aed
Aer Te eee want ee ee Pe

2 et ee Nt Baw ee
Sanne

ap et araene OAPOP IO FOTN Oe ETE LY Ane ea en ee
oO ee ane eee Ra Cena e ee ne

a
Coe hs ee eho oe re

a nee eee
eae eee

94 8 a ete ee

+
Seo

tee ee tee ete

7eeee se

Pie Gl ea eee
BAe an) Eee i a? ee St «+ are 1 Bi». 7

Tem |

THE

VOYAGE OF H.M.S. CHALLENGER.
ZOOLOGY.

REPORT on the ie scatae collected by H.M.S. Challenger during the
Years 1873-76/ By wok. Brooxs, Associate Professor of Zoology
and Director of the Marine Laboratory of the Johns Hopkins
University, Baltimore, U.S.A. .

INTRODUCTION.

THE Stomatopoda are restricted to shallow waters, and as the small collection which was
brought home by the Challenger, and entrusted to me for examination, contains no
startling novelties, my first feeling, after my preliminary examination, was disappoint-
ment at the absence of any unfamiliar type, but this soon gave way to a feeling of excited
interest after the discovery that the material in my hands furnished the most ample
opportunities for tracing out, with great completeness, the phylogenetic and ontogenetic
history of this small and compact order of Malacostraca.

The order Stomatopoda includes about sixty species of adults, and an equal or greater
number of larve, from the tropical, subtropical, and temperate waters of the Atlantic,
Pacific, and Indian Oceans. Some of the species, like Gonodactylus chiragra, range over
the whole of this area, while others, like Squilla nepa, are distributed over the bottoms
between the coast of Chili on the one side and the coasts of China and Africa on the other,
or like Squilla empusa, between Rhode Island, U.S.A., and Africa. They are usually
found in very shallow water, and, with the exception of the specimen of Squilla lepto-
squilla taken in the trawl by the Challenger in the Celebes Seas from a depth of 115
fathoms, and a specimen of Lysiosquilla armata which S8. I. Smith found in the stomach of
a Lopholatilus from 120 fathoms, they are all from very moderate depths, and the wide
distribution of many of the species is undoubtedly due to the great length of their

(ZOOL. CHALL., EXP.—PART XLY.—1886.) ¥yi

2 THE VOYAGE OF H.M.S. CHALLENGER.

larval life, during which they swim at the surface and are swept to great distances
by the oceanic currents.

There are, however, many species which are known from only a single restricted
locality, and almost one-fourth of the species which have been described are represented
by solitary specimens, and there is therefore every reason to believe that many species,
and possibly many genera, are still unknown, and that our knowledge of the group is
very incomplete. They are extremely active in their movements and retiring in their
habits, and they may remain undiscovered in a locality where they are abundant. A few
species are recorded as dwelling in crevices in coral reefs, but most of them are burrowing
animals. The living animals which form their prey are captured in the long raptorial
claws, and some species, like Squilla empusa, often venture to a great distance from their
burrows in their pursuit of prey, and are frequently captured in nets and trawls, although
others, such as Lysiosquilla excavatrix, are the Myrmeleons of the ocean, lying in wait
for their prey, covered with sand, with only the tips of their eyes exposed, at the mouths
of their very deep burrows, to the bottoms of which they dart at the least alarm. At
Beaufort, N.C., Lysiosquilla excavatrix is so abundant that the mouths of several burrows
may often be found in shallow water in a square yard of the bottom, yet during the six
summers I have passed there, I have obtained only one adult specimen which was
captured outside its burrow, and only one which was obtained by digging. It was not
until I devised the plan of holding near the mouth of the burrow with one hand a piece
of bait, such as a small fish or a crushed crab, while the other hand was held ready with
a trowel to cut off the retreat to the bottom of the burrow, that I was able to procure
them in abundance, and the movements of this species are so very rapid that most of the
specimens were so near escaping that they were cut in two when the trowel was plunged
into the ground.

The Challenger collection of adults is a very small one, consisting of only fifteen
species, but eight of them are new, while two of the others, Squilla fasciata and Proto-
squilla (Gonodactylus) guerinii, have been very inadequately described from single
specimens. The importance of the collection must not, however, be estimated by its
size, for it throws light upon many interesting problems, and furnishes the material for
a more exhaustive and satisfactory discussion of the phylogenetic relationship and the
natural classification of the various genera and species than has been possible hitherto.

The collection of pelagic Stomatopod larve is very rich, and it has yielded the
material for tracing the history of several of the larval types, and also for establishing,
in every genus except one, the connection between the adults and their larval types.

The larval history of the Stomatopoda is one of the most puzzling problems
in morphology, and most of our knowledge of the subject is derived from Claus’s well-
known memoir.*

1 Die Metamorphose der Squilliden, Abhandl. d. k. Gesellsch. d. Wiss. Gottingen, Bd. xvi. pp. 1-55, pls. i.—viil., 1871.

yi

REPORT ON THE STOMATOPODA. ove

While I have been able to show, from the study of the Challenger specimens, that
many of Claus’s conclusions are incorrect, and that he referred some of the most familiar
larval types to the wrong adult genera, I feel that I could not have accomplished this
alone, and that, while my results are in many cases directly opposite to his conclu-
sions, my ability to reach and to prove them is due, in a very great degree, to the study
of his memoir. The labour of tracing the history of the larve has been so simplified
through the accurate illustrations and ample and minute descriptions which he has
furnished, that the investigator who follows him in this field has much of the difficulty
removed, and, while I feel that I leave the subject in a much more satisfactory condition
than that in which I found it, I also feel that I could not have made the same progress
without the aid of his memoir.

The beautiful transparent, glass-like pelagic larvee of the Stomatopoda are familiar
to all naturalists who have had an opportunity to study pelagic life, and none of the
animals which are captured at the surface in the tow-net exceed them in interest to the
student, or in beauty and grace. Their perfect transparency, which allows the whole of
their complicated structure to be studied in the living animal, their great size and
rapacity, the graceful beauty of their constant and rapid movements, and the pro-
fundity of the morphological problems which they present for solution, cannot fail to
fascinate the naturalist. Unfortunately they are as difficult to study as they are
beautiful and interesting, and, notwithstanding their great abundance and variety, only
two or three of them have been traced to their adult form.

Unlike most Malacostraca, the Stomatopods, instead of carrying their developing eggs
about with them, deposit them in their deep and inaccessible burrows under the water,
where they are aerated by the currents of water produced by the abdominal feet of the
parent, which are so shaped as to form valves or paddles which exactly conform to the
outline of the cylindrical hole, The eggs quickly perish when deprived of this constant
current, and as it is very difficult to procure them at all, I know of no young Stomatopod
which has been reared from an egg outside the burrow or in an aquarium. The older
larvee are hardy, and they thrive in small aquaria and moult into the adult form, but
they are seldom found near the shore, and microscopic research is so difficult in mid ocean
that almost nothing has been accomplished in this way. The younger larve are common
near the shore, but they seldom pass through a moult in confinement, and the only way
to trace the life-history of the Stomatopoda is therefore by the comparison of the series
of larve which are collected in the ocean, and this is attended with peculiar difficulties,
for the number of larval forms which have been described is much greater than the

number of adults which are known, and many of them unquestionably belong to unknown
species, and possibly to unknown genera.

The growth of the larve is very slow, and the larval life long, and as they are as
independent and as much exposed to changes in their environment, and to the struggle

4 THE VOYAGE OF H.M.S. CHALLENGER.

for existence as the adults, they have undergone countless secondary modifications which
have no reference to the life of the adult, and are therefore unrepresented in the adult
organism ; and a comparison of the various larvee which are here figured and described
will show that they differ among themselves more than the adults, thus reversing the
general rule that larvee are less specialised and exhibit clearer evidence of genetic relation-
ship than mature animals. The problem which they present is very similar to, but more
difficult than, that presented by the Hydro-meduse, for young Meduse can be reared
from the hydroids in aquaria without difficulty, and it is also easy to rear young hydroids
from the eggs of Medusze, but the life-history of the Stomatopoda must be traced from
the internal and indirect evidence furnished by comparison.

The Stomatopod larve present differences among themselves, and they may be
arranged in genera and species, but unfortunately their generic characteristics are quite
different from those upon which the adult genera are based, and this is true in a still
greater degree of their specific characteristics. As the larvee undergo great changes during
their growth, different stages have been described as distinct species or even genera, and
it is not easy to select from the rich gatherings which are brought home by collectors, the
successive stages in the history of a single species. Like the adults, they are widely dis-
tributed, and a gap in a series from the North Atlantic may be filled by a specimen from
the coast of Australia or the Sandwich Islands, and the collection from a single locality
may contain the larvee of several widely separated species of adults in all stages of
growth.

The attempt to unravel the tangled thread of the larval history of the Stomatopoda
is therefore attended with very exceptional difficulties, and the earlier writers were
content to rest after the bestowal of generic and specific names upon the larvee, and the
first writer to approach the subject in a scientific spirit was Claus, whose classical mono-
graph not only abounds in fundamental generalisations of the greatest interest and value,
but also contains nearly all that we know regarding the relationship between the larve
and their adults; but the Challenger collections, especially the rich collections of Alima
larvee, a group in which Claus’s collections were very deficient, furnish the material for
revision of the subject, and enable us to determine, with much greater certainty than
before, the larval type which pertains to nearly every one of the genera of adult
Stomatopoda, and also to give a pretty complete picture of the developmental history of
each larval type. As the specific differences between the adults are very slight, the
specific identity of each larva can be determined only by rearing the adult from the
larvee, but this fact renders it the more important that the series collected by the
Challenger should be figured and described, as later investigators will thus be enabled to
complete the history by keeping the final stage in each series alive in an aquarium until
it assumes the characteristic of the adult. This can be done without difficulty, as the older
larvee are hardy, while the fact that the younger larve will not live in captivity

REPORT ON THE STOMATOPODA. 5

frustrates the attempt to trace the gradual growth and metamorphosis of the larve in
this way, and there is no other resource except comparison.

The first step in this direction is to trace the history of each larval type, by the
selection and comparison of those larva which belong to the same series. In accomplish-
ing this I have been guided in part by general resemblances, but more especially by
comparative measurements. After I had tabulated the measurements in millimetres of a
number of specimens, which resembled each other quite closely, and formed a tolerably
complete series, I failed at first to trace through the columns of the table any such
conformity to a general law as I had expected, but more careful examination indicated
that this might be due to the fact that the history of the larva consists of metamorphosis
as well as growth, and that the size of one organ might, when compared with that of
another organ, show a gradual decrease during the successive stages, while its absolute
size was actually increasing. I therefore reduced all my measurements to a common
standard, and expressed them in thousandth parts of the total length of the larva at
each stage instead of in millimetres, and I found that this at once introduced order
where all had before been confused, and that, when thus reduced, the measurements
usually enabled me to decide with confidence whether a given larva does or does not
belong to a certain series.

In a few cases these comparative measurements gave proofs of specific identity which
could hardly be made more conclusive by rearing the larve. Thus the lengths of the
series of Coronis larvee shown in Pl. XIII. figs. 1-8 are as follows, and if the length of the
first stage be successively multiplied by five-fourths of itself, and this number by five-
fourths of itself again, and so on, we obtain the series of numbers given in the second line,
and as it is not conceivable that an accidental collection of larvee should exhibit such
exact conformity to a numerical law, we may feel certain that these larvae are genetically -
related, that they belong to one species or else to closely related species, and that the
series is consecutive, with the exception of one missing stage before the last.

4°16 mm, 5-29 mm. 6°49 mm. | Soe 10°21 mm.
£16 ”,; 5-20 ,, 6-50 ,, | 8-13 mm. 10-16 ,,

After one or two series had been traced out in this way, the general character of the
metamorphosis itself became a trustworthy guide for establishing the series for a closely
related species, and thus simplified the labour, and the next step was the reference of
each larval type to its proper adult genus.

If the differences between the larvz are due to secondary modification, we should not
expect the larve of two distinct adults to become modified in the same way, and
although it is of course possible that the larve of two closely related adults might

6 THE VOYAGE OF H.M.S. CHALLENGER.

become divergently modified, or that two adults might diverge from each other while
the larve remain alike, yet we should expect a natural or phylogenetic classification of
the larvee to stand in some definite and recognisable relation to the natural classification
of the adults.

My attempt to discover a relation of this sort at once brought me face to face with a
serious difficulty. In most of the published descriptions little attention is given to any
points which are not regarded as diagnostic, and the resemblances, which are of even
greater scientific interest than the differences, are often completely neglected; and
careful study of the published figures soon showed that they are untrustworthy so far as
relates to points which did not seem significant to the writers. Brevity and exactness of
diagnosis is of course desirable and essential to the ready identification of species, but
the description and identification of species is only a means for a more important end,
the ultimate discovery of the laws of life, and it is therefore desirable that every specific
description should consist of two parts, a brief diagnosis for purposes of identification,
and a complete description, or brief monograph, giving all the characteristics ; the points
of resemblance to allied forms, as well as the distinctive peculiarities.

The absence of this information renders the establishment of phylogenetic relation-
ships very difficult, and I soon found that the characteristics which are most significant
and of most scientific importance are by no means the ones which have been selected for
diagnosis. The analytical key which Miers’ gives is probably the best which could be
devised for ease of identification, and it expresses the general relationship between the
genera with sufficient accuracy for the purposes of the systematist ; but while most of the
genera which are usually recognised are natural ones, the points which are of the greatest
value in tracing the relation between the larve and the adults are entirely ignored in
most of the published diagnoses.

While there can be no doubt that the many differences between the Stomatopoda and
the other Malacostraca are of ordinal importance, all the species are included in a single
family, the Squillidee, and the differences between the genera are slight. Excluding the
genus Leptosquilla, Miers, which is very slightly known, and not represented in the
Challenger collection, six genera are usually recognised, Squilla, Chloridella, Lysiosquilla,
Coronis, Pseudosquilla, and Gonodactylus.

The study of the Challenger specimens shows the necessity for redistributing the
species which have been associated under the generic name Gonodactylus, and the
establishment in its place of three genera, Gonodactylus (sensu stricto), Protosquilla
n. gen., and Coronida n. gen., and also that it is impossible to draw any natural line
between Coronis and Lysiosquilla, or between Chloridella and Squilla, and I therefore
recognise seven genera, Protosquilla, Gonodactylus, Pseudosquilla, Coronda, Lysio-
squilla (including Coronis), and Squilla (including Chloridella). My comparison of the

1 On the Squillide, Ann. and Mag. Nat. Hist., ser. 5, vol. v. p. 2, 1880,

REPORT ON THE STOMATOPODA. 7

adults with the larvee shows that the marginal spines of the telson present features which
are of the greatest significance, and as I shall make frequent reference to these structures
it will be convenient to give in this place a short description of them. The telson of a
Stomatopod is usually furnished with six marginal spines (see Pl. I. fig. 3), which are
arranged in three pairs, and which I shall designate as the primary marginal spines.
The two nearest the middle line are the submedians ; the two nearest the anterior edge,
usually the farthest from the middle line, are the laterals; and the one between the
lateral and the submedian on each side is the intermediate. Between these six
primary marginal spines there are others which are equally large and prominent
in the young larvee, but minute or absent in the adults; these I call the secondary
marginal spines.

ANALYTICAL Key, GIVING THE MORE PROMINENT DraGNnostic CHARACTERISTICS
OF EACH GENUS.

I. Sixth abdominal somite fused with telson ; rostrum with acute median and antero-
lateral spines.
a. Dactyle of raptorial claw dilated at the base and unarmed ; hind body narrow
and thick ; marginal spines of telson crowded towards posterior edge.
Genus Protosquilla (Pl. XV1.).

II. Sixth abdominal somite distinct ; rostrum without antero-lateral spines.
a. Hind body narrow and thick. :

1. Dactyle of raptorial claw dilated at base, and unarmed; primary
marginal spines of telson very large, with one or two secondary spines
on each side between the submedian and the intermediate.

Genus Gonodactylus (Pl. XIV. fig. 1).

2. Dactyle of raptorial claw not dilated at base, usually armed with
marginal spines; submedian spines of telson tipped with movable
spinules; from one to three secondary spines between the sub-
median and the intermediate.

Genus Pseudosquilla.

b. Hind body depressed and wide.
1. Dactyle of raptorial claw dilated at the base and armed with marginal
spines.
Genus Coronida.

8 THE VOYAGE OF H.M.S. CHALLENGER.

2. Dactyle of raptorial claw not dilated at the base, but usually armed
with marginal spines.

(1) Primary marginal spines of telson small, with no more than
four secondary spines between the submedian and the inter-
mediate; outer spine of basal prolongation of uropod usually
longer than the inner; dactyle of raptorial claw with not
less than six marginal spines.

Genus Lysiosquilla (Pl. X. figs. 8-16).

(2) Primary marginal spines of telson large, with more than four
secondary spines between the intermediate and the sub-
median ; inner spine of basal prolongation of uropod longer
than outer; dactyle of raptorial claw usually with no
more than six marginal spines.

Genus Squilla (Pls. I., IL, TIT.).

In each of these genera there are certain characteristic or typical species, which are
sharply cut off from all other genera, but it is difficult to give any absolutely diagnostic
generic characteristics, as, in addition to the divergent and typical species, each genus
also contains a few species which are more primitive, with the characteristics of the
genus very slightly developed, and with features of resemblance to the primitive
species in other genera.

The form of the dactyle of the raptorial claw affords a ready means for distinguishing
species, and most of the genera are based upon peculiarities of this organ, which furnishes
a tolerably satisfactory index of relationship, but gives no clue to the wider and more
deep-seated affinities; but as soon as we ignore the preponderating importance which has
been attached to the big claw, and take the whole organisation into consideration, we
find that there are, in each genus, species which exhibit evidences of relationship to a
common type or ancestral form, from which the various genera have diverged, and which
was characterised by the possession of small, subcylindrical eyes, an acutely pointed
rostrum, a smooth hind body, a short wide smooth carapace, very small antennary scales
and uropods, and a telson which was wider than long, with the marginal spines crowded
backwards, and the posterior border transverse, or nearly so.

From this primitive form, which is represented at the present day, probably with
slight secondary modifications, by the various species which I have associated under the
generic name Protosquilla, the various genera have diverged, and while it is not at all
probable that any species which we know is the actual stem form of the order, yet there
is ample evidence to show that this was characterised by the features which the various
species of Protosquilla have in common, and that it must have been more nearly allied to
them than to any other species with which we are acquainted. Thus, for instance, the

REPORT ON THE STOMATOPODA. 9

more typical species of the genus Squilla, such as Squilla nepa, have broad triangular eyes,
an elongated carinated carapace with acute spines at its antero-lateral angles, longitudinal
carinz on all the abdominal somites, the appendages of the exposed thoracic legs filiform,

and the telson longer than wide, with well-developed marginal spines, of which only the
submedians are on the posterior border, and the others lateral. In Squilla lata (PI. III.
figs. 1, 2,3) the eyes are narrow and only slightly enlarged at their tips, the appendages of
the exposed thoracic limbs are flat, and the submedian carinz are absent or obsolete on
the first five abdominal and all the thoracic somites. In Squilla fasciata (PL TEE
figs. 4, 5) the eyes are cylindrical, not at all dilated at the tips, the appendages of the
exposed thoracic limbs are flat and strap-shaped, and the submedian dorsal carinze are
completely absent on the first five abdominal somites; while in Squilla chlorida
(Pl. IL figs. 1-5), which is not usually regarded as a Squilla, but is placed in a distinct
genus Chloridella, the eyes are constricted at the tips, the appendages of the exposed
thoracic appendages broad and flat, and the submedian carinze absent from all the
exposed somites except the sixth abdominal; and in Squilla (Chloridella) microphthalma
the eyes and appendages are like those of Squilla chlorida, but the dorsal surfaces of
the exposed somites are smooth, and their lateral edges but faintly carinated, and the
telson is wider than long, and the uropods small. In all these species the antero-lateral
angles of the carapace are acute, but in Squilla (Chloridella) rotundicauda we have, in
addition to the wide telson the small uropods and eyes and antennary scales of Squilla
microphthalma, a loosely articulated hind body, a carapace which is rounded in front as
well as behind, and the posterior margin of the telson rounded.

The series is so complete that it is quite impossible to draw any line to separate the
genus Squilla from the genus Chloridella, and all the species must therefore be
associated in a single genus Squalla. It is easy to find an answer to the question, which
term in the series of species of Squil/a is most primitive, for while the large multi-
earinated Squille are so different from the Lysiosquillz that there can be no question as
to their distinctness, it is difficult to discover any characteristics which shall separate
Squilla microphthalma from the latter genus; but we find that this species does not
resemble all species of Lysiosquilla to an equal degree, for the genus includes species,
like Lysiosquilla (Coronis) eacavatrix (Pl. X. figs. 8-16), which have, like Squilla microph-
thalma, small eyes and uropods, and flat wide appendages to the exposed thoracic
limbs; as well as species which, like Lysiosquilla maculata, have these appendages
linear and the eyes broad and triangular. We may therefore state with confidence
that Lysiosquilla maculata and Squilla nepa are more divergent than Lysiosquilla
excavatric and Squilla microphthalma, and that the two genera are divergent branches
from a common type, from which both genera have inherited the flat hind body and the
many-spined raptorial claw; and that this ancestral form had small eyes, antennary
scales and uropods, and a transverse telson.

(ZOOL, CHALL, EXP.—PART XLvY,.—1886.) Yy 2

10 THE VOYAGE OF H.M.S. CHALLENGER.

This ancestral form must have been very similar to two closely related living species,
Squilla bradyi, Milne-Edwards, and G'onodactylus trachurus, Miers, which are referred
by Miers to the genus Gonodactylus on account of the enlargement of the base of the
dactyle, and to Sguwlla by Milne-Edwards on account of the presence on the same
organ of marginal spines, and the flatness of the hind body. Their poimts of
resemblance to Gonodactylus are also points of resemblance to Protosquilla, and as they
differ from all the species of Gonodactylus in the flatness of the hind body, and the
presence of spines on the dactyle, there can be no doubt of the propriety of placing
them in a distinct genus, for which I propose the name Coronda. Coronida has, like
the convergent species of Lysiosquilla and Squilla, small eyes, antennary scales and
uropods, a flat hind body, an armed dactyle, and a wide rounded telson, and there can
therefore be no doubt of its close relationship to the ancestral type of these genera.

The species of Coronida, Pseudosquilla and Gonodactylus, are closely related, but not
in such a way as to indicate that any one genus is the ancestor of the others. The two
latter resemble each other, and differ from the first, in the fact that the hind body is
convex and narrow and bent downwards at the tip, while it is straight and flat and wide
in Coronida.

Gonodactylus and Coronida resemble each other and differ from Pseudosquilla in the
presence of an enlargement at the base of the dactyle of the raptorial claw, while
Pseudosquilla and Coronida resemble each other and differ from Gonodactylus in the
presence of marginal spines on the dactyle.

This triangular relationship can be accounted for only on the hypothesis that they are
the divergent descendants of an ancestral form which each one of them resembles in
certain features, to which, in each genus, secondary differences have been added.

As there is no reason to suppose that this divergence is recent, we should not expect
to find this ancestral form still represented by living species, but as the living species of
Protosquilla exhibit, like this hypothetical stem form, features of resemblance to each of
these genera, it is not only probable, but almost certain, that they are much more directly
descended from the ancestral form than any of the species of Gonodactylus or Pseudo-
squilla or Coronida; and, of course, than any of the Lysiosquille or Squillz. +

Protosquilla resembles Gonodactylus in the small size and flatness of its carapace, in
the presence of an acute spine on the rostrum, in the unarmed dactyle dilated at its base, —
in the height of the narrow hind body, the terminal somites of which are bent down-
wards, and also in the small size of most of the species.

It also resembles Pseudosquilla in most of these features, but the dactyle of
Pseudosquilla is without the basal enlargement, and is usually armed, like that of Coronida
and the Squill# and Lysiosquillz.

Coronida resembles Protosquilla in the small size of its eyes, antennary scales and
uropods, the flatness of the small carapace, the enlargement of the base of the dactyle,

REPORT ON THE STOMATOPODA. 11

and the presence of a median spine on the rostrum, while Protosquilla differs from all
other Stomatopoda in the length of the acute median spine of its rostrum, and in the
presence of acute long spines on its antero-lateral angles, and the union of the sixth
abdominal somite with the telson; and as it also exhibits, in its long rostrum, rudimentary
uropods, and in the absence of a distinct sixth abdominal somite, a closer resemblance
than any other adult to the Stomatopod larva, there can, I think, be no doubt that it is
the most primitive genus in the order.

I have attempted to give a graphic representation, in the following diagram, of the
relationships between the genera of adult Stomatopoda, the heavy lines indicating their
convergent relationship to each other and to Protosquilla. The characteristics which
are joined by brackets to two or more genera are the features of resemblance between
those genera, while those characteristics which are thus joined to only one genus are
confined to this genus.

The diagram is intended to express the relationship between the genera as established
by the comparative study of all the species; and as it often happens that a feature which
is highly characteristic of a genus as a whole may be absent or modified in a few
exceptional species, the diagram cannot be used as a means of diagnosis. For example,
the Lysiosquille have, as a rule, the outer one of the two spines on the ventral prolongation
from the posterior edge of the basal joint of the uropod longer than the inner, and I have
therefore given this feature as characteristic of the genus; for the study of the larvae shows
that it is very significant, although it happens that there are one or two species in which
the inner spine is the longest.

While I trust that the diagram is quite intelligible, a word of explanation may be
desirable ; thus, the genus Coronida is shown, by the brackets, to share with Chloridella,
Coronis, and Protosquilla, its small eyes, uropods, and antennary scales; with Uhloridella,
Coroms, and Pseudosquilla the armed dactyle; and with Protosquilla the small flat
carapace and acute rostral spine.

THE VOYAGE OF H.M.S. CHALLENGER.

12

te

——_
‘y[emMs speutuy

‘oynoe ATTensn wn74
“SOY yey pure 41oys
soedeieg *paarno
pues xeauoo fpoq
pay ‘pewreun pues
peyepp af Aqouqy
SS
‘snphyoppowoy
snuey)

wos]o} Josourds [urs
-reu ATeuI
-uoa Apoq putzy

; “WOS]O4
YIM =pasng oqimos = yrurut
“opqe qyxXIg ‘sourds eaoqey
“O1OJU OFNOV YATM WN.A4S0Iy

—<$< __________

“mynbsojoug snua)

ce
‘Teotrpurpoqns

Teus sodny yey

‘eu0U IO Moy) PUL yLOYs oovdeirn
‘eyeTpoul1ey UL
uvipouqns weeagoq|-woo Apoq purfy

pue] ‘pedo pue xoa

*XOA

Mica a
—

‘outds uvrpou
eynow’ UL WII uwMaysoy
‘yey «puke qeus oord
vlet) ‘Tews spodorn pure
seqvos Auvuueyuy “Two
“upuyAoqns ‘Tews soo
‘aseq 4v poyepipopAyouq

sourds yeuisavu Are ; a : We ;
-puoveg -oBrey £10 Dypnbsopnasy pore effjoug DpPwology
2 snes

snuoy)

SaaS See

“OpIA
pue yey Apoq pury
“pote ep Aqoeqy

Tews seyeos Areu
“ueyue pu spodoiq

‘xaunanoxa (s2u010)) nyynbsorshry. “SUT Uweyy dept ‘nuyoyrydosonu (nyaprLojy) nyjanBey

KL
‘qsosuoy AT7ensn podoan yo
eurds 103nQ + “wWerpomqns pue
oJIpPOMIEjUI weamjoq sourds
Avepuoses imoz ueyy ox0u
ON ‘Su0CT UeyY Opt wosToT,
‘qe pue yoous sovderen
‘yjoous Lpoq pur ‘sourds
XIS UvY} oLoML YYIM op{youq

Wosay, *[eorrpurpAo
-qus {yews sods

‘pedeys
«| -deaqs sqm orovr093

‘peyzeyp you
‘pourre eAqonqT = ‘opLd
‘nynbsorshT | pue yey poq pury

snue
9 “1eoul]

SUNT o1ovs0yy Jo sose
-pueddy “poyepyp sokgy

Js
.

“pyynby
snuex)

co
"4se5u0] podoan yo
autds Jouuy ‘uerpomqns pur
ayeIpowiezUT woeaajoq soutds
Avepuodses ino wey} oor TATA
uosjey, ‘seurds xis wey} erour
ou YIM Ayyensn opAjoeq

‘Oph WeTy TeSu0y
Wosfay, “peyeysoo Apoq purpy
\ poyeysoo ‘peyesuoye oovdviep

REPORT ON THE STOMATOPODA. 13

THE Accrssory CopuLaTory ORGANS OF THE MALE STOMATOPOD.

Before I end the discussion of the phylogenetic relationship between the genera of
adult Stomatopoda, I wish to call attention to the importance of figures and descriptions
of the complicated structure on the first abdominal appendage of the male. If each
description of a new species contained a figure of this structure, the tracing out of the
genetic relationship between the species would be greatly simplified; although a knowledge
of this organ would, of course, be of no help in our present undertaking, the discovery of
the connections between the pelagic larve and the adults. Unfortunately the Malaco-
logists have given little attention to this structure, as it cannot be studied to advantage
without removing it from the body and mounting it as a microscopic object.

In the case of the Challenger specimens Mr. G. B. Haldeman has kindly done this
for me, and has also made the drawings of this organ which are here given.

The endopodite of the first abdominal appendage of the male Stomatopod is furnished,
on its inner edge near the tip, with a complicated grasping organ which probably serves
for seizing the female, like the grasping forceps of many of the lower Crustacea and some
few of the Malacostraca, although I am acquainted with no observations as to their use
in the Stomatopoda. The structure has been figured by several observers, but no careful
comparative description has ever been published, although it often presents character-
istics of specific value, and differs conspicuously in the different genera. One of the most
interesting and valuable peculiarities of the organ is that it seems to illustrate the degree
of relationship between the various genera, and an exhaustive study of its modifications
in the different species will be a very great aid in tracing the phylogeny and genetic
relationship of the various genera and species. Unfortunately, several of the new species
in the Challenger collections are represented only by female specimens, and the material
is too scanty to afford an opportunity for the exhaustive study of the subject, but I am
able to give, with the descriptions of several of the new species, figures and descriptions
of this structure, and I hope that these descriptions will serve to incite in some one who
has access to larger collections a desire to make a more thorough study of it.

The endopodite of the first abdominal appendage of the male Stomatopod consists of
a large flat basal joint (Pl. I. fig. 2, A) separated by a movable transverse suture c¢
from the terminal joint B, which is notched or bilobed at the tip, and thus completely
or incompletely divided into an external leaflet a, and an internal one b. The latter
carries on the anterior surface of its inner edge, at the line of articulation with the proximal
joint, the petasma, or forceps, which consists of three portions: (1) the retinaculum or
appendix interna d, which is rounded at its proximal end, prolonged into a long
acute or subacute point distally, and with a straight internal edge which is closely
set with several crowded rows of hooked spines which interlace on the middle line
with those of the corresponding appendage on the opposite side of the body; (2) the

14 THE VOYAGE OF H.M.S. CHALLENGER.

movable limb f of the forceps, which is furnished with a large musele which runs up into
its hollow base, while its tip is spoon-shaped on its external surface which is opposed to
the third portion, the fixed limb, e, of the forceps. This is a raised ridge on the flat
surface of the appendage, with its tip free, and, in most cases, bent into a hook, which is
curved towards the spoon-lke hollow at the tip of the movable limb.

In Gonodactylus (Pl. XV. fig. 4) the terminal joint A is divided into two lobes, a
and b, by a marginal notch, but the inner lobe is not separated from the outer one by a
suture. Pseudosquilla has justly been regarded, by nearly all writers on the Stomato-
poda, as very closely related to Gonodactylus, and the modified appendage (Pl. XV.
fig. 10) is very similar indeed to that of the latter genus, as will be seen by comparing
the figure with that of Gonodactylus graphurus (P]. XV. fig. 8). In both, the inner
lobe of the terminal joint is larger than the outer one, and separated from it by a marginal
notch but not by a suture.

The close relationship between Squilla and Chloridella is unquestioned, and I have
shown that the species which have been divided between the two genera really form a
single phylum or stem which cannot be divided into two genera. A comparison of the
modified abdominal appendage of the male Squilla leptosquilla (Pl. I. fig. 2), or that
of Squilla quinquedentata (Pl. Il. fig. 6), with that of Squilla (Chloridella) chlorida,
will show that there is the closest resemblance. In all these the outer lobe a of the
terminal joint is separated by a suture from the inner lobe, the tip of which is elongated
considerably in advance of the tip of the outer lobe, although the more primitive rank of
Squilla chlorida is shown by the fact that the separation between the two lobes is much
less pronounced than it is in the more typical and highly modified Squilla. I have
shown that the species of Lyszosquilla and Coronis also form a single phylum, which it is
not practicable to divide into two genera, and the character of the modified appendage
is as similar in Lysiosquilla maculata (Pl. X. fig. 6) and Lysiosquilla (Coronis)
excavatrix (Pl. X. fig. 12) as it is different from that of Gonodactylus and Squilla. The
terminal joint B is subtriangular, and the suture ¢ which separates it from the proximal
joint is transverse, while it is oblique in the other forms. The outer lobe a@ is very
much larger than the inner b and triangular, while the inner lobe is very small and
nearly circular, and separated by a movable suture from the outer. The movable limb
of the forceps fis greatly elongated and is not bent at the tip, while the fixed limb e is
very small and rounded at the tip.

While no classification can ever be accepted as final, the one which I have expressed
in the preceding diagram is at least an approximation to the truth, and gives, so far as
our present knowledge enables us to trace it, the genealogy of the species of Stomatopoda,
so far as this can be established from the study of the adults.

In most animals we look to the larvee or embryos for evidence of genetic relationship,
more satisfactory and intelligible than any which is furnished by the adults, for it is a

REPORT ON THE STOMATOPODA. 15

general rule, in the animal kingdom, that the larvee or young of related species are less
divergent than the mature animals. Even if we were able to rear the larvee of the Stomato-
pods, and thus to use the evidence which they supply, this rule would not apply in this
case. The larval life is so long, and forms such a considerable part of the total life of
each individual, and the larve are so perfectly developed, and their relations to their
environment so complex, that there are about as many species of larva as of adults, and the
specific differences between them are fully as pronounced ; while the differences between
different genera of larvee are often greater than those between the genera of adults. The
fully grown larvz are in no sense embryonic or generalised ; they have no reproductive
organs, but in all other particulars they are just as highly organised as the mature
animals, and if the animals were to become sexually mature while retaining the organisa-
tion which fits them for their pelagic life, and if the final sedentary stage were then
dropped, we should then have an order of pelagic Crustacea of as high organisation, and
with as many well-defined genera and species, as the order Stomatopoda.

The larve may thus be treated exactly as if they were adults, and a natural or
phylogenetic classification of them established by the comparative study of their
organisation exactly as we have done for the adults.

As each larva is only an immature adult, or each adult only a fully grown larva, the
genetic history of each specific adult must be identical with that of some specific larva,
namely, its own larva.

If, then, comparative anatomy enables us to trace from the study of the adults of an
order or family or genus, their natural or genealogical classification, it must of course be
possible to do the same thing with the larvee, and if the classification which is established
is natural, there must be a discoverable relation between the one derived from the larvee
and the one derived from the adults.

In most cases this is unnecessary, as we are able to trace the young to its
adult form, and to use the whole life history as a basis for classification, and in most
cases it would also be extremely difficult, on account of the embryonic or generalized
character of young animals, and the absence of conspicuous specific differences, but it
fortunately happens that in the Stomatopoda, where we are compelled to resort to ~
this or some other indirect method for discovering what larva pertains to what adult, it
is also much more easy than usual, owing to the high specialisation and great diversity
of the larvee.

We cannot expect absolute agreement between the two classifications, for the sources
of our evidence can never be complete. We knew nothing of the larval types which may
have existed in the past, and next to nothing of the fossil adults, and it is very probable
that some of the larvee belong to unknown adults, and also that the larve of some of
the known adults are as yet undiscovered, and it is very probable that two allied adults
may have remained alike, while their larvee have been modified in two divergent directions,

16 THE VOYAGE OF H.M.S. CHALLENGER.

and that two allied species of adults may have originated in the same manner from a
common type, while their larvee have remained alike.

Still, after all these allowances, it still remains true that, inasmuch as the larve of
two closely related species are themselves more closely related by blood than the larvze of
more widely separated species, their bodily structure must exhibit a record of this relation-
ship which can be discovered by study and comparison, and which will agree to some
extent with the record presented by the organisation of the adults; the degree of
agreement depending upon the completeness of the two records and the correctness of
our interpretation.

What then is the natural or phylogenetic classification of the Erichthide or
Stomatopod larvee when they are studied by themselves and treated as adult animals ?

The genera of Erichthidee which have been recognised by the systematists are
Erichthoidina, Erichthus, Squillerichthus, and Alima, and of these four the first,
Lrichthoidina, is simply a young Erichthus, and the third, Squillerichthus, a fully grown
larva of the Erichthus type, so that the genera become reduced to two, Hrichthus and
Alima. Of these two genera, one, Alima, is much more sharply defined than the other,
Erichthus, which contains a number of divergent types which admit of definition.
Three of these types are represented in the collection by numerous species, and are well
known. As it will be convenient to have names for them I shall use names which
indicate the adult genera or subgenera to which they are to be referred.

The Gonerichthus type, which Claus correctly refers to the genus Gonodactylus; for
reasons which receive added weight from the study of the Challenger specimens, is shown
in Pl. XV. fig. 6.

The Pseuderichthus type, which Claus has given very conclusive reasons? for
regarding as the young of Pseudosquilla, is shown in Pl. XII. fig. 6.

The Lysioerichthus (Pl. XI. figs. 1-5), which Claus erroneously regards as_ the
young of Squilla,’ is, as I shall show, the young of Lysiosquilla. These three groups,
together with Alima (Pl. I. figs. 4, 5), which Claus regards as the young of
Lysiosquilla,* but which, as I shall show, is confined to the genus Squilla and diagnostic
of this genus, include nearly all the Stomatopod larve, although there are a few larve
which have a more isolated position, such as the one shown in Claus’s fig. 14, which I
shall designate as Hrichthalima, and others which are intermediate between the three
Erichthus types. The statement on p. 610 and footnote in Claus’s Grundziige der
Zoologie, that he has shown from the study of alcoholic specimens that Alima is the
young of Squilla will seem to conflict with my own statement that he regards Alima as
a young Lysiosquilla, but a reference to pp. 182, 138, 184, 185, 136, 138, and 154 of his
Monograph will show that he refers a number of Hrichthus larve to the genus Squilla,

1 Metamorphose der Squilliden, pp. 138 and 139. ? Metamorphose der Squilliden, pp. 140-146.
3 Metamorphose der Squilliden, pp. 132, 133, 134, 135, 136, and 138. + Metamorphose der Squilliden, p. 154.

REPORT ON THE STOMATOPODA. 17

and he gives at length, on p. 154, his reasons for regarding Alima as the young of
Iysiosquilla. It is true that he regards Lysiosquilla as a branch from the Squilla stem,
and that in this sense he does hold that Alima is a Squilla larva, but I suppose no one
would now regard Lysiosquilla as a Squilla.

A comparison of these larvee with each other indicates that they are all derived from a
primitive larval type which was hatched from the egg as an Hrichthoidina, and reached its
final form by gradual growth, and an increase in the number of somites and appendages,
without any sudden change or the retrograde development of any of its appendages. It
was furnished with a deep carapace, which, however, was not folded inwards at its
ventral edges, and it was probably armed with a number of secondary spines between
the submedian and intermediate-spines of the telson, and the edge of the carapace was
probably serrated. The most primitive among the recent adult Stomatopoda might be
expected to retain the most primitive larval type. We know of no fully grown larva
which can safely be referred to the genus Protosquilla, but the primitive larva must
have been very similar to what we should have if the Hvichthoidina shown in figs. 1
and 2 of Pl. XII. were to grow up and acquire its full number of somites and append-
ages, while the carapace and telson remained without change. The Gonerichthus larva
(Pl. XV. figs. 1, 3, 6,11) passes through an Hrichthoidina stage, its appendages undergo
no retrograde metamorphosis, the hind body is convex, the carapace is deep but not folded
inwards, there are no secondary spines, or only one or two on the telson between the
submedian and the intermediate, the primary spines of the telson are long, and the
outer spine of the basal prolongation of the uropod is very much longer than the inner,
and no specimens have ever been found with marginal spines on the edge of the long
slender dactyle. The last four characteristics are also characteristics of the adult
Gonodactylus.

In Pseuderichthus (Pl. XII. fig. 6) the carapace is deep and very slightly infolded
along its lateral edges, the hind body is convex, the dactyle of the raptorial claw of the
older larvee sometimes show traces of marginal spines, the outer spine of the uropod is very
much longer than the inner, and the primary marginal spines of the telson are very
long, the submedians are tipped with movable spinules, and there are only one or two
secondary spines between the submedian and the intermediate. None of these character-
istics are absolutely diagnostic of the genus Pseudosquilla, but as all except the last are
true of the adults of this genus, it is probable that all these larvze are Pseudosquille,
like the one the history of which has been traced by Claus. If this larva hatches as an
Erichthoidina it must undergo retrograde metamorphosis, since the younger larve have
no appendages on the third and fourth or fifth thoracic somites.

In Erichthalima’ the carapace is deep, and its lateral edges are infolded over the
ventral surface, and serrated, the hind body is flat, the telson is wider than long, there

* Claus, Metamorphose der Squilliden, fig. 14.
(ZOOL, CHALL. EXP.—PART xLv.—1886.) Yy 3

18 THE VOYAGE OF H.M.S. CHALLENGER.

are numerous secondary spines on the telson between the submedian and the inter-
mediate, and the dactyle of the raptorial claw is armed with several teeth. In this last
respect, and in the flatness of the hind body, it resembles Squilla and Lysiosquilla, and
it also resembles all the Squille, and differs from all the Lysiosqwille in the great
number of secondary spines on the telson, while it resembles the Lystosquille and differs
from all the Squille except the most primitive in having the telson wider than long.
It resembles the most primitive species of both genera in the small size of its
uropods, and Claus’s description renders it probable that the adult form to which it
gives rise has a long acutely pointed rostrum, thus differing from both Sgwilla and
Lysiosquilla.

The fact that some of its characteristics are shared by the adults of both these
genera, while others are confined to one and still others to the other, while still others
are not found in either of them, indicates that its adult may be equally related to but
distinct from both of them. Its relation to the Lysioerichthus and Alima larvee is of
precisely the same character.

The Lysioerichthus larva probably passes through an Erichthoidina stage, with
retrograde metamorphosis of the third, fourth, and fifth thoracic appendages. The
carapace is very deep, and in the older larve its lateral edges are folded inwards, and
they are serrated in the younger larvee; the dactyle usually bears traces of more
than six marginal spines, the hind body is flat and wide, the outer spine of the uropod is
nearly always longer than the inner, the telson is wider than long, and there are from
one to four spines between the submedian and intermediate, and sometimes a larger
number in the very young larva.

There is no evidence that the Alima larva ever leaves the egg as an Hrichthoidina.
The third, fourth, and fifth, as well as the three following thoracic appendages, are
wanting in the youngest larvae. The carapace is shallow and flat, and its lateral edges
are serrated throughout the whole larval life. The dactyle bears traces of marginal spines
which are never more than six in number, the hind body is flat and wide, the telson
is longer than wide, except in Alimerichthus, and there are numerous secondary spines
between the submedian and the intermediate.

In the flatness of the hind body and the presence of spines on the dactyle, Evrich-
thalmia resembles both Lysioerichthus and Alima. It resembles the very young
Lysioerichthus, and Alima at all stages, in the serration of the lateral edges of the
carapace, while it resembles all Lysioerichtht and differs from all Alimex except
Alimerichthus, in having the telson wider than long. It resembles all the fully grown
Lysioerichthi and differs from all Alimx in having the carapace very deep, with its
lateral edges infolded, and it differs from all Lysioerichthi and resembles all Alime in
having more than four spines between the submedian and intermediate spines of the
telson. Lysioerichthus has the outer spine of the uropod usually longest, while the

REPORT ON THE STOMATOPODA. 19

inner is always longest in Alima. In Hrichthalima they are equal, and both very short,
as is also the case in some of the more primitive Lysiosquillx, and in Protosquilla.

It will thus be seen that Erichthalima has certain characteristics which are found
nowhere else except in Alima or Squilla, and others which are found nowhere else except
in Lysioerichthus or Lysiosquilla, and others which are common to both, and others
which are found in neither. We must therefore regard it as a more primitive larva than
either, equally related to both.

In the Lysioerichthus series we have forms which, like Alima, have the rostrum and
postero-lateral spines long, and others which have them short. And in the Alima series
we have Alimerichthus which has its carapace deep and its telson wider than long, as in
Lysioerichthus, while the carapace is flat and the telson longer than wide in all the other
Alime.

The relationship between these various larvee may then be expressed in a diagram as
follows :—

Alima. Lysioerichthus with short spines.
\ Mh
\ Ve
\ WA
Ks Vs
Alimerichthus. Lysioerichthus with long spines.
Ve
i

Unknown Erichthoidina-like larva.

This classification exactly matches the one given for the adult Stomatopoda on
page 12, and, added to the fact that the few larvae which have been traced to their

20 THE VOYAGE OF H.M.S. CHALLENGER.

adult Stomatopods fall into their proper places, it furnishes very satisfactory proof of
the relationship between the larvee and their adults.

Faxon has reared a Squilla empusa from an Alima larva. The Challenger collections
show that the larva of Lysiosquilla maculata is one of the short-spined Lysioerichthi; I
have reared Lysiosquilla (Coronis) excavatrix, one of the more primitive Lysiosquille,
from a long-spined Lysvoerichthus. Claus has figured a series which shows beyond
question that at least one species of Pseuderichthus becomes a Pseudosquilla, and the
Challenger collections furnish equally good proof that the Chiragra group of Gonodactyli
come from Gonerichthus larvee, and I therefore believe that we may very safely assume
that all the Lysioerichthus larve are young Lysiosquille, all the Alima larve young
Squile, Alimerichthus one of the lower Squille or Chloridelle, all the Gonerichthi
young Gonodactyl, Erichthalima very probably a young Coronida, and all the
Pseuderichthi very probably young Pseudosquille. The evidence for this conclusion can
be much better estimated after the examination of the special descriptions which follow.

REPORT ON THE STOMATOPODA. 21

LIST OF SPECIES IN THE COLLECTION.

The small collection of adult Stomatopoda consists of the following fifteen species,
of which eight are new, while two of the others, Sgwilla fasciata and Protosquilla
(Gonodactylus) guerinii, have been only imperfectly described from single specimens.

As there are few adults which have been traced to their larve, and as I am able to
give, from the study of the living specimens, an account of the metamorphosis of an
undescribed species, Lysiosquilla (Coronis) eacavatriz, I have added a systematic
description of it, with illustrations, and hence the Report contains notes on the following
sixteen species:—

Squilla nepa, Latreille. Five males and two females from Pokoska, Japan, taken
with the trawl in 5 to 25 fathoms; one male from Kobé, Japan,
8 to 50 fathoms; one female from Reefs at Honolulu; one female
taken with the trawl in 15 fathoms, on the coast of Japan, at
Station 2338, lat. 34° 18’ N., long. 133° 35’ E.

Squilla quinquedentata, n. sp. One male from 28 fathoms, at Station 188, in the
Arafura Sea, lat. 9° 59’ 8., long. 139° 42’ E.

Squilla leptosquilla, n. sp. One male taken with the trawl in 115 fathoms in
Celebes Sea, near the Philippine Islands, in lat. 12° 28’ N., long.
22° 15' HB.

Squilla chlorida, n. sp. One male specimen from a depth of 15 fathoms at
Amboina, Moluccas.

Squilla fasciata, De Haan, One male and. one female from 15 fathoms, at Station
2338, in the Inland Sea, Japan, lat. 34° 18’ N., long. 133° 35’ E.

Squilla lata, n. sp. Two males and one female from 49 fathoms in the
Arafura Sea, New Guinea, Station 190, lat. 8° 56’ N., long.
136° 5’ E.

Lysiosquilla maculata, Miers. One mature male from Amboina, one half-grown
male from Samboangan, and a mature male and female from
Samboangan.

Pseudosquilla ciliata, Miers. One male and one female from 2 fathoms at St.
Thomas; one male from Reefs at Honolulu.

22

THE VOYAGE OF H.M.S. CHALLENGER.

Gonodactylus chiragra, Latreille. Numerous males and females from St. Thomas ;
one male from Bermuda; one male from Station 36, lat. 32° 7’ 25”
N., long. 65° 4’ W., near Bermuda; two from Samboangan, and
one from Samboangan Reefs.

Gonodactylus chiragra (var. minutus). Numerous male and female specimens from
near Cape St. Roque.

Gonodactylus (juv.). Perhaps a young Gonodactylus chiragra, from Station 208,
near Philippine Islands; 18 fathoms.

Gonodactylus graphurus, White. One male and one female from 8 fathoms, at
Station 186, near Cape York, in lat. 10° 30’8., long. 142° 18’ E.

Gonodactylus glabrous, nu. sp. One female specimen from the Reefs at Samboangan.

Protosquilla (Gonodactylus) elongata, n. sp. One female specimen from St. Vincent,
Cape Verde Islands.

Protosquilla (Gonodactylus) cerebralis, n. sp. One specimen from Fiji.
Protosquilla (Gonodactylus) guerinii, White. One female specimen from Honolulu.

Lysiosquilla (Coronis) excavatrix, n. sp. From Beaufort, N.C., U.S.A.

DESCRIPTIONS OF GENERA AND SPECIES.

Order STOMATOPODA.

Family SquiLtip&.

Genus Squilla, Fabricius.

Diagnosis.—Stomatopoda with the sixth abdominal somite separated from the
telson by a movable joint; the hind body depressed and wide; the dactyle of the
raptorial claw without a basal enlargement, but with a series of spines, which are not
usually more than six in number, on its inner edge; more than four secondary spines
between the intermediate and submedian spines of the telson, which is usually longer
than wide ; inner spines of ventral prolongation from basal jomt of uropod longer than
the outer. Larva an Alima with the ocular and antennulary somites not covered by
the carapace ; the lateral edges of the carapace not reflected, but usually bordered by
small spines or serrations; the inner spine of the ventral prolongation from the basal
joint of the uropod longer than the outer, and the telson with more than four secondary
spines between the intermediate and submedian spines.

Remarks.—I have had an opportunity to study the first abdominal appendage of the
male in only a few species of Squilla, but this organ has well defined common character-
istics in all these species, and we may probably add to the diagnostic characteristics
given above, the statement that the inner lobe of the terminal joint of the first abdominal
appendage of the adult male is longer than the outer, and separated from it by a suture.

While recognising the very close relationship between the genus Chlorida (Eydoux
and Souleyet) and the genus Squwilla, Miers thinks that the former genus is a natural
one, which should be retained under the modified name Chloridella’ to include the
Stomatopoda which have the dactyle of the raptorial claw armed with marginal teeth
and without a basal enlargement; the carapace and hind body without longitudinal
carine ; the eyes dilated in the middle and contracted at the tip; the rostrum short ;
the carapace small and short, and the appendages of the thoracic limbs strap-shaped,

1 Miers, On the Squillide, p. 13.

24 THE VOYAGE OF H.M.S. CHALLENGER.

while the genus Sqwilla is restricted by him to those species which, with a similar
raptorial claw, have longitudinal carne on the carapace and hind body; the eyes not
constricted at the tips; the carapace elongated, and the appendages of the thoracic
limbs slender and styliform.

The forms which he includes in the genus Chloridella are certainly less specialised
than the higher Squille, but the Challenger collections show that they are connected
with the latter by intermediate forms in such a way that it is impossible to draw a line
between them, and that they do not form two divergent branches, but a single series.
Squilla lata, n. sp. (Pl. III. figs. 1, 2, 3), is a Squclla, according to Miers’s definition,
while Squilla chlorida, n. sp. (Pl. II. figs. 1-5), is a Chloridella, but Squilla fasciata
(Pl. IIL. figs. 4, 5) is so very similar to both of these species that it is very hard to
distinguish from them, and it is intermediate between them in respect to the very
characteristics upon which Miers bases his genera. We must therefore enlarge the genus
Squilla to include the Chloridelle.

Ontogeny.—The Alima larva is one of the most sharply defined larval types, and we
have every reason to believe that all the larve in this group pertain to closely related
adults. As one of them has been kept by Faxon in an aquarium until it changed into
a young Squilla, and as all the species of the genus Squilla agree with each other in
several features which are not united in any other adult Stomatopod ; the flatness of the
hind body, the small number of marginal spmes on the dactyle, the great number of
secondary spines on the telson between the intermediate and the submedian marginal
spines, and the greater length of the inner one of the two spines on the basal prolongation
of the uropod; and as all the Alima larve, including Alimerichthus, agree with each
other, and differ from all other Ervichthide except the anomalous Erichthalima, in
similar features, we can state with confidence that all Alima larve are young Squille,
and that all Sqwilla larvee are Alime.

While the Alima is a highly specialised larva it is, in a certain sense, embryonic, for
the fully grown Alima closely resembles the young Lysioerichthus larva, as may be
seen by comparing fig. 4 of Pl. I. with fig. 5 of Pl. XII. The Hrichthus, in some
cases and probably always, hatches from the egg as an EHrichthoidina, while it is
probable that all the Alimz leave the egg in the Alima stage; but this is so similar to
the young Erichthus that Claus was disposed to regard his Erichthus multispinosus as
an Alima, although the fully grown EHrichthus is very different from the Alima at any
stage of its development. Apparently the stage which the Lysioerichthus passes
through, immediately after the Hrichthoidina stage, has proved to be so well adapted to
the needs of the Sqwilla larva that it has been lengthened at both ends of the larval
life until both the initial Hrichthoidina stage and the final Squillerichthus stage have
been crowded out of its larval life, and the Alima hatches as an Alimo and remains an
Alima until it changes into a Squilla.

REPORT ON THE STOMATOPODA. 25

The relation between the metamorphosis of Lysiosquilla and that of Squilla is
therefore something like this—

Lysiosquilla Squilla
adult. adult.
|
Squillerichthus.
| |
Young Alima-like Erichthus. , Alma.

Erichthoidina.
!
|

Ege. Ege.

Squilla nepa, Latreille.

The Challenger collection includes ten specimens of this species: seven of them, five
of which are males and two females, from Pokoska, Japan, captured in the trawl in from
5 to 25 fathoms of water; one male specimen from Kobé, Japan, in 8 to 50 fathoms; one
female specimen from the Reefs at Honolulu ; and one female specimen taken in the trawl
at Station 2338, Inland Sea, Japan; May 26, 1875, lat. 34° 18’ N., long. 133° 35’ E
15 fathoms ; bottom, blue mud.

The tip of the first abdominal appendage of an adult male, Squilla empusa, is
shown in Pl. II. fig. 7; and that of Squilla nepa is almost the same.

3

Measurements.

No. 1, Large male. No. 4, Smaller male.
No. 2, Male a little smaller. No. 5, Large female.
No. 3, Small male.

(ZOOL, CHALL, EXP,—PART XLV. —1886.) Yy'4

26 THE VOYAGE

OF H.M.S. CHALLENGER.

Measurements.

Total length from tip of rostrum to tip of telson,

Rostrum,
Carapace,

Carapace including rostrum,

From posterior edge of carapace to posterior edge of
second thoracic somite, .

From posterior edge of second thoracic to posterior

edge of third thoracic somite,

From “posterior edge of third thoracic to posterior
edge of fourth thoracic somite, .

From “posterior edge of fourth thoracie to posterior
edge of filth thoracic somite, : .

First abdominal somite,

Second abdominal somite,

Third abdominal somite,

Fourth abdominal somite,

Fifth abdominal somite,

Sixth abdominal somite, .

Telson on middle line,

Total length of hind body,

Total length on middle line from tip of
rostrum to tip of telson,

Greatest length of telson,

Width of carapace at antero-lateral angle,
Width of carapace at postero-lateral angle,
Width of second thoracic somite, 5
Width of third thoracic somite,

Width of fourth thoracic somite,

Width of fifth thoracie somite,

Width of first abdominal somite,

Width of second abdominal somite,

Width of third abdominal somite,

Width of fourth abdominal somite,

Width of fifth abdominal somite,

Width of sixth abdominal somite, 4
Width of telson between postero- -median spines,
Width of telson between postero-lateral spines,
Width of telson (greatest), : , 5
Total length of first antenna, . .
Length of scale of second antenna,

Length of eye,

Width of eye, .

.|No, 2.

5°05

‘14
113

1:27

16
lz

Squilla quinquedentata, n. sp. (Pl. I. fig. 3; Pl. IL. fig. 6).

Diagnosis.—Dactylus of raptorial claw with five teeth, including terminal one.

In inches and decimals. In thousandths of total length.
No. 3. |No. 4. |No. 5. {No. 1.|/No. 2.)No. 3. |No. 4.|No. 5.
4°26 | 2°56 | 5°53 waa Bd ted
13 “09 14 028) :031 "35 | °026
97 boy) Lez 223 228 | -207| -219
1:10 62 | 1°31 2°51 | --259 242 | +245
13 10 20 032; °-031] :040| :037
13 10 20 034; °031| °040) -037
18 ny) 24 040] -042) -048] -085
‘18 12 "24 039} °042| *048|] °035
‘27 ‘20 "45 073 | °065 | *079| -085
27. 20 “40 O71} °065| ‘079| °075
27 18 “40 070} °065} -070) -075
31 16 | | 438 ‘070 | ‘073} *064| -080
85 22 47 082! -°082) ‘070|] ‘088
28 16 “Bl 066 | *066| *064| “058
“719 “38 88 184] -186| °160} ‘165
3°16 | 1°94 | 4°22 ‘761 748 | °762| -770
4°26 | 2°56 | 5°53 | 1°078 | 1°012 | 1-007 | 1:004 | 1-015
92 46 | 112] 190} -216) ‘179| °207
“oz 28 “64 114] 122] :109) 120
“84 | “44 | 1:00 180} 197} ‘172]| 187
SOLE gar 90 6 GAB een 168
“74 rs 96 176 | 174 Ad 184
‘78 45 | 1:00 194} +183) °176| -187
FO (ets Oe “1sd)| 178)" =.. 172
*89 250) | 1:20 210] 209) °195| °224
an ae 1:21 we us aad 233
Sh - 1°24 ans Bo 233 |
mee as 1-25 aoe aay 49 "224

1°02 “59 | 1°26 "238 | °240] ‘231] °226
88 "48 | 1:11 *202| :207) 187] :208
"24 12 30 059 | +056] -047| ‘056
80 | °30 “90 174] 188] ‘117]| ‘168
90 “49 | 1°10 206} :211)) “191 |) +206
oe ve 1°80 "AAD se. i 336
66 “38 e's "147 | °155) :148 a
tas sis 24 a ne a0 045
| "25 047 |
Free

somites of hind body with submedian dorsal carinze, which are absolute in the four
exposed thoracic and the first four abdominal somites, as are also the next pair in the

thoracic and first three abdominal somites.
Lateral margins of first three exposed thoracic somites bilobed.

General Description.—Rostral plate nearly rectangular, a little longer than wide,
with antero-lateral angles rounded, and lateral edges raised into ridges or marginal

Appendages of exposed thoracic legs slender.

REPORT ON THE STOMATOPODA. 27

carinze which are continued forwards as far as the rounded angles. Between these
marginal ridges the dorsal surface of the rostrum is smooth, without a median carina,
with the posterior half pigmented (in the alcoholic specimen) and separated by a
semicircular line, convex in front, from the anterior colourless half. Antero-lateral
angles of carapace ending in very short acute spines ; postero-lateral angles rounded and
not prominent, and without a marginal notch. The carapace has a median longitudinal
carina which is very slightly raised, although its dark colour renders it conspicuous. It is
bifurcated at its anterior end and is interrupted at the transverse cervical suture, behind
which it is bifurcated to its posterior end, where the two divisions meet in a round median
tubercle on the posterior edge of the carapace. A lateral carina, marked with dark
pigment, runs from the antero-lateral spine on each side nearly parallel to and close to
the lateral edge of the carapace, as far as the postero-lateral angle, on which it ends
abruptly, without uniting with the transverse pigmented ridge which fringes the posterior
border of the carapace.

A second pigmented carina runs backwards from the antero-lateral angle for about
half the length of the carapace, internal to the submarginal keel, and nearer to it
anteriorly than posteriorly. At its posterior end it is about half way between the edge
of the carapace and a strongly marked convex ridge which lies in the longitudinal gastric
suture, between the elevated convex median gastric area and the lateral area. The
gastric sutures and their ridges are slightly divergent posteriorly, and they are interrupted
at the transverse cervical suture, behind which the cardiac sutures are marked by similar
elevated ridges, which are slightly incurved at their anterior ends, behind which they are
slightly convergent posteriorly. External to the cardiac sutures there is, on each side, a
short pigmented carina, which fuses with the pigmented border of the posterior edge of
the carapace, which latter is slightly emarginated, with a short acute median tooth.

The second, third, and fourth thoracic somites have each a median slightly impressed
white line, and on the sides of this a pair of pigmented submedian longitudinal carine,
which are very short on the second, and which are clearly marked by their dark colour,
but are scarcely elevated above the general surface. The fifth thoracic somite has no
impressed median line, and the submedian carinz are like those on the preceding somite.
The submarginal carine of the thoracic somites are scarcely elevated, and they are
marked by dark pigment, as are also the posterior borders of all the thoracic and
abdominal somites. The lateral processes of the second, third, and fourth thoracic
somites are bilobed; the anterior lobe of the first is elongated, curved forwards and
acute, while the posterior lobe is much shorter and subacute. On the lateral process of
the third thoracic somite the anterior lobe is shorter and subacute, and on the fourth
somite it is still shorter, while in both the third and the fourth somites the posterior lobe is
the larger, and its postero-lateral angle is acute on the third, and subacute on the fourth.
The fifth thoracic somite has only a single subacute lobe on each side. ‘The first five

28 THE VOYAGE OF H.M.8. CHALLENGER.

abdominal somites have eight keels each, including the marginal ones, and the sixth
abdominal somite has six. In all the abdominal somites the postero-lateral angles end
acutely in spines, as do the posterior ends of all the keels on the sixth and fifth, and all
except the submedians of the fourth. There are no median keels between the carapace
and the telson, but the third, fourth, and fifth abdominal somites have median tubercles.
The submedian keels of the fifth abdominal somite diverge posteriorly, while all the keels
of the sixth abdominal somite converge posteriorly. The first lateral keels of the first,
second, and third abdominal somites are slightly sinuous. The telson has a very prominent
median dorsal carina, which is situated upon the middle line of an elongated convex
protuberance. The carina ends posteriorly in a prominent acute spine, and it is bordered
by dark pigment at each end and in the middle. The margin of the telson carries six
acute spines, with a short dorsal carina running forwards from the base of each of them,
and a similar carina, without a spine, on the anterior portion of the lateral margin. The
submedian spines are straight, the intermediate curved, and the externals curved and
very short. There is a single rounded lobe or tooth between the base of the external
spine and the base of the intermediate spine on each side, and eight on each side between
the base of the intermediate spine and the base of the submedian, and six between the
submedians, three of them on each side of the deep median notch, which is closed
behind and converted into a foramen in the single specimen which was obtained. On each
side of the broad median longitudinal ridge the dorsal surface of the telson is marked by
shallow pits symmetrically arranged in curved lines, running outwards and backwards
from the median ridge to the notches between the rounded lobes on the posterior border
of the telson. The ventral prolongation from the basal joint of the uropod ends in a
short curved outer branch, and a much stouter and longer imner branch which carries a
blunt rounded tooth on its outer edge about half way between the base and the tip, and
a number of irregular serrations along its inner edge.

The outer edge of the second joint of the exopodite of the uropod is fringed by a
series of seven movable spines, the first six of which increase regularly in size towards
the tip, while the seventh is much larger than the sixth, and is terminal. The eye is
broadly triangular, and its length from its union with the short stalk to the rounded
posterior angle of the cornea is about equal to its width between the two rounded angles
of the cornea, which is divided into two halves by a transverse vertical furrow. The
ocular somite is exposed on the middle line. The dactylus of the raptorial second
maxilliped is divided into five curved acute teeth, and the keel on the anterior edge of
the third or carpal joint is smooth and entire.

The movable ramus f (Pl. II. fig. 6) of the forceps on the first abdominal
appendage of the male is very large, and four times as long as the fixed ramus, which
is bent into a single hook at its tip.

REPORT ON THE STOMATOPODA. 29

In inches and | In thousandths
Mipesarsmaenis decimals. of total length.
Total length from tip of rostrum to tip of median spine of telson, : 5°37
Measurements along middle line :—
Rostrum, ; é : 30
Carapace, . > : : 2 ; : : 1:28 238
Total length of rostrum and carapace, . : : i 1-44 268
From posterior edge of carapace to posterior edge of second thoracic
somite, . ‘13 24
From posterior edge of second to ‘posterior edge of third thoracic
somite, . “24 45
From posterior edge of third to posterior “edge of fourth | thoracic
somite, 22 4]
From posterior edge of fourth to posterior edge of fifth "thoracic
somite, . ; : é : : 24 45
First abdominal somite, ; 4 : F : ‘ “40 74
Second abdominal somite, . : ; : : : 36 67
Third abdominal somite, . ; ; : : ‘ “36 67
Fourth abdominal somite, . 5 ; ; : ; ‘40 75
Fifth abdominal somite, . : : : ‘ : 50 93
Sixth abdominal somite, . é F F : : “34 63
Telson on middle line, d : ‘ F : ; ‘74 138
Total length of hind body, : - : , 3°93 732
Total length on middle line, ; ; - : 5°37 1000
Greatest length of telson, . ; : : 96 179
Width of carapace between antero- lateral spines, ; : : D8 98
Greatest width of carapace, .. ; ; : : 2 “94 175
Width of second thoracic somite, ‘ : : , , 88 164
Width of third thoracic somite, ; ; : ; f “90 168
Width of fourth thoracic somite, : < ; . : 92 171
Width of fifth thoracic somite, : F ‘ : : 92 171
Width of first abdominal somite, F F ‘ : 3 1:08 | 201
Width of second abdominal somite, . ‘ ‘ : a 1:16 216
Width of third abdominal somite, . j 3 ; 3 1:16 214
Width of fourth abdominal somite, . : : : : 1:16 216
Width of fifth abdominal somite, : , ; : : 1:16 216
Width of sixth abdominal somite, . ; ) 5 ‘98 183
Width of telson between postero-median spines, : : : "22 4]
Width of telson between postero-lateral is : : : 82 153
Width of telson (greatest), . : : ; 1:04 194
Length of first antenna (shaft), : ; - 5 , 1:00 189
Length of first antenna (flagellum), . : : : A 90 168
Length of first antenna (total), . 2 1:90 357
Length of flagellum of second antenna, including shaft, : é “96 179
Length of scale of second antenna, . : : 88 164
Length of eye, - : : : : , : “20 37
Width of eye, : ‘ ‘ : ; 2 ; "23 43

30 THE VOYAGE OF H.M.S. CHALLENGER.

Habitat.—Station 188, September 10, 1874, Arafura Sea; lat. 9° 59’S., long. 139° 42’ E.;
depth, 28 fathoms; bottom, green mud. A single male specimen.

Remarks.—Notwithstanding the fact that it has five teeth instead of six on the
dactylus of the raptorial claw, and the very slight development of the longitudinal carinz
on the hind body, this species bears such a very close general resemblance to the well
known Squilla nepa, that I was at first inclined to regard it as a smooth specimen of this
species, with an accidental variation in the number of teeth on the dactylus, and the more
especially as Squalla nepa is known to occur in the region where the specimen was
obtained. More careful examination, however, reveals so many minute points of difference
in all parts of the body that I think there can be no doubt of its distinctness. The
following, which are the most conspicuous differences, will serve to distineuish it with
readiness from the latter species. Squwilla nepa has six and Squilla quinquedentata five
spines, including the terminal one, on the dactylus of the raptorial claw. The carine on
the anterior edge of the third joint or carpus of the raptorial claw of Squilla nepa is
irregularly dentated, while it is entire and smooth in Squilla quinquedentata. The
rostrum of Sqwilla nepa narrows slightly towards the tip, while that of Squilla quinque-
dentata is more nearly rectangular. The longitudinal cardiac sutures on the posterior
portion of the carapace are divergent posteriorly in Squilla nepa, and convergent in
Squilla quinquedentata. The submedian carine of the fifth abdominal somite are
parallel in Squilla nepa, and divergent posteriorly in Squilla quinquedentata.

Squilla leptosquilla, n. sp. (Pl. I. figs. 1, 2).

Diagnosis.—Ocular segment in front of rostrum. Eyes broad and sub-tri-
angular. Exposed segments of hind body with submedian carine. Dactylus of raptorial
claw with four spines, including the terminal one. Antennulary somite very long,
reaching to tip of rostrum on middle line, and with its antero-lateral angles prolonged
forwards into acute spines. First thoracic somite partially exposed dorsally. Lateral
margins of second, third, and fourth thoracic somites acute. Telson with a long slender
median spine on dorsal surface, and three pairs of marginal spines, of which the second or
intermediate pair are much the longest, and the external pair shorter than the submedians,
and blunt, while the others are acute. First antennee very long.

General Description.—Rostrum nearly twice as long as wide, subtriangular, with a
rounded anterior end. Width of carapace between antero-lateral angles about 5%, of its
greatest width, which is equal to about two-thirds (44) of its length. The length of the
carapace is 7%%%5 of the total length of the body, and its antero-lateral angles are produced
forwards as long acute spines, while the postero-lateral angles are prominent and broadly
rounded. The gastric area of the carapace is convex and elevated, with an obscure

REPORT ON THE STOMATOPODA. 31

median carina, which is continued forwards on to the rostrum, and is interrupted behind
by the strongly marked cervical sutures, posterior to which the median carina is
continued to the posterior edge of the carapace, where there is no median tubercle. There
are two longitudinal carine on each lateral portion of the carapace, a submarginal one,
which lies near the base of the antero-lateral spine, along nearly the whole length of the
carapace parallel to and near the thickened lateral margin, and a second or internal one,
which lies about midway between the first and the gastric suture, and comes to an end
in front of the transverse cervical suture. The thickened ridge which borders the lateral
margin of the carapace and the rounded postero-lateral lobe bends inwards around the
posterior margin of the latter, and running forwards on the dorsal surface of the carapace,
forms a submedian carinz parallel to and just outside the cardiac suture, and extend-
ing forwards as far as the transverse cervical suture. The posterior edge of the carapace,
which does not completely cover the first thoracic somite, is nearly transverse.

The second, third, fourth, and fifth thoracic somites have each four longitudinal
dorsal carinz, the submedian on the second being very short and concave towards the
middle line ; the first five abdominal somites have each six longitudinal carinz besides
the marginal ones, and the sixth sumite six im all, which all end posteriorly in acute
spines, while there are no spines on the submedians of the thoracic and first five
abdominal somites, nor on the first lateral carinze of the thoracic and first abdominal.

The marginals and submarginals of the first five abdominal somites, and the inter-
mediates of all except the first abdominal, end in spines. The submedian dorsal carinz
are parallel in the first eight exposed somites, divergent posteriorly in the fifth abdominal
somite, and convergent posteriorly in the sixth abdominal somite. While the somites
increase slightly in width from in front backwards the space included between the second
pair of carine is of uniform width from the second thoracic to the fifth abdominal somite,
and these carine form the prominent parallel ridges running along the whole length of
the abdomen, although the carina itself is interrupted by a notch near the anterior edge
of each abdominal somite except the sixth. The second, third, fourth, and fifth abdominal
somites have each a median dorsal double tubercle. The lateral processes of the thoracic
somites are much like those of Sguwilla mantis. The lateral margin of the second is
prolonged with a long straight acute lobe, while the third and fourth are obliquely
truncated, the third more obliquely than the fourth, and end behind in acute points.
The fifth is furnished at its antero-lateral angle with a small subacute process or tooth.
The telson ends in six marginal spines, of which the submedians are acute and triangular,
the intermediates very long, curved, slender, and acute, and about twice as long as the
submedians, and the laterals very short, about half the length of the submedians, rounded
and subacute. The tips of the long median spines are in the same transverse line with
- the tips of the submedians, while their bases are in the same transverse line with the
base of the unpaired median spine. There is a single rounded dentation between the

By) THE VOYAGE OF H.M.S. CHALLENGER.

base of the lateral spine and the base of the intermediate, and between the base of the
intermediate and the tip of the submedian eleven or twelve small acute serrations, while
there are about the same number of similar but much smaller serrations between the tip
of the submedian spine and the middle line. The dorsal surface of the telson has a
median longitudinal ridge which ends behind in a long slender acute spine. On each side
of the ridge there is a longitudinal row of five or six shallow circular pits, from each one
of which a faintly marked curved line runs outwards and backwards to the posterior
margin. The endopodite of the uropod is long, slender, curved, and about eight times as
long as wide, and the elongated oval paddle of the exopodite is about equal in length to
the second joint. The ventral prolongation from the posterior edge of the basal joint ends
in two slender acute curved spines, of which the inner is much (about 3) longer than the
outer, with a small tooth on its outer, and a series of minute serrations on its inner edge.

The eyes are broadly subtriangular, and the distance from the rounded posterior
angle of the cornea to the point where the eye joins the constricted stalk is equal to the
distance between the angles of the cornea, which latter is divided into two nearly equal
lobes by a vertical groove. The ocular somite lies entirely in front of the rostrum.
The first antenna has a very long slender shaft, and its total length to the tip of the
internal or longest flagellum is a little more than half (45>) the total length of the body
from the tip of the rostrum to the tip of the telson. The anterior edge of the somite
which carries the first antenne is under the tip of the rostrum, and its antero-lateral
angles are produced forwards as slender acute spines. The raptorial second maxilliped is
long, and its dactylus has on its inner edge four spines, including the terminal one, which
are acute, curved, and gradually increasing in size distally. On the outer edge of the
dactylus near the base there is a small process. The three pairs of exposed thoracic legs
are short and slender, and their appendages are filiform. The endopodite of the first
abdominal appendage of the male (Pl. I. fig. 2) is elongated and nearly rectangular,
and the inner edge is straight. The outer lobe a of the terminal joint is much shorter
than the inner lobe b, which is separated from it by a suture. The fixed limb of the
petasma e is very long, reaching nearly to the tip of the mner lobe, and its tip carries
on its inner edge two hooks or lobes, one of which points towards the tip and the other
towards the base of the appendage.

Habitat.—The Challenger collection includes only one specimen, a male, from
Station 2048, lat. 12° 46’ N., long. 122° 10’ E., in the Celebes Sea, near the Philippine
Islands, taken with the trawl from a depth of 115 fathoms; bottom, green mud.

REPORT ON THE STOMATOPODA. 33

In thousandths
of total length.

In inches and
decimals.

Measurements.

On middle line :—

Rostrum, 36
Carapace, 209
Carapace, including rostrum, 245
From posterior edge of carapace to posterior edge of second thoracic
somite, 36
From posterior edge of second to posterior edge of third thoracic
somite, . 47
From posterior edge of third to posterior ‘edge of fourth thoracic
somite, 46
From posterior edge of fourth to Posterior reat of fifth “thoracic
somite, . : : 50
First abdominal somite, 72
Second abdominal somite, . 72
Third abdominal somite, (fe
Fourth abdominal somite, 65

Fifth abdominal somite,
Sixth abdominal somite,
Telson on middle line,

Total length of hind body,
Total length on middle line,

Greatest length of telson,

Width of carapace between antero- lateral angles,
Width of carapace (greatest), .

Width of second thoracic somite,

Width of third thoracic somite,

Width of fourth thoracic somite,

Width of fifth thoracie somite,

Width of first abdominal somite,

Width of second abdominal somite,

Width of third abdominal somite,

Width of fourth abdominal somite,

Width of fifth abdominal somite,

Width of sixth abdominal somite,

Width of telson between submedian terminal spines,
Width of telson between intermediate terminal spines,
Width of eyes between angles of cornea,
Length of eyes, :

Length of basal joint of first antenna,

I ength of second joint of first antenna,

Length of third joint of first antenna, .

Length of flagellum of first antenna,

Length of flagellum of second antenna,

Length of scale of second antenna,

J.ength of somite of first antenna,

Length of labrum,

Remarks.—This species differs from Squilla miles, Hess., in the presence of acute

spines on the antero-lateral angles of the carapace, and in the absence of triangular
(ZOOL, CHALL. EXP. —PART. XLY.—1886.) Yy 5

34 THE VOYAGE OF H.M.8. CHALLENGER.

processes on the four somites of the hind body, and of movable spinules on the sub-
median spines of the telson. It differs from Mier’s genus Leptosquilla in the enlarge-
ment of the tips of the eyes, and in the presence of distinct submedian carinz on all the
five somites of the hind body.

Squilla lata, n. sp. (PI. III. figs. 1, 2, 3).

Diagnosis.—The dactylus of the raptorial limb has six teeth, and the pectinations
on the inner edge of the second joint are arranged in an undulating line. Eyes
directed forward, nearly cylindrical, with corneal portion hardly wider than base.
Appendages of exposed thoracic limbs strap-shaped. Hind body gradually increasing in
width backwards to the fifth abdominal somite, which is equal in breadth to one-fourth of
the totallength. Inner spine of basal joint of sixth abdominal appendage with a rounded
tooth on its outer edge, and, on its inner edge, a series of acute dentations gradually
increasing in length distally. Rostrum elongated, nearly twice as long as wide.
Submedian dorsal carinze absent on the first and second thoracic and second and third
abdominal somites, and obsolete on all the others except the sixth abdominal, where they
are well developed. Submedian spines of telson curved at their tips; and flattened, broad
and rounded at their bases, with no intermediate dentations. Dorsal surface of telson
obscurely marked by curved lines. Paddle of exopodite of sixth abdominal appendage
as long as second jomt. There is no spine on the dorsal surface of the basal joint of the
sixth abdominal appendage, Lateral edges of second thoracic somite acute, those of the
third and fourth obliquely truncated.

Special Description.—Rostrum twice as long as broad, and ending in a rounded
point. Antero-lateral angles of carapace acute, posterior angles rounded, anterior border
subacute, posterior border nearly transverse. No median carina on rostrum or carapace,
the gastric area of which is only slightly convex. On each side of carapace there is a
carina which reaches nearly to the cervical suture, behind which there is a short carina
on each postero-lateral lobe. A median dorsal tubercle close to the posterior edge of the
carapace. Length of rostrum 7835, and of carapace 72935, and of the two 75% of the
total length. Hind body gradually and uniformly increasing in width from in front back-
wards to the fifth abdominal somite, which is more than twice (74) as wide as the second
thoracic somite. The second thoracic somite has no dorsal carinee; the third has one lateral
carina on each side, but no submedians; the fourth and fifth thoracic somites have very
obscure and short median carinee which converge anteriorly, as well as lateral carine.
The first five abdominal somites have three lateral carinee on each side. The first three
have no submedian carinze, while there are short obsolete submedian carinz on the fourth
and fifth, converging anteriorly. The third; fourth, and fifth abdominal somites have
each a median dorsal tubercle. The sixth abdominal somite has two well marked sub-

REPORT ON THE STOMATOPODA. 30

medians and two pairs of lateral ones, all six ending in spines, while none of the other
carine end in spines. The telson has a broad convex median dorsal ridge which ends
behind in a short acute spine, and on each lateral portion there are six obsolete
symmetrical curved lines. The posterior border of the telson carries six lobes or teeth, the
laterals and postero-laterals acute and straight, while the acute points of the submedians
are curved inwards. There is a single rounded lobe between the lateral and the postero-
lateral, six between the postero-lateral and postero-median, aud none between the postero-
medians, the space being filled up by the flattened bases of the spines which meet on the
middle line. The sixth abdominal appendage is very large, and the terminal paddle of
the exopodite is oval and equal in length to the proximal joint. The endopodite is long,
curved, and narrow, and the prolongation from the lower surface of the basal joint ends
in a short curved acute outer spine, and a much longer inner spine, which has a rounded
tooth on its outer border and on its inner border six or seven acute teeth which increase
in size distally. There is no dorsal spine on the basal joint.

The ocular segment is wholly anterior to the long rostrum, and the eyes are directed
forwards side by side, and are twice as long as wide and swollen in the middle, although
the retinal portion, which is divided by a depression into halves, is a little wider than
the base. The first antenne are short (2%, of the total length) and their somite ends
dorsally in a pair of lateral spines which point forward. The flagellum of the second
antenna, including the three-jointed shaft, is nearly (32) as long as the first, and the
scale is yp; as long as the animal.

The width of the carapace between the antero-lateral spines is about half (49) its
greatest width, and almost exactly half its length. The dactylus of the raptorial claw of
the second maxilliped is armed with six curved acute teeth, gradually increasing in size
distally. The second joint has three movable spines on its inner edge, and the pectinations
of its outer edge are arranged in an undulating line, with a convexity under each tooth of
the dactylus (see fig. 2 Pl. III). The appendages of the three pairs of exposed
thoracic legs are flat, strap-shaped, and dilated at the tips. The lateral edges of the
second thoracic somite are acute and a little curved forward, and those of the third and
fourth somites are obliquely truncated, with subacute anterior and posterior angles. The
fifth has no angular process. Sexes alike.

Habitat—Station 190, in the Arafura Sea, south of New Guinea, September 12,
1874; lat. 8° 56’ S., long. 136° 5’ E.; 49 fathoms; two males and one female taken
in the trawl.

Remarks.—This and the following species, Squilla fusciata, closely resemble the
genus Chloridella in the shape of the eyes, and in the flattened strap-like form of the
appendages to the exposed thoracic limbs. They agree in many points, as will be shown
in the description of the second species, but there are so many well-marked differences
that there can be no doubt of their distinctness.

36 THE VOYAGE OF H.M.S. CHALLENGER.

In inches and | In thousandths
Meseuremente decimals. of total length.

Total length from tips of eyes to tips of spines of telson, 3°500
On middle line :—
From tip of rostrum to tip of telson, ; . : ; 3°225 1000
Rostrum, . : : d ; : : ; 150 47
Carapace, . : j ; - ; : ‘ 660 205
Carapace, including rostrum, . 3 : : : 810 252
From posterior edge of carapace to posterior edge of second thoracic
somite, . 070 22
From posterior edge of second to posterior edge of third thoracic
somite, . 145 45
From posterior edge of third to posterior ‘edge of fourth thoracic
somite, . 125 39
From posterior edge of fourth to posterior edge of fifth thoracic
somite, . ; : ; 145 45
First abdominal somite, : : ; 3 : ; 255 79
Second abdominal somite, . ; ; : : 1 215 67
Third abdominal somite, . : , : 5 ; 220 68
Fourth abdominal somite, . : ; : : ; 220 68
Fifth abdominal somite, . x : ; . : °280 87
Sixth abdominal somite, . : " ‘ : 2 230 71
Telson on middle line, 3 é : : : : 510 158
Length of free somites, : : ; : ; 2-415 749
Total length on middle line, : : ; : 3°225 1000
Length of telson to tips of spines, . : : 3 620 192
Width of carapace between antero-lateral spines, 5 : : 340 105
Width of carapace (greatest), . 5 : : . : 58d 181
Width of second thoracic somite, ; : : : : 495 153
Width of third thoracic somite, ; i : ; , 035 166
Width of fourth thoracic somite, : : : , : "595 184
Width of fifth thoracic somite, : : Z F . 585 181
Width of first abdominal somite, ; 5 , j ; ‘740 229
Width of second abdominal somite, . ; : ; : ‘780 242
Width of third abdominal somite, . f . , “ 805 250
Width of fourth abdominal somite, . : ‘ : : $10 251
Width of fifth abdominal somite, : : ; : f 810 251
Width of sixth abdominal somite, , : ‘ ‘ 690 217
Width of telson between postero- median spines, : . é 130 40
| Width of telson between postero-lateral spines, : : : ‘470 146
| Width of telson (greatest), . ‘ : 5 ; 635 oi,
Total length of sixth abdominal appendage, ‘ ; : : 426 132
Distance between tips of abdominal appendages, ; 1-600 496
Length of paddle of exopodite of sixth abdominal appendage, : 265 82
Length of base of exopodite of sixth abdominal appendage, . : 265 82.
Length of endopodite of sixth abdominal appendage, . : : ‘408 126
Length of eye, ; : ; : : J : 135 42
Width of eye, 5 ; 095 29
Length of first antenna, measured from tip of rostrum, : : 925 287
Length of flagellum of second antenna, : : ‘ . 770 239
Length of scale of second antenna, . : 360 112
Distance from tip of rostrum to cervical suture, on middle line, é 675 209

REPORT ON THE STOMATOPODA. 37

In its general characteristics this species conforms pretty exactly to Mier’s description
of Squilla dubia of Milne-Edwards (Squilla rubrolineata, Dana and v. Martens), but it
may readily be distinguished from this species, which is recorded only from the western
shore of the Atlantic, by the absence of submedian carine on most of the free somites,
and by the absence of dentations between the submedian spines of the telson.

Squila fasciata, De Haan (PI. UI. figs. 4, 5; Pl. II. fig. 8).

Squilla fasciata, De Haan, Fauna Japon., Crustacea, p. 224, p. 2, 2¢, tig. 4, 1849.
Squilla fasciata, Miers, Squillide, Ann, and Mag. Nat. Hist., ser. 5, vol. v. p. 29, 1880.

Diagnosis.—Six teeth on dactylus of raptorial claw. Dentations on inner edge
undulating. Eyes directed forward, nearly cylindrical, with conical portion hardly
wider than the base. Appendages of exposed thoracic limbs strap-shaped. Hind body
gradually increasing in width from in front backwards to the fifth abdominal somite,
which is equal in breadth to more than one-fourth of the total length. Inner spine of
basal jomt of sixth abdominal appendage longer than outer, with a rounded tooth on its
outer and on its inner edge a series of acute dentations gradually increasing in length
distally. Rostrum only a little longer than wide. Submedian dorsal carinz entirely
absent in all the exposed somites of the hind body except the sixth abdominal. Sub-
median spines of telson acute, with about eight intermediate acute dentations. Dorsal
surface of telson with four or five sharply defined longitudinal carine on each side of the
median ridge. Paddle of exopodite of sixth abdominal appendage about half as long as
second joint, and an acute spine on the dorsal surface of the basal joint of the appendage.
Lateral edges of second thoracic somite acute, those of the third and fourth rounded.

Special Description.—Rostrum only a little longer than wide, subtriangular, with a
broadly rounded tip. Antero-lateral angles of carapace acute; postero-lateral angles
rounded ; anterior border nearly transverse ; posterior border deeply notched on middle
line. No median carina on the rostrum or on the carapace, the gastric area of which is
conspicuously convex. Lateral carinee of carapace very faintly indicated. No median
dorsal tubercle on posterior edge of carapace. Length of rostrum +425, and of the
carapace yoo, and of the two 473; of the total length. Hind body gradually and
uniformly increasing in width from in front backwards to the fifth abdominal somite,
which is one and one-half times as broad as the second thoracie somite. There are no
submedian dorsal carine on any of the exposed somites of the hind body except the
sixth abdominal ; the second thoracic somite has no dorsal carinz. The third, fourth, and
fifth have each a pair of lateral carina, and the first five abdominal somites have each
three pairs of lateral caring, including the marginal ones; all of which, except the inner
ones_of the first abdominal somite, end in spines. The sixth abdominal somite has well-
marked submedian carine converging posteriorly, and two pairs of lateral ones, all six

38 THE VOYAGE OF H.M.S. CHALLENGER.

ending in spines. The telson has a broad convex median dorsal ridge which ends behind
in a short acute spine, and on each lateral portion there are four or five sharply defined
parallel longitudinal carinze. The posterior border of the telson has three pairs of acute
straight spines, and there is a single rounded tooth between the lateral and postero-
lateral, eight acute curved teeth between the latter and the submedian, and four between
this and the middle line, which is deeply notched. The sixth abdominal appendage is
very large, and the terminal paddle of the exopodite is oval and half as long as the
proximal joint. The endopodite is long, narrow, and very slightly curved, and the
prolongation from the ventral surface of the basal jot ends in a short curved acute
outer spine, and a much longer acute inner spine which has a rounded tooth on its outer
border, and on its inner border seven or eight acute teeth which increase in size distally.
There is an acute spine on the dorsal surface of the basal joint.

The rostrum reaches to the base of the ocular segment, and the eyes are directed
forward, side by side, and nearly cylindrical and about twice as long as wide. The first
antenne are long (445 of the total length of the body), and their somite ends dorsally in
a pair of acute spines which are slightly divergent. The flagellum of the second antenna,
including the shaft, is about three-fourths (}§) as long as the first, and the scale is y'g as
long as the animal. The width of the carapace between the antero-lateral angles is half
(39) its greatest width and a little less than half (33) its length. The dactyle of the
raptorial claw is armed with six acute curved teeth gradually increasing in size distally.
The second joint has three movable spines on its inner edge, and the pectinations on its
inner edge are arranged in an undulating line, with a convexity under each tooth. The
appendages of the three pairs of exposed thoracic limbs are flat, strap-shaped and more
dilated and rounded at the tip than they are in Squilla lata. The lateral edges of
the second thoracic somite are acute, and strongly curved forward, while the lateral
edges of the third and fourth are rounded. The fifth has a subacute prominence on
each side. Males and females alike, except as regards the structures concerned in
reproduction.

Habitat.—Station 2338, Inland Sea, Japan, May 26, 1875; lat. 34° 18’ N., long.
133° 35’ E.; depth 15 fathoms; bottom blue mud.

Two specimens, a male and a female, were obtained. The only other specimen known
is also from Japan, and is described by De Haan as the type of the species.

Remarks.—This species is very similar to the one last described, agreeing with it in
the presence of six teeth in the raptorial claw, the cylindrical shape of the eyes, the
almost total absence of median and submedian dorsal carine, the great width and the
uniform increase in width of the hind body, the flattened strap-like shape of the appendages
of the exposed thoracic legs, the great length of the curved endopodite of the sixth
abdominal appendage, and the presence of acute teeth on the inner margin of its inner
spine. There are so many well-marked points of difference, however, that there can be

REPORT ON THE STOMATOPODA. 39

In inches and | In thousandths

Mes E i
Measurements. decimals. of total length.

Total length from tip of rostrum to tip of telson,
Measurements along the middle line :—

Rostrum, 42
Carapace, 231
Total length of carapace, including rostrum, . 273

From posterior margin of carapace to posterior edge of second thoracic

somite, 52
From posterior edge of second to posterior edge of third thoracic

somite, 37
From posterior edge of third to posterior ‘edge of fourth thoracic

somite,
From posterior edge of fourth to “posterior edge of fifth thoracic

somite, . : ° 2

First abdominal somite,
Second abdominal somite, j
Third abdominal somite,
Fourth abdominal somite,
Fifth abdominal somite,
Sixth abdominal somite,
Telson on middle line,

Total length of hind body,
Total length on middle line from tip of rostrum, .

Greatest length of telson,
Width of carapace between antero- lateral spines,
Width of carapace (greatest), . :
Width of second thoracic somite,

Width of third thoracic somite,

Width of fourth thoracic somite,

Width of fifth thoracic somite,

Width of first abdominal somite,

Width of second abdominal somite,

Width of third abdominal somite,

Width of fourth abdominal somite,

Width of fifth abdominal] somite,

Width of sixth abdominal somite,

Width of telson between postero-median spines,
Width of telson between postero-lateral ae
Width of telson (greatest),

Length of sixth abdominal appendage,

Distance from tip of rostrum to end of eye,
Distance from tip of rostrum to tip of first antenna,
Length of flagellum of second antenna,

Length of scale of second antenna,

no doubt of their specific distinctness, nor of the identity of the second species with
the single specimen upon which De Haan based his description of Squilla fasciata.

The many differences between the two species are noted in the general description,
but they are briefly as follows -—

40 THE VOYAGE OF H.M.S. CHALLENGER.

The species may be readily distinguished by the following features. The rostrum of
Squilla lata is twice as long as wide, and it gradually tapers to the narrow rounded tip,
while it is trigonal in Squilla fasciata, hardly longer than wide, with a broad tip. The
lateral edges of the third and fourth thoracic somites are obliquely truncated in Squwilla
lata, and rounded in Squilla fasciata. The paddle of the exopodite of the sixth
abdominal appendage of Squilla lata is as long, and that of Squilla fasciata half as long,
as the second joint. The dorsal surface of the telson of Squilla lata is obscurely marked
by curved lines, and the submedian posterior spines are curved at the tip and broad at
the base, with no intermediate dentations ; while the dorsal surface in Squilla fasciata
is marked by longitudinal carinee, and the submedian spines are acute with intermediate
dentations.

In addition to these more conspicuous points, the following minute differences serve to
show the distinctness of the two forms :—

The antenne are longer than in Sgwilla lata; the carapace is nearly transverse in
front and deeply notched behind, while in Squlla lata it is subacute in front and nearly
transverse behind. All the carinze of the abdominal somites, except the imner pair on
the first, end in spines in Squilla fasciata, while in Squilla lata only those on the sixth
somite end in spines, and there is no dorsal spine on the basal joint of the sixth abdominal
appendage, while there is an acute spine in Squilla fasciata.

The following measurements in thousandths of the total length also exhibit the
difference between the two.

Squilla | Squilla | Squilla

chlorida. | lata. fasciata.
Length of rostrum, : : , ; : : P 24 47 42
Length of carapace, ; . : F : 5 : 196 205 231
Length of hind body, : ; : ; ; 5 ; 781 749 726
Width of third thoracic somite, . : : : : ; 168 166 161
Width of fifth abdominal somite, . : : : ; : 264 251 278
Length of first antenna, measured from tip of rostrum, ; : F 288 287 401°
Length of flagellum of second antenna, . , : 3 ; 232 239 227
Length of scale of secon] antenna, : : : ; : 96 112 170

Squilla chlorida, n. sp. (Pl. I. figs. 1-5).

1 Clorida decorata, Wood-Mason, J. Proc. Asiatic Soc., Bengal, Dec. 1875. On some new species
of Stomatopoda Crustacea, Ann. and Mag. Nat. Hist., ser. 4, vol. xvii., 1876, p. 263.

1Chloridella decorata, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v. p. 15, 1880.

Diagnosis.—Eyes directed forward, with basal portion swollen, and a little
wider than corneal portion. Appendages of thoracic limbs strap-shaped. Five teeth on

REPORT ON THE STOMATOPODA. 41

dactylus of raptorial claw. All the exposed somites of the hind body with lateral dorsal
carine, but no submedian dorsal carine on any of the exposed somites, except the
sixth abdominal. Pectinations on inner edge of second joint of raptorial limb undulating.
Width of fifth abdominal somite equal to one-fourth of the total length. Inner spine of
basal joint of uropod longer than the outer, with a rounded tooth on its outer edge, and on
its inner edge four or five acute teeth gradually increasing in size distally. Rostrum
wider than long. Dorsal surface of telson with numerous rounded tubercles on each side
of median ridge ; ventral surface smooth.

Special Description—Rostrum without a median ridge, a little longer than wide,
subtriangular and nearly semicircular. Antero-lateral angles of carapace acute ; postero-
lateral angles broadly rounded, and prominent ; anterior and posterior borders nearly
transverse. Carapace without a median carina, and with very faint lines indicating the
lateral carinz ; gastric area convex and prominent, with a strongly marked cervical
suture. Length of rostrum +375, and of the carapace 7%, and of the two 345, of the
total length. Exposed thoracic somites increasing gradually in width. First five
abdominal somites nearly equal in width, and about equal in width to one-fourth of the
total length. The width of the fifth is Yovo Of the total length, and it is 2¢ as wide as
the second thoracic somite. There are no submedian dorsal caring on any of the exposed
somites of the hind body except the sixth abdominal. The first thoracic somite is
exposed, and, like the second, has no dorsal carine ; the third, fourth, and fifth have
each a pair of short lateral caring, and the first five abdominal somites have each three
pairs of lateral carine, including the marginal ones, none of which, except those on the
sixth abdominal somite, end in spines. The sixth abdominal somite has well-developed
submedian, and two pairs of lateral carine, all six ending in spines. The first four
abdominal somites have each a small median tubercle, which can be detected with
difficulty with a lens. The telson has a convex median dorsal ridge, which ends behind
in a short acute spine, on each side of which there are numerous scattered rounded
tubercles, which obscurely exhibit an arrangement in rows. The ventral surface of the
telson is smooth, and the posterior border ends in six pairs of acute spines, with a
tuberculated ridge over the base of each postero-median and postero-lateral ; a single
acute tooth between the lateral and postero-lateral ; six or seven acute teeth between
each postero-lateral and submedian, and three or four on each side between the
tip of the submedian and the deep notch which occupies the middle line. The
sixth abdominal appendage is long, and the: paddle of the exopodite is oval, about as
long as the second joint; the endopodite is long, narrow, and curved; the basal
joint carries an acute dorsal spine, and its ventral prolongation ends in a short acute
outer spine and a very much longer acute curved inner spine, which has a rounded
tooth on its outer margin, and on its inner margin four or five acute curved teeth, which

increase in size distally. The rostrum reaches nearly to the base of the ocular somite,
(ZOOL. CHALL. EXP,—PART XLv.—1886.) Yy 6

42 THE VOYAGE OF H.M.S. CHALLENGER.

and the eyes are directed forward, side by side, and swollen near the base. The first
antenne are about equal to 4 (3%%5) of the total length of the body, and their somite
ends dorsally in a pair of acute spines which point almost directly forward. The
flagellum of the second antenna, including the shaft, is more than 3 (32) as long as the
first, and the scale is only ;4 as long as the animal. The width of the carapace
between the antero-lateral angles is about half ($+) of its greatest width and about 4
(23) of its length. The dactylus of the raptorial claw of the second maxilliped is armed
with five acute curved teeth, which are finely serrated on their outer edges and increase
uniformly in length distally. The second joint has three movable spines on its inner
margin, and the pectinations on its outer margin are arranged in an undulating line,
with a convexity under each tooth. The appendages of the three pairs of exposed
thoracic limbs are strap-shaped and dilated at their tips. The first thoracic somite is
entirely exposed dorsally ; the second ends laterally in acute spines which are straight
and transverse; and the lateral edges of the third and fourth are rounded, with the
postero-lateral angles of the fourth a little more acute than the antero-lateral angles.
The fifth is obliquely truncated behind. The endopodite of the first abdominal
appendage of the male is broad and rounded (Pl. II. fig. 5), and the outer lobe a
of the terminal joint is subtriangular, and much larger, but shorter, than the inner,
which is separated from it by a suture. The fixed limb of the forceps e is as long as the
movable limb 6, and it ends in two hooks on its internal edge, one of which points
towards the tip and the other towards the base of the appendage.

Habitat.—Amboina, 15 fathoms; one male.

Remarks.—This species bears a close resemblance to Chlorida decorata, Wood-Mason,
from the Andamans, and to Chlorida microphthalma, Eydoux and Souleyet (Squilla
microphthalma, Milne-Edwards), from India, but it is readily distinguished from the
former by the fact that its telson is smooth below, and with scattered rounded tubercles
above, while the telson of Chlorida decorata is described as vermiculated above and
below by granulated ridges.

The presence of four instead of five teeth on the dactyle of the raptorial claw, and
the shortness of the rostrum, distinguish it from Sguilla (Chlorida) microphthalma, as
briefly described by Milne-Edwards ; and the presence of five teeth on the raptorial claw,
of long acute spines on the lateral edges of the second thoracic somite, and the fact that
the postero-lateral angles of the first four abdominal somites are obtuse, distinguish it
from Chloridella microphthalma (depressa) of Miers. The outline and dorsal surface of
the telson are also quite different from Miers’ figure, while the whole anterior end of the
body, especially the rostrum and eyes, is so different from his figure that there can be
no doubt of its distinctness.

The genus Chlorida, Eydoux and Souleyet (Chloridella, Miers) is characterised by the
fact that the eyes are constricted at the tip and dilated near the base. It includes four

REPORT ON THE STOMATOPODA. 43

In inches and | In thousandths

a 2s decimals. of total length.

Total length from tips of eyes to tip of telson,
Measurements on middle line :—
Total length from tip of rostrum to tip of telson,

Rostrum, z 24
Carapace, . ' 196
Total length of carapace, including rostrum, 220
From posterior edge of carapace to posterior edge of second thoracic
somite, 45
From posterior edge of second to posterior edge of third thoracic
somite, 40
From posterior edge of third to posterior "edge of fourth thoracic
somite, 43
From posterior edge of fourth to "posterior edge of fifth thoracic
somite, : ¢ 54
First abdominal somite, 72
Second abdominal somite, : 80
Third abdominal somite, 72
Fourth abdominal somite, 80
Fifth abdominal somite, 94
Sixth abdominal somite, 56
Telson on middle line, 145
Total length of hind body, 781
Total length on middle line, 1000
Width of carapace between antero-lateral spines, 112
Width of carapace (greatest), . 206
Width of second thoracic somite, 174
Width of third thoracic somite, 168
Width of fourth thoracie somite, 190
Width of fifth thoracic somite, 200
Width of first abdominal somite, 240
Width of second abdominal somite, 241
Width of third abdominal somite, 245
Width of fourth abdominal somite, 250
Width of fifth abdominal somite, 264
Width of sixth abdominal somite, 254
Width of telson, between postero-median spines, 53
Width of telson, between postero-lateral spines, 157
Width of telson (greatest), E . 248
Greatest length of telson, 182
Length of sixth abdominal appendage, 225
Distance between same, 512
Length of eye, 64
Width of eye, 34
Total length of first antenna, measured from tip of rostrum, 288
Total length of flagellum of second antenna, : 232
Total length of scale of second antenna, 96

species, one of which, Chloridella rotundicauda, Miers, is represented by a single female
specimen ; another, Chlorida latreilii, E. and §., by a single mutilated specimen ; while

44 THE VOYAGE OF H.M.S. CHALLENGER.

a third, Chlorida decorata, Wood-Mason, is very briefly and inadequately described,
and the fourth, Chlorida microphthalma, K. and 8., by no means satisfactorily known.

Our species would belong to the genus if the shape of the eyes were the criterion, but
it is so perfectly connected with Squwilla fasciata, a true Squilla, through the species
which was last described, Squilla lata, that the propriety of retaining the genus Chlori-
della must remain very doubtful for the present. A comparison of this species (PI. II.
fig. 1), Squilla lata (Pl. III. fig. 1), and Squilla fasciata (Pl. UI. fig. 4), will show
that no one of these three species should be placed in a genus which does not include
the other two, and as the last is clearly a true Sqgwilla I have included all three in this
genus.

The comparative table of measurements of the three species which is given at the
end of the description of Squilla fasciata will also serve to show the close resemblance
much better than a description.

Genus Lysiosquilla, Dana.

Diagnosis.—Stomatopoda with the sixth abdominal somite separated from the
telson by a movable joint ; the hind body depressed, loosely articulated and wide; the
dactyle of the raptorial claw without a basal enlargement, but with more than six
marginal spines; no more than four secondary spines, and often only one, between the
intermediate and submedian spines of the telson, which is usually wider than long; and
the outer spine of the ventral prolongation from the basal joint of the uropod usually
longer than the inner. The larva is an Erichthus or Squillerichthus, with the ocular
and antennulary somites covered by the carapace ; the lateral edges of the deep carapace
folded inwards over the ventral surface; the bases of the postero-lateral spines distant
from the dorsal middle line; the hind body flat and wide; the telson wider than long,
and with few spines or only one between the intermediate and submedian spines; and
the dactylus of the raptorial claw with numerous marginal spines.

Special Description.—I have examined the first abdominal appendage of the males
of twa species, Lysiosquilla maculata (PI. X. fig. 6) and Lysiosquilla excavatria (PI.
X. fig. 12), and find such great and characteristic difference from Squilla, that I
do not hesitate to add to the diagnostic characteristics of the genus the statement that
Lysiosquilla is distinguished by the fact that the terminal joint of the exopodite of the
first abdominal appendage of the adult male is subtriangular, with its large outer lobe
separated by a suture from the very small inner lobe, and the fixed limb of the petasma
very small and not ending in a hook.

Like the genus Squilla the genus Lysiosquilla includes two minor groups, a highly
specialized one and a more primitive and slightly modified one. The single specimen of
Lysiosquilla (Coronis) scolopendra upon which Latreille based his genus Coronis, was

REPORT ON THE STOMATOPODA. 45

very imperfectly described, and it has unfortunately been lost ; but his description and
the figures which were afterwards published by Milne-Edwards, show such a close resem-
blance to Lysiosquilla excavatrix, n. sp., described below, that there can be no doubt of
their very intimate relationship or of the importance of the differences between them and
Dana’s Lysiosquilla mornata, upon which this author based his genus Lysiosquilla, which
has broad triangular eyes, large antennary scales, filiform appendages to the last three
thoracic legs, and the larva of the closely related Lysiosquilla maculata has a short rostrum
and postero-lateral spines; while Lysiosquilla (Coronis) eacavatria has small subcylindrical
eyes, minute antennary scales and uropods, dilated appendages to the last three thoracic
limbs, and the rostrum and postero-lateral spines of its larva are very long.

Notwithstanding these important differences the various species agree in many
features, such as the presence of numerous spines on the dactyle, and of very few between
the submedian and intermediate spines of the telson, in the loose articulation of the wide
flat hind body, and the absence of dorsal carine. They cannot be arranged in two
divergent groups, and it is impossible to draw any abrupt line between them. There
can, therefore, be no doubt of the propriety of including them in one genus, as Miers’
does, retaining Dana’s term Lysiosquilla for all of them.

The higher and lower forms stand in precisely the same relation to each other as do
Squilla and Chloridella, and there is also a most suggestive similarity in the character of
the differences. In fact it is almost as difficult to detect generic differences between
Coronis and Chloridella as between the latter and Squilla, or between Coronis and
the higher Lysiosquilla.

The two genera Squilla and Lysiosquilla are divergent stems from a common stalk, .
and while the higher forms are quite distinct, the resemblance between the lower forms
is no more than we should expect.

While there can be no doubt that all the very wide Hrichthi with a deep infolded
carapace are Lysiosquilla larve, we cannot state with confidence that all Lysiosquille
have larvee of this type, for there are no strongly marked and constant differences between
the Lysioerichthus and the Gonerichthus and Pseuderichthus, and it is not impossible
that some of the narrow elongated Pseuderichthi may be Lysiosquilla larvee.

In some of the Lysiosquillx, and cam in all of them, there are marked secondary
differences between the sexes.

Lysiosquilla maculata (Fabricius) (Pl. X. figs. 1-7).

The Challenger collection includes four specimens of this well-known and widely
distributed species; one full-grown male, No. 1, presented by the king of Amboina;
a half-grown male, No. 3, from Amboina, and a mature male and female, No. 2, from

1 Miers, On the Squillide, p. 3.

46 THE VOYAGE OF H.M.S. CHALLENGER.

Samboangan. Miers has called attention’ to the fact that in the single female with the
large raptorial limbs which he has been able to examine, the spines arming the
inner margin of the dactylus, instead of being strong and elongated as in the males, are
very short, and towards the base are reduced to little more than small serrations or teeth.
The fact that there is the same difference between the raptorial claws of the males and
that of the female in the Challenger series would seem to indicate that this difference
between the sexes is constant. The raptorial claw of a large male is shown in Pl. X.
fig. 1, and that of the female in fig. 2. This difference is the more remarkable as
secondary structural differences between the sexes are extremely rare among the
Stomatopods. The male of Pseudosquilla ciliata is said to be more brilliantly coloured
than the female, and the female Lysiosquilla excavatriz is larger and darker coloured than
the male, but there are no structural differences except the modification of the endopodite
of the first abdominal appendage which is found in all male Stomatopods. The sterna of
the free thoracic somites of the female Lysiosquilla maculata are much thicker and stouter
than those of the male, and the median ventral carine larger.

The first abdominal appendage of the male is shown in Pl. X. fig. 6. The endo-
podite is elongated and shorter than the exopodite ex, and its terminal joint B is triangular,
wider than long, with the inner lobe h very much smaller than the outer lobe a, from
which it is separated by a suture. The inner lobe is nearly circular, with a deep notch
on its distal edge; the appendix interna is long and prominent; the movable limb / of
the forceps is acute and long, while the immovable limb is very small and scarcely visible.

Ontogeny.—The largest Erichthus larva in the Challenger collection, a Lysio-
erichthus, is the same as the one shown in Claus’s figure 16, and probably identical with
Erichthus duvaucellei, Guér. It is shown in Pl. X. fig. 7. It is very widely distri-
buted, and the collection contains numerous specimens from various localities in the
West Pacific. Specimens were taken in the tow-net between Api and Cape York,
between Admiralty Island and the coast of Japan, in the Straits of Mindoro, and at
other points, while Claus’s specimens were obtained in the Indian Ocean, and Guérin’s in
the Gulf of Bengal.

Its resemblance to the larva from which I have reared Lysiosquilla excavatrix
shows that it is a Lysiosquilla, and, like Lysiosquilla maculata, it is somewhat excep-
tional, since the imner spine of the uropod is longer than the outer, while most of the
adult Lysiosquillz, and most of the Lysioerichthus larvae, have the outer spine longest.
The raptorial claw is long and slender, and it exhibits traces of eight marginal spines,
while the adult female Lysiosquilla maculata has seven or eight and the adult male
nine.

The abdomen of the oldest larva is marked by the transverse dark bands which are
so characteristic of the adult Lysiosquilla maculata, and I cannot doubt that this

1 Miers, On the Squillide, p. 6.

REPORT ON THE STOMATOPODA. 47

In thousandths of

WV pi i .
Measurements In inches total length.

. | No. 1.| No. 2.] No. 3.
Total length on middle line from tip of rostrum to tip of

telson, . : . | 1133 | 933 | 539
Rostrum, : ‘ : / L , =r 15 037 | -032 04
if)

Carapace, : ; ; : 3 ; 14% | 122 164 | -168 | -166

From tip of rostrum to middle of posterior edge of

carapace, 2,5 | 143 _ 201 | -200 | -206
From posterior edge of carapace to posterior edge of
second thoracic somite, ‘ re 048 | -039 ‘04
From posterior edge of second to posterior edge ‘of third
thoracic somite, ts | we 048 | -045 05
From posterior edge of third to “posterior edge of fourth
thoracic somite, ‘ tt | ws 058 | -058 05
From posterior edge of fourth to posterio edge of fifth
thoracic somite, - : | 063 | -058 ‘05
First abdominal somite, 3 j : ; te | 24 OFS OTT ‘07
Second abdominal somite, ; : : ‘ ts | # 074 | ‘071 07
Third abdominal somite, ‘ ; : ‘ +5 | 32 ‘074 | ‘077 07
Fourth abdominal somite, : : : ts | 22 ‘O74 | -077 07
Fifth abdominal somite, ; 2 i ‘ ea |e 095 ‘09 ‘08
Sixth abdominal somite, ; 3 ; : i | 22 063 | -077 | -067
Telson on middle line, . : : ; : 15 | 1 122 | -129 | °145
Total length of hind body, - ; ; 97, | 732 ‘798 | -798 | -762
Total length on middle line from tip of rostrum
to tip of telson, , : : . | LI] ot 999 | 998 | -968
Width of carapace at anterior end, 14, | 1 Ht SLOG) 2103), |)) 117,
Greatest width of carapace, - 144) 139 | 1 159 | -168 | -168
Width of second thoracic somite, . 17; | 1 qr 090 | -103 | :095
Width of third thoracic somite, 1 | LS +8 ET) |esl35) ) 105
Width of fourth thoracic somite, 13, | 15 32 132 | -155 | :126
Width of fifth thoracic somite, 17% | 133 a4 143 | 174 | -147
Width of first abdominal somite, . 2, | 135 | 1 185 | -200 | -189
Width of second abdominal somite, 27, | 27, | 1% 185 | -213 | -200
Width of third abdominal somite, 25, | 27s | 15 196 | -213 |) -200
Width of fourth abdominal somite, 23; | 27s | 13 196 | -213 | -200
Width of fifth abdominal somite, . 23, | 2 1 190 | -206 | -200
Width of sixth abdominal somite, ; 275 | 144 | 13 180 | -194 | -189
Width of telson between postero-median spines, +3 | 1 Te 080 | 103 | *105
Width of telson between sage gee 143 | 1% | 1 159 | -161 | :179
Greatest width of telson, . 275 | 143 | LS 180 | :187 | :189
we stalk, : is 3 ss ive ie
) T > Uos
Width of eye, | He 042 | -045
First antenna, first joint, a ee ‘037
First antenna, second joint, qs 032
First antenna, third joint, os 032
First antenna, longest flagellum, 15, 127
Flagellum of second antenna, 2 166
Scale of second antenna, . flee ‘116
Antenna of i ey ae | oe SOoTiIe eee
Labrum, . ; 1 a4 084 | -090

1 No, 1, adult male; No. 2, adult female; No. 3, half-grown male.

48 THE VOYAGE OF H.M.S. CHALLENGER. -

large, widely distributed, transversely banded larva is the young of Lysiosquilla
maculata, which is one of the largest and most widely distributed Stomatopods. The
value of this identification can be much better estimated after examination of the
section on Stomatopod larvee.

Lysiosquilla (Coronis) excavatrix, n. sp. (Pl. X. figs. 8-16).

Diagnosis.—Body depressed and loosely articulated ; whole dorsal surface smooth.
Carapace, with rostrum, making 74% of the total length from tip of rostrum to
tip of telson. Eyes nearly cylindrical, with hemispherical corneze. Scale of second
antenna about +é5 of total length. Raptorial claw of second maxilliped oval, with
fourteen or fifteen short curved pointed teeth and a much longer terminal tooth on
dactylus, and three movable spines on inner side of base of second joint. Second thoracic
somite with a blunt rounded lobe on each side. Appendages of third, fourth, and fifth
pereiopods ovate, that of the fourth being largest. Sixth abdominal somite with a long
acute process on each side, near anterior edge. Telson smooth, nearly rectangular, with
one median and two pairs of lateral subacute lobes. Endopodite of sixth abdominal
appendage triangular, basal prolongation ending in two simple acute curved spines, the
outer much the larger. Larva a long-spined Lysioerichthus.

General Description.—The carapace is smooth on its dorsal surface, with the gastric
sutures distinct from the anterior edge to the transverse cervical suture, behind which no
longitudinal sutures are visible. The space included between the gastric sutures is wide
and makes up rather more than two-thirds of the total width of the carapace. This
is elongated and slightly narrowed anteriorly, with the antero-lateral angles nearly
right angles and the postero-lateral angles broadly rounded. The anterior edge of the
carapace is nearly transverse, and it is very deeply emarginated on the middle line
behind. Its length on the middle line is 74%, and its greatest breadth about +12, of the
total length, The rostrum is a little longer than wide and the lateral edges are not
angulated, but are strongly convex in outline and regularly curved to the base of the
short acuminate tip. The rostrum completely covers the ocular somite, and hides the
base of the eyes. The first antennary somite ends laterally in long slender acute
spines, which curve outwards and forwards.

The first antennee are very short and their shafts are hidden nearly to the
tips below the eyes, which are small, subcylindrical, with terminal hemispherical
corneze, and are usually directed forwards side by side. The second antennz are almost
as long as the first, and the oval scale is very narrow and short, its length being only
rééo of the total length of the body.

The second joint or propodus of the raptorial claw is broad, with an obscure longitu-
dinal ridge on its outer surface, and when the dactylus is closed the claw is oval and nearly

REPORT ON THE STOMATOPODA. 49

flat. The dactylus (Pl. X. fig. 11) is armed with fourteen or fifteen sharp curved
teeth, besides the terminal tooth, which is more than three times as long as any of the
others. There are three stout movable spines on the inner edge of the second joint near
the base, while the outer edge is fringed with numerous small immovable pectinations.

The carapace does not completely hide the dorsal surface of the first thoracic somite.
The second thoracic somite is prolonged on each side near its posterior edge into a
rounded lobe which projects backwards, while the lateral edges of the third, fourth, and
fifth thoracic somites are entire and longitudinal, with rounded antero- and postero-
lateral angles.

The thoracic somites are much narrower than the carapace, and they increase
regularly in width from the second to the fifth, which is less than two-thirds as wide as
the first abdominal. The first five abdominal somites are nearly equal in width and also
in length, their transverse diameter being about +445 and their average length about
too Of the total length of the body. They are a little (%) wider than the carapace.
The sixth abdominal somite is considerably narrower than the fifth, and slightly
narrower than the carapace, its transverse diameter being less than 519, of the total
length of the body. It is about twice as wide as long. On the lateral margin on each
side, close to the anterior edge, there is a prominent acute curved process, but its
posterior edge is entire, and the suture which separates it from the telson is very
obscure, but movable. The whole dorsal surface of the rostrum, carapace, hind body,
and telson is smooth and highly polished, and without carinz or spines.

The telson is as wide as the sixth abdominal somite, nearly rectangular, and 43, as
long as wide, its length being about +7, of the total length of the body. Its posterior
edge is transverse, as long as the anterior edge, and nearly straight, and it is divided, in
the adult, by four concave notches into five subacute lobes, of which one is on the
middle line, and two on each side, all nearly equidistant. In the young the median lobe
is represented by a pair of submedian spines, with minute sete on the margin of the
telson between them.

The exposed thoracic limbs are short and their appendages are broad, oval, and
membranous. The uropods are small and little used in locomotion; the endopodite
is triangular, and the ventral prolongation from the basal joint ends in two acute,
curved, unarmed spines, of which the outer is much the larger. The anterior process
of the mandible is bordered by two rows of irregular obtusely rounded dentations,
which are continued to the tip where there are two terminal dentations. The
endopodite of the first maxilla, fig. 9, ends in a stout acute curved spine, which
carries on its outer surface two stout movable hairs and one slender one. The outer
surface of the second maxilla, fig. 10, is smooth, and is not divided into lobes by a
median furrow.

The first five pairs of abdominal appendages are furnished with very large exopodites

(200L, CHALL. EXP.—PART XLV.—1886.) ; Yy7

50 THE VOYAGE OF H.M.S. CHALLENGER.

which overlap on the middle line. The endopodite of the first abdominal appendage of
the male is shown in Pl. X. fig. 12. It is very much smaller than the exopodite,
triangular, and as wide as long. The inner lobe 6 of the terminal joint is very small,
nearly circular, and separated by a suture from the broad triangular outer lobe a. The
movable limb of the petasma is long, nearly straight, and armed on its outer edge, near
the base, with an acute spine, while the fixed limb is very short and without a hook at
the tip, which is rounded, as is the case in Lysiosquilla maculata. The sexes of this
species are quite different from each other, and it seems probable that secondary sexual
differences may occur in all the flat loosely articulated Stomatopods.

The character of the difference between the sexes is however quite different from
what we find in Lysiosquilla maculata. In the latter species the sexes are alike in size
and colour, and they differ in structural features, while in this species there are no
structural differences except those which are concerned in reproduction, but the female is
much larger than the male, and of an opaque olive brown, almost black colour, while the
male is of a transparent grey.

Habits.—As our information regarding the habits of the Stomatopoda is very
scanty, I give all that I have been able to learn of the habits of this species. It is found
in the sand of the ocean beach just below low-tide mark, where it is exposed to the full
force of the ocean swell, and it inhabits a very deep cylindrical burrow which is nearly
vertical and goes down for several feet. While watching for its prey the animal stations
itself at the mouth of the burrow, which is arched over with sand, so that only the tips
of the eyes are exposed. The food consists of small Crustacea, fishes and other small
animals, and when one approaches within reach the Coronis darts out of the burrow,
knocking away the loose sand, and seizing it in its raptorial claw it darts backwards with
it and retreats to the bottom of the burrow. When hungry it often captures prey at a
distance of six or eight inches, but, as a rule, it waits until it is near enough to be caught
without leaving the opening. The food which is captured is usually stored away at the
bottom of the burrow, and the animal returns to the mouth and resumes its watch. In
excavating its burrow the animal begins by stretching its body out on the sand, which is
then swept away from under it by the action of the abdominal appendages, until all of
the body except the eyes and telson are buried. It then forces its head into the loose
sand which has been stirred up by the action of the abdominal appendages, and dragging
its body down it quickly becomes buried vertically, head downwards, and it continues to
burrow until it reaches the hard undisturbed sand, when it bends upon itself, and passing
the head up on the ventral side of the swimmerets it reverses its position and works
upwards to the surface, hardening and compressing the sand by the pressure of the dorsal
surface. After the upper end of the ‘burrow is thus rendered firm and circular it again

doubles upon itself, and going to the bottom gathers an armful of sand, which is clasped .

against the ventral surface of the body between the large second maxillipeds, where it is

REPORT ON THE STOMATOPODA. 51

held in place by the flat oval chele which are tightly clasped over it. At the opening it
stretches out as far as it can reach without leaving the burrow, and dropping the armful
of sand it smooths it down until it is level with the surrounding surface. This process is
then repeated until the burrow reaches a great depth, for I have dug for three or four
feet without reaching the end, and all the specimens which I kept in confinement
burrowed to the bottom of the aquarium.

When the burrow is finished the animal spends most of its time near the top, and as
the semicircular exopodites of the abdominal appendages complete the outline formed by
the convex dorsal surface, it completely fills the circular tube, into which the constantly
vibrating scoop-like abdominal appendages carry a continuous current of water, which
escapes through the loose sand.

The whole organisation of the species,—the convex body, the semicircular swimmerets,
the small closely approximated eyes, and the broad flat claws,—adapt it for its mode of
life, and it is doubtful whether any other species is more completely subterranean in its
habits. Although it is very common at Beaufort, I have captured only one specimen
while swimming, and it very rarely ventures more than a few inches from the burrow.

Its movements when seizing its prey are so rapid that the eye can scarcely follow
them, and the attempt to cut off its retreat with a trowel usually results in cutting the
animal in two, although this is the only method of capturing them which I have found
at all successful.

This species, when kept in confinement, makes a faint stridulating noise by rubbing
the uropod against the lower surface of the telson. Sguwilla empusa stridulates
vigorously in the same way, and its habits, which I have also had an opportunity to
observe, are quite different from those of Lysiosquilla excavatrix. It is very active,
swimming swiftly through the water, and pursuing its prey to a great distance from its
burrow, so that it is frequently captured in the water by the trawl or sein. It inhabits
muddy rather than sandy bottoms, and its burrow is a shallow U-shaped tube, open at
both ends, and excavated entirely by the action of the current of water which is set up
by the abdominal appendages.

Ontogeny.—Lysiosquilla excavatrix is one of the few Stomatopods which have been
traced through their metamorphoses, and the fully grown larva, which is a long-spined
Lysioerichthus, is shown in Pl. XI. figs. 1-3.

I shall show in the section on Stomatopod larve that it probably hatches as an
Erichthoidina, with five pairs of biramous thoracic appendages; the sixth, seventh, and
eighth thoracic somites distinct but without appendages, and the telson joined to the last
thoracic somite with no intervening abdominal region. In the youngest Hrichthus stage,
however, there is a long segmented abdomen with four pairs of fully developed append-
ages, and the thoracic somites from the third to the eighth have no appendages, while
those of the first and second thoracic somites have their adult form. The lateral edges of

52 THE VOYAGE OF H.M.S. CHALLENGER.

the young Hrichthus are bordered with spines as in Alima, and the general structure of
the larva is very Alima-like ; a resemblance which totally disappears in the older larvee.

Habitat.—Common below low water mark on sandy beaches at Beaufort, N.C.,
U.S.A. .

In hundredths | In thousandths of
SUSE DED SE of inch. total length.
Large Large Large Large
male. | female. | male. | female.
Total length on middle line from tip of rostrum to tip of telson, . 2°41 3°02
Rostrum, é oes 5 : P : 10 ait! 041 036
Carapace, : : ‘ : : : : 33 iog 137 129
Total length of carapace and rostrum, . : : 43 00 ‘178 165
From posterior edge of carapace to posterior edge of second
thoracic somite, . 04 12 017 039
From posterior edge of second to posterior edge of third thoracic
somite, : 10 14 041 046
From posterior edge of third to posterior edge of fourth thoracic
somite, ; 13 cI Lif 054 056
From posterior edge of fourth to posterior edge of fifth thoracic
somite, é : : a . : 16 ‘20 066 073
First abdominal somite, , : 4 : : “24 28 100 092
Second abdominal somite, : ; Z : 5 25 29 104 095
Third abdominal somite, : ‘ : ; . 25 30 104 099
Fourth abdominal somite, ; , : , 26 30 108 099
Fifth abdominal somite, 4 : ; ; "25 30 104 ‘099
Sixth abdominal somite, : : ; s 13 18 054 059
Telson, ; 3 : ; : ; . lp 24 ‘070 079
Total length of hind body, : : ; : 1:98 2°52 822 836
Total length on middle line, : : : 2°41 3°02 1-000 1-001
Width of carapace at anterior end, : F : : "20 “24 083 ‘079
Greatest width of carapace, E ‘ é : : "30 33 1125 109
Width of second thoracic somite, ; ; : : 13 “22 054 073
Width of third thoracic somite, . , : ; : 22 30 091 099
Width of fourth thoracic somite, : ; ‘ : 24 32 100 106
Width of fifth thoracie somite, . : : z 25 34 104 ‘112
Width of first abdominal somite, : i 5 a “Siem “45 129 149
Width of second abdominal somite, . : ; : "3 “46 129 “152
Width of third abdominal somite, : ZA ; : 32 “46 133 152
Width of fourth abdominal somite, 5 : : ‘ “32 “46 133 152
Width of fifth abdominal somite, 6 : : “ol “46 129 152
Width of sixth abdominal somite, ; $ : : 26 34 108 ‘112
Greatest width of telson, : ; : 29 36 120 119
Length of first antenna from tip of rostrum to tip of longest
flagellum, i 5 ‘ ; 38 36 158 ‘119
Length of appendage of second antenna, : . E "35 “ioe 145 oe
Length of scale of second antenna, 3 : : : 08 “12 032 ‘04
Length of eye, . : E é : : 09 “09 037 029
Total length of swimmeret, : : z z : 38 42 | +158 139

REPORT ON THE STOMATOPODA. 53

Size.—Males from two to three inches long ; females from two and one-half to four
inches long.

Colour.—Males transparent, with small dark brown pigment spots uniformly distri-
buted, giving the body a greyish tint; females more opaque and of a dark olive brown,
nearly black, colour.

Remarks.—This species seems to be very closely related to Lysiosquilla polydactyla,
but there can be no doubt of its specific distinctions.

Genus Pseudosquilla, Guérin.

Diagnosis.—Stomatopoda with the sixth abdominal somite separated from the telson
by a movable joint; the hind body smooth, convex, and narrow; the dactyle of the
raptorial claw without a basal enlargement, and with few marginal spines or none; the
submedian spine of the telson long, and tipped with movable spinules, with usually a
single secondary spinule, sometimes two, three, or four, between the submedian and
intermediate marginal spines ; the terminal joint of the first abdominal appendage of the
male imperfectly divided by a marginal notch into an inner and an outer lobe; larva an
elongated narrow Hrichthus, with a short narrow carapace, with the postero-lateral spines
near the dorsal middle line, and the lateral edges slightly or not at all infolded ; the telson
longer than wide, with long submedian spines; the proximal joint of the exopodite of
the uropod with numerous spines, and the outer spine of the basal prolongation much
longer than the inner and longer than the telson.

Pseudosquilla ciliata, Miers (Pl. XV. fig. 10).
Pseudosquilla ciliata, Miers, Ann. and Mag. Nat. Hist., ser. 5, vol. v. p. 108, pl. iii. figs. 7, 8.

The Challenger collection includes two specimens of this well-known species, a male and
a female, from 2 fathoms depth at St. Thomas, and also a male from the reefs at Honolulu.

The raptorial claws, the spines of the telson, and the paddles and spines of the
uropods retain, in the alcoholic specimen, the bright cherry red colour which, according
to G. Clark, is exhibited by the living animal. The alcoholic specimens also have eye-
like spots of black pigment near the lateral edges of the third thoracic and first abdominal
somites, and another on the dorsal surface of the base of the telson on the middle line.
The occurrence of this Pacific species at St. Thomas is a remarkable fact in the
distribution of the Stomatopoda, but it will probably be found to be widely distributed
throughout the Atlantic as well as the Pacific, for Von Martens records it from Cuba.
The specimens from St. Thomas agree perfectly in measurements as well as in most other
respects with the one from Honolulu. The only differences which I have been able to
detect are the following: the paddle of the exopodite of the uropod is about as long as
the second joint in the specimens from the Pacific, while it is a little shorter in the two

54 THE VOYAGE OF H.M.S. CHALLENGER.

Measurements of male specimen from Honolulu. a Faia tena ie ere ae
Total length on middle line, from tip of rostrum to edge of telson F 2°13
Measurements on middle line :—
Rostrum, . : ‘ 5 : : ; . 12 56
Carapace, . : : : ; - 3 : D4 254
Carapace, including rostrum, . : : ; : 66 310
From posterior edge of carapace to posterior edge of third thoracic
somite, . 12 56
From posterior edge of third to posterior ‘edge of fourth thoracic
somite, . 10 47
From posterior edge of fourth to posterior edge of fifth thoracic
somite, . , : ; ; 4 : 10 47
First abdominal somite, t : : : : 16 75
Second abdominal somite, . 5 ‘ : x j 16 75
Third abdominal somite, . : 4 . ; 16 75
Fourth abdominal somite, . - , : ; : 16 75
Fifth abdominal somite, . 3 ; : : 5 "22 104
Sixth abdominal somite, . : ; ; 3 14 66
Telson on middle line, é : ; d : 5 15 70
Total length of hind body, ; : : : 1:47 690
Total length on middle line, , : : : 2°13 1000
Greatest length of telson, : : F ; ; ; 38 178
Width of rostrum, : . , . : “20 94
Width of carapace at antero- lateral angles, : : F : 32 150
Width of carapace (greatest), . : : 4 : “44 208
Width of third thoracic somite, ‘ ‘ ; P ; 42 197
Width of fourth thoracic somite, P ‘ 4 ; : “42 197
Width of fifth thoracic somite, 4 : : : z “42 197
Width of first abdominal somite, : ; : , “46 216
Width of second abdominal somite, . : : : ; “46 216
Width of third abdominal somite, ; : : : F “46 216
Width of fourth abdominal somite, . ; : : ‘ “46 216
Width of fifth abdominal somite, 5 R “46 214
Width of sixth abdominal somite between postero-lateral spines, 4 38 178
Width of telson (greatest), . ; : ; 3 3 38 178

specimens from St. Thomas; the specimen from Honolulu has the inner spine of the
ventral prolongation from the base of the uropod much longer than the outer, while
the two spines are of equal length in the two specimens from St. Thomas; in the
specimens from St. Thomas the postero-lateral angles of the fourth, fifth, and sixth
abdominal appendages end acutely in spines, while this is the case with the fifth and
sixth only in the specimen from Honolulu. In all the specimens the lateral lobes
of the last thoracic somite are bidentate.

Miers is undoubtedly correct in his decision that v. Marten’s Pseudosquilla stylifera
from Cuba is a Pseudosquilla ciliata, and it is probable that it ranges very widely over

REPORT ON THE STOMATOPODA. 55

both the Atlantic and the Pacific, with slight differences between the specimens from
widely separated localities, and Pseudosquilla oculata, Brullé, from the Canaries and
Madeira, may possibly prove to be one of these varieties of Pseudosquilla ciliata, but
the species of Pseudosquilla all need careful revision, as there is evidently considerable
variability. Peculiarities of colouration are unsafe guides in the study of preserved
specimens, and most of the other marks for the discrimination of species are known to
vary. Thus Heller says that his Pseudosquilla oculata (Pseudosquilla ornata, Miers)
may be distinguished from Pseudosquilla stylifera (Pseudosquilla ciliata, Miers) by the
fact that the spine of the uropod is longer than the endopodite, but Miers’ figure of
Pseudosquilla ciliata (Squillide, pl. ii. fig. 8) represents the spine as longer than the
endopodite uropod, and this is true of the Challenger specimens also.

Genus Gonodactylus, Latreille.

Diagnosis.—Stomatopoda with the sixth abdominal somite separated from the
telson by a movable joint; the hind body convex; and the dactylus of the raptorial
claw enlarged at the base, and without marginal spines. Larva, an Hrichthus, has the
postero-lateral spines of the carapace near the dorsal middle line, the lateral edges not
infolded, and hatches from the egg as an Hrichthoidina, which becomes converted into
an Hrichthus without the loss of any of its appendages.

Remarks.—All the Stomatopods with a dilated unarmed dactylus on the second
maxilliped are usually grouped in a single genus, Gonodactylus, but the collection of
species which are thus brought together is a very heterogeneous one, and little examina-
tion is necessary to show that the genus, as usually characterised, includes at least three
distinct assemblages of species. The species which have the sixth abdominal somite
immovably fused with the telson are obviously more closely related to each other than
they are to the other species of Gonodactylus, and as this fact should find its expression
in the systematic zoology of the group I have placed these species by themselves in a
distinct genus Protosquilla, retaining in the genus Gonodactylus only those species
which have the telson movable. In the genus as thus restricted, two species, Gonodac-
tylus (Squilla) bradyi, A. Milne-Edwards, and Gonodactylus trachwrus, Miers, differ
from all the remaining species in many features, such as the small size of the eyes, the
scales of the second antennze and the uropods; the depression of the hind body, and the
presence of dentations on the inner edge of the dactylus of the raptorial claw. The pro-
priety of separating these species from the true Gonodactyli seems obvious, and as they
present many points of resemblance to both Lysiosquilla and Squilla, especially to the
least specialised species in these two genera, I suggest for them the generic name,
Coronida, compounded from Coronis, the generic name proposed by Latreille for the
Lysiosquille with dilated appendages on the exposed thoracic limbs, and Chlorida,

56 THE VOYAGE OF H.M.S. CHALLENGER.

the generic name proposed by Eydoux and Souleyet for the small-eyed Squilla. The
genus Coronida will therefore include the Stomatopods which have the dactyli of the
raptorial claw dilated at the base and armed with spines on the inner margin, and the
hind body depressed ; while the genus Gonodactylus, as thus restricted by the removal
of the Protosquille and the Coronide, will include only the species which have the
hind body convex, the dactylus of the raptorial claw dilated at the base and unarmed, and
the telson distinct and movable.

In the genus Gonodactylus as thus restricted the terminal joimt of the first
abdominal somite of the male is imperfectly divided, by a marginal notch, into an outer
and an inner lobe, which are not separated by a suture (see Pl. XV. fig. 8).

The fixed limb of the petasma is short, and ends in a single acute hook, while the
movable limb is abruptly bent outwards near its base.

Gonodactylus chiragra, Latreille (Pl. XV. fig. 4).

This common and widely distributed species is represented in the Challenger collection
by numerous males and females from St. Thomas, one male from Bermuda, one male
‘from Station 36, near Bermuda (32° 7’ 25” N., 65° 4’ W.), by two specimens from
Samboangan, and one specimen from Samboangan Bank, besides numerous adult male and
female specimens of a closely related but minute variety from near Cape St. Roque. The
appendages of the exposed thoracic limbs of all the specimens of Gonodactylus chiragra
are slightly flattened, and twice as wide as thick, and their edges are parallel and not
dilated at the tip. The second joint of the exopodite of the uropod is more than twice as
long as the paddle, and it carries about eleven (ten in four specimens, eleven in seven
specimens, nine or ten in Heller’s specimen from Nicobars, twelve in two specimens)
movable spines and one terminal ventral immovable spine.

The terminal joint of the endopodite of the male Gonodactylus chiragra is divided
by a deep marginal notch into an outer lobe (Pl. XV. fig. 4) @ and an inner one 8,
which is not separated from the outer one by a suture. The fixed limb of the petasma e
is short, swollen at the base, and bent inwards at right angles at the tip, thus forming a
hook which ends in an acute point. The much longer movable limb / is bent outwards
in a prominent sharply defined obtuse angle near its base.

There seems to be no room for doubting that the specimens from various parts of the
ocean which have been described as Gonodactylus chiragra really belong to one species,
and that it is very widely distributed throughout the Atlantic, the Pacific, and the Indian
Oceans. E. v. Martens says* that although he has formerly published his opinion that
this species is confined to the Indian Ocean, and the Pacific from the Red Sea to Chili, the

1 Ueber cubanische Crustaceen, nach den Sammlungen Dr. J. Gundlach’s, Archiv f. Natwrgesch., 1872, Jahrg.
xxxiiL, Bd. ii. p. 147.

REPORT ON THE STOMATOPODA.

Measurements of a female, a, from Bermuda and a male, }, from In inches and In thousandths
Samboangan. decimals, of total length.
a b. b.
Measurements along middle line :—
Rostrum, ‘ : : , ‘ : : “12 ‘18 69
Carapace, : d : ; - : : “43 66 222
Carapace, including rostrum, é : : ( oh) ‘78 289
From posterior edge of carapace to aga edge of third
thoracic somite, 10 14 52
From posterior edge ‘of third to posterior “edge of fourth
thoracic somite, 12 14 52
From posterior edge ‘of fourth to posterior edge ‘of fifth
thoracic somite, . : : : ; 12 14 52
First abdominal somite, 5 3 F 5 ‘ 14 19 69
Second abdominal somite, i ; ; r - 14 22 81
Third abdominal somite, ; : : ; : 16 22 81
Fourth abdominal somite, 2 : - - 18 20 74
Fifth abdominal somite, ‘ , 3 : : a) 24 89
Sixth abdominal somite, : : z ? : 12 16 59
Telson on middle line, ; E : : : ‘18 27 100
Total length of hind body, ; é : : 1-45 1:92 709
Total length on middle line, - f : 2-00 2-70 998
Greatest length of telson, ‘ . 29 “40 148
Width of carapace between antero-lateral angles, : ; 25 84 126
Width of carapace (greatest), . - - > : “32 “46 170
Width of sixth thoracic somite, . : : F : 32 50 185
Width of seventh thoracic somite, : ; : ‘ 33 10) 185
Width of eighth thoracic somite, : : F 3 33 “50 185
Width of first abdominal somite, : ; ‘ ‘ 34 52 192
Width of second abdominal somite, : ; : : "35 52 192
Width of third abdominal somite, E : F F 35 52 192
Width of fourth abdominal somite, ; : : : 35 52 192
Width of fifth abdominal somite, : 35 D2 192
Width of sixth abdominal somite between postero- Jateral spines, . “32 “44 163
Width of telson between sub-median spines, . ; : ‘ll ‘14 52
Width of telson (greatest), : : : F : “30 “44 163
Length of eye, . : : : ‘14 eas fe
Length of first antenna from tip of rostrum, : : : ac “49 155
Length of second antenna from tip of rostrum, . ; ; ot 56 207
Length of scale, . ; : : : : : ec 32 118

absence of any decided difference between the specimen obtained in Cuba by
Gundlach and those which he has himself collected at Amboina compels him to
reverse this opinion and to recognise the occurrence of the species in the tropical
Atlantic. He says that the specimens from the Atlantic are smaller on the average
than those from the Indian Ocean, that the rostral spine is a little shorter. He

points out also a slight variation among the Cuban specimens in the terminal spines
(ZOOL. CHALL, EXP.—PART XLy.—1886.) Yy 8

58 THE VOYAGE OF H.M.S. CHALLENGER.

of the telson, as well as slight differences in the colour of the alcoholic specimens from the
two previous localities.

Smith says* that specimens from the West Indian and Florida Keys do not differ from
those collected at the Reefs of the Abrolhos at Caravella, Province of Bahia, Brazil, by
Hartt, but he says that “the American species is, however, very likely distinct from the
true Gonodactylus chiragra, from the Old World,” although he gives no reason for this
opinion. A. Milne-Edwards? records it from the Mediterranean, and coasts of America,
Africa, Asia, and Oceania. Miers says that he has examined a specimen four inches long
from New Guinea.

Gonodactylus graphurus, Miers (Pl. XIV. figs. 1, 4, 6; Pl. XV. figs. 3, 8)..

Gonodactylus graphurus, Miers, Edw. J., On some new or undescribed Crustacea from the
Samoan Islands, Ann. and Mag. Nat. Hist., ser. 4, vol. xvi. p. 344, 1875.

Gonodactylus graphurus, Miers, Edw. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v., 1880, p. 120, pl. iii. fig. 9.

Gonodactylus graphurus, White, List of Crustacea, Brit. Mus., 1849, p- 85 (no description).

Gonodactylus graphurus, Miers, Edw. J., Malaysian Crustacea, Ann. and Mag. Nat. Hist., ser. 5,
vol. v. p. 459, 1880.

Diagnosis.—Dactylus of raptorial claw without teeth and dilated at base. Rostrum
with a long slender acute median spine and subacute antero-lateral angles. Hind
body wider than the carapace, with its sides straight and parallel. Second thoracic
somite partially exposed dorsally. Lateral edges of all the abdominal somites with
triangular marginal carinze with the apices directed backwards. Dorsal surface of the
thoracic and first five abdominal somites without carine or spines. Sixth abdominal
somite with a short unarmed median carina, and three pairs of swollen convex lateral
carine, all ending in spines, the second or intermediate shorter than the first and third.
Telson ending in six acute spines with broad bases, each one with a broad convex longi-
tudinal dorsal carina. Middle of dorsal surface of telson with a protuberant hemi-
spherical eminence, upon which are five very broad convex carinz, of which the median
and submedian end posteriorly in acute spines, while the marginal ones are without
spines and shorter than the others. There is a triangular subacute prominence on the
anterior edge of the telson on each side, halfway between the middle line and the lateral
margin, with a broad convex carina. Paddle of exopodite of uropod half as long as the
second joint. Basal joint of uropod with two acute spines on dorsal surface, while its
ventral prolongation ends in two nearly equal acute spines, with a rounded tooth on
inner edge of outer one near the base. Tergal surfaces of the first five abdominal somites
with longitudinal sutures on sides, and a transverse dorsal suture.

General Description.—Carapace flat, hind body convex, and highly polished.

1 Brazilian Crustacea, Trans. Conn. Acad., vol. ii. p. 31, 1869. 2 Nouv. Arch. Mus. Hist. Nat., t. iv. p. 65, 1868.

REPORT ON THE STOMATOPODA. 59

Rostrum consisting of a transverse proximal portion more than twice as wide as long,
with subacute antero-lateral angles, and a long slender median acute spine which reaches
beyond the bases of the eyes (see fig. 6). Carapace nearly rectangular, three-fifths
as long as wide, leaving the dorsal surface of the second thoracic appendage almost
completely exposed, with broadly rounded antero-lateral and more acutely rounded
postero-lateral angles. The antero-lateral angles project forwards beyond the median
gastric area, which is very slightly convex and a little wider posteriorly than anteriorly,
and bounded by two nearly straight gastric sutures which run from anterior to posterior
edge. Transverse cervical suture almost invisible. . Posterior edge of carapace
transverse. Second thoracic somite as wide as carapace. The eight following somites
equal in width, and wider ({22) than the carapace. The lateral margins of the third
thoracic somite are straight, with rounded angles, and as wide as the dorsal portion ; the
fourth is narrowed a little towards the lateral edge, and the fifth still more. There is a
‘median impressed line on the dorsal surface of the third and one on the fourth thoracic
somite. All the abdominal somites have marginal carinee which are wide in front and
gradually become narrow towards the postero-lateral angle, which is rounded in the first
four segments, rectangular in the fifth, and ending im an acute spine in the sixth (see
fig. 1). There are no dorsal carinz on the first five abdominal segments. The sixth
(see Pl. XIV. fig. 4) has a short unarmed median carina and three pairs of lateral carine,
which terminate posteriorly in acute spines and are greatly swollen, so that they together
occupy more than half the dorsal surface of the somite, convexly rounded, and widening
anteriorly. The spines of the submedian and marginal carine project beyond the
posterior edge of the somite, while those of the intermediate carinz barely reach it.
The second, third, fourth, and sixth abdominal somites are equal in length, and longer
than the first and shorter than the fifth. A strongly impressed suture (see Pl. XIV.
fig. 1) crosses the middle line of each of the first five abdominal somites near the
posterior border, and bends forwards on the side of the somite, and each of these somites
has also a lateral longitudinal suture on each side, branching upwards at its posterior edge.

The telson is considerably wider than long (+2) and its dorsal surface is folded into
a very graceful pattern which can hardly be satisfactorily described, although it can be
understood by comparing the profile view, fig. 1, with the surface view shown in
fig. 4. The median portion is occupied by’a hemispherical prominence, upon which
are five meridional carine, which are convexly rounded and so greatly swollen that they
almost completely cover it. Three of these, the median and submedians, are abruptly
rounded and emarginated at their posterior ends, where they end in acute spines, while
the third or marginal pair do not end in spines and are much shorter.

On each side of the median prominence there is on the anterior edge of the telson a
triangular prominence with a broad rounded subacute carina, and outside this, and near
the antero-lateral angle, there is a transverse rounded prominence without a keel, and a

60 THE VOYAGE OF H.M.S. CHALLENGER.

short transverse keel at the antero-lateral angle. The posterior edge of the telson is
folded into six acute spines, and from the tip of each a broad convex carina runs forwards
on the dorsal surface of the telson. The marginal carina reaches to the anterior border, the
intermediate stops a little short of this border, and the third or submedian runs forwards
for about half the length of the telson, where it abuts upon the median prominence.

The marginal spines of the telson have acute tips and swollen bases with convexly
rounded outlines, and the deep sulci between them are bordered by convex ridges. The
outer spine is simple; the intermediate has its outer edge simple and a secondary spine
on its inner edge near its base; the submedian spine has a secondary spine on its outer
edge near its base, and twelve or thirteen acute dentations on its inner edge. The dorsal
surface of the basal joint of the uropod ends posteriorly in two acute spines, and there is
a rounded lobe outside the base of the outer one; its ventral surface ends posteriorly in
a process divided into two acute curved spines, the outer longer and with a rounded
tooth near its base on its inner edge.

The paddle of the exopodite is more than half (}$) as long as the second joint, which
has a rounded process on the inner edge of its base, about twelve movable spines on its
outer edge, and a ventral terminal immovable spine.

The eyes are cylindrical, with rounded cornez (fig. 6), and the first and second
antennz are about equal in length. The eyes hide all the shaft of the first antenna
except the terminal joint. The tip of the simple dactyle of the raptorial claw is curved
outwards, and its inner edge is barbed with minute serrations.

The endopodite of the first abdominal appendage of the male (Pl. XV. fig. 8) is
very similar to that of Gonodactylus chiragra. The terminal joint is divided by a
deep marginal notch into a small outer lobe and a large inner lobe, both of which
are rounded and not separated by a suture. The fixed limb e of the petasma is swollen
at the base and it ends in a single acute hook.

Remarks.—Notwithstanding the fact that this is a widely distributed species, no
minute description of it has been published, as Miers’ description’ gives little except the
points of difference from Gonodactylus chiragra, and the only figure, the telson shown in
Miers’ fig. 9, is misleading, as will be seen by comparison with our fig. 4, Pl. XIV.
In his figures, as well as in his descriptions, he represents the central area of the dorsal
surface of the telson as made up of three pairs of curved carinz on the sides of the
median one, whereas more careful examination will show that the third or outermost
pair do not belong to the central elevated convex system so characteristic of this and
related species, but to a distinct eminence on the anterior edge of the telson.

This species is very closely related to Gonodactylus chiragra. The species next
described, Gonodactylus glabrous, is in many respects intermediate between the two,
the three forming a sharply defined natural group or subgenus. It may readily be

1 Ann. and Mag. Nat. Hist., ser. 4, vol. xiv. p. 344, and ser. 5, vol. v. p. 120.

REPORT ON THE STOMATOPODA. 61

F In inches and | In thousandths
Measurements of a male specimen from near Cape York. aaa Giécial lanai:
Total length from tip of rostrum to middle of posterior edge of telson, 2°14
Measurements on middle line :—
Rostrum, : : : : ‘ > : ‘15 70
Carapace, . : ‘ : : : : é 51 238
Carapace, including rostrum, . : : : ; 66 308
From posterior edge of carapace to posterior edge of third thoracic
somite, . - ‘ 5 : : F : 14 65
From posterior edge of. third to posterior edge of fourth thoracic
somite, . . - - < - : 10 47
From posterior edge of fourth to posterior edge of fifth thoracic
somite, . ; - : ; : ; 2 10 47
First abdominal somite, . : 2 : : : 12 56
Second abdominal somite, . ; ; ‘ : : ‘16 75
Third abdominal somite, . F : : : ; ‘16 75
Fourth abdominal somite, . P ‘ ; ‘ : 16 75
Fifth abdominal somite, . : : 2 : : 20 93
Sixth abdominal somite, . : F ‘ ‘ : 16 75
Telson on middle line, 3 : : : ; ; 18 84
Total length of hind body, . ‘ F ‘ F 1:48 692
Total length on middle line, : : : : 2°14 1000
Total length of telson, : : : : : 31 145
Width of rostrum, . : : ae ; ; ; 16 75
Width of carapace between antero-lateral angles, ; : 3 "26 122
Width of carapace (greatest), . , : : : : 33 155
Width of third thoracic somite, : , ; : ; 38 179
Width of fourth thoracic somite, : 4 . , : 38 179
Width of fifth thoracic somite, é ; ‘ : : 38 179
Width of first abdominal somite, ; : : : : 38 179
Width of second abdominal somite, . : : ; : 38 179
Width of third abdominal somite, . : : ; : "38 179
Width of fourth abdominal somite, . : ‘ , 38 179
Width of fifth abdominal somite, ‘ ; ‘ : : 38 179
Width of sixth abdominal somite, between postero-lateral spines, : 33 155
Width of telson between submedian spines, : - “ 10 47
Width of telson between intermediate spines, . 5 : s "26 122
Width of telson (greatest), . - : ; : : “34 160
Length of paddle of exopodite of uropod, on dorsal surface, . Jat ‘13 61
Width of second joint of exopodite of uropod, on dorsal surface, : 22 103

distinguished from Gonodactylus chiragra by the presence of five instead of three carinee
on the median protuberance of the telson, and by the presence of acute spines on the
posterior ends of three of them, by the possession of six instead of four spines on the
posterior border, and by the presence of a tooth on the inner edge of the outer spine of
the ventral prolongation of the uropod. It is more difficult to distinguish it from
Gonodactylus glabrous, which it resembles very closely. The many minute differences
will be pointed out in the description of this species.

62 THE VOYAGE OF H.M.S. CHALLENGER.

Colour.—The alcoholic specimens have the carapace mottled with dark brown
pigment, with four sharply defined dark spots close to its posterior edge, six elongated
pigment spots on the third, six on the fourth, four on the fifth, and four on each of the
first five abdominal somites. Miers says that his specimens from the Samoan Islands were
dull olive green, with branchial appendages pink, but those colours are not preserved in
the specimens which I have seen.

Habitat.—The Challenger collection includes two specimens, a male and a female,
from 8 fathoms at Station 186, near Cape York, lat. 10° 30’ S., long. 142° 18’ E. Miers
records it from Torres Straits and from other points in Northern Australia, as well as
from the Red Sea, Ceylon, Gulf of Suez, Seychelles, Samoa Islands, Amboina, and other
localities in the Indo-Pacific region.

Size.—Largest male specimen, 2744 inches from tip of rostrum to middle of

telson.

Gonodactylus glabrous, n. sp. (Pl. XIV. fig. 5; Pl. XV. figs. 7, 9).

Diagnosis.—Dactylus of raptorial claw without teeth, and dilated at base. Ros-
trum with a long slender acute median spine, and subacute lateral angles. Hind
body wider than the carapace, with its sides straight and parallel. Second thoracic
somite slightly exposed dorsally. Lateral edges of all the abdominal somites with
triangular marginal carinze with the apices directed backwards. Dorsal surfaces of the
thoracic and first five abdominal somites without carine or spines. Sixth abdominal
somite without a median carina, and with three pairs of lateral carine, all ending in
spines, the submedian and intermediate being swollen and convex, while the marginal
is more narrow and linear. The second or intermediate is shorter than the first and
third. Telson ending in six acute spines with broad bases, and each with a sharply
defined narrow longitudinal dorsal carina. Middle of dorsal surface of telson with.a pro-
tuberant hemispherical eminence, upon which are five narrow sharply defined carine, of
which the median and the two submedian end posteriorly in spines, while the marginal ones
are without spines and shorter than the others. There is a triangular subacute promin-
ence on the anterior edge of the dorsal surface of the telson on each side, halfway between
the middle line and the lateral margin, with a narrow sharply defined carina. Paddles
of exopodites of uropods less than half as long as the second joint. The basal joint of
the uropod has only one acute spine on its dorsal surface, while its ventral prolongation
ends in two nearly equal spines, with an obsolete rounded tooth on the inner margin of
the outer one near its base. The tergal surfaces of the abdominal somites have no trans-
verse or longitudinal sutures.

General Description.—This species is so very similar to Gonodactylus graphurus in
every prominent characteristic that the description of that species will serve for this also,

REPORT ON THE STOMATOPODA. 63

except as regards the following features. The median spine of the rostrum of Gonodac-
tylus glabrous reaches halfway to the tips of the eyes (see fig. 7). The carapace almost
completely covers the fifth thoracic somite. The transverse cervical suture crosses the
middle line of the carapace close to its posterior edge. The postero-lateral angles of the
third and fourth as well as the fifth abdominal somite rectangular. The sixth abdominal
somite has no median carina. The marginal carine of this somite are not swollen, and
the intermediate and submedian carinze are much less swollen than they are in Gonodac-
tylus graphurus, especially towards their posterior ends (compare fig. 5 of Pl. XIV. with
fig. 4). The abdominal somites have no transverse or longitudinal sutures. The telson is
nearly (}#¢) as long as wide, and its various dorsal carinz are all of them more sharply
defined than they are in Gonodactylus graphuwrus, as will be seen by comparing fig. 5 with
fig. 4, and they are very little or not at all swollen, so that the exposed spaces on the
dorsal surface of the telson, between the carinz on the median elevation as well as
those between the carine on the lateral portions of the telson, are much wider than the
carine themselves. The outlines of the bases of the marginal spines of the telson are
nearly straight, and they lack the swelling convex curves of the graceful outline of the
telson of Gonodactylus graphurus. The dorsal surface of the basal joint of the uropod
has only one acute spine, and the paddle of the exopodite is considerably less than half
(42) as long as the second joint, when measured on the dorsal surface. The tip of the
second joint of the first antenna is exposed in front of the eye. In all other respects the
species conforms to the general description of Gonodactylus graphurus.

Locality —The Challenger collection contains one female specimen from Samboangan
Reefs.

Size.—The total length from the tip of the rostrum to the middle point of the
posterior edge of the telson is 2;$5 inches.

Colowr.—In the alcoholic specimen there is much less pigment than there is in
Gonodactylus graphurus, and this is restricted to small sharply-defined symmetrical
eye-like black spots, of which there are six in the cardiac area of the carapace, two on
the sixth, and two on the seventh thoracic somite, four on the eighth, and six on each of
the first five abdominal somites.

Remarks.—This species may be recognised by its very close and striking resemblance
to Gonodactylus graphurus, from which it may be distinguished without difficulty by
the absence of sutures on the abdominal terga, by the absence of a median dorsal carina
on the sixth abdominal somite, as well as by the fact that all the dorsal carinz on this
somite and on the telson are more sharply defined and less swollen and rounded than
they are in this species.

While it is a little closer than Gonodactylus graphurus to Gonodactylus chiragra,
there is little difficulty in distinguishing it from the latter species by the presence of six
well-developed marginal spines upon the telson, and of four carine upon the central

64 THE VOYAGE OF H.M.S. CHALLENGER.

In inches and | In thousandths

ees decimals. of total length.
Total length of female specimen from Samboangan Reefs, : : 2°05
Measurements on middle line :—
Rostrum, . . F ‘ E : : 5 i 83
Carapace, . : ; 5 : : : : 1 249
Carapace, including rostrum, . : : : : 68 332
From posterior edge of carapace to posterior edge of third thoracic
somite, . ‘10 49
From posterior edge of third to posterior “edge of fourth thoracic
somite, . ‘10 49
From posterior edge of fourth to posterior edge of fifth thoracic
somite, . 3 ; ; ; ; ; ; 10 49
First abdominal somite, . ; ; ‘ : 4 ‘12 59
Second abdominal somite, . : = 5 ; ; 13 63
Third abdominal somite, . : : : . é 13 63
Fourth abdominal somite, . ‘ : ; ; : 13 63
Fifth abdominal somite, . ; , : : : 18 88
Sixth abdominal somite, . : ; : P ? 16 78
Telson on middle line, ‘ 22 107
Total length of hind body, : : ; : 1:37 668
Total length on middle line, : ; ; 5 2°05 1000
Greatest length of telson, ’ : ‘ . : ; 30 146
Width of rostrum, ; , : ? : 15 73
Width of carapace at antero- lateral angles, e é : : "25 122
Width of carapace (greatest), . : ; ; : : 35 171
Width of third thoracic somite, ; 5 : : 5 39 190
Width of fourth thoracic somite, : j ; A : 39 190
Width of fifth thoracic somite, : : ; : : 39 190
Width of first abdominal somite, : ; ; ; . 39 190
Width of second abdominal somite, . : d : : “39 190
Width of third abdominal somite, : : : : : “39 190
Width of fourth abdominal somite, . : : : : 39 190
Width of fifth abdominal somite, ‘ : “39 190
Width of sixth abdominal somite between postero- Jateral spines, : 32 156
Width of telson between submedian spines, . : : é ‘08 39
Width of telson between intermediate spines, . 3 . 22 107
Width of telson (greatest), . : : “31 151
Length of paddle of exopodite of uropod on dorsal surface, ; : 12 58
Length of second joint of exopodite of uropod on dorsal surface, : 28 137

prominence, which is more protuberant both in G'onodactylus glabrous and in Gono-
dactylus graphurus than it is in Gonodactylus chiragra.

The very close resemblance between the three species shows their close relationship,
and it is quite possible that future discoveries may reveal so many transitional forms
that the sharp lines between the species will break down, but in the present state of our
knowledge of the group I think we are compelled to recognise the three as distinct
species.

REPORT ON THE STOMATOPODA, 65

Genus Protosquilla, n. gen. (Pl. XVI.).

Diagnosis.—Small Stomatopoda with the sixth abdominal somite more or less
completely fused with the telson; the rostrum furnished with long acute median and
antero-lateral spines ; the dactylus of the raptorial claw dilated at the base, and without
marginal spines, the eyes, antennary scales, and uropods small, and the hind body convex.
Ontogeny unknown.

General Description and Remarks.—The following small but very interesting and
remarkable Stomatopoda are very slightly known, and most of the species have, on account
of the absence of spines on the dilated dactylus of the raptorial claw, been placed by various
authors in Latreille’s genus Gonodactylus, although they resemble each other in many
features which are also points of difference from Gonodactylus as well as from all other
Stomatopoda.

While it is also true that they differ greatly among themselves, it is not difficult to
discover a common type of structure which runs through them all, and which should find
its expression in our system of classification. I therefore place them in a distinct genus,
and as many of their distinctive characteristics, such as the small size of their antennary
scales and uropods, the great length of the acutely pointed rostrum, and the union of the
sixth abdominal somite with the telson, are points of resemblance to the Stomatopod
larva, I propose for the genus the name Protosquilla. This name is the more appropriate
inasmuch as all the other Stomatopoda present evidences of divergent descent from a
common stem form, which, like the living representatives of the genus Protosquilla, was
characterised by the small size of its eyes, antennary scales, and uropods. In this genus
the dactylus of the raptorial claw is dilated at the base, with microscopic serrations, but
no spines, on its inner margin; the rostrum has a long slender acute median spine, and
one or more pairs of acute antero-lateral spines; the eyes are small and subcylindrical;
the carapace is smooth and flat, with well-marked longitudinal sutures, but with the
transverse cervical suture obsolete; the hind body is convex, without dorsal carinze on
the first four terga, with its posterior end curved downwards, and with the sixth
abdominal somite more or less completely fused with the telson. The telson and sixth
abdominal somite differ in texture from the general surface of the body, and are often
remarkably and elegantly ornamented ; the telson is usually wider than long, and its mar-
ginal spines are pushed backwards, and are either obsolete or else developed in a remarkable
manner. The eyes, antennary scales, and uropods are very small, and the prolongation
from the ventral surface of the basal joint of the uropod ends in two short stout spines.

All but three of the following species are known to exhibit all of these characteristics,
and while the published accounts of those three are so incomplete as to leave many points
in uncertainty, the close general resemblance between them and the better known species
indicates that future research will show that they share all the characteristics of the genus.

(ZOOL, CHALL. EXP.—PART XLV.—1886.) Yy 9

66 THE VOYAGE OF H.M.S. CHALLENGER.

As regards one of the most remarkable peculiarities of the genus, the pushing
backwards of the marginal spines of the telson, the various species present a most
interesting series of steps in the process of modification. In Protosquilla folinw* the
outline of the telson closely approximates the prevalent type among the Stomatopoda.
There are six marginal spines, and the submedians, which project furthest backwards,
are separated by a well-marked median notch, while the second pair are a little more
anterior, and the third pair still more so, all the marginal spines, however, being
further back than they are in ordinary Stomatopods. In the very closely related species,
Protosquilla elongata, Pl. XV. fig. 2, all the spines are fused into a pair of rounded lobes
on the sides of the median notch, and no traces of the separate spines are visible in a
dorsal view, although their rounded ends project slightly below the general surface of the
ventral side of the telson, as shown in fig. 12.

In Protosquilla trispinosa (Gonodactylus trispinosus, Miers, Squillide, pl. 11. fig. 10)
the median notch is faintly indicated, and all the spines are on the convexly rounded
posterior border of the telson. In Protosquilla cerebralis (Pl. XVI. fig. 2) the telson is
transversely truncated behind, and the spines are all on its posterioredge. In Protosquilla
excavata (Gonodactylus excavatus, Miers, Squillidee, pl. ii. fig. 12) the spines are obscurely
indicated, but the lateral ones are now the longest, and the edge of the telson is deeply
excavated between them, while in Protosquilla furcicaudata (Gonodactylus furcicaudatus,
Miers, Squillide, pl. iii. fig. 14) this excavation is carried so far that the median portion
of the telson has disappeared, and the two marginal spines have approached each other on the
middle line, evidently as the result of an infolding of the median portion of the posterior
edge of the telson of the larva. In Protosquilla guerinii (Pl. XVI. fig. 1), the marginal
spines have undergone a very remarkable modification in a somewhat divergent direc-
tion. They are on the posterior edge, and the laterals project as far backwards as the
submedians, but they have become greatly divided, forming a complicated pectinated
structure,

As regards the telson, Protosquilla folinii is at one end of a series at the other end
of which is Protosquilla furcicaudata, but it is difficult to decide which of these forms
is the one in which the process of modification began. As the telson is long, with a
convex border, in most Malacostraca, and in the Stomatopod larva, it is probable that
Protosquilla folinii is the primitive form, and Protosquilla furcicaudata a highly
modified form, and that the relationship to the other Stomatopoda is through Proto-
squilla folinii, which must therefore be regarded as one of the closest living allies of
the ancestral type of the order.

Outside the genus Protosquilla ; the two species for which I have proposed to establish
the genus Coronida, Coronida bradyi and Coronida trachura, are very similar to
Protosquilla, as are also, but in a less degree, the true Gonodactyli. On account of the

1 Gonodactylus folinit, A. Milne-Edwards, Nouv. Arch, Mus. Hist. Nat., t. iv., pl. xviii. fig. 10.

REPORT ON THE STOMATOPODA. 67

great interest of this genus, I give, in addition to the descriptions of the species which
are represented in the Challenger collection, brief diagnoses of the other known
species.

1. Protosquilla elongata, n. sp. (Pl. XV. figs. 2, 12; Pl. XVI. fig. 4).

Diagnosis.—Protosquilla with the median spine of the rostrum long and
slender, and the antero-lateral spines bifurcated at the tips. Carapace elongated, two-
thirds as.wide as long, with antero-lateral angles acute. Second thoracic somite
exposed and subacute laterally. Hind body wider than carapace. Lateral angles of
third, fourth, and fifth thoracic somites nearly straight, with rounded angles. Sixth
abdominal somite with a thickened transverse ridge along its posterior border,
from which four thickened carinz, two submedian and two submarginal, run forwards
to the anterior edge of the somite, the two outer carinz being obscurely divided
into three lobes. Suture between sixth abdominal somite and telson distinct but
immovable.

Telson a little wider than long, with a thickened median carina, and on each side of
this a very wide and prominent curved lateral carina. Telson with a deep notch on
the middle line behind, and on each side of this an obtusely rounded lobe, the outline of
which is continuous in a dorsal view, but with traces of three marginal spines on the
ventral surface. Endopodite of uropod triangular. First antennze short. An acute
spine on anterior edge of first joint of second antenna.

General Description.—Median spine of rostrum slender and acute, slightly dilated at
base, and reaching nearly to the tips of the eyes, which are cylindrical, with the corneal
portion inclined backwards externally. Antero-lateral angles of rostrum long, acute,
curved forwards, and divided at the tip into two spines one above the other. Carapace
narrower than hind body, much longer (22) than wide, and slightly emarginated. The
antero-lateral angles ending in short spines, the postero-lateral angles rounded. Gastric
area distinct, rectangular, slightly convex, with gastric suture continued nearly straight
to posterior edge. Second thoracic somite exposed, narrow, and subacute at lateral
edges. The following thoracic somites wider than the carapace and nearly as wide as
the hind body. The lateral edge of the third is longitudinally truncated, that of the
fourth obliquely truncated, so that the somite is wider at the antero-lateral than at the
postero-lateral angle. The fifth is produced into a subacute lobe. First five abdominal
somites smooth dorsally, and almost but not quite equal in width; the width of the
first being 7482,, and that of the fifth #§4,;, of the total length. The postero-lateral
angles of the abdominal somites are subacute. The sixth abdominal somite is immoy-
ably united to the telson, but the suture is distinct. The posterior edge of the dorsal
surface is elevated into a thick, obscurely defined, rounded transverse ridge, from which

68 THE VOYAGE OF H.M.S. CHALLENGER.

four thickened carinze run forwards nearly to the anterior edge of the somite, separated
from each other by interspaces which are about equal to the ridges in width.

The two submedian carinz are nearly parallel and longitudinal, while the two outer
or submarginal ones diverge anteriorly, and have their imner edge obscurely divided
into three rounded lobes. The telson is deeply notched on the middle line, and is
slightly wider than long. In a ventral view (Pl. XV. fig. 12) each of the rounded lobes
which borders the median notch is seen to consist of three marginal spines, crowded
backwards, but with the submedian longest, the intermediate shorter, and the external
shortest. No trace of these spines is visible in a dorsal view, and the dorsal outline of
the telson consists of a single obtusely rounded lobe on each side of the triangular
median notch, which is fringed with hairs.

The middle of the dorsal surface of the telson is occupied by a broad wedge-shaped
median ridge or carina, which is wide anteriorly and narrows posteriorly to a rounded
point which is scarcely elongated above the general surface of the telson. This carina
is surrounded on all sides, except at its anterior end, by a depression or groove, outside
which there is on each side a greatly thickened submedian carina or ridge, the general
course of which is longitudinal, although it is slightly curved, with the convexity away
from the middle line.

The tips of the extended uropods reach nearly as far backwards as the telson. The
paddle of the exopodite is very small, that of the endopodite longer and nearly
triangular. The basal joint has a single acute spine on its dorsal surface, and the two
ventral spines are short and stout, the outer being the larger and longer.

The eyes are cylindrical, with the convex curved portion obliquely placed and
running backwards on the outside of the base. The ocular segment is exposed, broad,
and deeply notched on the middle line under the median spine of the rostrum. The
antenne are short, and the basal joint of the second ends in an acute spine. The
dactylus of the raptorial claw is greatly swollen at the base, with a deep notch on its
outer edge. The tip of the dactylus is gently and regularly curved to the acute point,
and its inner edge is barbed with minute serrations, nearly to the base. The edge
of the second joint is bordered by minute dentations,

The exposed thoracic legs are long and slender.

The dorsal surface of the body, with the exception of the sixth abdominal somite and
telson and uropod, is highly polished, while these are rough.

Habitat.—One female specimen from St. Vincent, Cape Verde Islands.

Colour.—In the alcoholic specimen the telson and sixth abdominal somite and
uropods are dirty-white, while the rest of the body is highly polished and very dark
brown, except a transverse strip of light yellowish-brown .across the carapace, posterior
to the middle point of its length, and a corresponding light band across the enlarged
joint of the second maxilliped.

REPORT ON THE STOMATOPODA, 69

|
In inches and | In thousandths

Be eeemenky decimals, of total length.
On middle line :—
Rostrum, . : ¢ : : : : , ‘07
Carapace, . : ; P ; : : : 17 221
Carapace, including rostrum, . : C ; ? 24 312
From posterior edge of carapace to posterior edge of third thoracic
somite, . 042 55
From posterior edge of third to posterior "edge of fourth. thoracic
somite, . 043 56
From posterior edge of fourth to posterior edge of fifth thoracic somite, 042 55
First abdominal somite, . ‘ : ‘ : 043 56
Second abdominal somite, . , ; : : : 043 56
Third abdominal somite, . d 4 : : : 052 68
Fourth abdominal somite, . : : ; : A 060 78
Fifth abdominal somite, . : ! ‘ : ? 073 95
Sixth abdominal somite, . P ie : : é 057 74
Telson on middle line, : ; . : i : 072 94
Length of hind body, . ; : é ‘ ; 527 687
Total length on middle line, : : ; - ‘767 999
Greatest length of telson, : : 5 t 10 130
Width of carapace between antero-lateral spines, : ; : 092 120
Width of carapace (greatest), . q : > : ; 115 149
Width of third thoracic somite, ; : : . ; 128 165
Width of first abdominal somite, - F ; 4 ‘ 125 163
Width of fifth abdominal somite, ? : : ; ; ‘141 184

Size.—Length, 74% inch.

‘Remarks.—A comparison of this species with the descriptions of Gonodactylus folinii
which are given by A. Milne-Edwards* and Miers? would at first sight seem to indicate
that they are the same as this species, as there is a very close general resemblance, as
well as an almost perfect agreement in size, and Milne-Edwards’ single specimen was
obtained at St. Vincent, the place where the Challenger specimen was collected.
Neither of these authors notices the bifurcation of the antero-lateral spines of the
rostrum, the elongation of the carapace or the spines at its antero-lateral angles, but as
these pomts might easily have been overlooked on account of the great minuteness of
the specimens, I was at first inclined to believe that they are the same, but careful
examination of Milne-Edwards’ figures brings out so many important points of difference
that I am forced to conclude that there are at St. Vincent two closely related species of
this minute type, or else that the figures which are given by Milne-Edwards are so
inaccurate as to be of no value. Which of these alternatives is the true one can be
decided only by renewed examination, and as our specimen differs very essentially from

1 Now. Arch. Mus, Hist. Nat., t. iv. p. 65. 2 Ann. and Mag. Nat. Hist., ser. 5, vol. v. p. 123.

70 THE VOYAGE OF H.M.S. CHALLENGER.

Milne-Edwards’ figures of Gonodactylus folini, and as it also fails to agree perfectly with
his and Miers’ descriptions, it must for the present at least be regarded as a distinct
species, differing from Protosquilla (Gonodactylus) folinii in the following points :—
Antero-lateral angles of rostrum simple in Protosquilla folinii, double in Protosquilla
elongata. Posterior margin of telson with three rounded spines on each side of the deep
median notch in Protosquilla (Gonodactylus) folinu, and a single obtusely rounded lobe
in Protosquilla elongata. .

If A. Milne-Edwards’ figures are trustworthy the two species also differ in the
following features :—First antennze very long in Protosquilla folinii, short in Protosquilla
elongata. No spine on anterior edge of basal joint of second antenna in Protosquilla
folinit, a long spine in Protosquilla elongata. Carapace nearly as wide as long,
rectangular, and with rounded antero- and postero-lateral angles in Protosquilla folini ;
elongated, emarginated, narrow anteriorly, and with antero-lateral angles acute in
Protosquilla elongata. Thoracic and abdominal somites rounded in Protosquilla folinii ;
lateral margins of thoracic somites of Protosquilla elongata truncated, and postero-lateral
angles of abdominal somites acute.

The species is of great interest as it and the closely related Protosquilla folinii
are the most primitive forms which are known among the Stomatopods, exhibiting
the embryonic or larva] features which are characteristic of the genus, such as the
fusion of the telson with the sixth abdominal somite, the small size of the uropods, and
_ the presence of three long spines on the rostrum; without exhibiting any of the remark-
able modifications of the posterior end of the body which are so strikingly characteristic
of most of the species of the genus. The telson of an ordinary Squilla, that of Coronis,
or that of the most aberrant Protosquilla, such as Protosquilla guerinit, may all be
derived from that of Protosquilla folinii by very slight changes, and the carapace of
Protosquilla elongata might be readily modified into that of a true Squilla, while it is
also very similar to the flat square carapace of the other members of its own genus.

2. Protosquilla folinii (Milne-Edwards).

Gonodactylus folinii, A. Milne-Edwards, Obs. sur la Fauna Carcinol. des Iles du Cap-Vert,
Nouv. Arch. Mus. Hist. Nat., t. iv. 1868, p. 65, pl. xviii. figs. 8-11.

Gonodactylus Folinti, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5, vol. v.,
1880, p. 123.

Diagnosis.—Protosquilla with the rostrum ending in a long slender acute
median spine, and on each side of this a single acute antero-lateral spine which is directed
outwards and forwards and is a little shorter than the median spme. Carapace with
rounded antero- and postero-lateral angles. First five abdominal somites smooth,
Dorsal surface of sixth abdominal somite with four rounded tubercles, the outer ones

REPORT ON THE STOMATOPODA. 1

three-lobed. Dorsal surface of telson with a broad unarmed median ridge, and on each
side of this a more protuberant longitudinal and somewhat uniform prominence, which is
convex on its lateral or external side, and concave towards the middle line. The telson
has a median notch on its posterior edge, and on each side of this there are three obtuse
rounded marginal spines, of which the submedian is farthest back, and the lateral most
anterior.

Length.—About three-fourths of an inch.

Habitat.—A. Milne-Edwards records a single specimen of this species from St.
Vincent, and Miers a single male from Mauritius, and two small specimens, without
record of locality, from the “ Herald” collection,

Remarks.—In Milne-Edwards’ figures the carapace is represented rounded at its
_ anterior as well as at its posterior angles, and nearly rectangular, and the fifth thoracic
somite is represented with rounded lateral edges, as are all the other thoracic and
abdominal somites,

In my remarks on Protosquilla elongata I have given my reasons for regarding that
species as distinct from this one.

3. Protosquilla trispinosa (White).

Gonodactylus trispinosus, White, List Crust. Brit. Mus., 1847, p. 85.

Gonodactylus trispinosus, Dana, J. D., U.S. Explor, Exped., Crustacea, i. p. 623, 1852.

Gonodactylus trispinosus, Heller, C., Reise der N ovara, 1868, Crustacea, p. 126.

Gonodactylus trispinosus, Miers, E. J., Cat. New Zealand Crust., 1876, p. 126.

Gonodactylus trispinosus, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v., 1880, p. 121, pl. iii. fig. 10 (telson).

Gonodactylus trispinosus, Miers, E. J., Malaysian Crustacea, Ann. and Mag. Nat. Hist., ser. 5,
vol. v. p. 460.

Diagnosis.—Protosquilla with the two antero-lateral spines of the rostrum nearly as
long as the median spine. The carapace is Jong, and nearly rectangular, with nearly
rectangular antero- and postero-lateral angles. The fifth abdominal somite is longi-
tudinally corrugated. The sixth is immovably united to the telson, although the suture
is clearly indicated ; its dorsal surface is marked by six smooth rounded tubercles. On
the dorsal surface of the telson there are three smooth rounded tubercles disposed in a
triangle, with the median one anterior to the two laterals, The posterior margin is
slightly notched in the middle, ‘and is armed with very minute spines.

Size.—About one and one half inches long.

Habitat.—Fiji Islands, Dana; Auckland, Heller; Mauritius, ? Hoffmann; Swan River,
Australia, Miers; Shark’s Bay, Australia, Miers; Amboina, Miers; Ceylon, Miers (var.
pulchella).

Remarks.—According to Heller’s description the telson is rectangular and is armed

72 THE VOYAGE OF H.M.S. CHALLENGER,

with four marginal spines, while Miers represents it as broadly rounded posteriorly and
armed with six minute marginal spines. It therefore seems possible that there may be
two closely related forms instead of a single species.

Although this species is much better known than any other Protosquilla no figure
has ever been published except the drawing of the telson given by Miers. A minutely
accurate figure of the entire animal is very much needed. All the specimens the sex of
which is given are males, except the single female recorded by Miers from Amboina.

4, Protosquilla cerebralis, n. sp. (Pl. XIV. figs. 2, 3; Pl. XVI. figs. 2, 3).

Diagnosis.—Protosquilla with the body almost uniform in width from the anterior
edge of the carapace to the posterior edge of the telson. The two slender acute antero-
lateral spines of the rostrum are nearly as long as the acute median spine. The carapace
is rectangular, with rounded angles, and the fifth thoracic somite is slightly exposed
dorsally. No dorsal carinze on the carapace or on the first seven exposed somites.
Lateral processes of the sixth and seventh thoracic somites longer than those of the
eighth, and rounded. Postero-lateral angles of all the abdominal somites obtuse. Dorsal
surface of fifth abdominal somite with a median thickened carina, and, on each side of
this numerous sublongitudinal ridges, converging posteriorly. Sixth abdominal somite
immovably fused with the telson and ornamented by swollen convoluted ridges
symmetrically arranged in a complex pattern. Telson nearly twice as wide as long,
deeply notched on the middle line behind, with its posterior margin rounded and ending
in three pairs of short acute spines. Dorsal surface of the telson sculptured in a
complicated symmetrical convoluted pattern, consisting of a very complex central pattern
which is subtriangular and divided into five principal convolutions; two submedian
convex oval convoluted lobes, and a number of parallel longitudinal lateral ridges.
Dactylus of raptorial claw greatly swollen at base.

General Description.—The median spine of the rostrum is slender, acute, longer than
the antero-lateral spines, and reaching about halfway to the tips of the eyes, which are
subcylindrical, with the corneal portion inclined backwards externally. The antero-lateral
spines of the rostrum are slender and acute. The flat carapace is a little longer than
wide, slightly narrower than the exposed thoracic somites, and nearly rectangular with
the antero- and postero-lateral angles, rounded and nearly alike. The two longitudinal
sutures are prominent and continuous from the anterior to the transverse posterior edge,
and there is a faintly marked transverse cervical suture. The sixth and seventh thoracic
somites are wider than the eighth, and their lateral edges are produced backwards and
rounded, as seen from above, while in profile the anterior angle is obtuse and the
posterior one nearly a right angle. The middle of the lateral edge of the eighth is
produced into a subacute lobe over the basal joint of the appendage. The eighth is a

REPORT ON THE STOMATOPODA. 73

little longer and narrower than the sixth and seventh thoracic somites, and a little
shorter than the first four abdominal somites, which are about equal in length, as wide as
the sixth and seventh thoracic, smooth on the dorsal surfaces, and with marginal keels or
ridges on their lateral edges, the postero-lateral angles of which are nearly right angles.
The fifth abdominal somite is longer than, and quite different from, those in front of it, as
its dorsal surface is sculptured by numerous sublongitudinal ridges symmetrically
arranged on each side of a broad median ridge or carina.. The ridges are not quite as
long as the somite, which has along its anterior and posterior borders smooth bands
which are not broken into ridges. The lateral ridges are slightly convergent towards the
posterior edge, and their anterior ends are lobed and sometimes irregularly forked. The
sixth abdominal somite is shorter on the middle line than at the edges and immovably
fused with the telson which is transverse, and nearly twice as wide as long. The
posterior end of the body is bent downwards, so that the dorsal surface of the telson faces
backwards. The sixth abdominal somite is sculptured in a symmetrical pattern of broad
convoluted ridges and tubercles, which are so closely crowded together that the dorsal
surface is almost covered with them. The dorsal surface of the telson is also sculptured
in a complicated pattern, in which six elements may be recognised, first, a transverse
ridge lobed on its posterior edge running along the anterior border of the telson ; second,
a triangular median area, the broad anterior end of which is elevated above the general
surface of the telson, while its narrow posterior portion is sunk between two lateral
elevations. The median area is convoluted and divided into five lobes, a median anterior,
two antero-laterals which are again subdivided, and two posterior lobes or ridges, one
of which runs backwards on each side of the deep narrow postero-median notch ; third,
on each side of the median area an elongated oval elevated area, which is divided into
halves by a longitudinal furrow, and also divided by lateral furrows into a number of
secondary lobes; fourth, three thickened longitudinal marginal keels on each side
external to the oval elevation. The six short acute marginal spines are crowded together
in two sets, on the posterior edge of the telson, with a broad space fringed with small
spines between the submedians.

The uropods are small, and the basal prolongation (PI. XVI. fig 3) ends in two short
stout acute spines, the outer larger and longer than the inner, and with a slight enlargement
on its inner edge. The row of marginal spines, about ten in number, on the outer edge
of the second joint of the exopodite curves upwards on to the dorsal surface at its proximal
end. The inner edge of the dactylus (Pl. XIV. fig. 3) of the raptorial claw is finely serrated,
and the basal enlargement on the outer edge is semi-circular, abruptly limited distally,
and indented on its outer margin.

Size.—The length of the single specimen is 1,%%5 inches from the tip of the rostrum
to the middle line of the telson.

Habitat.—The single female specimen is from the reefs at Levuka, Fiji.
(z0oL. OHALL, EXP.—PART XLy.—1886.) Yy 10

74 THE VOYAGE OF H.M.S. CHALLENGER.

In inches and | In thousandths
Measusemicnts decimals. of total length.
Total length from tip of rostrum to tip of telson on middle line, ; 1:27

Rostrum, . : . : ; : : : 08 62
Carapace, . 4 ; : : : 5 : 30 234
Carapace Including rostrum, . ; : : : 38 296

From posterior edge of carapace to posterior edge of third thoracic
somite, . 05 39

From posterior edge of third to posterior ‘edge of fourth thoracic
somite, . 06 47

From posterior edge of fourth to posterior edge of fifth “thoracic
somite, . 5 : . : ; 5 08 © 62
First abdominal somite, " : : : : 5 ‘10 78
Second abdominal somite, . é : : : - ‘09 70
Third abdominal somite, . 5 3 i F ‘ 09 70
Fourth abdominal somite, . : ; ; ; : 12 93
Fifth abdominal somite, . . ; ; ; ; pil) 101
Sixth abdominal somite, . ‘ 5 , 7 ‘07 55
Telson on middle line, ; : F owe : F 10 78
Total length of hind body, , . ; ‘ “89 673
Total length on middle line, : : 5 ; 1:27 969
Greatest length of telson, . : 3 , ‘14 109
Width of carapace between antero-lateral angles, : : é 18 140
Width of carapace (greatest), . : F : ; ; 23 179
Width of third thoracic somite, F 3 , : - “25 195
Width of fourth thoracic somite, F 3 : é 2 26 199
Width of fifth thoracic somite, : : ; : ; 24 187
Width of first abdominal somite, : : : : : Di, 210
Width of second abdominal somite, . ; ; : 4 27 210
Width of third abdominal somite, j j ‘ , ; 27 210
Width of fourth abdominal somite, . 5 ; ; 5 27 210
Width of fifth abdominal somite, 5 : 3 : ‘ 27 210
Width of sixth abdominal somite, ; : : : 26 199
Width of telson between postero-median spines, ; 3 , 08 62
Width of telson (greatest), ; 26 199
Length of paddle of exopodite of sixth abdominal appendage, é 06 47
Length of second joint of exopodite of sixth abdominal appendage, . 10 78
Length of endopodite of exopodite of sixth abdominal appendage, : 10 78
Distance between tips of sixth abdominal appendage, . . ; ‘D7 444
Length of first antenna, measured from tip of rostrum, . é 3D 273
Length of flagellum of second antenna, : : : : 20 156
Length of scale of second antenna, . : : : : 13 101
Length of eye, : 5 ; ; : 5 : 09 70

Colowr.—The fifth and sixth abdominal somites and the telson and uropods of the
alcoholic specimen are colourless and rough, while the rest of the body is highly polished,
and mottled with brown pigment.

Remarks.—The only species which is at all similar to this is Wood-Mason’s Gono-
dactylus glyptocercus from the Nicobars, but his description, which is given below, is so

REPORT ON THE STOMATOPODA. 75

brief and imperfect that it is impossible to decide how closely related these two are. They
may be the same species, but_so far as the description shows, it is quite possible that the
resemblance is a superficial one only, and that they are not at all similar in their essential
characteristics.

5. Protosquilla glyptocerca (Wood-Mason).

Gonodactylus glyptocercus, Wood-Mason, Proe. Asiatic Soc. Bengal, p. 232, 1875.

Gonodactylus glyptocercus, Wood-Mason, Ann, and Mag. Nat. Hist., ser. 4, vol. xvii. p. 263, 1876.

Gonodactylus glyptocercus, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v. p. 123, 1880.

In his description of this species, which is very brief, he does not say whether the
sixth abdominal somite is fused with the telson. It is possible that it is not related in
any way to the species of Protosquilla, but I place it here provisionally as the only
points he mentions show a striking likeness to Protosquilla cerebralis, with which it
may possibly be identical.

His short notice simply says that it is like Gonodactylus trispinosus, but that the
terminal post-abdominal somite is ornamented with two oval tubercles, bounded by an
impressed inverted line, and with a medium cinquefoil-shaped one, and the two preced-
ing segments symmetrically engraved with fine lines.

6. Protosquilla guérinii (White) (Pl. XVI. figs. 1 and 6.)

: Gonodactylus guérinii, White, A., New Crustacea, Proc, Zool. Soc. Lond., January 22, 1861,
p. 43, pl. vii.
Gonodactylus guérinii, White, A., Ann. and Mag. Nat. Hist., vol. vii., 1861, p. 476.
Gonodactylus guérinii, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v., 1880, p. 121.

Diagnosis.—Protosquilla with eyes slightly enlarged at tips, carapace as wide as
long. Second thoracic somite exposed. First five abdominal somites with marginal
carine. Fifth abdominal somite with several transverse rows of short acute dorsal spines
on its posterior half. Sixth abdominal somite covered on its dorsal surface with
numerous long spines. Posterior edge of telson with two pairs of marginal spines pro-
jecting backwards. The submedian with a single row of secondary spines on each side,
and the laterals with a single row on the inner and a double row on the outer edge,
running from its tip along the outer edge of the telson to its anterior end, Dorsal
surface of telson with twenty-two long spines symmetrically arranged. Second
joint of exopodite of uropod nearly three times as long as paddle. Dilated base of
dactylus of raptorial claw truncated on its outer edge, and not notched. Endopodite of
uropod with five spines on its dorsal surface.

76 THE VOYAGE OF H.M.S. CHALLENGER.

General Description.—The Challenger collection contains one female specimen from
Honolulu of this species, which has previously been represented by the single specimen
which was obtained by the “Herald” from Matuku, Fiji Islands. It may be dis-
tinguished without difficulty from all other known Stomatopods, except A. Milne-
Edward’s Squilla brady, by the long spines which cover the telson and sixth abdominal
segment, and from this species by the length of the median spine of the rostrum, and
the absence of spines on the inner edge of the dactyle of the raptorial claw.

Transverse diameter of rostrum about equal to its length; antero-lateral spines of
rostrum acute, but very slightly produced forwards, and shorter than the median spine,
which ends in an acute point opposite the middle of the eye. The carapace is nearly
square, as long as wide, slightly narrowed anteriorly, and with antero- and postero-
lateral angles rounded and alike. The gastric area is slightly,convex’and its lateral
sutures are sharply defined, while the transverse cervical suture is almost obsolete.
Carapace wider than thoracic region, and about as wide as the abdomen. First thoracic
somite exposed and subacute at lateral edges. Third, fourth, and fifth thoracic somites
so articulated as to form a convex dorsal protuberance. Lateral edges of third and
fourth obliquely truncated converging posteriorly, with rounded angles. Lateral edges
of fifth subacute. The abdomen increases slightly in width from in front backwards,
but not so rapidly as in White’s figure. The abdominal somites have faintly marked
marginal carine or ridges, and the first four have on each side, at about one-third of the
distance from the lateral edge to the middle line, a small sharply defined indentation.
The sixth abdominal somite is immovably united to the fifth as well as to the telson,
although the sutures are distinctly visible. The first four abdominal somites are smooth
dorsally, as is“also the anterior half of the fifth, a transverse line separating it from the
posterior half, which is armed by about six transverse rows of numerous short spines,
which increase in length and in the definiteness of the transverse rows as they approach
the posterior margin of the somite. The dorsal surface of the sixth abdominal somite is
armed with very numerous (about fifty-six) long cylindrical spines, each of which ends
in a blunt rounded tip with a perforation from which a soft tubular fleshy process
protrudes. Eight of these spines are arranged in a transverse line along the posterior border
of the somite, and are arranged symmetrically in pairs with reference to the middle line,
about twenty more form an anterior transverse band which, starting at the postero-
lateral angle, runs obliquely forwards to the anterior edge of the somite, along which it is
continued across the middle line. The triangular antero-lateral area, external to this
line, is occupied by about fourteen crowded spines on each side.

The suture between this somite and the telson is distinct, but immovable, and
bordered by a raised ridge.

The centralarea of the telson is occupied by twenty-two similar long spines, with
fleshy appendages, symmetrically arranged with reference to the middle line, and so

REPORT ON THE STOMATOPODA. v7

In inches and | In thousandths

ESE UReen: decimals. of total length.
On middle line.

Rostrum, . 2 : , ; ; ; ‘07 62
Carapace, . é ; E : , ; : 23 205
Carapace including rostrum, . 30 267

From posterior edge of carapace to posterior edge of third tharacic
somite, . 05 45

From posterior edge of third to posterior ‘edge of fourth "thoracic
somite, . 07 62

From posterior edge of fourth to Posterior edge of fifth thoracic
somite, . : c 09 81
First abdominal somite, : : 5 : : ‘ ‘07 62
Second abdominal somite, : ‘ ‘ ; 3 - 08 72
Third abdominal somite, . : F : , : 07 62
Fourth abdominal somite, . : ; : F ; ‘08 72
Fifth abdominal somite, . : ; P ; F “15 ) 98
Sixth abdominal somite, . : F ‘ : : 05 45
Telson on middle line, : : : ; ‘ / 15 134
Total length of hind body, ‘ : : : 82 733
Total length on middle line, d : : : 1:12 1000
Telson to tips of spines, itor : 24 234
Width of carapace between antero-lateral spines, , : : 19 169
Width of carapace (greatest), . : : : F 24 234
Width of fourth thoracic somite, : ‘ : ; ; 20 178
Width of first abdominal somite, ; A $ : : ‘21 187
Width of fifth abdominal somite, : : : : 25 223
Width of telson between bars of lateral spines, i ; : * -28 249
Width of telson between tips of lateral spines, . 4 : lit 151

placed that they fall into rows longitudinally and transversely. Two are on the middle
line, and the others in pairs. There are five in the first transverse row, six in the next,
two in the third, five in the fourth, and four in the last. The free edge of the telson
carries two pairs of primary marginal spines which are pushed back into the posterior
border. The submedians are slightly longest, acute, separated by a deep median notch,
and with twelve long acute secondary spines on their inner and nine on their outer edge.
The two lateral spines are also acute, and have six acute long secondary spines on their
inner edge, and on their outer edge very numerous (twenty-four or more) long spines
placed in two rows, one above the other, and running along the outer edge of the telson
to its anterior end.

The uropods are very short and small; the paddle of the endopodite is elongated
oval with five acute immovable spines placed in a longitudinal row on its dorsal surface,
the paddle of the exopodite is very small, the second joint long with eight movable spines
on its outer edge. The spines on the basal prolongation are short and acute, and the
inner one very much shorter and more slender than the outer one.

78 THE VOYAGE OF H.M.S. CHALLENGER.

The basal enlargement on the dactyle of the raptorial claw is not notched, its outer
border is truncated, and at its distal edge the transition is abrupt to the long slender
gently curved acute tooth, which is barbed with minute serrations on its inner edge,
while the outer edge of the second joint is finely dentate, with acute teeth.

The eyes are subeylindrical, with the coronal portion a little wider than the basis.
No secondary differences between the sexes, except in the first abdominal somite,

Halbitat.—The Challenger collection contains a single female specimen from
Honolulu. The only other known specimen is a male from Matuku, Fiji Islands,
described by White and also by Miers.

Colour.—White describes his dry specimen as “marbled,” of a light yellowish-
brown colour, varied with a darker colour. The Challenger specimen, preserved in
alcohol, has a broad transverse light band across the carapace, while the rest of the
dorsal surface is marked with brown pigment.

Length, 1,1,3,ths inches.

7. Protosquilla excavata (Miers).

Gonodactylus excavatus, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v. p. 123, pl. iii. figs. 11, 12.

Diagnosis.—Antero-lateral angles of carapace subacute, antero-lateral angles of
rostrum acute, slender, shorter than elongated median spine. Fifth abdominal somite
smooth. Sixth with six irregular longitudinal prominences, separated by deep inter-
spaces, and confluent distally. Telson deeply excavated posteriorly, and with five
prominent smooth longitudinal obtuse keels, the submedians longest and reaching to
the postero-lateral lobe. First antennz very long.

Length, ?th inch.

Locality.—Miers describes this species from a single male specimen of unknown
origin. |

Remarks.—The published description does not state that the sixth abdominal
somite is fused with the telson.

8. (2) Protosquila furcicaudata (Miers).

Gonodactylus furcicaudatus, Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5,
vol. v., 1880, p. 124, pl. iii. figs. 13-16.

Although Miers does not state whether the sixth abdominal somite of this species
is fused with the telson, it exhibits so many features of resemblance to the Protosquille
that I place it here provisionally. It may be briefly described as follows. Eyes,
antenne, rostrum, carapace, exposed thoracic somites, and first five abdominal somites as

REPORT ON THE STOMATOPODA. 79

in Protosquilla. The sixth abdominal somite narrow, with six longitudinal promin-
ences. Telson very short and transverse, with two slender spines projecting backwards
from its posterior margin, with their bases in contact on the middle line, and their
tips slightly divergent.

Comparison with the other Protosqwillz indicates that the two long spines of this
species are the external marginal spines of the telson, which have become approximated
on the middle line by the disappearance of almost the whole of the posterior border
of the telson.

Genus Coronida, n. gen.

The two following species Gonodactylus bradyi of A. Milne-Edwards, and Gono-
dactylus trachurus, Miers, resemble each other very closely and present many points of
resemblance to the genus Protosquilla, but they are quite anomalous in other respects,
since they present a most interesting and suggestive resemblance to Squilla, Lysiosquilla
and Gonodactylus. Their features of resemblance to Gonodactylus, the enlargement of
the base of the dactylus of the raptorial claw, the flat rectangular carapace, and the pointed
rostrum, are at the same time features of resemblance to Protosquilla, and it seems
probable that they are the living representatives of an ancestral type which was closely
related on the one hand to Protosgwilla, while on the other hand it was very similar to
the common type of which Lysiosquilla and Squilla are the divergent descendants. As
the most primitive species of the genus Squilla are often placed in a distinct genus
Chlorida, while the lowest members of the genus Lysiosquilla are often placed in a
distinct genus, Coronis, and as the features of resemblance to Squwilla and Lysiosquilla,
exhibited by the species now under discussion, point toa relationship with the lower rather
than with the higher forms in these genera, I propose for them a generic name which
shall express this fact, and the generic term Coronida is a compound of the first two
syllables of Coronis, with the last two of Chlorida.

The two species of Coronida resemble Protosquilla in the minuteness of the antennary
scales and uropods ; in the fact that the anterior somites of the hind body are smooth,
and conspicuously different from the sixth abdominal somite and telson, as well as in the
presence of a median spine on the rostrum, and the enlargement of the base of the
dactylus of the raptorial claw, and the shape of the carapace.

It is not stated that the sixth abdominal somite is fused with the telson, but the
markings on the posterior end of the body are strikingly like those of Protosquilla
guerinu, since Milne-Edwards describes this part of the body of Coronida bradyi in
words which perfectly fit Protosquilla querinii, as uniformly covered with numerous
closely placed slender spines which are longest near the posterior margin, while Miers
says that the posterior half of the fifth abdominal somite of Coronida trachura is

80 THE VOYAGE OF H.M.S. CHALLENGER.

minutely spinulose on its posterior margin, a characteristic which is exactly,duplicated in
Protosquilla gueriu.

The two species differ from Protosqwilla in the absence of antero-lateral spines on the
rostrum, and the enlargement of the tips of the eyes, and from both Protosquilla and
Gonodactylus in the presence of spines on the inner edge of the dactyle of the raptorial
claw, and the depression of the hind body; resembling both Squilla and Lysiosquilla in
these features.

The relationships of Coronida are therefore expressed in the accompanying diagram.

Lysiosquilla. Squilla.

Coronis. Chlorida.

\

_ Coronida. Gonodactylus.

A

SY
Protosquilla.

Diagnosis.—Hind body depressed, dactylus of raptorial claw dilated at base, and
armed with spines on its inner edge. Rostrum ending in a small median spine.
Antennary scales and uropods very small. Terminal segments of hind body and telson
thickly set with small spines. Carapace flat and nearly rectangular.

Coromda bradyi (A. Milne-Edwards).

Gonodactylus bradyi, A. Milne-Edwards, Fonds de la Mer, vol. ix., 1869, p. 137, pl. xvii. fig.
11; Miers, E. J., On the Squillide, Ann. and Mag. Nat. Hist., ser. 5, vol. v., 1880, p. 117.

Gonodactylus bradyi, Miers, E. J., Crustacea from Mauritius, Proc, Zool. Soc, Lond., January
5, 1884, p. 16.

Coronida trachura.

Gonodactylus trachurus, Miers, E. J., Crustacea from Mauritius, Proc. Zool. Soc. Lond.,
January 5, 1884, p. 16.

REPORT ON THE STOMATOPODA. 81

The Alina Larva and the Metamorphosis of Squilla.

Although, as Claus has well shown in his classical monograph on the metamorphosis
of the Stomatopoda,’ the Alima larva is connected with the Erichthus larva by so many
intermediate forms that it is difficult to draw an absolute line between them, it is never-
theless true that the Alima is more different from the Hrichthus than any of the various
modifications of the latter type are from each other, and in the description and discussion
of the Alima larva which follows, I shall give my reasons for believing that all the Alime
are the larve of adults which belong to the genus Squilla, and that all the species of
this genus pass through an Alima stage, while all the other Stomatopods pass through
their larval life as Hrichthi. The Alima larva is undoubtedly a modified Erichthus, and
some species deviate much more widely than others from the Hrichthus type, but the
group is on the whole sharply defined, and the rich supply of Alima larvee brought home
by the Challenger furnishes us with a very complete series of stages in the growth and
development of several species of Alima, and thus shews that the history of all of them
is essentially the same, and that they differ from all the other Stomatopods in the
possession of numerous common characteristics which are also points of resemblance to
the adult Squilla, a conclusion which receives added weight from the fact that Faxon has
reared a young Squilla empusa from an Alima larva.

The complete history of the Alima which is furnished by the Challenger material is
all the more valuable since Claus, who has given us, in his paper above quoted, a very
complete history of the young stages of the Erichthus larva in all its more important
modifications, had access to much more scanty material for studying the Alima. It is
true that he gives figures and descriptions of many forms, but they are all well advanced
and have the same number of somites and appendages as the adult Stomatopoda, and the
fact that the Challenger collection contains consecutive series of several species of Alima
from a very early stage up to the mature larva, with unmistakable characteristics of the
genus Squilla, is therefore of great scientific interest.

The fully grown Alima is usually much larger than any of the Hrichthi, and among
the largest known pelagic larve. It leads an active swimming life, pursuing and
capturing with the greatest rapacity the Copepods and other small Crustacea which form
the chief part of its food. Its metamorphosis is slow, and the wide distribution of most
of the species of Squilla is undoubtedly due to the fact that the larva is carried to distant
localities by the winds and currents, but notwithstanding the great size, often 2 inches or
more, which is attained by the fully-grown larva, the young Alima, even of the largest
species, is very minute, and it is probable that all Alimex hatch from the egg in the Alima
form and that the Hrichthoidina stage has been entirely dropped from their metamorphosis.

1 Metamiorphose der Squilliden, Abhandl. d. hk. Gesellsch. d. Wiss, Gottingen, Bd. xvi. pp. 111-163, Tafs. i-viii.,

1871.
(ZOOL. CHALL. EXP.—PaRT xLy.—1886.) Yy 11

82 THE VOYAGE OF H.M.S. CHALLENGER.

Although Claus decided that they are young Lysvosquille they show their relationship
to the genus Squilla as distinguished from Lysiosquilla by the following characteristics,
all of which are shared by all fully-grown Alima larve. The dactylus of the raptorial
claw has on its inner edge a small number of marginal spines, usually about five or six; the
hind body is wide and flat, and the postero-lateral angles of the abdominal somites end
in acute spines. The outer edge of the proximal joint of the uropod is bordered by a
small number of spines, usually less than eight, and the inner one of the two spines on
the ventral process from the posterior edge of the basal joint of the uropod is longer
than the outer, and it has a tooth or lobe on its outer edge; and the telson has six
marginal spines with minute secondary spines between the submedians, and four or
more larger secondary spines between the submedian and the second or intermediate,
and usually a single one internal to the base of the third or lateral marginal spine.
While it is true that all of these characteristics are not exhibited by every adult Sqwilla,
there are no Stomatopods except those of this genus in which they are all united, and
they are all of them present in most Squillz and in all the Alime.

Joined to the fact that Faxon has reared a Squilla from an Alima, they indicate very
clearly that the Alimex are Squilla larve, and as we know of no other type of larva
which can be referred to this genus, the present state of our knowledge indicates that all
Squille have Alima larve. As it can be shown very conclusively that the Alimea is an
Erichthus, which has become accelerated in development and has dropped its Hrichthoi-
dina stage, and has become peculiarly adapted for a rapacious pelagic life, it is highly
improbable that this change has taken place more than once, and as I shall give reasons
for believing that it occurred very soon after the evolution of the genus Squilla, and that
the larva of the most primitive of the true Squillz is an Hrichthus-like Alima, the con-
clusion that all the Alime are Squilla larvee is warranted by the facts. The validity of
these general conclusions can be better estimated after reading the descriptions of the
larvee which follow, but these descriptions will be the more intelligible if an outline of the
generalizations to be drawn from them be kept in mind.

The Alima larva is characterised by the great elongation of the body, the possession
of a flattened elongated carapace, with the posterior median dorsal spine absent or rudi-
mentary, the elongated abdomen and usually several of the posterior thoracic somites
exposed behind the posterior edge of the carapace, by the very great elongation of the
region between the antennz and the labrum, and by the fact that the eyes and eye stalks
are usually exposed on the sides of the long slender rostrum. The carapace is narrow
and its width is usually one-third or one-fourth of its length, although its total length
makes a much smaller part of the total length of the body than it does in the Hrichthus
larva, as its increased length is more than overbalanced by the great elongation of the
hind body. The older specimens of Alma are usually much larger than most of the
Erichthus larvee, and the inner one of the two spines which project backwards from the

REPORT ON THE STOMATOPODA. 83

ventral surface of the basal joint of the uropod is usually much longer than the outer,
and is usually furnished with a tooth on its outer edge, and in the older specimens there
are indications, underneath the cuticle, of spines on the inner edge of the dactylus of the
raptorial claw. In his classical paper on the Metamorphosis of the Stomatopoda, Claus
pointed out that there are so many features of resemblance between the Alima and the
various forms of Hrichthus that the larval nature of Alima cannot be doubted. Milne-
Edwards and Dana had found it difficult to draw any line between the two genera; the
first-named writer placing in the genus all Erichthide in which the ocular segment is
exposed, while Dana includes in it those forms in which the distance from the anterior edge
of the carapace to the mouth is greater than the distance from the mouth to the posterior
edge. Claus shows that neither of these features serves to discriminate between
Alima and Erichthus in every case, and he figures and describes a larval type which is
intermediate between the two, having the elongated flattened carapace and the exposed
eyes, but the mouth well forward, and the thoracic region well covered by the carapace.
For this intermediate larval type he proposes the name Alimerichthus. Claus was not
able to connect any one of his Alimzx with a specific’ adult, but he shows that they
resemble the adults of the Squilla type very closely, and he correctly decides that
they are the larve of this type, although he erroneously believes that they belong to
the Lysiosquilla branch, rather than to the true Squille! He says there can be no
doubt that we must seek their adult representatives in the Squilla-group, and that the
Alima larva, as distinguished from Erichthus, belongs exclusively to the genus
Lysiosquilla, which is characterised, like the Alima larva, by the elongation and loose
articulation of the abdomen. The lower members of the genus Squilla are loosely
articulated, like the Lysiosquille, and the hind body is about as long in the one genus
as it is in the other, and there is therefore no reason for believing that any of these
larvee are young Lysiosquille, although later researches have shown that he is correct in
his surmise that they pertain to the Squilla-group.

In a paper which was published in 18792 I described a series of Alima larvee, which
were procured in-abundance in the Chesapeake Bay, a locality where Squilla empusa is
common, while no other Stomatopod is known to occur there, and I therefore advanced
the opinion that this larva, a young stage of which is shown in figs. 4 and 5 of
Pl. I., is a young Squilla. Three years before, Faxon reared from a similar but slightly
more advanced larva, a young Squilla, which had the characteristics of the adult Squilla
empusa, and although his results were not published until 1882° the proof that Alima
is a young Squilla is due to him. It is of course possible that some species
~ of Lysiosquilla may also pass through an Alima stage, but I shall show that, among

1 Metamorphose der Squilliden, p. 154. ? On the larval stages of Squilla empusa.
* Selections from Embryological Monographs compiled by Alexander Agassiz, Walter Faxon, and E. L. Mark,
I. Crustacea, Cambridge, 1882, Bull. Mus. Comp. Zool., vol. ix. No. L., pl. viii. figs. 2, 3.

84 THE VOYAGE OF H.M.S. CHALLENGER.

the lower Lysiosquille, in the subgenus Coronis, the larva is not an Alima
but a Squillerichthus, and if it be true that Squilla and Lysiosquilla represent two
divergent stems, and that their lower representatives are most closely related, it is
not at all probable that any species of Lysvosquilla passes through an Alima stage, for
if it were the case we should be forced to believe that the higher Lysiosquille have inde-
pendently acquired the same secondary larval form as the higher Squillz.

While Claus has given us a very complete history of the Hrichthus larva his
collections did not furnish a connected series of Alima larvee, and although he points out
the possibility that the very young larva which had been figured by Fritz Miiller,' as well
as avery similar one from Messina which he himself figures,* are young Alime, he was
unable to obtain any of the intermediate stages, and my paper on the larval stages of
Squilla enypusa is the only one in which a tolerably complete series of Alima larvee
are figured. In this paper I showed that the distinctive characteristics of the larva are
present at a very early stage of development, and that it is in all essential respects an
Alima at a time when the last three thoracic somites are not yet marked out, and when
there are no appendages between the large raptorial limbs of the second thoracic somite
and the first abdominal appendages. I also pointed out the great probability that this
larva leaves the ege as an Alima rather than as an Hrichthoidina or an Erichthus; a
probability which is strengthened by the fact that Fritz Miiller has figured an ege con-
taining a larva which is probably in this stage.

Squilla (Alima) gracilis.—The Challenger collection contains a number of larve
which were collected in the tow-net at St. Vincent, and from these I have been able to
select a series of Alimx, which give a much more complete history of the growth
and gradual modification of the larva than that which I obtained in 1879.

This series of larvee, Alima gracilis of Milne-Edwards (Alima angustata, Dana) is
shown in Pl. IV. figs. 4-6, Pl. V. fig. 3, Pl. VI. figs. 3-5, and Pl. VIII. figs. 4-6.

Its distinctive or specific characteristics are as follows:—The body is narrow and
greatly elongated, the exposed hind body making about half the total length as
measured from the tip of the long slender rostrum. The raptorial claw of the second
thoracic appendage (Pl. VIII. fig. 5) is narrow and greatly elongated, and the dactylus is
only about half as long as the second joint.

The telson is remarkably long and narrow, and in the older larve its length is three
times its width. It has six large marginal spines (Pl. VI. fig. 3 and Pl. VIII. fig. 6) with
minute spinules between the submedians, and also between the submedian and the inter-
mediate. The lateral edge of the greatly elongated narrow flat carapace is armed with
twelve or thirteen small spines and a larger spine projects from the side of the postero-
lateral spine near its base. There is a small median dorsal spine on the posterior ede of
the carapace, which exposes the last three thoracic somites, and is narrowed posteriorly

1 Archiv f. Naturgesch., Jabrg. xxviii. Taf. xiii. fig. 1. 2 Metamorphose der Squilliden, fig. 22B.

REPORT ON THE STOMATOPODA. 85

in the older larve, The inner spine of the prolongation from the basal joint of the
uropod is much longer than the outer one, and it has a rounded lobe on its outer edge.
The exopodite is armed on its outer edge with five or six marginal spines. The eyes
are pear-shaped and the stalks very long. The mouth in old larvee is under the posterior
fourth of the carapace ; the postero-lateral angles of all the abdominal somites are pro-
longed into acute spines, and the dactylus of the raptorial claw shows traces of five
marginal spines.

Alima gracilis is one of the best known and widely distributed Stomatopod larvee,
and any naturalist who has the good fortune to be becalmed in tropical waters should
be able to determine the adult to which it belongs without difficulty, as the older
Stomatopod larvee thrive and moult in small aquaria. The very close resemblance
between it and the Alima from which Faxon reared a young Squilla empusa shows
beyond doubt that the adult is one of the highly specialized carinate Squille, and
its wide distribution indicates that the adult also is very widely distributed. The
most striking difference between it and other Alime is the great elongation which
takes place in the telson during the latest stages. The late appearance of this
character indicates that it is shared by the adult, and as there is no known
species with a long narrow telson, and as it is hardly possible that an animal which
must be one of the largest and most widely distributed of the Stomatopods, should
have escaped discovery if it were littoral in its habits, it is probable that Alima
gracilis is the larva of an unknown, deep-water Squilla, with an elongated telson and a
long raptorial claw.

The smallest larva in the series from St. Vincent (No. 1) measures 5,26; mm. from
the tip of the rostrum to the middle line of the telson; the second (No. 2, Pl. IV. fig. 4)
measures 65%'5 mm., the third (No. 3, Pl. V. fig. 3) 9,48, mm., the fourth (No. 4, Pl. TV.
fig. 5) 11355 mm., the fifth (No. 5, Pl. IV. fig. 6) 17,38; mm., and the sixth (No. 6,
Pl. VI. fig. 3) 42;8%5, or a little less than Claus’ larva, which is a little less than 52
mm. long. This large larva is well known and widely distributed, and the Challenger
collection contains numerous specimens from St. Vincent, the west coast of Africa, the
Central Pacific, and the vicinity of Cape York. :

The youngest larva (No. 1) of the table was not figured as there is no difference,
except in size, between it and No. 2, which is shown in Pl. IV. fig. 4. In
this larva all the somites of the hind body, except the fifth and sixth abdominal, are
distinct, and the outline of the fifth is indicated. The appendages of the sixth
abdominal are entirely absent, those of the fifth are rudimentary bilobed pouches, while
the first four are well developed and functional, with a very long basal joint, and an
appendix interna on the endopodite.

There are no traces of appendages on the last six thoracic somites, and the third,

Metamorphose der Squilliden, pl. viii. fig. 35.

86 THE VOYAGE OF H.M.S. CHALLENGER.

Alima gracilis.
Measurements.

In millimetres.

In thousandths of total length.

Measurements along middle line:-—

Total length from tip of rostrum to tip of
telson, . 5 : 5 F >

Total length of carapace including rostrum,

Total length of rostrum, . é é :

From tip of rostrum to anterior edge of
ocular somite, . o 5 -

From anterior edge of ocular somite to line
joining bases of antero-lateral spines,

From line joining bases of antero-lateral
spines to tip of labrum, : 5 2

From tip of labrum to anterior edge of
third thoracic somite, ; :

Third thoracie somite,

Fourth thoracic somite,

Fifth thoracic somite,

Sixth thoracic somite,

Seventh thoracic somite, .

Eighth thoracic somite, . . F -

From anterior edge of fourth thoracic to
anterior edge of first abdominal somite, .

First abdominal somite, : 5

Second abdominal somite,

Third abdominal somite, .

Fourth abdominal somite,

Fifth abdominal somite, .

Sixth abdominal somite, .

Telson on middle line,

Total length of abdomen and telson, .
Total length of body on middle

line,

Measurements of carapace:—
Total length on middle line including
rostrum, > 5 . : z :
Length of rostrum, . 0 E
Length of antero-lateral spines,
Length of postero-lateral spines, 5
Width at bases of antero-lateral spines,
Width at bases of postero-lateral spines,
Width of seventh thoracic somite,
Width of eighth thoracic somite, .
Width of second abdominal somite,
Width of fifth abdominal somite,
Greatest width of telson, . ; 5 ;
Distance between submedian spines of telson,
First pleopod, basal portion, .
First pleopod, exopodite,
Second maxillipod, basal joint,
Second maxillipod, second joint, .
Second maxillipod, third joint,
Second maxillipod, fourth joint, .
Second maxillipod, sixth joint,
Second maxillipod, dactyle,
First antenna, first joint,
First antenna, second joint, .
First antenna, third joint,
First antenna, flagellum,
Eye stalk, . : 5
Eye, F 5

bo Or
© od

rm OO OD

i

.| No. 2.| No. 3.

5°82| 8:34 | 20°17
1°84) 1°92] 3°32
72 84) 1:16
1°84] 1°76) 2°49
133] 1°76} 3°65
2°88} 3°68] 6°64
‘51 48 | 1°00
“BG D1} 1°16
“76 64) 1°66
75 64] 1°10
1°60) 1°92] 3°57
77 54) 1°00
69 ‘74! 1:00
Ol 70) 1°49
“56 “38 | 1°00
1:28} 1°92) 4°98
1°60] 2°08) 5°23
“32 38 “66
2°32 | 3°20) 5°40
1:20] 1:92) 3°65

“43 *b4| 1°41

541
174

No. 4.| No. 5.| No. 6.] No. 1.| No. 2.) No. 3.| No. 4.

No. 5.

492
110

110

No. 6.

fourth, and fifth are short, and wider than long, while the sixth, seventh, and eighth are

elongated, and about as long as they are wide.

The second antenna has a two-jointed

REPORT ON THE STOMATOPODA. 87

shaft, and a broad oval exopodite or scale, but no flagellum, and the three-jointed shaft
of the first antenna carries two short unjointed terminal pouches. The eye stalks are
very long, slender, and transverse, and the eyes broad at the rounded conical end. The
long slender rostrum is about half (38) as long as the carapace, and the distance from its
base to its tip is about equal to the distance from its base to the tip of the labrum. The
carapace widens posteriorly, and its width between the bases of the antero-lateral spines
is more than half (3§) of its width between the bases of the postero-lateral spines. Its
lateral edges are straight, and would if prolonged meet at the tip of the rostrum. The
carapace and rostrum make up considerably more than half 554, of the total length, and
the posterior edge of the carapace, which is nearly transverse, lies on the posterior edge
of the sixth thoracic somite, while the tips of the divergent postero-lateral spines are in
the plane of the anterior end of the second abdominal somite. The telson is oval and its
length is a little (4%) greater than its width. The lateral marginal spines are about
midway between the anterior and posterior ends of the telson; the space between them
and the intermediate is a little shorter than the space between the intermediate and
the submedian. There are seven small teeth between the intermediate and the submedian,
and the distance between the submedians is +88, of the total length. There are only
three small spines on the outer edge of the carapace at this stage, and they are all behind
the middle.

Larva No. 3, 974% mm. long, is shown in Pl. V. fig. 3. The fifth abdominal somite
is now distinct, but much shorter than those in front of it, and its appendages are
perfectly formed but small and without the appendix interna. The thoracic somites and
appendages are like those of No. 2. The antenne are like those of No 2, except that the
two flagella of the first antenna are divided each into three joints. The rounded conical
end of the eye is broader than in the previous stage, and there have been important
changes in the relative length of the carapace and hind body. The rostrum is less than
half (34) as long as the carapace, and the distance from its base to its tip is only 7§ of
the distance from its base to the tip of the labrum. The carapace still has the same
general shape, and its lateral edges are straight, with three spines on the posterior half,
but its width between the bases of the postero-lateral spines is relatively less, and the
width between the bases of antero-laterals is to that between the postero-laterals as 15 to
31, or about 1 to 2. In most other respects this larva is very similar to No. 2.

Larva No. 4 is shown in Pl. IV. fig. 5. It may possibly belong to a different
species, as it has only one spine on the lateral edge of the carapace, but if distinct it
must belong to some very closely related species, as there are no essential differences in
the measurements. The third, fourth, and fifth thoracic somites are shortened and
crowded together, and their appendages have appeared as bud-like outgrowths, while the
appendages of the sixth, seventh, and eighth thoracic somites are also represented by
similar but much smaller buds. The sixth abdominal somite has not yet been separated

88 THE VOYAGE OF H.M.S. CHALLENGER.

from the telson, although its appendage is represented by a minute bilobed bud. This
larva therefore has all the appendages of the adult, either functional or represented by
buds, and all the somites of the hind body except the sixth abdominal. The outer ramus
of the flagellum of the first antenna is now bilobed, and the flagellum of the second
antenna is represented by a bud, aud the ocular somite is distinct and movable. The
rostrum is a little less than half as long as the carapace, and the labrum is still farther
back. The width of the carapace between the bases of the antero-lateral spines is $4 of
its width between the bases of the postero-lateral spines, and the carapace with the
rostrum makes up almost exactly half of the total length (3,05). In other respects this
larva closely resembles No. 3.

Larva No. 5, 1733%5 mm. long, is shown in Pl. IV. fig. 6. The sixth abdominal
somite is still absent, and its appendages and those of the three last thoracic somites
rudimentary, although those of the third, fourth, and fifth thoracic somites have assumed
nearly their final form, and the first antenna has its three-jointed flagella.

Although actually longer than that of No. 4, the rostrum is now relatively much
shorter, and only about one-fourth (44) as long as the carapace, and the distance from
its base to its tip is less than one-half (44) of the distance from its base to the tip of the
labrum.

The carapace is still shorter, as compared with the hind body, and with the rostrum, it
now makes up less than half (,4%2;) of the total length, although its lateral edges are still
straight, and its triangular shape is still retained. The greatest change is in the length
of the telson, which is now nearly twice as long as wide.

Larva No. 6, 42,8; mm. long, is shown in PI. VI. fig. 3. It has all the somites
and appendages of the adult, although the sixth pair of abdomimal appendages are
rudimentary. The carapace is still more elongated, and although the rostrum is actually
longer than it was in stage 5, it is much shorter both as compared with the total length
of the body of which it now makes +85, and also as compared with the carapace, which
is more than five times (494) as long as the rostrum. One of the most prominent
characteristics of the fully grown Alima gracilis is the great distance of the mouth from
the anterior end of the body, and in larva No. 6 the length of the rostrum is little more
than one-fourth (,8) the distance from its base to the tip of the labrum. The carapace
including the rostrum makes a slightly smaller portion of the total length (4/5) than at
stage 5, and its lateral edges are no longer straight but are incurved near their posterior
ends, so that there is no increase in width in the posterior third of the carapace.
In the still older larva figured by Claus, this peculiarity is still more marked, and the
broadest part of the carapace is some distance in front of its posterior margin; this is
more emarginated in stage 6 than it is in younger larvee, and it crosses the middle of
the sixth thoracic somite. The telson is still more narrow and elongated, and the sub-
median spines (Pl. VI. fig. 3), which have become more and more closely approxi-

REPORT ON THE STOMATOPODA. 89

mated to each other at each successive moult, are now separated by a space equal to only
r24y or less than jy of the total length, while in stage 1 the distance between them is
1335 of the total length. As shown in Pl. VIII. fig. 6, there are fourteen short acute
spines between the submedians, with still smaller spines between them, and there are
fifteen small acute dentations between each submedian and the intermediate of the: same
side, which is about equal in length to the submedian.

The uropod of a specimen of the same size, which was captured at the surface by the
Challenger expedition, between Api and Cape York, is shown in ventral view in Pl. VIII.
fiz. 4. The exopodite is now divided into a paddle and a second joint, and the latter
has five spines on its outer margin. The ventral prolongation from the basal joint ends in
a short outer spine and a much longer inner one, which has a rounded lobe on its outer
margin, near the base.

From the table of measurements given above the following measurements may
be selected as showing the character of the changes through which the larva passes
during its growth. They are all in thousandths of the total length from the tip of the

rostrum.

No. 1.| No. 2.| No. 3.| No. 4.) No. 5.) No. 6.
Rostrum, . F : : : 187 | 182 | 174 | 160 | 111 80
Carapaee, exclusive of rostrum, : : 357 | 350 | 367 | 345 | 381 | 404
| Width of carapace between bases of anterolateral spines, : 194 | 182 | 148 | 115 | 101 88
| Width of carapace between bases of postero-lateral spines, . 301 | 289 | 314 | 249 | 212 | 158
| Length of abdomen including telson, ; ; : 382 | 395 | 403 | 408 | 435 | 444
| Distance between submedian spines of telson, : ‘ 76 85 56 67 31 24

While Alima gracilis differs from ordinary Alimz in many respects, especially the
great elongation of the body, the shortness of the carapace, and the elongation of the
telson, Claus has figured a series of Alimew which shows that the shorter and wider
species are connected with the elongated ones by so many intermediate forms that there
can be no doubt that the adults which they represent are all closely related. The
Challenger collection also contains great numbers of these intermediate forms, and I
give in Pl. VII. fig. 7 and Pl. VIII. fig. 7 figures of the telson and the raptorial claw
of one of them, which resembles Alima gracilis in the shape of its carpus, while its
telson and the general outline of the body are so much like Faxon’s larva as to indicate
that it is the young of a species of Squilla very closely related to Squilla nepa. A
number of specimens of this larva were collected by the Challenger on April 13, 1876,
near Sierra Leone. The carapace makes about half the total length of the body, and
it exposes the posterior end of the sixth thoracie somite, while the tips of its postero-
lateral spines, each of which has a small secondary spine about midway between the
base and the tip, extend backwards to the plane of the suture between the first and

(ZOOL, CHALL, EXP,—PART XLV.—1886.) Yy 12

90 THE VOYAGE OF H.M.S. CHALLENGER.

the second abdominal somites. The lateral edge of the carapace has a few, five or six,
marginal spines, and a well marked dorsal spine. The postero-lateral angles of the abdo-
minal somites end in acute spines, and the inner and longer of the two basal spines of
the uropod has a rounded lobe on its outer edge close to its base. The carpal joint of
the raptorial claw is as much elongated as it is in Alima gracilis, although the telson is no
longer than its width. As the adult form is certainly a Squilla, closely related to Squilla
nepa and Squilla mantis, it does not seem necessary to give the larva a provisional name.

Squilla (Alima), empusa.—tn order to render the series of Alima larvee more
complete, I give two figures, a ventral and a side view (PI. I. figs. 4, 5), of a larva
which is smaller and probably one moult younger than any of those in the Challenger
collection.

The drawings were made from the living larva, which is probably the larva of Squilla
empusa, and the same as the one which is shown in Faxon’s pl. vi. fig. 17. It is 3°216
mm. long and the last six thoracic somites increase uniformly in length from in front
backwards, and have no traces of appendages ; the sixth abdominal somite is absent, and
the fifth very short and without appendages. The telson is longer than wide, narrowed
anteriorly and posteriorly, and the four spines between the submedian and the inter-
mediate are nearly as large as the latter. The rostrum is short, and not quite as long as
the slightly divergent postero-lateral spines of the carapace, which is oval in outline and
nearly as wide in front as behind. It crosses the posterior end of the seventh thoracic
somite in the middle line, and the tips of the postero-lateral spines are about opposite the
middle of the third abdominal somite. The tip of the labrum is close to the anterior
end of the carapace, and the eye-stalks are short, less than half as long as the eyes.

Alima bidens.—Claus has figured and described,’ under the provisional name Alima
bidens, an Alima larva which is of especial interest, as many of its organs, especially the
gills and the dactyles of the raptorial claws, undergo much more complete development
during the larval life than is usually the case, and it therefore presents more data than
the ordinary larvee for establishing an identity with some one of the genera of adult
Stomatopods.

Although Claus regarded it as the larva of some species of Lysvosquilla, I shall show
that Alima bidens must be referred to the genus Squilla, and there is, so far as I
am aware, no cther Stomatopod larva which exhibit clearer evidences of relationship to a
definite adult genus.

It is of course desirable that some one who has the opportunity should actually rear it,
and determine in this way the specific adult of which it is the larva, but in the absence
of this decisive proof the evidence that it is a Squilla could hardly be stronger than it
now is.

Claus obtained only a single larva 26 mm. long from the Indian Ocean, and as this

1 Metamorphose der Squilliden.
P q

REPORT ON THE STOMATOPODA. 91

Squilla (Alima), empusa. In inches and | In thousandths
Measurements. decimals. of total length.
ee et
Measurements on middle line :—
Rostrum, . . : : : 496 153
From base of rostrum to tip of labrum, : 3 320 101
From tip of labrum to anterior edge of third thoracic somite, P 512 158
Third thoracic somite, : : : ‘ i ? 064 20
Fourth thoracic somite, ‘ ; ; ; ; ; 064 20
Fifth thoracic somite, . 2 ‘ : ; : 064 20
Sixth thoracic somite, ; : 2 F : : 080 25
Seventh thoracic somite, . . , : ; : 096 30
Eighth thoracic somite, 3 ; : 2 : : 112 35
First abdominal somite, . ; : : : F 160 50
Second abdominal somite, . ; : : . : 160 50
Third abdominal somite, . ; : : : 3 160 50
Fourth abdominal somite, . t : : é ; 160 50
Fifth abdominal somite, . : : : : ‘ 048 15
Telson, : ; : : : : : : “720 223
Total length on middle line, . : : 3 : 3-216 1000
| Length of carapace on middle line, ‘ i 1648 411
Width of carapace between bases of antero-lateral spines, ; ; “704 218
Width of carapace between bases of postero-lateral spines, : ; ‘736 228
Width of telson between submedian spines, . : : s 320 99
| Width of telson (greatest), . ? , : : : 448 139
| Length of abdomen including telson, . ? 3 : ; 1-408 438

specimen, which is shown in Claus’s figure 34, has on the dactylus of the raptorial claw two
marginal spines besides the terminal one, he derives the provisional specific name from
this characteristic.

The Challenger collections contain two specimens of the same larval type; one from
the Cape of Good Hope (Pl. IX. figs. 1 and 2), a little older than Claus’s larva, and with
three spines on the dactylus besides the terminal one, and 25°39 mm. long; and another
from the Gulf of Penas, younger than Claus’s larva, with only one spine, and 17°57 mm.
long. The two specimens, which undoubtedly belong to a single species, differ very slightly
from Claus’s larva, which may possibly be the young of a distinct species, although the
differences are so very slight that it seems best to retain Claus’s specific name Alima
bidens for them all.

Alima bidens is characterised as follows :—The very short and narrow carapace, which
has no median dorsal spine, makes with the dorsal rostrum only 2 of the total
length, and its deeply emarginated posterior edge exposes the four posterior thoracic
somites. The long tip of the slender.rostrum is in front of the end of the shaft of the
first antenna, and the length of the rostrum is a little more than half the length of the
carapace, measured on the middle line from its base. The antero-lateral spines of the
carapace are unusually long and divergent; the tips of the long postero-laterals are opposite

92 THE VOYAGE OF H.M.S. CHALLENGER.

the middle of the first abdominal somite, and they have two or three small secondary
spines on their inner edge, while the lateral margin of the carapace, immediately anterior to
the base of the postero-lateral spine, has two well-marked lateral spines on each side.

The hind body is greatly elongated, making about +¥ or a little less than 3 of the
total length from the tip of the rostrum, In older larvee the abdomen is depressed, and its
width is equal to } of the total length. The inner spine of the basal prolongation of the
uropod is much longer than the outer, and it has a small tooth on its outer margin, about
half way between the base and the tip. The telson is shghtly elongated, with six
marginal spines, and with secondary dentations between the submedians, and also between
the submedian and the intermediate. The eye-stalks are nearly as long as the eyes, and
the mouth is a little behind the middle point of the carapace, as measured from the tip of
the rostrum.

In Claus’s larva there are three secondary spines on the inner edge of the postero-
lateral spine of the carapace, while there are only two such spines in each of the
Challenger specimens.

In the smallest of the Challenger specimens the dactylus of the raptorial claw has a
single marginal spine besides the terminal one, the uropods are very small, and the gills
are represented by simple pouches on the exopodites of the first five pairs of abdominal
appendages. The carapace, with the rostrum, makes 344%;, the hind body 3388,, and
the width of the abdomen -2,%% of the total length.

In the older specimen (PI. IX. fig. 1) the dactylus is armed with three well-developed
marginal spines besides the terminal one, the uropods are nearly half as long as the
telson, with the exopodite divided into a broad rounded paddle and a second joint which
is fringed with six marginal spines, and the prolongation from the posterior edge of the

ventral surface of the basal joint is as long as the exopodite, and ends in a short acute
curved outer spine and a much longer inner spine, which has a tooth on its outer margin
about half way between the base and the tip.

The carapace has a well-marked median carina; the postero-lateral angles of the broad
flat abdominal somites all end in acute spines, and there are two submedian carine,
ending posteriorly in spines, on the dorsal surface of the sixth abdominal. The telson
has a median longitudinal carina, and is slightly longer than wide. Its lateral edges,
from the anterior edge to the tips of the intermediate marginal spines, are nearly parallel;
between the intermediates and the submedians they are inclined inwards and backwards
at an angle of 45° with the long axis of the body, while between the submedians they are
inclined inwards and forwards. The lateral marginal spine (fig. 2) is nearer to the
intermediaté than to the anterior edge, and it has a single minute spine internal to its
base. The intermediate is longer than the lateral or the submedian, and there are
twelve or thirteen small spines between it and the submedian, and there are about twenty
very minute spines between the submedian and the middle line.

REPORT ON THE STOMATOPODA. 93

In this larva, which is 1 inch long, the carapace makes, with the rostrum, 74%, the
exposed hind body ,o%%5, and the telson ;2%% of the total length.

Even if the close resemblance between this larva and the one from which Faxon
reared a young Squilla empusa did not indicate that Alima bidens is also a Squilla,
this could be inferred with great certainty from the examination of the larva itself, since
it resembles the adults of this genus in the presence and in the small number of spines
on the inner edge of the dactylus of the raptorial claw, in the depression and width of the
hind body, in the presence of a median-dorsal carina on the carapace and on the telson,
in the presence of acute spines on the postero-lateral angles of the abdominal somites, in
the presence and in the small number of marginal spines on the outer edge of the proximal
joint of the exopodite of the uropod, in the relative length of the two spines of its ventral
process, in the presence of a secondary tooth on the outer edge of the inner and longer
spine, in the relative positions of the marginal spines of the telson, in the presence of a
single minute dentation inside the base of the lateral, and a number of dentations (more
than three) between the intermediate and the submedian. While it is true that there
are some adult Squwille which do not shew all of these characteristics, and while no one
of them is in itself perfectly diagnostic, it is also true that there are no Stomatopods in
which they are all united except members of the genus Squilla, and we may therefore
decide, with all the certainty which is possible in absence of direct proof, that Alima
bidens is the larva of one or perhaps of two species of the higher carinate Squillz.
The presence of three secondary spines on the inner edge of the postero-lateral
spine of the carapace of Claus’s larva, and of only two in our specimen, possibly corre-
sponds to a specific difference between the adults.

Alima macrophthalma.—The Challenger collection contains a number of specimens of
an Alima larva of a type which is quite different from that of which Alima gracilis is
an example, and I have selected from a surface gathering, made near Cape Howe, the
series which is shown in PI. VII. figs. 1-6; Pl. VIII. figs. 1—3.

It is possible that these are not all of one species, but the differences between them
are so slight that, if not the same, they must at least belong to adults which are
very closely related, and as I am not able to identify the larva with any of the published
descriptions, [ propose for it the provisional name Alima macrophthalma, on account of
the great size of the eyes as compared with the very small eye stalks.

The youngest larva which I have found, No. 1, is essentially like No. 2. The telson
is shown in Pl. VII. fig. 2; No. 2, which is shown in Pl. VIL fig. 1, is 4;% mm. long ;
No. 3, shown in fig. 4, is 6;4 mm. long; No. 4, shown in fig. 5, is 8; mm. long;
No. 5, shown in PI. VIII. fig. 1, is 8,43, mm. long; No. 6, shown in fig. 2, is 117445 mm.
long, and No. 7, shown in Pl. VIII. fig. 3, is 19 mm. long.

The most prominent diagnostic characteristics of Alima macrophthalma are as
follows. The eye-stalks are very short, and the eyes large with very broad tips; the

94 THE VOYAGE OF H.M.S. CHALLENGER.

labrum is near the middle of the carapace, which is narrow and elongated, exposing the
fifth, sixth, seventh and eighth thoracic somites, and very much shorter than the elongated
hind body. The carapace has a median dorsal spine, about as long as the antero-lateral
spine, on its posterior edge, and there is a single acute straight spine on the lateral margin
of the carapace, on each side, immediately anterior to the postero-lateral, the tip of which
reaches nearly to the line of the anterior edge of the first abdominal somite.

In the younger larvee the posterior edge of the carapace is transverse, and the telson
is narrowed anteriorly, while in older larvee the carapace is deeply emarginated, and the
lateral edges of the telson nearly parallel. In the oldest larva the inner spine of the
uropod is longer than the outer, with a rounded lobe on its outer edge, the abdomen is
depressed, about as wide as the carapace between the bases of the antero-lateral spines, and
the postero-lateral angles of all the abdominal somites end in acute spines.

This series, together with the gracilis series, gives a very complete picture of the
changes which the Alima larva undergoes during its larval life. In the youngest stage
which was observed (Pl. VII. figs. 1 and 2) the shaft of the first antennz has only two
joints, and the appendage ends in only two rami. There is no flagellum on the second
antenna, and no appendages between the raptorial second thoracic limbs and the first
pair of abdominal appendages, All the four thoracic somites are wider than long, and
the third, fourth, and fifth are equal in length, and more than half as wide as those which
follow. The abdomen is narrow, very slightly wider than the thorax, with five distinct
somites, the last of which is short without appendages, while the appendages of the fourth are
rudimentary. Those of the third, second, and first are functional, long and slender, with
a long appendix interna. The tip of the labrum, which has no spine, is slightly anterior
to the middle line, and the greatest width of the carapace is slightly less than its length.
It narrows a little posteriorly, and its width between the bases of the postero-lateral
spines is very slightly greater than its width between the bases of the antero-laterals.

The width of the telson at its anterior end is about half the greatest width (see figs.
1 and 2) and it is also slightly narrowed posteriorly. The posterior border, between the
submedian spines (fig. 2) is straight and transverse, with about eighteen small spines.

The intermediate spines are near the middle of the telson, separated by a sligh ;
interval from the laterals, and by a much longer interval from the postero-median,
with some intervening dentations which are about as large as the primary spines.

At the next stage (No. 3, fig. 4) the appendages are as before, except that the
fourth abdominal appendages are larger, and the fifth pair are now present as rudiments.
The position of the labrum and the shape of the carapace is essentially as in stage 2,
and the telson is also nearly the same except that its posterior border is slightly
emarginated. At the next stage (No. 4, fig. 5) the flagellum of the second antenna,
the appendages of the third thoracic somite, and those of the sixth abdominal somite, are
all represented by minute buds, and the third thoracic somite has become shorter than

REPORT ON THE STOMATOPODA. 95

the fourth and fifth. The outline of the carapace is about as before, except that it is
now deeply emarginated on the middle line. The posterior border of the telson is slightly
emarginated ; the submedian spines have approached each other, and the intermediates,
which are now much larger than the seven secondary dentations, have moved backwards,
as have also the laterals.

In the next stage (No. 5, Pl. VIII. fig. 1) more important changes have taken place.
The third, fourth, and fifth thoracic appendages are represented by buds, and their
somites have become reduced in length, so that the sixth is as long as all three of them.
The abdomen is now much wider than the thorax, and its somites all end posteriorly
in acute spines. The labrum is a little further back, and the anterior end of the carapace
narrower than at stage 4, but the telson is essentially like that of stage 4, although it
is a little more emarginated in the middle line. In stage 6 (fig. 2) the labrum is still
further back, the flagellum of the second antenna and the appendages of the third,
fourth, and fifth thoracic somites are elongated, and the latter are obscurely divided into
joints ; the appendages of the sixth, seventh, and eighth thoracic somites are represented
by buds, the abdomen is wider and more depressed, and the appendages of the sixth
somite are parallel but well developed. The space between the lateral and the
intermediate spines of the telson is now equal to the space between the intermediate and
the submedian, and the posterior border is deeply notched on the middle line. In the
oldest larva in this series (No. 7, fig. 3) the appendages all have essentially their adult
forms, and the more important changes are the lengthening of the hind body, the
flattening and widening of the abdomen, and especially the widening of the sixth
abdominal somite and the anterior end of the telson. This is now nearly rectangular, and
deeply notched on the middle line ; the submedian spines are more approximated, and the
intermediates further back.

Although the collection contains no specimens which serve to connect this larva with a
specific adult, its close resemblance to the more typical Alimz, especially Alima bidens,
gives every reason for believing that it is the young of one of the higher multicarinate
species of the genus Squilla.

The specimens are all from the coast of Australia, and all these which were drawn
were from Cape Howe and its vicinity.

Alimerichthus.—Inasmuch as the Erichthus type of Stomatopod larvee presents a very
much greater diversity of forms than the Alima type, and as it is preceded by a Erich-
thoidina stage which is absent in the Alima larve, there can be no doubt that the latter type
is a secondary modification of the Hrichthus type, and that the greatly elongated Alimex
like Alima gracilis (Pl. VI. fig. 3) are more divergent from the primitive larva than the
shorter and broader forms like Alima macrophthalma (P1. VIII. fig. 3). Although Milne-
Edwards and Dana have attempted to show that the two types are sharply separated,
Claus has pointed out that among the Alimz themselves there is a series of larve,

96 THE VOYAGE OF H.M.S. CHALLENGER.

which, starting with Alima gracilis, with its short carapace and elongated hind body,
leads, through many intermediate types, to a larva which is short and wide, and
almost completely covered by the elongated carapace, like an Hrichthus, although its
general structure is Alima-like. For this larva, which he justly regards as the ancestral
form from which the Alima larvee have been produced, Claus proposes the provisional
generic name Alimerichthus, as expressing its double relationship, to Alima on the one
hand, and Erichthus on the other.

He figures’ a single advanced larva of this type 18 mm. long from the Indian Ocean,
but as he gives no account of its early stages, the occurrence in the Challenger collection
of younger specimens of this type is a matter of great interest, as these younger larvee
show that the young Alimerichthus, like the young Alima, passes through a stage in
which the last six thoracic somites have no appendages, while the raptorial limbs of the
second thoracic somite and the first pairs of abdominal appendages are well developed
and essentially like those of the adult.

I have selected from a collection made, February 29, 1886, in the South Atlantic, off
the coast of South America, in lat. 36° 9’ 8” S., long. 48° 22’ W., the two specimens
which are shown in Pl. VIIL fig. 8, and Pl. IX. fig. 3, and which represent a form
which is very closely related to, but probably not identical with the one figured by
Claus. The youngest, shown in Pl. VIII. fig. 8, is 8°715 mm. long, and the next stage,
shown in Pl. IX. fig. 3, is 15°52 mm. long, while Claus’s Alimerichthus is 18 mm.
long, so that we probably have, in this series, three successive moults in the history
of the larva for which Claus’s generic name may be retained without a specific
name. Alimerichthus is characterised as follows—A short wide Alma with a short
hind body which is wide and flat in the older larve. The mouth is near the middle of
the carapace, and the rostrum is less than half as long as the carapace, which has a
median dorsal spine, and moderately long antero- and postero-laterals. There is a second-
ary spine on the inner edge of the postero-lateral close to its base, and a very prominent
acute spine projecting outwards from the lateral edge of the carapace, about midway
between the bases of the antero- and postero-lateral spines, and two or three smaller ones
projecting inwards between this and the one at the base of the postero-lateral. The
length of the carapace, measured on the middle line, from the tip of the rostrum, makes
much more than half the total length, and it covers all of the thorax except the tip of
the eighth somite, while the tips of the postero-lateral spines are in the line of the
anterior edge of the telson. The eye stalks are about as long as the eyes, which have
swollen globular tips, the width of the carapace equals about one-third the total length,
the telson is wider than long, with six marginal spines, and numerous secondary spines
between the submedians and also between each submedian and the adjacent intermediate.
The inner spine of the basal prolongation of the uropod is slightly longer than the outer,

1 Metamorphose der Squilliden, p. 147, Taf. viii. fig. 30.

REPORT ON THE STOMATOPODA. 97

and there are few (five or six) marginal spines on the outer edge of its exopodite. In
the older larvee the dactylus of the raptorial limb shows traces of six spines besides the
marginal one,

In the youngest Alimerichthus which has been observed (Pl. VIII. fig. 8) the
flagellum of the second antenna is represented by a bud, as are also the appendages of
the third thoracic somite, and the third, fourth, and fifth thoracic somites are crowded
together so that the sum of their lengths is about equal to the length of each one of the
three last thoracic somites, upon which there are no traces of appendages. There are five
pairs of fully developed and functional abdominal appendages, and five distinct abdominal
somites, the first four wider than the thorax, but the fifth very narrow and deeply con-
stricted off from the telson, upon the anterior edge of which the sixth pair of abdominal
appendages are represented by buds. The telson is wider than long, and nearly four
times as wide as the abdomen. It has six pairs of marginal spines with numerous
minute secondary spines between the submedians, ten or eleven on each side between the
submedian and adjacent intermediate, and a single one internal to the base of the lateral.
The margin of the telson, between the submedian and the intermediate, makes an angle
of about 45° with the principal axis of the body. A comparison of this larva with the
corresponding stage of Alima gracilis (Pl. IV. fig. 5), or of Alima macrophthalma
(Pl. VIII. fig. 1), shows that there is every reason for believing that it is preceded
by an earlier stage like the youngest observed stage of Alima gracilis (Pl. IV. fig. 4),
of Alima macrophthalma (Pl. VII. fig. 2), or Alima (Squilla) empusa (Pl. I. fig. 4),
and that at this time the third, fourth, and fifth thoracic somites are long and without
appendages like the sixth, seventh, and eighth; the fifth abdominal somite and its
appendages absent or rudimentary, the sixth absent, and the telson spatulate, with the
submedians wide apart, and the secondary dentations between the submedians and inter-
mediates about as large as the primary spines. As we know that some, and probably
all, of the Alima larve hatch from the egg in this condition, and do not pass through a
free Erichthoidina stage, this is undoubtedly true of the Alimerichthus also.

Tn the next stage which has been observed (Pl. IX. fig. 3) the appendages and
somites are all present, the exopodites of the first five abdominal appendages carry at
their bases the rudimentary buds which are to become the gills, the telson is shorter
and wider than before, and the intermediate marginal spines have travelled backwards
until the posterior margin of the telson is nearly transverse between them. The sixth
abdominal appendages are still small, but in the next stage, the one shown in Claus’s
fig. 30, they are nearly as long as the telson, with six or seven marginal spines on the
outer edge of the exopodite and with the inner spine longer than the outer, with an
obscure lobe on its outer margin near the base. The whole hind body is now wide and
flat, and there are indications of five or six marginal spines on the inner edge of the

dactylus of the raptorial claw.
(ZOOL. CHALL, EXP.—PART XLy,—1886.) Yy 13

98 THE VOYAGE OF H.M.8. CHALLENGER.

I have shown that the Alima larva resembles the adults of the genus Squilla in the
. depression of the hind body, the presence of marginal spines on the dactylus of the rap-
torial claw, in the small number of marginal spines on the outer edge of the exopodite of
the uropod, in having its inner spime longer than the outer, with a lobe or dentation on
its outer edge, and in the presence of numerous secondary spines between the submedian
spine and the intermediate marginal spine of the telson.

In all of these respects Alimerichthus resembles Alima, and it is, therefore, beyond
question, a Squilla larva, but it differs from Alima in the great width of its telson and
the absence of spines on its abdominal somites, as well as in its resemblance to the more
primitive Hrichthus larva.

The comparative study of the adult Stomatopoda teaches that the genera Lysio-
squilla and Squilla are two divergent branches from a common stem, and that the
primitive Squille were more like this stem-form, and therefore more like Lysio-
squilla than the more specialized species. In the genus Squilla, Squilla (Chlorida)
microphthalma and its allies are the closest living representatives of the stem-form,
and they resemble the lowest species of the genus Lysiosgwilla in the small size and the
approximation of the eyes, the small size of the antennz and uropods, the loose
articulation of the hind body, and the width of the telson. I shall show further on
that the larva of Lysiosquilla, as well as of the more primitive genera of Stomatopoda,
is an Hrichthus, and that all the true Alimx are Squilla larve. The common ancestor
of Lysiosquilla and Sqwilla must therefore have passed through an Evichthus stage. If
it be true that the characteristics of the Alima larva are the result of secondary
modification, it is of course quite possible that the most modified adult Squallzx
might have their larvee the least modified, but in the absence of any proof that this is
the case, it is more natural to believe that the most typical Alime are the young of the
most typical Squall, and that Alimerichthus, the most primitive and Hrichthus-like of
the Alimx, is the young of a smooth loosely articulated and primitive Squilla, like
Squilla (Chlorida) microphthalma. While this conclusion cannot be accepted without
question, in the absence of direct proof, there is much reason for believing that
Alimerichthus is the larva of a Squilla closely related to Squilla nuwerophthalma, and
this decision receives added force from the fact that several of the most conspicuous
peculiarities of Alimerichthus as distinguished from Alima, such as the width and
shortness of the telson, and the loose articulation of the himd body, are points of
resemblance to Squilla microphthalma. In the true Alima the postero-lateral angles
of the abdominal somites end in acute spines, which are not developed in Alimerichthus,
and, as Squilla mucrophthalma is the least costate of the true Squillz, this is another
point of resemblance. It is not probable, however, that Alimerichthus is specifically
identical with Squilla microphthalma, and future research may prove that its adult
form is an unknown and still more primitive Squilla.

REPORT ON THE STOMATOPODA. 99

The Lysioerichthus larva, and the Metamorphosis of Lysiosquilla.

If my decision that all the Alima larvee are young Squille be correct, we must look
for the larve of all the other genera of Stomatopoda among the Frichthi and
Squillerichtha; or, as Squillerichthus is simply an adyanced Erichthus, among the
Erichthi.

The series of Erichthus larve is so complete, and transitional forms are so
“numerous, that it is very difficult to divide the group into minor groups; and while it is
obvious that there are several distinct larval types, they are so intimately united by
intermediate forms that the attempt to study them is very puzzling. The genera
merge into each other in such a way that it is difficult to find any strictly diagnostic
characteristics, but this is no more than we should expect from the absence of sharply
limited genera among the adult Stomatopoda.

I have shown that the species of Lysiosquilla, in which genus I include Coronis,
and the species of Sqwilla including Chloridella, exhibit proofs of divergent descent
from a common stem form, which was more like Coronzs and Chloridella than it was
like the more divergent Lysiosquille and Squillz; and as I have also shown that the
larvee of all the species in the Sqwilla-branch from this common stem are Alimex, we
naturally turn to the Alima-like EHrichthi in our search for the larval type of the second
or Lysiosquilla-branch.

In addition to their features of relationship to the adult genus Squwilla, the Alima
larvee agree with each other in the general occurrence of marginal spines on the lateral
edges of the carapace, the length of the telson, which is almost always greater than its
breadth, the flatness of the hind body and the presence of marginal spines on the inner
edge of the dactylus of the raptorial claw. Squilla and Lysiosquilla agree with each
other in the flatness of the hind body, and in the presence of spines on the dactylus, but
the Alima larva shows its relationship to Squilla by the presence of numerous
secondary spines between the submedian and intermediate marginal spines of the telson,
by the small number of spines on its dactylus, and by the fact that the inner spine of
the uropod is always longer than the outer.

Now there is a group of Hrichthus larvee, of which Hrichthus duvaucellei (Lysiosquilla
maculata ?) (Pl. X. fig. 7), and Erichthus multispinosus (Lysiosquilla excavatriz) (Pl. XI.
figs. 1, 2 and 3) are examples, which show by the flatness of the hind body, and by the
presence under the cuticle of the dactylus, in the older larve, of traces of marginal spines,
that they are either Sqwilla or Lysiosquilla larve. Claus refers them to the genus
Squilla, but as the marginal spines are usually more numerous than they are in any
known Squilla or in the Alima larva, we must exclude the genus Squilla in our attempt
to trace them to their adult form. In some of these larvee there are as many as seventeen
of these rudimentary spines on the dactylus, and they are seldom less than six, and there

100 THE VOYAGE OF H.M.S. CHALLENGER.

is ample internal evidence, drawn from other structural features, to show that they are
Lysiosquilla larvee; and in addition to this indirect but satisfactory evidence, I am now
able to furnish more direct proof, as I have reared a Lysiosquilla, Lysiosquilla excavatria,
from one of those larve (Pl. X. figs. 14, 15), and one of the Challenger specimens
(Pl. X. fig. 7) exhibits the characteristic transverse pigment stripes and other specific
characteristics of Lysiosquilla maculata, while another advanced larva of this type
(Pl. XI. figs. 6, 8, 9) exhibits the flattened oval thi membraneous appendages to the
exposed thoracic limbs which is characteristic of the lower Lysvosquille.

This larval type, for which I propose the provisional generic name Lysioerichthus,
merges into the Hrichthus of Gonodactylus and that of Psewdosquilla in such a way
that it is often difficult to decide whether a certain larva is to be referred to the one or
the other of these groups, but the Lysioerichthus may usually be distinguished from the
other Erichthi, by the position of the postero-lateral spies of the carapace, which, when
seen in profile (Pl. X. figs. 7, 14; Pl. XI. figs. 2, 3) are separated by a wide interval
from the dorsal middle line ; they are either on the ventral edge of the carapace or else
they are much nearer to it than to the dorsal middle line, while the reverse is the case
in the Erichthus of Pseudosquilla (Pl. XII. fig. 6) and that of Gonodactylus (Pl. XV.
figs. 11,12). In addition to this feature, which is, I believe, strictly diagnostic, they are
also characterised among the Hrichthi by the width and flatness of the hind body, and
by the great depth of the carapace, the lateral edges of which are ventrally infolded, as
shown in the figures.

During their younger stages they exhibit a most striking resemblance to the Alima
larva (Pl. XI. fig. 1; Pl. XI. fig. 4), although this resemblance is entirely lost by
the older larvee.

In addition to the greater number of spines on the dactylus of the raptorial claw they
are also distinguished by other features, all of which indicate their identity with
Lysiosquilla. The telson is wider than long, like that of Lysiosquilla, and unlike that of
Alima and Squilla, where it is, almost without exception, longer than wide ; and between
its intermediate and submedian marginal spines there is usually only one secondary spine,
and never more than four. This is true of the adult Lysiosquwilla also, while in all the
adult Squille and all the Alimx the secondary spines are more numerous.

The outer one of the two spines at the end of the ventral prolongation from the base
of the uropod is, with rare exceptions, longer than the inner, as is the case, also with rare
exceptions, in Lysiosquilla, while the reverse is true without any exception in Alima and
in Squilla.

The Lysioerichthus is a true Erichthus, although it is the most Alma-like of the
Erichthi. Like the Alima the young Lysioerichthus has numerous marginal spines on
the lateral edges of the carapace, although this characteristic disappears as the larva
grows older, while it is retained by the fully grown Alima larva. In the very young

REPORT ON THE STOMATOPODA. 101

stages of both types the thoracic region is elongated ; there are no traces of appendages
on the last six thoracic somites ; the hind body is elongated and narrow, and the carapace
shallow and flat. This general resemblance to an Alima disappears with the growth of
the larva, and the fully grown Lysioerichthus is not at all Alima-like, as its body is short
and wide, and the carapace so deep that it covers the sides and part of the ventral surface
of the free body, while the hind body also may be bent forwards and entirely covered by
the carapace.

These Hrichthi with a deep carapace and a wide flat hind body are very numerous
and widely distributed, and as they are also among the largest of the Stomatopod larve
we may be confident that they pertain to adults which belong to a widely distributed
genus, including many species, some of which are among the largest Stomatopods.

While we cannot feel at all confident that all the genera of adult Stomatopoda are
known, it is highly improbable that these larvze belong to an unknown genus, and we may
safely refer them to one of the well known genera. Their large size and the presence of
marginal spines on the dactyle exclude Protosquilla and Gonodactylus, and the
depression of the hind body excludes Pseudosquwilla, and we must therefore refer these
larve to either Sqwilla or Lysiosquilla.

Claus, as I have pointed out, advocates the first view, but the description which
follows will show that there is ample internal evidence that they are all Lysiosquilla
larvee, and this indirect evidence is rendered all the more conclusive by the fact that I
have reared Lysiosywilla excavatriz from one of them, while the Challenger collections
enable me to trace another to Lysiosquilla maculata with nearly equal certainty.

I therefore feel sure that the examination of the descriptions which follow will furnish
convincing proof that all these Hrichthi are young Lysiosquille, and that all the
Lysiosquille pass through the Lysioerichthus stage.

The Erichthus larva shown in Claus’s figure 14 is very similar to Lysioerichthus, but
it differs from them all in features in which they all agree with each other, and I shall
give farther on my reasons for believing that it is a Coronida larva, and therefore equally
related to both Alima and Lysioerichthus.

As I shall soon show, there is reason for distrusting the accuracy of his drawing of
the telson of his Hrichthus multispinosus, which is also exceptional.

Lysiosquilla (Erichthus) excavatric.—At Beaufort, N.C., U.S.A., where only two
species of Stomatopoda, Squilla empusa and Lysiosquilla (Coronis) excavatriz, are known
to occur, and where both species are abundant, two types of Stomatopod larvee are also
abundant, and as there is ample evidence to show that one of these, shown in Pl. IL.
fig. 4, is the young of Squwilla empusa, it is natural to infer that the other (Pl. XI.
figs. 1, 2, 3) is the larva of the Beaufort Coronis, Lysiosquilla excavatrix, of which
Pl. X. fig. 8 represents the adult male. Nearly all the larvee which I found were in the
stage shown in Pl. XI. fig. 1, and all my attempts to rear them in captivity failed, as

102 THE VOYAGE OF H.M.S. CHALLENGER.

they all died in moulting. The newly hatched young are swept out to sea by the tide,
and are widely distributed, and the older larvee are seldom found at Beaufort, but I have
procured a few specimens which serve to connect the youngest stage with a larva
nearly an inch long, which changed in the house into a young Lysiosquilla excavatriz,
so that there is little doubt that all the larve of this type found there belong to this
species.

The youngest larva which is figured is very similar to Claus’s Hrichthus
multispinosus from the Indian Ocean, although it is much younger, nor do they belong
to the same species, for Claus’s specimen has eight or nine, and ours only three, secondary
spines between the intermediate and the submedian marginal spines of the telson,
although in other particulars there is the closest resemblance between the two.

It has four distinct well-developed abdominal somites with appendages, while the
short fifth abdominal somite is not yet separated from the telson. The flagellum
of the second antenna and the appendages of the last six thoracic somites are
absent, although the somites are present and equal in length, with the exception of the
eighth, which is longer than the others. The rostrum is about as long as the carapace,
and it has two or three small spines on its lower surface. The carapace covers
all the thoracic, but not the abdominal somites, and its general outline, in dorsal
or ventral view, is nearly square, but its length slightly exceeds its greatest width,
and its width between the bases of the antero-lateral spines is less than between
the bases of the postero-laterals. In profile view (fig. 2) the lateral edges are
bent downwards below the level of the ventral surface of the thorax, and the chamber
which is thus formed is deepest at its posterior edge, so that the mid-dorsal outline and
the lateral edge are wide apart posteriorly, and approach each other in an acute angle at
the base of the rostrum. On the middle line of the posterior edge there is a slender
spine a little shorter than the rostrum, and about equal in length to the curved,
divergent postero-laterals, each of which carries a secondary ventral spine near its base,
dorsal to which, on the posterior edge of the carapace, there is a small secondary spine on
each side, as well as one on the lateral edge posterior to the base of each of the long,
divergent and widely separated antero-laterals, and on the lateral edge about halfway
between the antero- and postero-laterals there are two secondary spines. The telson is
considerably longer than wide, with its posterior edge nearly transverse, its anterior edge
narrow, and with six marginal spines on each side, the first, which is longer than the
next five, becoming the lateral marginal spine of the adult, is separated from the second
or intermediate by a wider space than those between the others. The eye-stalks are about
as long as the eyes, which are narrow with globular tips. The raptorial claw is flat and
oval, and there is a large prominent spine close to the proximal end of the carpus.

The next larva which is shown as seen from the left and below (fig. 3) is con-
siderably older and larger, and the marginal spines of the rostrum and carapace have

REPORT ON THE STOMATOPODA. 103

increased in size and in numbers. The sides of the carapace are folded down over the
body more than they are in the younger larvee, and the median dorsal spine is relatively
shorter. The appendages of the third, fourth, and fifth thoracic somites are represented
by buds, as well as the fifth abdominal appendages, and the three last thoracic somites
are now longer than those in front. While I obtained a number of specimens of this
stage I found only one older one, which moulted, before I had an opportunity to draw it,
into the young Lysiosquilla excavatriz, shown in Pl. X. fig. 13, although I was able to
make from the moulted skin the drawings of the carapace and telson which are given
in figs. 14, 15, and 16, of Pl. X.

Although this larva, the carapace of which was $2 inch long on the middle line,
including the rostrum, and 3% inch wide between the bases of the postero-lateral
spines, is many stages older than the one last described, the differences are so very
great that I at first doubted whether they could belong to the same series, but the
consecutive series of stages in the growth of the closely related larva which is described
in the next section furnishes satisfactory proof that this is the case. The antero-lateral
and dorsal spines are now very short, although the rostrum and postero-laterals are of
about the same relative length as before, and none of the marginal spines were visible
with the hand-lens under which the drawings were made, except one in front of the base
of each postero-lateral and one in front of this by one-third of the distance to the small
antero-lateral. The telson (fig. 16) is now nearly rectangular, a little (7) wider than long,
with very long submedian spine, and a nearly transverse but notched posterior border
carrying thirty-six small secondary spines with very minute spinules between them. As
compared with the submedians the other marginal spines are very small; the laterals
are posterior to the middle line and have each a small spinule internal to the base,
and one of the three secondary spines which were present in the younger larva has
disappeared, while the second is very small, and the third much smaller than the
intermediate. The change which takes place between this stage and the next, in the
shape of the telson, is fully as great as the difference between the telson at this stage and
that of the very young larva, as immediately after the moult the young Lysiosquilla has a
telson essentially like that of the adult shown in Pl. X. fig. 8, transverse, and about twice as
wide as long, with no secondary spines, and with all six marginal spines on the transverse
posterior border, and the submedians united in a single median process, as shown in fig. 13.

I am inclined to believe that a small Erichthoidina larva, which is occasionally,
although very rarely, found on the eastern coast of the United States, is the larva of this
species. Faxon has found one of these larvee at Newport, R.I., and he has figured it in
pl. viii. figs. 11 and 12 of his selection from Embryological Monographs.' | Through the
courtesy of Professor Baird I have also had an opportunity to examine a sketch of another
specimen which Professor 8. J. Smith obtained at the United States Fish Commission

1 Bull. Mus. Comp. Zool., vol. ix. 1, 1882.

104 THE VOYAGE OF H.M.S. CHALLENGER.

laboratory at Wood’s Holl, and I have found one specimen at Beaufort, although sickness
in my family prevented me from studying, and my hope of rearing it from preserving it.
It died, however, without moulting, like all the young Stomatopod larvee which I have
tried to rear in captivity. It is essentially like the Evichthoidina, from Honolulu, shown
in Pl. XII. figs. 1, 2, and very similar to Claus’s Hrichthoidina gracilis! and Erich-
thoidina armata (fig. 3), and it differs from the Hrichthoidina, shown in Pl. XII.
fig. 3, and from Claus’s Hrichthoidina brevispinosa, which I hold to be young Gonodactyli,
in the absence of a spine below the base of the postero-lateral, and in the greater relative
distance between this spine and the dorsal spine.

I shall give reasons for believing that these differences are characteristic of the
Lysiosquilla Erichthus as distinguished from the Hvichthus of Gonodactylus, and as we
find two corresponding types of Erichthoidina it is natural to believe that one becomes
converted at last into a Lysiosqualla, and the other into a Gonodactylus.

I therefore regard Claus’s Erichthoidina gracilis and Erichthoidina armata, the
Challenger Erichthoidina from Honolulu (Pl. XII. figs. 1, 2), Faxon’s larva and Smith’s
larva as very young Lysiosquille, and Claus’s Erichthoidina brevispinosa and the
Challenger Erichthoidina, from St. Vincent (Pl. XII. fig. 3), as young Gonodactyli, for
reasons which will be more fully developed in the sequel. If this is true we have a
most striking corroboration of the correctness of the opinion so ably and ingeniously
advocated by Claus (Crustaceen System) that the Stomatopod larva without appendages
upon the last six thoracic somites, in which condition the Alima larva leaves the egg, is
the phylogenetic descendant of a larva with biramous feet on all these somites, for this
change must actually occur during the ontogenetic development of Lysiosquilla excava-
trix, since Faxon’s and Smith’s larvee have biramous appendages on the third, fourth, and
fifth of these somites, like those on the first and second, although the youngest
Lysioerichthus, shown in our figure has no traces of them on these somites, or upon
the sixth, seventh, and eighth.

The great rarity of Hrichthoidina larve may possibly be due to the fact that
during this early period of its larval life, the young Stomatopod remains within the
burrow of its parent, or it may be that the larva does not usually escape from the ege
until this stage is passed, and that the few specimens which are met with at rare intervals
are those which have been prematurely hatched.

The analogy of other Crustacea, the various species of Alpheus, for example, shows
that two closely related species may hatch in different stages, and it is therefore possible
that one Hrichthus may hatch as an Hrichthoidina, while another hatches in the Erichthus
stage. Itis not impossible that some Alimw may hatch as Hrichthoidinex, although there
is no evidence that this is the case.

Before I enter upon the general discussion of the Lysioerichthus I will describe

1 Metamorphose der Squilliden, Taf. i. figs, 1, 2.

REPORT ON THE STOMATOPODA. 105

a few larval formss elected from the numerous Challenger specimens which I refer to this
genus, or to the subgenus Corons.

Coronis (Erichthus) minutus.—From a collection of larvee taken in the tow-net at St.
Vincent I have selected the series which is shown in Pl. XII. fig. 4; Pl XIII.
figs. 1-8, and 11, all of which undoubtedly belong to a single species. This is shown
by their general. resemblance and also by the accompanying table of measure-
ments. The length of the four specimens, measured from the tip of the rostrum
to the tip of the telson, are as follows :—

No. 1. No. 2. No. 3. No. 4.
4:16 mm, | 5°29 mm. | 6°49 mm. 10°21 mm.
If the length of No. 1, which is 4°16 mm.
be successively multiplied by $, we shall
have the following series of numbers, . 4:16 5:20 6°50 8:13 10°16 |

and the close agreement between the measured and the computed length shows that
Nos. 1, 2, and 3, are successive stages, and that there is a stage which is not represented
in the collection between Nos. 3 and 4, and that the length of the larva increases
uniformly at each moult by one-fourth of its length before the moult. It is hardly
conceivable that an accidental collection of unrelated specimens should show such close
conformity to a numerical law, and we may safely decide that the larva shown in Pl. XII.
fig. 4 becomes transformed, through the stages described, into the one shown in Pl. XIII.
fig. 7. This latter larva is so similar to the larva of Lysiosquilla (Coronis) excavatria
that the series from St. Vincent may also be safely referred to an adult of the subgenus
Coronis of the genus Lysiosquilla. It does not seem to have been described, and on
account of its small size, I propose for it the provisional name Erichthus (Coronis)
minutus.

The diagnostic characteristics of this species are as follows:—A small Lrichthus
with a broad flat hind body, a rostrum less than half as long as the carapace, and with
the postero-lateral spines of the carapace ventral, and without a tooth ventral to their base;
the raptorial claw (Pl. XIII. fig. 11) is flat and oval, and there is one large spine on
the anterior edge of the carpus close to its proximal end. The telson has six primary
marginal spines, the intermediates larger than the laterals and having a minute
secondary spinule internal to their bases, and one small secondary spine between the
intermediate and the submedian. The outer margin of the exopodite of the uropod
has few spines, and the basal prolongation ends in a long slender acute outer spine
with a broad base, and a very small inner spine. The carapace with the rostrum makes
a little more than half the total length, and its posterior edge, which has a small

(200L, OHALL, EXP,—PART XLV.—1886.) Yy 14

106 THE VOYAGE OF H.M.S8. CHALLENGER.

median dorsal spine, crosses the middle line above the posterior edge of the first
abdominal somite. The eyes are nearly cylindrical, with very short stalks, and the
hind body is flat and wide.

The youngest larva (No. 1, shown in Pl. XII. fig. 4) is so similar to the one repre-
sented in Claus’s figure 22 B, that we must believe that they are closely related. The
Challenger larva from St. Vincent is 4:16 mm. long, and our measurements indicate
that, if there be a younger stage, its length should be 3°33 mm., while Claus’s larva,
from Messina, is about 3 mm. long, and it is highly probable that it belongs to
this species, and is therefore a Coronis, although this author held that it is either
a Pseudosquilla or an Alima.’ The spines of the carapace are longer, and the eyes
longer and narrower in our species than in Claus’s figure, but these are the greatest
differences.

In our larva No. 1 (Pl. XII. fig. 4) the first antennze have only two branches, the
second antenne have no flagellum, and the third, fourth, fifth, sixth, seventh, and eighth
thoracic somites have no traces of appendages, and they are equal in length, although
the eighth is wider than the others. The thoracic ganglia are marked off by fissures
but they are in contact with each other.

There are five distinct abdominal somites with separate gangla and functional
appendages, each with an appendix interna or retinaculum on the thoracic edge of its
endopodite (Pl. XII. fig. 4).

The sixth abdominal somite is obscurely indicated, and its appendage is represented
by a bud (PL XIII. fig. 1). The carapace has a small dorsal median spine on its
posterior edge, the rostrum is about half as long as the carapace, with three small spines
on its ventral surface, and about as long as the divergent postero-lateral spines. The
lateral edges of the oval carapace are fringed with numerous minute serrations, but
there is no large tooth on the ventral side of the base of the postero-lateral spine. The
raptorial claws, even at this early stage, are flat and wide, and the carpus has a single
prominent tooth close to its base, on its anterior edge.

A profile view of the next stage (No. 2) is given in fig. 11 of Pl. XIII A
small bud now represents the flagellum of the second antenna (PI. XIII. fig. 4), and the
appendages of the third, fourth, and fifth thoracic somites are also represented by buds
(Pl. XIIL fig. 5), and the ganglia of the sixth, seventh, and eighth thoracic somites have
separated from each other. The sixth abdominal ganglion and somite are now distinct,
and the rudimentary appendage (Pl. XIII. fig. 3) consists of a long acute simple spine
and two rounded lobes. The posterior edge of the telson, which was transverse in
stage 1 is now angulated on the middle line. The next stage is shown from below
in Pl. XU. fig. 6. All the appendages and somites are now represented, the lateral

1 Metamorphose der Squilliden, p. 144.

REPORT ON THE STOMATOPODA. 107

edges of the carapace are folded inwards, and the minute inner spine has appeared upon
the ventral process of the uropod. The abdomen is now flat and wide and the lateral
edges of the fifth somite are acute. ;

The single specimen of the oldest stage (No. 4) was so badly preserved that many
points were obscure. It is shown in ventral view in Pl. XIII. fig. 7. The abdomen
is wide and flat, the lateral edges of the carapace are folded in, and the basal
process of the uropod is as long as the telson, which is deeply angulated on the middle
line (Pl. XIII. fig. 8). The secondary single spine between the submedian and the
intermediate is now much smaller than the latter, which has a minute spinule internal
to its base. The raptorial claw is flat and oval as it is in the adult Coronis.

The striking resemblance which Claus points out between the early stages of this
larva and a young Alima larva are superficial, since it is a true Hvrichthus, with the
ocular and antennary somites covered by the rostrum, the carapace folded downwards
and inwards at the sides, the outer spine of the basal prolongation of the uropod longer
than the inner, and only one instead of several secondary spines between the submedian
and the intermediate spines of the telson.

While there are numerous specific differences between it and the larva of Coronis
excavatrix, a comparison of the various stages in the development of the two larvee
will show their very close similarity, and there can be no doubt that this also is a
Lysiosquilla, and it probably belongs to the subgenus Coronis also.

Claus figures several advanced larvee which agree with this one in the depression of
the hind body, the number and relative size of the marginal spines on the telson, the

No. 1.| No. 2.| No. 3.} No. 4.|| No. 1.| No. 2.) No. 3.) No. 4.
Measurements on middle line :—

Rostrum from tip to anterior edge of

ocular somite, : *80 | 1:04.) 1:25} 1°99 192; 197 | 193 | 195
From anterior edge of ocular somite to

tip of labrum, 32 CTA |y ace “75 77 ide | aka 73
From tip of labrum to posterior edge of

carapace, ; 1:12 | 1:44 | 2°50 | 2:49) 269} 272} 385) 244
From posterior edge of carapace to posterior :

edge of telson, 1:92 | 2°40 | 2°74 | 4:98 462 | 454] 422] 488

Total length on middle line, A 4:16 | 5:29 | 6:49 | 10°21 || 1000 | 1000 | 1000 | 1000

Width of carapace between bases of pos-

tero-lateral spines, : = tas one 2°16
Width of fourth abdominal somite, ‘ ave Be we 1:76

shape and character of the uropod, the width of the raptorial claw, the position of the
postero-lateral spines of the carapace and other features, and as he has also been able
to show that the dactylus of the raptorial claw of the older larve is bordered by

108 THE VOYAGE OF H.M.S. CHALLENGER.

numerous marginal spines, there is little doubt that they are Lysiosquilla, pretty closely
related to the two which have been described. One of these, an Hrichthus 14 mm. long,
from Bengal,’ he himself refers to the subgenus Coronis, and his Squillerichthus
triangularis (fig. 18) and possibly his Hrichthus latreillec (fig. 18) are closely related
species.

The Challenger collection contains many specimens of a number of species of this
larval type, some of them very broad and short, and others as narrow as a G'onodactylus
Erichthus, from which they are distinguished by the position of the postero-lateral spine
of the carapace, which is either on the ventral edge or nearer to it than to the dorsal
middle line, while the reverse is the case in the Gonodactylus larva, which also has its
hind body convex instead of flat.

One of these Lysioerichthus larve from a surface collection made by the Chal-
lenger at Papua, Japan, is shown in Pl. XI. figs. 10, 11, 12, and 13. It is ;% of an
inch long on the middle line, with a minute dorsal spine, a rostrum which is nearly as
long as the carapace, small antero-laterals, and postero-laterals which spring from the
carapace close to its ventral edge, and reach backwards as far as the middle of the
sixth abdominal somite, while all the abdominal somites except the first are exposed
on the middle line. The raptorial claw (fig. 13) is flat, broad and oval, and its
dactylus shows under the cuticle traces of eight marginal spines. The hind body is
as wide as the carapace, and the telson (fig. 11) is much wider than long, with
its posterior border angulated and fringed with numerous small spinules between
the submedians, the distance between which is about half the greatest width of the
telson. There is a small spinule internal to the base of each lateral and each inter-
mediate marginal spine, and one small secondary spine between the intermediate and
submedian.

Length of carapace including rostrum, ‘ 5 : : “48 inch.

Length of exposed somites and telson, : : : : 42 «Cs,
Total length on middle line, . : : . : 90

Length of postero-lateral spines, . $ : ; : PZ Ghee

The oval shape of the flattened carpus indicates that this larva is a Coronis. There
is a general resemblance between it and the one last described, but the raptorial
claw is less oval, and in a third closely related larva, collected between Sydney and
Wellington, the raptorial claw is still more elongated. This latter larva, which is shown
in Pl. IX. figs. 6, 7, 8, 9, and 10, is so similar to the one shown in Claus’s figure 20,
that they must represent related adults, and I am therefore compelled to dispute his
identification of his larva as a young Gonodactylus.

It is remarkable for the great length of the rostrum and postero-lateral spines, the

1 Metamorphose der Squilliden, fig. 19.

REPORT ON THE STOMATOPODA. 109

latter reaching beyond the tip of the extended telson. The carapace has a long median
dorsal spine, and it exposes the posterior half of the third abdominal somite.

Length of carapace on middle line including rostrum, . 2 ‘D0 inch.
Length of exposed hind body, : ; : : : py) Ui
Total length from tip of broken rostrum, : : ; cL Omas
Length from tip of rostrum to tip of postero-lateral spines, . > xiii
Width of carapace between bases of postero-lateral spines, . . aUT IN;
Width of hind body, : - A - : ? Gi

In addition to the larvee which have been described, the Challenger collection contains
numerous specimens from various localities, which must represent closely related adults.
It also contains specimens of two somewhat peculiar larval types, which join to their
distinctive characteristics so many features which are shared by all the Lysioerichthus
larvee that I place them in this group.

One of them, which is represented by several specimens from Rio Janeiro, is shown in
profile view in Pl. IX. fig. 11, and in dorsal view in Pl. XI. fig. 6, while the telson, the
raptorial claw, and the seventh thoracic limb are shown in figures 7, 8, and 9. Itis a
little younger than Claus’s Squillerichthus triangularis,' but it belongs to the same or a
closely related adult. The dactylus of the raptorial claw of the oldest specimen in
the Challenger collection is smooth, but in Claus’s larva six marginal spines were
visible underneath the cuticle. This, as well as the width and flatness of the hind
body, the depth of the carapace, and the ventral infolding of its lateral edges, and
the shape of its telson and uropods (fig. 7) show its close relationship to the
Lysioerichthus larvee which have already been described, and the flat oval raptorial
claw (fig. 8) and the dilated oval scale-like form of the appendages to the exposed
thoracic limbs of the older larvee (fig. 9), indicate that the adult is one of the lower or
Coronis-like species of the genus Lysiosquilla. In Claus’s larva, which is slightly more
advanced than the oldest one in the Challenger collection, the raptorial claw exhibits
under the cuticle indications of six marginal spines, and this author therefore regards it
as the young of one of the six-spined species of Squzlla (p. 131). The fact that the young
Lysiosquilla excavatrix has a smaller number of marginal spines than the adult male,
shows that the presence of six spines in the larva is no evidence that they are not
more numerous in the adult, and while it is true that most of the adult Lysiosquille have
more than six spines, and that some of them have less, there are several species in
which the adult has only six. Claus says (p. 131) that the telson of this larva exhibits
the Squilla-type, but as all known species of Squilla and all the Alima larve have
numerous secondary marginal spines between the submedian and intermediate marginal
spines of the telson, while our specimens, as well as the one figured by Claus, have only one
such secondary spine, its relationship is obviously with Lysvosquilla rather than Squilla.

1 Metamorphose der Squilliden, fig. 13.

110 THE VOYAGE OF H.M.S. CHALLENGER.

A second very remarkable larva, characterised by the great elongation of all the
spines of the carapace, especially the antero-laterals, I refer, with less confidence, to the
Lysioerichthus group. The single specimen, which is shown in PL. XI. fig. 14, was taken
in the tow-net off Kandavu Island, Fiji.

Lysiosquilla maculata (Erichthus duvaucellet) Guérin.—The largest Hrichthus in the
Challenger collection is shown in profile in Pl. X. fig. 7, and from below in Pl. XI.
fig. 4. The collection contains numerous specimens which differ from each other only
in the length of the dorsal spine, which is often entirely absent, sometimes present but very
short, and occasionally well developed, as in the specimen from which fig. 7 was drawn.
The various specimens are so much alike in all other particulars that I cannot believe
that the length of this spine can be taken as an index of specific identity. It is probable
that in the older larva at least it is often broken off in moulting, and that the differences
in its length are accidental.

The largest specimens, which are more than one inch long, were collected between
Api and Cape York, between Admiralty Island and Japan, in the Straits of Mendino,
and at other points in the West Pacific. It is apparently the same as the Hrichthus
duvaucellet which Guérin obtained in the Gulf of Bengal, and Claus in the Indian Ocean.
Although Claus states (p. 135) his opinion that it is a Squilla larva, it is clearly a
Lysioerichthus, closely related to the various larvee which have been described. In the
larger specimens the integument of the carapace and of the abdominal somites is soft,
flexible and leather-like, as in Lystosquilla maculata, and the edges of the somites of
the hind body exhibit the transverse dark bands which are so characteristic of this species.
The inner spine of the basal prolongation of the uropod is longer than the outer, a relation
which is somewhat exceptional among the Lysiverichthus larve, and also among the adult
Lysiosquillz, although the adult Lysiosquilla maculata is one of the exceptional species,
having the inner spine longer than the outer. The raptorial claw of the larva is
long and slender, with traces under the cuticle of eight marginal spines, and as the
adult female Lysiosquilla maculata has seven or eight, while this number is increased in
the adult male to nine or ten, I cannot doubt that this large, flat, soft, transversely striped,
widely distributed larva, is the young of the largest of the Stomatopoda, Lysiosquilla
maculata, which is also flat, transversely striped, soft, and very widely distributed.

The Larva of the genus Coronida.

I have shown that we are led by the comparative study of the adult Stomatopoda,
to believe that Lysiosquilla and Squilla are the divergent descendants of a Protosquilla-
like form, with an acutely pointed rostrum, and minute uropods, and with the base of the
dactylus of the raptorial claw dilated as in Protosquilla and Gonodactylus, but armed with
marginal spines, and with the hind body depressed as it is in Squilla and Lysiosquilla,

REPORT ON THE STOMATOPODA. 111

- and with a wide telson with the marginal spines on its posterior edge. For this genus [
have proposed the name Coronida, and I have shown that we are acquainted with two
species which are to be referred to it. If this phylogenetic generalisation be correct, we
should expect the larva of this genus to unite in itself characteristics of both Alima and
Erichthus, and to stand in somewhat the same relation to them as that which the adult
Coronida bears to Lysiosquilla and Squilla. We should expect it to be a stem-form from
which both of these larvee may be derived. The Challenger collection contains no larve
of this character, and so far as I am aware only a single specimen has been observed.
This remarkable and interesting form, from the Atlantic, is shown in Claus’s figure 14. It
is much more advanced than any other Hrichthus or Squillerichthus larva which has ever
been described, resembling in this respect an Alima larva, and like the advanced Alima
larvee it has well developed gills, a long annulated flagellum on the second antenne,
a mandibular palpus, and its first five pairs of abdominal fect are, like those of the Alima-
larva, more perfectly developed than in Evrichthus, and it resembles all Alima larvee and
differs from all Erichthus larvee in the presence of numerous (twelve) secondary spines
between the intermediate and submedian spines of the telson. Like all Alima larve,
and the young and a few of the old Lysiosquilla Erichthus larvee, the lateral edges of the
carapace are fringed with spines, but these edges are folded downwards and inwards, and
in all other respects it is an Hrichthus. The many points in which it resembles Hrichthus
and differs from all Alime, joined to many other points of resemblance to Alima and
difference from all Hrichthi, render it peculiarly interesting. It is so far advanced that it
undoubtedly assumes its adult form after the moult which follows the stage shown in
Claus’s figure, and the adult rostrum with a long acute median spine is visible under the
cuticle. As it has a wide flat hind body and spines in the dactyle it is not a Protosquilla,
or a Gonodactylus, or a Pseudosquilla, and the long spine on the rostrum shows that it is
neither a Squilla nor a Lysiosquilla. The telson is wider than long, its marginal spines
are crowded backwards, the figure indicates that the sixth abdominal somite is probably
fused with the telson, and the uropods are very small, and the two spines are very smal]
and equal.

All its characteristics indicate that it is a very primitive and synthetic type, and
while it may possibly belong to an unknown genus, all the indirect evidence which it
furnishes indicates very strongly that it is either the larva of Coronida, or else of some
closely allied form. I propose for it the provisional generic and specific names Erich-
thalima synthetica.

Pseuderichthus and the Metamorphosis of Pseudosquilla.

Claus has traced to the adult Pseudosquilla a long narrow Erichthus larva which
differs from Lystoerichthus in the shape of the carapace, which is narrow and short, and

2 THE VOYAGE OF H.M.S. CHALLENGER.

also in the great length of the narrow hind body. His collection contained numerous
specimens, all very similar in general appearance to the one which he traced to its adult
form, from various localities in the Indian and Atlantic Oceans. They are all narrow
and elongated and from 16 to 42 mm. long.

The Challenger collection also contains numerous specimens from many widely
separated localities; all so similar to the one which Claus studied that they must
pertain to closely related adults, and I therefore place them all in a provisional genus
Pseuderichthus. One of them 1-42, inches long, from a collection made between
Tenerife and St. Thomas, is shown in dorsal view in PI. XII. fig. 6, while various parts
of its body, more highly magnified, are shown in Pl. VI. figs. 2 and 6, and PI. XIII.
figs. 12 and 14. The telson and uropods of a very similar but much larger specimen,
12 inches long, from Volcano, are shown in Pl. VI. fig. 7. Another much smaller
species, which undoubtedly belongs in the same group, although it is much wider and
flatter than the one figured above, is shown in side view in Pl. V. fig. 4. It is
24 inches long, and is from a gathering made between Api and Cape York. In all the
older larvee of this group the submedian spines of the telson are very long and slender,
and are tipped with movable spinules, as in the adult Pseudosquilla; the proximal
joint of the exopodite of the uropod is bordered by numerous (six to twelve) spines, the
terminals are much longer than the others, and the dactylus of the raptorial claw often
exhibits traces of two or three marginal spines under the cuticle. As all these
characteristics are features of resemblance to the adult Pseudosquilla, and as Claus has
obtained a very complete series of stages connecting one of these larvee with an adult
of this genus, there can be little doubt that they are all Pseudosquilla larvee.

While it is very closely related to both Lysioerichthus and the Hrichthus of Gono-
dactylus, and united to both these larval types by intermediate larval forms, I believe
the following features may be relied upon as diagnostic of the Pseuderichthus larva. It
is distinguished from the Lysioerichthus by the position of the postero-lateral spines of
the carapace, which are near the dorsal surface; by the narrowness of the carapace,
which is at least twice as long as wide, shallower than in Lysioerichthus, and not at all,
or only very slightly, infolded along its lateral edges; by the elongation of the hind
body, the length of the submedian spines of the telson, the presence of numerous spines
on the outer edge of the proximal joint of the exopodite of the uropod, and the very
oreat elongation of the outer one of the two ventral spines on its basal joint. It is
distinguished from the Hrichthus of Gonodactylus by the fact that the postero-lateral
spines of the carapace are short, usually only one-fourth or one-third as long as the
carapace, while they are usually more than half as long in the Gonodactylus larva, and
also by the fact that the rostrum is usually short and compressed, and armed at about
the middle of its ventral edge by a large acute curved spine, in front of which there are
often two or three smaller spines.

REPORT ON THE STOMATOPODA. 115

The adult genus Pseudosquilla is not sharply limited, as the various species are very
closely related to Gonodactylus on the one hand, and to Coronida and Coronis on the
other, and we should not expect the larval type to be more definite than the adult, and
as a.matter of fact the collection contains larvee which closely resemble Pseuderichthus
although they may possibly be Gonodactylus larvee, and others which may possibly be
Lysiosquilla larvee.

The Challenger collection contains very few young larve of this type, and I have
not been able to trace its’ metamorphosis, although Claus has given reasons for
believing that it hatches as an Hrichthoidina, and afterwards undergoes a retrograde
metamorphosis, loosing and afterwards redeveloping all the thoracic appendages, except
the first and second pairs.

The Hrichthoidina from St. Vincent, shown in Pl. XII. fig. 3, may possibly be a
young Pseudosquilla, although it more closely resembles the Gonodactylus Erichthus.

The Gonerichthus larva and the Metamorphosis of Gonodactylus.

The last larval type which I shall discuss is represented in the Challenger collection
by numerous specimens, a few of which I have selected and drawn.

One of these, from St. Vincent, is shown in Pl. XII. fig. 5, while the telson and
uropods of another specimen, from the Celebes Sea, in the same stage of development,
and belonging to the same or a closely related species, are shown in Pl. XIII. fig. 9.
The fully grown larva, 1428; inches long, of another species from the West Pacific,
lat. 17° 29’ N., long. 141° 21’ E., is shown in profile in Pl. XV. fig. 11, and from above
in fig. 6, and the telson and uropods of a closely related species 14 inches long, from
Volcano Island, in the West Pacific, are shown, more enlarged, in fig. 7. ,

The carapace of another species, 36°; inch long, from the Celebes Sea, is shown in
Pl. XV. fig. 1, and the telson of a young Gonodactylus of the Chiragra type in the adult
condition is shown in Pl. XVI. fig. 5. This specimen, which is ,8, inches long, was taken
with the trawl in 18 fathoms, west of the Philippine Islands, at Station 208. This latter
specimer. had the raptorial claw fully developed, and it exhibited all the characteristics of
the adult Gonodactylus. It is certainly a young Gonodactylus closely related to, or
possibly a specimen of, Gonodactylus chiragra, and a comparison of its telson and uropods
with those of the various larvee which have just been noticed will show that there is the
closest agreement in every particular. In all of these larve, as in the young Gono-
dactylus, the sixth abdominal somite has a pair of submedian spines near its posterior
edge, and its postero-lateral angles are produced into acute spines. The telson‘is slightly
wider than long, its submedian spines are long and slender, but shorter than they are in
Pseuderichthus, and without the movable spinules of the latter. The telson is notched

on the middle line, and there are from fourteen to twenty small secondary spinules on
(ZOOL. CHALL, EXP.—PART xLY.—1886.) Yy 15

114 THE VOYAGE OF H.M.S. CHALLENGER.

its posterior edge, between the submedians. There is one small secondary spinule
internal to the base of the lateral marginal spine, another internal to the base of the
intermediate, and a third midway between this and the submedian. In figs. 7 and 8
as in the young Gonodactylus the outer edge of the proximal joint of the exopodite of
the uropod is fringed by nine marginal spines, the terminal one longest, and the outer
spine of the basal prolongation is much longer than the inner, but not as long as it is in
Pseuderichthus. A comparison of the telson of the young Gonodactylus with that of
the other larvae which are figured in this paper, will show that none of them except
those now under discussion exhibit this resemblance. The Gonerichthus larvee which
are here figured are all of them well advanced, and are furnished with large compound
gills on their abdominal feet; this, together with the perfect development of their
uropods, shows that they are nearly mature, and about ready to moult into the adult
form, and as none of them exhibit any traces of marginal spines on the dactylus of the
elongated slender raptorial claw, we may feel confident that the adults belong to a genus
in which the dactylus is unarmed. It is not probable that a larval type which is so
common pertains to an unknown adult genus; the larve are not Protosquillz, as the
sixth abdominal somite is well developed, and as they have no movable spinules on the
tips of the submedian spines of the telson they are not Pseudosquillz. The only
remaining genus is Gonodactylus, and their structural characteristics .all indicate that
they are the young of species in this genus. Claus, who has figured two of these larvee
in his figures 21 A. and 21 A’. correctly refers them to Gonodactylus (p. 139) although he
also refers to this genus two larve, figures 20 and 21 B. which are not Gonerichthi but
Lysioerichthi, as he indeed suspects may be the case with the second one, 21 B.

The Gonerichthus larva may be distinguished from the Lysioerichthus by the
’ shallowness of its carapace, which is not at all infolded, and by the position of its postero-
lateral spines, which arise very close to the dorsal middle line. It is distinguished from
the Pseuderichthus larva by the length of these spines, which are at least half as long as
the carapace, and also by the fact that the telson is wider than long, and longer than the
long outer spine of the uropod.

INDEX.

Where more than one reference is given, the figures in darker type indicate the most important page.

Alima, 4, 16, 17, 18, 19, 20, 23, 24, 25, 52, 81, 82, 83,
84, 85, 89, 90, 93, 94, 95, 96, 97, 98, 99, 100,
101, 104, 106, 107, 109, 111 (PL iv. figs. 4-6 ;
PL vis PE px PL xa.)

Alima angustata, Dana, 84.

Alima bidens, 90, 91, 93, 95 (Pl. ix. figs. 1, 2).

Alima empusa, 90, 91, 97.

Alima gracilis, 84, 85, 88, 89, 90, 93, 94, 95, 96, 97
(Pl. iv. figs. 4-6 ; Pl. v. fig. 3; Pl. vi. figs. 3-5 ;
Pl. viii. figs. 4-6).

Alima macrophthalma, 93, 95, 97 (Pl. vii. figs. 1-6 ;

} Pl. viii. figs. 1-3).

Alimerichthus, 18, 19, 20, 24, 83, 95, 96, 97, 98
(Pl. viii. fig. 8; Pl. ix.).

Chiragra, 20, 113.

Chlorida, Eyd. et Soul., 23, 42, 55, 79, 80.

Chlorida decorata, Wood-Mason, 40, 42, 44.

Chlorida latreilii, Eyd. et Soul., 43. k

Chlorida microphthalma, Eyd. et Soul., 42, 44, 98.

Chloridella, Miers, 6, 9, 11, 14, 20, 23, 24, 35, 42,
44, 45, 99.

Chloridella chlorida, 14.

Chloridella decorata, Miers, 40.

Chloridella depressa, Miers, 42.

Chloridella microphthalma, 9, 12, 42.

Chloridella rotundicauda, Miers, 9, 43.

Coronida, n. gen., 6, 7, 10, 11, 12, 20, 55, 56, 66, '79,
80, 101, 110, 111, 113.

Coronida bradyi (A. M.-Edw.), 66, 79, 80.

Coronida trachura, 66, 79, 80.

Coronis, 5, 6, 11, 14, 44, 45, 50, 55, 70, 79, 80, 84,
99, 105, 106, 107, 108, 109, 113 (PL v. fig. 6).

Coronis excavatrix, n. sp., 9, 12, 14, 20, 21, 22, 45,
48, 101, 105, 107 (Pl. x. figs. 8-16; Pl. xi.).

Coronis minutus, 105 (PI. xii. fig. 4; Pl xiii. figs, 1-8,
i);

| ‘Coronis scolopendra, 44.

Erichthalima, 16, 17, 18, 19, 20, 24.

Erichthalima synthetica, 111.

Erichthoidina, 16, 17, 18, 19, 24, 25, 51, 55, 81, 82,
84, 95, 97, 103, 104, 113 (Pl. iv. figs. 1-3;
Pl. v. figs. 1, 2, 5, 7, 8; Pl. vi. fig. 1.; PL. xii).

Erichthoidina armata, 104.

Erichthoidina brevispinosa, 104.

Erichthoidina gracilis, 104.

Erichthus, 16, 24, 25, 44, 45, 46, 51, 52, 53, 55, 81,
82, 83, 84, 95, 96, 98, 99, 100, 101, 104, 105,
107, 108, 111, 112.

Erichthus duvaucellei, Guér., 46, 99,110 (PL. x. fig. 7 ;
Pl. xi. fig. 4).

Erichthus excavatrix, 101.

Erichthus latreillei, 108.

Erichthus minutus, 105 (PL xii. fig. 4; Pl xiii. figs.
1-8, 11).

Erichthus multispinosus, 24, 99, 101, 102 (Pl. xi. figs.
1-3).
Gonerichthus, 16, 17, 19, 20, 45, 113, 114 (Pl. v.
fig, 45 Plyxity;) Ply xi figs 10); Ply xy),
Gonodactylus, Latr., 6, 7, 10, 12, 14, 16, 17, 20, 22,
55, 56, 65, 66, 79, 80, 100, 101, 104, 108, 110,
111, 112, 113, 114 (PI. iv. figs. 1-3; Pl v.
figs 7); Ply xiig;) Plo sais: fig..93) Plo xiv. Pl xv;
PL xvi.).

Gonodactylus bradyi, A. M.-Edw., 55, 79, 80.

Gonodactylus cerebralis, 22.

Gonodactylus chiragra, Latr., 1, 22, 56, 58, 60, 61,

63, 64, 113 (PL. xv. fig. 4).
Gonodactylus chiragra (var. minutus), 22.
Gonodactylus elongatus, 22.”

Gonodactylus excavatus, Miers, 66, 78.
Gonodactylus folinit, A. M.-Edw., 66, 69, 70.
Gonodactylus furcicaudatus, Miers, 66, 78.

116

Gonodactylus glabrous, n. sp., 92, 60, 61, 62, 63, 64
(Pl. xiv. fig. 5; Pl. xv. figs. 7, 9).

Gonodactylus glyptocercus, Wood-Mason, 74, 75.

Gonodactylus graphurus, Miers, 14, 22, 58, 62, 63,
64 (Pl. xiv. figs. 1, 4, 6; Pl. xv. figs. 3, 8).

Gonodactylus guerinit. 2, 21, 22, 75.

Gonodactylus trachurus, Miers, 10, 55, 79, 80.

Gonodactylus trispinosus, Miers, 66, 71, 75.

Leptosquilla, 6, 34.

Lopholatilus, 1.

Lysioerichthus, 16, 18, 19, 20, 24, 45, 46, 48, 51, 99,
100, 101, 104, 108, 109, 110, 111, 112, 114
(Pl. v. fig. 6; PL vi; Pl. ix.; 1G sahf leh seth
Pl. xiii.).

Lysiosquilla, Dana, 6, 8, 9, 10, 11, {2, 14, 16; 17, 18,
19, 20, 25, 44, 45, 46, 55, 79, 80, 82, 83, 84,
90, 98, 99, 100, 101, 103, 104, 105, 107, 108,
109, 110, 111, 113 (Pl. v. fig. 8; Pl vi. fig.
fe Plant Phas Pl x5 Plixivy EL xiii.).

Lysiosquilla armata, 1.

Lysiosquilla excavatrix, n. sp., 2, 9, 12, 14, 20, 21, 22,

; 44, 45, 46, 48, 51, 99, 100, 101, 102, 103,
104, 105, 109 (Pl. x. figs. 8-16; Pl. xi. figs. 1-3).

Lysiosquilla inornata, Dana, 45.

Lysiosquilla maculata (Fabr.), 9, 14, 20, 21, 44,
45, 46, 48, 50, 99, 100, 101, 110, (Pl. x. figs.
127; Plex. iig.4))

Lysiosquilla scolopendra, 44.

Protosquilla, n. gen., 6, 7, 8, 10, MDs 7, 05 bb;
56, 6D, 66, 70, 72, 75, 78, 79, 80, 101, 110,
114 (Pl. xiv.; Pl. xv.; PL xvi.).

Protosquilla cerebralis, n. sp., 22, 66, 72, 75 (Pl. xiv.
figs. 2, 3; Pl. xvi. figs. 2, 3).

Protosquilla elongata, n. sp., 22, 66, 67, 70, 71 (PL xv.
figs. 2, 12; Pl. xvi. fig. 4).

Protosquilla excavata (Miers), 66, 78.

Protosquilla folinii (M.-Edw.), 66, ‘70.

Protosquilla furcicaudata (Miers), 66, '78.

Protosquilla glyptocerea (Wood-Mason), ‘75.

Protosquilla guerinii (White), 2, 21, 22, 66, 70, '75,
79, 80 (PL xvi. figs. 1, 6).

THE VOYAGE OF H.M.S. CHALLENGER.

Protosquilla trispinosa (White), 66, '71.

Pseuderichthus, 16, 17, 19, 20, 45, 111, 112, 113, 114
(BU vie hes: 2506; wgesel sextia):

Pseudosquilla, Guér., 6, 7, 10, 11, 12, 14, 16, 17, 20,
53, 55, 100, 101, 106, 111, 112, 113, 114
(PL. vi. figs. 2, 6, 7; Pl. xii; Pl. xiii. figs. 12,
13, 14).

Pseudosquilla ciliata, Miers, 21, 46,53, 54, 55 (PI. xv.
fig. 10).

| Pseudosquilla oculata, Brullé, 5D.

Pseudosquilla ornata, Miers, 55.

Pseudosquilla stylifera, v. Mart., 54, 55.

Squilla, Fabr., 6, 8, 9, 10, 12, 14, 16, 17, 18, 19, 20,
23, 24,.25, 44, 45, 55, 56, 70, 79, 80, 81, 82,
83, 84, 85, 89, 90, 93, 95, 98, 99, 100, 101,
109; 110; Ae (Pla. fies: Ly 25 Peis Plas
BU ivarsb Le vals ee levixe see othe).

Squilla bradyi, A. M.-Edw., 10, 55, 76.

Squilla chlorida, n. sp., 9, 14, 21, 24, 40 (Pl. ii.
figs. 1-5.).

Squilla dubia, M.-Edw., 37.

Squilla empusa, 1, 2, 20, 25, 51, 81, 83, 84, 85, 90,
91, 93, 97, 101 (Pl. i. figs. 4, 5; Pl. ii. fig. 7).

Squilla fasciata, de Haan, 2, 9, 21, 24, 35, 87, 39,
40, 44 (Pl. ii. fig. 8; Pl. iii. figs. 4, 5).

Squilla gracilis, 84 (Pl. iv. figs. 4-6; Pl. v. fig. 3;
Pl. vi. figs. 3-5; Pl. viii. figs. 4-6).

Squilla lata, n. sp., 9, 21, 24, 34, 38, 40, 44 (Pl. ii.
figs. 1-3).

Squilla leptosquilla, n. sp., 1, 14, 21, 30 (Pl i.
figs. 1, 2).

Squilla mantis, 31, 90.

Squilla microphthalma, M.-Edw., 9, 12, 42, 98.

Squilla miles, Hess., 33.

Squilla nepa, 1, 9, 21, 25, 30, 89, 90.

Squilla quinquedentata, n. sp., 14, 21, 26, 30 (Pl. i.
fig. 3; Pl. ii. fig. 6).

Squilla rotundicauda, 9.

Squilla rubrolineata, Dana and v. Mart., 37.

Squillerichthus, 16, 24, 25, 44, 84, 99, 111.

Squillerichthus triangularis, 108, 109.

PLATE I,

SQUILLA.

Fig, 1. Squilla leptosquilla, nu, sp., from the Celebes Sea. Dorsal view of a male
specimen, twice the natural size (p. 30).

Fig. 2. Squilla leptosquilla. Endopodite of first abdominal appendage of male; magnified.
A, basal joint ; B, terminal joint; a, outer lobe; b, inner lobe; ¢, suture ;
’ d, retinaculum ; e, fixed limb of petasma ; f; moveable limb of petasma.

Hea
iQ

ig. 3. Squilla quinquedentata, n. sp., from the Marquesas Islands. Male specimen;
natural size (p. 26).

Fj

_

g. 4, Squilla empusa, from Beaufort, N.C. Ventral view of young Alima larva,
3°216 mm. long (pp. 90, 101).

Ki

—

g. 5. Profile view of the same larva.

The Voyage of H.M.S. "Challenger"

PLATE

*

T

Fig.

Fig.
Fig.
Fig.
Fig.

Fig.

Fig.

Fig.

PLATE II.

Squitia (AxiMa).

. Squilla chlorida, nu. sp., from Amboina. Dorsal view of male specimen, one and

one-half inches long (p. 40).

. Squilla chlorida. Outline of telson ; more enlarged.
. Squilla chlorida. Anterior end of body ; enlarged.
. Squilla chlorida. Raptorial claw ; enlarged.

. Squila chlorida. Endopodite of first abdominal appendage of male; enlarged,

from a drawing by Mr. G. B. Haldeman. A, basal joint; B, terminal joint;
a, outer lobe; 6, inner lobe; c, suture ; d, retinaculum: e, fixed limb of
petasma; f, moveable limb of petasma.

. Squilla quinquedentata, from the Marquesas Islands. First abdominal appendage

of male, from a drawing by Mr. G. B. Haldeman (reference letters as in fig. 5).

. Squilla empusa, from Beaufort, N. C. First abdominal appendage of a male,

from a drawing by Mr. G. B. Haldeman.

. Squilla fasciata, from Japan. Raptorial claw of male ; enlarged (p. 37).

St omatopoda. PLATE II

.W.K. BROOKS Dal.

AHOEN 8 CO Lith
SOU: A ewe (ALIMA)

PLATE II.

(Zoob. OHALL. EXP.—PaRT XLV.—1886.)—Yy.
‘* » ey ia j

PLATE IIL.

SQUILLA.

. Squilla lata, nu. sp., from Arafura Sea, Dorsal view of a female specimen ; twice

the natural size (p. 34).

. Squilla lata. Raptorial claw ; enlarged.
. Squilla lata. Anterior end of body; enlarged.

. Squilla fasciata, from Japan. Dorsal view of a male specimen; three times

natural size.

. Squila fasciata. Submedian spine of telson; enlarged to show secondary

spinules.

Stomatopoda.PLATE ‘Til

"The Voyage of H.M.S. "Challenger.

- W. & BROOKS.0>1.

AHOEN & CO Lith

<t
al
al
S
CG
un

=
g
a

PLATE IV.

Squriis (ALIA).

. Gonodactylus (Erichthoidina). Telson of the specimen shown in PI. XII. fig. 3

(pp. 55, 118).

. Mandible of the same larva.

. Maxilla of the same larva.

. Squilla (Alima) gracilis, from St. Vincent. Stage No. 2, 6°91 mm. long (p. 84).
. Squilla (Alima) gracilis, from St. Vincent. Stage No. 4, 11°54 mm. long.

. Squilla (Alima) gracilis, from St. Vincent. Stage No. 5, 17°38 mm. long.

The Voyage of H.M.S. 'Challenger”

Stomatopoda. PLATE TV

; W.K BROOKS Dol, AHGEN BCO. Lith

S2Q0O fh ACA LDAP A)

4 _ a Se =
we > »
a a = - a fy 7 °
.
‘
:
‘
w
, -
‘
.
,
a
.
oe 7
‘
.
;
’ - .
.
- 4
ta
-
7 «
j
ho i a
4
(
» “
s
__
.
*
*
* " .

PLATE V.

. First maxilla of the larva shown in fig. 5.

. Second maxilla of same.

. Squilla (Alima) gracilis, from St. Vincent. Stage No. 3; 9°18 mm. long (p. 84).
. Gonerichthus larva, from St. Vincent (p. 113).

. Erichthoidina. Ventral view of the anterior end of body of larva shown in

Pl. XII. fig. 1.

. Abdominal appendage of the Coronis larva shown in Pl. XII. fig. 4.
. First antenna of the larva shown in Pl. XII. fig. 3.

. Telson of the larva shown in Pl. XII. fig. 2.

‘ The Voyage of H.M.S. Challenger” .

Stomatopoda. PLATE V

W. K. BROOKS.Dol. AHOEN & CO. Lith.

d Pel ih 7 oe 7
: .
-
4
®e
’
.
.
¥
‘ | |
.
$ ’
‘ | | |
'
\ ,
.
j 3
.
5 ,
| .
PLATE VI |
; rr re le
.
, 3 i
'
. ; f ‘ .
.
.
.
‘
'
.
. | | |
| J
| .
.
‘
*.
* a7
i
LT]
. 4
» |
é
. ‘ .
"

PLATE VI.

LyYSIOERICHTHUS.

. Lysiosquilla (2) (Erichthoidina). Telson of the larva shown in Pl. XII. fig. 1.
. Seventh thoracic limb of the larva shown in Pl. XII. fig. 6.

. Squilla-(Alima) gracilis, from St. Vincent. Stage No. 6; 42°83 mm. long.

. Squilla (Alima) gracilis. Stage No. 6. Mouth parts more highly magnified.
. Squilla (Alima) gracilis. Stage No. 6. Anterior end of body more magnified.
. First abdominal appendage of the larva shown in Pl. XII. fig. 6.

. Telson of a Pseuderichthus larva from Volcano, similar to the one shown in

Pl. XII. fig. 6, but 13 inches long.

ere « Ye P , ,
The Voyage of H.M.S. 'Challenger”

Stomatopoda. PLATE VI.

AHOEN &CO. Lith.

LYSIOERICHTHUS

’
i

i ay Aeery i, Tal oe
eo — iy .
| ae apne f ¥
nee, a :
i ; “— aay, ' —- uf, id
5 ea i
i : rae r Wu ’
a ay a ; r
\
° ‘ *
«
°
PLATE VII.
-
a .
(ZOOL. OHALL. EXP. —PART XLV.—1886.)—Yy,
. 2 .
i é :

5 ‘ ‘
& SPA
* > ae eA oh
ie ei ipal ae e
OS ee bi '
” do PO / *
rn ‘

PLATE VII.

Squr~ya (ALIMa).

. Squila (Alima) macrophthalma, from Cape Howe. Stage No. 2; 4°9 mm. long

(p. 93).
Squilla (Alima) macrophthalma, from Cape Howe. Stage No. 3; 6'4 mm. long.
Squilla (Alima) macrophthalma, trom Cape Howe. Stage No. 4; 8°4 mm. long.

Squilla (Alima) macrophthalma, from Cape Howe. Outline of telson of stage
No. 1; greatly magnified.

Squilla (Alima) macrophthalma, from Cape Howe. Outline of telson of stage
No. 4.

. Squilla (Alima) macrophthalma, from Cape Howe. Mouth parts of stage No. 1;

greatly magnified.

. Raptorial claw of an Alima larva, 14 inches long, from the coast of Africa.

Stomatopoda. PLATE VII

The Voyage of H.M.S. "Challenger™

i
}

iy wwriioany

PLATE VIII.

Squriia (ALIMA).

. Squilla (Alima) macrophthalma, from Cape Howe. Stage No. 5; 8°75 mm. long.

. Squilla (Alima) macrophthaima, from:Cape Howe. Stage No. 6; 11°11 mm.

long.

. Squilla (Alima) macrophthalma, from Cape Howe. Stage No. 7; 19 mm. long.

. Squilla (Alima) gracilis. Uropod of a fully-grown larva, collected between

Api and Cape York.

. Squilla (Alma) gracilis. Raptorial claw of a fully-grown larva, from St.

Vincent.

. Squila (Alma) gracilis. Telson of larva shown in Pl. VI. fig. 4.
. Telson of an Alima, 14 inches long, from the coast of Africa.

. Alimerichthus from the South Atlantic ; 8°715 mm. long.

“Sar 4 The Voyage of ‘HLM.S. "Challenger.

S00) iL GALA EMA )

|
:

PLATE IX.

—Pant xtv.—1886,)—Yy.

es
wa |

7

g
; <4
q

igs
Fig. 2
Hig, 3
His,’
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11

PLATE IX.

Squitita (Atma) (ALIMERIcHTHUS) (LystosQuiLLA) LysioERICHTHUS.

. Squilla (Alima) bidens, 1 inch long, from the Cape of Good Hope (p. 90).
. Telson and uropods of same larva.

. Alimerichthus, 15°52 mm. long, from the South Atlantic.

. Telson and uropods of a younger Alimerichthus ; 8°715 mm. long.

. Lysiosquilla (Lysioerichthus). A larva of the subgenus Coronis, 4% ich
long, collected between Sydney and Wellington.

. Telson and uropods of same larva.

. Telson and uropods of a younger specimen.

. Side view of the carapace of the larva shown in fig. 6.
. Raptorial claw of the same larva.

. Lysiosquilla (Lysioerichthus) triangularis, from Rio Janeiro.

Stomatopoda. PLATE IX

WK. BROOKS.Dsl. AHOEN & CO. Lith,

SQOUILLA (ALIMA) ALIMERICHTHUS)
ists OW IOLA) Paper stoAta us

Fig.
Fig.

my LLL

5 Ie

5 Use

r, 14.

15.

16.

PLATE X.

(LysrosQuriLia).

. Lysiosquilla maculata. Raptorial claw of male; one-half natural size (p. 45).
. Lysiosquilla maculata. Raptorial claw of female ; one-half natural size.

. Lysiosquilla maculata. Mandible.

. Lysiosquilla maculata. First maxilla.

. Lysiosquilla maculata. Second maxilla.

. Lysiosquilla maculata. First abdominal appendage of adult male; slightly

enlarged. e, exopodite; A, basal joint of endopodite; B, terminal joint
of endopodite.

. Lysiosquilla (Lysioerichthus) maculata ; magnified 5 diameters.
. Lysiosquilla (Coronis) excavatrix, n. sp.; male, magnified 2 diameters.
. Lysiosquilla (Coronis) excavatrix. First maxilla.

. Lysiosquilla (Coronis) excavatrix. Second maxilla.

Lysiosquilla (Coroms) excavatrix. Raptorial claw of male.

Lysiosquilla (Coronis) excavatrix. Endopodite of first abdominal appendage
of adult male; greatly enlarged (reference letters as m fig. 6; and, in
addition, a, outer lobe; b, inner lobe; ce, suture; d, retinaculum; e, fixed
limb ; f, movable limb of petasma).

Lysiosquilla (Coronis) excavatrix. Young specimen; enlarged 5 diameters.

Lysiosquilla (Coronis) excavatrix. Side view of the carapace of the larval
skin, 3% inch long, from which the young specimen shown in fig. 13
moulted.

Lysiosquilla (Coronis) excavatrix. Ventral view of same.

Lysiosquilla (Coronis) excavatrix. Telson and uropods of the same larva,
#5 inch long, 34; inch wide.

VEN SPOS Ow lL PA)

Stomatopoda. PLATE ¥

AROEN 8 CO. Lith.

_

. 13.

is

PLATE XI.

LySIOERICHTHUS.

. Lysiosquilla (Coronis) excavatric. Ventral view of a young Hrichthus larva

(Erichthus multispinous, Claus).

. Lysiosquilla (Coronis) excavatriz. Side view of same larva.
. Lysiosquilla (Coronis) eacavatrix. Side view of older larva.

. Lysiosquilla maculata. Ventral view of Squillerichthus larva.

Lysiosquilla maculata. Marginal spinules of telson.
Lysiosquilla (Lysioerschthus) triangularis, from Rio Janeiro.

Telson and uropods of same.

. Raptorial claw of same.

Seventh thoracic appendage of same.

. Lysiosquilla (Lysioerichthus), 85 of an inch long, collected between Papua and

Japan.

. Telson and uropods of same.

. Side view of carapace of same.

Raptorial claw of same.

Lysioserichthus (2), from Kandavu Island, Fiji.

a sexe es.
Lia 1.

‘5
-

_ The Voyage of H.M.S. Challenger”

Stomatopoda. PLATE XI

f

. K. BROOKS Del.

LYSIOERICHTHUS AHOEN ACO, Lith.

fi *
e =
‘ ,
we
7 d
.
7 4
= 43° (
j 7
ey res Mara
: i
~ 5 _- PT
i i) h AA) 24 . ‘ ry a ; . Pasta
lacey reir i PAE RSH) f 1a ae Bd ust a A ah
: . 3 iff pi ™
s ve . : u : »- yr 7, ri -
, ; o “gM, - he AV ee
> Li . », 7 : ranien S ei sn Le) 1 "34 . |
) 4 4 7 ” a . ‘ 4 te =
\ [ yon rou a ae ap tons a aiee 2
Ve PLATE XI. Cty
= : L tuaAtit i
i 7° . » : spe ; is
* < a oa, “4 1 Tr 1
a ¥ te oe j { 3 sly - y
1 e*- j 43, + iP Ad 7 7 ¢ ce} ary d int] aes iy PY, i
| ee :
hg Patt ca ee
Ph vy rk 12% @ rai "7 Fgh ey (Aah -
7 ; S r 7 4 ~ oO
¥ by
‘
pi
wie ou |
Pra? t' 4
-t
F. .
+
- ]
. F '
y
¥
A
. “ La -
7 ’ > has
é ‘ f
re a : - “ i
r . :
“- , ; Ad

PLATE XII.

1-3. EricHTHompina. 4. LysiogRICHTHUS. 5. GONERICHTHUS. 6. PSEUDERICHTHUS.

Fig.

. Erichthoidina (Lysiosquilla 2), from Honolulu. Zeiss, A, 2.

Older stage of same.
Erichthoidina (Gonodactylus ?), from St. Vincent, Cape Verde. Zeiss, A, 2.

Lysiosquilla (Lysioerichthus) minutus, from St. Vincent, Cape Verde. Stage No. 1;
416 mm. long.

. Gonodactylus (Gonerichthus), from St. Vincent, Cape Verd. Zeiss, A, 2.

Pseudosquilla (Pseuderichthus), 1:14 inches long, collected between Tenerife and
St. Thomas.

ee

The Voyage of H.M.S. 'Challenser”

Stomatopoda. PLATE XII

W.K. BROOKS Dol.

1-3 ERICHTHQIDINA
5 GONERICHTHUS

AHMOEN &CO. Lith
* LYSIOERICHTHUS
S6PSEUDERICHTHUS

be |

” A
oe . oe ™ he a Re “

'
&
a “

*

5

PLATE XIII.

Fig. 1
Fig. 2
Rigs 3
Fig. 4
Fig. 5
Fig. 6
Hig. “7.
Fig. 8
Fig. 9
Fig. 10
Bigs
Fig, 12
Fig, 13.
Fig, 14.

PLATE XIII

Squrtta (ALIMA).

. Corons (Erichthus) minutus, from St. Vincent. Hind body of the larva shown

in Pl. XII. fig. 4, without the appendages (p. 105).

. Mouth parts of same larva.

. Telson of a slightly older larva. Stage No, 2; 5°20 mm. long.
. Antenna of same larva.

. Nervous system of same larva.

. Coroms (Erichthus) minutus, from St. Vincent. Stage No. 3; 6°46 mm.

long.

. Coronis (Erichthus) minutus, from St. Vincent. Stage No. 4; 10°16 mm.

long.

. Telson and uropods of same larva.

. Telson of a larva similar,to the one shown in Pl. XII. fig. 5.
. Telson of the larva shown in Pl. V. fig. 4.

. Profile view of the carapace of the larva shown in fig. 3.

. Anterior end of the body of the larva shown in Pl. XII. fig. 6.

Thoracic limb of same larva.

Raptorial claw of the larva shown in Pl. XIL. fig. 6.

ly Nod ee 5 ee ee

The Voyage of H.M.S. Challenger”

Fig. 14

W.K. BROOKS Dol.

SQUILLA (ALIMA)

Stomatopoda. PLATE XIII.

COCO!

=

“

wv
I.

= (yey
ei

'_
Lo
I)

AHOEN BCO Lith.

eed:

PLATE XIV.

GONODACTYLUS. PROTOSQUILLA.

. Gonodactylus graphurus (Miers). Side view of a male specimen, 234, inches

100

long, from Cape York (p. 58).

. Protosquilla cerebralis, n. sp. Side view of a female specimen, 1,22; inches

long, from Fiji.

. Protosquilla cerebralis. Raptorial claw (p. 72).

. Gonodactylus graphurus. Dorsal view of sixth abdominal somite of a male

specimen, with telson and uropods.

. Gonodactylus glabrous. Dorsal surface of sixth abdominal somite, with telson

and uropods (p. 62).

. Gonodactylus graphurus. Anterior end of body, greatly enlarged.

a Sel a gd eee ale i _

Stomatopoda. PLATE XTYV

AHOEN 3B CO Lith.
GONODACTYLUS BOTs Omi fw

at Mey

¥

key
(=e
jo}

bo

Fig. 3
Fig. 4
Hig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12

PLATE XV.

GONODACTYLUS. PRoTosQuILia.

. Gonodactylus (Gonerichthus). Carapace of a Gonerichthus-larva, +8, inch lone,

smaller than the one shown in fig. 11, but older, from the Celebes Sea.

2. Protosquilla elongata, n. sp. Dorsal view of a specimen, yoo inch long, from

St. Vincent (p. 67).

3. Gonodactylus graphurus. Raptorial claw.

1. Gonodactylus chiragra. Endopodite of the first abdominal appendage of a male

specimen. A, distal joint ; a, its outer lobe; b, its inner lobe; B, basal
joint ; c,c, suture; d, retinaculum ; e, fixed limb of petasma: 4, movable
limb.

- Gonodactylus (Gonerichthus). Telson and uropods of the larva shown in fig. 1.

- Gonodactylus (Gonerichthus). Dorsal view of a Gonerichthus-larva, 1,28, inches

100

long, from the West Pacific,

. Gonodactylus glabrous, n. sp. Dorsal view of a specimen, 2;j9 inches long,

TOC

from Samboangan Reef (p. 62).

. Gonodactylus graphurus. Petasma of a male specimen ; reference letters as in

fig: A,

. Gonodactylus glabrous. First abdominal appendage of a male; reference letters

as in fig. 4.

|
. Pseudosquilla ciliata. Endopodite of first abdominal appendage of a male;

reference letters as in fig. 4.

} °
- Gonodactylus (Gonerichthus). Side view of a Gonerichthus-larva, 1,28, inches

long, from the West Pacific.

. Protosquilla elongata. Magnified view of the ventral surface of the telson of

the specimen shown in fig. 2.

ee ee ee
" ive =
The Voyage of H.M.S. Challenger”

Stomatopoda. PLATE XV

EN
Wo
Kg 10

GONODACTYLUS PROTOSOUILLA ten oe:

.
.
cl
.
> s
‘
‘
’
é
ns ¢
a!
‘

PLATE XVI.

PROTOSQUILLA.

. Protosquilla guerinii (White). Dorsal view of a female specimen, 13, inches

long, from Honolulu (p. 75).

. Protosquilla cerebralis, n. sp. Dorsal view of a female specimen, 152%, inches

long, from Fiji (p. 72).

. Protosquilla cerebralis. Uropod of the specimen shown in fig. 8, from below.

. Protosquilla elongata. Anterior end of the body of the specimen shown in

PL XV. fig. 2.

. Gonodactylus. Telson of a young specimen of a Chiragra-like species, from 188

fathoms, at Station 208, near the Philippine Islands.

. Protosquilla querinu. Ventral view of the telson and uropods of the specimen

shown in fig. 1.

The Voyage

WK. BROOKS Dol

of H.M.S. Challenger”

PIROTOSQ@UILEA

Stomatopoda. PLA

n &CO

eS ob. tay ee aa
oe Ve

4

>