168.
B.M. No. 1969.8.26.43-44.

2 specimens, 159 and 273 mm. in length. Tarawa.

The smaller specimen was taken on the reef off Tarawa at night. The larger was
taken from Tarawa lagoon over sand and rock. Live colours are stated to have been
a cream background mottled with brown.

Gymnothorax flavimargimata (Riippell)

Muraena flavimarginata Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres: 119,
pi. 30, fig. 3.

B.M. No. 1969.8.26.45-51.

7 specimens, 117-283 mm. in length, taken on the outer reef, Tarawa.
Live colour brown, mottled with darker brown. A bright greenish margin to
dorsal fin and around caudal.

216 G. PALMER

Gymnothorax sp.
B.M. No. 1969.8.26.52-54.

3 specimens, 52-80 mm. in length, taken in Tarawa lagoon.

I have been unable to assign these specimens to any particular species.

* Uropterygius leucurus Snyder

Uropterygius leucurus Snyder, 1904, Bull. U.S. Fish. Comm. 22: 521, pi. 6 fig. 12.
B.M. No. 1969.8.26.55-65.

ii specimens, 87-123 mm. in length. Taken on Betio, Tarawa at low water in a
tide-pool with rotenone.

This appears to be a new record for the species from the Gilbert Islands. The type
was taken in the Hawaiian Islands, and more recently has been recorded by Schultz
in Schultz et al (1953) from the Marshall and Marianas Islands.

Trunk 42 mm., tail 45 mm. About 94 pores in the lateral line.

Colour in alcohol overall brownish. On the smaller specimens there are some 70
alternating vertical bars, sometimes incomplete, of brown and white. These are
visible from the rictus of the jaws back to the end of the body.

Uropterygius micropterus (Bleeker)

Muraena micropterus Bleeker, 1852, Natuurk. Tijdschr. Ned. Indie 3: 298.
B.M. No. 1969.8.26.66-69.

4 specimens, 200-233 mm. in length. Taken at Bairiki, Tarawa on outer reef,
close inshore in tide-pools. Very common in this area.

I follow Randall (1955) in considering this to be distinct from U. marmoratus,
although some authors believe these two species to be synonymous.

The maxillary teeth are in two series, the inner being two to three times as long as
those of the outer row. They are canine-like, depressible and extend the whole length
of the maxillary. The mandibular teeth are also in two series, the inner row again
being the longer and extending almost to the rictus of the jaws. There are 6-7
vomerine teeth in a single series.

Live colour is stated to be white, speckled with greyish-black. Colour in alcohol is
a greyish-white, speckled with darker markings along the dorsal side of the body.
Ventral surface an unrelieved greyish-white.

Uropterygius marmoratus (Lacepede)

a Lacepede, 1803, Hist. nat. Poissons 5: 648.
-77-

8 specimens, 40-460 mm. in length, from Tarawa lagoon.

Gymnomuraena marmorata Lacepede, 1803, Hist. nat. Poissons 5: 648.
B.M. No. 1969.8.26.70-77.

TELEOST FISHES FROM THE GILBERT ISLANDS 217

* Anarchias allardicei Jordan & Starks

Anarchias allardicei Jordon & Starks, in Jordan & Scale, (1905) 1906, Bull. U.S. Bur. Fish. 25:
204, fig. 9.
B.M. No. 1969.8.26.68.

1 specimen, 73 mm. in length. Taken in bomb-crater tide-pool on the southern
reef, Betio, Tarawa.

This specimen is almost certainly a young example of A. allardicei. It is of a
plain, darkish brown colour, much lighter around the caudal fin. The snout and
lower jaw are very pale. The tooth pattern agrees very closely with the diagram in
Schultz in Schultz et al (1953) for this species. Previously recorded from Samoa
and the Marshall and Marianas Islands, it is not surprising to find that it occurs in
this region, from which it has not previously been recorded.

OPHICHTHIDAE

Randall (1955) follows Myers and Storey (1929) and Gosline (1950) including in
this family the genus Muraenichthys. I have followed this course here.

Myrichthys maculosus (Cuvier)

Muraena maculosa Cuvier, 1817, Regne Animal (ist. Ed.) 2: 232.
B.M. No. 1969.8.26.79-80.

2 specimens, 98 and 265 mm. in length. Taken on the outer reef at Bairiki, Tarawa,
on sand with rocks.

The live colour of the larger specimen was a deep cream, with reddish-brown spots.
There are 28 of these in a lateral series, with a few smaller ones along the dorsal edge
of the body in the interspaces. The smaller example shows only a few groups of
melanophores along the sides of the body.

Myrichthys colubrinus (Boddaert)

Muraena colubrina Boddaert, 1781, Neue nord. Beytr. 2: 56, pi. 2 fig. 3.
Ophisurus fasciatus var. semicincta Bleeker, 1864, Atlas ichthyol. Indes orient. Neerl. 4: 64.
Chlevastes elaps Fowler, 1912, Proc. Acad. nat. Sci. Philad. 64: 13.
B.M. No. 1969.8.26.81-82.

2 specimens, 398 and 628 mm. in length, taken off the reef at night, Betio, Tarawa.

Schultz in Schultz et al (1953) states that the species M . semicinctus (Bleeker) and
M. elaps (Fowler) are distinct from M . colubrinus, primarily on the form and com-
pleteness of the vertical bands, with the presence or absence of additional dark spots
or markings in the interspaces. Randall (1955) and Smith (1962), however, con-
sider these two forms to be synonymous with M. colubrinus. It may be noted that
of the two specimens listed above the smaller keys out, according to Schultz, as
semicinctus, whilst the larger represents elaps. Smith (1962) gives an illustration
showing both forms under the name colubrinus. Here again, the smaller of these is a
semicinctus according to Schultz, whilst the larger shows the wider bands and many
dark spots and partial bands in the interspaces, of an elaps.

2i8 G. PALMER

Callechelys melanotaenia Bleeker

Cattechelys melanotaenia Bleeker, 1864, Atlas Ichth. Ind. Neerl. ... 4: 66, pi. 193 fig. 2.
B.M. No. 1969.8.26.83-86.

4 specimens, 449-543 mm. in length. Taken with rotenone in bomb-crater tide-
pool, Betio, Tarawa.

Callechelys sp.

B.M. No. 1969.8.26.87.

i specimen, 86 mm. in length, taken in bomb-crater tide-pool, Betio, Tarawa.
This specimen shows no colour pattern, other than a few circular groups of melano-
phores scattered along the sides of the body.

* Muraenichthys macropterus Bleeker

Muraenichthys macropterus, Bleeker, 1857, Act. Soc. Sclent. Ind. Neerl. 2: 91.
B.M. No. 1969.8.26.88.

I specimen, 235 mm. in length, taken in Tarawa lagoon.

Although this species has been recorded from the Tuamotu Archipelago and from
the Marshall and Marianas Islands (Schultz in Schutz et al 1953), it has not previously
been listed from the Gilbert and Ellice Islands.

Greatest depth 7 mm. ; head length 24 mm. ; trunk (from snout tip to vent) 90 mm. ;
length of tail 143 mm. ; origin of dorsal fin 39 mm. anterior to vent.

Live colour was a sandy pink, with pale fins. Colour of preserved specimen a light
brown.

Muraenichthys sp.

B.M. No. 1969.8.26.89.

i specimen, 118 mm. in length. Taken in bomb-crater tide-pool on southern reef,
Betio, Tarawa.

I am unable to assign a specific name to this specimen. Measurements are as
follows: depth of body 3 mm.; head 12 mm.; trunk 43 mm.; tail 75 mm.; dorsal
origin is 4 mm. in advance of vent. Maxillary and mandibular teeth in single series,
vomerine teeth biserial.

Phaenomonas

Phaenomonas Myers and Wade, 1941, Allan Hancock Pac. Exped. 9: 77, pi. 12.
B.M. No. 1969.8.26.90.

This genus has been known only from the type species, P. pinnata, taken off the
Pacific coast of tropical America, notably from Colombia and Isabel Island, Mexico.
The specimen reported on here is considered to be distinct from P. pinnata and is
described below. It is necessary, also, to modify the generic diagnosis given by
Myers and Wade by altering the statement concerning the length of the dorsal fin.
The amended description is given herewith.

Body elongate, cylindrical, very wormlike. Head and trunk much longer than

TELEOST FISHES FROM THE GILBERT ISLANDS 219

tail. A short, well-developed dorsal fin is present commencing on the head approx-
imately half way between the snout and gill-opening; its entire length is not more
than three times the head length. No pectoral or anal fins present. Snout blunt,
lower jaw included. Eye small and almost hidden beneath skin, closer to angle of
jaw than to tip of snout. Anterior nostrils tubular, on underside of snout. Posterior
nostrils on underside of upper lip, the pore covered by a small membrane. Gill
openings nearly vertical, low lateral. Teeth uniserial. 1

* Phaenomonas cooperae sp. nov.

Fig. i

HOLOTYPE: Brit. Mus. (Nat. Hist.) 1969.8. 26. 90. 513 mm. in length.
LOCALITY : Lagoon reef at Betio, Tarawa.

Body extremely elongate, cyclindrical. Tail ending in a blunt, conical point.
Depth of body contained 128 times in total length. Tail very short, contained 2-3
times in length of head and trunk. Dorsal fin present, originating midway between
tip of snout and gill-opening and extending back to above the 39th vertebra; it is
57 mm. long and almost 3 times the head length. No anal or pectoral fins. There
are about 260 pores in the lateral line from above the gill-opening to the end of the
tail. The anterior nostrils are tubular, placed on the underside of the snout. The
posterior nostrils are under the edge of the upper lip, covered by a flap of skin, and
lie below the posterior half of the eye. The eyes are small, about 1/3 snout length and
almost completely hidden beneath the integument and probably virtually non-
functional. The gill-openings are oblique, low lateral in position, their length of

3 mm. being equal to the width of the isthmus. Ventral surface of snout grooved,
with an enlarged premaxillary tooth in the groove. Teeth uniserial. Maxillary
teeth, as in P. pinnata, running in a straight line from posterior insertion to the head
of the vomer. Vomerine teeth in a single row, widely spaced, about 5 in number.
Only three premaxillary teeth present, one on either side of the groove in the snout
and partially hidden by a flap of skin. An enlarged median tooth protruding from
the groove itself. Mandibular teeth in a single row on either side, becoming smaller
posteriorly. All the teeth are conical.

Colour in alcohol a sandy brown, with no trace of pigment.

Summary of counts and measurements. Total length 513 mm.; depth of body

4 mm. ; head and trunk 357 mm. ; tail 156 mm. ; head 18 mm. ; snout 3 mm. ; eye i mm.
Depth in length 128-25; head in trunk 19-8; tail in trunk 2-3; eye in snout 3; dorsal
fin 3 times head length ; length of dorsal fin 57 mm. ; number of lateral line pores 260 ;
number of vertebrae 270.

1 After this paper had gone to press, Dr. McCosker of the Scripps Institution of Ocean-
ography informed me that he had 32 adult examples from the western Caroline Islands that
appeared to be a new species of Phaenomonas, and offered some of this material to me for
study. He also told me that Mr. Wayne Baldwin had obtained an adult specimen of this genus
from Hawaii. For a full treatment of this genus, therefore, reference should be made to a joint
paper which is to be published by Dr. McCosker and Mr. Baldwin.

220 G. PALMER

Through the courtesy of the authorities of the Los Angeles County Museum, I
have been enabled to examine the paratype of P. pinnata, which is 233 mm. in length.
P. cooper ae differs from the type species in having a proportionately longer dorsal fin,
3 times head length compared with less than head length ; a larger number of lateral
line pores, 260 as against 175, and the eye contained 3 times in snout as against 5-6
times in P. pinnata.

4**"* **S<.

FIG. i. X-ray of Phaenomonas cooperae. D-Di = position of dorsal fin.

The species is named in honour of Mrs. Jane Cooper, without whose enthusiasm this
collection would not have been made.

MORINGUIDAE

Randall (1955) states that he follows Gosline and Strasburg (1956) in classifying
the worm eels. I am following the same course here, so that material which keyed
out in Schultz (1953) as M. abbreviata and M. macrocephala is here listed as M.
macrochir. P. H. J. Castle (in litt.) states that he is convinced from pore and vertebral
counts that M. macrochir and M. macrocephala are identical.

TELEOST FISHES FROM THE GILBERT ISLANDS 221

Moringua macrochir Bleeker

Moringua macrochir Bleeker, 1855, Natuurk. Tijdschr. Ned. Indie 9: 71.
Aphthalmichthys abbreviates Bleeker, 1863, Ned. Tijdschr. Dierk. 1: 163.
Aphthalmichthysmacrocephalus Bleeker, 1863, Ned. Tijdschr. Dierk. 1: 165.
B.M. No. 1969.8.26.91-125.

35 specimens, 144-368 mm. in length. Taken at Betio, Tarawa in bomb-crater
tide-pools, with rotenone.

Colour in alcohol a uniform light yellowish brown. The number of lateral line
pores from above the gill-opening to the end of the tail varies between 93-120.

Moringua javanica (Kaup)

Aphthalmichthys javanicus Kaup, 1865, Cat. Apodal Fish. Brit. Mus. 105, fig. 71.
B.M. No. 1969.8.26.126-129.

4 specimens, 340-740 mm. in length. Taken at Betio, Tarawa in bomb-crater
tide-pool.

Colour in alcohol a light greyish brown. The number of lateral line pores varies
from 138-148.

CONGRIDAE
* Conger cinereus cinereus Riippell

Conger cinereus Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres: 115.
B.M. No. 1969.8.26.130-131.

2 specimens, 538 and 735 mm. in length. Taken off Tarawa reefs at night.

These specimens represent a new record for the Gilbert Islands. The sixth
infraorbital pore is above the rictus of the jaw. Pectoral fin with 15-16 rays. A
black spot is present on this fin in both specimens.

According to Kanazawa (1958), this subspecies ranges through the Marshall,
Phoenix and Christmas Islands, Palmyra, New Caledonia, Guam, Okinawa, Japan,
the Philippines westward through the Indian Ocean, the Red Sea and the east and
south coasts of Africa.

It is believed to be nocturnal in its habits and this is borne out by the fact that both
these examples were taken at night.

HEMIRHAMPHIDAE
Hyporhamphus dussumieri (Val.)

Hemirhamphus dussumieri Valenciennes, 1846, Hist. nat.Poissons 19: 33, pi. 554.
B.M. No. 1969.8.26.132-135.

4 specimens, 73-96 mm. in S.L. Taken in Tarawa lagoon.

In June, 1968, 3 additional specimens of this species, collected by Dr. D. N. F.
Hall, were received from the same locality. These were 245-263 mm. in S.L.

222 G. PALMER

SYNGNATHIDAE

Doryrhamphus melanopleura (Bleeker)

Syngnathus melanopleura Bleeker, 1858, Natuurk. Tijdschr. Ned. Indie 15: 464.
B.M. No. 1969.8.26.136.

i specimen, 48 mm. in S.L., from Tarawa lagoon.

SERRANIDAE

* Epinephelus kohleri Schultz

Fig. 2

Epinephelus kohleri Schultz, 1953, Bull. U.S. Nat. Mus. No. 202 Pt. i : 336.

Serranus cyanopodus Richardson 1846. Whitehead, 1969, Bull. Br. Mus. nat. Hist. (Zool.) 3: 212,

pi. 3 fig. a.

B.M. No. 1969.8.26.137.

i specimen, 280 mm. in S.L. Taken in deep water at the entrance to Tarawa
lagoon.

D. XI 16, the fourth spine being the longest. A. Ill 8, the third spine the longest.
Pect. I 16; depth 105 mm. ; head in nim. ; eye 18 mm. ; Depth in S.L. 2-6; head in
S.L. 2-5; eye in head 6-2; interorbital space 21 mm. Gillrakers on first arch 9+16.
Scales ctenoid posteriorly, cycloid anteriorly in front of a line drawn between the
origins of the dorsal fin and the pelvic fins.

FIG. 2. Photograph of Epinephelus kohleri, to show markings.

Live colours pale blue all over, with many brown spots and larger blotches.

This species is apparently quite rare in these waters, but it is known to most of the
deep-water fishermen, who fish at night. They say that this species may be poison-
ous in certain localities, as in the entrance to Tarawa lagoon. This particular
specimen is not so deep-bodied as was a previous one seen by Mrs. Cooper, nor had it
so many of the larger blotches on the body. Chan (1968) has recently reported this

TELEOST FISHES FROM THE GILBERT ISLANDS 223

species for the first time from the Macclesfield Bank and Pratas Reef in the South
China Sea.

This is a new record for the Gilbert Islands, the species previously being known
from Rongarik and Bikini Atolls in the Marshall and Marianas Islands and from New
Caledonia.

Epinephelus hexagonatus (Schneider)

Holocentrus hexagonatus Schneider, 1801, Syst. Ichth. Block. : 323.
B.M. No. 1969.8.26.138-141.

4 specimens, 63-161 mm. in S.L. Taken at Betio, Tarawa in bomb-crater tide
pool.

The markings on the two smaller examples tend to be more distinct than in the
larger specimens. Gill-raker counts on first arch, including rudiments, are between
24-25. Pectoral rays 18.

Epinephelus merra Bloch

Epinephelus merra Bloch, 1793, Naturg. ausland. Fische 7: Pt. 10 : 17.
B.M. No. 1969,8.26.142-144.

3 specimens, 125-153 mm. in S.L. Taken at Betio, Tarawa in bomb-crater tide
pool.

Gill-raker counts on first arch, including rudiments, 21-23.

Epinephelus corallicola Val.

Epinephelus corallicola Valenciennes, 1828, Hist. nat. Poissons2: 336.
Epinephelus macrospilus Bleeker, 1873-1876, Atlas ichthyol. Indes orient. Neerl. 7: 33.
B.M. No. 1969.8.26.145-146.

2 specimens, 132 and 138 mm. in S.L. Taken at Betio, Tarawa in bomb-crater
tide pool with the previous two species.

According to Schultz in Schultz et al (1953) these specimens would be E. macro-
spilus (Bleeker), which he considers a valid species. I follow Randall (1955) in this
instance and consider E. macrospilus to be a synonym of E. corallicola.

* Epinephelus summana (Forsskal)

Perca summana Forsskal, 1775, Descriptiones Animalium . . . : xi, 42.
B.M. No. 1969.8.26.147.

I specimen, 96 mm. in S.L. Taken at Betio, Tarawa in lagoon reef.

D. XI 16; A. Ill 9; pectoral 16; head 39 mm. ; eye 7-5 mm. ; interorbital 5 mm.

Live colour light brown, with pale blue-green spots scattered over the body and
unpaired fins.

This species does not appear to have been recorded previously from the Gilbert
Islands.

224 G. PALMER

Epinephelus sp.

B.M. No. 1969.8.26.148-149.

2 specimens, 39 and 43 mm. in S.L. taken at Betio, Tarawa.

I have not found it possible satisfactorily to assign either of these two juveniles to a
definite species.

Cephalopholis argus Schneider

Cephalopholis argus Schneider, 1801, Syst. Ichth. Block. : 311.
B.M. No. 1969.8.26.150-153.

4 specimens, 70-90 mm. in S.L. Taken at Betio, Tarawa in bomb-crater tide pool.

PSEUDOCHROMIDAE

For the purpose of this paper, I am following Schultz (1953) and Randall (1955) in
placing Plesiops in the family Pseudochromidae, with the genera Pseudochromis,
Pseudogramma and Pseudoplesiops.

Plesiops nigricans (Riippell)

Pharopteryx nigricans Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres: 15,
pi. 4 fig. 2.

B.M. No. 1969.8.26.154.

i specimen, 68 mm. in S.L. Taken at Betio, Tarawa in bomb-crater tide pool.
Live colour was a dark brown, with numerous light blue-green spots on the body
and unpaired fins.
D. XII 7; A. Ill 8.

* Plesiops melas Bleeker

Plesiops melas Bleeker, 1849, Verh. batav. Genoot. 22: 9.
B.M. No. 1969.8.26.155-157.

3 specimens, 30-40 mm. in S.L. Taken at Betio, Tarawa in lagoon reef.

D. X-XI 7-8; A. Ill 8. This species, which is a new record for the Gilbert
Islands, may be distinguished from P. nigricans by the lower number of dorsal fin
spines and by the lack of an ocellated blackish spot on the operculum.

APOGONIDAE

Apogon novemfasciatus Cuvier

Apogon novemfasciatus Cuvier, 1828, Hist. nat. Poissons 2: 154.
B.M. No. 1969.8.26.158-162.

5 specimens, 38-76 mm. in S.L. Taken at Betio, Tarawa in tide pool.

TELEOST FISHES FROM THE GILBERT ISLANDS 225

CARANGIDAE
Alectis ciliaris (Bloch)

Zeus ciliaris Bloch, 1788, Naturg. ausland. Fische: 29.
B.M. No. 1969.8.26.163.

1 specimen, 329 mm. in S.L. Taken in Tarawa lagoon, where it was netted.

This appears to be only the second record of this species from the area, the first
having been that of Giinther (1876), listed as from the Kingsmill Islands. This fish
is seen in shoals at infrequent intervals at Tarawa, when it is caught in nets in the
lagoon. At the time that the present specimen was taken, several examples had been
caught within the space of a few months. This caused great interest amongst the
younger Gilbertese, most of whom had never before seen this fish. Live colour was
bluish above and plain silvery below, with dorsal and anal fins banded with black and
white.

CHAETODONTIDAE

Chaetodon lunula (Lacepede)

Pomacentrus lunula Lacepede, 1802, Hist. nat. Poissons 4: 507.
B.M. No. 1969.8.26.164-166.

3 specimens, 64-91 mm. in S.L. from Tarawa lagoon.

This species has a broad black ocular band extending from the head to the border
of the preoperculum. Posterior to this is a pale band. There is a triangular shaped
dark band extending from above the pectoral base to the middle of the spinous dorsal,
where it is narrowest. There is a large black blotch on the caudal peduncle, narrow-
ing to a thin line as it extends forward along the base of the soft dorsal. In smaller
examples there is a black blotch, surrounded by a paler band, at the base of the
middle of the soft dorsal.

Chaetodon vagabundus Linnaeus

Chaetodon vagabundus Linnaeus, 1758, Syst. Nat. (Ed. 10) : 276.
B.M. No. 1969.8.26.167-168.

2 specimens, 51 and 81 mm. in S.L. from Tarawa lagoon.

There is an ocular band commencing just forward of the dorsal origin and extending
through the eye to the ventral edge of the preoperculum. Dorsally, this is narrower
than the eye, ventrally broader than the eye. Dorsal fin with a broad marginal
black band, not continued on to the last few rays. There is a broad dark band run-
ning along the base of the soft dorsal and extending across the caudal peduncle on to
the last few anal rays. Caudal fin with a broad, crescent shaped dark band across the
centre of the fin. Posterior to this is a narrow dark submarginal band, with the
margin of the fin hyaline. On the body are about 6 thin black lines running obliquely
backwards and upwards from behind the head and 10 to n similar dark lines running
obliquely backwards and downwards towards the anal base.

226 G. PALMER

POMACENTRIDAE

Amphiprion bicinctus Riippell

Amphiprion bicinctus Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres : 139.
B.M. No. 1969.8.26.169.

1 specimen, 79 mm. in S.L. from Tarawa lagoon.

LABRIDAE

Stethojulis axillaris (Quoy & Gaimard)

Julis axillaris Quoy & Gaimard, 1824, Voy. de I'Uranie et de laPhysicienne Zool. : 272
B.M. No. 1969. 8.26.170-178.

9 specimens, 49-61 mm. in S.L. from Tarawa lagoon.

This species appears to be quite common throughout Oceania.

Halichoeres marginatus Riippell

Halichoeres marginatus Riippell, 1835, Neue Wirbelth. Fische: 16.

Julis notopsis Valenciennes, 1839, in Cuvier & Valenciennes, Hist. nat. Poissons 13: 485.
B.M. No. 1969.8.26.179-180.

2 specimens, 37 and 42 mm. In S.L. from bomb-crater tide pool at Betio, Tarawa.
The prominent ocellated black spot on the dorsal fin is still very clear in both these

young specimens. Both Randall (1955) and Schultz in Schultz et al (1960) consider
that Julis notopsis Cuvier and Valenciennes represents the juvenile form of this
species.

Halichoeres trimaculatus (Quoy & Gaimard)

Julis trimaculata Quoy & Gaimard, 1834, Voy. "Astrolabe" Zool. 3: 705.
B.M. No. 1969.8.26.181-207.

27 specimens, 26-103 mm. in S.L. from bomb-crater tide pool, Betio, Tarawa.
Very abundant in this area.

Thalassoma lunare (Linnaeus)

Labrus lunaris Linnaeus, 1758, Syst. Nat. (Ed. 10) : 283.
B.M. No. 1969.8.26.208-211.

4 specimens, 53-63 mm. in S.L. from Tarawa lagoon.

Thalassoma amblycephalus (Bleeker)

Julis amblycephalus Bleeker, 1856. Natuurk. Tijdschr. Ned. Indie 11 : 83.
B.M. No. 1969.8.26.212.

i specimen, 30 mm. in S.L. from Tarawa lagoon.

This is the first record for this species from the Gilbert and Ellice Islands, although
it is widely distributed throughout the south west Pacific and the Philippines.

TELEOST FISHES FROM THE GILBERT ISLANDS 227

The colour pattern in the young of this species is somewhat different from that of
the adults, consisting of a black band extending from the snout, through the eye, to
the base of the caudal fin, its width being about equal to eye diameter. There is a
second black band commencing on the head between the eyes and extending back
along the side of the dorsal fin, disappearing about halfway along the base of the soft
dorsal fin.

D. VIII 13; A. Ill 14. Pores in lateral line 19+2+6.

ELEOTRIDAE

v-,

Eleotriodes sexguttatus (Valenciennes)

Eleotris sexguttatus Valenciennes, 1837, Hist. nat. Poissons 12: 254.
B.M. No. 1969.8.26.213-214.

2 specimens, 40 and 76 mm. in S.L. from bomb-crater tide pool, Betio, Tarawa.
This species appears to be widely distributed throughout the western South Pacific,
the Philippines and the Indian Ocean to Ceylon.

* Asterropteryx semipunctatus Riippell

Asterropteryx semipunctatus Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres: 138
B.M. No. 1969.8.26.215.

I specimen, 23 mm. in S.L., taken at Betio, Tarawa from bomb-crater tide pool on
the south reef.

Although widely distributed throughout the west Pacific, Hawaii and in the
Indian Ocean, this is the first record of this species from the Gilbert Islands.

D. VI 8; A. I 7. Lateral line scales 24. This specimen has three spines at the
angle of the preoperculum and has the second and third rays of the first dorsal pro-
longed beyond the membrane.

GOBIIDAE

Bathygobius fuscus (Riippell)

Gobiusfuscus Riippell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres : 137.
B.M. No. 1969.8.26.216-222.

7 specimens, 28-53 mm. in S.L. Taken in bomb-crater tide pool at Tarawa.

* Acentrogobius cauerensis (Bleeker)

Gobius cauerensis Bleeker, 1853. Natuurk. Tijdschr. Ned. Indie 4: 269.
. B.M. No. 1969.8.26.223-237.

15 specimens, 29-41 mm. in S.L. Taken from bomb-crater tide pool at Betio,
Tarawa.

A new record for the Gilbert and Ellice Islands, although known from a wide area
of the Pacific.

228 G. PALMER

D. VI ii ; A. I ii. Pect. 16. Lateral line scales 29-30. Cheek and operculum
scaled. No free silk-like rays in pectoral fin.

Acentrogobius sp.

B.M. No. 1969.8.26.238-239.

2 specimens, 18 and 20 mm. in S.L. from bomb-crater tide pool, Tarawa.

Callogobius sclateri (Steindachner)

Eleotris sclateri Steindachner, 1880, Sitz. her. Akad. Wiss. Wien 80: 157.
B.M. No. 1969.8.26.240

i specimen, 27 mm. in S.L. from tide pool, Betio, Tarawa.

D. VI 1 9 ; A. I 8 ; Pect. 16. Lateral line scales ca. 31 . Scales are cycloid anteriorly,
ctenoid posteriorly. There is a raised row of papillae on the snout, two rows on each
cheek and one row transversely behind each eye. There are no free silk-like rays in
the pectoral fin and no basal membrane on the pelvic fins.

This species was provisionally recorded from the Gilbert Islands by Randall (1955)
as Mucogobius sclateri.

Quisquilius eugenius Jordan & Evermann

Quisquilius eugenius Jordan & Evermann, 1904, Bull. U.S. Fish Comm. 22: 203.
B.M. No. 1969.8.26.241.

i specimen, 16 mm. in S.L. from Tarawa lagoon.

Not previously recorded from the Gilbert Islands. Apart from the number of pre-
dorsal scales, which appear to be absent in this specimen, this fish agrees in all
respects with the description of Q. eugenius, as well as with an example sent for
comparison by the late Professor J. L. B. Smith.

* Quisquilius inhaca (Smith)

Gobius inhaca Smith, J. L. B., 1949, Ann. Mag. nat. Hist. (12) 2: 103.
B.M. No. 1969.8.26.242-244.

3 specimens, 19-23 mm. in S.L. from Tarawa lagoon.

D. VI 10 ; A I 8; pect. 16-17. Lateral line scales 28.

These specimens show the characteristic reticulate appearance of the body caused
by the edges of the scales being outlined by a dark line, and by the three to four light
vertical bars on the head region. This species has not previously been recorded from
the Gilbert Islands, and its range must now be extended from Inhaca, Mozambique
to this area.

TELEOST FISHES FROM THE GILBERT ISLANDS 229

BLENNIIDAE
Istiblennius edentulus (Schneider)

Blennius edentulus Schneider, 1801, Syst. Ichth. Bloch. : 172.
B.M. No. 1969.8.26.245-246.

2 specimens, 56 and 68 mm. in S.L. from a tide pool at Bairiki, Tarawa.

D. XIII 20; A. II 22. Live colour grey and green on the back. Dorsal fin grey-
green, with lines of darker spots. Outer edge of fin a brilliant blue-black. Anal fin a
pale greenish grey, with two pale blue lines shading to blue-black. Extreme edge of
fin a pale blue. Pectoral fins yellowish, pelvics hyaline.

Istiblennius lineatus (Valenciennes)

Salarias lineatus Valenciennes, 1836, Hist. nat. Poissons 11: 314.
B.M. No. 1969.8.26.247.

1 specimen, 99 mm. in S.L. from Tarawa tide pool.
D. XII 22; A. 1124.

Istiblennius paulus (Bryan & Herre)

Salarias paulus Bryan & Herre, 1903, Occ. Pap. B. P. Bishop Mus. 2: 136.
B.M. I969.8.26.248.-249.

2 specimens, 27 and 40 mm. in S.L. from outer reef, tide pools at Bairiki, Tarawa,
where it is quite common.

It is with some hesitation that I include the smaller of these two specimens in this
species.

Live colour, upper 2/3 of body pale brown, spotted with red, the lower 1/3 being a
pale blue. Six dark brown inverted Y-shaped bars, with two turquoise dots per bar.
Head brown with red spots. Preoperculum yellow. First dorsal spotted with red;
soft dorsal rays cream, the membrane brown. The first half of this fin is also spotted
with red. Anal fin becoming a dusky brown distally. Pectorals hyaline, pelvic fins
white.

* Istiblennius cyanostigma (Bleeker)

Salarias cyanostigma Bleeker, 1849, Verh. Batav. Genoots. 22: 5.
B.M. No. 1969.8.26.250-254.

5 specimens, 43-58 mm. in S.L. from Tarawa lagoon.

D. XIII 20; A. II 20; pect. 14. A well developed nuchal crest present. Supra-
orbital cirrus simple. Nasal cirrus multifid. No nuchal cirrus. A well developed,
recurved canine present on each side of the lower jaw. No well defined colour
pattern evident on any of these specimens. Although fairly widely distributed in the
central and south tropical Pacific, this species does not appear to have been recorded
previously from the Gilbert Islands.

230 G. PALMER

* Petroscirtes mitratus Ruppell

Petroscirtes mitratus Ruppell, 1828, A Has Reise nord. Afrika. Fische des rothen Meeres : 1 1 1 .
B.M. No. 1969.8.26.255-256.

2 specimens, 17 and 43 mm. in S.L. Taken on the reef edge in the lagoon at
Bairiki, Tarawa.

D. 26; A. 17; pect. 14. Origin of dorsal fin between hind edge of eye and pre-
operculum. Colour in alcohol brownish, with dark blotches or transverse bands on
body. Five dark edged ocelli along the dorsal part of the body.

Although widely distributed throughout the Indo-Australian region from the Red
Sea, down the coast of East Africa, across the Indian Ocean to the Philippines,
Australia and the western South Pacific, this species has not previously been recorded
from the Gilbert Islands.

CLINIDAE

Tripterygion minutus Giinther

Tripterygium minutum Giinther, 1877, /. Mus. Godeffroy 4 : Pt 13: 211.
B.M. No. 1969.8.26.257.

1 specimen, 20 mm. in S.L. from tide pool on edge of outer reef, Tarawa.
Live colour stated to have been banded with green and cream.

SGORPAENIDAE

Randall (1955) listed only six species of scorpion fishes from Onotoa. Three of the
five species dealt with here represent new records for the area.

Scorpaena albobrunneus Giinther

Scorpaena albobrunneus Giinther, 1873, /. Mus. Godeffroy 2 : Pt 3 : 77.
B.M. No. 1969.8.26.258-259.

2 specimens, 44 and 48 mm. in S.L. Taken on the west reef at Betio, Tarawa
amongst branches of coral.

Live colours brown, with patches of bright yellowish green. Pectorals yellow.

* Scorpaenodes guamensis (Quoy & Gaimard)

Scorpaena guamensis Quoy & Gaimard, 1824, Voy. de I'Uranie etde laPhysicienne, Zool. : 326.
B.M. No. 1969.8.26.260-270.

ii specimens, 39-63 mm. in S.L. from the S.W. outer reef, Betio, Tarawa.

D. XIII 7-8 ; A III 5 ; pect. 18. Lateral line pores 24-25.

Live colour a rich mottled brown. Pectoral and caudal fins clear, with bright
orange spots.

Although widely distributed throughout the Indo-Australian archipelago and the
Philippines, this species has not been recorded before from the Gilbert Islands.

TELEOST FISHES FROM THE GILBERT ISLANDS 231

* Sebastapistes nuchalis (Giinther)

Scorpaena nuchalis Giinther, 1874, /. Mus. Godeffroy 2-3: Pts 5-6: 76.
B.M. No. 1969.8.26.271-279.

9 specimens, 17-43 mm. in S.L., from the S.E. reef at Betio, Tarawa.

D. XII 8-9; A. Ill 5; pect. 15. Lateral line with 22-24 tubular scales. Teeth
present on vomer and palatines. Body scales mostly ctenoid.

Not previously recorded from this group of islands. Stated by Smith (1957) to be
abundant throughout East African waters and at all the islands beyond to the
Seychelles.

There is a further specimen, 38 mm. in S.L. which I include with this species,
although it possesses 13 dorsal spines. In all other respects it agrees with the nine
other specimens identified as S. nuchalis and was taken with them.

Dendrochirus zebra (Quoy & Gaimard)

Pterois zebra Quoy & Gaimard, 1824, Voy. de I'Uranie et de la Physicienne, Zool. : 329.
B.M. No. 1969.8.26.280-282.

3 specimens, 65-112 mm. in S.L. from bomb-crater tide pool at Betio, Tarawa.
D. XIII 9; A. Ill 5. Scale rows above lateral line 48-50. Middorsal spines longer
than depth of body.

* Pterois antennata (Bloch)

Scorpaena antennata Bloch, 1787, Naturg. ausldnd. FischeS: 21.
B.M. No. 1969.8.26.283-284.

2 specimens, 28 and 70 mm. in S.L. from the S.W. outer reef, Betio, Tarawa.
D. XII ii ; A. Ill 6; pect. 17. Scales above lateral line about 50.

This species is widely distributed throughout the Western Pacific and the Indo-
Australian region, although according to Smith (1957), it is nowhere abundant. It
has not previously been recorded from the Gilbert Islands.

There are two juveniles, 14 and 15 mm. in S.L., which I am provisionally identify-
ing as this species.

SYNANCEJIDAE

Synanceja verrucosa Schneider

Synanceja verrucosa Schneider, 1801, Syst. Ichth. Bloch. : 195.
B.M. No. 1969.8.26.285-287.

3 specimens, 104-165 mm. in S.L., taken from the bomb-crater tide pool at Betio,
Tarawa.

D. XIII 7; A. Ill 6; pect. 18.

Live colour greyish green, the pectorals edged with orange.

Widely distributed throughout the Indian Ocean from the east coast of Africa
across to India and Ceylon, through Indonesia, the South China Sea and the Philip-
pines to the western South Pacific, and the Indo- Australian region.

232 G. PALMER

Not previously recorded from the Gilbert Islands, although Randall (1955) stated
that it was clear that this species was present in the islands. It may be distinguished
by the presence of a deep pit behind each eye, the latter being separated by an
elevated bony area. There is also a small, deep groove below each eye.

CARACANTHIDAE

Caracanthus maculatus (Gray)

Micropus maculatus Gray, 1831, Zool, Misc. : 20.
B.M. No. 1969.8.26.289.

1 specimen, 42 mm. in S.L., taken on the west reef at Betio, Tarawa, between
branches of coral.

D. VII 12 ; A. II ii ; pect. 14. Live colour grey, with a pink cast. Body covered
with black spots.

BOTHIDAE
Bothus mancus (Broussonet)

Pleuronectes mancus Broussonet, 1782, Ichthyologia, fig. and text (no pagination.)
B.M. No. 1969.8.26.299-302.

4 specimens, 92-167 mm. in S.L. from Tarawa lagoon.

D. 99; A. 78.

The colouration of this species is somewhat similar to B. pantherinus, but the two
may be distinguished, as stated by Randall (1955), by the difference in the squama-
tion of the interorbital. In B. mancus the anterior part of this area is naked,
whereas in B. pantherinus it is scaled.

Bothus pantherinus (Riippell)

Rhombus pantherinus Riippell, 1828, A tlas Reise nord. Afrika. Fische des rothen Meeres: 121.
B.M. No. 1969.8.26.303-305.

3 specimens, 71-145 mm. in S.L., from Tarawa lagoon.
D. 90; A. 71.

BALISTIDAE

* Pseudobalistes flavimarginatus (Riippell)

Balistesflavimarginatus Ruppell, 1828, Atlas Reise nord. Afrika. Fische des rothen Meeres: 33.
B.M. No. 1969.8.26.306-307.

2 specimens, 59 and 390 mm. in S.L., taken in Tarawa lagoon, near the reef.
D. Ill 2 24; A. 24; pect. 15. Lateral line ca. 35 pored scales.

Live colour of the smaller specimen was orange yellow on the body, with many
brown spots; the larger specimen was dark greenish brown on the body, the head

TELEOST FISHES FROM THE GILBERT ISLANDS 233

being a pale orange except above the eye, where it was brown. Fins darkish basally,
with two orange lines separated by a dark line on the distal edge.

Not previously recorded from the Gilbert Islands, although apparently not un-
common in Tarawa lagoon, where it is caught both in nets and on a hook.

ANTENNARIIDAE

Antennarius altipinnis Smith & Radcliffe

Antennarius altipinnis Smith & Radcliffe, 1912, in Radcliffe, Proc. U.S. Nat. Mus. 42: 204.
B.M. No. 1969.8.26.308-310.

3 specimens, 28-28 mm. in S.L. Taken from the S.W. outer reef at Betio, Tarawa.

The largest of these three specimens is abnormal in possessing only 9 dorsal rays
instead of the usual 12. In all other respects it agrees with the description of A.
altipinnis.

ACKNOWLEDGEMENTS

My thanks are due to the late Professor J. L. B. Smith for the loan of specimens of
Quisquilius eugenius and Q. inhaca; to Dr. R. J. Lavenbergand the authorities of the
Los Angeles Co. Museum for lending me a paratype of Phaenomonas pinnata; to Dr.
P. H. J. Castle, for information on the Ophichthidae, and to my colleagues in the Fish
Section for help and criticism during the preparation of this paper.

REFERENCES

CHAN, W. L. 1968. Marine fishes of Hong Kong. Gov. Press, Hong Kong : i-xxii ; 1-129.
GOSLINE, W.A. 1951. The osteology and classification of the Ophichthid eels of the Hawaiian

Islands. Pacific Sci. 5 : 298-320.
& Strasburg, D. W. 1956. The Hawaiian fishes of the family Moringuidae : another eel

problem. Copeia : 9-18.

GUNTHER, A. 1876. Die Fische der Siidsee. /. Mus. Godeffroy 11 : 134.
KANAZAWA, R. H. 1958. A revision of the eels of the genus Conger with descriptions of four

new species. Proc. U.S. nat. Mus. 108 : 219-267.
McCANN, C. 1967. A new species of eel of the genus Moringua (Pisces) from Manihiki Atoll,

Northern Cook Islands. Trans. R. Soc. N.Z. 8 : 211-213 6 figs.
MYERS, G. S. & WADE, C. B. 1941. Four new genera and ten new species of eels from the

Pacific coast of tropical America. Allan Hancock Pac. Exped. 1932-1940 9 Pt 4 : 65-111.
RANDALL, J. E. 1955. Fishes of the Gilbert Islands. Atoll Res. Bull. No. 47 : 1-243.
1964. Note on the groupers of Tahiti, with description of a new serranid fish genus.

Pacific Sci. 18 : 281-296.
& BROCK, V. 1960. Observations on the ecology of Epinepheline and Lutjanid fishes of

the Society Islands, with emphasis on food habits. Trans. Amer. Fish. Soc. 89 : 9-16.
SCHULTZ, L. P. 1943. Fishes of the Phoenix and Samoan Islands. Bull. U.S. nat. Mus. 180 :

22.

I957- The frogfhhes of the family Antennariidae. Proc. U.S. nat. Mus. 107 : 47-105.

et al. 1953 & 1960. Fishes of the Marshall and Marianas Islands. Bull. U.S. nat. Mus.

No. 202 Pt i : 1-685 ; Pt 2 : 1-438.
SMITH, J. L. B. 1957. The fishes of the family Scorpaenidae in the western Indian Ocean

Pts. i & 2. Ichthyol. Bull. Rhodes Univ. No. 4 : 49-69 and No. 5 : 75-87.

234 G. PALMER

SMITH, J. L. B. i959(a). Gobioid fishes of the western Indian Ocean. Ichthyol. Bull. Rhodes

Univ. No. 13: 185-225.
1959(0). Fishes of the families Blenniidae and Salariidae of the western Indian Ocean.

Ichthyol. Bull. Rhodes Univ. No. 14 : 229-252.
WEBER, M. & DE BEAUFORT, L. F. 1916-1962. Fishes of the Indo-Australina Archipelago.

3-11 E. J. Brill, Leiden.
WHITEHEAD, P. J. P. (In press). The Reeves Collection of Chinese fish drawings. Bull. Brit.

Mus. nat. Hist. (Zool.) 3 No. 7 : 212, pi. 3, fig. a.
WILLIAMS, F. 1958. Fishes of the family Carangidae in British East African waters. Ann.

Mag. nat. Hist. (13) 1 : 369-430.

GEOFFREY PALMER

Department of Zoology

BRITISH MUSEUM (NATURAL HISTORY)

CROMWELL ROAD

LONDON, S.W.7.

Printed in Great Britain by

Alden & Mowbray Ltd
at the Alden Press, Oxford

A REVIEW OF THE SOUTH

AMERICAN IGUANID LIZARD

GENUS PLICA

R. ETHERIDGE

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 7

LONDON: 1970

A REVIEW OF THE SOUTH AMERICAN
IGUANID LIZARD GENUS PLICA

BY

RICHARD ETHERIDGE

Pp. 235-256; 4 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

ZOOLOGY Vol. 19 No. 7

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they
become ready. Volumes will contain about three
or four hundred pages, and will not necessarily be
completed within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 7 of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 21 July, 1970 Price Fifteen Shillings

(75P-)

A REVIEW OF THE SOUTH AMERICAN
I IGUANID LIZARD GENUS PLICA

By RICHARD ETHERIDGE

SYNOPSIS

The nomenclatorial history, taxonomy, distribution, and external morphology of the iguanid
lizard genus Plica are reviewed, and the genus is compared with related iguanid genera. Plica
plica and Plica umbra are recognized as the only valid species. Geographic variation in Plica
umbra is discussed, and two races are recognized : the nominate form in Venezuela and the
Guianas, and P. u. ochrocollaris Spix in the Amazonian basin of northern Bolivia, Peru, Ecuador,
Colombia, and northern Brazil. Use of the generic name Uranoscodon Boie is discussed, and
Lacerta superciliosa is designated as its type species.

INTRODUCTION

Plica plica and Plica umbra are among the most abundant, widespread, and earliest
known lizards of South America. Their range includes most of northern South
America east of the Andes as far to the south as northern Bolivia, and save for the
northern parts of Venezuela and Trinidad, from which P. umbra is absent, they are
often to be found in the same locale. Both species are conspicuous, arboreal in-
sectivores, and they are exceedingly similar in choice of habitat, diet, movements, and
general reactions.

In physical appearance the two species are quite different, and they are easy to
distinguish in the field. Plica plica has small dorsal and ventral scales, tufts of
spines on the sides of the neck, and a colour pattern of dark brown bands or marbling
on a background of light to olive green, a white chin with a black throat and a
narrow, black collar. Plica umbra has large dorsal and ventral scales, no spiny scales
on the neck, and a colour pattern that is highly variable as well as metachromatic,
consisting basically of wide bands or blotches of red, brown, or yellow on a background
in some shade of green. Another curious distinction is the colour of the lining of the
mouth : white in plica, and deep violet in umbra.

The nomenclatorial history of Plica and related genera has been long and extremely
confused. Two other species are now incorrectly recognized : P. tuberculatum
Andersson (1918 : 2), and P. stejnegeri Burt & Burt (1930 : 24), and two currently
recognized species of Tropidurus, T. holotropis Boulenger (1912 : 420), and T.
unicarinatus Werner (1899 : 480) are actually synonyms of Plica umbra. The genus
Plica itself has never been adequately characterized or diagnosed, and in mots
European museums it is still catalogued under the name Uraniscodon Kaup.

In this study the history of the nomenclature of Plica is presented, the genus is
defined and diagnosed, its species are redescribed and accompanied by synonomies,
and geographic variation in Plica umbra is discussed.

238 R. ETHERIDGE

ACKNOWLEDGEMENTS

It is a pleasure to acknowledge the kindness and cooperation of many museum
curators at whose institutions I have had the good fortune to study : Dr. C. M.
Bogert, American Museum of Natural History, New York ; Dr. E. E. Williams,
Museum of Comparative Zoology, Harvard University ; Dr. R. F. Inger, Field
Museum of Natural History, Chicago ; Dr. A. Leviton, California Academy of
Sciences, San Francisco ; Dr. J. Wright, Los Angeles County Museum ; Miss A. G. C.
Grandison, British Museum (Natural History), London ; Monsieur J. Guibe", Museum
National d'Histoire Naturelle, Paris ; Dr. K. Klemmer, Natur-Museum und For-
schungs-Institut Senckenberg, Frankfurt a.M. ; Dr. J. Eiselt, Naturhistorischen
Museum, Wien ; Dr. C. Edelstam, Naturhistoriska Riksmuseet, Stockholm ; Dr. H.
Wermuth, Staatliches Museum fur Naturkunde, Ludwigsburg ; Dr J. A. Peters,
United States National Museum, Washington ; Dr. G. Peters, Zoologisches Museum
der Humboldt Universitat, Berlin ; Dr. W. Ladiges, Zoologisches Staatsinstitut
und Zoologisches Museum, Hamburg.

This study was financed, in part, by a grant from the Systematics Section of the
National Science Foundation of the United States.

I am especially grateful to Miss A. G. C. Grandison and her staff at the British
Museum (Natural History) for their generous contributions of time, facilities, and
good counsel during the course of this study.

The following abbreviations are used :

A.M.N.H. American Museum of Natural History, New York.

B.M.N.H. British Museum (Natural History), London.

C.A.S. California Academy of Sciences, San Francisco.

F.M.N.H. Field Museum of Natural History, Chicago.

L.A.C.M. Los Angeles County Museum, Los Angeles.

M.C.Z. Museum of Comparative Zoology, Harvard University, Cambridge.

M.H.N.P. Museum National d'Histoire Naturelle, Paris.

N.M.W. Naturhistorischen Museum, Wien.

N.R.M.S. Naturhistoriska Riksmuseet, Stockholm.

S.M.F. Natur-Museum und Forschungs-Institut Senckenberg, Frank-
furt a.M.

S.M.N.L. Staatliches Museum fur Naturkunde, Ludwigsburg.

S.N.M. Stanford Natural History Museum, Stanford.

U.S.N.M. United States National Museum, Washington.

Z.M.B. Zoologisches Museum der Humboldt Universitat, Berlin.

Z.M.H. Zoologisches Staatsinstitut und Zoologisches Museum, Hamburg.

HISTORICAL REVIEW

Linnaeus (1758) described the two valid species currently referred to the genus
Plica as Lacerta Umbra (pg. 207) from " Meridionalibus, " and Lacerta Plica (pg.
208) from " Indiis. " Although Linnaeus made no reference to it a figure of plica
had been included by Seba (1734 : 11.65.5), a figure which later formed the basis for
Laurenti's (1768 : 48) Iguana chalcidica from " Gallaccia. "

Latreille placed umbra in the genus Iguana (i8o2a : 263), and plica in the genus

IGUANID LIZARD PLICA 239

Stellio (i8o2b : 27). Later that same year Daudin (1802) employed the combina-
tions Agama umbra and Agama plica. However, his characterization of Agama
umbra, based upon specimens from Cayenne and Surinam, is clearly that of Linnaeus'
Lacerta plica, while his reference to Agama plica is based solely on the authority of Lin-
naeus. Merrem (1820 : 54) , and Wied (1825 : 12 5) a l so included these species in Agama.

Spix (1825 : I0 ) described Lophyrus ochrocollaris from " in sylvis fluminis Ama-
zonum, " and on the following page described Lophyrus Panther a from " sylvis ad
pagum Ecga. " The types of both forms, formerly in the Zoologische Sammlung des
Bayerischen Staates, Miinchen, were destroyed during World War II. However, it
is clear from the descriptions and coloured figures of Spix that ochrocollaris was
based upon a sombre coloured adult of Plica umbra, and panther a was based upon a
brightly coloured juvenile of Plica, plica.

At this point the nomenclatorial history of Plica becomes confused with that of two
unrelated genera, Enyalius and Uranoscodon. Kaup (1825 : 590) described the genus
Uranoscodon, based upon Linnaeus' (1758 : 200) Lacerta superciliosa, although he
included as well both plica and umbra. Boie (1825 : 1090), in commenting upon
Kaup's work, indicated that Agama catenata (Wied, 1821 : 247), and Agama picta
(Wied, 1925 : 125) should be included in the same genus with superciliosa, for which
he suggested that either his own name Ophryessa, or Uraniscodon (emended from
Kaup's Uranoscodon) might be used. Boie specifically excluded plica and umbra
from this genus. In the following year Kaup (1826 : 89) adopting Boie's spelling of
Uraniscodon, gave a more detailed description of the genus, to which he referred eight
species. In addition to superciliosa, plica, and umbra he included Lophyrus rhombifer
(Spix, 1825 : 9)> Lophyrus margaritaceus (Spix, 1825 : I0 ) Agama picta, Agama
catenata, and Agama hispida (Spix, 1825 : I2 )- L- rhombifer and L. margaritaceus
are synonyms of Agama catenata, and both catenata and picta are currently referred to
as Enyalius (Etheridge, 1969). Agama hispida Spix is now regarded as a species of
Tropidurus. Later that year Boie (1826 : 119), in dealing with most of the same
species, placed in the genus Ophryessa the species superciliosa, catenata, margaritaceus,
ochrocollaris, panthera, and auronitens. As pointed out above, margaritaceus is a
synonym of catenata, ochrocollaris a synonym of umbra, and panthera a synonym of
plica ; auronitens, as recognized by Kaup (1826 : 90) is a synonym of superciliosa.

Most of the same species were referred to by Fitzinger (1826 : 48), who recognized
Boie's Ophryessa, in which he included superciliosa, catenata, margaritaceus, and
umbra. In the genus Ecphymotes Fitzinger placed plica, undulatus, pictus, and
acutirostris. Kaup (1827 : 612) then proposed the subdivision of Uraniscodon into
three subgenera : Uraniscodon, containing Agama hispida ; Pneustes, containing
picta, umbra, and plica ; and Ophryessa, containing superciliosa, catenata, and
margaritaceus.

Wagler (1830 : 150) described the genus Hypsibatus, including in it the species
plica, umbra, and picta. Wiegmann (1835 : 289) replaced Hypsibatus with Hyp-
silophus, pointing out that the former was preoccupied by Hypsibates Nitzsch
(1827 : I 5) a genus of birds. 1

1 Under Article 56 (a) of the current edition of the International Code of Zoological Nomenclature
Hypsibatus Wagler and Hypsibates Nitzsch would not now be considered homonyms, but under Article
23 (b) Hypsibatus Wagler is to be considered a nomen oblitum.

240 R. ETHERIDGE

Gray (1827 : 208) described Lophyrus agamoides, which he later (Gray, 1831 : 41)
recognized as a synonym of Plica plica. Gray (1831 : 40) described Plica as a sub-
genus of Ophyessa (an emendation of Ophryessa Boie), including in the subgenus
brasiliensis , picta, and plica. Brasiliensis is unrecognizable from its description, but
probably is a synonym of Enyalius catenatus (Etheridge, 1969 : 244). The charac-
terization of Ophyessa (Plica] picta is clearly that of Plica umbra, although Gray lists
as its synonyms Agama picta Wied, Lophyrus ochrocollaris Spix, and L. panther a Spix.
The characterization of plica is that of Plica plica, and listed as its synonyms were
Lacerta plica Linnaeus, Agama plica and Agama umbra of Daudin, and Lophyrus
agamoides Gray.

The confusion between Wied's Agama picta (= Enyalius pictus fide Etheridge,
1969) and Plica umbra was continued by Schinz (1833 : pi. 25 ; 1835 : 86) who figured
Wied's species of Enyalius under the name Calotes pictus, and referred to its synonymy
Lophyrus ochrocollaris and L. panthera of Spix.

Dumeril & Bibron (1827) used Wagler's name Hypsibatus for two species, H.
agamoides, and a new form, H. punctatus, both of which are clearly synonyms of
Plica plica. In the same work Dumeril & Bibron (1837 : 2 47) proposed the new genus
Uperanodon, based on Plica of Gray in part, and Hypsibatus of Wagler in part.
Included in Uperanodon were ochrocollorae, with Lacerta umbra listed as its synonym,
and pictum, with Lophyrus panthera as a synonym.

Fitzinger (1843 : 58) included umbra and pictus in the genus Hypsibatus, with
umbra designated as type of the subgenus Hypsibatus, and pictus the type of the sub-
genus Uperanodon. On the following page he included punctatus (of Dum6ril &
Bibron) and plica in the genus Ptychosaurus , designating punctatus as the type of the
subgenus Ptychosaurus, and plica as the type of the subgenus Ptychopleura.

Gray (1845 : 223) referred two species to Uraniscodon : umbra, based on a speci-
men from Pernambuco that is clearly Plica umbra, and pictum, based on Wied's
description of Agama picta. He then refers to Plica umbra, based on a specimen
of uncertain locality and referable to Plica plica, and Plica punctata (p. 224)
based on the description of Hypsibatus punctatus.

Cope (1876 : 170) described Hyperanodon peltigerus from Peru, another synonym
of Plica umbra.

Boulenger (1885 : 179-180) did a remarkably good job of untangling the confused
history of the genus, correctly recognizing the existence of only two valid species, and
erring only in the consideration of Agama picta as a synonym of umbra, and in using
the generic name Uraniscodon rather than Plica.

Stejneger (1901 : 182) pointed out that Kaup's original description of Urano-
scodon was based upon Lacerta superciliosa, although plica and umbra were included
by Kaup in the genus, and that Boie (1825 : 1090) specifically excluded plica and
umbra from the genus. Although this may not be considered a formal designation of
the type species of Uranoscodon, as Stejneger believed, to do so now would certainly
seem to be the desirable course. I therefore designate Lacerta superciliosa Linnaeus
(1758 : 200) as the type species of the genus Uranoscodon Kaup (1825 : 590). Note
that although Fitzinger (1843 : 16) designated superciliosa as type of the subgenus

IGUANID LIZARD PLICA 241

Ophryoessa (emended from Ophryessa), Boie did not publish the description of
Ophryessa until several months after the publication of Kaup's Uranoscodon.

Following the appearance of Boulenger's catalogue, three additional names have
been proposed for Plica umbra : Tropidurus unicarinatus from Surinam (Werner,
1899 : 480), Tropidurus holotropis from Peru (Boulenger, 1912 : 420), and Plica
tuber culatum from Bolivia (Andersson, 1918 : 2) . An additional new species, described
as Plica stejnegeri (Burt & Burt, 1930 : 24), is a synonym of Tropidurus spinulosus.

A review of the recent literature, and of the catalogues of numerous museums in
the United States and Europe indicates that at the present writing Tropidurus
holotropis, T. unicarinatus, Plica tuberculatum, and Plica stejnegeri are all considered
valid.

PLICA Gray, 1831

In Cuvier edit Griffith, The Animal Kingdom, 9 : 40. Type species : Lacerta
plica Linnaeus 1758.

Hypsibatus Wagler, 1830 (preoccupied by Hypisibates Nitzsch), p. 150. Type species : Lacerta

umbra Linnaeus.

Hypsilophus Wiegmann, 1835 (replacement name for Hypsibatus Wagler), p. 289.
Uperanodon Dumeril & Bibron, 1837, p. 247. Type species : Lophyrus ochrocollaris Spix.
Pytchopleura Fitzinger, 1843, p. 59. Type species : Lacerta plica Linnaeus.
Ptychosaurus Fitzinger, 1843, p. 59. Type species : Hypsibatus punctatus Dumeril & Bibron.
Hyperanodon Agassiz, 1847 (replacement name for Uperanodon Dumeril & Bribon), p. 190.

DEFINITION AND DIAGNOSIS. No adequate characterization or diagnosis of the
genus Plica has been published. Boulenger (1885 : 179), using the name Urano-
scodon, listed a number of characteristics that the species umbra and plica have in
common, by which, supposedly, they may be distinguished from other iguanids.
Most of the characters listed by Boulenger are shared with one or more other iguanid
genera. In his key to the genera of iguanid lizards Boulenger (1885 : 4) employed
the characteristics of " digits strongly bent at the articulations ; anterior maxillary
teeth longest " to distinguish Uranoscodon (that is Plica in current usage) from
Tropidurus : " digits straight ; no canine-like teeth. " The digits are bent at their
articulations to at least some degree in Tropidurus and Uracentron, and in some
species of Tropidurus, e.g. spinulosus, pictus, and bogerti, the digits are as strongly
bent as they are in Plica. The enlargement of anterior maxillary teeth is onto-
genetic. Even though they are relatively longer in adult Plica than in Tropidurus,
the condition is approached by some species of Tropidurus. Although enlarged
anterior maxillary teeth and bent digits will not distinguish Plica from other iguanid
genera, several other characteristics will.

Plica belongs to a group of South American-West Indian iguanids that is dis-
tinguished by the presence of a large sternal fontanelle and the absence of femoral
pores, the " tropidurines " (Etheridge, 1964 : 629). Within the tropidurines Plica
is a member of that section in which the interparietal scale is very much larger than
the adjacent scales, a group which also includes Tropidurus, Uracentron, Strobilurus,
and Uranoscodon. The characteristics by which Plica may be distinguished from
Tropidurus are as follows :

242

R. ETHERIDGE

Plica

1. Orbit bordered below by 5 to 7 sub-
equal suboculars ; no enlarged pre-
ocular.

2. Loreal scales small and numerous ;
several rows of loreolabials.

3. No enlarged postmentals or lateral
gulars.

4. Nostral in the approximate centre of
nasal scale, directed anterolaterally.

5. Superciliaries short and broad, form-
ing a sharp crest.

Tropidurus

1. Orbit bordered below by a single,
wide, elongate subocular that extend
to the anterior corner of the orbit,
or is preceded by a large preocular.

2. Loreal scales large and few ; i or 2
rows of loreolabials.

3. One or two pairs of enlarged post-
mentals, and usually a row of en-
larged lateral gulars.

4. Nostral in the posterior part of nasal
scale, directed posterolaterally.

5. Superciliaries long and narrow, not
forming a sharp crest.

The characteristics by which Plica may be distinguished from Uranoscodon are as
follows :

Plica

Upper head scales of various sizes ;
supraoculars enlarged ; supraorbital
semicircles distinct ; scales flat or
convex, or a few conical.
Interparietal scale much larger than
distance between orbits.
Superciliary crest moderately pro-
jecting, the last superciliary not en-
larged.

Tail long and slender, rounded, with-
out a serrate middorsal crest.

Uranoscodon

Upper head scales small, subequal ;
supraoculars small ; supraorbital
semicircles not distinct ; scales irreg-
ularly pyramidal.

Interparietal scale about as wide as
distance between orbits.
Superciliary crest strongly project-
ing, the last superciliary conspic-
uously enlarged.

Tail long and thick, distinctly com-
pressed, with a serrate middorsal
crest.

Plica plica (Linnaeus)

Lacerta Plica Linnaeus, 1758, p. 208 (type locality, " Indiis "; here restricted to the vicinity of

Paramaribo, Surinam).

Iguana chalcidica Laurenti, 1768, p. 48 (type locality, " Gallaccia ").
Agama umbra Daudin, 1802, p. 375.
Agama plica Daudin, 1802, p. 412.
Stellio plica Latrielle, 18020, p. 27.

Lophyrus Panthera Spix, 1825, p. n (type locality, " sylvis ad pagum Ecga, " Brazil).
Ecphymotes Plica Fitzinger, 1826, p. 49.
Uraniscodon plica Kaup, 1826, c. 91.
Uraniscodon (Pneustes) plica Kaup, 1827, c. 610.
Lophyrus (Ophryesa) Agamoides Gray, 1827, p. 208.
Hypsibatus Plica Wagler, 1830, p. 150.
Ophyessa (Plica) Plica Gray, 1831, p. 40.
Hypsibatus agamoides Wiegmann, 1834, P- I 5-

IGUANID LIZARD PLICA 243

Hypsibatus punctatus Dumeril & Bibron, 1837, p. 258 (type locality, none given).

Hypsilophus plica Wiegmann, 1835, p. 289.

Ptychosaurus (Ptychosaurus) punctatus Fitzinger, 1843, p. 59.

Ptychosaurus (Ptychopleura) Plica Fitzinger, 1843, p. 59.

Plica umbra Gray, 1845, p. 223.

Plica punctata Gray, 1845, p. 224.

Hypsibatus punctatus Peters, 1877, p. 408.

Uraniscodon plica Boulenger, 1885, p. 180.

Plica plica Burt and Burt, 1930, p. 19.

HOLOTYPE. N.R.M. No. KaF 1900 : 112. Linnaeus obviously erred in listing
India as the origin of the holotype. Although the ranges of both species of Plica
in South America are vast, geographic variation in Plica umbra is sufficient to narrow
the origin of the syntypes of that species to the general region of the Guianas. I
have elsewhere (Etheridge, 1968 : 51) pointed out that many of the South American
species of plants and animals described by Linnaeus were brought to him from Suri-
nam by Carl Gustaf Dahlberg and Daniel Rolander, the latter a student of Linnaeus,
and that most of their collections were made in Paramaribo and the nearby country-
side. I, therefore, propose the restriction of the type locality of Plica plica to the
vicinity of Paramaribo, Surinam.

CHARACTERISTICS. Rostral band-like, about four times wider than high. Nasals
convex, above and anterior to canthal ridge, separated from rostral and upper labials
by two rows of small scales ; nasal opening directed dorsolaterally. Internasals,
frontonasals, and prefrontals small, imbricate, irregularly arranged. Supraorbital
semicircles moderately distinct anteriorly, scarcely distinct from adjacent parietal
scales posteriorly, medially separated by a row of small scales or with one or two
scales in contact medially between the orbits. Four to seven large, transversely
widened supraoculars, medially separated from supraorbital semicircles by two rows
of small scales, laterally separated from superciliaries by several rows of smaller
scales. Interparietal much larger than adjacent scales, a third to a fifth as wide as
head, separated from supraorbital semicircles by smaller scales, narrowing in front,
with a central " eye. " Scales of parietal and upper temporal region small, irregu-
larly arranged, imbricate, and sharply keeled. An elongate, sharply keeled canthal,
followed by four or five elongate, keeled superciliaries, in turn followed by two similar
scales overlapping in the opposite direction ; the canthal and superciliaries together
forming a projecting crest. About 15 imbricating scales in loreal region, not marked-
ly distinct from adjacent loreolabials and scales of preocular region. Orbit bordered
below by a well defined arc of large scales of more or less equal size, each with a sharp
keel near the upper margin ; no enlarged preocular. Loreolabials in several irregular
rows, separating nasals, loreals, and suboculars from upper labials. Temporals small,
rhomboidal, imbricate, obtusely keeled ; those bordering anterior margin of ear with
a laterally projecting spine. Four to six rectangular upper labials. Mental sub-
triangular. Five or six lower labials similar to upper labials. No enlarged post-
mental or lateral gulars. Gulars small, imbricate, smooth anteriorly, becoming
smaller still and obtusely keeled near the transverse gular fold ; a medial row of very
slightly enlarged scales along margin of dewlap.

A row of enlarged scales aligned middorsally extending from occiput to proximal

244 R - ETHERIDGE

third of tail, forming in adults a low serration on the neck and anterior back. Dorsal
nuchals conical, each with a stout central spine, not distinctly imbricate, becoming
smaller still on the sides of the neck ; three patches of markedly enlarged, spinose
scales on side of neck, one posteroventral to, and another posterodorsal to the ear
opening, the third posterior to the anterodorsal patch. Dorsal body scales keeled,
imbricate, the keels forming oblique lines that converge toward the midline posteri-
orly, each keel ending in a short, outward projecting spine ; dorsolateral body scales
becoming gradually smaller until they meet the dorsolateral body fold, the latter
provided with slightly larger and more distinctly spinose scales ; below the dorso-
lateral fold scales becoming gradually larger and less strongly keeled, grading into
smooth, rhomboidal, imbricate ventrals.

FIG. i. Dorsal view of the head of Plica plica (A.M.N.H. 61315).

Two transverse gular folds, the posterior one rising on each side as an antibrachiaJ
fold enclosing granular scales, arching above the forelimb insertion, and continuing
back as a dorsolateral body fold to end just above the hindlimb insertion. Sides o]
neck with strong, irregular folds.

Caudal scales anterior to first autotomy segment similar to body scales above and
below ; autotomy annuli distinct, with five scale rows in the first few segments, six
rows in segments of the midportion of the tail, and thereafter seven rows in each
segment.

IGUANID LIZARD PLICA 245

Upper limb scales mostly rhomboidal, keeled, imbricate, those of the proximal
limb segments also mucronate ; ventral limb scales mostly smooth, imbricate.
Supracarpals, supratarsals, and supradigitals keeled, imbricate ; infracarpals and
infratarsals each with a median keel ending in a stout spine. Infradigital lamellae
each with a sharp, median keel ending in a stout spine, flanked on each side by a
much weaker keel and spine. Proximal lamellae of first and second fingers, and
first, second and third toes somewhat enlarged dorsoventrally.

COLOUR. The pattern is basically one of cross bands or bold, irregular blotches
of black, dark brown, or grey on a background of some shade of green. The anterior
cross band usually forms a distinct, very dark collar. Upper surfaces of the limbs
are similarly patterned. Small individuals are whitish below ; adult males have a
black throat narrowly bordered with yellow, shading into orange that covers the
rest of the ventral surfaces of the body, limbs and tail. In life the epithelial lining
of the mouth is white. A detailed description of the colour in life of an individual
from Kartabo, British Guiana may be found in Beebe (1944 : 204-205). The pattern
usually remains distinct even after many years in preservative, with only the back-
ground fading to bluish or olive grey.

GEOGRAPHIC VARIATION. Unfortunately the poor state of preservation of most
of the specimens examined does not permit a useful analysis of geographic variation
in Plica plica. The most deviant sample I have seen is from Puerto Ayachuco on
the Orinoco River, between Venezuela and Colombia. Specimens from this locality
differ from others in having a more acuminate snout, larger dorsal snout scales,
slightly smaller dorsal body scales, less well developed lateral neck spines, and a less
densely spotted throat in males. Specimens from the vicinity of Esmeralda,
Venezuela also have large dorsal snout scales but are otherwise typical. Specimens
from northern Venezuela, Trinidad, and the Guianas appear to reach the greatest
maximum size, and because of greater development of scale mucrons have overall a
more spiny appearance.

RANGE. (Text-fig. 2.) Trinidad and northern South America east of the Andes,
including northern Brazil, British Guiana, Surinam, Cayenne, northern Bolivia,
eastern Peru, eastern Ecuador, eastern Colombia, and Venezuela except in the
northwest.

MATERIAL EXAMINED. British Guiana : Kartabo A.M.N.H. 14117, 21331, 21290 ;
Kamakuska A.M.N.H. 25071, 25113-5, 25089 ; Kamaria Landing, Cuyuni River
A.M.N.H. 8090 ; Greets Creek, Essequibo River A.M.N.H. 8091 ; Kuyuwini Landing
A.M.N.H. 61316 ; 30 mi. from Bartica M.C.Z. 81189. Surinam : Pomeroon River,
Pickersgill M.C.Z. 32492 ; Orleane Kreek M.C.Z. 66887 ; Zanderij B.M.N.H.
1946.4.3.43 ; Paramaribo Z.M.H. 3548 ; no specific locality S.M.F. 11198, M.H.N.P.
2385. Trinidad : Tucker Valley A.M.N.H. 64461, 64529 ; Guayaguayase A.M.N.H.
70999 ; hts. of Guanape A.M.N.H. 72808 ; Caura Valley A.M.N.H. 72813-5 ;
Mayaro A.M.N.H. 72816 ; Port of Spain, St. Ann's Valley M.C.Z. 9001-3 ; Salibea

246

R. ETHERIDGE

Beach M.C.Z. 12064 J Nariva Swamp at Cascadoux Trace M.C.Z. 60825 ; near
Maracas Valley M.C.Z. 60826 ; no specific locality M.C.Z. 79829, 6102, S.M.F. 11200,
B.M.N.H. 66.6.16.11-12. Venezuela : 6 mi. N. Esmeralda, Mt. Duida region
A.M.N.H. 36661-62 ; Raudal de Atures, 10 km S. Puerto Ayachucho A.M.N.H.
81813-4 ; Serro de Cucuchy, La Varre A.M.N.H. 36530-1 ; Rio Pescada A.M.N.H.
36638 ; Mt. Duida A.M.N.H. 36640, M.C.Z. 101841 ; 10 mi. N. Esmeralda A.M.N.H.
36658-9 ; 6 mi. N. Esmeralda A.M.N.H. 36650 ; Estado Bolivar, Bajo Cuchivero,

FIG. 2. Map of northern South America showing localities for Plica plica.

Piedra Pintada C.A.S. 94658 ; Territorio Federal Amazonas, Alto Ventuari, Cacori
C.A.S. 94659 ; Paria Peninsula, Yacua M.C.Z. 43861-5 ; Puerto Ayacucho M.C.Z.
58335-6, U.S.N.M. 80629-31, F.M.N.H. 177924-36 ; nr. Rio San Juan, Los Mangos,
nr. Caripito M.C.Z. 81185 ; Chacaracual, Estado Sucre M.C.Z. 81186-8 ; Puerto La
Cruz U.S.N.M. 121184 ; Brazo Casiquiare U.S.N.M. 83612 ; Cerro Ya-pacana
U.S.N.M. 83943 ; Iguapo, Alto Orinoco S.M.F. 11201, M.C.Z. 101837, 101840.
Colombia : Lower Rio Guayabero, Cerro de las Pinturas F.M.N.H. 165207 ; Comisaria
Vaupes, Cano " Agua Bonita," 8 km S.W. San Jos< F.M.N.H. 75691. Peru :

IGUANID LIZARD PLICA 247

Utoquina region, Peru-Brazil border A.M.N.H. 56458 ; Monte Alegre, Rio Pachitea
system A.M.N.H. 56463 ; east of Contamana on trail to Contaya A.M.N.H. 56463 ;
Barranca, Rio Maranon Valley A.M.N.H. 56465 ; Rio Tapiche, trib. of Rio Ucayali
A.M.N.H. 56470 ; Rio Panaya mouth, middle of Rio Cushabatay Valley A.M.N.H.
56481, 56483 ; Pan de Azucar, 39 km NNE of Oxapampa, Rio Izcozazin L.A.C.M.
49139 ; Roaboya, Rio Ucayali Valley A.M.N.H. 56441-2, 56471 ; Colonia Calleria,
Rio Calleria, 15 km from Ucayali C.A.S. 93243-5 ; mouth of Rio Santiago M.V.Z.
16899 A.M.N.H. 56448, 56462 ; Pebas S.M.N.H. 15142, 12469, A.M.N.H. 8731-2,
B.M.N.H. 67.9.17.5 ; Boca de Amigo, Madre de Dios F.M.N.H. 81451-2 ; Rio
Samiria, Sta. Elena, Nanta F.M.N.H. 109821 ; Rio Morona, Quebrada Pushaga
F.M.N.H. 109826 ; Iquitos F.M.N.H. 45492-5, N.M.W. 13902, A.M.N.H. 56437 ;
Yarinacocha, Rio Ucayali F.M.N.H. 56011-20 ; Boca de Rio Inambari, Madre de
Dios F.M.N.H. 83305 ; Cayaria B.M.N.H. 81.5.13.23-4 ; lower Cushabatay
A.M.N.H. 56431, 56434 ; Rio Tamaya A.M.N.H. 56433, 56473 ; Pilluana, Rio
Mishqulyacu A.M.N.H. 56435 ; Rio Ayendama, trib. of Rio Cenipa, Maranon
system A.M.N.H. 56438, 56443 ; Rio Itaya, near Iquitos A.M.N.H. 56444, 56484 ;
Pampa Hermosa, near mouth of Rio Cushabatay A.M.N.H. 56445-7, 56451-4, 56456,
56460, 56466-9, 56472, 56474-7, 56480, 56488, 56485-6 ; mouth of Rio Napo, Lago
Mirano region A.M.N.H. 56449-50 ; Achinamisa, Rio Huallaga system A.M.N.H.
56436, 56457. Ecuador : Sarayacu Z.M.B. 9984 ; Rio Pastaza M.C.Z. 37268 ;
Santiago, between Rio Santiago and Rio Pastaza, S. of Macas and N. of Gualaquiza
F.M.N.H. 42502-4 ; Palatanga B.M.N.H. 80.12.8.55. Brazil : Rio Manjuru
A.M.N.H. 101938-40, 101942-3 ; Maue"s A.M.N.H. 101941, 101944 ; Cairari Uaupes,
Tatu, on Rio Negro A.M.N.H. 36626 ; Rio Madeira, Abuna C.A.S. 49782 ; Para, no
specific locality M.C.Z. 2890 ; mouth of Rio Canabury U.S.N.M. 83535 ; Sao Gabriel,
Rio Negro U.S.N.M. 83579-81 ; Salto do Hua, on Venezuelan border U.S.N.M.
83566 ; Serro do Cucuhy A.M.N.H. 36530-1 ; Tabatinga F.M.N.H. 83595 ; Cameta,
Rio Tocantins N.M.W. 13901 ; Boco do Rio Tucano, Roraime M.C.Z. 87289-90 ;
Reserva Ducke, Manaus M.C.Z. 92750 ; no specific locality S.M.F. 11199. No specific
locality ; M.H.N.P. 2387 (holotype Hypsibatus punctatus) .

Plica umbra (Linnaeus)

Lacerta Umbra Linnaeus, 1758, p. 207 (type locality, " Meridionalibus " ; here restricted to the

vicinity of Paramaribo, Surinam).
Iguana umbra Latreille, iSoaa, 263.
Agama umbra Merrem, 1820, p. 54.

Lophyrus ochrocollaris Spix, 1825, p. 10 (type locality, " sylvis fluminis Amazonum, " Brazil).
Ophryessa Umbra Fitzinger, 1826, p. 48.
Uraniscodon Umbra Kaup, 1826, c. 91.
Uraniscodon (Pneustes) umbra Kaup, 1827, c. 612.
Hypsibatus umbra Wagler, 1830, p. 150.
Uperanodon ochrocollare Dum6ril & Bibron, 1837, p. 247.
Hypsibatus (Hypsibatus) Umbra Fitzinger, 1843, p. 58.
Uraniscodon umbra Gray, 1845, p. 223.

Hyperanodon peltigerus Cope, 1876, p. 170 (type locality, Peru).
Hyperanodon umbra Peters, 1877, p. 408.

248 R. ETHERIDGE

Hyperanodon panthera Peters, 1877, p. 408.

Uraniscodon umbra Boulenger, 1885, p. 179.

Tropidurus unicarinatus Werner, 1899, p. 602 (type locality, Surinam).

Tropidurus holotropis Boulenger, 1912, p. 420 (type locality, Alpayaca, Rio Pastaza, Ecuador).

Plica tuberculatum Andersson, 1918, p. 2 (type locality, San Fermin, Bolivia).

Plica umbra Burt & Burt, 1930, p. 24.

SYNTYPES. N.R.M. No. 1900 : in (2 exs). The syntypes exhibit those charac-
teristics of scalation that are typical of individuals from the region of the Guianas
(see discussion of geographic variation below) . The reasons applied to the restriction
of the type locality of Plica plica are equally applicable to this species, and I therefore
propose the restriction of the type locality of Plica umbra to the vicinity of Para-
maribo, Surinam.

CHARACTERISTICS. Rostral strap-like or subtriangular. Nasals moderately large,
convex, above and anterior to canthal ridge, separated from rostral and anterior
upper labials by one or two small scales ; nasal opening directed dorsolaterally.
Scales of internasal, frontonasal, and prefrontal region moderately large, more or
less symmetrically arranged. Supraorbital semicircle scales large, usually three
pairs in broad contact medially between the orbits. A row of large, transversely
widened supraoculars, usually four in number, medially separated from supraorbital
semicircles by two rows of small scales, laterally separated from superciliaries by
two rows of smaller scales anteriorly ; posterior enlarged supraocular in broad contact
laterally with posterior superciliary. Interparietal scale large, a third to half as
wide as head, narrowing anteriorly, in broad contact with posterior scales of supra-
orbital semicircles, with a central " eye." Posterior parietals bordering interparietal
small to large, slightly to very strongly convex. A single, elongate canthal, followed
by four or five elongate, overlapping superciliaries, in turn followed by three super-
ciliaries overlapping in the opposite direction, the last much the longest ; canthal and
superciliaries together form a continuous, projecting crest. Scales of loreal region
small, juxtaposed, scarcely distinct from loreolabials or scales of preocular region.
Orbit bordered below by an ill denned arc of slightly enlarged, strongly keeled scales
of nearly equal size. Loreolabials in several irregular rows, separating loreals and
suboculars from upper labials. Temporals small, subimbricate, weakly and obtusely
keeled, those bordering the ear anteriorly scarcely projecting. Upper labials four
or five, more or less rectangular. Mental subtriangular. Lower labials five or six,
similar to upper labials. No enlarged postmentals or enlarged lateral gulars.
Gulars small, imbricate, becoming a little smaller, feebly keeled, and bluntly
mucronate posteriorly as they approach the transverse gular fold.

A row of scales aligned middorsally beginning at the occiput and extending to the
lumbar region, or past the sacral region and onto the tail ; middorsal scales of neck
and anterior body distinctly enlarged, forming a denticulate crest. Dorsal scales of
body and neck rhomboidal, imbricate, with a blunt posterior keel and a short, blunt
mucrone, the keels forming oblique rows that converge toward the midline posteriorly.
Lateral neck scales, except those within skin folds, smaller, but otherwise similar to
dorsal neck scales ; lateral body scales similar to dorsals, grading into distinctly

IGUANID LIZARD PLICA 249

keeled and mucronate ventrals. Dorsal, lateral and ventral body scales about equal
in size, save the anterior laterals which are a trifle smaller.

Strong antibrachial folds containing granular scales, terminating dorsally about
halfway between forelimb insertion and median crest, and joining a very strong
transverse gular fold ventrally. No lateral body folds and no other lateral neck
folds other than antibrachial and gular ; no patches of enlarged scales on sides of
neck.

Caudal scales anterior to first autotomy annulus similar to body scales both above
and below ; median crest disappears before first autotomy annulus. Autotomy
annuli distinct, four scale rows in the first few annuli, five in those of the midportion
of the tail, and thereafter six rows. Males lack enlarged postanal scales.

Forelimb scales equal to or a little smaller than body scales, imbricate and keeled,
except for proximal postbrachials which are smooth, the keels forming lines more or
less parallel to long axis of limb ; postantibrachials more sharply keeled and more
strongly mucronate than other forelimb scales. In males, a patch of smooth posterior
infrafemorals three or four scale rows wide and about 20 scale rows long ; all other
hindlimb scales (as well as all posterior infrafemorals in females) keeled, imbricate,
a little larger to a little smaller than body scales. Infracarpals and infratarsals each
with a single, stout mucrone. Supradigitals unicarinate ; subdigital lamellae with a
strong median keel and mucrone flanked on each side by a much weaker keel and
mucrone.

Scale counts are given in Table I. Maximum snout-vent lengths and body-tail
proportions of both sexes are given in Table II.

COLOUR. The colour is highly metachromatic, and the pattern exceedingly
variable. The pattern is basically one of wide, dark cross-bands or blotches on the
body, limbs, and tail, light coloured on the throat and belly, with or without streaks
of darker colour, and very dark colour within the antihumeral fold. The background
is usually some shade of green, ranging from yellowish-green, to bright green, bluish-
green, olive, or olive brown ; the bands and blotches are usually some shade of brown,
yellowish-brown, black, red, or orange. In adult males the centre of the throat may
be orange, and there is a patch of yellowish colour anterior to the vent and on the
lower surface of each femur. In life the epithelial lining of the mouth is deep violet.
Beebe (1944 : 207-8) has provided us with a most excellent and detailed description
of the colour and pattern of four living animals at Kartabo, British Guiana, and
Spix's (1825 : pi. 12, fig. 2) illustration of Lophyrus ochrocollaris is a good representa-
tion of an adult. In preservative specimens usually become dark purplish brown
with still darker markings, although the paler markings of the head and neck
frequently persist.

GEOGRAPHIC VARIATION. The very great majority of specimens may be readily
identified as one of two distinct types. One of them occurs in the Guianas, the other
in the Amazonian Basin of Brazil, Colombia, Ecuador, Peru, Bolivia, and in southern
Venezuela. Specimens from intermediate localities are unavailable so that the
question of whether two subspecies or two distinct species are involved cannot be

250

R. ETHER1DGE

resolved with certainty. Tentatively, however, the two forms may be considered
subspecies. The syntypes of Plica umbra are of the form that occurs in the Guianas.
The oldest name available for the Amazonian form appears to be Lophyrus ochro-
collaris Spix.

FIG. 3. Dorsal view of the head of Plica umbra ochrocollaris (L.A.C.M. 49140).

The two forms may be compared as follows :

ochrocollaris

i. Vertebral scale row prominent from
occiput to base of tail, forming a
distinct denticulation on the neck.

2. Dorsal head scales swollen, with
blunt, irregular surface.

3. A pair of large, distinctly pyramidal
scales flank the posterolateral corners
of the interparietal.

4. All scales of body and limbs more
sharply keeled and distinctly mucron-
ate.

5. Scales larger (Table I).

umbra

1. Vertebral scale row prominent on
neck and anterior body, becoming
indistinguishable from adjacent
scales on posterior back, and forming
only a slight denticulation on the
neck.

2. Dorsal head scales smooth or nearly
so.

3. Scales flanking posterolateral corners
of interparietal smaller, and flat or
nearly so.

4. Scales of body and limbs less sharply
keeled, less distinctly mucronate,
those of the belly only feebly keeled.

5. Scales smaller (Table I).

IGUANID LIZARD PLICA 251

Synonyms of Plica umbra ochrocollaris are Hyperanodon peltigerus Cope, Tropidurus
holotropis Boulenger, and Plica tuber culatum Andersson. A synonym of Plica umbra
umbra is Tropidurus unicarinatus Werner. These synonyms are further discussed in
the following section.

REMARKS. Boulenger (1885 : 179-180) adequately dealt with the synonomies of
names proposed for Plica umbra prior to the appearance of his Catalogue. Subse-
quently two species of Tropidurus have been proposed that are in fact also synonyms
of Plica umbra : Tropidurus unicarinatus (Werner, 1899 : 480), and Tropidurus
holotropis (Boulenger, 1912 : 420). In their type descriptions both species were
compared only with other forms of Tropidurus, with no mention of the genus Plica.
The reasons for their generic allocations were unspecified ; however, the authors
apparently based their decision to place these species in Tropidurus upon the presence
of a very large interparietal scale. In my studies at a large number of American and
European museums I was unable to find any specimens, other than the syntypes,
catalogued as Tropidurus unicarinatus. However, many Ecuadorian and Peruvian
specimens of Plica umbra are catalogued as Tropidurus holotropis, and that name has
appeared in such recent faunal lists as Mertens (1956 : 120), and Peters (1967 : 35).

Tropidurus unicarinatus was based upon a subadult female and a male (N.M.V.
18728-9) from Surinam. Both syntypes are in good condition, and are identical in
all details of scalation, colour pattern, and proportions with specimens of Plica umbra
of comparable size from northeastern South America. The number of scales in the
vertebral crest from occiput to the anterior margin of the thigh (53, 56), the number
of paravertebral scales in the same length (74, 78) and the number of scales around the
middle of the body (59, 59) are within the limits of variation for specimens from this
region (Table i).

TABLE i

Subspecies Sex N Vertebral Paravertebral Midbody

umbra M 19 49-(54'6)-6i yo-(78-o)-9o 49-(5yi)-7i

F 23 4 6-( 54 -7)-63 66-( 74 -8)-9i 48-(

ochrocollaris M 26 32-(39'8)-46 47~(58'5)-69 43-(49'6)-57
F 17 3 6-( 4 o-6)-47 4 8-( 57 -2)-67 4H5O'3)-55

Scale counts of the subspecies of Plica umbra. N = number of specimens
examined for these counts. Vertebral scales are counted along the midline
from occiput to a line even with the anterior margin of the thighs held at
right angles to the body axis ; paravertebral scales are counted for the same
distance ; midbody scales are counted around the body midway between the
limb insertions. Mean figures are in parentheses.

Tropidurus holotropis was based upon a subadult individual from Alpayaca, Rio
Pastaza, Ecuador (B.M.N.H. 1912.11.1.29 [RR 1946.8.29.64]). It is identical in all
details of colour pattern, scalation, and proportions with specimens of Plica umbra
of comparable size from eastern Ecuador. The number of scales in the vertebral
crest (35), the paravertebral scale row (55), and around the middle of the body (45)
are within the range of variation for specimens from this region (Table 2) .

252 R. ETHERIDGE

TABLE 2

Subspecies Sex N Snouth-vent Tail/Snouth-vent Length

Length
Range in mm.

umbra M 15 72-100 2'O2-(2.24)-2'44

F 14 6l- 94 2'O2-(2.l8)-2'42

ochrocollaris M 23 67- 90 2-oo-(2.i6)-2'5i

F 18 42- 90 i'9o-(2.o6)-2'33

Snout-vent lengths and tail/snout-vent length ratios for Plica umbra.
N = number of specimens examined. Mean figures are in parentheses.

Andersson (1918 : 2) described Uraniscodon tuberculatum from San Fermin,
northwestern Bolivia, and subsequently Burt and Burt (1930 : 19) referred to the
species as Plica tuberculatum. Andersson compared tuberculatum with umbra, from
which it was said to differ in the lepidosis of the head, especially the wider interpari-
etal scale, the absence of a feeble, longitudinal fold on the chin, a more strongly raised
supracilliary ridge, and a different colour, especially on the head. The holotype
(N.R.M. 3256) is a juvenile, 55 mm in snout-vent length, in fair condition except that
the skin of the lateral neck region is badly lacerated. Andersson apparently com-
pared the holotype of tuberculatum with adult specimens of Plica umbra from the
northeastern portion of its range, for the differences he noted are clearly due to
ontogenetic differences between juveniles and adults, and to geographic variation
within the species. The feeble longitudinal fold on the throat of adults is not
apparent in juveniles, and the vivid head pattern of juveniles becomes obscure in
adults. Although not specifically mentioned by Andersson as a diagnostic character,
the large interparietal scale, " much broader than half the breadth of the head," was
used by Burt and Burt (1930 : 19) to distinguish tuberculatum from umbra, "...
about half as wide as the head." The head width of the holotype is n-o mm, the
interparietal width 6-4 mm ; thus the interparietal is only slightly more than half
(58) as wide as the head, and within the range of variation of umbra. The strongly
projecting supracilliary ridge is characteristic of specimens of umbra from the western
portion of its range. In these, and in all other details of scalation, colour pattern,
and body proportions the holotype of Uraniscodon tuberculatum is typical of juvenile
specimens of Plica umbra from the western portion of its range.

RANGE. (Text-fig. 4.) Most of northern South America east of the Andes,
including British Guiana, Surinam, Cayenne, southern Venezuela, eastern Colombia,
eastern Ecuador, eastern Peru, extreme northern Bolivia, and northern Brazil. The
race P. u. umbra is confined to the Guianas ; the remainder of the range of the species
is occupied by P. u. ochrocollaris.

MATERIALS EXAMINED. Plica umbra umbra. British Guiana : Kartabo A.M.N.H.
21316, 14115-6 ; Kalacoon A.M.N.H. 8097 ; Kamakuska A.M.N.H. 25118 ;
Georgetown A.M.N.H. 17690-1 ; Demeara A.M.N.H. 18200, B.M.N.H. 72.10.16.66-7;

IGUANID LIZARD PLICA

253

Maccasseema B.M.N.H. 87.1.22.7 ; Moraballi Creek, Essequibo River B.M.N.H.
1930.10.10.67-80 ; Mazaruni, Penal settlement B.M.N.H. 1934.11.1.80 ; near
Bartica, upper Cuyuni River B.M.N.H. 1934.11.1.79 ; New River B.M.N.H.
1939.1.1.72-3 ; no specific locality A.M.N.H. 61276-8, 61291. Cayenne : La Mana
M.H.N.P. 2584 ; no specific locality M.H.N.P. 2581, 6828. Surinam : Paramaribo
Z.M.H. 2229 (2 exs), N.M.W. 13898 (4 exs), 13897 (2 exs), B.M.N.H. xxiii.no.6 ;
near Saramaca River, 91.5 kms, B.M.N.H. 1946.4.3.42 ; no specific locality N.M.W.
13899 (5 exs), 18754, 18728-9 (syntypes, Tropidurus unicarinatus) , S.M.F. 11195.

FIG. 4. Map of northern South America showing localities for Plica umbra umbra
(triangles), and Plica umbra ochrocollaris (circles).

No Specific Locality : N.R.M. 1900 (2 exs) (syntypes, Lacerta umbra). Additional
records from Surinam have been provided to me (in litt.} by Mr. Marinus S. Hoog-
moed : Wilhelminagebergte, Fallawatra-Coppename, Matta, Cultuurtuin, Coeroeniei-
land, Sipaliwini, Surinamerivier Powakka, Maguire, Moengo Tapoe, Wanekreek,

254 R - ETHERIDGE

Djaikreek, Calibi, Tibiti, Kaborkreek, Marowyne Nassaugebergte, Kaiserberg air-
strip, Bigiston dist. Marowyne, Boven Nicherie.

Plica umbra ochrocollaris. Ecuador : Montalbo, Rio Bobonaza A.M.N.H. 60607 '
San Jose de Sumaco A.M.N.H. 28877, 28893 ; Riobamba A.M.N.H. 14573, 23331,
M.C.Z. 29291 ; Sarayacu B.M.N.H. 80.12.8.60 ; Rio Pastaza, Alpayaca B.M.N.H.
1912.11.1.29 [RR 1946.8.29.64] (holotype, Tropidurus holotropis) ; Santiago, between
Macas and Gualaguiza F.M.N.H. 42511 ; Avila, Rio Napo. Prov. Orienta S.N.M.
8269 ; Rio Pastaza M.C.Z. 37269 ; Limon Cocha, Napo M.C.Z. 86685. Colombia :
Santa Rosa de los Cofanes, Rio Guamues, W. of Puerto Asis F.M.N.H. 165209-10 ;
Caqueta, Rio Mecaya F.M.N.H. 69671 ; Puerto Narino M.C.Z. 61151. Venezuela :
Esmeralda M.C.Z. 101748 ; La Culebra, upper Rio Cunucunuma M.C.Z. 101749.
Peru : Rio Pisqui, trib. of Rio Ucayali A.M.N.H. 57183 ; Rio Tamaya, trib. of Rio
Ucayali A.M.N.H. 57184 ; Rio Utoquinia, Tapoche region A.M.N.H. 57185 ; Rio
Itaya, Iquitos region A.M.N.H. 57186, 57188, 57192, 57197 ; lower Rio Cushabatay,
trib. of Rio Ucayali A.M.N.H. 57187 ; Rio Cenipa, trib. of Rio Maranon A.M.N.H.
57184, 57194, 57196 ; Rio Ayendama, trib. of Rio Cenipa A.M.N.H. 57190 ; upper
Rio Cushabatay A.M.N.H. 57191 ; Rian Rian, Rio Suhua, Contamana region
A.M.N.H. 57193 ; mouth of Rio Santiago A.M.N.H. 57195 ; Guache, Rio Pastaza, at
Ecuadorian border A.M.N.H. 60577 > Moyobamba B.M.N.H. 74.8.4.90-94 ; Iquitos
F.M.N.H. 45498 ; Yarinacocha, Rio Ucayali F.M.N.H. 56010 ; Cerro Azul, Dept.
Loreto F.M.N.H. 56008 ; Pedrera, Dept. Loreto F.M.N.H. 45497 ; Puerto Mayo
F.M.N.H. 3672 ; Pucalpa F.M.N.H. 56009 ; Colonia Calleria, Rio Calleria, 15 km
from Rio Ucayali C.A.S. 93241-2 ; Pebas, S.N.M. 8356, 15142, M.C.Z. 16360 ; Pan
de Azucar, 39 km NNE Oxapampa L.A.C.M. 49140-2. Bolivia : Guayaramerin
A.M.N.H. 91857 ; San Fermin N.R.M.S. 3256 (holotype, Uraniscodon tuberculatum).
Brazil : Manaos A.M.N.H. 64859 ; Manacapuni, Rio Solimoes A.M.N.H. 4484 (4 exs)
S.M.F. 11196-7 ; Cahabouca, Rio Solimoes, Parana do Jacare Z.M.B. 36945 ; Rio
Tanantins, upper Amazonas Z.M.B. 30975 (2 exs) ; Rio Capim, SE of Belem Z.S.B.S.
76/1929 ; Piexe-boi, on Braganca road, east of Belem Z.S.B.S. 151/1911 (4 exs)
Faro Yamunda Z.S.B.S. 91/1960 ; Col. de Vrado Obidon Z.S.B.S. 92/1960 ; Auawin-
dena Z.S.B.S. 93/1960 ; below Torarritius Z.S.B.S. 94/1960 ; Baiao, Rio Tocantins,
S.W. of Belem N.M.W. 13904 (2 exs) ; Cameta, Rio Tocantins N.M.W. 18179 (2 exs)
Belem U.S.N.M. 159184-93, M.C.Z. 86685 ; Salto do Hua, on Venezuelan border
U.S.N.M. 83567 ; Tabatinga F.M.N.H. 83596-7 ; Codajas, Amazonas B.M.N.H.
1965.1315 ; Rio Guama, Para B.M.N.H. 1926.5.5.6 ; no specific locality M.H.N.P.
2383 (2 exs), B.M.N.H. 1961.1083, 53.2.4.106.

Species Erroneously Referred to Plica

Burt and Burt (1930 : 19) described Plica stejnegeri, based upon a specimen
(U.S.N.M. 73505) from " Argentina." Their reasons for referring this species to
Plica were not specified ; however, they were probably led to this allocation because
the skin of the neck is strongly plicate and pouched, and on the sides provided with
many tufts of small, erect, spinose scales, a condition that closely resembles that of
Plica plica.

IGUANID LIZARD PLICA 255

I have compared the holotype of Plica stejnegeri, an adult male, with series of
Tropidurus spinulosus from Brazil, Bolivia, and Argentina, and find that it is identical
in all important details of scalation, colour pattern, and proportions. In this species
the scales of the middorsal crest become elongate, two to four times higher than long,
in large adults, and the spininess of the body scales, especially those of the lateral
neck region and the posterior tibials, becomes more pronounced. Plica stejnegeri
is thus clearly a synonym of Tropidurus spinulosus.

In a recent check list of the lizards of Venezuela Donoso-Barros (1968 : 114)
placed Tropidurus bogerti Roze (1958 : 247) in the genus Plica without comment.
Roze discussed in detail the superficial resemblance of bogerti to Plica plica and
correctly concluded that the species should be placed in Tropidurus. To this I can
only add that in every respect bogerti shows the characteristics of Tropidurus that I
have used to distinguish that genus from Plica.

REFERENCES

AGASSIZ, L. 1847. Nomenclatoris zoologici index universalis. Soloduri. vii+393 pp.
ANDERSSON, L. G. 1918. New lizards from South America. Ark. Zool., 2 16 : 1-9.
BEEBE, W. 1944. Field notes on the lizards of Kartabo, British Guiana and Curipito, Vene-
zuela. Zoologica, N.Y., 29 4 : 195-216.

BOIE, H. 1825. Zu Kaup's Aufsatz iiber Lurche. Isis, Jena, 17 : 1089-1090.
1826. Bemerkungen iiber die von Dr. von Spix abgebildeten brasilianischen Saurier.

Isis, Jena, 18 : 117-120.
BOULENGER, G. A. 1885. Catalogue of the lizards in the British Museum (Natural History),

vol. 2, ed. 2, London, xiii+497 pp.
1912. Descriptions of new reptiles from the Andes of South America preserved in the

British Museum. Ann. Mag. nat. Hist., (10) 1912 : 420-424.
BURT, C. E. & BURT, M. D. 1930. The South American lizards in the collection of the United

States National Museum. Proc. U.S. natn. Mus., 76 6 : 1-52.
COPE, E. D. 1876. Report on the reptiles brought by Professor James Orton from the middle

and upper Amazon, and western Peru. /. Acad. nat. Sci. Philad., (2) 8 : 159-193.
DAUDIN, F. M. 1802. Histoire naturelle gdndrale et paticuliere des reptiles, vol. 4, Paris. 385

pp.
DONOSO-BARROS, R. 1968. The lizards of Venezuela, (Check list and Key). Carrib. J. Sci.,

8 : 105-122.
DUMERIL, A. M. C. & BIBRON, G. 1837-1839. Erpe"tologie generale ou histoire naturelle complete

des reptiles, vol. 4, Paris. ii+57i pp.
ETHERIDGE, R. 1964. The skeletal morphology and systematic relationships of sceloporine

lizards. Copeia no. 4 : 610-631, 6 figs.
1968. A review of the iguanid lizard genera Uracentron and Strobilurus. Bull. Br. Mus.

nat. Hist. (Zool.), 17 2 : 47-64.
1969. A review of the iguanid lizard genus Enyalius. Bull. Br. Mus. nat. Hist. (Zool.),

18 8 : 231-260.
FITZINGER, L. 1826. Neue Classification der Reptilien nach ihren natiirlichen Verwandtschaften.

Wien. viii + 66 pp.

1843. Sy sterna Reptilium. Fasciculus Primus. Amblyglossae. Wien. vi+io6pp.

GRAVENHORST, J. L. 1838. Beitrage zur Genauren kenntnis einiger Eidechsengattungen.

Nova A eta Leopoldina, 18 2 : 713-783.
GRAY, J. E. 1827. A description of a new species of saurian reptiles ; with a revision of the

species of chamaeleons. Phil. Mag., 2 : 207-214.
1831. in CUVIER, edit. GRIFFITH, The Animal Kingdom. Reptilia. A synopsis of the

species of the class reptilia. London, no pp.

256 R. ETHERIDGE

GRAY, J. E. 1845. Catalogue of the specimens of lizards in the collection of the British Museum.

London. xxviii + aSg pp.
KAUP, J. 1825. Einige Bemerkungen zu Merrems Handbuch. Isis, Jena, 16 : 589-593.

1826. Behtrage zu Amphibiologie und Ichthinologie. Isis, Jena, 18 : 87-91.

1827. Zoologische Monographien. Isis, Jena, 19 : 610-625.

LATREILLE, P. A. i8o2a. in SONNINI & LATREILLE, Histoire naturelle des reptiles, vol. i,

Paris, xxii + 280 pp.

i8o2b. in SONNINI & LATREILLE, Histoire naturelle des reptiles, vol. 2, Paris. 332 pp.
LAURENTI, J. N. 1768. Specimen medicum, exhibens synopsin reptilium emendatum cum experi-

mentis circa venena et antidota reptilium Austriacorum. Wien. 214 pp., 5 pis.
LINNAEUS, C. 1758. Sy sterna naturae per regna tria naturae, secundum classes, ordines, genera,

species cum characteribus differentiis, synonymis, locis, ed. 10, vol. i, 11+824 PP-
MERREM, B. 1820. Versuch eines Systems der Amphibien. Tentamem systematis Amphi-

biorum. Marburg. xv+iSgpp.
MERTENS, R. 1956. Zur Kenntnis der Iguaniden-Gattung Tropidurus in Peru. Senckenberg.

biol., 37 1/2 : 101-136.

NITZSCH, C. L. 1827. in ERSCH & GRUBER. Allgem. Ency., 16 : 1-386.
PETERS, J. A. 1967. The lizards of Ecuador, a check list and key. Proc. U.S. natn. Mus.,

119 3545 : 1-49.
PETERS, W. 1877. Uber die von Spix in Brasilien gesammelten Eidechsen des Koniglichen

Naturalien-Kabinets zu Miinchen. Mber. K. Akad. Wiss. preuss., 407-423.
ROZE, J. A. 1958. Resultados zoologicos de la expedicion de la Universidad Central de

Venezuela a la region del Auyantepui en la Guayana Venezolana, Abril de 1956. 5. Los

reptiles del Auyantepui, Venezuela. Acta biol. venez., 2 22 : 243-270.
SPIX, J. B. 1825. Animalia nova sive species novae lacertarum, quas in itinere per Brasiliam

annis MDCCCXVII-MDCCCXX jussu et auspiciis Maximiliani Josephi I. Tomus I.

Lacertae. Monachii. 26 pp., 28 pis.
STEJNEGER, L. 1901. An annotated list of the batrachians and reptiles collected in the vicinity

of La Guaira, Venezuela, with descriptions of two new species of snakes. Proc. U.S. natn.

Mus., 24 : 179-192.
WAGLER, J. G. 1830. Natiirliches System der Amphibien mit vorangehender Classification der

Sdugthiere und Vogel. Miinchen. vi + 354 pp., 9 pis.

WERNER, F. 1899. Beschreibung neuer Reptilien und Batrachier. Zool. A nz., 22 602 : 479-484.
WIED, M. 1825. Beitrage zur Naturgeschichte von Brasilien, vol. i, xxii + 6i4 pp.
WIEGMANN, A. F. A. 1834. Herpetologica Mexicana. I. Saurorum Species. Berlin.

vi + 54 pp., 10 pis.

1835. in MEYEN, Beitrage zur Zoologie, gesammelt auf einer Reise um die Erde. Siebente
Abhandlung. Amphibien. Nova Acta Leopoldina, 17 i : 183-268.

Professor R. ETHERIDGE, Ph.D.
Department of Zoology
SAN DIEGO STATE COLLEGE
SAN DIEGO, CAL.
U.S.A.

Printed in Great Britain by

AJden & Mowbray Ltd
at the Alden Press, Oxford

ON THE GENUS LYCOPTERA
AND ITS RELATIONSHIP WITH

THE FAMILY HIODONTIDAE
(PISCES, OSTEOGLOSSOMORPHA)

li-o*'

%RAL^ :

P. H. GREENWOOD

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 8

LONDON: 1970

ON THE GENUS LYCOPTERA AND ITS

RELATIONSHIP WITH THE FAMILY
HIODONTIDAE (PISCES, OSTEOGLOSSOMORPHA)

BY

PETER HUMPHRY GREENWOOD

Pp. 257-285 ; 10 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

ZOOLOGY Vol. 19 No. 8

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 8 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 21 July, 1970 Price

ON THE GENUS LYCOPTERA AND ITS

RELATIONSHIP WITH THE FAMILY
HIODONTIDAE (PISCES, OSTEOGLOSSOMORPHA)

By P. H. GREENWOOD

CONTENTS

Page

INTRODUCTION ........... 259

A COMPARISON OF Lycoptera middendorffi WITH Hiodon

AND Eohiodon SPECIES ......... 260

Study material .......... 260

Neurocranium .......... 260

Nasals ........... 261

Median ethmoid ......... 261

Orbital region .......... 261

Temporal fenestra ......... 263

Otico-temporal region ........ 266

Skull base .......... 267

Otoliths 267

Circumorbital bones ......... 270

Opercular series . . . . . . . . .271

Hyoid arch .......... 272

Palatopterygoid arch . . . . . . . . .273

Vertebral column ......... 273

Caudal fin skeleton . . . . . . . . .275

Fins and fin girdles ......... 279

Scales ........... 280

DISCUSSION ........... 280

Summary of anatomical features ....... 280

Taxonomic considerations ........ 282

Zoogeographical implications ....... 283

ACKNOWLEDGEMENTS ......... 283

SUMMARY ........... 283

REFERENCES ........... 284

INTRODUCTION

THE Lycopteridae, a family of small (< 12 cm) upper Jurassic-lower Cretaceous
teleosts, are widely distributed in lacustrine deposits of eastern Siberia, Mongolia
and China (from 36-56N, and 96-i34E).

Until Yakovlev's (1965) recent revision, the systematics of this family were in some
disorder, with at least sixteen nominal species referred to the genus Lycoptera. In
addition, a number of rather ill-defined genera were also recognized. Yakovlev's
critical study suggests that there are only two species of Lycoptera (L. middendorffi
Miiller, 1848, and L. fragilis Hussakof, 1932), and that three other and monotypic

260 P. H. GREENWOOD

genera should also be recognized (Manchurichthys, Sinolycoptera and Mesoclupea).
Compared with Lycoptera these genera are poorly studied, and contribute little to
understanding the phyletic relationships of the family. Thus, for the purposes of this
paper references to the lycopterids are, in fact, to the genus Lycoptera itself.

Apart from Reis' (1910) observations on the otoliths of Lycoptera middendorffi
(and some notes on vertebral structure by Saito [1936]), no really comprehensive
anatomical study appeared until Berg's revision of the family was published in 1948.
Thus it is not surprising that the systematic position of the Lycopteridae has been in
some doubt ever since the family was denned by Cockerell (1925). Yakovlev's (op.
cit.) remarks that " Lycoptera has to this day remained the most interesting and
enigmatic fish of the Asian Mesozoic ", are fully justified, particularly when it is
recalled that Lycoptera has been considered the ancestor or near ancestor of the
Cyprinidae. If this relationship could be accepted, then the lycopterids might also
throw some light on the origin of the largest and most diverse group of freshwater
fishes, the Ostariophysi.

It was mainly from this viewpoint (see Rosen and Greenwood, 1970) that I began
to study the Lycopteridae in detail. Soon it became clear, however, that the lycop-
terids and the cyprinids (or for that matter any member of the Ostariophysi) were in
no way related, a conclusion reinforced by the appearance of Gaudant's (1968) paper
dealing at length with the anatomy of Lycoptera davidi ( = L. middendorffi).

Several features in the osteology of Lycoptera suggested a relationship with the
osteoglossomorph family Hiodontidae. It is the object of this paper to review the
osteology of Lycoptera (particularly since I disagree with Gaudant [op. cit.] on
certain points), and in the light of that review to compare the two families and
reconsider the various proposed relationships of the Lycopteridae.

The only fossil hiodontid, Eohiodon rosei (Hussakof), from the middle Eocene beds
of British Columbia (see Cavender, I966a) closely resembles Hiodon in most anatomi-
cal features. References to Hiodon thus include Eohiodon as well as the two extant
species of Hiodon, H. alosoides and H. tergisus. Where Eohiodon resembles, or
departs from Lycoptera more than does either Hiodon species, then it will be given
particular mention.

A COMPARISON OF LYCOPTERA MIDDENDORFFI WITH HIODON AND EOHIODON

SPECIES

Study material : Information on Lycoptera osteology was derived principally from
eight well-preserved specimens of L. middendorffi in the collections of the British
Museum (Nat. Hist.) Palaeontology Department (specimen numbers Pi6967a and b,
and P20930) ; forty-eight less well-preserved specimens from the same slabs also
provided some information, as did three specimens kindly loaned to me by the
Munich Museum.

Most of the Hiodon material examined is of H. alosoides (27 specimens) but some
H. tergisus (n specimens) were also studied ; dissections, dried skeletons, alizarin
transparencies and radiographs were used. The size range of these fishes is 2-5 to
26-0 cm standard length. All are from the B.M. (N.H.) collections.

Neurocranium (Text figs, i to 4). The nature of the Lycoptera fossils is such that

ON THE GENUS LYCOPTERA 261

it is impossible to prepare a complete, three-dimensional neurocranium. Conse-
quently, any comparison is restricted to superficial regions of the skull. It is particu-
larly regrettable that no observations could be made on the otic regions of the skull
in these fishes since the otic region provides a highly characteristic and specialized
feature of the hiodontids (Ridewood, 1904 ; Greenwood 1963, and in preparation).

Nasals. These bones are not well-preserved in most specimens. The few that I
have been able to examine are narrow, slightly curved, gutter-like elements closely
associated with and separated by the median ethmoid.

In Hiodon, the nasals are tubular, and more strongly curved (indeed, are angled).
As in Lycoptera they are associated with the median ethmoid, and posteriorly are in
contact with the frontals.

Gaudant (op. cit.) likens the nasals of Lycoptera to those of Elops saurus, but in
my opinion they are more like those of the Mormyridae. In Elops (and other
elopoids, see Nybelin 1956 and 1967) the sensory canal is contained within a tube
provided with four pores, and the base of the nasal is expanded to a greater (Elops
and Megalops) or lesser (Tarpon) degree. In contrast, the mormyrid nasal is an
open gutter (personal observations ; Taverne 1968 and 1969 ; Ridewood, 1904).

Yakovlev (1965) on the other hand, describes the nasals of Lycoptera as " . . . con-
tiguous and not divided by the frontal bones . . . , even in the posterior third ..."
I would agree that the frontals probably do not separate the nasals, but in no speci-
mens can I find evidence of the nasals being contiguous at any point. Judging from
the shape of the anterior frontal tips it is even possible that a short region of each
frontal lies between the posterior part of the nasals (personal observation ; and see
Gaudant's plate 4, figs, i and 2). The supposed contiguity of the nasals is one of the
reasons why Yakovlev (op. cit.) thought that Lycoptera might be related to the
osteoglossids (see page 282).

Median ethmoid : Lying between the nasals is a relatively small, almost rectangular
bone identified as a rostral by Berg (1948), and as a rostro-postrostral by Gaudant
(op. cit.). Yakovlev (op. cit.) refers to a rostral in Lycoptera (but gives no description
of the bone), and uses its presence as part of an argument to show that the ethmoid
region in Lycoptera is more primitive than in Leptolepis coryphaenoides.

From my own observations (on admittedly few and, for that region, poorly pre-
served specimens) and from Gaudant's figures and plates, I can see no reasons for
identifying the bone as a rostral (see Gardiner, 1963 and Nybelin 1967 for discussion).
There is no trace of a transverse rostral commissure (pace Berg, op. cit.), nor does the
arrangement of the circumorbital bones suggest that one was present. Gaudant (op.
cit.) tentatively identifies, in one of his specimens, an ethmoidal pit line, but in none
does he mention a tubular commissure ; none is visible in any of his figures.

Thus, I would propose that this bone be identified as a median ethmoid. Its shape
and spatial relationships are like those of the median ethmoid in Hiodon.

Orbital region (Text figs, i and 2). Little can be learned about the orbital region
in Lycoptera. The orbitosphenoid is somewhat smaller than in Hiodon ; the ptero-
sphenoid and basisphenoid have not been clearly defined although faint indications of
both are sometimes visible.

A long, splint-like supraorbital bone has been identified by Berg (1948) and by Liu

262

P. H. GREENWOOD

et al (1963). Gaudant (op. cit.}, however, failed to find this element, and I would
agree with his conclusion that it is wanting in Ly copter a. Hiodon, too, lacks a supra-
orbital (as do all the Osteoglossomorpha) .

Soc

Exsp

Pt

Dp

lop

Dent

Lach

Max'

SmX

FIG. i. Lycoptera middendorffi. Skull slightly restored ; based mainly on specimen B.M.
(N.H.) P2OQ3O I. Lateral line canals of extrascapular from Berg (1948) and Gaudant
(1968) ; canals in dentary omitted. The endopterygoid is not shown. The margins of the
2nd infra-orbital bone are indicated by dotted lines.

Ang. angular ; Ant. antorbital-prefrontal ; Art. articular ; Bs. basisphenoid ; Dent,
dentary ; Dp.? dermopalatine ; Ds. dermosphenotic ; Epi. epiotic ; Exsp. extrascapular ;
Fr. frontal ; Hy. hyomandibula ; Ic. intercalar ; lop. interoperculum ; Lach. lachrymal ;
Max. maxilla ; Meth. median ethmoid ; Mth. metapterygoid ; Nas. nasal ; Op. operculum ;
Os. orbitosphenoid ; Par. parietal ; Pmx. premaxilla ; Pop. preoperculum ; Pt. post-
temporal ; Pto. dermopterotic ; Pts. pterosphenoid ; Q. quadrate ; Smx. supramaxilla ;
Soc. supraoccipital ; Sop. suboperculum ; Sy. symplectic ; Tf. temporal fenestra. 2-5.
2nd to 5th infraorbitals.

Frontals : (Text figs. 3 and 4). In their general outline and relative contribution
to the dorsicranium, the frontals of Lycoptera and Hiodon are similar (see below, p.
271, regarding the sensory canal systems). Whether or not in Lycoptera the anterior
tips of the frontals lie between the nasals, is uncertain (see above, p. 261).

ON THE GENUS LYCOPTERA

263

Parietals (Text figs. 3 and 4). The entire outline of these bones is not clearly
defined in any of the B.M. (N.H.) specimens. However, from the observations I could
make, I would agree with the restorations figured by Gaudant (op. cit.). The shape

Soc

1cm

Pt

Pmx

Sop

Dent

Lach

Pop

Max

Art

FIG. 2. Hiodon alosoides. Skull. Endoptery gold, and lateral line canals of dentary omit-
ted. The extrascapular has been removed to show the parietal, and the temporal fenestra.
Abbreviations as in fig. i.

of the parietals in Lycoptera and Hiodon is not markedly different. But, in Hiodon
the supraoccipital is a much larger bone with a greater anterior extension ; conse-
quently the area of medial contact between the parietals is much less than in Lycoptera
(cf. text figs. 3 and 4).

Temporal fenestra. A large temporal fenestra is visible in all B.M. (N.H.) speci-
mens of Lycoptera in which that region of the skull is well-preserved. Gaudant
(op. cit.} also found this opening in his material. Liu et al (1963), however, illustrate
the dorsicranium of a specimen (size not stated) in which no fenestra is present.
Gaudant suggests that Liu's fish may be a very large individual, and that with growth
the bones surrounding the fenestrae encroach upon, and eventually obliterate the
openings completely. No specimens with fenestrae in an intermediate stage of

26 4

P. H. GREENWOOD

closure have been illustrated or noted, and there remains the possibility that Liu
et al overlooked the fenestrae in their material. Even in the best specimens the
fenestral outline is faint.

Gaudant (op. cit.} describes the temporal fenestra as being bordered by the
parietals, dermopterotic, extrascapular and supraoccipital. Some of these bones are
undoubtedly associated with the fenestra, but since the extrascapular lies above the
opening (text figs, i and 3) it can hardly be considered one of the delimiting bones.
Also, from my own observations I am uncertain whether the supraoccipital is
involved. A small area of the medial margin, which Gaudant identifies as part of the

Meth

Mas

Soc

Exsp

Pt

cd 2mm

FIG. 3. Lycoptera middendorffi. Dorsicranium. Modified after Gaudant (1968). Abbre-
viations as in fig. i, except Pro. pterotic.

supraoccipital, appears to be a narrow, posterior prolongation of the parietal (see
text fig. 3). Finally, I would identify the posterior boundary of the fenestra as being
the epiotic ; from Gaudant's figure and description he implies that the extrascapular
forms that border. In at least two B.M. (N.H.) specimens the extrascapular is dis-
placed and the epiotic can be seen forming the hinder limit of the fenestra.

The significance and homology of the temporal fenestra are difficult to establish.
As Ridewood (1904) noted, it would appear to correspond with the preepiotic fossa
found in most clupeoid fishes. This fossa is a small but deep pit lying immediately
anterior to the epiotic, and is formed by an invagination from the parietal, epiotic and
autopterotic. There are, however, differences in the two structures. The clupeoid
preepiotic fossa usually ends blindly against the deeper part of the supraoccipital

ON THE GENUS LYCOPTERA

265

and is a distinct pit in the bones involved. The temporal fenestra does not involve
invagination of the bones, and if it were not closed by cartilage, would open directly
into the cranial cavity.

A temporal fenestra similar to that of Lycoptera, (albeit less extensive) occurs in
Hiodon, where it is occluded by a plug of cartilage. In H. alosoides the fenestra is
largest in small fishes, but is still clearly visible in the largest specimen examined
(26 cm S.L.). It is filled by a slip of epaxial body musculature and the whole area is
covered by the greatly expanded extrascapular (cf. Lycoptera where the extrascapular
only covers the posterior quarter of the fenestra.)

Meth

Mas

Ant

Asph

Ic

Exo

FIG. 4. Hiodon alosoides. Dorsicranium. The plane of orientation is such that a line drawn
through the centre of the basioccipital facet to the vomer, would be horizontal. Asph.
autosphenotic ; Exo. exoccipital. Others as in fig. i.

As Gaudant (op. cit.} points out, the temporal fenestra is an unusual character, and
one not found in any other Mesozoic teleost. Indeed, Lycoptera and Hiodon would
appear to be the only primitive teleosts in which a temporal fenestra is developed (see
also Greenwood, 1963) . The large cartilaginous area in the temporal region of salmon-
oid and galaxioid fishes (the 'temporal fossa' of McDowall, 1969) resembles the tempo-
ral fenestra but is a more extensive, less well-defined area bordered by more (and

266 P. H. GREENWOOD

often intrasubordinally different) bones ; it would seem to be a correlate of the
generally reduced ossification of the salmoniform skull.

Undoubtedly the development of a temporal fenestra (Lycoptera and Hiodon) or
preepiotic fossa (Clupeoidei) is associated with the invasion of that part of the skull
by epaxial body muscles. That a fossa is developed in clupeoids and a fenestra in
Hiodon and Lycoptera may indicate that the structures are more probably analogous
than homologous. Certainly there is no evidence from other characters to suggest a
close relationship between the clupeoids and hiodontids (Greenwood, 1963 ; Green-
wood et al, 1966).

In Lycoptera only the posterior part of the temporal fenestra is covered by the
extrascapular, whilst in Hiodon the entire fenestra (and most of the parietal) lies
beneath the expanded, scale-like extrascapular. Thus both the relative size, and
the shape of this bone differ in the two genera. Another difference is the arrangement
of the extrascapular latero-sensory canals. Lycoptera, according to Berg (1948) and
Gaudant (op. cit.}, had the posterior section of the temporal canal (i.e. the section
between the posttemporal and the pterotic) passing through the lower part of the
extrascapular (see text fig. i). The branch forming the transverse commissure is
given off near the hind margin of the bone, and runs medially in that position. (None
of the B.M. (N.H.) specimens has this region of the skull sufficiently well-preserved
to allow for first-hand observation on these points) . In Hiodon, the branch from the
posttemporal connects with the dermopterotic directly, and the extrascapular carries,
near its posterior margin, the transverse commissure only. Both genera are alike in
having the bony tube for the transverse commissure interrupted in the midline because
the extrascapulars are separated by the supraoccipital. The gap is wider in Lycoptera
than in Hiodon.

The otico-temporal region. The pterotic is an elongate and substantial bone in both
genera. As far as I can determine, the temporal latero-sensory canal in Lycoptera is
entirely bone enclosed ; in Hiodon the ventromedial aspect of the tube is open
(Greenwood, 1963).

Virtually nothing is known about the occipital and lower otic regions of the Lycop-
tera skull, since that region is invariably covered by the collapsed operculum and
associated bones.

From Gaudant's description and figures (especially plate 3) it is clear that the
supraoccipital in Lycoptera is a smaller bone than in Hiodon, and that it does not
separate the parietals for such a distance anteriorly. The supraoccipital crest in
Lycoptera appears to be lower and probably does not have the characteristically " T "
shaped section of Hiodon. However, judging from Gaudant's plate the crest may
have a short-armed " T " section.

The presence of a moderately large supraocciptal in the Lycoptera middendorffi
specimens figured by Gaudant (op. cit.} makes it the more difficult to understand the
minute bone (not separating any part of the parietals, and hardly separating the
extrascapulars) illustrated by Liu et al (1963, fig. 8). It will be recalled (p. 263) that
these authors do not show a temporal fenestra in this fish, which Gaudant suggested
might be a large individual in which the fenestrae had closed. Unfortunately Liu
et al give no indication of the size of the specimen (nor any magnification for the

ON THE GENUS LYCOPTERA 267

figure), neither do they discuss the supraoccipital in the text of their paper ; no
dorsicranium is figured in the plates. That a fenestra might be obliterated by growth
is a reasonable assumption, but it is improbable that parietal growth would almost
cover the supraoccipital. I am inclined to think that Liu et al, probably influenced
by the relationships of these bones in Leptolepis, have misinterpreted the situation in
Ly copter a (cf. their drawings of these skulls in figure 8).

Skull base (Text figs. 5A and B). The parasphenoid is toothed in both genera with
the teeth relatively larger and extending further posteriorly in Hiodon (and probably
in Eohiodon). A small basipterygoid process is present in Ly copter a, but is absent in
Hiodon. Gaudant (op. cit. plate 5, figs. 3 and 4) illustrates the posterior half of the
parasphenoid in Lycoptera middendorffi, and clearly demonstrates its sharp upward
angling (ca. 45), a little posterior to the foramen for the internal carotid (see text fig.
56) . His photographs also show the relatively short, paired and near vertical ascend-
ing processes of this bone. These come into contact with the prootic, while the section
running obliquely upwards forms the ventral, unpaired part of the base underlying
the otico-occipital region of the skull.

.Oca

Ora

FIG. 5. A. Hiodon alosoides. Posterior part of a parasphenoid, viewed from the right.
B. Lycoptera middendorffi. Posterior part of parasphenoid, viewed from the right (after
Gaudant, 1968 : plate 5, fig. 4). Asa. ascending arm ; Fie. foramen for internal carotid
artery ; Oca. occipital arm ; Ora. orbital arm.

The shape of the parasphenoid and the details of its morphology (except for the
basipterygoid processes) are very similar in Lycoptera and Hiodon (see text fig. 5).
Thus it seems reasonable to conclude that the ventral skull profile in Lycoptera had the
same, peculiar " stepped " appearance as in Hiodon.

Otoliths (Text figs. 6 A-C). The peculiar otoliths of Lycoptera middendorffi were
described and illustrated with great thoroughness by Reis (1910). In most of his
specimens two of the three otoliths from each side (and sometimes from both sides)
were preserved, if not in situ, at least closely associated with the skull. Accordingly
Reis was able to determine the approximate spatial relationships of the otoliths, and
on this basis to show that the asteriscus (lagenar otolith) is much larger than the
sagitta (saccular otolith). From Reis' drawings and descriptions, the asteriscus is
about 4 to 6 times longer than the sagitta. It also has an unusual, trapezoidal or
unequally hexagonal outline. The shape of the sagitta is difficult to determine from
the figures, but it appears to be trianguloid and without any marked sculpturing.
The asteriscus, by contrast (text fig. 6C) has each face radially fluted over about half

268 P. H. GREENWOOD

its area, and there is at least one, short, crescentic groove at a point about two-thirds
distant from the lower border (as orientated in the figures). Ornamentation seems
to be more obvious on one face than on the other. No otolith was identified (or
shown in a drawing) as a lapillus (utricular otolith) and Reis (op. cit.} makes no com-
ments on its absence.

As Reis points out, it is most unusual among teleosts for the asteriscus to be larger
than the sagitta, a condition otherwise found only in the Ostariophysi (and to a
somewhat less marked degree in the Mormyridae, see Taverne, op. cit.} In both these
groups, the shape of the sagitta and asteriscus is quite unlike their presumed counter-
parts in Lycoptera. Indeed, a large, hexagonal-trapezoidal asteriscus has, to the
best of my knowledge, never been described from any other teleost group. Further-
more, the peculiar radial fluting of the Lycoptera asteriscus has not been described
elsewhere.

Later workers (Saito, 1936 ; Takai, 1944 ; Berg, 1948 ; Liu et al 1963) do not
describe otoliths from their material. When the otoliths are mentioned by these
authors, it is always with reference to Reis' (1910) paper and figures. Gaudant (op.
cit.} found otoliths in only one of the specimens he examined (L. middendorffi, B.M.
(N.H.) reg. no. P273 [now ?47792] ; but see below). His comments on Lycoptera
otoliths, like those of his predecessors, are also based on Reis. (It should be noted
that Gaudant reproduces one of Reis's drawing of an asteriscus but in the legend to
the figure calls it a sagitta).

None of the B.M. (N.H.) specimens has the otoliths preserved, and I was unable to
find them in specimen ?47792 (where, according to Gaudant [op. cit.], they should be
visible). I have tried to locate Reis' material, but without success. Apparently it is
now either lost or destroyed. The absence of suitable material for otolith studies is
the more regrettable because certain features of the otoliths described by Reis (op.
cit.} suggest that his identification of the larger element as an asteriscus may be in
error.

Compared with the sagitta, far less is known about intergroup variation in the
shape of the teleost lapillus and asteriscus. Nevertheless, some generalizations on
this subject are possible (based on Retzius, 1881 ; Frost, 1925-1930 ; Taverne, 1968
and 1969 ; and personal observations).

Apart from in the Ostariophysi and the Mormyriformes, the asteriscus is generally a
flattened, slightly elongate, often kidney-shaped otolith with all or most of its margin
deeply crenulate (often almost serrate) ; there is a deep and extensive sulcus on the
median face. In the Ostariophysi and Mormyriformes the asteriscus is usually
discoidal, the broad sulcus horseshoe-shaped, and the margin variously crenulate,
serrate or smooth (or a combination of all three). A few ostariophysans have an
elongate asteriscus which closely resembles the " typical " sagitta of other groups.

The angular, trapezoidal to hexagonal outline of the so-called asteriscus in Lycoptera
has not yet been found in any teleost. Neither does the Lycoptera " asteriscus "
have the prominent sulcus characteristic of that otolith in most (if not all) teleosts.
Reis (op. cit.} attempts to identify a sulcus, but it is clear from both his description
and figures that the feature he is dealing with is more in the nature of an elongate
pit than the typically deep, extensive groove of the asteriscal sulcus.

ON THE GENUS LYCOPTERA 269

A trapezoidal, or at least a noticeably angular outline appears to be more charac-
teristic of the teleost lapillus (utricular otolith) . Furthermore, that otolith does not
show a well-developed sulcus, and its marginal ornamentation is often in the form of
rather widely spaced radial grooves extending towards the centre of the otolith. On
the upper face of the lapillus (near the region where the radial grooves converge)
there is frequently a shallow, depressed area lying close to a low dome or ridge
(text fig. 6A).

Altogether, the so-called asteriscus of Lycoptera shows more of the characters of a
lapillus than an asteriscus. I am especially impressed by the resemblance between
the Ly copter a " asteriscus " and the lapillus of Hiodon alosoides (cf. text figs. 6A and B
with 6C, and with the figures in plate 2 of Reis, 1910). This resemblance includes the
outline shape as well as many details of ornamentation, including the spaced radial
grooves (more numerous over one half of the otolith) and the absence of a well-
defined sulcus. Thus, I suspect that Reis misidentified the lapillus in Ly copter a as
an asteriscus.

<a.1mm

FIG. 6. A, B. Hiodon alosoides, lapillus in dorsal and ventral views. C. Lycoptera midden-
dorffi, " asteriscus " (after Reis, 1910, plate 2, fig. n)

There remains, however, the problem posed by the disposition of the otoliths as
Reis has figured them in relation to the skull (Reis, op. cit., plate 2). The problem is
complicated by the lack of a third otolith in any specimen figured. In Hiodon the
lapillus lies above and well anterior to the sagitta and asteriscus. One would,
therefore, expect that in a fossil (even allowing for some post mortem dislocation) the
lapillus should lie ahead and slightly above the other two otoliths (or the sagitta
alone if that is the only lower otolith visible). In Reis' figures, the larger " asteris-
cus " is always shown behind the smaller otolith (his sagitta), and generally above it.
The labyrinth of Hiodon is unusual in having the lapillus so far in advance of the
sagitta ; in most fishes the lapillus lies above or even slightly behind the sagitta. If
this latter condition was found in Lycoptera, then it would be possible for the sagitta
to come to lie in front of the lapillus. Also, if Lycoptera had a large auditory fenestra
(as in Hiodon) lateral to the sacculus, it seems even more likely that the sagitta could
be displaced anteriorly during fossilization. The bone enclosed lapillus would be less
subject to horizontal displacement.

I have of course assumed, as did Reis, that the smaller otolith is a sagitta. This
point also remains to be clarified ; the figures and descriptions given by Reis (op. cit.}
are not sufficiently clear for a decision to be made. The sagitta of Hiodon alosoides

270 P. H. GREENWOOD

is a most peculiar otolith and it should be possible to recognize its salient features in
well-preserved fossils. All that can be said at present is that no features visible in
Reis' drawings would immediately preclude the smaller otolith from being of the
Hiodon sagitta type.

The so-called lycopteroid otoliths from the Mesozoic of Germany (Weiler, 1954,
1957 and 1965) are not closely similar to those of Ly copter a middendorffi (see page 283) .

Circumorbital bones (Text fig. i). Because the circumorbital series is generally not
well preserved, little agreement has been reached on the number or the shape of these
bones (cf. Berg, 1948, fig. i ; Liu et al, 1963, fig. 8 ; Gaudant, 1968, fig. 3.).

Berg (op. cit.} illustrates a single, elongate suborbital bone lying between the
canal-bearing infraorbital series and the preoperculum. From Gaudant's (op. cit.}
study, and from my own observations it is certain that Berg was in error. No sub-
orbital is present ; the circumorbital series in Lycoptera is typically teleostean and is
composed solely of canal-bearing infraorbital bones.

Judging from the two B.M. (N.H.) specimens (both on slab P2093O) and from Gau-
dant's plate 2, fig. i, Liu et al's (op. cit.) restoration is the most accurate so far pub-
lished. Certainly Gaudant's text fig. 3 is inaccurate in some respects. For example,
the postorbitally situated infraorbitals are considerably wider than he depicts them,
and the lachrymal (ist infraorbital) is much longer (see text fig. i).

From my material I cannot be certain whether there are one or two bones lying
below the orbit (Berg figures two and part of a third, Liu et al three, and Gaudant
three and part of a fourth) . Be that as it may, the first infraorbital (lachrymal) is a
narrow, elongate bone deepening a little (or more definitely if there is no small
second infraorbital) posteriorly. Behind the orbit there are two deep and wide
bones, which extend backwards almost to the margin of the preoperculum. Above
these it is difficult to determine the number or shape of the succeeding infraorbitals.
In specimen P2093O there appears to be a short, barrel-shaped bone followed by an
elongate, almost boomerang-shaped dermosphenotic.

Anteriorly, a large, shield-shaped antorbital is apparently fused with the pre-
frontal. Berg (1948) and Liu et al (1963) identify this bone as the lateral ethmoid,
but it is clearly a canal bearing bone with another ossification lying medial to it (which
I identify as being the prefrontal). My reason for thinking that the antorbital and
prefrontal are fused is simply that the two bones (in all specimens I have examined)
always have the same relative positions ; the other infraorbital bones are invariably
displaced and variously disarticulated.

The bone Berg (op. cit.} identifies as a dermosphenotic (i.e. the uppermost infra-
orbital) is roughly triangular, as is the dermosphenotic in Gaudant's (op. cit.} restora-
tion. A bone occupying a similar position in my best preserved specimen (P. 20930)
is much narrower and is boomerang-shaped, the concave face directed anteriorly
(text fig. i). But, as there is only one specimen with a presumed dermosphenotic
visible, I could well be mistaken on this point.

Regarding the total number of infraorbital bones (including the dermosphenotic
and antorbital), Berg (op. cit.} finds 7 (i.e. interpreting his lateral ethmoid as an antor-
bital), Liu et al 6 (again interpreting the lateral ethmoid as an antorbital), and
Gaudant (op. cit.} 6. However, Liu et al do not specifically identify a dermos-

ON THE GENUS LYCOPTERA 271

phenotic (which they do in the Leptolepis they figure) ; from their drawing it seems
likely that a 7th infraorbital (i.e. a dermosphenotic) is present. The count from
specimen B.M. (N.H.) P20930 could be either 6 or 7 depending on whether or not the
small " second " infraorbital is a preservation artefact.

Hiodon has a total of 6 infraorbital bones (Nelson, 1969). Although there is an
overall similarity between Hiodon and Lycoptera in the shape of the entire series
(narrow suborbitally, expanded postorbitally) there are differences in detail (cf.
text figs, i and 2). The expanded postorbital bones are deeper in Hiodon, and the
infraorbital preceding the dermosphenotic is expanded (short and narrow in
Lycoptera). The antorbital in Hiodon, like that of Lycoptera, is fused with the pre-
frontal, and the dermosphenotic is an elongate, angled bone (larger than, but
basically similar to Lycoptera).

Cephalic lateral line system (Text figs. I to 4). Superficially, the cephalic lateral line
system is virtually identical in both Lycoptera and Hiodon. It is of the primitive
type in which there is no connection between the supraorbital and temporal canals ;
the infraorbital canal is connected, through the dermosphenotic, with the temporal
canal in the dermopterotic, but the supraorbital system ends in the parietal. The
supraorbital canal has medial and lateral branches in Lycoptera, but these are not
developed in Hiodon.

Other differences concern branching of the infraorbital system. In Hiodon this
canal divides in the dermosphenotic, one branch curving backwards to join the
temporal canal, the other and longer branch running forward in the dermosphenotic
below and parallel with the supraorbital canal in the frontal. No such dichotomy
has been described in Lycoptera.

In Hiodon, the pterotic portion of the temporal canal is open medially for much
of its length (Greenwood, 1963) and forms part of what appears to be a pseudo recessus
lateralis (in connection with which other bones, including the hyomandibula, are
modified ; personal observations). Unfortunately the nature of the pterotic canal
cannot be determined in Lycoptera (or in Eohiodon).

Opercular series (Text figs, i and 2). The preoperculum of Lycoptera, compared
with that of Hiodon, and especially Eohiodon, has a more pronounced posterior
prolongation of its lower, horizontal arm which is also shallower than in Hiodon. The
vertical arm is longer, and reaches the pterotic ; in Hiodon there is a distinct gap
between the two bones. Expressed in another way, the preoperculum in Lycoptera
extends vertically beyond the hyomandibula head, whereas in Hiodon it barely reaches
the level of the opercular process on the hyomandibula.

The operculum of Lycoptera is a relatively larger bone and extends further dorsally
than it does in Hiodon ; in that genus its upper margin is excavated, not rounded as in
Lycoptera.

The inter operculum of Lycoptera appears to be more ventrally situated because it is
visible in lateral view, whereas in Hiodon it is hidden behind the ventral limb of the
preoperculum. There are also differences in the shape of the inter operculum. In
Hiodon the bone is almost rectangular, is narrow and nearly as long as the horizontal
arm of the preoperculum. In Lycoptera it is relatively shorter (about half the length
of the preoperculum), is somewhat deeper, and has a curved lower margin.

272

P. H. GREENWOOD

The suboperculum is similar in both genera.

Hyoid arch (Text figs. 7 A and B). The hyoid arch of Ly copter a middendorffi has
been described in great detail by Gaudant (op. cit.}. The hyomandibula differs from
that of Hiodon in several respects. Both, however, have a very well-developed
opercular process, but this is probably of little significance since a large process is
found in such unrelated groups as the esocoids and gadoids, as well as in the osteo-
glossoids and notopteroids. Ly copter a has a single articular head to the hyomandi-
bula ; in Hiodon this surface is double with a large but bridged gap between each
head (an arrangement probably correlated with the pseudo recessus lateralis men-
tioned above).

hH

PhH

FIG. 7. A. Lycoptera middendorffi. Right hyomandibula, lateral aspect (after Gaudant,
1968). hH. head of hyomandibula ; Op. opercular process. B. Hiodon alosoides.
Right hyomandibula, lateral view. AhH. anterior head ; PhH. posterior head ; Op.
opercular process.

The shape of the ceratohyal, as well as the shape, number and disposition of the
branchiostegal rays (especially the three, curved, broad upper rays) are closely similar
in both Lycoptera and Hiodon. It may be noted in passing that in these characters
neither genus resembles either the elopoid or clupeoid types ; perhaps the closest
resemblance is to the notopterid and mormyrid types (McAllister, 1968).

Gaudant (op. cit.} describes and figures a single hypohyal in his specimens, but Berg
(1948) found two small hypophyals. Unfortunately, this area of the hyoid arch is
not clearly visible in any B.M. (N.H.) specimens. Hiodon has a pair of hypophyals
on each side.

Both genera have a well-developed basihyal dentition, but the shape of the basihyal
and its tooth plate are not known in Lycoptera. This is regrettable, since Nelson
(1968) has drawn attention to the peculiar (and apparently characteristic) shape of

ON THE GENUS LYCOPTERA 273

the tooth plate in Hiodon ; a similarly shaped plate is otherwise only known from the
related Notopteridae (Nelson, op. cit.}.

Palatopterygoid arch (Text figs, i and 2). Again, it is from Gaudant's (op. cit.)
work that the most detailed descriptions are available (see his text fig. 8).

The metapterygoid of Hiodon is relatively smaller than that in Lycoptera, and is
kidney-shaped, without the antero-dorsally directed flange which overlaps the endop-
terygoid in Lycoptera.

Gaudant (op. cit.} identifies a small, toothless bone lying above the ectopterygoid
(and immediately anterior to the endopterygoid) as a dermopalatine. If he is correct
(which I doubt), then Lycoptera differs from Hiodon where the dermopalatine is
toothed, and lies anterior to and is in a linear series with the ectopterygoid. One
B.M.(N.H.) specimen, P20Q3O, shows medially and a little above the dorsal margin of
the maxillary head, a short length of toothed bone (see text. fig. i). It seems reason-
able to identify this bone as the dermopalatine ; if this is correct, then it is similar to
the dermopalatine of Hiodon.

The ectopterygoid is toothed in Hiodon, toothless in Lycoptera ; the endopterygoid
is toothed in both genera.

The quadrate has a similar shape and proportions in both Hiodon and Lycoptera.

Jaws (Text figs, i and 2) There is an overall similarity in the upper and lower jaw
elements of both genera.

The premaxilla is a small bone without traces of an ascending process ; it is rela-
tively longer in Hiodon, and its articulation with the maxilla is a more intimate
junction in that genus.

The dentary in Hiodon is more slender, and the coronoid region is shallower than in
Lycoptera. I do not agree with Gaudant's (op. cit., fig. 9) figure of the articular. In
B.M. (N.H.) specimens the joint surface of the articular is a deep and clearly defined
notch (not, as he shows it, an elongate concavity) closely fitting around the quadrate
head.

The angular is a small bone in both genera. According to Ridewood (1904) there
is no angular in Hiodon, but it is present in all the alizarin preparations I have
examined.

A single supramaxilla is present in Lycoptera, but absent in Hiodon.

Gular plate. A large gular plate is present in apparently both species of Lycoptera.
No trace of this bone was found in any of the many Hiodon specimens examined.

Vertebral column. Saito (1936) has given a detailed description of the vertebrae
in L. davidi ( = L. middendorffi) to which little can be added. Saito was also the
first to detect ontogenetic changes in the centra, which are diplospondylous in small
individuals but mostly monospondylous in larger fishes (> 50 mm S.L.).

According to Saito (op. cit.} and Gaudant (op. cit.}, the neural and haemal arches of
caudal vertebrae are fused with the centrum (neither author mentions the condition
in abdominal vertebrae). In my material of L. middendorffi (size range 40-55 mm
S.L.) I find that the first four preural vertebrae have autogenous haemal arches, but
that haemal arches on the more anterior vertebrae are fused with the centra in
fishes > 45 mm S.L. ; in smaller fishes all the haemal arches appear to be autogenous.

Gaudant (op. cit.} describes both epineural and epicentral intermuscular bones (the

274 p - H. GREENWOOD

latter he calls epipleurals although they are in no way associated with the pleural
ribs). Other authors (except Takai [1944] who failed to locate intermuscular bones
in some specimens) mention only one pair of intermuscular bones per vertebra, but
differ in their opinions as to whether the bones are epineural or epicentral in position ;
all agree that intermuscular bones are not present posteriorly beyond the first four or
five caudal vertebrae. Takai's confused nomenclature for structures associated
with the vertebrae has been clarified by Yakovlev (1965), who shows that, in fact,
one pair of intermuscular bones is present in all the Ly copter a " species " studied by
Takai (op. cit.}

From the B.M. (N.H.) material studied, I would identify the intermuscular bones
as epineurals. This material also suggests how Gaudant (op. cit.} came to describe
both epineurals and epicentrals. The epineurals in Lycoptera are very elongate
bones, and thus extend backwards across several vertebrae. If the distal end of the
epineural is displaced ventrally (or the tip is broken off and slips downwards) it can
look much like a short intermuscular bone arising from a centrum behind that from
which it actually stems. Careful examination, however, usually enables one to trace
the " epicentral " back to its origin as an epineural on an anterior vertebra.

The pleural (ventral) ribs articulate with the tips of the broad-based, triangular
parapophyses fused with the centra.

In Hiodon, the vertebral centra are somewhat stouter structures than those of
Lycoptera, but are essentially of the same primitive type with a relatively large
notochordal foramen. No trace of diplospondyly can be detected, even in the small-
est specimen examined (H. alosoides 25 mm S.L.) ; it would be of great interest to
know something about the ontogeny of the vertebrae in Hiodon.

More neural arches are autogenous in Hiodon than in Lycoptera (the condition is
unknown in Eohidon], since the arches of all abdominal vertebrae are free from their
centra, as are the first two caudal neural arches. All haemal arches except those of
the first and second preural centra and the first ural centrum, are fused to the
centrum. As in Lycoptera, the left and right halves of each abdominal neural spine
are usually separate, even in large fishes (i.e. 25 cm S.L.) ; in caudal vertebrae,
however, the spine is always a single structure (separate halves are common in the
anterior caudal vertebrae of Lycoptera, and occasionally occur in the posterior
vertebrae as well).

The parapophyses in Hiodon are fused with the centra but are elongate and dagger-
shaped. In contrast with Lycoptera, the pleural ribs articulate directly with the
centrum a little behind the shaft of the parapophysis (Greenwood, 1963).

Yakovlev (op. cit.} tabulated the various counts obtained for the number of
vertebrae in Lycoptera " species " and warned of the difficulties in obtaining exact
figures (diplospondyly, state of preservation, etc.). The counts for total number of
vertebrae range from 43-50 (No indication is given of whether or not all the vertebrae
associated with the caudal fin skeleton are included in these totals) ; 18-24 anterior
vertebrae carry pleural ribs, the first rib being associated with the third centrum.
Hiodon (i.e. both species) has 55-58 vertebrae (excluding the two ural centra), of
which 27 or 28 bear ribs (Data from 9 specimens) ; the first rib is associated with the
third vertebra. Even allowing for difficulties in obtaining accurate counts in

ON THE GENUS LYCOPTERA 275

Ly copter a it is reasonable to conclude that Hiodon has more vertebrae and ribs.
(Another lycopterid, the monotypic genus Sinoly copter a langshanensis resembles
Hiodon in having 58 vertebrae [Yakovlev, op. cit.]). Eohiodon, with 49 vertebrae has
a lower count than either Hiodon species, and thus more closely resembles Ly copter a,
although it may have slightly more caudal vertebrae (ca. 25, cf. 20-23 in Lycoptera} ;
no information is available on the number of pleural ribs in Eohidon (Cavender,
ig66a).

Both Lycoptera and Hiodon have slender and elongate supraneurals lying between
the neural spines of all vertebrae from in front of the first, until about the second
vertebra behind the dorsal fin origin.

Caudal fin skeleton (Text figs. 8, 9 and 10) Besides a few comments by Berg (1948)
on the number of hypurals and epurals, no author before Gaudant (op. cit.} paid
much attention to the caudal fin skeleton in lycopterids. The figures in Reis (1910),
Saito (1936), Takai (1944) and Liu et al (1963) are schematic and provide no critical
information. Liu et al (op. cit.} describe the caudal fin in L. muroii (a synonym of L.
middendorffi) as a "... somewhat advanced character " (relative to other " species "
of Lycoptera} but do not list any specific reasons for their statement.

Gaudant (op. cit., pp. 26-28, fig. 14) gives a detailed figure and description, as well
as an excellent photograph of the specimen on which these were based (see plate 2,
fig. 2). As figured by Gaudant, the caudal skeleton of L. middendorffi resembles the
elopoid type (Nybelin, 1963) but with certain differences.

Salient features in Gaudant's restoration are : a separate second ural centrum,
seven hypurals, three epurals, a fully-developed neural spine on the first pre-ural
vertebra, a reduced spine on the first ural vertebra, and the uroneurals (called
urodermals by Gaudant) arranged in two sets (a lower one of three elongate bones,
and an upper set of two [one moderately long, and other short]). This uroneural
arrangement is peculiar. It is neither of the elopoid type (three lower and a single
upper bone lying at an angle to, and overlapping the former), nor of the leptolepid
type (lower set of four and an upper of three short bones with at least the first over-
lapping the lower set) ; see Nybelin (1963) and Patterson (1967) . The alignment and
size of the upper uroneural (Ur 5) in Gaudant's figure does not accord with any
known teleost caudal skeleton type (see Monod, 1968), nor does the arrangement of a
small upper uroneural lying below an elongate one.

I have examined six well-preserved caudal fin skeletons of L. middendorffi, ca.
45-50 mm S.L., and find that I disagree with Gaudant on two major points. It
might be added that I find the specimen illustrated in Gaudant's plate (op. cit.,
plate 2, fig. 2) to be essentially like those I have examined.

In the B.M. (N.H.) fishes there is a single epural (not 3) closely associated proximally
with the short neural spine on the first ural vertebra (see text fig. 9). The second
point of disagreement concerns the number, shape and disposition of the uroneurals.
In all B.M. (N.H.) specimens I find only a single series composed of three or four
elongate bones, the first three being of approximately the same length, the fourth
somewhat shorter.

I suspect that Gaudant has confused with epurals the proximal ends of caudal fin
rays, the bifid neural spines of preural vertebrae, and possibly, fragments of uro-

276

P. H. GREENWOOD

neurals. The uroneural arrangement he describes is probably the result, in part,
of the uroneurals being broken and displaced. His " urodermal 5 " in plate 2, fig. 2,
and in text-fig. 14, certainly looks like the distal portion of the 2nd or 3rd uroneural.
In other respects Gaudant and I agree on the anatomy of the caudal fin skeleton.
We both find that there is a total of 18 principal caudal fin rays (i unbranched and 8
branched rays in each lobe). The articulated fin rays of the upper lobe are preceded
by about lo-spine-like procurrent rays ; there are 8 procurrent rays in the lower lobe.

H6

NsPU1

UN 1-3

U2

FIG. 8. Hiodon tergisus. Caudal fin skeleton, drawn from a radiograph of B.M. (N.H.)
reg. no. 1862. n. 18 : 45. For abbreviations see fig. 9.

Two of the six B.M. (N.H.) specimens have an interesting condition of the verte-
brae associated with the ist and 2nd hypurals (text fig. 10). Instead of there being
a single first ural centrum carrying both hypurals, each hypural articulates with a
separate centrum. The first caudal centrum carries a neural arch and short spine,
the second an arch and very reduced spine. A third specimen (of about the same
length as the other two, ca. 45 mm S.L.) has these two centra fused but with the
fusion line still visible ; it resembles a caudal vertebra in the intermediate stage
between diplo- and monospondyly. Those specimens with two distinct parts to the

ON THE GENUS LYCOPTERA

277

" first " ural centrum also have a greater degree of displospondyly in the caudal
vertebrae. Presumably the double " first " ural represents an early stage in the
ontogeny of the usual compound first ural centrum. At first sight the double " first "
ural centrum condition might be interpreted as comparable with the diplospondyly
which occurs throughout the vertebral column in small fishes, and becomes progres-
sively confined to the caudal region in larger individuals. However, there is an
important difference between the two ural elements and a pair of hemicentra

UN 1-3

NslM

NaPU1

FIG. 9. Lycoptera middendorffi. Caudal fin skeleton, drawn from a latex cast of B.M. (N.H.)
Pi 6967!?. Ep. epural ; Hi-7 ; hypurals ; naPUI. neural arch of 2nd preural vertebra ;
NsPUi. neural spine of ist preural vertebra ; NsUi. neural spine of ist ural vertebra ;
PH. parhypural ; PUi ; PU 2 , ist and 2nd preural centra. Ui ; U 2 , ist and 2nd ural
centra. UN 1-4, uroneurals.

anteriorly. As Saito (op. cit.} demonstrated, in a diplospondylous caudal vertebra
one hemicentrum of a pair (the anterior) is arch-less, while the other carries a neural
and haemal arch and spine. The ural pair each has a neural arch and spine, as well
as a hypural (modified haemal arch and spine). This suggests that each " first "
ural element is a complete vertebra.

There is no appreciable size difference between individuals with double and single
first ural vertebrae. Perhaps, as Patterson (1967) suggested, Lycoptera shows
individual variability in this character. Gaudant (op. cit.} believed that Patterson's
observations could be attributed to poor preservation of the material. But I can
certainly confirm that this is not so. Indeed, on one slab (Pi6Q67a) there are two

2 7 8

P. H. GREENWOOD

specimens, overlapping one another, and both equally well-preserved ; in one fish
there is a double " first " ural vertebra, in the other a single centrum.

When the first ural is a single element, only one neural arch and short spine is
present, and the same condition is found on those apparently single centra with a
clear-cut line of fusion still persisting. Also, when the centrum is a single element, its
overall length is somewhat less than the two portions together. These observations
seem to indicate that, if fusion is involved in the usual growth process, then there is a
certain amount of resorption and remodelling of the bones.

Ep UN4

NsUlp

UN1

FIG. 10. Lycoptera middendorffi. Caudal fin skeleton (from P2O93O) to show the double
" first ural " centrum. The first branched principal caudal ray of each lobe is indicated
by an arrow. For abbreviations, see fig. 9 and NsUia, NsUip. neural spine on anterior
and posterior elements of " first ural " centrum ; Uia and Uip. anterior and posterior
components, respectively, of " first ural " centrum.

In all teleosts (as defined by Patterson, 1968) the first ural centrum always
carries two hypurals and is therefore to be considered a compound element. The
double condition found in certain Lycoptera middendorffi individuals would seem to
be the retention of a primitive condition, a stage already lost in the other upper Juras-
sic and lower Cretaceous teleosts (eg. LeptoUpis, Thrissops, Allothrissops). Other
characters of the caudal fin skeleton in Lycoptera, eg. perichordally ossified centra,
elongate uroneurals extending forward to the preural centra, two hypural support for
the lower caudal lobe preclude arguments that the skeleton is not teleostean.

Returning now to the overall morphology of the caudal skeleton in Lycoptera. If

ON THE GENUS LYCOPTERA 279

my interpretation of this structure is acceptable, there is a definite resemblance
between the caudal skeleton in Lycoptera and Hiodon ; particularly in three important
characters (cf. text fig. 8 with figs. 9 and 10).

First, there is agreement in the number of branched principal rays (16) ; other prim-
itive teleosts (including Leptolepis) have 17 (i.e. 9+8). Sixteen rays are otherwise
known from the Hiodontidae and most Mormyridae (see Nelson, 1969 for further
discussion ; certain Galaxiidae with 16 rays [see McDowall, 1969] are irrelevant since
the caudal skeleton in these fishes eliminates them from consideration as truly primi-
tive teleosts) .

Second, there is close similarity in the uroneural arrangement. In Hiodon and
Eohiodon the 3 or 4 long uroneurals are arranged in a single series (that is, there is no
upper series of one or more short uroneurals, as is typical for the lower teleosts).
Certain other primitive fishes (eg. Allothrissops, Thrissops, ichthyodectids) also have
only the lower set of uroneurals, but these are far more numerous (up to 7 ; see
Patterson, 1967 ; Cavender, I966b) than in Hiodon or Lycoptera.

Finally, in both Lycoptera and the hiodontids there is only one epural, a most
unusual feature in lower teleosts. All other primitive groups (except perhaps the
Osteoglossoidei, Notopteridae and Mormyridae, where the epural is probably absent,
see discussion in Nelson, 1969) have three epurals (Patterson, 1968).

There is considerable variation is some aspects of the Hiodon caudal skeleton,
especially in the number and development of neural superstructures associated with
the first ural and preural vertebrae. For example, a fully developed spine on the
first preural vertebra (present in Lycoptera) is not always developed, and the neural
spine on the first ural is smaller than in Lycoptera. Nevertheless, the overall picture
of variation for these characters in Hiodon and Eohiodon encompasses the condition
found in Lycoptera. Most specimens of H. alosoides and H. tergisus have two neural
arches on the first ural centrum, a condition reminiscent of that seen in some Lycop-
tera, but probably of no importance other than as a further indication of the com-
pound nature of that vertebra.

Not only in general features, but in at least three critical characters the caudal fin
and its supporting skeleton in Lycoptera is closer to that of Hiodon than to that of
any other genus or group of fishes. Resemblances to the Thrissops ichthyodectid
caudal skeleton will be discussed elsewhere (but see Patterson, 1968, and Greenwood
et al, 1966).

Fins and fin girdles. There is not enough information about the pectoral and
pelvic girdles in Lycoptera for any meaningful comparisons to be made.

The pectoral fin of Hiodon spp., and Eohiodon is more expansive and falcate than
that of Lycoptera, and has more rays (11-14, an d ca. 13, cf. 9). Lycoptera has both
the first and last rays enlarged, thickened and unbranched. In Hiodon only the
first ray is unbranched and noticeably enlarged ; the lowermost ray has, however,
a broad, fleshy extension of the membrane along its trailing edge. No details are
available for this fin in Eohiodon.

Pelvic fin size, shape and area are similar in the three genera ; Hiodon species have
7 pelvic rays (the first unbranched), Eohiodon about 7 (no other details) and Lycoptera
spp., 7 to 9.

28o P. H. GREENWOOD

The median fins in Ly copter a, Hiodon and Eohiodon are set well-back on the body ;
fin outline is similar in the three genera, although the anal is longer in Hiodon (23-32
rays) than in Ly copter a (13-20, according to accounts in Yakovlev, 1965). Dorsal
fin length is approximately equal (9-12 rays in Hiodon, 10-12 in Lycoptera). The
dorsal fin ray count for Eohiodon (12 to 13) lies within the range for Lycoptera, as
does the count for the anal (15 to 16 rays). In these characters Eohiodon shows a
somewhat closer resemblance to Lycoptera than to Hiodon.

Scales. Cockerell (1925) used scale morphology in Lycoptera as a basis for defining
the family Lycopteridae (as distinct from the Leptolepidae, in which Lycoptera had
been classified previously). Although the scales of Lycoptera are distinctive vis a vis
those of Leptolepis, Cockerell did note their close resemblance to the scales of the
European Minnow Phoxinus phoxinus (thereby initiating the idea that lycopterids
might be cyprinid ancestors).

Lycoptera scales are small, circular or slightly ovoid, with regularly arranged
concentric circuli, with radii present basally and apically (the latter less well-defined
than the former), and with a distinct, relatively large nucleus.

Hiodon scales are not similar. In outline they are almost thumbnail-shaped, with
the anterior margin concave at each corner but convex medially (more so in scales
from the flank than the dorsum) and with the laterobasal corners sharply rounded.
The circuli are not arranged in circles since they bend to follow the outline of the
margin. The nucleus is relatively small. Primary radii are confined to the anterior
field, and are more numerous than in Lycoptera scales.

Little is known about the scales of Eohiodon (see Cavender, I966a, page 317.)

DISCUSSION

SUMMARY OF ANATOMICAL FEATURES. Many characters shared by Lycoptera and
Hiodon are characteristics of primitive teleosts, and thus are of little value in estab-
lishing relationships. Among these are :

1. The arrangement and relationship of bones in the dorsicranium (but see below
regarding the temporal f enestra) .

2. The pattern of the cephalic lateral line system.

3. Presence of teeth on the parasphenoid and the palatopterygoid arch.

4. Mode of articulation between upper jaw elements.

5. Similarities in the morphology of the vertebrae.

The diplospondylous condition of the centra in (presumably) near adult Lycoptera
would seem to be more primitive than the monospondyly of Hiodon. Nelson (1969)
considers the absence of the lower intermuscular bones a specialized feature. Since
these bones are not developed in pholidophorids (Lund, 1966), nor in primitive
teleosts like the leptolepids and ichthyodectids (including Thrissops and Attothrissops) ,
I cannot agree with Nelson's conclusion.

Certain characters are less readily classified as primitive or derived. Among these
perhaps should be included the infraorbital bones. Nelson (1969) suggests that the
six infraorbitals in the Osteoglossomorpha (including Hiodon) probably represent a
specialized condition Lycoptera seems to have the primitive number of seven (see
page 270) but the elongate and slender lachrymal (ist infraorbital) is not typically

ON THE GENUS LYCOPTERA 281

that of the more primitive teleosts, and closely resembles the lachrymal of Hiodon.
In these other fishes the lachrymal is stouter and deeper than in Lycoptera (see for
example the figures of Leptolepis and Thrissops in Patterson [1967], and those of
Flops and Albula in Nelson [1969, text fig. 7]). The infraorbitals of Lycoptera could
well provide the basic type from which the Hiodon pattern evolved by the fusion of
bones 3 and 4 (see Nelson, op. cit.} and the posterior extension of the 5th bone. The
specialized condition of the Ly copter a antorbital is discussed above. The absence
of a supraorbital bone in both Hiodon and Lycoptera is an unusual and probably a
phyletically significant feature.

The reduced number of branchiostegal rays (ca. 8-10) in Lycoptera is certainly an
advanced character. That this number is identical with the number occurring in
Hiodon, and that both genera have the ceratohyal and branchiostegal rays of a
similar shape, can I think be taken as evidence indicating relationship. A review of
branchiostegal and ceratohyal shape in lower teleosts (including fossil taxa where this
bone is preserved) brings to light no other species with branchiostegals and ceratohyal
as similar to those of Lycoptera (personal observations, and McAllister, 1968).
Hiodon has two hypophyals on each side (a primitive feature) but Lycoptera has only
one (see page 272).

There is a marked similarity in the shape of the parasphenoid of Lycoptera and
Hiodon, particularly with respect to the steep angle made between the orbital and
occipital limbs of the bone. Such a sharply angled skull-floor is probably unique
among primitive teleosts. The presence of a basipterygoid process on the para-
sphenoid of Lycoptera is, of course, a primitive feature lost in Hiodon.

Parapophyses fused with the vertebral centra is a specialized characteristic. In
Hiodon the pleural ribs articulate with the centrum or with the parapophysis base
where it is fused with the centrum. Lycoptera, although having fused parapophyses,
retains the more primitive condition of ribs articulating with the tips of the para-
pophyses. It is noteworthy that in both Hiodon and Lycoptera the vertebral organ-
ization in general is primitive, yet the parapophyses are firmly fused with the
centrum.

If my identification of the large otolith in Lycoptera as a lapillus is correct (see page
269), then there is a remarkable degree of similarity between the two genera in a
most unusual specialization. Little is known about this otolith in lower teleosts, but
no living elopomorph, osteoglossomorph or primitive euteleostean has a lapillus as like
that of Lycoptera as is the lapillus of Hiodon.

There remain two characters, the caudal fin skeleton and the lateral temporal
fenestra. These I consider to have great import in establishing a phyletic relationship
between the lycopterids and hiodontids.

Although primitive characters persist in the caudal skeleton of Hiodon and Eohio-
don (as for example a separate second ural centrum and seven hypurals) there are two
other features which can only be considered derived ones. These are the single
epural, and a reduction in the number of principal branched caudal rays to sixteen.
In the primitive caudal fin there are three (rarely two) epurals, and seventeen
branched rays (for further discussion see Greenwood and Patterson [1967], and
Nelson [1969]). Another unusual feature, but one shared with some Mesozoic Ichthyo-

282 P. H. GREENWOOD

dectidae (Patterson, 1968), is the arrangement of the uroneurals into a single series
of elongate elements (see page 277).

In combination, the single uroneural series, the single epural, and the sixteen
caudal rays of Ly copter a are otherwise found only in Hiodon and Eohiodon.

Likewise, a temporal fenestra bordered by the parietal, pterotic and epiotic is
found only in Ly copter a and Hiodon (the condition in Eohiodon is unknown).

TAXONOMIC CONSIDERATIONS. On the basis of their shared specialized characters,
I would conclude that the Lycopteridae are more closely related to the Hiodontidae
than to any other group of teleosts. Several primitive features, lost in the Hiodonti-
dae, are present in the Lycopteridae, which should therefore, be considered the plesio-
morph sister group of the Hiodontidae.

In turn, the lycopterids and hiodontids together constitute a plesiomorph sister
group to the more specialized Notopteridae. These relationships are perhaps best
expressed by uniting the Hiodontidae and Lycopteridae in one superfamily (Hio-
dontoidea ; new taxon), and by creating another superfamily (Notopteroidea) for the
Notopteridae. Both superfamilies can be included within the suborder Notopteroidei
(order Osteoglossiformes) of Greenwood et al, 1966.

These conclusions are at variance with the opinions expressed by previous students
of the Lycopteridae. Of these workers, five suggested (with varying degrees of con-
fidence) relationship with the Cyprinidae (Cockerell [1925] ; Gregory [1933] ; Saito
[1936] ; Takai [1944] ; Liu et al [1963]), three favoured affinity with the basal
Cmpeiformes sensu Berg (1940) (Berg [1948] ; Lehman [1966], Gaudant [1968]) and
one, Yakovlev (1965), with the Osteoglossidae.

Evidence for relationship between the Lycopteridae and the Cyprinidae stems
chiefly from similarity in scale morphology (Cockerell op. cit.} and from Reis' (1910)
identification of the largest otolith as an asteriscus, thus implying a cyprinoid type
of inner ear. However, no other anatomical or gross morphological features can be
adduced to support this postulated relationship, and doubt can be thrown on the
otolith evidence (see above, page 267). In all features generally thought to be im-
portant indicators of phyletic relationship (especially the caudal fin skeleton) the
cyprinids and lycopterids are very distinct, and the same may be said for the other
ostariophysan families (see Rosen and Greenwood, 1970).

A relationship between the Lycopteridae and certain basal teleosts (i.e., the more
primitive members of Berg's [1940] order Clupeiformes, such as the elopoids and
leptolepoids) seems, at first sight, to be more reasonable. But again, apart from
similarities in basic, primitive features the lycopterids differ in many important ways
(caudal skeleton, number of branchiostegals, morphology and relative size of the
otoliths, presence of a temporal fenestra, etc.) To classify the lycopterids with Berg's
Clupeiformes would seem totally to disregard their phyletic affinities.

Yakovlev's (op. cit.} suggested relationship of Ly copter a with the osteoglossid
Arapaima is of particular interest. However, the shared characters which Yakovlev
lists are not critical ones ; most are primitive features found in all basal teleosts (a
medioparietal skull, basipterygoid process on the parasphenoid, teeth on para-
sphenoid and basihyal plate). Of the others, I would doubt whether the nasals are
contiguous in Ly copter a (see page 261), and a large opercular process occurs in other

ON THE GENUS LYCOPTERA 283

and unrelated lineages (see page 272). That Ly copter a is a member of the Osteo-
glossomorpha seems certain, but the nature of its caudal skeleton militates against
its inclusion in the Osteoglossoidei (Greenwood, 1967 ; Nelson, 1969).

ZOOGEOGRAPHICAL IMPLICATIONS. Nelson (1969) has discussed at length evi-
dence relating to the phyletic and geographical history of the Osteoglossomorpha.
Important lacunae in this evidence concern the primary area of distribution, and the
interrelationships, of the Hiodontidae. The lycopterid hiodontid relationship
proposed above does little to fill this gap except to strengthen the hypothesis that the
hiodontid lineage was probably of north-eastern Asiatic origin. The problem of
hiodontid-notopterid interrelations, and other aspects of osteoglossomorph geo-
graphy will be reviewed in a forthcoming paper.

Weiler (in Martin and Weiler, 1954, 1957 and 1965) has referred a number of differ-
ent otoliths, from the Mesozoic of Germany, to the suborder Lycopteroidei (sensu
Berg, 1940). Like other authors, Weiler based his comparisons on Reis' (1910)
figures. Even allowing for the difficulties inherent in this procedure, I cannot see a
close resemblance between the German and Siberian otoliths ; none of the specimens
figured by Weiler shows the characteristic outline and sculpturing of the larger
otolith in Lycoptera. Thus, on the evidence of otoliths alone, it would by unwise to
assume that lycopterid fishes occurred in Europe during the Mesozoic.

ACKNOWLEDGEMENTS

I am indebted to a number of people. As usual to Dr. Colin Patterson for many
discussions on all aspects of the work ; to my colleagues in the American Museum of
Natural History, Drs. Donn E. Rosen and Gareth J. Nelson, for their hospitality,
loan of material and, again, for many profitable discussions ; to Mr. James Chambers
for preparing several alizarin transparencies and for making the radiographs ; to Mr.
Ian Macadie of the Palaeontology Department B.M. (N.H.) for preparing the Lycop-
tera fossils, and for his skill in producing latex casts of these fishes ; to Sharon Lesure
for her patience and skill in making the drawings, often from my own crude sketches ;
and to my assistant Mr. Gordon Howes for his constant help in many ways. To them
all, my best thanks.

I am also greatly indebted to Dr. Frielinde Obergfell of the Bayerische Staatssamm-
lung fur Palaontologie und historische Geologic, Munich for the loan of casts from
that Museum, and for her help in trying to locate Reis' original material.

SUMMARY

A comparative osteological study of the Mesozoic genus Lycoptera and members
of the family Hiodontidae, indicates a close relationship between the two taxa. Of
particular importance in this regard are the nature of the caudal fin skeleton, the
number of caudal fin rays, the presence of a temporal fenestra, and details of the
hyoid arch skeleton.

The peculiar otoliths of Lycoptera seem to resemble those of Hiodon and it is sug-
gested that, contrary to earlier opinions, the inner ear of Lycoptera was probably of
the Hiodon-type and not the cyprinid type.

284 P. H. GREENWOOD

Previous views on the relationships of the Lycopteridae are reviewed. No evi-
dence can be found to support the idea that the Lycopteridae are ancestral cyprinids
or that there is any relationship between the elopoid lineages and the Lycopteridae.

It is suggested that the Lycopteridae and Hiodontidae be placed in one superf amily,
the Hiodontoidea, and that the apomorph sister group of this taxon is the superfamily
Notopteroidea (containing only the Notopteridae).

REFERENCES

BERG, L. S. 1940. Classification of fishes both recent and fossil. Trud. Inst. Acad. Nauk.

S.S.S.R., 5, 2 : 87-517.
1948. On the lower Cretaceous fish Lycoptera (family Lycopteridae). Trud. Inst. Acad.

Nauk. S.S.S.R., 7, 3 : 58-75.
CAVENDER, T. ig66a. Systematic position of the North American Eocene fish " Leuciscus "

rosei Hussakof. Copeia, 1966. 311-320.

19660. The caudal skeleton of the Cretaceous teleosts Xiphactinus, Ichthyodectes, and

Gillicus, and its bearing on their relationship with Chirocentrus. Occ. Pap. Mus. Zool. Univ.
Mich., 650 : 1-15.

COCKERELL, T. D. A. 1925. The affinities of the fish Lycoptera middendorffii. Bull. Am. Mus.

nat. Hist., 51, art. 8 : 313-317.
FROST, G. A. 1925-1930. A comparative study of the otoliths in neopterygian fishes. I-

XXVII. Ann. Mag. nat. Hist., (9 and 10), 15-5 for the series respectively.
GARDINER, B. G. 1963. Certain palaeoniscoid fishes and the evolution of the snout in actino-

pterygians. Bull. Br. Mus. nat. Hist. (Geol), 8, 6 : 255-325.
GAUD ANT, J. 1968. Recherches sur 1'anatomie et la position systematique du genre Lycoptera

(Poisson teleosteen). Mem. Soc. Geol. Fr., 40 : 1-41.
GREENWOOD, P. H. 1963. The swimbladder in African Notopteridae (Pisces) and its bearing

on the taxonomy of the family. Bull. Br. Mus. nat. Hist (Zool), 11, No 5 : 377-412.

1967. The caudal fin skeleton in osteoglossoid fishes. Ann. Mag. nat. Hist., (13) 9 :

581-597.

GREENWOOD, P. H., ROSEN, D. E., WEITZMAN, S. H., & MYERS, G. S. 1966. Phyletic studies

of teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat.

Hist., 131 : 339-456.
GREENWOOD, P. H. & PATTERSON, C. 1967. A fossil osteoglossoid fish from Tanzania (E.

Africa). /. Linn Soc. (Zool), 47 : 211-223.
GREGORY, W. K. 1933. Fish skulls : a study of the evolution of natural mechanisms. Trans.

Amer. phil. Soc., 23 : 75-481.
LEHMAN, J. P. (1966). Actinopterygii. In J. Piveteau, Traite de Paleontologie, 4, fasc. 3.

Paris, Masson et Cie.
Liu, H. T., Su, T-T., HUANG, W-L., & CHANG, K-J. 1963. Lycopterid fishes from north China.

Mem. Inst. vertebr. Palaeont., Palaeoanthrop. Peking, No. 6 : 1-53.
LUND, R. 1966. Intermuscular bones in Pholidophorous bechei from the Lower Lias of England.

Science, N.Y., 152 : 348-349.
MCALLISTER, D. E. 1968. Evolution of branchiostegals and classification of teleostome fishes.

Bull. natn. Mus. Can., Biol., ser., 77 : 1-239
McDowALL, R. 1969. Relationships of galaxioid fishes with a further discussion of salmoni-

form classification. Copeia, 1969 : 796-824.
MARTIN, G. P. R. & WEILER, W. 1954. Fisch-Otolithen aus den deutschen Mesozoikum

(Dogger bis Wealden). Senkeb. Leth., 35 : 119-192.

- 1957. Das Aldorfer Otolithen " Pflaster " und seine Fauna. Senkenb. Leth., 38 :

211-250.
- 1965. Neue Untersuchungen an Fisch-Otolithen aus dem alteren Jura N.W.-Deutsch-

lands. Senckenb. Leth., 46 : 35-72.

ON THE GENUS LYCOPTERA 285

MONOD, T. 1968. Le complexe urophore des poissons te!6osteens. Mem. Inst. fond. Afri.

Noire, No. 81 : 1-705.
NYBELIN, O. 1956. Les canaux sensoriels du museau chez Elops saurus (L). Notice pre-

liminaire. Ark. ZooL, (2), 10 : 453-458.

- 1963. Zur Morphologic und Terminologie des Schwanzskelettes der Actinopterygier.
Ark. ZooL, (2) 15 : 485-516.

1967. Notes on the reduction of the sensory canal system and of the canal bearing bones

in the snout of higher actinopterygian fishes. Ark. ZooL, (2) 19 : 235-246.
NELSON, G. J. 1968. Gill arches of teleostean fishes of the division Osteoglossomorpha. /.
Linn. Soc. (ZooL), 47 : 261-277.

- 1969. Infraorbital bones and their bearing on the phylogeny and geography of osteo-
glossomorph fishes. Am. Mus. Novit., No. 2394 : 1-37.

PATTERSON, C. 1967. Are the teleosts a polyphyletic group? Collogues int. Cent. natn. Rech.
scient., Paris, 163 : 93-109.

- 1968. The caudal skeleton in Lower Liassic pholidophorid fishes. Bull. Br. Mus. nat.
Hist. (Geol.), 16, no. 5 : 201-239.

REIS, O. M. 1910. Die Binnenfauna der Fischschiefer in Transbaikalien. Rech. geol. et min.

long ch. def. Siberie, 29 : 1-82.

RETZIUS, G. 1881. Das Gehororgan der Wirbelthiere, 1. Stockholm, Samson and Wallin.
RIDEWOOD, W. G. 1904. On the cranial osteology of the fishes of the families Mormyridae,

Notopteridae and Hyodontidae. /. Linn. Soc. (ZooL), 29 : 118-217.
ROSEN, D. E. & GREENWOOD, P. H. 1970. The origin of the Weberian apparatus and the

relationships of the ostariophysan and gonorynchiform fishes. Am. Mus. Novit., In press.
SAITO, K. 1936. Mesozoic leptolepid fishes from Jehol and Chientao, Manchuria. Rep.

ist scient. Exped. Manchuokuo, sect. 2. pt. 2 :. 1-23.
TAKAI, F. 1944. A monograph on the lycopterid fishes from the Mesozoic of eastern Asia.

/. Fac. Sci. Tokyo Univ., sect. 2, 6 : 207-270.
TA VERNE, L. 1968. Osteologie du genre Gnathonemus Gill sensu strictu (Pisces Mormyriformes)

Annls. Mus. r. Afr. cent. Ser. 8., no. 170 : 1-91.
1969. Etude ost^ologique des genres Boulengeromyrus Taverne et G6ry, Genyomyrus

Boulenger, Petrocephalus Marcusen (Pisces Mormyriformes). Annls. Mus. r. Afr. Cent.

Ser. 8., no. 174 : 1-85.
YAKOVLEV, V. N. 1965. Systematics of the family Lycopteridae. Paleont. Zh. no. 2 : 80-92.

(translation used, Internat. Geol. Rev. 8, no. i : 71-80, 1966).

P. H. GREENWOOD, D.Sc.

Department of Zoology

BRITISH MUSEUM (NATURAL HISTORY)

CROMWELL ROAD

LONDON, S.W. 7

Printed in Great Britain by

AJden & Mowbray Ltd
at the Alden Press, Oxford

A REDESCRIPTION OF THE

SPECIES OF EUPOLYBOTHRUS

VERHOEFF S.STR.

PRESERVED IN THE BRITISH MUSEUM

(NATURAL HISTORY) AND THE HOPE

DEPARTMENT OF ZOOLOGY, OXFORD

(CHILOPODA, LITHOBIOMORPHA)

E. H. EASON

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 9

LONDON: 1970

A REDESCRIPTION OF THE SPECIES OF
I EUPOLYBOTHRUS VERHOEFF S. STR.

PRESERVED IN THE BRITISH MUSEUM (NATURAL HISTORY)

AND THE HOPE DEPARTMENT OF ZOOLOGY, OXFORD

(CHILOPODA, LITHOBIOMORPHA)

BY

EDWARD HOLT EASON

Pp. 287-310; 12 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 19 No. 9

LONDON: 1970

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is issued
in five series corresponding to the Departments of the
Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four hundred
pages, and will not necessarily be completed within one
calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 19 No. 9 of the Zoological series.
The abbreviated titles of periodicals cited follow those of
the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Zool.).

Trustees of the British Museum (Natural History), 1970

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 21 July, 1970 Price Seventeen Shillings

(0.85)

A REDESCRIPTION OF THE SPECIES OF
EUPOLYBOTHRUS VERHOEFF S. STR.

BY E. H. EASON

SYNOPSIS

Lithobius fasciatus Newport has long been regarded as the senior synonym of Lithobius
grossipes C. L. Koch, a widespread Alpine species. Lithobius litoralis L. Koch, hitherto known
only from its original description, is another previously accepted synonym. The type material
belonging to these three species has been critically examined ; they are fully redescribed and are
shown to be three distinct species of the subgenus Eupolybothrus Verhoeff s.str. Lithobius
fasciatus graecus Verhoeff is redescribed ; it is regarded as a subspecies of Eupolybothrus litoralis.

INTRODUCTION

CHAMBERLIN (1925) designated Lithobius grossipes C. L. Koch as the generotype of
Eupolybothrus Verhoeff (1907) and the nominate subgenus, as emended by Jeekel
(1967), contains this species which was previously known as E. fasciatus (Newport),
and its close relatives. Although the latter comprise a large number of nominal
species, subspecies and varieties, most of these are based on scanty descriptions of
inadequate material and depend for their definition on unstable characters of little
taxonomic importance. Many of these forms will undoubtedly prove to be identical
with one another and the subgenus is in need of revision.

The material belonging to Eupolybothrus s. str. in the British Museum (Natural
History) and the Hope Department of Zoology is referable to three species, E.
fasciatus (Newport), E. grossipes (C. L. Koch) and E. litoralis (L. Koch) and includes
the type specimens of each of them. The conclusions to be drawn as to their identity
are quite different to those previously accepted and it therefore seems desirable,
pending a thorough revision of the subgenus, to describe these three species in
detail from the above material, particularly as they are likely to prove to be the
predominant surface-dwelling species of the subgenus, and indeed of the genus, in
southern Europe and the Balkans.

TAXONOMIC CHARACTERS

In order to separate the three species under review the following features of the
legs are of special importance.

Chaetotaxy

In addition to the general setae irregularly disposed on all the articles there are, on
the ventral aspects of the tarsi and metatarsi, clearly differentiated somewhat
decumbent setae arranged in linear series parallel to the long axis of the article . These
series of setae have been called "pectines" by Crabill (1958) who uses their arrange-
ment on the first 13 legs to differentiate between American species of Eulithobius ;

2QO E. H. EASON

although they are also evident on the first 13 or 14 legs of many of the larger species
of Lithobius they have not attracted much attention from European workers. In the
present species of Eupolybothrus there are two such series on each of the tarsi and
metatarsi of the I3th and more anterior legs, one or two on the I4th metatarsus and
one on the I4th tarsus ; they are not found on the I5th tarsus but on the I5th meta-
tarsus a single series may or may not be present according to species.

These seriate setae are only a little stouter than the general setae but at the distal
end of the ventral aspect of the I4th tarsus and in a similar position on the tarsi and
tibiae of the I3th and more anterior legs are usually found stout spinous setae, easily
mistaken for small spines in the positions VaTa, VpTa and VpT ; those on the I4th
tarsus differ in arrangement in each species. Since the general setae of the I4th and
1 5th legs also show marked interspecific variation there are three distinct kinds of
setae, general, seriate and spinous, all of which are useful in separating the species.
Finally, mature males usually bear a distinctive tuft of setae on the internal basal angle
of the I5th femur.

Sculpturing of male i$th legs

Although there are one or two longitudinal dorsal sulci on the prefemur, femur and
tibia of each of the I4th and I5th legs in both sexes, only the sculpturing of the adult
male I5th femur is really distinctive. On this article much of the dorsum of the base is
occupied by a pit, continuous with the internal dorsal sulcus. On the internal aspect
of the distal end of the article is an area of greater or lesser extent which is clearly
demarcated from the rest of the shaft by being free of glandular pores and of a rather
paler colour ; this is the "pore-free area" of Chamberlin (1952 : 212) and may bear
special features such as a swelling, a concentration of setae or a small circumscribed
group of very fine pores corresponding to the " fine pore-sieve " of Chamberlin (1952 :
214).

REDESCRIPTIONS

Eupolybothrus (E.) fasciatus (Newport)
Figs, i to 3

Lithiobius fasciatus Newport, 1844, p. 365.

Lithobius (Polybothrus) fasciatus bosniensis: Verhoeff, 1900, p. 162.

not Lithobius fasciatus Muralewitsch, 1929, p. 102.

MATERIAL EXAMINED. Three dried specimens in a fair state of preservation, all
males, are present in the Newport Collection of Myriapoda in the Hope Department
of Zoology, Oxford. I have designated these specimens A, B and C. Specimen A, 23
mm long, is pinned through the 5th tergite and lacks the I4th and I5th legs ; it bears a
label " Lithobius fasciatus Newp. " but no indication of locality. Specimens B and C,
though unlabelled, are arranged in the cabinet immediately on either side of specimen
A. Specimen B, 23 mm long, is pinned through the 7th tergite and has the I4th and
I5th legs intact on the right side only. Specimen C, 22 mm long, is pinned through
the 6th tergite and lacks the I4th and I5th legs.

Although much of the original description of this species is based on specimens up

SPECIES OF EUPOLYBOTHRUS

291

0-5 mm

ventral external
spinous seta

distal seriate
seta

0-25 mm

ventral internal
spinous seta

0-25 mm

FIGS. I to 3. Eupolybothrus fasciatus. i. Femur and distal part of prefemur of I5th leg
(dorsal view). 2. Metatarsus of I5th leg (external view). 3. Region of tarso-meta-
tarsal articulation of I4th leg to show seriate and spinous setae (ventral view) . All figures
taken from Newport's specimen B (t? 23 mm. long)

B

292 E. H. EASON

to 32 mm long with 9+9 prosternal teeth, Newport (1844) states that his material
was deposited in the Hope Museum and careful comparison of the writing on the
label attached to specimen A with an example of Newport's handwriting in the Sher-
borne Collection of Manuscripts in the British Museum (Natural History) leaves
little doubt that the label is, in fact, in Newport's hand. These three specimens,
therefore, seem to be the only survivors of the syntypical series. Specimen A,
bearing the label, should be regarded as the lectotype but we must depend for the
characters of the I4th and I5th legs on specimen B.

The following specimens, preserved in spirit, are present in the Koch Collection of
Arachnida and Myriapoda in the British Museum (Natural History) and are labelled
"grossipes, Rom" in L. Koch's hand (Reg. no. 13. 6. 18. 301-302) :

A male 22 mm long lacking the I4th and I5th legs and a female 23 mm long.

The following specimens, preserved in spirit, are present in the Verhoeff Collection
of Myriapoda in the British Museum (Natural History) and are labelled "Lithobius
fasciatus bosniensis Latz. " "Bosnia" (Reg. no. 03. 8. 25. 41-46) :

A well-preserved male 27 mm long, a mutilated male 24 mm long lacking most of
the legs and four females 23 to 26 mm long.

It seems reasonable to assume that these specimens were among those on which
Verhoeff (1900 : 162) based his records of L. fasciatus bosniensis from Bosnia (Yugo-
slavia) .

DIAGNOSIS OF ADULT. Length 22 to 32 mm. Antennae of 36 to 42 articles. Glan-
dular pores of 1 5th leg concentrated on internal aspects of femur, tibia, tarsus and
metatarsus only. I5th metatarsal general setae up to three-quarters the diameter of
the article in length. I5th metatarsal seriate setae present. Basal pit of male I5th
femur extensive and deep. Internal dorsal sulcus of male I5th femur not extending to
margin of pore-free area which bears a prominent globular swelling. No coxolateral
spines.

DESCRIPTION OF ADULT. Length : present specimens from 22 to 27 mm (Newport
gives i inches = 32 mm) ; I5th legs about half body-length. Colour : Newport's
dried specimens vary from pale (specimen A) to dark (specimens B and C) brown ;
those preserved in spirit are intermediate in colour with a trace of a darker dorsal
median band. Antennae : about half body-length ; of 36 to 42 articles, the distal 12
or so often elongate. Ocelli: 1+4, 5, 4, i ; 1+4, 3, 4, 3 ; i+4, 4, 5, 4, 3 etc.; those
of the superior row horizontally oval, widely spaced and larger than those of the
other rows ; posterior superior ocellus almost as large as isolated posterior ocellus.
Prosternum : with 6+6 to 9+9 small teeth : lateral spines variable in structure and
position, usually peg-like but sometimes dentiform or setiform and often of different
structure on either side of the same individual, usually placed well lateral to the
external teeth but sometimes immediately postero-lateral or rarely postero-medial.
A specimen from Rome has 10+7 teeth, possibly due to abnormal development.
Tergites : posterior borders of large tergites slightly emarginate ; posterior angles of
T.5 rounded, those of T.8 blunt without angulation, those of T.io somewhat angu-
lated, those of T.I2 and 14 distinctly angulated but without projections ; posterior

SPECIES OF EUPOLYBOTHRUS 293

angles of 1.4 rounded without trace of projections ; posterior projections on T.6
rounded, those on T.7 short and broad, those on T.Q, n and 13 well-developed with
internal borders very slightly sinuous ; posterior border of intermediate tergite
sinuous in male, with rounded or trapezoidal emargination in female. Coxal pores :
from 15 in smaller specimens to 35 in larger specimens, in about four rows on each of
the I2th to I5th coxae. Glandular pores of i^th legs (Fig. i) : concentrated on in-
ternal aspects only of femur, tibia, tarsus and metatarsus. On the femur the pores
are limited by the internal dorsal sulcus ; on the pref emur they are sparse as figured by
Verhoeff (1937 : 172, fig. i) for Polybothrus apenninigenus.

Chaetotaxy of i^th and i$th legs

General setae : on the metatarsus numerous, up to three-quarters the diameter of
the article in length (Fig. 2) ; of much the same density and absolute length on the
tarsus, about a quarter the diameter of the article in length ; on the tibia and femur
(Fig. i) sparser and rather shorter ; on the prefemur rather denser, longer and
stouter than on tibia or femur, particularly along the ventral aspect of the article.
Seriate setae : a single row on the distal half to three-quarters of the I5th metatarsus
(Fig. 2), a well-defined external row and a rather sparse internal row along the
length of the I4th metatarsus and a single row on the distal half to three-quarters of
the I4th tarsus (Fig. 3). Spinous setae : on the I4th tarsus the distal seriate seta and
the ventral external seta (VaTa) are of much the same size, rather stouter than the
other seriate setae, while the ventral internal seta (VpTa) is much stouter (Fig. 3).
Setae of tuft (males) : long and numerous in Newport's specimen B (Fig. i) ; absent in
Verhoeff's specimen.

Sculpturing of male i$th legs (Fig. i)

These characters are taken from the right leg of Newport's specimen B and both
legs of Verhoeff's larger male : prefemur with internal dorsal sulcus distinct on right
leg of Verhoeff's specimen only, otherwise faint, no external sulcus ; basal femoral
pit deep and extensive, occupying about three-quarters the diameter of the base of
femur, rather sharply attenuated distally, continuous with the fairly broad internal
femoral sulcus which becomes gradually shallower distally to disappear before
reaching the margin of the pore-free area ; external femoral sulcus absent ; pore-free
area occupying distal one third of the internal aspect of femur, bearing a prominent
globular swelling on which the general setae are of the same structure as those on the
rest of the shaft but a little denser ; on the right leg of Verhoeff's specimen is a well-
circumscribed slightly raised oblong fine pore-sieve at the centre of the globular
swelling about a quarter the diameter of the distal end of the femur, bearing a rather
ill-defined tuft of minute setae ; on the left leg of the same specimen only a small
faint group of fine pores can be seen on the surface of the globular swelling whereas
in Newport's specimen B this character was not confirmed, possibly being obscured
by the slight shrivelling of the integument which tends to occur in dried material.

Sculpturing of male i^th legs

Internal and external dorsal sulci distinct on both prefemur and femur.

294 E - H - EASON

Spinulation

Ventral Dorsal

CtPFT CtPFT

14 m amp amp am a amp a-p a-p

15 m amp amp am a amp p p

15 VpF may be absent ; no coxolateral spines ; a well-developed i5th accessory
apical claw.

Genitalia

Male : posterior border of genital sternite with a median notch and marginal setae
showing no obvious differentiation ; gonopods long and slender, basal article less than
half the length of distal article. Female : two cylindro-conical spurs on each gono-
pod, separated from one another at their insertion by rather more than their own
diameter ; claw of gonopod sharp, without denticles ; dorso-lateral setae of gonopod
short, in an irregular band of about five setae on the first article, about twelve on the
second and none on the terminal article.

DISCUSSION. Latzel (1880) suspected Lithobins fasciatus Newport of being the
same as Lithobius grossipes C. L. Koch and in his paper on Ligurian centipedes
Pocock (1890), after examining the type material of L. fasciatus in the Hope Museum,
says definitely that these two species are identical. Ever since, L. fasciatus has been
regarded as the senior synonym of L. grossipes. But it is now clear that Newport's
specimens are distinct, not only from the holotype of L. grossipes and most of the
other specimens in the Koch Collection assigned by L. Koch to this species, but also
from the Ligurian specimens on which Pocock presumably based his conclusion (see
p. 297).

Newport (1844) gives Florence and Naples as the type localities of L. fasciatus,
mentioning that specimens from Naples are usually of a much lighter colour than
those from Florence. If we assume that specimen A, which is markedly paler than
either B or C, comes from Naples and that specimens B and C come from Florence
it is just possible, since the two species are very hard to separate without reference
to the structure of their I5th legs, that specimen A belongs to Eupolybothrus grossipes
and that Pocock examined it before its I5th legs were lost, omitted to examine speci-
mens B and C, and came to the correct conclusion. But Newport states that speci-
mens from the two localities agree with one another precisely except in their colour
and he describes the " metatarsal joints " of the posterior legs as " very hairy " which
he would hardly have done had his specimens included mature examples of E.
grossipes. It therefore seems reasonable to regard Newport's three specimens as
conspecific.

The difference between these specimens and those referred by Verhoeff to L. fascia-
tus bosniensis seems only to be due to individual variation. Colour is altogether un-
reliable as a taxonomic character and, in any case, dried specimens and those pre-
served in spirit are not comparable with one another in this respect. The prominent
tuft of setae on the internal basal angle of the male I5th femur, absent in Verhoeff' s
specimen, was noted by Latzel (1888) in his original description of bosniensis but

SPECIES OF EUPOLYBOTHRUS 295

Attems (1902) remarks on its absence in some specimens. It seems that these setae,
which are absent from all but the fully mature examples of E. grossipes and . litoralis
that I have examined, may fail to appear even in the adult.

There is therefore little doubt that Verhoeff's specimens are indeed identical with
L. grossipes var. bosniensis Latzel (1888) and this form, originally recorded from a
number of localities in Bosnia and Hercegovina, will almost certainly prove to be
identical with E. fasciatus and was only named because Latzel was unfamiliar with
Newport's species. The same may be said of L. fasciatus var. apenninigenus Brole-
mann (1894), originally recorded from Borgotaro in the northern Apennines, which
agrees with Latzel's rather scanty original description and only differs from Attems
(1902) expanded redescription of bosniensis in quite unimportant characters. It is
also safe to assume that Verhoeff (1934 and 1937) had specimens of E. fasciatus before
him when he wrote his descriptions of Polybothrus apenninigenus.

The labelling of specimens of E. fasciatus as "grossipes" by L. Koch is hardly
surprising as the characters separating these two species are unlikely to have
attracted his attention.

Eupolybothrus (E.) grossipes (C. L. Koch)
Figs. 4 to 7

Lithobius grossipes C. L. Koch, 1847, p. 146; L. Koch, 1862, p. 32, fig. 4.
Lithobius montanus C. L. Koch, 1847, p. 148; L. Koch, 1862, p. 27, fig. i.
Lithobius festivus L. Koch, 1862, p. 29, fig. 2.
Lithobius (Eulithobius] fasciatus: Pocock, 1890, p. 61.

MATERIAL EXAMINED. The following specimens, preserved in spirit, are present in
the Koch Collection of Arachnida and Myriapoda in the British Museum (Natural
History) ; except where otherwise stated the labels appear to be in L. Koch's hand :

"Lithobius grossipes C. Koch, festivus L. Koch" "Idria" (Reg. no. 13.6.18. 262).

This specimen, a female 30 mm long, although preserved in spirit has obviously
been dried. An unusual feature is that the i4th leg appears stouter than the i5th ;
this appearance was noted, not only by L. Koch (1862) who undoubtedly used this
specimen for his description of L. grossipes, but also by C. L. Koch (1847) in his
original description. I therefore believe it to be the holotype and have labelled it
accordingly. The fact that C. L. Koch gives Trieste as the type locality is consistent
with the same specimen having been used for both descriptions since Idria (Idrija,
Yugoslavia) is in the neighbourhood of Trieste.

"Lithobius grossipes C. K." "Hopfgarten, 26 August, 1868" (Reg. no. 13.6.18.
263-265).

Three females 28 to 30 mm long.

" Lithobius grossipes C.K. " " Seiseralpe [an alpine hut in Italy], *[leg.] Gredler "
(Reg. no. 13.6.18. 266).

This specimen, a male 35 mm long, agrees in detail with L. Koch's (1862) descrip-
tion of L. montanus C. L. Koch, a description based on a single specimen sent him

* Words in brackets throughout this paper are the author's additions to the original labels.

2Q6

E. H. EASON

O'Smm

l-Omm

0'5 mm

I'Omm

FIGS. 4 to 7. Eupolybothrus grossipes. 4. Femur and distal part of prefemur of i5th leg
(dorsal view). 5. Metatarsus of i5th leg (external view). 6. Region of tarso-meta-
tarsal articulation of i4th leg to show seriate and spinous setae (ventral view) . 7. nth
to i4th tergites (dorsal view). Figs. 4 to 6 taken from a <$ from Busalla (43 mm long) ;
fig. 7 taken from a stadium agenitalis i from Busalla (10 mm long).

by Prof. P. Gredler from "Seiseralpe" in South Tyrol. This is undoubtedly the
specimen in question and was presumably relabelled by L. Koch when he realised its
identity with L. grossipes, of which L. montanus is now an accepted synonym.

"grossipes, Eisarkthal [Valle dell'Isarco, Italy] 1870" (Reg. no. 13. 6.18. 267-269).

Two females and a male 30 to 31 mm long.

SPECIES OF EUPOLYBOTHRUS 297

" Lithobius grossipes C.K., festivus L.K. " " Sudtirol " (Reg. no. 13.6.18. 270-282,
part).

Two females 35 and 37 mm long, a mature male 34 mm long and two immature
females 17 and 21 mm long.

"grossipes, Sudtirol" (Reg. no. 13.6.18. 270-282, part).

A single female 37 mm long.

" Lithobius grossipes C.K., festivus L.K. " " Sudtirol " (Reg. no. 13.6.18. 270-282,
part).

Four mature females 28 to 31 mm long and three immature females 16 to 21 mm
long.

" Lithobius grossipes C.K., festivus L.K. " " Sudtirol " (Reg. no. 13.6.18. 270-282,
part).

Three mature males 28 mm long and two immature males 12 and 14 mm long.

"Lithobius grossipes C. Koch." "Meran [Merano, Italy], [leg.] Milde" (Reg. no.
13.6.18. 283-292, part).

Three females 27 to 30 mm long, six mature males 27 to 38 mm long and an
immature male 19 mm long.

"grossipes, Meran [Merano, Italy], [leg.] Ausserer" (Reg. no. 13.6.18. 283-292,
part).

An immature male 23 mm long.

"Lithobius grossipes C.K. " "Garmisch" (Reg. no. 13.6.18. 293).

An immature male 24 mm long. Although not agreeing in every detail with L.
Koch's (1862) original description of L. festivus from Garmisch (Bavaria) this may be
one of the specimens on which it was based.

"grossipes, Ballino [Italy] 12. IX. 69" (Reg. no. 13.6.18. 298).

An immature female 17 mm long. The identity of this specimen is doubtful ; it
may possibly belong to E. fasciatus.

"L. grossipes." "Susa, Oberitalien" (Reg. no. 13.6.18. 299-300).

Two immature males 18 and 24 mm long. The locality label is in L. Koch's hand
but the identity label has been rewritten.

"leptopus? Hopfgarten" (Reg. no. 13.6.18. 365).

An immature female 13 mm long.

The following specimens are in the British Museum (Natural History), preserved in
spirit in a jar labelled "Lithobius fasciatus Newp. Liguria, O. Thomas" in Pocock's
hand (Reg. no. 89.3.8. 1-2) :

A total of 31 specimens ranging in size from fourth larval stadia 9 mm long to a
mature male 43 mm long. These specimens almost certainly come from Busalla
(Liguria, Italy) both because Pocock (1890) mentions Busalla as the locality and Mr.
Oldfield Thomas as the collector of examples of " L. fasciatus " in his paper on Ligurian
centipedes, and because they were mixed with two specimens of Lithobius doriae
Pocock ; in the same paper this author describes L. doriae as a new species not only
from the same locality (Busalla) but also from the same habitat as " L. fasciatus"
(Pocock, 1890 : 59 & 64) and he must mistakenly have included specimens of the two
species under the same label. I have relabelled these two examples of L. doriae and
placed them in a separate tube.

298 E. H. EASON

The following specimens were contained in the above jar along with the Ligurian
specimens but in separate tubes :

A badly mutilated " stadium agenitalis " 12 mm long labelled " Eulith. grossipes,
Grande Chartreuse [Isere, France], [leg.] A. Dollfus " (Reg. no. 89.3.10. i, part) and a
badly mutilated immature male 15 mm long labelled " L. grossipes Koch, Portofino
[Italy] " (Reg. no. 89.3.10. i, part). Both labels appear to be in Pocock's hand ; in
neither case is it possible to identify the specimens other than as species of the
subgenus Eupolybothrus.

DIAGNOSIS OF ADULT. Length 27 to 43 mm. Antennae of 40 to 58 articles.
Glandular pores of I5th leg concentrated on internal and dorsal aspects of prefemur
and all aspects of other articles. I5th metatarsal general setae rarely exceeding a
quarter the diameter of the article in length. I5th metatarsal seriate setae present.
Basal pit of male I5th femur small and shallow. Internal dorsal sulcus of male I5th
femur extending to margin of pore-free area which is not swollen. No coxolateral
spines.

DESCRIPTION OF ADULT. Length : 27 to 43 mm ; i5th legs up to two-thirds of body-
length. Colour : pale to dark brown, often with a darker dorsal median band.
Antennae : half to two-thirds of body-length, sometimes shorter in female ; of 40 to
50 (female) and 47 to 58 (male) articles, the distal 10 to 14 often very elongate in large
specimens. Ocelli : number and general pattern as described for E. fasciatus ; in
some specimens, however, the ocelli of the superior row may be round rather than
oval and little larger than the others, but they are always relatively widely separated
from one another. There is a tendency for the ocelli of the second, third and fourth
rows to be more numerous than those of the superior row but this is by no means in-
variable. Prosternum : with 6+7 to 8+9 small teeth, usually 7+7 or 8+8 ; minute
setiform lateral spines immediately postero-lateral to the external teeth in some
specimens, but these spines are often absent from one of both sides, particularly in
large specimens, possibly due to damage. Tergites : emargination of posterior bor-
ders of large tergites variable, particularly that of T.I4 which may be quite markedly
emarginate or may be almost straight ; posterior angles of T-5 rounded or blunt,
those of T.8 blunt or angulated sometimes with traces of projections, those of T.io
blunt or angulated often with slight projections, those of T.I2 sometimes blunt but
usually angulated with slight projections, those of T.I4 blunt or angulated, sometimes
rounded ; posterior angles of T.4 usually rounded, sometimes very slightly projecting.
Posterior projections on T.6 always well-developed, usually blunt, sometimes sharp ;
those on T.7 well-developed, sometimes very broad and short, with or without some
sinuosity of their internal borders ; those on T.g, n and 13 well-developed with
internal borders more or less sinuous ; in some specimens this sinuosity is so marked,
particularly on T.n and 13, that the tip of the projection is narrow and pointed as
in the figured immature female of E. litoralis (Fig. 12). Posterior border of inter-
mediate tergite sinuous in male, with rounded or trapezoidal emargination in female.
Coxal pores : 35 to 65 in four to six irregular rows on each of the I2th to I5th coxae,
most numerous on I4th. Glandular pores of i$th legs (Fig. 4) : concentrated on all
aspects of femur, tibia, tarsus and metatarsus, and also on internal and dorsal

SPECIES OF EUPOLYBOTHRUS 299

aspects of prefemur as figured by Verhoeff (1937 : 172, fig. 2) for " Polybothrus
fasciatus ".

Chaetotaxy of i^th and i$th legs

General setae : on the metatarsus sparse and short, not more than a quarter the
diameter of the article in length, usually much less in males (Fig. 5), sometimes a little
longer in females ; on the tarsus, tibia, femur (Fig. 4) and prefemur shorter and
sparser than on metatarsus, sometimes a little longer in females than in males.
Seriate setae (Figs. 5 & 6) : as in E. fasciatus, but in females those of the I5th meta-
tarsus may be reduced to about three at the distal extremity of the article. Spinous
setae : on the i/j-th tarsus the ventral external and ventral internal setae are of much
the same structure (Fig. 6) and correspond to the " spines " VaTa and VpTa des-
cribed by Brolemann (1930 : 246) in " Bothropolys fasciatus " ; in addition to the
ventral internal seta there are, in large specimens, a few more spinous setae at intervals
along the ventro-internal border of the tarsus. On the I3th and more anterior legs the
spinous tarsal setae are similar to those on the i4th but rather more numerous along
the ventro-internal border ; the tibia of each of these legs also bears a linear series of
spinous ventro-internal setae, the most distal of which corresponds to the " spine "
VpT of Brolemann. Setae of tuft (males) : short and few, but present in all adults
examined (Fig. 4)

Sculpturing of male i^th legs (Fig. 4)

Prefemur with dorsal sulci distinct, internal one continuous with basal femoral pit,
external one not reaching the distal end of article ; basal femoral pit shallow,
occupying about a quarter to a third the diameter of the base of femur, continuous
distally with the narrow internal femoral sulcus which runs to the margin of the
pore-free area and is of the same width and depth throughout ; external femoral sul-
cus similar in width, starting level with the basal pit and extending further distally
than the internal sulcus ; pore-free area occupying distal one third or less of the
internal aspect of femur, with no swelling, almost glabrous, the minute setae on its
surface being just as sparse as on the rest of shaft.

Sculpturing of male i^th legs

Internal and external dorsal sulci usually fairly distinct on both prefemur and
femur.

The i^th leg of holotype

The femur (R. only) is unusually broad so that the I4th leg appears stouter than
the T_5th ; this appearance is not due to defective development of the i5th leg as
suggested by Latzel (1880 : 48).
Spinulation

Ventral Dorsal

CtPFT CtPFT

14 m amp amp am a amp a-p a-p

15 m amp am am a amp p p

14 VpF and 15 DpT may be absent ; 15 VpF and 15 DaF may be present ; no coxo-
lateral spines ; a well-developed I5th accessory apical claw.

300 E. H. EASON

Genitalia.

Male : posterior border of genital sternite with a median notch, more distinct in
large specimens, and long marginal setae on either side, the medial setae adjacent to
the notch being often shorter than the more lateral setae but showing no sharp
differentiation ; gonopods long and slender, basal article less than half the length of
distal article. Female : two cylindro-conical spurs on each gonopod separated from
one another at their insertion by about their own diameter, the internal pair being
often rather smaller than the external pair even in mature specimens ; claw of gono-
pod sharp, without denticles ; dorso-lateral setae of gonopod short and stout, in an
irregular band of about six setae on the first article, about twelve on the second and
none on the terminal article. In large specimens the spurs of the gonopods may be
short and blunt and the claw blunt, possibily due to wear and tear.

IMMATURE STADIA. There is sufficient available material to describe the last larval
stadium and five further stadia which probably cover the complete post-larval life-
history. It is difficult to make an exact comparison between each of these stadia
and those described by Verhoeff (1905) for Lithobius forficatus ; Verhoeffs terms are
only roughly applicable and are therefore placed in parenthesis.

Fourth larval stadium

Length : about 9 mm. Antennae : 19 to 24 articles. Prosternal teeth : usually
6+6, sometimes 6+5 or 6+4. Tergites : posterior border of last tergite (T.I2)
deeply emarginate ; posterior projection on T.6, 7, 9 and n well-developed. Coxal
Pores : one on I2th coxa.

First post-larval stadium (agenitalis i)

Length : about 10 mm. Antennae : broken. Ocelli : 1+2, i. Prosternal teeth :
6+6 or 6+7. Tergites (Fig. 7) : posterior borders of large tergites deeply emargin-
ate ; posterior projections on short tergites well-developed and much narrower than
in adults. Coxal pores : 2, 2, 2, i. i^th and i$th legs : missing.

Second post-larval stadium (agenitalis 2)

Length : about 12 mm. Antennae : broken. Ocelli : 1+2, 2. Prosternal teeth :
6+6 or 6+7. Tergites : as in last stadium. Coxal pores : one relatively large
external pore and 4 to 6 much smaller pores on each of the I2th to I5th coxae.
and i$th legs : missing. Genitalia : undeveloped.

Third post-larval stadium (immaturus)

Length : 12 to 14 mm. Antennae : 38 articles. Ocelli : 1+3, 3, 2. Prosternal
teeth : 6+6 to 7+7. Tergites : posterior borders of large tergites as in last stadium ;
posterior projections on short tergites rather less narrow ; posterior border of inter-
mediate tergite straight. Coxal pores : 7 to 10 on each of the I2th to I5th coxae.
General setae of i^th leg : those of metatarsus exceeding the diameter of the article in
length. Seriate setae of i^th leg : as in adult. Spinous setae of i^th leg : not fully

SPECIES OF EUPOLYBOTHRUS 301

developed. Spinulation of i^th legs : as in adult. i$th legs : missing. Male
genitalia : posterior border of genital sternite without a notch, with or without one or
two marginal setae on either side ; gonopods appears as unsegmented slender buds.
Female genitalia : gonopods small with indefinite segmentation and neither spurs
nor claw.

Fourth post-larval stadium (praematurus)

Length : 15 to 18 mm. Antennae : 39 or 40 articles. Ocelli : 1+3, 3, 3 or 1+3,
3, 2, 2. Prosternal teeth 6+6 to 7+7. Tergites : approaching the shape found in
adults, but posterior borders of large tergites tend to be more deeply emarginate and
posterior projections on short tergites tend to be narrower and sharper ; posterior
border of intermediate tergite straight or slightly emarginate in either sex. Coxal
pores : 10 to 19 on each of the I2th to I5th coxae, usually many more on I2th than on
I5th. Glandular pores of i$th legs : as in adult. General setae of i^th and i^th legs :
much longer than in adult, those of metatarsus being about as long as the diameter of
the article. Seriate setae of 14^ and i$th legs : those of I4th tarsus and metatarsus
as in adult ; those of I5th metatarsus reduced to a few at the distal extremity of the
article. Spinous setae of 14^ leg : not fully developed. Setae of tuft (male) : absent.
Sculpturing of i$th legs : indistinct in males as in females. Spinulation of i^th and
i$th legs : as in adult. Male genitalia : posterior border of genital sternite with or
without a trace of median notch, with about 3 or 4 marginal setae on either side ; basal
article of gonopod as long as distal article. Female genitalia : gonopods fairly well-
developed, completely segmented, with or without minute spurs, with a small
claw.

Fifth post-larval stadium (pseudomaturus)

Length : 19 to 25 mm. Antennae : 40 to 49 articles. Ocelli : i +3, 4, 3 to i + 4, 4, 3, 2.
Prosternal teeth : 7+7 to 8+9. Tergites : as in last stadium but posterior border of
intermediate tergite very slightly sinuous in male, with rounded or trapezoidal
emargination in female. Coxal pores : 20 to 35 on each of the I2th to I5th coxae.
General setae of i^th and i$th legs : those of metatarsus up to half and those of
tarsus up to one third of the diameter of the article in length ; others relatively
shorter than in last stadium but longer than in adult. Seriate setae of i^th and i$th
legs : as in adult but sometimes slightly reduced on I5th metatarsus. Spinous setae
of i^th leg : as in adult, or the ventral external seta more slender. Setae of tuft
(male) : absent. Sculpturing of i$th legs : femoral sulci more distinct in males than
in females, but basal femoral pit and pore-free area absent or ill-defined. Male
genitalia : posterior border of genital sternite usually with a feeble median notch
and about 6 to 8 marginal setae on either side ; basal article of gonopod about
half the length of distal article. Female genitalia : gonopods with small unequal
spurs and a well-developed claw.

DISCUSSION. This is the species which, ever since the publication of Pocock's
(1890) synonymy (see p. 294), has been regarded by most authors as the typical form of
Lithiobius fasciatus Newport. But Chamberlain (1925) designated L. grossipes C. L.
Koch as the type species of Eupolybothrus and only he, among modern authors, uses

302 E. H. EASON

C. L. Koch's name either because he was not familiar with the current European
literature or because he regarded Newport's description of L.fasciatus as inadequate.

It seems that when L. Koch redescribed L. grossipes in 1862 he had only the holo-
type before him and identified his other available specimens of the species either as
L. montanus C. L. Koch or as a new species, L. festivus, based on immature
examples. Most of the material in the Koch Collection was probably named subse-
quent to 1862 when L. Koch must have realised the identity of these three forms with
one another and labelled them "L. grossipes" of "L. grossipes, festivus." His reason
for the continued use of the name " festivus " on some of the labels is quite obscure
and was not based on immaturity.

E. grossipes has been fairly adequately described under either grossipes or " fas-
ciatus " by a number of authors. Verhoeff (1937 : 178) describes T.8, 10, 12 and 14
as having rounded posterior angles and T.g, n and 13 as having posterior projections
with almost rectilinear borders. Whereas a few of Koch's specimens have rounded
angles on the large tergites as figured by Verhoeff (1937 : 174, fig. 7, oben) for "fas-
ciatus ", the majority of Koch's and all Pocock's specimens have sharp or even
slightly produced angles, particularly on T.io and 12, much as figured by Verhoeff
(1937 : 174, fig. 7, unten) for Polybothrus baldensis. The posterior projections on the
short tergites are more often as described and figured by Verhoeff (1937 : 173, fig. 6)
for " fasciatus " but there is considerable variation and those of T.n and 13 may
have their internal borders so sinuous and emarginate, and their extremities so narrow,
that they resemble the figure of E. litoralis (Fig. 12) or even Verhoeff's (1937 : 173,
fig. 4) figure of T.I3 in P. baldensis. The exact shape of the trunk tergites is so
variable that it is of little use as a taxonomic character, and the above description as
well as the figures of E. litoralis (Figs. 10 & 12) might apply equally to each of the
three species under consideration.

But this tendancy of the tergites to differ from the shape described by Verhoeff
for "fasciatus " does occur more often in immature specimens both as regards the
narrowing of the projections on the short tergites and the sharper angulation of the
large ones ; sharper angulation of the latter is often associated with increased con-
cavity or emargination of their posterior borders as described by Verhoeff (1934 : 72)
for Polybothrus fasciatus albanicus. Verhoeff's statement, therefore, when describing
albanicus, that tergal projections are weaker (blunter) in immature than in mature
specimens is not true of E. grossipes although it may be true of many species of Litho-
biidae. These variations are well shown by Pocock's specimens, the majority of
which are immature. For example, angulation and emargination of T.I4 is most
marked in the earliest post-larval stadium (Fig. 7) and the last tergite (T.I2) of the
fourth larval stadium is even more deeply emarginate.

Although in mature males the seriate setae of the i5th metatarsus are extensive
(Fig. 5) and the general setae of this and other articles of the I4th and I5th legs are
very short, in immature specimens (and sometimes in mature females) the seriate
setae tend to be reduced in number and the general setae are relatively longer. There
actually seems to be an inverse relationship between the length of the general setae
relative to the diameter of the article on which they are borne and the degree of
maturity of the specimen.

SPECIES OF EUPOLYBOTHRUS 303

We thus have two sets of characters, the shape of the tergites and the length and
arrangement of the setae on the legs, particularly those of the I5th metatarsus,
which are dependant to a very great extent on degree of maturity. The validity of
such species and subspecies as P. baldensis Verhoeff (1937), P. fasciatus albanicus
Verhoeff (1934) and P. fosciatus storkani Verhoeff (1934), all depending on one or
both of these characters for their definition, is therefore open to question.

Eupolybothrus (E.) litoralis (L. Koch)

The material which can undoubtedly be assigned to the typical form of E. litoralis
is certainly conspecific with the specimens Verhoeff has labelled " Lithobius fasciatus
graecus", but since the latter may represent a geographical race they should be
regarded as belonging to a distinct subspecies unless it can be shown that their dis-
tinctive character is altogether unstable.

Eupolybothrus (E.) litoralis litoralis (L. Koch)
Figs. 8 to 12

Lithobius litoralis L. Koch, 1867, p. 899.

? Lithiobius (Polybothrus) fasciatus graecus var. fasciatograecus Verhoeff, 1901, p. 347.

not Lithobius litoralis Muralewitsch, 1906, p. 67.

MATERIAL EXAMINED. The following specimens, preserved in spirit, are present
in the Koch Collection of Arachnida and Myriapoda in the British Museum (Natural
History) ; except where otherwise stated, the labels appear to be in L. Koch's hand :

" L. litoralis " " Tinos Erber " (Reg. no. 13. 6.18. 368).

A single immature female 24 mm long. The locality label appears to be in L.
Koch's hand but the identity label has been rewritten. As L. Koch's original
description was based on a single immature female from Tinos (Aegean archipelago)
I have no doubt that this specimen is the holotype and have labelled it accordingly.

" Lithobius grossipes C.K. " " Tinos Erber " (Reg. no. 13.6.18. 297).

A single male 38 mm long.

" Lithobius grossipes C.K. " " Rhodus Erber " (Reg. no. 13.6.18. 294-296).

A mature male 28 mm long, an immature male 18 mm long and an immature
female 23 mm long.

DIAGNOSIS OF ADULT. Length up to 38 mm or more. Antennae up to 57 articles or
more. Glandular pores of I5th leg concentrated on internal aspects of femur, tibia,
tarsus and metatarsus only. I5th metatarsal general setae up to three-quarters the
diameter of the article in length. I5th metatarsal seriate setae absent. Basal pit
of male I5th femur extensive and deep. Internal dorsal sulcus of male I5th femur
not extending to margin of pore-free area which is barely swollen. Coxolateral spines
usually absent.

DESCRIPTION OF ADULT. Two males only. Length : 28 and 38 mm ; i5th legs two-
thirds of body-length. Colour : uniform yellow. Antennae : two-thirds of body-
length ; of 56 and 57 articles. Ocelli : 1+4, 5, 5, 3 with those of the superior row

304 E. H. EASON

oval, and 1+4, 4, 4, 3 with those of the superior row almost round ; general pattern
as described for E. fasciatus. Pro sternum : with 8+8 small teeth ; lateral spines
peg-like, placed immediately lateral to external teeth. Tergites : there is a greater
tendency towards angulation and slight projection of the posterior angles of the large
tergites, and towards sinuosity of the internal borders of the posterior projections on
the short tergites (Fig. 10) than in most specimens of either E. fasciatus or E. grossipes
but the difference in these respects is hardly consistent enough to be of diagnostic
value ; posterior border of intermediate tergite sinuous. Coxal pores : 30 to 55 in
four to five rows on each of the I2th to I5th coxae, most numerous on I4th. Glandu-
lar pores of i$th legs (Fig. 8) : concentrated on internal aspects only of femur, tibia,
tarsus and metatarsus ; sparse on prefemur.

Chaetotaxy of i^th and i$th legs

General setae : on the metatarsus numerous, up to three-quarters the diameter of
the article in length (Fig. 9) ; of much the same density and absolute length on the
tarsus, about a quarter the diameter of the article in length ; on the tibia and
femur (Fig 8) sparser and rather shorter ; on the dorsal aspect of the prefemur much
the same as those on tibia and femur but on the ventral aspect of the article the
setae are stout, almost spinous. Seriate setae : none on the I5th metatarsus ; as in
E. fasciatus on the I4th leg. Spinous setae : on the i4th tarsus the ventral external
seta (VaTa) is replaced by two small setae (Fig. n), while the ventral internal seta
(VpTa) is stout. Setae of tuft (Fig. 8) : long and numerous in both specimens.

Sculpturing of i$th legs (Fig. 8)

Prefemur with dorsal sulci distinct, internal one broad and deep, continuous with
basal femoral pit, external one finer, not reaching the distal end of article ; basal
femoral pit deep and extensive, scaphoid, occupying about half the diameter of the
base of femur and a quarter of its length, not attenuated distally as in E. fasciatus but
becoming shallower before leading to the internal femoral sulcus which then deepens
and becomes gradually narrower and shallower distally to disappear before reaching
the margin of the pore-free area ; external femoral sulcus finer, extending to the mar-
gin of the pore-free area ; pore-free area occupying about the distal one-fifth of the
internal aspect of femur, slightly elevated without producing any marked swelling,
bearing at its centre a well-circumscribed circular or slightly oval fine pore-sieve
about a quarter the diameter of the distal end of femur ; setae on the pore-free area no
denser than on the rest of the shaft and whether or not there are any actually on the
pore-sieve seems quite fortuitous.

Sculpturing of i^th legs

Internal and external dorsal sulci rather indistinct on prefemur and femur.

Spinulation :

Ventral Dorsal

CtPFT CtPFT

14 m amp amp am a amp a-p a-p

15 m amp amp am a amp p p

SPECIES OF EUPOLYBOTHRUS

305

fine
pore-sieve

O-Smrm

2*0 mm

0'5 mm

2'0mm

FIGS. 8 to 12. Eupolybothrus litoralis litoralis. 8. Femur and distal part of prefemur of
I5th leg (dorsal view). 9. Metatarsus of i5th leg (external view). 10. 6th to I4th
tergites (dorsal view), u. Region of tarso-metatarsal articulation of I4th leg to show
seriate and spinous setae (ventral view). 12. Posterior projection of I3th tergite
(dorsal view). Figs. 8 to n taken from a <$ from Tinos (38 mm long) ; fig. 12 taken from
an immature $ from Rhodes (23 mm long).

3 o6 E. H. EASON

15 VpF and 15 DpT may be absent ; a small coxolateral spine on right I5th leg of
the male from Rhodes only ; a well-developed I5th accessory apical claw.

Genitalia : as in E. fasciatus.

IMMATURE SPECIMENS. The male from Rhodes (lacking the i4th and I5th legs),
18 mm long with 41 antennal articles, 1+3, 4, 3 ocelli, 7+7 prosternal teeth, setiform
lateral spines, about 15 pores on each of the posterior coxae and no coxolateral spines
seems, judging by the development of the genitalia, to belong to the fourth post-
larval stadium.

The female holotype from Tinos, 24 mm long, with 44 antennal articles, 1+4, 4, 3
ocelli, 7+7 prosternal teeth, setiform lateral spines, about 18 pores on each of the
posterior coxae, no coxolateral spines and adult spinulation on the I4th and I5th
legs, has the i5th metatarsal general setae rather longer than the diameter of the
article and the tarsal setae about half the diameter of the article in length. In spite
of its size this specimen seems also to belong to the fourth post-larval stadium because
each gonopod bears only a single minute spur and a small claw.

The female from Rhodes, 23 mm long, resembles the above specimen but for a pair
of incipient teeth adjacent to the median prosternal notch (making a total of 8+8),
30 to 35 pores on each of the posterior coxae and more advanced genitalia with two
small unequal spurs and a well-developed claw on each gonopod. It certainly
belongs to the fifth post-larval stadium.

The tergites of all these specimens tend to show features of immaturity comparable
to those found in E. grossipes and the posterior projection of the i3th tergite of the
last specimen is figured (Fig. 12).

DISCUSSION. Latzel (1880) regarded Lithobius litoralis as a synonym of L.
grossipes ( = " L. fasciatus " of more recent authors) and this synonymy has not
hitherto been disputed. This is hardly surprising because L. Koch's description of
the single immature female might well apply to E. grossipes. But examination of
the holotype shows it to be distinct from this species owing to the distribution of con-
centrated glandular pores on the I5th legs, and also from E. fasciatus owing to the
absence of i5th metatarsal seriate setae. Further, the adult male from Tinos (the
type locality), although labelled " L. grossipes " by Koch, is clearly of the same
species as the holotype and the sculpturing of its I5th femur is characteristic. Koch
defines litoralis by means of immature characters only, notably the rather un-
developed gonopods, and was unaware of its true nature ; otherwise he would not
have identified other specimens of litoralis in his collection as L. grossipes. He may
thus be said to have named E. litoralis accidentally.

Lithobius fasciatus graecus Verhoeff , which is described below, is considered here
to be conspecific, though not consubspecific, with E. litoralis, differing only in the
constant possession of coxolateral spines. L. fasciatus graecus var. fasciatograecus
Verhoeff, which was originally described from Naxos and Crete (Verhoeff, 1901) as a
variety without coxolateral spines, is therefore almost certainly identical with the
nominate subspecies of E. litoralis.

SPECIES OF EUPOLYBOTHRUS 307

Eupolybothrus (E.) litoralis graecus (Verhoeff), n. comb.

Lithobius (Polybothrus) fasciatus graecus Verhoeff, 1899, p. 434, fig. la.

MATERIAL EXAMINED. The following specimens, preserved in spirit, are present
in the Verhoeff Collection of Myriapoda in the British Museum (Natural History) and
are labelled " Lithobius fasciatus graecus Verhoeff. " " Greece " (Reg. no. 03.8.25.

47-49) :

An immature male 23 mm long in a fair state of preservation, a mutilated mature

female about 26 mm long in two separate fragments lacking most of the appendages,
and a mutilated immature female 22 mm long lacking most of the legs. Verhoeff's
(1899) original description of L. fasciatus graecus was based on a number of examples
from several localities in southern Greece and it is probable that these specimens are
part of the syntypical series.

DIAGNOSIS OF ADULT. As nominate subspecies except for the presence of one or two
well-developed coxolateral spines on each of the I5th and sometimes also on each of
the I4th or I2th to I4th legs.

DESCRIPTION OF ADULT. A mutilated and fragmented female only ; detached telopo-
dites of I4th and I5th legs seem to belong to this specimen and their characters are
included in the following description.

Length : 26 mm ; I5th leg 15 mm. Colour : uniform yellow. Antennae : broken.
Ocelli : 1+4, 4, 4, 2. Prosternum : with 8+8 small teeth, the internal pair being
only incipient ; lateral spines peg-like, lateral to external teeth. Tergites : as figures
for E. 1. litoralis (Fig. 11). Coxal pores : 20 to 30 very unequal in size in four rows on
each of the I2th to I5th coxae. Glandular pores of i$th legs : as in E. I. litoralis.

Chaetotaxy of i^th and i$th legs

General setae : as in E. 1. litoralis. Seriate setae : none of the 15th metatarsus ; on
the I4th metatarsus the external row well-defined but the internal row absent ; on
the I4th tarsus rather sparse. Spinous setae : as in E. I. litoralis.

Spinulation :

Ventral Dorsal

CtPFT CtPFT

14 a m amp amp am a amp a-p a-p

15 a m amp amp a a amp p p

Coxolateral spines (VaC) present on each of the I2th to I5th legs; a well-developed
I5th accessory apical claw.

Genitalia : as in E. fasciatus
IMMATURE STADIA

Male fifth post-larval stadium (pseudomaturus)

Differs from the adult in the following characters. Length : 23 mm. Antennae :
three-fifths of body-length ; of 45 articles. Ocelli : 1+4, 4, 2, 2. Prosternum :

3 o8 E. H. EASON

with 7+7 teeth ; lateral spines setiform, lateral to external teeth. Coxal pores :
13 to 18. General setae of i^th and i$th legs : on the metatarsus about as long as the
diameter of the article, on the tarsus about half the diameter of the article in length.
Senate setae of the i^th leg : on the metatarsus the internal row present but reduced
to a few small setae ; on the tarsus very sparse. Setae of tuft : absent. Scuplturing
of i$th leg (left only) : dorsal sulci of prefemur indistinct ; basal femoral pit and in-
ternal femoral sulcus quite distinct but less well denned than in adults of . /. litoralis ;
external femoral sulcus indistinct ; pore-free area occupying about the distal one-fifth
of the internal aspect of femur with neither swelling nor fine pore-sieve. Sculpturing
of i^th legs : dorsal sulci on prefemur and femur very faint. Spinulation : 15 VmT
present ; 14 DaC absent ; two coxolateral spines on left I5th leg, one on right I5th
leg, none on I2th to I4th legs. Genitalia : posterior border of genital sternite with a
feeble median notch and about 8 marginal setae on either side ; gonopods very slender,
basal article about half the length of distal article.

Female fifth post larval stadium (pseudomaturus)

Differs from the male in the following characters : Length : 22 mm. Antennae : 42
articles. Ocelli : 1+4, 3, 3, i. Coxal pores : 13 to 16. Coxolateral spines : one on
each of I4th and i5th legs. Genitalia : two small unequal spurs and a well-developed
claw on each gonopod.

DISCUSSION. In his original description of Lithobius fasciatus graecus, Verhoeff
(1899) mentions the relatively pale colour, the marked posterior emargination of the
large tergites and the presence of one or two coxolateral spines on each I5th leg.
The projecting posterior angles of the large tergites in Koch's specimens of E.
litoralis (Fig. 10) accentuate the general appearance of emargination and this charac-
ter together with the pale colour and also the presence of a coxolateral spine on one
I5th leg of a male from Rhodes leads to the suspicion that graecus may be identical
with litoralis. However, Verhoeff (1899) describes the male I5th femur as resembling
that of "fasciatus " and makes no mention of the chaetotaxy or pore-distribution on
the I5th legs ; later the same author (Verhoeff, 1941 : 105), when discussing these
characters, states that he had no material referable to graecus available for examina-
tion. We would therefore be uncertain as to the identity of graecus were it not for
the specimens purchased from Verhoeff by the British Museum which show the same
pore-distribution and almost the same chaetotaxy as E. litoralis ; the I5th femur of
the single immature male shows just the sort of sculpturing to be expected in a com-
parable stadium of E. litoralis so there is little doubt that Verhoeff's specimens of
graecus and Koch's specimens of litoralis are conspecific. The slight deficiency in
the seriate setae and the number of coxal pores of the former as compared with the
latter is probably fortuitous but the presence of well-developed coxolateral spines in
both mature and immature examples of graecus contrasts with their absence from all
but one example of litoralis, and this specimen shows a small spine on the right side
only.

In the present state of our knowledge of the distribution of these forms it cannot be
more than a matter of opinion as to whether we regard the coxolateral spines as

SPECIES OF EUPOLYBOTHRUS 309

altogether unstable and graecus as identical with litoralis, or whether these spines
should be regarded as characteristic of a subspecies from the mainland of Greece,
occurring only occasionally in the nominate subspecies from the islands. The latter
view has been adopted here although further collecting may well prove it to be in-
correct.

ACKNOWLEDGEMENTS

My sincere thanks are due to the Trustees of the British Museum (Natural History)
for allowing me to borrow specimens ; to Dr. G. O. Evans, Dr. J. G. Sheals and Mr.
K. H. Hyatt of the Arachnida Section of the Department of Zoology of the Museum for
their personal help ; to Professor G. C. Varley and Mr. E. Taylor of the Hope Depart-
ment of Zoology, Oxford for enabling me to examine specimens from the Newport
Collection ; and to Mr. D. Macfarlane of the Commonwealth Institute of Entomology
for reading and criticising my manuscript.

REFERENCES

ATTEMS, C. G. (1902). Myriopoden von Kreta, nebst Beitragen zur allgemeinen Kenntnis

einiger Gattungen. Sber. Akad. Wiss. Wien (Abt. i), 111 : 527-614.
BROLEMANN, H. W. (1894). Contribution a la faune myriapodologique M6diterran6enne, 30

note. Mem. Soc. zool. Fr. 7 : 437-462.

(1930). Myriapodes. Chilopodes. Faune Fr. 25 : 1-405.

CHAMBERLIN, R. V. (1925). The Ethopolidae of America north of Mexico. Bull. Mus. comp.

Zool. Haw. 57 : 383-437.

(1952). On the Chilpoda of Turkey. Istanb. Univ. FenFak. Mecm. (B) 17 : 183-258.

CRABILL, R. E. (1958). A new Eulithobius, with a key to the known American species (Chilo-

poda : Lithobiidae). /. Wash. Acad. Sci. 48 : 260-262.
JEEKEL, C. A. W. (1967). On two Italian Lithobius species described by Silvestri, with

taxonomic notes on the genus Eupolybothrus Verhoeff (Chilopoda, Lithobiidae) . Beaufortia

14 : 165-175.
KOCH, C. L. (1847). System der Myriapoden. In Kritische Revision der Insectenfauna

Deutschlands 3. Panzer, G. & Herrich-SchafEer, A. (eds.). Regensburg.
KOCH, L. (1862). Die Myriapodengattung Lithobius. Niirnberg : J. L. Lotzbeck.

(1867). Zur Arachniden-und Myriapoden-Fauna siid-Europas. Verh. zool.-bot. Ges.

Wien, 17 : 857-900.

LATZEL, R. (1880). Die Myriapoden der Osterreichisch-Ungarischen Monarchic, i. Die Chilo-

poden. Wien : Holder.
(1888). Die vom k. k. Oberarzte Herrn Dr. Justyn Karlinski im Jahre 1887 in Bosnien,

der Herzegowina und in Novibazar gesammelten Myriopoden. Verh. zool.-bot. Ges. Wien,

38 : 91-94.
MURALEWITSCH, W. S. (1906). Myriapoden, gesammelt von der Expedition nach der Halbinsel

Kanin im Jahre 1902. Zool. Anz. 30 : 66-69.

(1929). Scutigeridae et Lithobiidae de la faune du Caucase. Memu. Obshch. Lyub.

Estest. Anthrop. Etnogr. Zool. 4 : 1-120.

NEWPORT, G. (1844). Monograph of the class Myriapoda, order Chilopoda. Trans. Linn.

Soc. Lond. 19 : 265-439.
POCOCK, R. I. (1890). Contributions to our knowledge of the Chilopoda of Liguria. Annali

Mus. civ. Stor. not. Giacovno Doria, 29 : 59-68.
VERHOEFF, K. W. (1899). Beitrage zur Kenntniss palaarktischer Myriopoden. XI. Auf-

satz : neue und wenig bekannte Lithobiiden. Verh. zool.-bot. Ges. Wien, 49 : 451-459.

(1900). Beitrage zur Kenntniss palaarktischer Myriopoden. XV. Aufsatz : Litho-
biiden aus Bosnien, Herzegowina und Dalmatien. Berl. ent. Z. 45 : 153-179.

3 io E. H. EASON

VERHOEFF, K. W. (1901). Beitrage zur Kenntniss palaarktischer Myriopoden. XVI.
Aufsatz : zur vergleichenden Morphologic, Systematik und Geographic der Chilopoden.
Nova Acta Acad. Caesar. Leop. Carol. 77 : 369-465.

(1905). t)ber die Entwicklungsstufen der Steinlaufer, Lithobiiden, und Beitrage zur

Kenntnis der Chilopoden. Zoo/. Jb. (Supplement), 8 : 195-298.

(1907). Chilopoda. Bronn's Kl. Ordn. Tierreichs, 5, Abt. 2, Buch i : 1-725.

(1934)- Beitrage zur Systematik und Geographic der Chilopoden. Zoo/. Jb. (Syst.),

66 : I-H2.

(1937). Chilopoden Studien. Zur Kenntnis der Lithobiiden. Arch. Naturgesch.

N.F. 6 : 171-257.

(1941). Asyanin zoogeografiyasi ve hay van sistematige hakkinda. Asiatische Beitrage.

Istanb. Univ.FenFak. Mecm. (B) 6 : 85-117.

Dr. E. H. EASON,
BOURTON FAR HILL,
MORETON-IN-MARSH,
GLOUCESTERSHIRE.

INDEX TO VOLUME 19

The page numbers of the principal references and the new taxonomic names are printed in bold type

abbreviata, Amphiura .... 8
abbreviata, Moringua . . . .220
abbreviatus, Aphthalmichthys . . 221

abdita, Triodia .... 67-6g

abditus, Amphioplus . 4, 6, 46, 47, 48, 56

abernethyi, Amphiura . . . 21, 22

acanthinus, Amphioplus 33

Acaridae .... 83-i 18,4 pis.
Acarina .... 83-1 18,4 pis.
Acarus .... .83-1 18, 4 pis.

Acentrogobius .... 227-228
aciculatus, Amphioplus . . . 45. 56

Acrocnida ..... 6, 26, 27
acuticeps, Gymnodraco .... 202
acutirostris, Ecphy motes . . .239

acutispina, Amphiodia . . . .27
acutus, Amphioplus .... 56
aestuarii, Amphiura . .11, 14, 15, 17, 18
aestuarii, Ophinoephthys ... 20

affinis, Amphiodia . 24, 25, 27, 39, 42, 46, 54

affinis, Amphioplus . . . 50, 54

affinis, Amphiura ..... 48
affinis, Ophiophragmus . . . .52
africana, Fellaria ..... 20
africana, Ophionephthys . 13, 14, 15, 18, 20

africana, Tyroglyphus . . . 113, 117
africanthiforme, Mesacanthion 139, 142-145
Africanthion . . . 143, 145

Agama . . . 239, 240, 242, 247

agamoides, Hypsibatus . . . 240, 242
agamoides, Lophyrus . . . 240, 242
agamoides, Ophryesa .... 242
agassizi, Amphioplus .... 58
agassizii, Amphioplus . . . 70, 71

ailsaclarki, Amphioplus . . . 51, 57

Ailsaria .... 21,23,36,49,50

alba, Amphiura . . . . .12
albanicus, Polybothrus fasciatus . 302, 303
albobrunneus, Scorpaena . . . 230

Albula ..... 130, 133, 281

Alectis 225

allardicei, Anarchias .... 217
Allothrissops . . . 278, 279, 280

alosoides, Hiodon . . 260, 263, 265, 267,

269, 274, 279

altipinnis, Antennarius .... 233
amblycephalus, Julis .... 226
amblycephalus, Thalassoma . . 226-227
amphacantha, Amphiacantha ... 34
amphacantha, Dougaloplus ... 34
amphacantha, Ophiocnida 33
Amphiacantha ... 6, 28, 33, 34

Amphiactis . 3,41,42,58,62,69,70,71,73,74

Amphichilus .

Amphichondrius
Amphilepididae
Amphilepis .

36, 38, 39, 44, 45, 46, 48,
49, 56, 57, 60, 63, 64
. 6,31
69-70

3,4.37.38,39,40,43,
44, 60, 69, 70-72
Amphilimna . . . . . .34

Amphinephthys . . . 7,9,10,11,12
Amphiodia . 3, 4, 5, 6, 7, 19, 20, 23-28, 29, 36,

37. 39, 4. 4 2 , 43, 44. 4^, 52, 56, 57, 62
Amphioplus . 3, 4, 5, 6, 14, 19, 20, 21, 22, 23, 27,

28, 32, 33, 35, 36-66, 68, 70, 71, 73, 74
Amphipholis . . 3, 4, 6, 7, 26, 28-31, 33,

35, 46, 50, 52, 63
Amphiprion ...... 226

Amphispina .... 23, 24, 25-26, 52

Amphistigma

Amphiura ....

Amphiuridae

ampullaceus, Leptolaimus

Anarchias ....

ancistrotus, Amphioplus

andreae, Amphipholis

angularis, Amphiura

anomala, Amphiura

antarctica, Amphilepis .

antarctica, Amphiodia .

Antennarius ....

antennata, Pterois

antennata, Scorpaena

apenninigenus, Lithobius fasciatus

apenninigenus, Polybothrus

Aphthalmichthys .

Apogon ....

Arapaima ....

archeri, Amphioplus

arcystata, Amphiura

argus, Cephalopholis

assimilis, Amphiodia

astarte, Amphiactis

aster, Amphiura

asterictus, Amphioplus

Asterropteryx

asymetricus, Theristus

atlantica, Amphiura

atlanticus, Tarpon

32,33

. l-8i

. 1-8 1

139, 151-153

. 217

48, 56, 58

52,53,55

12
12

atra, Amphiodia
atra, Amphipholis .
atra, Ophiolepis
Aulorhynchus
aurensis, Amphioplus
auronitens, Ophryessa

72
28

. 233
. 231

. 231

295
293, 295

221

. 224

. 282

4, 46, 47, 49, 50, 57, 60
10, ii
224
27
7, 71, 73, 74

12

56

. 227
. 161
13

121, 122, 125-129,

131, 133. 134, PI- 3

26
26
26

. 198
51,57

239

312

INDEX

axillaris, Julis ..... 226
axillaris, Stethojulis .... 226
Axiognathus . . . . 29, 30, 31

Axonolaimus . . . 139, 148-1 51

baldensis, Polybothrus . . . 302, 303
Balistes ...... 232

Banister, K. E . 179-2O9

barbarae, Amphiura . . . .26
barbatus, Camacolaimus . 139, 158-156
basilicus, Amphioplus .... 56
bathybia, Ophiomonas . . -37, 4, 44

Bathygobius 227

Bathythrissus . . . . 130, 131
belgicae, Amphiura . . . 12, 72

beta, Opsanus ..... 203
bicinctus, Amphiprion .... 226
bidentata, Amphiura . . . 38, 39, 43
bocki, Amphioplus . . . 52, 53

bogerti, Tropidurus .... 255
borealis, Amphiura . . . 13, 17

borealis, Ophiopeltis . . . 13, 14

bosniensis, Lithobius fasciatus 290, 292, 294

bosniensis, Lithobius grossipes . . 295

bosniensis, Polybothrus fasciatus . . 290
Bothropolys ...... 299

Bothus 232

brachiostictus, Amphioplus . . 27, 57

brachyactis, Amphiura . . . . 8, n
brachyactis, Ophiophragmus . . 24, 25, 26
brasiliensis, Ophyessa .... 240
brasiliensis, Plica . . . . .240
brummeri, Muraena . . . .214
brummeri, Pseudechidna . . . 214

cesareus, Amphichilus
cesareus, Amphioplus
Chaetodon ..

. 39, 45, 63
40, 45, 63, 64
.. 225

caelatus, Amphioplus
calcarabellus, Acarus .
Callechelys .
Callogobius .
Calotes
Camacolaimus
canescens, Amphiactis
canescens, Amphiura
canescens, Histampica .
capax, Amphioplus
Caracanthus .
catenata, Agama .
catenata, Ophryessa
catenatus, Enyalius
cauerensis, Acentrogobius
cauerensis, Gobius
caulleryi, Amphiodia
caulleryi, Amphiura
causatus, Amphioplus
Cephalopholis
ceramis, Amphiura
cernua, Amphiura .
cernuus, Amphioplus
cesarea, Amphioplus
cesarea, Amphiura

52, 55

85,93, 107-no, PI. 2

. 218

. 228

240

139, 153-156
62, 70, 71

73
74

45, 57. 66

. 232

239

239
240

227-228
227

27,39,43

42, 43, 48

5i,57
. . 224

22, 23, 58

. 48

37. 4. 42, 44. 49

63
. 45, 63, 64

chaetoxysilos, Acarus 85,93,105-107, 112, PI. 3
chalcidica, Iguana .... 238, 242

chiajei, Amphiura . . . 4, 5, 6, 7, 12, 29
chilensis, Amphiodia . . . 27, 28

Chlevastes . . . . . .217

chordatus, Stylophorous . . . 200

Chromaspirina . . . 139, 165-I&7
ciliaris, Alectis ..... 225

ciliaris, Zeus ...... 225

cinctus, Amphioplus . . 46,47,50,51,57
cinereus, Conger . . . . .221

cinereus, Conger cinereus . . . 221

cinereus cinereus, Conger . . . 221

Clark, A. M l-8i

Clupeomorpha
colubrina, Muraena
colubrinus, Myrichthys .
concolor, Amphiura
condita, Amphiura
conditus, Amphioplus
conditus, Unioplus
conductus, Amphioplus .
confinis, Amphioplus
confinis, Amphiura
confinis, Unioplus .
congensis, Amphioplus .
congensis, Amphiura
Conger

coniortodes, Amphioplus
connexus, Enoplolaimus
consors, Amphioplus
Cooper, M. J.
cooperae, Phaenomonas
corallicola, Epinephelus .
correcta, Amphiura
coryphaenoides, Leptolepis
crassa, Amphiodia
craterodmeta, Amphiodia
crossota, Amphiura
Ctenamphiura
cuneatus, Amphioplus
cyanopodus, Serranus
cyanostigma, Istiblennius
cyanostigma, Salarias
cyclaspis, Amphiodia
cyclospis, Amphiodia
cyprinoides, Megalops . 121, 122-125, I2 6, 127,
128, 129, 131, 133, 134, Pis. i, 2
cyrtacanthus, Amphioplus . 20,21,46,47,56
cythera, Amphiactis . . .58, 70, 71
cythera, Amphioplus ... 58, 74

cythera, Histampica .... 74

132-133
. 217
. 217

12,38,39,41,42,43
49,58

38, 39. 4. 46, 58

. 38, 39. 58

- 56
. 37,40,42,58-59

48, 58

58-59

35

52

. 221

21,23,46,47,49,56
139, 142

57

213

219-220

. 223

21-22

261

. . 24,25,27

27,28

9, 10, 12

27
. 56

. 222

. 229

. 229

28

dacunhae, Amphiura
dalea, Amphioplus
dalea, Amphiura .
daleus, Amphioplus

12

43.68

43, 48, 67, 68
37. 38, 39. 4. 42, 44. 45

INDEX

313

daleus, Unioplus .
Dalophis

davidi, Lycoptera .
debilis, Amphioplus
debita, Amphiodia
decacantha, Fellaria
decacantha, Ophionephthys
deficiens, Amphiura
Dendrochirus
denticulata, Amphiura .
denticulatus, Enoplolaimus
denticulatus, Theristus
depressa, Amphipholis .
depressus, Amphioplus .
depressus, Lymanella
derjugini, Amphiacantha
diacritica, Amphiura
diacritus, Amphioplus
Diamphiodia
diastata, Amphilepis
diastata, Amphiura
didymus, Amphioplus
difficilis, Amphioplus
digitata, Amphiodia
digitata, Amphispina
digitula, Amphiodia
digitula, Amphiura
diomediae, Amphiura

dispar, Amphioplus
dissidens, Ophiactis
diversispiculum, Theristus
dividua, Amphiacantha .
dividua, Amphiodia
doriae, Lithobius .
dorsalis, Bathythrissus .
Doryrhamphus
Dougaloplus
duplicata, Amphiactis
duplicata, Amphiodia
duplicata, Amphispina .
duplicata, Amphiura
duplicata, Histampica
duplicatus, Ophiophragmus
dussumieri, Hyporhamphus
dybowskii, Hypoptychus

Eason, E. H.
Echidna

echinata, Amphiacantha
echinata, Ophiocnida
echinata, Ophiophragmus
echinatus, Dougaloplus .
echinatus, Ophiocnida
echinatus, Ophiophragmus
echinulatus, Amphioplus
ecnomiotata, Amphiura .
ecnomiotata, Fellaria
ecnomiotata, Ophionephthys
Ecphymotes .

38, 39

. 215

260, 273

56,66

27

20

. 13, 14, 15, 18, 20

II

. 231

12

139, 140-142

139, 158-i6i

. 52

4, 46, 47, 50, 54, 55

4 6 .47

34

9

. 58

23, 25, 26, 27

. 40, 70, 72

72

56
57

25

25
27

40, 48

. 38, 39, 40, 41, 42,

43, 44, 48, 70

45,57

73. 74

. 161

33

27
. 297

130, 131
. 222

. 32,33-36,58,60
74

24.25

24,25

73

74

25

221

287-3io

213-214

. 36

34.35

34

32,33,35-36,58

35

28, 34, 35

49, 50, 56
ii, 14, 15, 17, 18

20

13,20
239, 242

edentulus, Blennius

edentulus, Istiblennius .

elaps, Chlevastes

elegans, Amphipholis

elegans, Ophiura .

Eleotriodes .

Eleotris

Elopomorpha

Elops ....

enoploidiformis, Enoplolaimus

enoploidiformis, Epacanthion

Enoplolaimus

Enyalius

Eohiodon . . . 260,

Epacanthion
epigrus, Ophiochytra
Epinephelus .
erectus, Theristus .
Etheridge, R.
eugenius, Quisquilius
Eulithobius .
Eupolybothrus
euryaspis, Amphiodia
Exocoetus

exsecrata, Amphioplus .
exsecrata, Amphiura
exsecratus, Amphioplus .

130

139

267,
279

217

. 227

227, 228

132-133

133,261,281

. 148

139
140-142, 148

239, 240

271, 274, 275,

280, 281, 282

139, 145-148

. 69

222-224

. 165

235-256

228, 233

. 289

. 287-3io

25,27

. 198

59

59

51, 56, 59-6i

falcatus, Amphioplus 37, 40, 43, 44, 45, 46, 48, 51
falcatus, Amphiura .... 49

famula, Amphioplus . . . -57
farinae, Tyroglyphus
farris, Acarus

farris, Tyroglyphus

85, 89, 93, 95. 99. 1 08,
114, 115, 116, PI. 2
. . .113
fasciatograecus, Lithobius fasciatus graecus

var ....... 303, 306

fasciatograceus, Polybothrus fasciatus

graecus var. . . . . 303

fasciatus, Bothropolys .... 299

fasciatus, Eupolybothrus 289, 290-295, 297,

298, 299, 302, 304, 306, 307

fasciatus, Lithobius . 289, 290, 294, 295,

297, 301, 302, 306

fasciatus, Polybothrus .... 299
fasciatus albanicus, Polybothrus . 302, 303
fasciatus apenninigenus, Lithobius . . 295
fasciatus bosniensis, Lithobius 290, 292, 294

fasciatus bosniensis, Polybothrus . . 290
fasciatus graecus, Lithobius 289, 306, 307, 308
fasciatus graecus, Polybothrus . . 307

fasciatus graecus var. fasciatograecus,

Lithobius ..... 303, 306
fasciatus graecus var. fasciatograecus,

Polybothrus ..... 303
fasciatus semicincta, Ophisurus . .217
favaginea, Gymnothorax . . 215

favagineus, Gymnothorax . . -215
Fellaria . . . . 16, 17, 18-ig, 20

314

INDEX

festivus, Lithobius
fibulata, Amphiura
filiformis, Amphiura
filograneus, Ophiophragmus .
fimbriata, Gymnothorax
fimbriata, Muraena
firma, Amphioplus
firma, Amphiura .
firmus, Amphioplus
fissa, Amphiodia
flavimarginata, Gymnothorax
flavimarginata, Muraena
flavimarginatus, Balistes
flavimarinatus, Pseudobalistes
florifera, Amphiura
forficatus, Lithobius
formata, Amphioplus
formata, Amphiura
formatus, Amphioplus . 44,
formosa, Ophiostigma
formosus, Dougaloplus .
fosciatus storkani, Polybothrus
fragilis, Amphiura .
fragilis, Lycoptera
frigida, Amphiodia
furcatus, Amphioplus
fuscoalba, Amphiodia
fuscus, Bathygobius
fuscus, Gobius

295. 297. 32

. . 12

4,7, 13,27,29

3i
. 215

. 215
61
61

56, 61
27, 28
. 215
. 215

232
232-233

7

. 300
61
61

58, 61-62, 70, 71
33
35

303
12, 16, 17

259

27

50. 52

27
. 227

227

Gasterosteus . . . 186, 193, 198

gastracantha, Amphiura . . n, 33

geminata, Amphipholis . . . .26
gentilis, Amphioplus 37, 38, 39, 40, 43, 62, 70, 71
gentilis, Amphiura . . . 48, 62

gentilis, Unioplus . . .38, 39, 62, 70, 71
georginaus, Parachaenichthys . . . 202
gibbosus, Amphioplus .... 50
gibbosus, Ophiophragmus ... 52
gissu, Pterothrissus . . . .130

glaucus, Amphioplus . . . 44, 57

Gobius 227, 228

gorgonocephalum, Epacanthion 139, 145-148
gracilis, Acarus . . . 87, 91, 92, PI. i
gracillima, Amphipholis .... 26
graecus, Eupolybothrus litoralis . 307-3OQ
graecus, Lithobius fasciatus 289, 306, 307, 308

graecus, Polybothrus fasciatus
granulosus, Amphichondrius
grata, Amphiodia .
grata, Amphiura .
gratus, Amphioplus
gratus, Unioplus .
Greenwood, P. H. .
Griffiths, D. A.
grisea, Amphiodia .
grossipes, Eupolybothrus

37

grossipes, Lithobius

27, 37, 4, 62
4 8 ,62

3 8 , 39, 62-63
38, 39, 62-63
119-135, 3 pis., 257-285
83-i 18,4 pis.
26

289, 294, 295-303,
304, 306
. 289, 294, 295, 296, 301,
302, 306

grossipes bosniensis, Lithobius

guamensis, Scorpaena

guamensis, Scorpaenodes

Gymnodia .

Gymnodraco

gymnogastra, Amphiura

Gymnomuraena

Gymnoplus .

gymnopora, Amphilepis .

Gymnothorax

gyraspis, Amphiodia

Halichoeres . . . .

Halinema

Halocypselus

hastata, Amphioplus . .

hastata, Amphipholis

hastatus, Amphioplus . 6

hastatus, Lymanella
Hemilepis . . . .
Hemipholis .
heptacantha, Fellaria
heptacantha, Ophionephthys .
heterospiculoides, Theristus
heterospiculum, Theristus
hexacantha, Ophionema
hexacanthus, Amphioplus
hexactis, Ophionema
hexagonatus, Epinephelus
hexagonatus, Holocentrus
hinemoae, Amphiura . .
Hiodon .

Hiodontidae .
hispida, Agama . .
hispida, Ophiocnida
hispida, Ophiolepis
Histampica

holotropis, Tropidurus . 237,
Hope Department of Zoology,
Hoplichthys . . . .
Hyperanodon
Hypoptychus
Hyporhamphus
Hypsibates .
Hypsibatus .
Hypsilophus .

295

230
. 230

23, 25, 26, 27

202
10

2l6
23

72

214-216

26

. 226

; . 158

. 198

51

52

. 4 6 . 47, 5, 5i, 52,

54, 55, 59

. 46, 47, 59

7, 9, 10, ii, 12

3

20

13, 14, 15, 18, 20

. 161

. 161

14, 19

. 56

14, 15, 17, 18, 19

. 223

223

12

260-283
257-285

239
28,31,32,33

28

73-74

241, 248, 251, 254

Oxford 287-310

202

. 240, 241, 247, 240, 251
200
. 221
239, 241

239, 240, 241, 242, 243, 247
239, 241, 243

Icalia .
Iguana
immobilis, Acarus

impressa, Amphipholis
impressus, Amphioplus
incisa, Amphiura .
incisus, Amphioplus
incisus, Unioplus .
incompleta, Amphiura
Indostomiformes .
Indostomus .

7, ii, 12
238, 242, 247

86, 89, 90, 93, 101, 109, no,
112, 113, 114, 116, Pis. i, 4

63
45,46,47,51,57,63-64

. . . . .48

38, 39, 4- 43

. . - 38,39

. ... ii

179-204

. . . 179-209

INDEX

315

inepta, Amphiura .
infantile, Mesacanthion .
infolnana, Amphilepis
inglisi, Chromaspirina
ingolfiana, Amphilepis
inhaca, Gobius
inhaca, Quisquilius
insidiator, Platycephalus
integer, Amphioplus
integra, Amphipholis
intermedia, Amphioplus
intermedia, Amphiura
intermedius, Amphioplus
interstitialis, Theristus
intricata, Amphiura
intricata, Ophionema
iranica, Ophionephthys .
iranica, Ophiopeltis
Istiblennius .
iuxtus, Amphioplus

januarii, Amphipholis
japonicus, Amphioplus .
japonicus, Ophiophragmus
japonicus, Platycephalus
javanica, Moringua
javanicus, Aphthalmichthys
Julis ....

145
72

139, 165-167

69, 72

. 228

. 228, 233

202

4. 50, 55
52

. 64
. 64

45,46,51.57.64

139, 161-I&2

16, 17

13, 14, 15, 16, 17, 19

14, 15, 18, 20

20

. 229
51,57

28, 29, 30
50, 53, 54- 55
50, 52. 54

202

. 221

221
226

kohleri, Epinephelus . . . 222-223
koreae, Amphiura . 37, 38, 39, 40, 41, 42, 43, 48

Labrus ...... 226

Lacerta 237, 238, 239, 240, 241, 242, 247, 253
lacerta, Elops . . . . .130

laevis, Amphioplus . 46, 47, 50, 52, 53, 55

laevis, Amphiura ..... 52

laevis, Lymanella . . . .46, 47, 55

langsdorni, Hoplichthys . . . 202

langshanensis, Sinolycoptera . . -275
latispina, Amphiura .... 9

legatus, Amphioplus . . . .56

Leptolaimus . . . . 139, 151-153

Leptolepis . 261, 267, 271, 278, 279, 280, 281
leucotaenia, Echidna? .... 214

leucurus, Uropterygius .... 216

libera, Ophiocnida . . . -33

limbata, Ophiolepis . . . .26

limicola, Ophionephthys . 13, 14, 16, 17,

19, 20, 21, 60
lineatus, Istiblennius .... 229

lineatus, Salarias ..... 229

Lithobius . 289, 290, 294, 295, 296, 297, 300,
301, 302, 303, 306, 307, 308

litoralis, Eupolybothrus . 289, 295, 298, 302,

303, 306, 308

litoralis, Eupolybothrus litoralis 303-3o6, 308
litoralis, Lithobius . . 289, 303, 306

litoralis graecus, Eupolybothrus . 307-309
litoralis litoralis, Eupolybothrus 303-3O6, 308

lobatus, Amphioplus .... 56
longipeda, Ophiocnida 35

longirima, Amphioplus . . . -57
Lophyrus . 239, 240, 241, 242, 247, 249, 250
lorioli, Amphioplus . . . 22, 58

lorioli, Amphiura .... 22-23, 5 8
lorioli, Ophiactis . . . . .22
loripes, Amphipholis . . . 29, 30

loveni, Ophiocnida . . . 28, 31

lucidus, Amphioplus . . . 20,21,56
luctator, Amphioplus 20, 21, 32, 33, 35, 36, 58
luetkeni, Ophiophragmus ... 26
lunare, Thalassoma .... 226
lunaris, Amphiura . . . 38, 39, 40, 43
lunaris, Labrus . . . . .226
lunula, Chaetodon ..... 225
lunula, Pomacentrus
lycidas, Amphiactis
lycidas, Histampica
Lycoptera
Lymanella

225

70, 7L 73. 74
74

257-285
46, 47, 51-56, 59, 60, 66

machnata, Elops . . . . .130
macilentus, Amphioplus . . 46, 47

macraspis, Amphiodia .... 20
macraspis, Amphioplus . . .20, 44, 56
macrocephala, Moringua . . . 220

macrocephalus, Aphthalmichthys . .221
macrochir, Moringua . . . 220, 221
macrocoryne, Acarus 85, 90, 93, 102-105, PI. 2

macropterus, Muraenichthys
macrospilus, Epinephelus
maculatus, Caracanthus .
maculatus, Micropus
maculosa, Muraena
maculosus, Myrichthys .
magellanica, Ophionephthys
magellanicus, Amphioplus
magnificus, Amphioplus .
Manchurichthys
mancus, Bothus
mancus, Pleuronectes
margaritaceus, Lophyrus
margaritaceus, Ophryessa
marginatus, Halichoeres
marginatus, Ophiophragmus
marmorata, Gymnomuraena
marmoratus, Uropterygius
mediterranea, Amphiura
Megalopidae .
Megalops

. 218

223
. 232

232
. 217
. 217

14, 2O, 21

56

57

260

. 232

232

239

239
. 226

26
216
216

. 6,8

119-135, 3 pl s -
121, 122-125, I2 6, 127, 128,
129, 131, 133, 134, 261
megapomus, Amphioplus . 46, 47, 52, 53, 54
megapomus, Lymanella . . . 46, 47

melanopleura, Doryrhamphus . . 222

melanopleura, Syngnathus . . .222
melanotaenia, Callechelys . . . 218
melas, Plesiops ..... 224
merra, Epinephelus .... 223
Mesacanthion . . . 139, 142-145

INDEX

Mesoclupea ....

260

ochrocollaris, Ophyressa

239

Microlaimus ....

139, 167- 1 69

ochrocollaris, Plica umbra 237, 250,

251,252,253,

Micropholis ....

26, 29, 31

254

microplax, Amphiodia

25

ochrocollorae, Uperanodon

240

micropterus, Muraena

216

ochroleuca, Amphiodia .

27, 39, 4, 4 6

micropterus, Uropterygius

. 216

ochroleuca, Amphiura

. 49

Micropus ....

. 232

ochroleucus, Amphichilus

38, 39

microsoma, Amphiura

8,9,11

ochroleucus, Amphioplus

38, 39

middendorffi, Lycoptera 259, 260, 262, 264, 266,

octacantha, Amphiura .

. 16, 17, 18

267, 268, 269, 270, 272, 273,

275. 277, 278

octacantha, Fellaria

16, 17, 20

minuta, Amphiodia

27

octacantha, Ophionephthys . 13,

14, 15, 18, 20

minuta, Amphistigma

32,33

oerstedi, Amphiodia

26

minutum, Tripterygium .

. 230

oliffi, Epacanthion

139, 148

minutus, Tripterygion

. 230

olivacea, Amphiodia

27

mirabilis, Acarus .

91-92

opheliae, Paracanthonchus

139, 172-176

mitratus, Petroscirtes

. 230

Ophiactidae ....

73-75

miyadii, Amphioplus

52, 53

Ophiactis . . . -3,

22, 69, 74, 74

mobilis, Amphilepis

72

Ophiocentrus

3- 6, 69

modestus, Amphioplus .

. 58

Ophioceramis

42

modestus, Ophiochiton .

. 58

Ophiochiton .....

- 58

Monamphiura

. 7, 12

Ophiochytra ....

. 69

Monopholis ....

. 28,29,31

Ophiocnida . . 3,6,28,31,32,

33, 34, 35, 73

montanus, Lithobius

295, 296, 302

Ophiocnidella

28

Moringua ....

220, 221

Ophiocomella . . .

" 33

Mucogobius ....

. 228

Ophiocopa ....

74

Muraena . . . 213,214,

215, 216, 217

Ophiodictys ....

73

Muraenichthys

. .218

Ophiolepis ....

26, 28

Muraenophis

. 215

Ophiomastix

37

Murenophis ....

. 214

Ophiomonas . 23, 37, 38, 39, 40, 41,

44, 60, 70, 72

murex, Amphiura .

3i

Ophionema . 3, 7, 13, 14, 15, 16,

17, 18, 19, 20

murmanica, Amphipholis

28, 29

Ophionephthys . . 3, 6, 7, 9,

10, 13, 14, 15,

muroii, Lycoptera .

275

16, 17,

1 8, 19-ai, 60

Myrichthys ....

. 217

Ophiopelte ....

7, r 3, 17

nebulosa, Echidna

213-214

Ophiopeltis . . 3, 7, 9, n,
1 6,

12, 13, 14, 15,

17, 18, 19, 20

nebulosa, Muraena
neglecta, Amphiura
nereis, Amphioplus

. 213

7
. 58, 7- 7i

Ophiopholis ....
Ophiophragmus . 3, 6, 19, 24,

3

25, 26, 27, 28,

nidicolous, Acarus 85, 86, 87, 90

, 93-102, 105,

T T A V>} 1

3 1 ' 34
Ophiophrura

5, 5 2 > 57, 6
74

nigricans, Pharopteryx .
nigricans, Plesiops
norae, Amphiura .
norvegica, Amphilepis 4, 38, 39, 40,
norvegica, Amphiura
notacantha, Amphiura .
notopsis, Julis
novemfasciatus, Apogon
nuchalis, Scorpaena
nuchalis, Sebastapistes .
nudipora, Amphipholis .

114, **
224
224

21, 22
60, 69, 70, 72
. 7

33
226

. 224

. 231
. 231

29, 31

Ophiostigma . . 3,6,
Ophiozona ....
Ophiozonella
Ophisurus ....
Ophiura ....
Ophryesa ....
Ophryessa . . . 239,
Ophryoessa ....
Ophyessa ....
Opsanus ....
orcombensis, Axonolaimus .

32, 33, 35, 69

74
74
. 217

31

242
240, 241, 247
. 241
240, 242
203
139,148-151

nudus, Africanthion
Nullamphiura

M3
7, ii, 12

pachybactra, Amphipholis
Palmer, G. .

30
. 211-234

Nullopholis ....

. 28, 29, 31

Pandelia ....

7, n, 12

obtecta, Amphiodia

25,27

panthera, Hyperanodon .

. 248

occidentalis, Amphiodia .

4, 25, 26, 27

Panthera, Lophyrus

239, 240, 242

occidentalis, Amphioplus

57

Panthera, Ophryessa

239

ochrocollare, Uperanodon

247

pantherinus, Bothus

. 232

ochrocollaris, Lophyrus . 239,

240, 241, 247,

pantherinus, Rhombus .

. 232

249, 250

papillatus, Amphioplus .

58

INDEX

317

papyracea, Amphilepis .
Paracanthonchus .
Parachaenichthys .
Paracrocnida
paradoxus, Indostomus .
Paralinhomoeus
parata, Ophiactis .
pardalis, Muraena .
pardalis, Murenophis
partita, Amphiura
partita, Ophiactis .
parviclypeus, Amphioplus
parviscutata, Amphiura .
parviscutata, Ophiopeltis
patens, Amphilepis

patula, Amphiura .

patulus, Amphioplus 37, 38, 39

patulus, Unioplus .

paucispinus, Ophiophragmus .

paulus, Istiblennius

paulus, Salarias

pectinatus, Amphioplus .

pectorale, Ophiactis

Pegasus .

peloria, Amphiodia

peltigerus, Hyperanodon

pentacantha, Amphilimna

pentacantha, Amphiphdis

peregrinator, Amphioplus

periercta, Amphiodia

persica, Paracrocnida

personata, Amphioplus .

personata, Amphiura

personatus, Amphioplus

petelli, Gymnothorax

petelli, Muraena

Petroscirtes .

Phaenomonas

phalerata, Ophionephthys

phalerata, Ophiopeltis

philippinensis, Ophionema

philohelminthius, Amphioplus

Phoxinus .

phoxinus, Phoxinus

picta, Agama

picta, Gymnothorax

picta, Hypsibatus .

picta, Muraena

picta, Ophyessa

picta, Plica . . . .

picta, Pneustes

picteti, Ophiocnida

pictum, Uraniscodon

pictus, Calotes

pictus, Ecphymotes

pictus, Enyalius

pictus, Hypsibatus

pictus, Uperanodon

pinnata, Phaenomonas .

72 planispina, Amphiodia .... 26
139. 172-176 platyacanthus, Amphioplus . . -57

198, 202 Platycephalus ..... 202

6, 27, 46, 47 platydisca, Amphipholis ... 26

179-209 platyspina, Amphiodia ... 23, 26

139, 156-158 platytata, Amphilepis .... 72

73, 74 Plesiops ...... 224

214 Pleuronectes ..... 232

. 214 Plica 235-256

73 plica, Agama . . . 239, 240, 242

74 plica, Ecphymotes . . . 239, 242
57 plica, Hypsibatus . . . . .239

15, 17, 18 plica, Hypsilophus .... 243

15, 17, 18 Plica, Lacerta . . 238, 239, 240, 241, 242

37, 38, 39, 40, 43, plica, Ophyessa .... 240, 242

44, 60, 70, 72 plica, Plica 237, 239, 240, 242-247, 248, 254, 255

48, 73 plica, Pneustes .... 239, 242

,40,41,42,44,49 plica. Ptychopleura .... 240

38, 39, 44 plica, Ptychosaurus . . . 240, 243

26 plica, Stellio ..... 238, 242

229 plica, Uraniscodon . . 239, 242, 243

229 plica, Uranoscodon . . . .241

56 Pneustes .... 239, 242, 247
73 polyacantha, Amphiura . . . 10, n

201 Polybothrus . 290, 293, 295, 299, 302, 303, 307

26, 27 polydonta, Pomponema . . . 170, 172

240, 247, 251 Pomacentrus ..... 225

34 Pomponema .... 139, 169-172

28 pontica, Chromaspirina . . . 139, 167

57 potens, Amphioplus . . . 52, 55
26, 27 praestans, Amphioplus . . 46, 47, 52, 53
46, 47 praestans, Amphiura .... 52

64 problematica, Theristus . . . .161

64 profundi, Amphiura . . . .11

49. 57. 64-65 protecta, Amphilepis . . . 40, 70, 72

. 214 protecta, Ophiomonas . . 37, 38, 39, 40,

. 214 41, 44. 7. 72

. 230 prytherchi, Camacolaimus . . 139, 155

218-220, 233 psammoides, Theristus . 139, 163-I&5

14, 15, 17, 20 psammoides, Trichotheristus 139, 163-I&5

20 psara, Amphiodia .... 24, 25, 26

19 Pseudechidna ..... 214

44, 57 Pseudobalistes ..... 232

. 280 psilochora, Amphiodia . . 23, 24, 25, 27, 57

280 psilochora, Amphioplus . . . 57, 60

239,240 psilopora, Amphiura . . . 11,12

. 214 Pterois . . . 231

. 239 Pterothrissus . . . . 130, 133

214 Ptychopleura .... 240,241

240 Ptychosaurus . . . 240, 241, 243

. 240 pulchella, Amphiodia . . .23, 26, 27

239 pulchella, Amphiura .... 23
73 pulcher, Ophiophragmus . . 26, 28

. 240 punctata, Plica .... 240, 243

240 punctatus, Hypsibatus . 240, 241, 243, 247
. 239 punctatus, Ptychosaurus . . 240, 243
. 240 puntarenae, Amphipholis . . 26

240 pycnostoma, Amphiura .... 72
. 240

219, 220, 233 Quisquilius ..... 228, 233

INDEX

reducta, Pomponema .
refecta, Amphioplus ' .
refecta, Amphiura .
refectus, Amphioplus
relicta, Amphiura .
relictus, Amphioplus
remittens, Amphilepis
repens, Amphiura .
reposita, Amphiodia
reposita, Amphiura
repositus, Amphioplus
repositus, Unioplus
rhabdota, Amphiodia
rhadinobrachius, Amphioplus .
rhombifer, Lophyrus
Rhombus ....
rhynchonemoides, Theristus
roscoffiensis, Theristus .
rosea, Amphiura
rosei, Eohiodon
rossica, Amphiodia
riipelli, Gymnothorax
riippelliae, Dalophis
riisei, Amphiodia .

Salarias ....

sarsi, Amphiura
saurus, Elops
scabriuscula, Amphiura .
sclateri, Callogobius
sclateri, Eleotris
sclateri, Mucogobius
Scorpaena ....
Scorpaenichthys
Scorpaenodes
sculptilis, Amphiodia
scutata, Amphilepis
securigera, Amphiura
securigera, Ophiopeltis
semicincta, Ophisurus fasciatus
semiermis, Amphiura
seminudus, Amphioplus .
semipunctatus, Asterropteryx
Serranus ....
servata, Amphiodia
servata, Amphiura
servatus, Amphioplus
servatus, Unioplus
sesquipedalis, Ophionephthys .
setiger, Leptolaimus
sexguttatus, Eleotriodes
sexguttatus, Eleotris
sexradia, Ophiocnida
signalis, Amphioplus
Silax

simonsi, Amphiura
sinensis, Ctenamphiura
singularis, Ophiocopa
sinicola, Amphiura
Sinolycoptera

139, 169-172

65
. 65

57.65

46. 47. 52

50,54
72

23, 26
39, 40, 46, 65

43. 49. 65
. 38, 39, 65

38, 39, 65

26, 28

57

239
232

. 161
. 161

12
. 260

28

. 215

. 215
26

229

. 13, 16, 17
130, 261
28

. 228
. 228
. 228
230, 231
. 187
. 230
26
72

. 12, 13, 16, 17, 18
13, 14, 15, 16, 17, 18
. 217
8,9,12
. 20,21,57
. 227

222

27, 37. 4. 6 5
48, 65

38, 39, 65-66
38, 39, 65-66
19

153
. 227

227
33
57
36, 37, 38, 39. 4. 43. 60

12

27

74

. ii, 14, 15, 17, 18
260, 275

siro, Acarus . 85, 86, 88, 89, 92, 93, 94, 95, 96,
97. 98, 99, 101, 102, 103, 104,
113, 114, 115, 116, 117, Pis. i, 4

spinulosa, Amphioplus
spinulosa, Amphiura
spinulosus, Amphioplus
spinulosus, Amphiura
spinulosus, Lymanella
spinulosus, Tropidurus
spirifer, Microlaimus
squamata, Amphipholis
squamata, Ophiura
steineri, Axonolaimus
stejnegeri, Plica

Stellio

stenaspis, Amphioplus . .
stepanovi, Amphiura
stepanovii, Amphiura
Stethojulis .
stewartensis, Amphioplus
stewartensis, Ophionephthys .
stimpsoni, Amphiura . .
stomachor, Pomponema .
storkani, Polybothrus fosciatus
stratus, Amphioplus
Strobilurus . . . ,
strongyloplax, Amphiodia
strongyloplax, Amphioplus
Stylophorous
subtilis, Amphipholis
summana, Epinephelus .
summana, Perca
sundevalli, Amphiura
superciliosa, Lacerta
superciliosa, Ophryessa .
superciliosa, Ophryoessa
superciliosa, Uraniscodon
Synanceja .
syntaracha, Amphiura .

tabogae, Amphiodia

Tarpon 121, 122, 125-129, 131,

tau, Opsanus

tenue, Ophiostigma

tenuis, Amphilepis

tenuis, Ophionephthys .

tenuis, Ophiopeltis.

tergisus, Hiodon

tesselatus, Amphioplus .

tetracantha, Amphipholis

textilis, Amphioplus

Thalassoma .

Theristus . . . .

Thrissops

thrombodes, Amphioplus .

thyrsoidea, Muraena . .

thyrsoideus, Gymnothorax

timsae, Amphiplus . .

transacta, Amphiacantha

trepidus, Amphioplus

52
52

46, 47, 50, 52, 66
66

46, 47, 66

241,255

139, 167-i6g

4. 7, 29, 30, 31

3i

139. 151

237. 241, 254. 255
239, 242

46, 47, 5L 57

. 226

21,57
14, 20

12
. 170

33

57

. 241
20

20, 44, 46, 47, 57
200
26

. 223
223

12
237, 239, 240

239
240

239
231-232

ii, 14, 16, 17

23, 25, 26
132, 133, 134,261

203
32, 33

72

14, 15, 17, 20
2O
260, 276, 279

57
3i

57
.226

139, 158-165
278, 279, 280, 281

44. 57
. 215
. 215

. 48, 5, 57
34
37. 58

INDEX

319

trichoides, Amphichilus .
Trichotheristus
trimaculata, Julis .
trimaculatus, Halichoeres
Triodia
Tripterygion
Tripterygium
Tropidurus .

trychna, Amphiodia
tuberculatum, Plica
tuberculatum, Uraniscodon
tumida, Amphiura
tumidus, Amphioplus
tymbara, Amphiodia
Tyroglyphus farris
tyrophagoides, Acarus

umbonata, Amphiactis
umbonata, Histampica
umbra, Agama
umbra, Hyperanodon
umbra, Hypsibatus
umbra, Iguana
umbra, Lacerta
Umbra, Ophryessa
umbra, Plica 237, 2;
umbra, Plica umbra
umbra, Pneustes .
umbra, Uraniscodon
umbra, Uranoscodon

umbra umbra, Plica
uncinatus, Ophiodictys
undulata, Gymnothorax
undulatus, Ecphymotes
undulatus, Muraenophis

39, 45, 4 6 > 49, 64

unicarinatus, Tropidurus 237,

241,248, 251, 253

139, 163

Unioplus . 36, 38, 39, 44

, 45, 48-49, 56, 57,

. 226

58-59, 60, 62-63, 65, 70, 71

. 226

uniovarium, Paralinhomoeus

139, 156-158

67-6Q

Uperanodon .

240, 241, 247

. 230

Uracentron .

. 241

230

Uraniscodon . . -237,

239, 240, 242, 243,

2 37> 2 39, 241, 242, 248,

247, 248, 252, 254

251, 253, 254, 255

Uranoscodon . -237,

239, 240, 241, 242

26

Uropterygius

. 216

237, 241, 248, 251, 252

urtica, Amphiodia .

26

>n . . 252, 254

vadicola, Amphiura

. ii, 14, 15, 17, 18

36, 56

vadicola, Fellaria .

20

36, 37, 46, 47, 50, 57

vadicola, Ophionephthys

20

26

vagabundus, Chaetodon .

. 225

113, 114, 117

verrili, Amphioplus

36

. 91, 92, 105

verrili, Silax .

37, 38, 39, 4, 43

verrucosa, Synanceja

231-232

69, 73, 74

verticillata, Amphiura

II, 12

74

vicina, Amphiodia

26

239, 240, 242, 247

vicinus, Theristus .

139, 165

247

vicinus, Trichotheristus .

139, 165

239, 240, 247

violacea, Amphiodia

26

238, 247

violacea, Amphiura

25

238, 240, 241, 247, 253

vitax, Amphiopholis

29

239, 247

vitax, Amphiura .

31

i, 240, 241, 243, 247-254

volitans, Exocoetus

. 198

251,252,253

vulpes, Albula

. 130

239, 247

Warwick, R. M. .

137-177

239, 240, 247, 248

wimmeri, Theristus

. 161

. 241

wurdemanni, Ophiophragmus

24, 25, 26, 50

237, 250, 251, 252, 253,
254

Xenopoecilus

199

251,252,253

zebra, Dendrochirus

. 231

73

zebra, Echidna

. 213

. 215

zebra, Gymnothorax

213

239

zebra, Pterois

. 231

215

Zeus ....

. 225

Printed in Great Britain by

AJden & Mowbray Ltd
at the Alden Press, Oxford