75/9

O ;

a !
a '•

a •
a >

tr ;
tr •

Ln |

JT

«\

II. Lucifer : a Study in Morphology.

By W. K. BROOKS, Associate in Biology and Director of the Chesapeake Zoological
Laboratory of the Johns Hopkins University, Baltimore, Md., U.S.A.

Communicated by Professor HUXLEY, Sec. R.S.

Received April G,— Read April 28, 1881.

[PLATES 1-11.]
CONTENTS.

Page.

SECTION I. — Introductory , 58

SECTION II. — Segmentation of the egg, and formation of the food-yolk and primitive digestive cavity 64

SECTION III. — Metamorphosis of Lucifer 72

SECTION TV. — History of each appendage of Lucifer 92

SECTION V. — The Metamorphosis of Acctes 101

SECTION VI. — Relation between the larvae of Lucifer, Acetes, Sergestes, Peiut-us, and Eaphausia, and

the significance of the Decapod Zoea, and the Crustacean Nauplius 109

SECTION VII. — Serial homology and bilateral symmetry in the Crustacea 125

SECTION VIII. — Explanation of the plates , 129

SECTION I. — INTRODUCTORY.

THE general anatomy of the adult Lucifer has been satisfactorily made known by
the observations of SOULEYET, HUXLEY, HENSEN, DANA, SEMPER, GLAUS, DOHRN, and
FAXON ; and the only facts which I have to add relate to the structure of the
reproductive organs.

The earliest recorded observations upon this subject are by DANA (' United States
Exploring Expedition during the Years 1838, 1839, 1840, 1841, and 1842,' under the
Command of CHARLES WILKES, U.S.N., vol. xiii., part 1). In plate 44, fig. 9, b, h, and
m", he gives a very correct representation of the male reproductive organs and sperma-
tophore of an adult male specimen of Lucifer (acestra) ; but his description of these
figures (p. 670) shows that he was completely at a loss for an interpretation of the
parts which he has represented, and had no idea of their true function.

Later students have entirely overlooked these figures by DANA, and there has been

MDCCCLXXXH. I

58 MR. W. K. BROOKS ON LUCIFER:

some doubt whether Lucifer is an adult animal at all, rather than the young form of
some other Decapod.

In 1801, SEMPER (Reisehericht des Herrn Dr. SEMPERS. Ein Schreiben an
A. KOLLIKER ; Zeit. f. Wiss. Zool., xi., 1861, pp. 100-108) re-discovered and described
the male organs, and also the female organs, of a large, transparent, and probably new
species which he found at Zamboanga.

He gives no figures, and his short account, which is in the main correct, is as follows : —

" Die Geschlechtsoffnung ist einfach, liegt bei beiden Geschlechtern in der Mittel-
linie des Baucb.es, dicht liinter dem letzten Brustfusse. Der Hode besteht aus einer,
in der Mittellinie des Thorax, dicht unter dem Magen liegenden Samendrtlse, an
deren hiuteres Ende, dort wo der kurze Samenleiter entspringt, sich mehrere Neben-
driisen ansetzen. Der Same wird, noch unentwickelt, in einen birnformigen grossen
Spermatophor eingeschlossen. Das hmterste Ende dieser mannlichen Driise' reicht bis in
die Mitte des ersten Hinterleibgliedes, das vorderste bis ziemlich dicht an den Schlund.

" Das Weibchen hat zwei Eierstocke, die vom Ende des sechsten Hinterleibgliedes
an dicht unter dern Darm, sich bis in die Mitte des Thorax erstrecken, hier biegen sich
die beiden Samenleiter nach unten, und schwellen dann zu zwei grossen Taschen an,
die eine kleine Tasche umfassen ; die Geschlechtsoffnung ist einfach ; ein einziger
Spermatophor steckt mit seinem spitzen Ende darin. Entwickelte Zoospermien
habe ich nicht beobachtet. WeibHche Begattungsorgane fehlen. Die Entwickelungs-
geschichte ist mir unbekanut geblieben."

The male organ has two external openings ; they are not on the median line, and
their position in the body does not correspond to that of the female orifice ; but in
other respects my own observations show the correctness of this description.

As SEMPER does not give any account of the general structure of these sexual
individuals, GLAUS (" Ueber einige Schizopoden und niedere Malacostraken Messinas,"
Zeit. f. Wiss. Zool., xiii., 1863, pp. 433-437) held that the adult nature of Lucifer
must still be a matter of uncertainty ; but in 1871 DOHRN verified SEMPER'S account
from alcoholic specimens ('' Untersuchungen iiber Ban und Entwickelung der Arthro-
poden," von Dr. ANT. DOHRN, Zeit. f. Wiss. Zool., xxi., 1871, p. 357), and showed
that the mature animals have the form which had been described by THOMPSON
('Zoological Researches and Illustrations,' 1829, p. 58) as characteristic of the genus.

In the following year SEMPER ("Zoologische Aphorismen, von G. SEMPER : I. Einige
Bemerkungen iiber die Gattung Lucifer," Zeit, f. Wiss. ZooL, xxii., 1872, p. 305)
published a second paper, in which he gave two good figures of the male and female
reproductive organs (taf. xxii., figs. 3 and 4), but added nothing to his earlier
description.

During my own studies upon the development of the larva I found an abundant
supply of adult specimens of both sexes, and am thus enabled to give a more complete
account of the structure and relations of the reproductive organs.

Plate 9, fig. 75, is a side view of the carapace (c) and the first abdominal

A STUDY IN MORPHOLOGY. 59

somite (a) of an adult male, showing the first abdominal foot (PI. 1) and the basal
joints of the third maxilliped (J//>. 3), and the first, second, and third thoracic limbs
(Pi: 1, Pi: 2, Pr. 3). The testis (t) consists of a series of about eight pouches or
follicles, which hang down into the body cavity under the anterior end of the intestine
(/). The body of the animal is so thin that it is almost impossible to get a good
dorsal view without crushing the specimen; but a very careful examination of the side
view seems to show that there is only a single organ on the median line of the body,
as SEMPER states. On each side of the intestine, along the line where the testis joins
its wall, a small tubular vas deferens (vd) arises, and runs backwards along the side of
the intestine nearly to the end of the first abdominal somite, to which it seems to be
attached (at I) by a ligament. It then bends outwards and forwards upon itself to
form a second much larger portion (sp), which is parallel to and outside of the first
portion, and reaches nearly to the anterior edge of the first abdominal somite. The
third or terminal portion (sv) has a large cavity, thick walls, and it runs down to an
external opening which is situated on the outer edge of the sternal surface of the
thoracic region, behind the basal joint of the third pereiopod, and therefore in the
position which would be occupied by the basal joints of the fourth or fifth pereiopods
if they were present.

There is a vas deferens, made up of these three portions, on each side of the body,
and the ventral nerve chain (ty) passes between their external openings.

The more anterior follicles of the testis are almost perfectly transparent, but the
development of the male cells in the posterior ones gives to them a faint granulation.
The first division of the vas deferens (vd) has a small cavity, thin walls, and as it
usually seems to be entirely empty it is probable that the passage of the male cells
from the testis through it to the second division (sp) takes place quickly. The second
division (sp) has a very large cavity, and in it the male cells become arranged in a
single layer around the surface of a central core, which is formed of some dense
transparent adhesive substance.

The sperraatophore appears to pass into the third chamber (sv) before it is completely
formed, as all those which were seen in the second chamber consisted only of a central
core and a layer of male cells, while those which were contained in the thick-walled
third chamber had an outer enveloping capsule.

I found several specimens with a fully-developed spermatophore on one side of the
body and none on the other side, and was thus enabled to thoroughly satisfy myself
of the presence of two vasa deferentia, and two external openings.

I was unable to discover how the spermatophore is transported to the body of the
female, or what part the clasping organ (c) upon the first pleopod of the male performs
during the act of copulation.

Upon several occasions I observed a male clinging to the basal joints of the first
antennas of a female, but as I never succeeded in getting the pair under a lens with-
out separating them, I made no careful examination. Copulation usually takes place

GO MR. W. K. BROOKS ON LUCIFER:

during the daytime, or at least this was the case in every instance which I observed.
In several cases I found female specimens with a single fresh spermatophore attached
to the opening (Plate 9, fig. 74, o) of the seminal receptacle (sr). This opening is
situated between and a little anterior to the basal joints of the third pair of thoracic
limbs (Pr. 3 of fig. 74). As the spermatophore gradually discharged its contents, it
was easy to see that both the central core and the investing layer of spermatozoa
escaped from the outer sheath and passed into the seminal receptacle. In all the
breeding females which I observed the spermatozoa filled the posterior, and the trans-
parent core of the spermatophore the anterior half of the spermatic receptacle, as shown
in fig. 74. The ovary is very long (fig. 74, or), and it lies under the intestine,
reaching from the fifth abdominal somite to the posterior edge of the carapace, where
it bends upon itself at right angles and runs down to its external opening, which is
upon or close to the median line of the ventral surface, a little in front of the third
pair of pereiopods. The wall of the ovary is so very thin and delicate that I was not
able to detect it at all except when it was filled with ripe ova. These are very much
elongated, granular, and slightly opaque ; and there does not seem to be any shell
around them. They are very elastic, and undergo great changes of shape as they pass
through the small oviduct.

Oviposition occurs between 9 and 10 o'clock in the evening, a.nd occupies only a few
minutes.

After the eggs are laid they are spherical, transparent, and each one has a rather
thick shell. They are attached, in a loose bunch of twenty or more, to the last pair of
thoracic limbs, and in order to save space I have shown them in fig. 74, although the
specimen from which the figure was drawn had not laid any of its eggs.

As I obtained very few ripe females, I was not able to sacrifice one of them to study
the reproductive organs under pressure, and I am therefore unable to decide whether
any parts of this system are double ; but I feel confident that there is only one sper-
matic receptacle, and the opening of the oviduct seems to be upon the median line.

We found a few adult specimens out at sea, but, while I was able to learn little
about their habits, I think that they are not strictly pelagic, but that their proper
home is the salt marshes close to the ocean.

They Avere met with in the greatest abundance about half-a-mile inside Old Topsail
Inlet, near a large marsh, during the first hour of the ebb tide, on calm evenings when
the tide turned between 7 and 8 P.M. ; and I infer that they leave the marshes at this
time to breed in the ocean. All the mature females which we found, with one excep-
tion, were captured under these peculiar conditions ; and we never failed to find them
at this spot when the tide turned about sunset and the water was calm.

Owing to this singular limitation there are only a very few favourable evenings for
procuring the eggs in a single season ; and until the animals can be made to thrive
and multiply in confinement, it must always remain an extremely difficult matter to
procure the eggs in abundance.

A STUDY IN MORPHOLOGY. 61

Up to the present time our knowledge of the early stages of Lucifer has been
extremely meagre.

In his report on the Crustacea of the United States Exploring Expedition, DANA
described (p. 634) an organism under the name of Erichthina demissa, and figured
it in plate 42, fig. 3. GLAUS ('Crustacean System,' p. 13) gives a figure of the same
organism at a latter stage of development, and calls attention to the numerous features
of resemblance to the Protozoea stage of development of Penceus.

Only a few months before his death, the lamented WiLLEMoES-SuHM collected a
number of specimens of Erichthina in the South Pacific, and, associated with them,
a sufficient number of later stages to assure him that Erichthina is the larva of
Lucifer. His account (" Preliminary Ptemarks on the Development of some Pelagic
Decapods," by R. von WILLEMOES-SUHM, Ph.D., Proc. Eoyal Soc., Dec. 9, 1875,
p. 132-4) is very brief, and as it contains all that is known about the metamor-
phosis of this extremely interesting form, I quote it in full : —

" Very similar to that of Sergestes is the development of Leucifer. Here the earliest
Zoea of a species from the Western Pacific has got at first no eyes, then sessile ones
came out, and the animal then presents the form which DANA has called Erichthina
demissa, and which GLAUS suspected to be not a Stomatopod but a Schizopod larva.
After the second moulting this Erichthina gets stalked eyes, and very long setse
on all its appendages, becoming a rather long, very delicate Zoea. It now enters the
Amphion stage, but never gets more than four pairs of pereiopods, and loses another
pair of these when it moults for the youngest Leucifer stage, in which two pairs of
pei^eiopods are absent.

" The next question after having found this out, was, of course, whether Amphion,
Sergestes, and Leucifer leave the egg as a Zoea, or whether there is a preceding Nauplius
stage. My own impression is that in the two first-named genera this is not the case,
as the youngest Zoeas which I ca,ught had all the same size, and as none of them was
without the large lateral stalked eyes. As for Leucifer, the question appears to me to
be doubtful ; for it is, from what I have seen, quite possible that my youngest Zoea,
which has only got a central eye, may be preceded by a Nauplius. Of course, the
simplest thing would be to get the eggs ; but there is the difficulty, for Amphion is
caught very rarely, and has never been obtained at any other time but between
8 and 12 P.M., when it is extremely difficult by lamplight to find out the youngest
stages. Senjcstes larvae are commoner, appearing also in the daytime, and Leucifer is
sometimes caught in abundance. I hope, therefore, that I shall succeed in completing my
researches about this question, especially as far as the two latter genera are concerned.

" H.M.S. 'Challenger,' Honolulu, Sandwich Islands.
"July 30, 1875."

As the sad death of this lamented naturalist, only a short time after, put an end to
tins as well as to his other researches in all departments of zoology, I take pleasure in

G-2 MR, W. K. BROOKS ON LUCIFER:

stating that I have fortunately been able to complete his observations upon Lucifer,
and to furnish a very perfect account of its entire metamorphosis, as well as a few
important facts with reference to its development in the egg.

At the end of April, 1880, I found a single specimen of Lucifer with two eggs
attached to one of its appendages, and I was led by the great importance and interest
of the subject to make every effort to trace its life-history. For four months I met
with no success whatever, but about the 1st of September I found a few advanced
larvas, and traced them to the adult, and I then succeeded in finding earlier stages
and tracing them as far as the stages which I had previously found, but it was not until
the last week of my season at the sea-shore that I succeeded in hatching the Navplius
from the egg, and the last gap in my series was bridged by a moult which occurred
only a few hours before my departure.

As the result of my four months' efforts I can now state that I have seen the eggs
of Lucifer pass out of the oviduct. I have seen the Nauplius embryo escape from the
same egg which I had seen laid, and I have traced every moult from the Nauplius to
the adult in isolated specimens. There is therefore no Crustacean with the meta-
morphosis of which we are more thoroughly acquainted than we now are with that of
this extremely interesting genus.

Not only is it true, as WiLLEMoES-SuHM has pointed out, than DANA'S Erichthina
demissa is a larval stage of Lucifer, but that DANA'S Sceletina armata is a later stage
in the same series, while some of the forms which he includes in his genus Furcilta
are also, in all probability, Lucifer larvre.

The occurrence of a free Nauplius stage of development in the life-history of one of the
higher Crustacea is a matter of such profound significance in the scientific discussion
of the phenomena of embryology in general, that it can hardly be accepted without
question so long as there is any possibility of error. Two of the observers who have
testified to its occurrence have based their conclusions upon evidence which would be
perfectly satisfactory in any ordinary case, but as they did not actually trace all the
stages of development their statements do not stand the severe analysis which the
importance of the case demands, and certain naturalists have therefore refused to
give them unqualified acceptance.

The third observation was made so many years ago, and the larva is so briefly
described, that it would not be safe to assume, in the absence of all corroborative
evidence, that it is a Nauplius at all.

In December, 1838, DANA found in the harbour of Ptio de Janeiro great numbers of
specimens of a Schizopod, which he described (' United States Exploring Expedition
during the Years 1838-1842,' under the command of CHARLES WILKES, U.S.N.,
vol. xiii., part i., p. G54) as Macromysis gracilis. In the brood-pouches of some of
his specimens he found an abundance of eggs and developing embryos, several of
which are shown in his plate 45, fig. 5. He made no careful study of their structure ;
his notice of them in the text is only a few words ; and his figures are very small,

A STUDY IN MORPHOLOGY. 63

and show the embryos in dorsal view, as seen under a very low magnifying power,
but they are so much like FRITZ MULLER'S figures, that we must acknowledge that
the credit of the first discovery of a Malacostracan Nauplius belongs to DANA, and
that up to the present time this is the only case in which a Nauplius has been traced
to an egg which could be definitely identified as that of a specific adult Mala-
costracan, although his account is so imperfect that in itself it is certainly not
sufiicient to prove the existence of the Nauplius stage at all.

In 1861 FRITZ MULLER found, at Desterro, in Brazil, a single specimen of a Nanp-
lius ("Die Verwandlung der Garneelen." Erster Beitrag von FRITZ MULLER, in
Desterro, Arch. f. Naturgeschichte, 1863, p. 9), which he traced, through other speci-
mens winch were also collected in the ocean, to a form which he believed to change
into the youngest Zoea of a species of Penceus. The series of stages is so satisfactory
that there is no reason for doubting the accuracy of his conclusion, but the chances
for error, in the attempt to trace Crustacean development from isolated specimens, are
so very great that the statement has not received unqualified acceptance.

The only other recorded observation of a Malacostracan Nauplius is not among the
Decapods, but in the more embryonic Schizopods. These observations, which were
made by METSCHNICKOFF, would tend to corroborate those by MULLER, but they are
unfortunately open to the same criticism. He did not actually rear the larvae and
trace them to a specific adult, and although there would in ordinary cases be no doubt
of the correctness of his conclusion, a careful analysis of his papers will show that
there certainly is a possibility of error.

In the spring of 1868 he collected from the surface of the ocean at Messina a few
early stages in the development of a Crustacean, which he believed to be Euphausia
mulleri (GLAUS), and showed (" Ueber ein Larvenstadium von Euphausia " von
EL. METSCHNICKOFF in Petersburg, Zeit. f. Wiss. Zool., xxix., 1869, p. 479, taf. xxxvi.)
that it passes through a well-marked Nauplius stage, of which he gives three figures.

The following year, at Villafranca, he collected a good supply of young Iarva3 and
floating eggs in advanced stages of development, and was thus enabled to supplement
his first paper by a second (" Ueber die Naupliuszustande von JHuphausia," von ELIAS
METSCHNICKOFF, Zeit, f. Wiss. Zool.7 xxi., 1870, p. 380, taf. xxxiv.) in which he gives a
minute account of the Nauplius from the time it leaves the egg until it changes into a
form somewhat similar to the youngest stage of Euphausia, which had been previously
described by GLAUS (" Ueber einige Schizopoden und niedere Malacostraken Messinas,"
von Prof. Dr. C. CLAUS, Zeit. f. Wiss. Zool., xiii., 1863, p. 422). CLAUS had supposed
this to be the stage in which the larva leaves the egg, and he says (p. 450), " Diese
Larve bin ich geneigt fur die jiingste aller freieren Entwickelungsformen der Euphausia
anzusehen." He subsequently learned, however (" Untersuchungen zur Erforschung
der Genealogischen Grundlage des Crustacean-Systems," p. 9), that he had been in error,
since he afterwards found, in an Atlantic and also in a Mediterranean species, an earlier
Protozoea stage, which changed into the Zoca described in his first paper. It therefore

64 MR. W. K. BROOKS ON LUCIFER:

follows that METSCHNICKOFF studied something else, or that he was in error in
believing that he had traced his Nauplius directly to what CLAUS has shown to be a
somewhat late stage in the development of Euphausia. METSCHNICKOFF'S only reason
for believing that his Nauplius is a young Euphausia is its resemblance to CLAUS'S
larva, and as there is certainly an error here, we are not justified in giving unqualified
acceptance to his statement that it is an Euphausia larva. It seems very probable,
indeed, that this is the case, but in the absence of the direct evidence which could only
be afforded by actually tracing it back to an Euphausia egg, or forwards to the adult
Euphausia, I do not think that the existence of a Malacostracan Nauplius can be said
to be established by these observations, for they do not stand the severe test which is
demanded by their unusual importance, and I think the facts justify the statement
that, up to the present time, there has been no unquestionable evidence of the
occurrence of such a stage of development in the higher Crustacea.

The present series of observations is complete at both ends, and I have not relied
upon surface-collecting to fill a single gap, but have traced every stage in isolated
captive specimens, and the possibility of error seems to be entirely out of the
question.

The close resemblance between the Nauplius of Lucifer, and MULLER'S and
METSCHNICKOFF'S larvae, renders it almost certain that they also are Malacostracan
larvte, but before this corroborative evidence was furnished, it was certainly quite
possible, although hardly reasonable, to doubt whether this was true of either of them.

II. — THE SEGMENTATION OF THE EGG, AND FORMATION OF THE FOOD-YOLK AND

PRIMITIVE DIGESTIVE CAVITY.

Unusual difficulties attend the study of the early stages in the embryology of
Lucifer, and the observations which I have been able to make are incomplete, and
leave many gaps to be filled and many interesting points to be decided by future
investigations ; but the facts which I have made out are so novel, so different from all
that was previously known of the early stages of Arthropod development, and they
throw so much light upon the relation of the peculiar and greatly modified form of
segmentation characteristic of the group to the less modified form of segmentation
presented by the more normal eggs of other animals, that it seems best to give my
results in their present incomplete state.

I am the more willing to do this, because the peculiar difficulties of the subject leave
little hope for the attainment of more complete results in the future.

The eggs are so loosely attached to the appendages of the female that they are
broken off by the slightest roughness of handling, and it is very difficult to obtain them
by collecting the egg-bearing females. Even when great numbers of mature specimens
are captured in the breeding season, with the greatest care and delicacy, very few of

A STUDY IN MORPHOLOGY. 6.3

them, much less than 1 per cent., are found to have eggs attached to their limbs when
the collection is examined.

If the mature animals could be induced to thrive and multiply in confinement, there
would be no difficulty in obtaining a sufficient supply of eggs, but until this can be
done it must be extremely difficult to procure them in sufficient numbers for exhaustive
study.

During the early stages the eggs are so delicate that they are soon destroyed by
the confinement and compression to which it is necessary to subject them while they
are under examination, and it is therefore impossible to watch very many stages
in a single egg.

When we add to this that the eggs are laid about 9 o'clock in the evening, and
must be studied between this time and daylight, after several hours of laborious
collecting, by eyes that have been already severely taxed with looking over the collec-
tions and picking out the transparent and almost invisible adults by an artificial
light, and examining each one of them with a lens to find those which carry eggs, the
difficulty of the subject will be appreciated.

The eggs are spherical, transparent, and they contain extremely little food- material.
This is uniformly distributed over the whole egg in minute globules, which have nearly
the same colour and refractive index as the sunminding protoplasm.

The egg undergoes total regular segmentation, and a true segmentation cavity
occupies the place filled by the large central yolk-mass in the eggs of other Arthropods.

It first divides into two equal portions (Plate 1 , fig. 1 ) ; then, by a cleavage at
right angles to the first, into four (fig. 5) ; then into eight (fig. 8) ; then into sixteen
(fig. 10) ; and so on.

At the stage shown in fig. 10 the inner ends of the sixteen spherules are seen to be
separated from each other by a central space, the segmentation cavity, which persists,
and is shown at later stages in figs. 11, 13, 15, 16, 17, 19, and 20, at b.

In fig. 10 the egg will be seen to be spherical, and all the segments have their
broad ends at the surface ; but in the next stage one pole of the egg becomes a little
flattened, and in an optical section the spherule (c), which occupies the centre of the
flattened area, is seen to have its broad end nearest the centre of the egg.

Most of the food-material has meanwhile disappeared from the other spherules,
which are now quite transparent, while the spherule (c) still contains as much as
ever, but apparently no more than there was contained at an earlier stage in an equal
area of any part of the egg. In an optical section of the same egg, in a plane at right
angles to that of fig. 11, the spherule (c) shows a trace of a fissure, which a little later
divides it into two (see fig. 12, c).

Plate 2, fig. 13, is an optical section, like the one given in fig. 11, of a somewhat
older egg; and fig. 14 an optical section of the same egg at right angles to fig. 13.
The outline is a little more flattened on one side than it is in fig. 11, and the

MDCCCLXXXII. K

66 MR, W. K. ERODES ON LUCIFER:

spherule (?) is completely divided by a radial fissure into two, and these project into
the segmentation cavity (b) a little more than they did before.

In fig. 15 the flattening has become a deep pit (d), and the spherules (c) have been
pushed quite into the segmentation cavity, and the adjacent cells have begun to move
in the same direction. This change is more marked in fig. 16 ; and in fig. 17 the egg
consists of a double wall of cells, the ectoderm and the endoderm, surrounding a
primitive digestive cavity (d), and separated from each other by the segmentation
cavity (b), in which the two cells (c) are situated. Each of these also shows traces of
a division into two.

These changes are more marked in fig. 19 ; and in fig. 20 the opening of the
primitive digestive cavity is much reduced in size, and the cavity itself does not lie
exactly in the axis of the egg, but at one side of it.

A more minute examination of the segmentation brings out a number of interesting
points ; one of them is the rhythmical character of the process, which is not a con-
tinuous uniform change, but a series of stages of activity, separated from each other by
periods of rest.

The egg shown in Plate 1, fig. 1, was laid about 10 o'clock P.M., and about 10.35 it
was in the condition which is represented in the figure. As I had not been watching
it I did not observe the first division, and when first seen it was in the resting
condition, and the two spherules were not sharply defined, but pressed together.

During the next fifteen minutes no external change was visible, and the drawing
was made at 10.50 P.M. It then entered upon the second period of segmenting activity,
and in five minutes the two spherules were well defined, as shown in fig. 2 ; and in
five minutes more (fig. 3) one of them showed traces of division into two. In ten
minutes more (fig. 4) this division was completed, and traces of a similar change had
made their appearance in the other spherule, which was also perfectly divided into
two at the end of five minutes more (fig. 5). This stage ended the second period
of activity, which was twenty-five minutes long.

During the whole of this tune the egg showed gradual and uniform change, which
was sufficiently rapid to be distinctly visible. Although four so-called stages are
figured, there was no division into stages, but a continuous change without interruption.

The four spherules now began to flatten down, and in five minutes the egg was in
the condition which is shown in fig. 6, and it then remained without any external
change for more than ten minutes. The second period of rest, measured from the time
when the four spherules began to shrink together to the time when they began to
swell out and enter upon the third period of active segmentation, was therefore more
than fifteen minutes long.

At 11.40 the four spherules were once more sharply defined (fig. 7), and changes
went on uniformly until, at 12.15 A.M., each was perfectly divided into two, as shown
in fig. 8, which marks the end of the third period of activity, thirty-five minutes long.

I was not able to watch this egg pass into the next rusting stage, as it had been so

A STUDY TN MORPTTOMHJY. 07

long under observation (1 hour 45 minutes) that its development was arrested at this
point ; but another egg in this stage of development was seen to pass into the resting
condition, as shown in fig. 9, and it then remained quiet for about fifteen minutes,
showing no external indications of change during this time.

At the end of the third period of rest the spherules again became prominent, so that
the outline of the egg was exactly like that of fig. 8, and the egg entered upon the
fourth period of activity, soon dividing into sixteen spherules (fig. 10), arranged around
a segmentation cavity.

In about twenty-five minutes from the beginning of this period of activity the
spherules began to flatten down once more, and the egg passed into the fourth resting
stage, but it was not observed beyond this point.

The alternation of activity with rest was observed at much later stages, but after
the gastrula invagination makes its appearance the cells of the endodermic portion of
the egg do not undergo active change at the same time with those of the ectoderm,
and the egg has one set of periods of activity for each layer. As development goes on
the periods of rest grow longer and the periods of activity shorter, and the spherules do
not flatten down while at rest.

The egg which is shown in optical longitudinal section in fig. 1G was in the field of
the microscope for nearly twenty minutes, while I was examining another specimen.
An occasional look at it showed that it was not changing, but at the end of this time
I noticed that the outer ends of the ectoderm cells directly opposite the orifice of
invagination were notched, as is shown in the figure. Activity spread in all directions
from this point, and in less than five minutes all the cells were notched, and those
nearest the centre of the area of activity were perfectly divided into halves. In about
five minutes more all the ectodermal cells had divided, and this layer had the
appearance shown in fig. 17 — which, however, was drawn from another specimen.

This last egg remained in the condition shown in the drawing for fifteen minutes
from the time it was first observed, and a movement of the appendage to which it was
fastened caused it to roll over and present its formative pole for examination before the
beginning of the next period of activity, which is shown in surface view in fig. 1 8.
The manner of division was simply a repetition of that which has just been described.

The cells nearest the centre of the formative area became notched, and then divided
into halves ; and the activity gradually spread over the egg in all directions, until, in
a few minutes, all the cells which were visible were at some stage of division.

The rapidity and uniformity with which this change spread over the egg rendered it
an extremely interesting and impressive sight, and I know of no other case in which
segmentation is so perfectly regular at such an advanced stage of development.

The activity did not affect the endoderm cells in either of these cases, but at a later
stage (fig. 20) they were seen to be in an active segmenting condition at a time when
the ectoderm cells were at rest. I was not able to keep this egg alive long enough to
watch the completion of the process, for it had been under the microscope for some

K 2

68 MR. W. K. BEOOKS ON LUCIFER:

time before the stage shown in the figure was reached ; but the division of the
endoderm cells appears to go on much more slowly than that of the ectoderm cells.

This phenomenon, the alternation of periods of rest with the periods of active
segmentation, does not seem to have received from embryologists the attention which
it deserves. A number of observers have pointed out that in many animals, among
the Mollusca especially, the distinctness of the spherules becomes more or less com-
pletely obscured after each division, and that this state persists until just before the
next division, when the spherules swell out and again become conspicuous. The
change of form does not seem to be at all general, and in most accounts of segmentation
nothing of the kind is recorded.

I believe that it is a secondary phenomenon, and that the essential thing is the
alternation of rest with activity ; and I am confident that careful time records of
segmentation will show that this occurs in nearly every case, sometimes with and
sometimes without the accompanying change of form.

I have observed it in Physa, Limncsus, and Planorbis, where segmentation is total
and nearly regular ; in the Oyster, where the egg has a rudimentary food-yolk and
segmentation is irregular ; in a bony fish with a large food-yolk and a discoidal seg-
mentation ; and in Lucifer, Other investigators working under my guidance have
observed it in Amllyxtoma and in oligochsetous and polychretous Annelids. These
are all the cases in which I have been able to test the matter since my attention has
been attracted to the subject ; and as the alternation was found to occur in every case,
although the animals are so widely separated and present such diverse modes of
segmentation, I feel justified in assuming that the phenomenon is general, and will be
found in all eggs which can be properly examined by watching and timing them while
segmentation is going on.

The cause of rhythmical physiological change is an extremely interesting question ;
and as the segmenting egg exhibits the phenomenon in the greatest possible simplicity,
it would seem to be a peculiarly favourable subject for investigation.

The phenomena which have been described seem to show that segmentation is not
clue to the action of any purely molecular force, like polarity, but is essentially a vital
activity, and in a paper on the embryology of the fresh-water Pulmonates (' Studies
from the Biological Laboratory of the Johns Hopkins University,' vol. i., part ii.)
I have ventured the following explanation.

During the period of segmentation the protoplasm of the whole egg (of Physa)
gradually becomes more and more transparent, on account of the gradual disappear-
ance of the granular food-material which it contains, and the rhythmical character
of the process of segmentation would seem to admit of a simple explanation on the
supposition that the physical properties of the protoplasm offer a resistance which
must be overcome before the force which is set free by the assimilation and reduction
of the food-material can exert itself to bring about the active changes of segmentation.
During a period of rest the process of digestion and assimilation accumulates a store

A STUDY IN MORPHOLOGY. 09

of energy which, at length, becomes sufficient to overcome this resistance, and to
initiate a period of activity which lasts until the whole of this reserve of force has been
expended in the rearrangement of the protoplasm. The physical properties of the
protoplasm now reassert themselves, and tend to reduce the whole egg as nearly as
possible to a spherical form once more, and the egg then remains inactive until the
supply of energy again becomes great enough to overcome the resistance.

If this is the true explanation we should expect to find the alternation of rest and
activity much more general than the change of form, for the degree of consistency of
the protoplasm or the amount or character of the food-material, or the way in which
it is distributed through the egg, may prevent the second set of changes from showing
themselves. This is precisely what we do find, and in the bony fishes, where the large
food-yolk would prevent any marked change of form, we find the first set of changes
well marked, but with no trace of the second set.

Leaving this subject for the present, I wish to say a few words about another
interesting phase of the early stages of Lucifer. We cannot fail to be impressed by
the very remarkable departure from ordinary Arthropod segmentation, nor can we
overlook the fact that in all the points of difference from the eggs of allied forms, the
eggs of Lucifer show a most suggestive resemblance to the ordinary unspecialized ova
of other Metazoa.

In an ordinary Arthropodan egg we have, as the outcome of the process of segmen-
tation, a central mass of food-yolk, which may or may not be divided into segmentation
products, and which completely fills the segmentation cavity ; and an outer investing
layer of blastoderm cells; that is, the egg undergoes a centrolycethal segmentation."5'

In most Crustacea the early stages of segmentation are regular, and apparently
total, but the lines of cleavage do not pass entirely through the egg, and the spherules
are united to each other by a central mass of food-yolk. When segmentation is
somewhat advanced the products of segmentation become more or less pyramidal,
with the bases of the pyramids at the surface, and their apices fused together at the
centre of the egg. The outer ends of the pyramids then become transparent and
separate off as a blastoderm, while the inner portions usually fuse together, more or
less perfectly, to form a central food-yolk, which fills the space which in ordinary eggs
constitutes the segmentation cavity. A small portion of the blastoderm then becomes
invaginated to form the primitive digestion cavity, and the remainder becomes the
ectoderm.

The centrolycethal type of segmentation presents great variations in the different
groups of Arthropods, but in nearly all cases its peculiarities are so well marked

f The whole subject of segmentation has been so ably and exhaustively reviewed by BALFOUR in his
recent work on ' Comparative Embryology,' that it does not seem necessary to burden this paper with
a long list of references to the literature of Arthropod segmentation, or to enter into an exposition of the
present state of our knowledge of the subject. All the essential facts and opinions may be found on
pages 79-99, 317-379, and 425-433 of vol. i. of the ' Comparative Embryology.'

L I 3 R A R Y
W

70 MR. W. K. BROOKS ON LUCIFER:

that it is difficult to trace any resemblance to the various forms of segmentation which
occur in other groups of animals. In Lucifer the case is reversed, and we have a type
of segmentation which is obviously similar to that of the Echinoderms, Annelids,
Molluscs, Tunicates, Vertebrates, &c., but is less obviously related to that of the eggs
of closely allied forms. The resemblance to what may be called " normal " segmenta-
tion is so plain that it need not be dwelt upon, but the relation between the egg of
Lucifer and an ordinary centrolycethal egg is by no means clear.

It seems probable, however, that since the food -material which has not been assimi-
lated becomes centralised, after segmentation is somewhat advanced, in the single
spherule c, of fig. 11, this spherule must correspond to one of the yolk-pyramids of an
ordinary Crustacean egg. This then divides, by radial fission, into two portions (fig.
13, c), and it seems probable that the food-material then becomes restricted to their
central ends, while the outer protoplasmic ends separate off as a pair of blastoderm
cells (fig. 15), thus leaving the two masses of food-yolk (c) inside the segmentation
cavity. While I was investigating the subject I regarded the spherule c, of fig. 11,
as a primary mesoblast, which became pushed into the segmentation cavity, and then
divided up to form the mesoderm ; and I expressed this view without comment
in a preliminary abstract of the subject (" Embryology and Metamorphosis of the
Sergestida?," Zoologischer Anzeiger, iii., p. 563). In most cases where the origin of the
mesoderm has been most carefully studied, it originates by the separation of the inner
ends of the cells which are to give rise to the endoderm, either before or during or
after the invagination takes place ; the mode of origin of these spherules in Lucifer
and their position in the egg agree with what we should expect if thej belong to the
mesoderm, but the great quantity of food -material which they contain would hardly be
looked for in this case, and favours the view that they are yolk-pyramids rather than
mesoblasts.

As I examined no eggs between fig. 20 and fig. 21, the later history is uncertain, but
a reference to figs. 21, 22, 23, and 24, which are about twenty hours later than fig. 20,
shows that the region of the digestive tract of the Nauplius is marked by the presence
of a number of large polygonal masses of what appears to be food-yolk, and it seems
probable that these are the derivatives of the spherules c, of fig. 20. I was not able
to actually witness the change from fig. 11 to fig. 15, and cannot state with absolute
certainty that the spherules c divide into a central and a peripheral portion. Fig. 1 5
seems to indicate that this is the case, but in the absence of direct observation of the
change, it is possible that the two cells which in fig. 15 lie below the cells c, are the
ones which were at its sides in fig. 1 1 .

If each of the cells c gives rise to a blastoderm cell, we shoiild expect to find two
more cells in fig. 15 than in fig. 18, but the number is the same. This is hardly a
safe guide, however, for while the drawings are careful copies from Nature, they are
not from the same egg, and the cells are so wedged together that vertical sections in

A STUDY IN MORPHOLOGY. 71

different planes would not intersect the same number in all cases, and there may have
been two more in fig. 15 than in fig. 13.

I think, then, that the facts indicate that c of fig. 11 is a yolk-pyramid, rather than
a primary mesoblast, and that after it divides into two, as in fig. 13, each part gives
rise to a central portion c, and a peripheral endoderm cell.

If we accept this view and regard the cell c as a yolk-pyramid, two views as to the
relationship between the egg of Lucifer and an ordinary Crustacean egg at once
suggest themselves.

We may hold that Lucifer presents the primitive or ancestral form of segmentation,
of which centrolycethal segmentation is a secondary modification. In this case we may
suppose that as the supply of food-material gradually increased, new food-bearing cells
or yolk-pyramids were added until all the cells were included, and the segmentation
cavity was entirely filled and obliterated by them.

According to the other view, we may hold that the segmentation of the Lucifer egg
is a secondary modification, which has been brought about by the gradual reduction of
the amount of food-material, and its restriction, at last, to a single one of the cells of
the segmenting egg.

There does not seem to be much difficulty in deciding which of these views is most
satisfactory and probable. Lucifer is undoubtedly a very primitive Malacostracan, but
it can hardly be regarded as a primitive Crustacean ; and the occurrence of perfectly
centrolycethic segmentation in the Copopods, Phyllopods, Amphipods, and Isopods, as
well as in the Decapods — forms below as well as forms above Lucifer — forbids us to
believe that the egg of Lucifer is ancestral, or the unmodified descendant of an ances-
tral type of egg ; and we must therefore believe that the egg of Lucifer has been
simplified by the loss of the greater part of its food-yolk.

A change of this kind is not without a parallel, and I have shown (" The
Acquisition and Loss of a Food- Yolk in Molluscan Eggs," ' Studies from the
Biological Laboratory of the Johns Hopkins University/ vol. i., part iv.) that the
resemblance between the segmenting egg of the Oyster and a Molluscan egg with a
food-yolk can only be explained by the supposition that the Lamellibranchs have
inherited a rudimentary food-yolk which was functional at some past time, and that
the assumption gives an explanation of all the peculiarities of oyster segmentation.

If we accept this view, and regard the egg of Lucifer as simplified by secondary
change, it is extremely instructive to note that the loss of a food-yolk has brought it
back to a type of segmentation which is directly comparable with that of ordinary
Metazoan eggs, and we must therefore believe that a segmentation cavity is poten-
tially present in all centrolycethic eggs, or else that the segmentation cavity of the
egg of Lucifer is not homologous with that of ordinary eggs.

72 MR. W. K. BROOKS ON LUCIFER:

III.— GENERAL ACCOUNT OF THE METAMORPHOSIS OF LUCIFER.

The most instructive method of studying the metamorphosis of Lucifer is to trace
each part of the body through the series of changes which it undergoes from its first
appearance until it assumes the adult form ; but as this method of comparing the
successive stages in the development of each organ necessarily involves references to
other organs, it seems best to give first a general account of the whole structure of the
larva at each stage of development, and afterwards to go over the same ground more
rapidly in a different way, and to trace the history of each appendage.

The egg Nauplius.

About thirty hours after oviposition the eye spot and appendages of the Nauplius
became visible inside the egg-shell, as shown in a ventral view in Plate 2, fig. 21,
and in a dorsal view in fig. 22. If the egg-shell is torn at this stage the embryo
escapes, and swims about quite vigorously for a short time, but soon dies. The various
parts of the body are much better shown in the swimming embryo than while it is
contained in the egg, and I therefore give, for comparison with figs. 21 and 22, a
dorsal view (fig. 23) and a ventral view (fig. 24) of an embryo which has thus been
set free.

Fig. 23 shows an embryo of exactly the same age as those in figs. 21 and 22, while
fig. 24 was drawn from an embryo a few hours older. The difference in the outline of
the body is not due to this difference in age, however, but to a slight change in the
point of view. In all four figures the letter e marks the anterior end of the body, and
fig. 22 is a view directly opposite to fig. 21. Fig. 23 is in the same position as fig. 22,
but the embryo shown in fig. 24 was in such a position that more of the anterior
surface and less of the posterior surface was visible than in the other figures.

On the median line of the ventral surface the labrurn (figs. 21 and 24, L) is very
conspicuous at the anterior end of the body, and behind it there is a double row of
four pairs of bud-like eminences, arranged in a longitudinal series. The first pair
(figs. 21 and 24, //) are much larger than the others, and the depression which separates
them on the median line is less marked than it is in the three pairs which lie behind.
It is rather difficult to decide with certainty what this pair of buds becomes, but in
the larva which METSCHNICKOFF studied the changes were more gradual than they
are in Lucifer, and he was therefore able to trace their history more satisfactorily,
and to show that they become the rnetastoma. Their position with reference to other
parts indicates that they have the same history here, and that the other three pairs of
buds are the first and second maxilla? and the first pair of maxillipeds (Mx. 1, MX. 2,
and M'p. 1).

Three pairs of much larger appendages are folded down on to the sides of the body,
within the egg ; and when the embryo is set free they are seen to be the first antenna?

A STUDY IN MORPHOLOGY. 73

(A), the second antennas (An], and the mandibles (J/). They are not divided into
joints or rings, although the second antennae and the mandibles are bi ram cms, and
consist of a basal portion or protopodite, an expedite, and an endopodite. All three
pairs have hairs projecting from their tips, and these lengthen considerably within a
few minutes after the embryo is freed from the egg. The first antennae are nearly as
long as the second, and both pairs, as well as the mandibles, are organs of locomotion,
to row the animal through the water. The motions of the larva are very erratic and
violent, and consist of a series of quick leaps produced by vigorous backward strokes
of the appendages.

The outline of the body will be understood by a reference to the figures. When
the second maxillae are in the centre of the field of view, as in fig. 21, the outline is
pear-shaped, with the broad end of the pear at the posterior end of the body ; but
when the metastoma is in the centre this is reversed, and the broad end is in front.
This difference is due to the fact that the dorsal region is much wider than the labrum
and series of buds, which together form a ridge along the ventral surface.

In a dorsal view the simple eye (Oc) is seen as a black spot on the middle line, near
the anterior end of the body. It did not show any traces of a division into halves at
any stage of development which was observed.

The ocellus lies upon a large rounded granular body, which is imperfectly divided
into halves by a notch upon its posterior margin. This body consists of the fused
cerebral ganglia.

The dorsal portion of the posterior region of the body is swollen and rounded, as
shown in figs. 21 and 23 ; and near its lateral margins there are a pair of small, but
very conspicuous, dark pigment-spots («'), which might easily be mistaken at this stage
for ocelli, since they have almost exactly the same size and colour. These two pigment-
spots are very conspicuous during all the early stages of the metamorphosis, and their
position during the later stages (figs. 25, 26, 27, 34, 35, and 47, p) shows that the
portion of the Nauplius body which bears them becomes the thoracic, not the abdominal,
region of the adult.

In the interior of the enlarged posterior portion of the body there is a huge mass of
polygonal highly-refractive bodies, which appear to constitute a food-yolk, and which
surround the digestive tract of the embryo. I have already given my reasons for
believing that those bodies are derived from the spherule which becomes pushed into
the segmentation cavity during the early stages of development. If this is their
origin they must increase in size between the stage shown in fig. 20 and that shown
in fig. 21. This is not at all an unusual occurrence, and in the fresh -water Pulmonates
the yolk-spherules which surround the digestive tract continue to grow until a very
advanced stage of development. I found so few eggs at this stage that I was afraid
to sacrifice any of them by attempting to study their internal structure under pressure,
and I am not able to give an account of the digestive tract or of the other internal
organs.

MDCCCLXXXII. L

74 MR. W. K. BROOKS ON LUCIFER:

When the embryo is set free from the egg it is seen to be inclosed by a delicate
cuticle, which is shown, around the antennae, in figs. 23 and 24. It is soon stripped
off by the vigorous movements of the larva, and in fig. 24 it has been torn from all
the appendages except the first antennae (A).

In a dorsal view a number of muscular fibres are seen to extend outwards and
forwards from the median line of the body to the basal joints of the antennae.

The posterior end of the body is not notched, the anus is absent, and there is no
trace of the telson or of the carapace.

The first free Nauplius stuye.

About thirty-six hours after oviposition the larva escapes from the egg as a Nauplius,
T~oou ulcn l°ng> which is shown in side view in Plate 3, fig. 25. There is now no
difficulty in keeping it alive and rearing it, and it swims very actively by vigorous
strokes of its two pairs of antennae. Its movements are very characteristic, and much
like those of a Copepod or Cirrhiped Nauplius.

The most important differences between it and the egg Nauplius are the segmenta-
tion of the locomotor appendages, the lengthening of their hairs, the increased size
and dendritic form of the pigment-spots (p), and the appearance of the telson (77), as
a projecting fold furnished with two pairs of short spines or hairs, in the ventral
surface of the posterior end of the body.

As regards the more minute structure of the appendages, the first antennas (fig. 25, .4)
are five jointed, and the hairs, which are more than half as long as the limb, are
borne on the terminal joint.

The second antenna consists of a two-jointed basal portion or protopodite which
carries two rami, one of which (fig. 25 ex), is obscurely divided into three nearly equal
joints, while the other (fig. 25, en), is divided into eight very distinctly marked joints.
Both at this stage and later the appendage possesses considerable power of rotation,
and sometimes the branch ex, and sometimes the branch en, is on the outer surface.
It is therefore very difficult to decide from an examination of this appendage alone
which branch is the exopodite and which the endopodite ; but, as I shall show further
on, a comparison with other appendages at a later stage indicates that the eight-
jointed ram us is the endopodite, although the limb is frequently, and perhaps
generally, carried in a position which brings this branch on to the outside. At this
stage the locomotor hairs of both branches are confined to the tips of the terminal
joints. The first and second joints of the endopodite are quite short, while the other
six are longer and nearly equal in length.

The mandible consists of a short unjointed basal segment, which carries a one-
jointed endopodite, and an obscurely three-jointed exopodite. Each branch carries
three hairs, which are somewhat longer than the limb, and the entire length of the
appendage, including the hairs, is about equal to the length of the first or second

A STUDY TN MORPHOLOGY. 75

antenna, without its hairs. There are no cutting blades or hooks upon the basal
joints of either pair of antennte or the mandibles.

The labrum (L) is somewhat larger and more prominent than it was at the stage
before, and the anus is still absent.

The second free Nauplius stage or ??u-?a-Nauplius.

In about twelve or fourteen hours the Nauplius sheds its skin and assumes the
form shown in Plate 3, fig. 26. From the prominence of the region of the hind body,
and the presence of a carapace, GLAUS has distinguished this stage of development, in
allied forms, by the name of meia,- Nauplius.

I did not actually witness the change, and am not sure of the exact length of the
first free Nauplius stage, bnt it is not more than eighteen, and probably no more than
twelve hours long. A Nauplius which had hatched from the egg some time during
the latter part of Monday night was placed, alone, in a watch-glass of sea-water, and
changed into the one from which fig. 26 was drawn before 9 P.M. on Tuesday evening.

The differences between this and the preceding stage are sufficiently great to
attract the attention at first sight. The length, as measured from the ocellus to the
posterior end of the body, has increased from y-jjol) hich to J-Q-Q-Q inch. The labrum (L)
is longer and more prominent. The first antennae (A) are unjointed, and the joints of
the second antennae (An) and mandibles (M) are almost absent.

The hairs at the tips of the endopodites of the second antenna? and mandibles (en)
are irregularly plumose, and a long slender slightly curved hair is carried by each of
the larger joints of the endopodite of the second antennae.

On the inner posterior edge of the basal joint of the mandible, a short stout curved
hook or blade has made its appearance. The four pairs of buds on the ventral surface,
posterior to the labrum, are in the same condition as before, but the telson (T) is quite
prominent, notched or forked, and furnished with two pairs of short stout spines, the
inner pair being much longer than the outer. A well-marked fold (c) of the surface
of the body now marks the posterior and the lateral edges of the carapace, but this
line is not continued on to the anterior end of the body, and the posterior edge is not
yet raised or separated from the hind body as it is, according to METSCHINCKOFF, in
the last Nauplius stage of Eupliausia.

The pigment-spots (p) are drawn out in such a way as to surround a large rectan-
gular area, at the posterior end of the carapace, and in the region where the heart
is placed at the next stage.

The digestive tract is now visible in a side view. The oesophagus (a?) runs upwards
and forwards from the mouth, situated under the overhanging tip of the labrum, and
then bends backwards and upwards to open into the floor of the stomach (s) ; the side
walls and top of the stomach could be made out without difficulty, but I was not able
to decide whether its ventral wall is complete or not. It is divided by a fold or flap

L 2

6 MR. W. K. BROOKS ON LUCIFER:

in its dorsal wall into a small rounded anterior chamber, into which the oesophagus
opens, and a longer posterior chamber, with its dorsal wall very thick, which gives
rise at its posterior end to the intestine (i). The greater part of the anterior chamber
lies in front of the cesophageal opening. On each side of the stomach there is a group
of polygonal yolk-cells (/), which are by no means as conspicuous as they were at an
earlier stage. The intestine is small, with thin walls, and it follows the dorsal curva-
ture of the body to the anus, which was visible in a ventral view just in front of the
spines of the telson, at the point marked (a) in fig. 2G. The cerebral ganglia (go),
and the ocellus (oc), are still visible, and underneath the stomach there is an elongated
granular body (n), obscurely divided into segments, which is, without doubt, the rudi-
mentary ventral nervous system.

As it was necessary to keep this larva alive I did not dare to use much pressure
whilst examining it, and was therefore unable to make a very thorough study of its
internal structure.

The first Protozoea stage.

On Tuesday evening, September 28th, at 9.30 P.M., the Nauplius which has just
been described was placed alone in a watch-glass of sea- water, and at 9 A.M. on
Wednesday, the 29th, it had changed into the larva which is shown in dorsal view in
Plate 3, fig. 27. The number of segments and appendages of this larva and its
general form and proportions are like those of the Euphausia, Pcnceus, and Sergestes
larvae at the stage of 'development which GLAUS has proposed to call a Protozoea
('Crustacean System,' p. 2). The precise time when the change took place could not be
learned, but there is reason to believe that it was not much later than the middle of
the night, On September 1 4th I obtained, by dipping with a surface-net, a Protozoea,
which I studied and drew. It was of exactly the same size (T§TO °f an inch measured
from the tip of the rostrum to the bases of the spines of the telson) as the one which
moulted from the Naiyjlius, and it agreed with this in every respect except that the
free segments of the hind body, shown in fig. 27, were wanting. It hardly seems
probable that there are two stages of exactly the same size between 9.30 P.M.
and 9 A.M., and it is much more probable that the body segments do not become
distinct until some time after the moult, and as the larva had them at 9 A.M., I infer
that it was nearer the end than the beginning of the first Protozoea stage, and that
the change had taken place some hours before I examined it.

GLAUS is inclined to believe that the difference between FRITZ MULLER'S last figure
of the Nauplitis of Penceus and his first figure of the Protozoea is so great that there must
be a gap in the series of observations. The isolated Nauplius of Lucifer passes through
quite as great a change in twelve hours, and its length increases from 10900 to Y§§^J,
or mure than 100 per cent., and there does not seem to be any necessity for supposing
that FRITZ MULLER has missed a stage in order to account for the change in his
larva,

A STUDY IN MORPHOLOGY. 77

In the case of Lucifer the actual increase in size is not very great, but the carapace
becomes folded out over the body, and the thick posterior portion of the body of the
Nauplius becomes pulled out into the long free movable hind body of the Protozoea,, so
that the length is more than doubled, while the vertical thickness of the body is
correspondingly reduced. The shape of the larva when seen from ' one side will
be understood by a reference to Plate 4, fig. 35, for although this figure was drawn
from an older larva, it correctly represents a side view of fig. 27 in all essentials.

The most marked differences between the meta-NaupUus of Lucifer and the Protozoea
are due to the development of the carapace and the hind body. The carapace
(fig. 27) is horse-shoe shaped, with smooth lateral and posterior edges, and it forms
about one-half of the total length of the body. On the median line of the anterior
edge it is drawn out into a long rostrum (R), at the base of which are the cerebral
ganglia (yet) and the ocellus (Or). On the median line of the posterior edge of
the dorsal surface there is a shorter dorsal spine (ds), and at the outer angles of
the posterior edge a pair of lateral spines (Is), which are a little longer than the dorsal
one. The side view (fig. 35) shows that the sides of the carapace have folded
down on to the sides of the body, and all the appendages, except the antennas,
are almost completely covered by it. The appendages are so nearly alike in this and
the next stage that it will be most convenient to describe them together.

The stomach (s) is now divided into a pair of anterior or cephalic, and a pair
of posterior or hepatic lobes, and between the cephalic lobes a number of muscular
fibres run upwards and forwards from the oesophagus to be attached (at m) to the
carapace. The intestine is small and straight (i), but it is not of uniform character,
and is divided into a series of small enlargements separated from each other by
constricted portions.

The last of these enlai-gements is much more constant than the others, and its walls
are attached to the integument of the abdomen by a number of small muscles.

It exhibits regular pulsations, which seem to draw water into and out of the anus
(a), which is on the ventral surface of the telson.

The heart (h) is compact, short, situated near the posterior edge of the carapace,
and it gives rise to a single median and two lateral anterior arteries.

The hind body is about as long as the carapace, and it is divided into four somites
and a long unsegrnented region (al>d). The study of the appendages shows that the
four somites are those which carry the third pair of maxillipeds (Mp. 3), and the first,
second, and third thoracic somites ( T 1 , T 2, and T 3). There are no traces of appendages
on any of them. The end of the unsegmented region of the hind body forms a well-
marked flattened telson (J1), which is slightly notched on the median line, and carries
four pairs of stout spines, and one pair of very small ones. The small ones are nearest
the median line ; the third pair are the longest and largest, and the fifth pair spring
from the edges of the telson, some distance from the end.

78 MR. W. K. BROOKS ON LUCIFER:

The second Protozoea stage.

As my season's work at the sea-shore ended the day the Nauplius shown in fig. 2G
turned into the Protozoea shown in fig. 27, I was not able to trace the development
of that specimen ; but on September 14th I had captured and drawn a larva in the
same stage, and this moulted, while isolated in a watch-glass, into the second
Protozoea which is shown from above in Plate 3, fig. 34, and from the right side in
Plate 4, fig. 35.

This larva measures y§ij-o inch from the tip of the rostrum to the fork of the telson.
The appendages are like those of the first Zoea in number and structure, but there is
a well-marked difference in the shape of the body. The carapace is somewhat
elongated, its anterior edge is less perfectly rounded than before, and a pigment-spot
(fig. 34, E) represents the future compound eye.

The pouches of the stomach (s) are much more conspicuous than before, and the
oesophagus (fig. 34, cc) is visible in a dorsal view, between its anterior or cephalic
lobes. The four somites of the hind body (Mp. 3, T I , T 2, and T 3) have become short,
but there is, as yet, no trace of their appendages. The unsegmented portion of the
abdomen (abd) has increased in length, as have also the spines of the telson (T). The
two pairs of antennfe have substantially the same form that they had during the
Nauplius stage, and they are still the chief locomotor organs. The larva swims by
jerks like a Nauplius or a Copepod.

The appendages at this as well as at the preceding stage are as follows (see Plate 4,
fig. 35) : the long uniramous first antennae (A) ; the biramous second antennas (An.);
the cutting mandibles (J/); the biramous first and second maxilla3 (M.r. 1, M.c. 2) ; and
two pairs of biramous maxillipeds (Mp. 1, Mp. 2).

The first antennse consist at both stages (figs. 27, 34, and 35, A) of a long
cylindrical basal joint which carries a few short hairs, and a short pointed terminal
joint or flagellum, which ends in two long rather thick sensory hairs.

The second antenna) (figs. 27, 34, and 35, An ; and fig. 36) are the chief locomotor
organs, and are made up of a short stout two-jointed basal portion, a longer unjointed
exopodite (ex), with four long terminal swimming hairs, and a longer endopodite (en),
which is made up of two short proximal rings, and a series of six longer joints, each
of which carries one, and the terminal one four, long slender swimming hairs.

Underneath the rostrum (fig. 35, R) there is a little elevation upon which the ocellus
(Oc) is situated.

The labrum (fig. 35, L) has been carried on to the ventral surface of the body, and
its anterior angle has become produced into a short .stout, sharp spine, which is
extremely small during the first Protozoea stage.

As has been stated, the compound eye is represented at the second stage by a
pigment-spot (fig. 35, E).

The mandibles (J/), (figs. 27, 34, 35), have become reduced to cutting blades, which
are visible in a dorsal view, and all traces of the Nauplius limb have disappeared.

A STUDY IN MORPHOLOGY. 79

During the first Protozoiu stage (Plate 3, figs. 28 and 29) it lias only one denticle,
which is large and pointed, and situated at the posterior angle of the cutting edge ; Imt
at the second Protozoea stage (Plate 4, fig. 37) a number of smaller denticles have
appeared in front of the long one. The mandibles are never quite symmetrical, but
the outline of the left always differs a little from that of the right.

The external surface of the first maxilla of the first Protozoea is shown in fig. 30,
and the posterior surface of that of the second Protozoea in Plate 4, fig. 38. It
consists, at both stages, of a basal portion made up of two joints with cutting hairs
(fig. 38, i and 2) ; a two-jointed endopodite (en), with three long slender hairs ; and an
exopodite or scaphognathite (figs. 30 and 38 sc), with three long slender hairs. In
the first stage (fig. 30) the hairs of the scaphognathite are simple, but in the second
stage (fig. 38) they are plumose.

The posterior surface of the second maxilla of the first Protozoea is shown in Plate
3, fig. 31, and that of the second Protozoea in Plate 4, fig. 39. It consists of a
many-jointed basal portion (b), a two-jointed endopodite (en), and a scaphognathite or
exopodite (sc). The whole inner edge of the appendage carries short stout hairs ; the
tip of the endopodite a few somewhat longer hairs ; and the scaphognathite three
slender plumose hairs, which are much longer in the second than in the first stage.

The first maxilliped (figs. 32 and 40) is very similar to the second antenna, and
consists of a two-jointed basal portion, a four-jointed endopodite, and an unjointed
exopodite. The inner edge is set with short stout hairs, which are simple in the
first, but irregularly plumose in the second Protozoea stage. The terminal joint of the
endopodite" carries four long slender simple hairs, and the tip of the exopodite four
long straight slender hairs, which are plain in the first but regularly plumose in the
second stage.

The second maxilliped of the first Protozoea is shown in fig. 33, and that of the
second Protozoea in fig. 41. It is essentially like the first maxilliped in structure, but
much smaller, and apparently of little functional importance.

In the second stage there is a small convoluted shell gland (fig. 35, sg), which
appears to open at the base of the first maxilla ; but the constant and violent
movements of the limbs render it difficult to decide with confidence exactly what its
relation to them is, and it is possible that its opening is upon the base of the second
instead of first maxilla.

In the second Protozoea stage the two pigment-spots (/>) on the carapace become
extremely dendritic, and a pair of a,nal pigment-spots (Plate 3, fig. 34, pj>) make
their appearance on the telson on each side of the anus.

At this stage the area, when the oesophageal muscles are attached to the carapace,
is somewhat peculiarly marked by six little circles arranged in a pentagon, as shown,
highly magnified, in fig. 35o.

SO MR. W. K. BROOKS ON LUCIFER:

The hitst Protozoea stage (Erich tliina).

The change from the last stage to the next one in the series was actually observed
in several specimens, and more than fifty larvae passed through it in the laboratory.

After the moult the larva, which is shown from the ventral surface in Plate 4,
fig. 42, and in outline in fig. 42a, has the characteristics of DANA'S genus Erichthina.

Its length, from the tip of the rostrum to the end of the telson, has increased
to about T§^Q- inch, and most of the increase is in the hind body. The carapace also
is somewhat elongated (it was a little flattened by pressure in the specimen which
was drawn), and the outline of the antei'ior edge is no longer regularly curved.
At the base of the rostrum there is a slight eminence where the integument is pushed
out a little by the optic ganglion, and at the outer angle there is a much larger
eminence which is the rudimentary cornea of the compound eye. The eye itself is
now represented by a large conspicuous pigment-spot (fig. 42a, E).

The appendages have undergone extremely little change, and they are, as before, as
follows : the first antennas (A), the second antennas (An), the mandibles (M), the two
pairs of maxillas (Mx. I and MX. 2), and two pairs of maxillipeds (Mp. 1 and Mp. 2).
The second antennas are still the chief organs of locomotion.

The hind body is much longer than it was at the stage before, and it is now some-
what longer than the carapace. It now consists of nine free segments and an unseg-
mented portion (A 5, 6). The first of the free segments (fig. 42, Mp. 3) is much
narrower than any of the others, and its outer edges are marked by enlargements
which appear to be the rudimentary appendages, the third pair of maxillipeds. None
of the segments which follow it show a trace of the appendages, and the thoracic
and abdominal ganglia are not yet visible.

The four segments which follow next after the one with the bud-like processes have
rounded posterior edges, while the posterior edges of the next four are pointed. The
later history seems to show clearly that those with rounded edges are the first, second,
third, and fourth thoracic somites, and that the following ones are the first, second,
third, and fourth abdominal somites. It will be seen, then, by a comparison of this
with the earlier and later stages, that the somites of the body are all developed in
regular order, from in front backwards, but that the first abdominal somite follows
immediately after the fourth thoracic, while the fifth thoracic is never developed. At
this stage the long unsegmented region (A 5, 6), represents the fifth and sixth abdominal
segments and the telson. The two anal pigment-spots are larger than they were
during the stage before, and from this time to maturity their colour is a dirty reddish-
brown instead of black.

The "Zoea" stage (Elaphocaris stage of Sergestes.)

After the next moult, which was observed in a great number of specimens, the
larva passes into a stage which is directly comparable, so far as the appendages are

A STUDY IN MORPHOLOGY. 81

concerned, with the Elaphocaris stage of Scrgestcs, although the most conspicuous
features of the Elaphocaris larva, the long compound spines, are not present in Lucifer.
It is now about ytHfo inch long, and it is shown in a dorsal view in Plate 5, fig. 44,
and, more highly magnified, from below in Plate 4, fig. 43. In a side view (fig. 45)
it still agrees pretty^ closely with fig. 35 ; its body is carried in the same attitude,
and the antennae are still the chief organs of locomotion. The fully-developed
appendages are, as before, the first and second antennae, the mandibles, two pairs of
maxillae, and the first and second pairs of maxillipeds, but the third pair of maxillipeds,
four pairs of thoracic appendages, and the swimmerets or appendages of the sixth
abdominal somite are now present as rudimentary buds.

The compound eye (figs. 43 and 45, E] is now well advanced in development, although
there is as yet no trace of a stalk, and the cornea is simply a modified portion of the
integument of the carapace.

The carapace is longer, narrower, and more rectangular in a dorsal view than it was
at the last stage, and it makes only about one-third of the total length of the body of
the larva. Its pigment-spots are very large, dendritic, and conspicuous, but their
colour has changed from black to dark reddish-brown.

The anterior lobes of the stomach (fig. 44, s) have lengthened and approached each
other on the median line, and they now reach forwards nearly to the optic ganglia.

The appendages which were present during the Protozoea stage have essentially
the same structure now, and the differences are very slight. The number of cutting
hairs on the basal joints of the first maxilla (fig. 46) has increased ; the hairs on
its endopodite are plumose, and one of those carried by the scaphognathite is much
longer than the other two. This is the case also with the second maxilla (fig. 47),
and the hairs along its inner edge have become almost as long and slender as those at
its tip. The first maxilliped (fig. 48) is almost exactly like that of the Protozoea;
but the second (fig. 49) is much more developed, and the hairs on its exopodite are
plumose.

The hind body is now divided into its full number of segments ; that of the third
pair of maxillipeds (Mp. 3) ; the first, second, third, and fourth thoracic somites (T 1,
T 2, T3, and 7*4); and the six abdominal somites, but the telson (T) is not yet com-
pletely distinct from the last abdominal somite. The thoracic somites are shortened
and crowded together, and each of them carries a pair of bilobed buds, the rudi-
mentary thoracic appendages. These buds are crowded together in a double row on
the median line of the ventral surface of the body, and outside them is a pair of
much larger buds (figs. 43 and 45, Mp. 3), bilobed also, but pointing backwards ; the
rudimentary third pair of maxillipeds.

The future history of the larva seems to show conclusively that the inner set of
buds are, as indicated in fig. 43, the first four pairs of thoracic limbs or pereiopods.
The side view (Plate 5, fig. 45) shows that there is no other pair in front of or

MDCCCLXXX1I. M

ME. W. K. BROOKS ON LUCIFER:

behind them, and the fifth thoracic somite is entirely wanting, nor are its appendages
present at any stage in the development of Lucifer.

The abdomen is much longer than it was at the last stage, and all its segments
(fig. 43, A 1, A 6) are present, although the last one (A 6) and the telson (T) are not yet
entirely separated.

The ventral surface of the sixth abdominal somite is armed with a pair of long
stout spines over the base of the swimmeret, or sixth abdominal appendage, which
is shown in fig. 43 as a long, bilobed pouch or bud, which reaches nearly to the tip
of the telson. The third, fourth, and fifth abdominal somites carry, close to the anterior
edge of the ventral surface, irregular groups of reddish-brown pigment -spots, which
do not seem to be present in all specimens. The thoracic spots (fig. 44) and the
anal spots (fig. 45) are usually a little more red than before, but they are nearly
black in some specimens. The abdominal ganglia, which could not be distinctly made
out in the last Protozoea, are now very conspicuous, as shown in the ventral view
(fig. 43). They lie near the posterior edges of the somites, and their halves are
united in the median line, although the commissures between the ganglia are quite
widely sepai'ated.

The spines on the telson have lengthened, but their number, arrangement, and
relative size is the same as before. Their pi-oximal ends from the base about half-way
to the tip are marked by fine serrations, which appear to be short hairs, which have
not been perfectly extended.

Schizopod or Sceletina stage (Acanthosorna of Sergestes).

Up to this time the mode of locomotion has been by means of short, jerking Naup-
lius leaps, and the two pairs of antennae have been, as they were when the larva left
the egg, the chief organs of locomotion. The structure of these appendages has
remained extremely constant through all the moults, but they now change their
character entirely, and lose their locomotor function.

The change which is undergone by the larva at the end of the Zoca series is very
much greater than it has been at any preceding moult, except that between the Naitp-
lius and the first Protozoea, and in some respects it is even greater than it was at that
time. After the moult it is a Schizopod (Plate 6, fig. 50), about y^g^ inch long,
with seven pairs of long jointed biramous swimming feet, fringed with long slender
hairs. The swimmerets are also present as functional appendages, with long fringing
hairs.

This stage differs from those which have gone before in this, that it persists with
slight change for several moults, while there has been considerable change at each of
the preceding moults. It is shown from below in fig. 50, as it appears immediately
after the moult which follows the stage shown in fig. 43.

The figure was drawn from a Zoca which was captured at the surface of the ocean,

A STUDY IN MORPHOLOGY.

carefully examined and compared with fig. 43, and found to agree with it exactly.
It was then placed alone in a small beaker of sea-water. The next day it was found
to be moulting, and the drawing (fig. 50) was made from it immediately after the
completion of the moult. Other specimens, like fig. 50, were kept until they changed
their skins, and assumed a form a little larger than fig. 50, but similar to it in all
respects except that the abdominal appendages were now present as small buds.
Some of these were kept until they changed into larva) like the one which is shown,
less highly magnified, from the side, in fig. 54. The abdominal appendages were now
quite long, but still rudimentary, and the general form of the larva from above or
below, as well as the form, number, and arrangement of the thoracic appendages and
mouth parts, was like fig. 50.

When seen from above or below (fig. 50) the carapace has nearly the same shape
that it had during the Zoea stages, but it now makes less than one-third of the
total length of the body, and a side view (fig. 54) shows that it is now only a little
deeper than the body, so that the basal joints of the thoracic limbs and maxillipeds
are exposed below its inferior border. The posterior dorsal spine and the two postero-
lateral spines have disappeared, and a pair of long antero-lateral spines (fig. 54, .s),
nearly half as long as the rostrum, have made their appearance underneath the eyes.
The rostrum (fig. 50, R) has the same shape and about the same relative length as
before, and the ocellus (Oc) is still present at its base.

The compound eye (E) is mounted upon a movable stalk, which is quite short during
the first Schizopod stage, but it soon lengthens, as shown in fig. 55, which is a dorsal
view of the anterior end of the carapace of the larvoe shown in fig. 54.

The first antenna has undergone more change at this than at all the previous
moults together. It is now about as long as the carapace, and each of the two long
cylindrical joints (fig. 50), which make up its basal portion, carries on its inner edge
three long slender two-jointed delicately plumose hairs. The base of the proximal
joint is swollen and carries a small hook-like process on its inner edge. The two
long sensory hairs have disappeared from the tip, which is unsegmented, pointed, and
ends in a bunch of short hairs. This appendage changes slightly with each moult,
and in the third Schizopod stage (fig. 54) the distal half of the proximal joint (fig. 56)
has separated from the proximal joint, so that the shaft is made up of three instead
of two portions. The hook is still present on the swollen base of the first joint, and
behind it the otocyst (e] has made its appearance. The terminal joint or flagellum has
now lengthened, and it carries three long sensory hairs which spring from about the
middle of its outer surface.

The changes which the second pair of antenna undergo at this moult are even
greater than those which take place in the first pair. Their locomotor function is
lost ; the long swimming hairs have disappeared ; and in the first Schizopod stage
(fig. 50) the appendage is quite rudimentary, unjointed, less than one-half as long as
the first antenna, and divided into an exopodite and an endopodite which are nearly

M 2

84 ME. W. K. BROOKS ON LUCIFER:

equal in length, although even at this stage the endopodite is a little the longest.
Each ramus ends with a pair of very short hairs.

The appendage now changes with each moult, and in the third Schizopod stage
(fig. 54) the exopodite has become a scale (fig. 57, ex) while the endopodite (en) has
elongated, and now forms a seven-jointed flagellum, about as long as the first
antennse or the carapace. The basal joint (fig. 57, b) is thick and swollen, the two
proximal joints of the flagellum (2 and 3) are short ; the next (4) long, and the other
four about equal in length, and about half as long as the joint (4).

Through all the Schizopod stages the structure of the labrum (L) is about as it was
in the Protozoea and Zoea, and its interior angle is still produced into a short stout
sharp spine.

The mandibles are cutting jaws with no trace of a palpus, and at the first Schizopod
stage (fig. 51) the denticles are numerous and of nearly uniform size. In the last
Schizopod stage (fig. 58) a second set of denticles has appeared on the outer surface
of the blade a short distance from the cutting edge.

The first maxilla (fig. 52) is very much like that of the Protozoea and Zoea, but the
cutting hairs upon the two basal joints (1 and 2) are more numerous, and a small
slender plumose hair has appeared near the edge of each joint. The scaphognathite
is small and has only two hairs, which are less regularly plumose than before.

The scaphognathite of the second maxilla (fig. 53, sc) is now rudimentary and has
no hairs. The hairs on the inner edge of the appendage are shorter than they were
during the Zoea stage, and all of them are plumose and about equal in length.

The first maxilliped (fig. 50, Mp. 1) has not changed very much, although its joints
are nearly absent. The exopodite is about as long as the endopodite, and all the hairs
on the appendage are short and plumose.

The second and third maxillipeds and the four pairs of thoracic appendages are well
developed, as a series of long biramous or Schizopod feet, which are essentially alike in
form and structure, and, with the telson and swimrnerets, now form the locomotor
apparatus of the larva, which no longer swims by jerks but darts through the water
with great rapidity, and is able to offer considerable resistance to the suction of a
dipping tube. Each swimming foot consists of a two-jointed basal portion or protopo-
dite, a long four-jointed endopodite, and a much shorter exopodite. The exopodite is
flat, pointed, and its outer or distal half is marked by a series of six pairs of notches,
or annulations, close together. The terminal joint carries a pair of long slender
unplumose hairs, and a pair of similar hairs springs from each annulation, so that there
are fourteen hairs in all on each exopodite, arranged so as to form a large fan -shaped
paddle at the tip of the limb. The terminal joint of the endopodite is much shorter
than the others, and it carries six long plumose hairs. The first appendage in this
series, the second maxilliped (fig. 59, Mp. 2), is somewhat rudimentary: the endopodite
is scarcely longer than the exopodite, and its hairs are short. The next or third

A STUDY IN MORPHOLOGY. 85

maxilliped (Mp. 3) is more like those which follow, but its hairs are shorter. The
first, second, and third pereiopods are about equal in length, and they have the typical
structure which has just been described ; but the endopodite of the fourth (Pr. 4),
like that of the second maxilliped, is shorter than the exopodite, although its hairs are
very long.

At the last Schizopod stage (fig. 54) the series of limbs, shown from above in
fig. 59, is about as it is in the first stage, but the hairs on the endopodites of all the
appendages, except the last, are short. A comparison of one of these appendages
with the second antenna of the Nauplius or Protozoea or Zoea shows great similarity,
and I am therefore disposed to believe that the long jointed ramus of the antenna is
homologous with the long ramus of the thoracic limb, and consequently the
endopodite.

The abdomen is very much longer in proportion to the carapace than it was at the
"Zoea" stage, and a comparison of figs. 50 and 54 with fig. 43 will show that it has
become flattened from side to side, while its vertical thickness has greatly increased.
All six somites are distinct, but at the first Schizopod stage there are no traces of any
abdominal feet except the swirnnierets, which are large and perfect. In the second
Schizopod stage the first five pairs of pleopods are represented by short buds, and in
the last Schizopod stage (fig. 54) they have nearly or quite their full size, but are still
rudimentary.

The posterior edge of the ventral surface of each abdominal somite carries a couple
of spines (fig. 50) pointing backwards. They are small on all the somites except the
last, and they appear to correspond to those which, from their great size, have given the
name Acanthosoma to the larva of Seryestes at the same stage of development. The
sixth abdominal somite also has a small median dorsal spine.

The telson (T) is movable, greatly elongated, three times as long as wide, and its
spines have become very small, although in number, arrangement, and relative size
they agree with those of the Zoea and Protozoea.

The sixth pleopod or swimmeret consists of a short thick basal joint, a long flat
exopodite which is serrated along its inner edge and free extremity, but smooth
along its outer edge ; and a flat endopodite serrated on both sides. Each serration
carries a long slightly curved plumose hair, and the outer edge of the exopodite has a
small tooth at its outer end. From the base to the tooth the outer border is nearly
straight and parallel to the inner border, but the end of the appendage is prolonged
into a rounded tip which reaches beyond the tooth. In the first Schizopod stage there
are eight hairs on the inner border and four on the end of the exopodite, or twelve in
all ; and there are eight hairs on the endopodites, but the number of serrations and
hairs increases rapidly with each moult, on each division of the limb, and they are
much more numerous in the last Schizopod stage, as shown in fig. 54.

A large reddish-brown pigment-spot (fig. 54, p) has now appeared on each side of

WAT

&/P

/CO

80 ME, W. K. BROOKS ON LUCIFER:

the fourth abdominal segment, and the anal spots are large, with a dull red tinge.
The spots on the carapace disappear at the end of the Zoca series.

The Mastigopus stage.

After the next moult the larva (Plate 7, fig. 60) assumes a form which is essentially
like that of the adult, but with numerous slight differences, the most important of
which are the shortness of the nagellum of the first antenna and the absence of
the neck or elongation of the carapace. In these respects, as well as in the number,
character, and relative size of the appendages, it now agrees very closely with the young
Sergestes or Mastigopus.

The size of the thorax is reduced, while the abdomen has grown larger and longer.
The exopodites of the maxillipeds and first three pairs of pereiopods have disappeared,
together with every trace of the fourth pereiopod. The abdominal appendages are
perfect ; the first is made up of an elongated basal joint, which carries a single terminal
branch of about the same length as the basal joint, but pointed and fringed with long
slender swimming hairs. The four appendages which follow are each furnished with
two terminal branches instead of one, but are similar in other respects. The larva now
sheds its skin several times, and grows with each moult ; but the process of change
into the adult is, with the exception of the elongation to form the neck, simply a
process of growth, as the appendages and somites ah1 have essentially their adult
character.

A larva about one-fifth of an inch long, two moults after the last Schizopod stage, is
shown from the side, magnified about fifty diameters, in Plate 7, fig. GO. The first
antenna (A) is a little more than twice as long as the eye-stalk, and consists of a stout
three-jointed basal portion, which forms about two-thirds of the total length of the
appendage, and a short, thin, two-jointed flagellum. The scale (ex) of the second
antenna is only a little longer than the eye, while the flagellum (en) is more than half
as long as the body of the animal, measured from rostrum to telson, and is made up of
thirteen small joints and two thicker basal joints.

The carapace has elongated considerably, and the neck (n) makes nearly half its
length. The anterior end of the carapace has a dorsal rostrum (R), two much shorter
lateral spines (Is), and a very small spine on each side close to the anterior edge and
about half way between the rostrum and the lateral spine. The cephalic lobes of the
stomach extend into the neck, and reach nearly to the basis of the eye-stalks. The
coiled autennal gland (g) has made its appearance. The carapace proper (c) has a pair
of anterior spines, but none on its posterior margin. The labrum (L) has a much
greater relative size than it had during the Schizopod stages, but its spine disappears
at the end of the last Schizopod stage. The mouth parts and thoracic limbs have
their adult character, and will be noticed at length in the description of the adult.
A reddish -brown pigment-spot has now appeared between the bases of the eye-stalks;

A STUDY IN MORPHOLOGY. 87

another at the base of the telson ; and the dorsal surfaces of the fifth, fourth, and
sometimes the third abdominal somites are irregularly marked, near their posterior
edges, by patches of the same colour. The anal pigment-spots are of a dirty red
colour.

The Lucifer stage.

The specimen from which fig. 61 was drawn was a little more than half an inch long,
or about half as large as an adult specimen. It differs in several particulars, besides
size, from an adult male, but in all respects except size and the presence of reproduc-
tive organs it is exactly like a mature female. Its appendages are like those which
are shown in figs. 63 to 70, although these were drawn from an adult female specimen.

The adult structure of our American species has been described by FAXON (' Studies
from the Biological Laboratory of the Johns Hopkins University,' vol. i., part iii.) ;
but as he had only a single male specimen, which had been preserved in alcohol, his
account was necessarily somewhat incomplete.

The first antenna (Plate 7, fig. 61, and Plate 8, fig. 66, A) is about as long as the
carapace and neck, and it is divided into two nearly equal portions, the base (fig. 66, i)
and the flagellum (fig. 66, 2). The base is divided into three joints, the first about as
long or a little longer than the eye, the second much shorter, and the third still
shorter. The large ear occupies the centre of the proximal end of the first joint. On
the outer end of the first joint and on the second there is a row of six short, equal,
plumose hairs, three on each joint. The flagellum is made up of ten joints ; the first
and second are thicker than the others, and the first carries two and the second three
sensory hairs. The terminal joint of the flagellum is much longer than the other, and
carries a few very short hairs at its tip.

The second antenna (figs. 61 and 66, An] is, in the fully-grown specimen, almost
twice as long as the first, and nearly or quite as long as the body. It consists of a
very short basal joint (fig. 66, a), which carries the scale (ex) and the flagellum (en).
The scale is somewhat longer than the eye, flat and narrow, and its inner edge carries
nine and its tip three long, slender, plumose haii-s, which are about half as long as the
scale itself. The flagellum tapers gradually from the base to the tip, and is made up
of twenty-four joints, each of which carries a pair of very short hairs. The joints at
the tip of the flagellum are a little longer than those at the base. The living animal
usually carries these appendages extended before it, and diverging a little at their tips.
It occasionally throws them back along the sides of the body, but only for an instant
at a time.

The eye-stalk tapers gradually from the base to the tip, and there is no abrupt
distinction between the stalk and the eye proper, as there appears to be in other
species. The length of the eye, with its stalk, is a little less than that of the true
carapace.

The neck makes a little more than three-fifths of the total length of the carapace,

MR. W. K. BROOKS ON LUCIFER:

and its vertical diameter is more than half that of the thorax. It has a median dorsal
rostrum (fig. 61, .R), which is much smaller relatively in the adult than in the young,
and two antero-lateral spines (Is). About half-way between the rostrum and the
lateral spine the anterior edge of the neck has an extremely minute spine on each
side, as in the younger stage last described. The cerebral ganglia (cy) occupy the
ventral half of the anterior end of the neck, and the long commissures can be seen
at co. running back to join the ventral nervous system. The cephalic lobes of the
stomach (s) and the antennary gland (g) occupy the dorsal portion of the neck.

The true carapace (c) does not reach down on to the sides of the body as far as the
basal joints of the thoracic limbs and mouth parts, and both these and their ganglia
(figs. 75 and 76, trf) are visible below its free edge. Its edges are smooth, but there
is a small spine at its anterior end.

The labrum (fig. 61, L) is massive and prominent, but there is no trace of a spine.

The inner surface of the mandible (fig. 62) is marked by a number of parallel
ridges, one for each denticle; and there is a second, and a faint trace of a third, series
of denticles on the outer surface (fig. 63). There is no trace of a mandibular palpus.

The scaphognathite of the first maxilla (fig. 64) has disappeared, the endopodite is
rudimentary, while the second basal joint is very much larger than the first, and carries
about fifteen stout short hairs arranged in three rows. The first joint has four much
larger unequal hairs, which are serrated. The outer edge of the first and both edges
of the second joint carry a single delicate plumose hair each. Fig. 65 shows the
inner surface of the second basal joint.

The second maxilla (Plate 8, fig. 67) is more like that of the larva. There is a
three-jointed inner portion with short stitY hairs, and an extremely large scaphognathite
(fig. 67, sc), which is long and narrow, and united to the body of the appendage by a
very narrow stalk. The outer end carries three rather stiff, short, plumose hairs, and
five similar but somewhat longer hairs arise from the inner surface between the oxiter
end and the area of attachment. The inner end carries four plumose hairs, three of
which are almost as long as the scaphognathite itself, while the fourth appeared to be
broken off in the four specimens which I dissected out.

The first rnaxilliped is a short, stout, two-jointed appendage (Plate 8, fig. 68),
convex on its outer but flat on its inner surface, and fringed with short, stout, plumose
hairs.

The second maxilliped (Plate 7, fig. 61, Mp. 2, and Plate 8, fig. 70) is a long
jointed limb, bent into a knee, and formed of six joints. It is fringed by long plumose
hairs, which, on all the joints except the first and second, are arranged in a single row.
The first and shortest joint has no hairs ; the next, or second, has one row of five and
one row of three ; the next, or third, has six hairs ; the next, or fourth, and the fifth
have ten each ; and the terminal joint has six.

The next or third maxilliped (fig. 61, Mp. 3) is a long, slender, six-jointed limb, with
a double row of short hairs.

A STUDY IN MORPHOLOGY. 89

The first pereiopod (fig. 01, Pr. 1) is four-jointed, and shorter than the last
maxilliped.

The second and third pereiopods (Pr. 2 and Pr. 3) are nearly equal, and twice as
long as the first ; they are four-jointed, have a double row of small hairs along the
anterior edge, and the last ends in a small curved hairy claw.

They exhibit no trace of gills or of endopodites, and there is no stump to indicate the
position of the fourth pereiopod, which disappeared at the end of the Schizopod period.

The first abdominal appendage of immature specimens or of mature females
(Plate 9, fig. 74, PL 1) is made up of a thick basal portion, which is unjomted
in young specimens but two-jointed in mature ones, and a pointed annulated terminal
portion which is fringed with swimming hairs. In the nearly grown but immature
male (fig. 76) there is a little bud or projection (a) near the base of the anterior
surface of the long basal joint. In the sexually mature male (fig. 75) this bud has
become the clasping organ which has been described by MILNE-EDWARDS, DANA,
SEMPER, DOHRN, GLAUS, FAXON, and others ; and another smaller process or tooth
has appeared upon the distal one of the two joints into which the base of the
limb has now divided.

The second, third, fourth, and fifth pleopods consist, in the young of both sexes,
and in the mature females, of a long unjointed basal portion and two hairy terminal
branches. In the adult male the second pleopod has a third and smaller terminal
branch, as GLAUS has pointed out (Zeit. f. Wiss. Zool., xiii., 434).

The first, second, third, fourth, and fifth abdominal somites end below in short spines,
and they are all about equal in length, except the fifth which is nearly twice as long as
any of the others. It has a median dorsal spine on its posterior edge, and the very
young specimens also have a pair of postero -lateral spines, as shown in Plate 7,
fig. GO. In older specimens this pair of spines disappears, as shown in Plate 9, fig. 72,
and in the adult female the somite undergoes no further change. When the male
reaches sexual maturity, however, the lower edge of the somite becomes produced, as
described by DANA, on each side into the hooks shown in fig. 73. In our species
the smaller one of these hooks is near the middle of the somite, and the larger
one about halfway between it and the posterior edge.

As shown in figures 72 and 73, the telson of an adult specimen is only about half
as long as the swimmerets. The tip of the telson of an adult female is shown from
above in fig. 71.

In the male the telson becomes somewhat bent (fig. 73, T) as maturity is reached,
and a rounded anal papilla becomes developed in its lower surface, while the telson of
the adult female remains like that of immature specimens of both sexes.

The exopodite of mature specimens usually has about twenty hairs, and the endopodite
sixteen. The exopodite is longer and wider than the endopodite, and it is alike in
both sexes until maturity is reached, when it becomes somewhat modified in the male,

MDCCOLXXXII. N

90 MR. W. K. BROOKS ON LUCIFER:

This sexual difference has been pointed out by DOHRN (Zeit. Zool., xxi., 1871, p. 358),
but it seems to have escaped the notice of all other observers.

In the young and in the mature female (fig. 72) the rounded tip projects beyond the
tooth («), but as the male approaches maturity the outer edge lengthens, thus pushing
the tooth out, as shown in fig. 73, until the end of the appendage becomes square
instead of rounded. It is extremely interesting to notice that in Lucifer, as in so
many other animals, the adult female is infantile in all the secondary points of
difference from the male.

General view of the metamorphosis of Lucifer.

A review of the facts which have been described in this section indicates that some
of the changes are much more significant than others, since the number of moults is
much greater than the number of distinct larval type.

The meia-NaupUus is obviously a Nauplius with the rudiments of structures which
are to appear after the moult, and it must therefore be regarded as a Nauplius
prepared for the change into a Protozoca, rather than a distinct stage of development.

There is no such break between the first Protozoca and the last Zoca as there
is between the first Protozoca and the Nauplius. The rudimentary pereiopods and
swimmerets of the so-called Zoea are nothing but a preparation for the next stage of
development, and the supposed necessity for finding a stage which can be directly
compared with the Zoca of the higher Decapods does not justify us in making two
larval types out of the unbroken series of Protozoca and Zoca forms.

It is obvious that the three Schizopod stages are modifications of a single larval
type, and the presence of rudimentary pleopods in the second and third stages must
be regarded as a preparation for the next stage of development.

There is no abrupt break between the so-called Mastigopus and the young Lucifer
when it is a little older and the neck has appeared.

On the other hand, there is a real break between the Nauplius and the Protozoca,
and the change from one to the other is accompanied by profound structural changes.
This is the case also with the transition from the Zoca to the Schizopod stage ; and
with that from the Schizopod stage to the young Lucifer stage. The same thing is
true to a lesser degree of the change from the immature Lucifer to the adult male.

The metamorphosis may then be divided into the following well-marked stages, each
of which except the last, and in all probability the last also, persists through more
than one moult :—

1. A Nauplius stage.

2. A Protozoea stage.

3. A Schizopod stage.

4. An immature Lucifer stage, which persists in the female.

5. An adult male stage.

A STUDY TX MORPHOLOGY. PI

If we neglect the features which, at the end of each stage, make llieir appearance
as preparation for the next, we may describe each stage as follows : —

The Nauplius has three pairs of locornotor appendages, the first antenna?, the second
antennae, and the mandibles ; and there is a large labrum without a spine, and the
carapace and telson are absent. There is an ocellus, but no compound eyes.

The Protozoca has two pairs of antennae, which are like those of the Nanplius. The
mandible is reduced to a cutting blade. There are two pairs of biramous maxillee,
with scaphognathites, and two pairs of biramous maxillipeds. There is a long hind
body, ending in a flat telson. The labrum has a spine. The carapace is large, and
has a rostrum, a median dorsal and two lateral posterior spines ; and its free edges
reach down beyond the basal joints of the appendages. There is an ocellus, but no
stalked eyes.

The Schizopod stage is characterized by the great change in the two pairs of
antennae, which are no longer like those of the Nauplius, but have the characteristics
of those of the adult. All the mouth parts and four pairs of thoracic limbs are present,
and all posterior to the first pair of maxillipeds are biramous and locomotor. The
abdomen has six somites and a movable telson. The swimmerets are present, but the
other abdominal appendages are not.

The ocellus persists, but the stalked eyes are also present. The carapace has a
rostrum and two antero-lateral spines, but those at the posterior edge have disappeared.
The edges of the carapace do not reach over the basal joints of the thoracic limbs, and
the body is flattened vertically. The labrum still has a spine.

The young Lucifer and the adult female have a long flagellum on the first antenna,
a flagellum and scale on the second ; the ear and antennary gland are present ; the
neck is elongated. The fourth pereiopod has disappeared, and the others, as well as
the maxillipeds, have lost their exopodites. The first pleopod has one terminal branch,
the next four two branches each ; the sixth abdominal somite has a smooth lower
edge. The telson is straight and the outer end of the exopodite of the swimmeret is
rounded.

The adult male has a clasping organ on the first pereiopod, three rami on the second,
two teeth on the lower edge of the sixth abdominal somite, a square end to the
exopodite of the swimmeret, and a bent telson.

It is true that these five stages merge into each other somewhat, and that they are
complicated by the presence of the rudiments of organs which are be functional at the
next stage ; but after ah1 these secondary modifications are allowed for, it will be seen
that each stage is sharply and definitely marked, and separated by a pronounced gap
from the stages before and after.

The significance of these five stages can be best inquired into after the corresponding
stages of other Sergestidce have been examined, and I will return to the subject further
on, in a section on the general relationships^ the group.

N 2

92 MR. W. K. BROOKS ON LUCIFER:

IV. HISTORY OF THE APPENDAGES OF LUCIFER.

For convenience of reference I will now describe the changes which each appendage
undergoes at each stage of development, going over the same ground once more, but in
a different way.

The first antenna.

In the egg Nauplius (figs. 21, 23, and 24,^4) this appendage is unjointed, more than
half as long as the body, and it carries a terminal tuft of hairs.

In the first free Nauplius (6g. 25, A) it consists of five nearly equal joints; it is
nearly as long as the body of the second antenna, and its tip carries two long simple
hairs and two much smaller hairs.

In the last Nauplius stage, or metti-Nauplius (fig. 2G) the joints have disappeared ;
it is only about two-thirds as long as the body, and it carries only the two long hairs
at the tip.

In the first Protozoea stage (fig. 27, A) it is made up of a long cylindrical basal joint
with a few short hairs, arid a much shorter terminal joint, which is pointed, and carries
the two long hairs as before.

The structure of the appendage does not change until the end of the Zo&a series,
and it is shown at A in figs. 34, 42, 43, and 44.

At the first Schizopod stage (fig. 50, A) the basal portion is made up of one very
long cylindrical joint, with a hook near its swollen base, and a much shorter distal
joint. Three long, two-jointed, plumose hairs spring from the inner edge of the second
joint, and three more from the inner edge of the distal third of the basal joint. The
terminal portion has lost the two long hairs which it had at earlier stages.

In the last Schizopod stage (fig. 54, A, and fig. 56) the distal third of the basal
joint has separated off as a distinct joint (fig. 56, 2) upon which the tliree hairs are
situated. The ear has made its appearance, behind the hook, on the swollen base of
the first joint. The terminal joint (4) carries three sensory hairs, which arise upon its
outer surface about half way between its tip and base.

In the Mastigopus stage (fig. 60) the terminal joint has lengthened to form a two-
jointed flagellum, and the appendage is more than twice as long as the eye.

In the young specimens which have attained to the adult form (fig. 61, A) the
appendage is about as long as the carapace and neck, and in the adult (fig. 66, A) the
flagellum (2) is about as long as the basal portion (1). It consists of ten joints, the
terminal one longest, and the first and second thick. The first carries two and the
second three sensory hairs.

The basal portion is thick, cylindrical, three-jointed, with six plumose hairs, and the
ear nearly fills the enlarged base.

A STUDY IN MORPHOLOGY. '.»3

The second antenna.

In the egg NaupJ/t/s (fig. 24, An) this is unjointed, more than half as long as the
body, divided into two nearly equal rami, with hairs at their tips.

In the first free Nauplius stage (fig. 25, An) a two-jointed basal portion carries a
three-jointed exopodite and an eight-jointed endopodite. The appendage is nearly as
long as the body, the two rami are about equal in length, and each has three long
simple hairs at its tip. In the last Nauplius stage (fig. 26) the joints are obscure ;
the endopodite is longer than the exopodite ; it has long hairs along its side, and
those at the tip are plumose. In the Protozoea stages (figs. 27, 34, 35, 42, 43, and
45, An, and fig. 36) it consists of a two-jointed basal portion (fig. 3G), which carries an
unjointed exopodite (ex) with long, slender, non-plumose terminal hairs, and an eight-
jointed endopodite (en) with eight long hairs arranged along its side and tip. The first
and second joints are very short, while the other six are longer and nearly equal.

In the first Schizopod stage (fig. 50, An) the appendage is rudimentary, its joints
are absent, and the exopodite is almost but not quite as long as the endopodite. The
appendage is only half as long as the first antenna. In the last Schizopod stage
(fig. 54, An, and fig. 57) the exopodite has become a scale, which is only half as long
as the seven-jointed flagellum which has become developed from the endopodite ; the
basal joint is simple, very large, and the appendage is as long as the first antenna.

The flagellum now grows rapidly, and in the adult (fig. 61, An, and fig. 66, An)
it has twenty-four joints, and is more than half as long as the body. The antennal
gland opens into its base, and the scale is longer than the eye, and carries twelve long
plumose hairs.

The mandible.

In the egg Nauplius (fig. 24, M) this is biramous, unjointed, and tipped with hairs.
In the first free Nauplius (fig. 25, M) it is short, and made up of a stout basal joint ; a
two-jointed exopodite with three long slender hairs, two of which are carried by the
terminal and one by the proximal joint ; and a shorter endopodite with three long
simple hairs. In the last Nauplius stage (fig. 26, M) the joints of the exopodite have
disappeared, the three hairs on the endopodite have lengthened and become plumose,
and the inner edge of the basal joint carries a hook or blade. From the beginning of
the Protozoea series to maturity the mandible is a cutting blade, with no trace of a
palpus, and the number of its denticles gradually increases with age.

The metastoma.

The manner in which the metastoma originates in the Nauplius as a pair of buds
similar to those which become the maxillee, as well as the fact that it persists in
closely-allied forms as a pair of limb-like structures, seems to show, as CLAUS has

94 MB. W. K. BROOKS ON LUCIFER:

pointed out (' Untersuchungen,' &c., p. 15), that the Decapod metastoma is moi*pho-
logically a pair of appendages ; that it has been formed by the simplification and union
of structures homologous with the limbs; and that this pair of appendages was originally
furnished with a body-somite and a pair of ganglia. CLAUS'S reason for the homology
is the resemblance between the Decapod Protozoca, and the larva of Phyllopods and
Copepods, and the manner in which these parts are developed in the Nauplii of Lvcifer
and Eitphausia seems to be an additional reason for accepting his view.

The first maxilla.

This appendage is rudimentary during the Nauplius stages, but, as shown in
fig. 21, MX. 1, it is represented by a pair of buds several hours before birth.

In the Protozoea and Zoea series it has the form shown in fig. 46, which was drawn
from the appendage of a larva in the last Zoea stage. Its characteristics are developed
gradually, and it is somewhat simpler during the earlier Protozoea stages than it is in
fig. 4(3. Fig. 30 shows it as it appears in the first Protozoea when seen from the
outside. It consists of a basal portion (fig. 46) made up of two joints (1 and 2), which
carries a short obscurely-jointed endopodite (en) and a scaphognathite (.*•). In my
description of this and the other mouth parts of Lucifer I have accepted CLAUS'S
homology ('Untersuchungen,' &c., p. 16), and regard the two basal joints as the
equivalent of the basal portion of the antenna, or of one of the thoracic limbs ; the
jointed palpus as the homologue of the inner ramus of the antenna, or the limb proper
of one of the thoracic appendages ; and the scaphognathite as the homologue of the
exopodite of one of the thoracic appendages, or of the antennte. In all these appen-
dages the exopodite is shorter than the endopodite, unjointed, and set with long hairs,
the plumose character of which is well marked. The scaphognathite of the maxilla
agrees with the exopodite of the second antenna and of the other appendages in this
respect, while the palpus of the maxilla agrees with the endopodite of the second
antenna, and with that of the mandible of the JVfmj;>?n<sand with the thoracic limbs of
the adult Lucifer, in consisting of several joints with one or more, usually simple,
hairs at each joint.

The inner edges of the basal joints of the maxilla, carry cutting hairs, and the
second joint is largest. The endopodite carries five long slender hairs which are simple
in the earlier and plumose in the later stages. The scaphognathite carries three hairs
which are equal and simple in the earlier Protozoea, but plumose in the Zoca, where one
is very much longer than the other two.

The structure of this appendage undergoes extremely little change from the time it
appears in the Protozoea to maturity. In the Schizopod larva (fig. 52) the second
basal joint (2) has become much larger than the first (1), and its cutting hairs are more
numerous than before ; a small slender plumose hair has made its appearance on the
edge of each joint. The endopodite (en) is obscurely three-jointed, and the scaphogna-
thite (sc) has only two long plumose hairs.

A STUDY IN MORPHOLOGY. U5

In the adult (fig. t>4) the seaphognathite is absent ; the endopodite is rudimentary
and the second joint of the base (2) is very much wider than the first (l), and has
fifteen cutting hairs arranged in three rows, while the first joint has only four very
much longer serrated cutting hairs. The basal joint has only one plumose hair as
before, but the second joint has one on each side of the blade.

The second maxilla.

The second maxilla is present as a bud (fig. 24, MX. 2) in the egg, and it becomes
functional in the first Protozoea, and persists without very much change to maturity.

In the first Protozoea (fig. 31) it has a long, many -jointed basal portion (h), with slender
simple hairs on its inner edge ; a two-jointed endopodite (en) with three simple hairs
on its tip, and two on the second joint ; and a small seaphognathite with plumose
hairs.

In the last Zoea (fig. 47) the hairs on the inner edge are plumose, and one of
the three hairs on the small seaphognathite is much longer than the others.

In the Schizopod stage (fig. 53) the limb is thick and long, the seaphognathite is
rudimentary, and the endopodite is small, and has no terminal hairs.

In the adult (fig. 67) the endopodite and all but three of the joints of the basal
portion are absent. The first of these (3) is the largest and has a broad edge, with a
number of cutting hairs, while the others (2 and 1) are narrow and have three hairs
each. All these hairs are simple. The seaphognathite is elongated, and is now about
as long as the body of the appendage, to which it is joined by a narrow neck. The
inner end has four plumose hairs, three of which are about as long as the appendage,
while the fourth was short and apparently broken in all the specimens which I
examined. The outer half of the seaphognathite has three short straight plumose
hairs on its outer end, and five somewhat longer ones on its inner side.

Thejirst maxilliped.

The first maxilliped is represented by a bud in the. egg Nauplius (fig. 21 J\Jj>. 1) and
it becomes functional in the first Protozoea., and then consists (fig. 32) of an unjointed
exopodite (ex) with four long terminal hairs ; a four-jointed endopodite (en), with three
long terminal simple hairs, and a shorter hair springing from the inner edge of each
joint ; and an obscurely two-jointed basal portion with short simple hairs on its inner
edge.

In the Zoea stage (fig. 48) the hairs on the inner edge and on the exopodite are
plumose, and the endopodite is long and six-jointed.

In the Schizopod stage (fig. 50, Mp. 1) the joints are obscure ; the exopodite
is nearly as long as the endopodite ; all the hairs are plumose, and about equal in
length, and there is a double row along the inner edge of the appendage.

96 MR, W. K. BROOKS ON LUCIFER:

In the adult the appendage (fig. 68) is extremely simple, short, stout, two-jointed,
flattened on its inner and rounded on its outer surface, with a fringe of short, stout,
equal, plumose hairs around the edge of the flattened surface.

The second maxUliped.

It is difficult to decide with certainty whether this appendage is represented by
a bud in the Nauplius or not. If the first pair of buds become the nietastoma,
as seems probable from their position with reference to the mandibles and from the
analogy of the Euphausia nauplius, the second pair of maxillipeds are not represented,
but if the first pair of buds are the rudimentary first maxillas the last pair are the
second maxillipeds. At any rate the appendages are present in the first Protozoea
(fig. 33), and they are essentially like the first pair, but much smaller.

In the last Zoca stage (fig. 49) they are larger, although still smaller than the first,
and their inner edges carry only three short hairs which are not plumose.

In the Schizopod stage (fig. 59, Mp. 2) a long basal joint carries a four-jointed
endopodite and an unjointed exopodite of nearly equal length. The outer half of the
exopodite is fringed by fourteen long, simple hairs, and the terminal joint of the
endopodite has a few short plumose hairs.

In the next stage the exopodite is absent, and the long six-jointed limb (fig. 70) is
bent into the shape which is so characteristic of the adult Sergestidce.

The basal joint (1) is quite short and stout. The next joint (2) is longer and has
five plumose hairs, almost as long as the joint, on one side and three on the other.

The next joint (3) is the longest, and carries six plumose hairs. The next (4) is
about as long as the second, and the bend in the limb occurs in this joint and between
it and the third. It carries ten plumose hairs about as long as those in the other
joints, and arranged in a single close rank. The fifth and sixth joints are shorter than
any of the others except the tirst ; they are about equal in length, and the fifth carries
ten, the sixth six long plumose hairs.

T/te third maxilliped.

This appendage makes its appearance as a bilobed rudiment (figs. 43 and 45, Mp. 3),
at the end of the Zoca series, and it becomes developed into a Schizopod foot, at the
next or first Schizopod stage (fig. 59, Mp. 3). A stout basal portion which appears to
be two-jointed, carries an unjointed exopodite, and a four-jointed endopodite. The
latter branch is the longest, and its tip carries four rather short plumose hairs. The
outer half of the exopodite carries fourteen long simple hairs.

At the end of the Schizopod period the limb loses its exopodite entirely, lengthens
and becomes a slender six-jointed leg, fringed by a double row of short hairs, as
shown in fig. 61, Mp. 3.

A STUDY IN .MORPHOLOGY. D7

The history of this appendage in Lucifer shows that that there is no reason, except
the arbitrary system borrowed from the higher Decapods, for classing this appendage
with the mouth parts, instead of with the thoracic limbs.

It appears much later than the first and second pairs of maxillipeds, or at the same
time with the thoracic limbs. It agrees with these latter in all its subsequent changes
and in its adult structure, and must be regarded as forming one of the thoracic series.
I have employed the recognised name, third maxilliped, to prevent confusion, but the
appendage is in no sense a mouth part. In fact, the only reason for holding that the
missing appendage in Lucifer is the fifth pereiopod, instead of the last maxilliped, is
the tacit assumption that the appendages must follow a definite serial order from in
front backwards. We do not know that this assumption is justifiable in all cases, and
it is therefore perfectly possible that the appendage which is usually called the third
maxilliped of Lucifer may really be the first pereioped. I think the probability is in
favour of the accepted homology, but the use of the term " third maxilliped " in the
present paper for the appendage in question must not be regarded as evidence that
the homology is accepted without question.

The pereiopods.

At the end of the Zoea series four pairs of pereiopods, the first, second, third,
and fourth, are represented by buds (figs. 43 and 45), while the fifth is entirely absent,
as DANA pointed out in the ' Report on the Crustacea collected by the United States
Exploring Expedition,' p. 634. WiLLEMOES-SuHM (Proc. Roy. Soc., vol. 24, p. 134),
calls attention to the same fact : the total absence of this somite at all stages of
development. In the Schizopod stage each of these appendages is biramous (fig. 59),
and similar to the last maxilliped, although the first three pairs (fig. 59, Pr. 1, Pr. 2,
and Pr. 3) are longer.

At the end of the Schizopod series of stages the entire fourth pair and the exopo-
dites of the other three pairs disappear, and the endopodites lengthen to form the
long slender limbs of the adult (fig. 61, Pr. 1, Pr. 2, and Pr. 3). They are four-
jointed, with a double row of short hairs along the anterior edge, and the first is only
half as long as the second and third, which are nearly equal, and almost as long
as the carapace and neck. The third ends in a short, curved hairy claw, too small
to be shown in the figure.

The Jirst abdominal appendage.

This is present as a rudimentary bud at the end of the Schizopod series, but does
not become functional until the Lucifer form is reached. In the young it consists
of a long uujointed base, and a single pointed tip, fringed with swimming hairs
(fig. 61, PL 1). In older specimens the basal portion divides into two joints, and in
the young male or the young or mature female the appendage has the form shown
in fig. 74. As the male approaches maturity a small process, shown in fig. 76,

MDCCCLXXXII. o

98 MR. W. K. BROOKS ON LUCIFER:

appears on its anterior face, and becomes modified in the mature male into the
clasping organ (rig. 75, c), while a second process (d) appears a little nearer the tip of
the limb.

The second abdominal appendage.

This appears at the same time with the first, and developes two terminal branches.
In the mature male a third shorter one is added.

The third, fourth, and fifth abdominal appendages,

These all develop at the same time with the first and second ; they have two
terminal branches and are alike in both sexes.

The sixth abdominal appendage.

This is present as a rudiment in the last '•' Zoca" and it becomes fully developed in
the first Schizopod larva.

It consists of a basal joint which carries a long, wide, and flat exopodite, and a
narrower shorter endopodite.

In the young and in the mature female the outer end of the exopodite is rounded,
but it is nearly square in the mature male.

The labrurn.

The labrum is large and conspicuous in the Nauplins, but it has no spine. The
spine is present from the first Protozoea stage to the last Schizopod stage, but it is
absent in the adult.

The compound eyes.

These make their appearance as rudiments in the last Protozoea, but they are
not perfectly developed or stalked until the last Schizopod stage. The homology of
the stalked eyes of the Malacostracan has been a matter of some uncertainty. They
are usually enumerated in the list of appendages, and the typical Crustacean is sup-
posed to have a corresponding somite. GLAUS has pointed out (" Zur. Kenntniss
der Malacostrakenlarva," Wiirzb. Zeitschr. ii., 180 1. p. 33) that no especial taxonomic
importance can be attached to their presence or absence ; and their mode of origin in
Lucifer certainly gives no support to the view that they have been produced, like the
mandibles, by the gradual specialisation of a, pair of ordinary appendages. They do
not resemble ordinary appendages at any stage, but are formed directly, and the fact
that the period of their development is spread over several moults renders their
history quite different from that of the appendages. As I shall show further on,

A STUDY IN MORPHOLOGY. 99

serially homologous organs do not necessarily owe their resemblances to inheritance
from the unspecialisecl organs of a remote ancestor, and I think that the presence of a
distinct occular segment in Squilla compels us to recognise an homology between the
stalked eye and an ordinary appendage, although it is no doubt true that all the
groups in which stalked eyes occur cannot be traced back to a common stalked-eyed
ancestor, and also true that the stalked eyes themselves cannot be traced back to
ordinary appendages.

The ocellus.
This is present from the first Nauplius stage to the end of the Schizopod series.

Explanation of Table I.

This table is designed to show at a single view the condition of each appendage at
each stage of development.

For convenience I have included the compound eyes, the ocellus, and the labrum,
but do not wish to imply that these structures are or are not homologous with ordinary
appendages, and I have omitted the metastoma, although I have no doubt that this
should be included in a list of the appendages.

In the table the word " same " indicates that the condition of the appendage is the
same as it was at an earlier stage, and does not refer to other appendages in the same
vertical line.

100

MR. W. K. BROOKS ON LUCIFER:

aj

13

a

3
•<

43

| J |

02 <! 02

0)

a

02

o o o o

a a a a

0202 02 CO

t* t* ^3 £

(DOj<D<i>eDa><BcDE1<»co(i>iB°«3
| | | | | | | | J jj | | | 1" g'-J'

O2 CO O2 02 CO CO 02 CO O EH 02 02 02 &3

S

• '-gVg

'S

o -S

a,

a

^ 0) 4^

s g S

43

a)

<u .

Q y ^Q C^

"e "S "S "§ "S 'g

'is *"»

02 & yi •!) vi 3

Young of hot
fe

-p

O g 03

a s a

C« r^ tf

02 <!co

to

a

c3

J 1 I'g'a'p.
^ co co a

§ o> eo o M o>

| i ;i if 4i 1 J ' a 1 1 1 1 ;|

p^p^HHWW^t/iOE^^HE"'^

^3
0

• ' a

g

.§

'eg

•g

00

0) CD 22

a a ,5

cS c3 ^

02 CO <J

"3

O Q) Q} O

925 S
ob co ob OQ

llllllllllllll

i

S

•

' • • be

o

1

• \a

^

g

o 53

0 .

* ' ' a
^ c

Is3

^

2

1

05 S

-4-3 ^

1

1

r^* ' £

a

o
•3 'S

O ,-jf

ill It

'33333 ff
o o o o o o

MO^ <s

t£ OQ CQ W

B

"ii

J

ua CD o>

-g a a

tx". ,^n [/^

S

CD O Q) CJ

a a a a

ce cS c3 cS
02 02' O2 02

llllllllllllll

Is

S

|i|

O 3 E3
**-« o O

o o . ....

w

0) 0) 0)

a a a

1

02

® ^ s a

lla|||||aaaaa|

ffiWcecOCO02COCOO2CO02O2COO2

" en

CQ

J

a

Q

—

- *

cS

0) 3 •

1

li

a p

a'E,a •

• '-ggggggggggg

J £ S

r^r2 '£, f, S J§ S £ J§ ja j§ ,0 JD J§

^fj p i Pq

^M"

^5 CQ n*i rn

mpH«l«l^«^-<<i«i«^<i<i

Compound eye .
Ocellus
First antenna .

Second antenna

Labrtim .
Mandible .
First maxilla

Second maxilla .

I? ^ ? ^ S-"^ E.'X ^ r- ^

00

<D

9P

0)
OH

S1

<HH

O

o

4J

03

CO

o

-s

-p

O

O

w

HH
ffl

A STUDY IN MORPHOLOGY. 101

V. THE METAMORPHOSIS OF ACETES.

While I was studying the development of Lncifcr, I found during the month of
September a few specimens of the very similar larva which is shown from above in
Plate 9, fig. 79, and from the side in fig. 78.

Several specimens were placed by themselves in tumblers of sea-water, where they
passed through the stages shown in Plate 11, figs. 84, 85, and 90. Only one of my
specimens reached this last stage, and as this one moulted on the last day of the
season I was not able to trace it any further, and as I collected no adult specimens of
the same kind, its precise systematic position must at present remain in some
uncertainty. The close similarity which I shall point out between its larval stages
and those of Lucifer and Sergestes renders it very probable that it is a Sergestid, and
the analogy of these forms also indicates that the larva shown in fig. 90 has in all
probability nearly or quite attained to the mature form. This larva differs from the
other two forms in the possession of small claws at the tips of the last three pairs of
pereiopods, and as this is characteristic of MILNE-EDWARDS' genus Acetcs, and only
three genera of Sergestidse — Lucifer, Acetcs, and Sergestes — have been described, I
think we may conclude that we have to do with the development of an American
species of this genus. At any rate, whatever the systematic position of the adult may
be, the fact that the Protozoea is in most respects intermediate between the simple
Protozoea of Lucifer and the extremely modified Protozoea of Sergestes, gives this form
so much interest that it seemed best, for the sake of comparison, to embody all that
I was able to learn about its metamorphosis in the present paper.

At the earliest stage which was observed, the larva (figs. 77, 78, and 79) is a "Zo'e'a"
1000 inch long, and a comparison of fig. 79 with fig. 44, or of fig. 77 with fig. 43, will
show that it is essentially like the last Zoca of Lucifer, although the minor differences
are both numerous and conspicuous.

The number of somites and appendages is alike in both forms, and the appendages
are alike in most respects, although each one of them shows distinctive characteristics
of its own.

The carapace (fig. 79) makes about one-half the length of the body, and it is much
more flattened than it is in Lucifer. It has a rostrum (R) and a median dorsal spine,
but the postero-lateral spines (sp.) point outwards and backwards, instead of directly
backwards, and there are a pair of anterior spines as long as the rostrum, projecting
over the eyes. The two large pigment-spots which give such a characteristic appear-
ance to the carapace of Lucifer are entirely absent, and the thoracic segments and
appendages are covered by its posterior edge.

The eyes are mounted upon distinct stalks, while they are sessile in Lucifer at the
same stage.

The abdominal somites are short and wide, and coloured by bright-red pigment-
spots, and their lower edges are produced into strong projecting spines.

102 MR, W. K. BROOKS ON LUCIFER:

The telson is deeply cleft, and its halves diverge from each other like a swallow's
tail feathers, so that the posterior ends of the rudimentary swimmerets are visible
between them in a dorsal view, as shown in figs. 77, 78, and 79. The spines on the
telson are similar in arrangement to those of Lucifer, but much longer.

A comparison of the Lucifer Zoea (fig. 44), the Acetes Zoea (fig. 79), and the
Sergestes Zoea (CLAUS'S 'Crustacean-System,' taf. vi., fig. 1) at the same stage of develop-
ment, brings out the extremely interesting fact that the Acetes larva stands between
the very simple Zoea of Lucifer and the remarkably complicated Elaphocaris larva of
Sergestes in nearly every feature in which, the two differ. In Lucifer the eyes are
sessile ; in Acetes they have short stalks ; and in Sergestes the stalks are very long.

In Lucifer the spines over the eyes are absent ; in Acetes they are present and
simple ; and in Sergestes they are very long and compound.

In Lucifer the postero-lateral spines are parallel to the long axis of the body ; in
Acetes they are oblique, so that they project a little beyond the outline of the body ;
and in Sergestes they are at right angles to the long axis, and compound.

The carapace, including the rostrum, makes about one-third of the total length of the
body of the Lucifer Zoea ; about one-half of that of the Acetes Zoea ; and more than
two-thirds of the total length of the Sergestes Zoea. The abdominal somites of the
Acetes Zoea are shorter and wider than those of the Lucifer Zoea, and this change is
carried still further in the Sergestes Zoea. In the Lucifer Zoea the sixth abdominal
somite is the only one which has ventral spines, and these point backwards. All the
abdominal somites of Acetes have spines, and they point backwards and a little
outwards, while in Sergestes they all point directly outwards.

The telson is slightly notched in Lucifer ; deeply forked in Acetes ; and in Sergestes
the prongs of the fork diverge so much as to form a right angle.

These facts are extremely interesting, as they seem to show that the Elaphocaris is a
larva essentially like that of Lucifer, which has passed through a remarkable process
of secondary modification, resulting in the acceleration of the development of the
eyes, and the production of a forked telson, and a very 'spiny body. The larva of
Acetes has been modified in the same direction but to a much less degree. It may be
asked why we are to assume that the Lucifer Zoea is the primitive form, and the
Elaphocaris larva the secondary modification rather than the reverse ; but a little
thought will show that the distinctive features of the Elaphocaris stand in direct
relation to the environment, as weapons of defence, sense organs, or locomotor
apparatus, while the distinctive marks of the Lucifer Zoea are features of general or
typical resemblance to the corresponding larva of Euphausia and Penceus.

I did not succeed in finding the Proiozoea from which the Zoea shown in fig. 79 is
derived, but I think it extremely probable that future research will show that an
unknown larva which has been figured by DOHRN and GLAUS is the Protozoea of Acetes,
or else of a new closely-related genus of the Sergestidae.

In his " Untersuchungen iiber Ban und Entwickelung cler Anthropoden" (Zeit. f.

A STUDY IN MORPHOLOGY. 103

Wiss. Zool., xxi., 1871), DoHEN describes the " Larve eines unbekannten Krebses"
from the Indian Ocean (p. 377), which is shown in his plates 29 and 30, tigs. 62
to 67. In his ' Crustacean- System ' (taf. iv., figs. 2 to 7) CLAUS gives much more
satisfactory figures of what appears to be the same larva, and speaks of it as a
" Phyllopodenahnlichen Protozoea unbekannter Herkunft." Its close resemblance to
the Protozoea of Lucifer renders it extremely probable that it is the Protozoea of a
Sergestid, and as the Protozoea of Lucifer and that of Seiyestes-a,re known, this must
be the larva of Acetes, or of some closely-related unknown form.

The carapace is nearly smooth, rounded, and there is no trace of a rostrum, and it
makes more than three-quarters of the total length of the body.

The compound eyes are present and well developed, but they are sessile, and there
is no indication of the stalk. The first antenna is seven-jointed, and the two tei'minal
joints are thin and long.

The second antenna is nearly twice as long as the first, and very thick. Its short
stout basal portion consists of two joints, and carries a short two-jointed exopodite, with
three long terminal non-plumose swimming hairs, and a very large twelve-jointed
endopodite with a long swimming hair at each joint.

CLAUS'S figures show that the appendages at the back of the antenna? are very much
like those of Lucifer, and the same ones are present ; that is, the mandibles, first and
second maxillse, and first and second maxillipeds.

The hind body is segmented, and ends in a broad, fiat, deeply-cleft telson, with six
pairs of irregularly plumose hairs, the third pair very much longer and thicker than
the others.

A comparison of CLAUS'S figure with tig. 27 of this paper will show that most of the
ditferences between this unknown larva and the first Protozoea of Lucifer are of the
same kind as the ditferences between the Acetes Zoea (fig. 79) and the corresponding
stage of Lucifer (fig. 44).

At a time when the eyes of Lucifer are rudimentary and sessile they are perfect and
stalked in Acetes, and at a time when they are entirely absent in Lucifer DOHEN'S
larva has them sessile and rudimentary but distinct.

The Zoea of Acetes, like this larva, has its telson deeply forked ; its hairs are
plumose, and the third is much longer than the others. These resemblances, and the
great length of the carapace, render it very probable that this unknown larva is the
Protozoea of Acetes.

I will now continue my description of the appendages of the Zoea.

The first antenna (fig. 77, A) is uniramous, and it consists of a long, cylindrical,
two-jointed shaft, and a single short flagellum, which shows obscure traces of a division
into three joints. The basal joint of the shaft is a little more than half as long as the
second joint, and it carries a single short sharp hair on the inner side of its distal end.
The second joint has two much longer hairs on its distal end, and one about half way
between its ends. The flagellum makes about one-fifth of the total length of the

104 MR, W. K. BROOKS ON LUCIFER:

appendage, and it carries four terminal hairs, two of them about as long as those on
the terminal joint of the shaft, and two nearly three times as long.

The second antenna is the chief locomotor organ, and (as shown in fig. 77, An) it
consists of a thick two-jointed basal portion, which carries a two-jointed exopodite (ex)
and a ten-jointed endopodite (en).

The proximal joint of the exopodite is about twice as long as the terminal joint, and
it carries two long hairs on its outer end, and two more near the base. The terminal
joint has, at its tip, one short hair, and four which are about as long as the limb. The
endopodite consists of four short rings, and a series of six joints like those of the
corresponding organ of the Nauplius, Protozoea, and Zoca of Lucifer. The terminal
joint carries four, and each of the five other joints one long swimming hair, and none
of these hairs are plumose.

On the basal portion of the appendage there is a large bright-red pigment-spot,
which forks and runs along the exopodite and endopodite, about half way to their
tips.

The labrum (fig. 77, L) is smaller than that of Lucifer, with a spine and a large red
pigment-spot.

The mandible (fig. 77, M, and fig. 80) has small irregular denticles along its cutting
edge, and these reach to the tip of the long tooth which occupies the posterior angle
of the blade. The mandibles of two specimens were dissected out, and in each case
there was a little hairy pad (in) upon the posterior surface. It could also be seen
in the entire animal (as shown in fig. 77). It is possible that this pad is the man-
clibular palpus, but it seems much more probable that it is half of the lower lip or
metastoma, for no palpus is present on the mandible of Lucifer.

The first maxilla (fig. 77, MX. 1, and fig. 81) is quite different from that of Lucifer
(fig. 46) at the same stage, but the difference is in minor points, and there is essential
agreement in general structure. The two basal joints or blades are long and slender,
and their hairs are also longer and thinner than they are in Lucifer. The endopodite
(en) is placed nearly at right angles to the base, and is distinctly three-jointed. It
carries five hairs as it does in Lucifer, and they are similarly placed, but longer. The
three hairs on the scaphognathite are about equal in length, and the plumules on
their sides are short and irregular.

The second maxilla (fig. 77, MX. 2, and fig. 82) is much like that of the Lucifer
Zoea (fig. 47), but the three hairs at the tip are more than twice as long as those on
the inner edge of the appendage, and they are irregularly plumose, while they are
simple in Lucifer.

The first maxilliped (fig. 77, Mp. 1, and fig. 83) differs from that of Lucifer (fig. 48)
in the same way, and the exopodite carries seven instead of four haii-s, and these are
as long as the appendage, and two-jointed.

The second maxilliped (fig. 77, Mp. 2) is about as long as the first, but it does not
seem to be of much functional importance. It is usually carried stretched back along

A STUDY IN MORPHOLOGY. 105

the hind body, as shown in the figure, and its hairs are short. As in Lucifer at the
same stage, the exopodite is as long as the endopodite.

The third pair of maxillipeds, and the first, second, third, and fourth pairs of
thoracic limbs are represented by buds, as in Lucifer at the same stage. The bud for
the third maxilliped (fig. 77, Mp. 3) is bilobed, longer than the others, and it points
backwards outside the other buds. The buds for the first three pairs of periopods
are bilobed, in contact on the median line, and about equal in size. Those for the
fourth pair are much smaller, and are hidden in a ventral view by the buds for the
third pair, but they can be seen in side view (as shown at T4 in fig. 78). There is no
trace of the fifth pair of pereipods either at this or at any later stage. GLAUS figures
buds for the fifth pair in the Zo'ca of Sergestes, and also in the next or Acanthosoma
stage of Sergestes ; but the study of the Zoea of Acetes shows even more satisfactorily
than is the case in Lucifer that these appendages are entirely absent, and it seems
safe to believe that this is the case in Sergestes also until the larva of the latter has
been carefully examined with reference to this particular point.

The abdominal appendages, with the exception of the fifth pair, are entirely absent ;
but each abdominal somite has a pair of long ventral spines. The swimmerets are
represented by long bilobed buds, which project beyond the fork or notch in the telson.
The abdominal ganglia are very much more conspicuous than they are in Lucifer.

The distribution of pigment is somewhat different from what we find in the Lucifer
Zoea, and nearly all the pigment-spots are bright-red. There is a large spot of red and
one of reddish-yellow on the eye stalk, a red spot on the labrum, a large red and very
dendritic spot on the second antenna, red spots on the dorsal surface of the posterior
edge of the third, fourth, and fifth abdominal somites on the median line; red spots
on the ventral surface of the first, second, and third at the bases of the spines ; a red
and a brown spot at the base of the spine on the fifth ; a brown spot at the base of the
spine on the sixth, and a red spot on the base of the swimmeret. The anal spots are
large and bright-red.

On September 20th I found several specimens of the stage which has just been
described. Fig. 79 was made from one of them, which was then placed in a glass of
water by itself, and the next day it was found to be moulting. In the evening
the moult was found to be finished, and the larva was swimming actively. The'
drawing given in fig. 84 was made from it without injuring it, and later stages were also
drawn from the same specimen.

The larva, T^-Q inch long, has undergone very great change, and although it is an
Acanthosoma, it presents many important differences from both Lucifer and Sergestes.

The abdomen has lengthened so that the carapace makes less than one-third the
total length of the body, and the dorsal and postero-lateral spines have disappeared.

The abdominal spines stand out from the body, and the swimmerets have become
the chief locomotor organs. The spine has disappeared from the labrum ; the two
pairs of antennae have changed from the larval to the adult form ; the endopodite of the
fourth pereiopod, and the first three pairs of pleopods are represented by long buds.

MDCCCLXXXII. p

106 MR. W. K. BROOKS ON LUCIFER:

The first antenna (fig. 84, A) consists of a three-jointed shaft about as long as the
carapace, and two terminal flagella. The basal joint of the shaft makes half the total
leno-th of the appendage, and the other two are about equal to each other. On the
inner edge of the distal two-thirds of the shaft there are eight long, similar equidistant,
plumose hairs, and there are two short spines on the outer edge. The inner flagellum
is short, and carries one long slender terminal hair. The outer one is more than twice
as long, and carries two thick sensory hairs.

The exopodite of the second antenna (fig. 84, ex) has become a scale, only one-third
as long as the endopodite, which is now a ten -jointed flagellum about as long as the
carapace.

The second and third maxillipeds (fig. 84, Mp. 2 and Mp. 3) and the first, second
and third pereipods (fig. 84, T I, T -2, and T 3), are Schizoped-like, but they are of very
slight functional importance, and their endopodites are folded forwards on the ventral
surface, like the maxillipeds of Squilla, so that it is impossible to study the mouth
parts without dissection. The endopodite of the fourth pereiopod has entirely
disappeared, and the limb is represented only by its exopodite. The five exopodites
are about alike, and they all end in long slender swimming hairs : those of the four
pereiopods (T 1 ex, T 2 e'x, T 3 ex, and T4 ex) are bent outwards and upwards towards
the dorsal surface, as in the maxillipeds of a Crab Zoca, but those of the second and
third maxillipeds (Mp. 2 ex and Mp. 3 ex) are more nearly parallel to the endopodites.
The endopodite of the second maxilliped (Mp. 2) is free and movable, but those of the
third maxillipeds (Mp. 3) and of the first, second, and third pereiopods are covered by
a delicate cuticle, and are almost immovable.

I did not actually witness the next moult, but four days later the larva, ToW incn
long, was in the stage shown in fig. 85. The exopodites of the thoracic limbs have
become reduced to rudiments, the limbs themselves have stretched out and are now
functional, as are the three pairs of abdominal feet.

The first antenna (fig. 86) has not changed much, but its base is swollen and the
otocyst has appeared.

The second aiitenna is now half as long as the whole body, its flagellum is ten-
jointed, and red pigment has appeared at its base and tip (fig. 87). The outer end of
the scale carries nine long plumose hairs arranged on the tip and inner edge.

The second maxilliped (fig. 88, Mp. 2) is bent into a knee, and is fringed by sixteen
plumose hairs. Its exopodite is rudimentary, but longer than in any of the appendages
which follow.

The third maxilliped (fig. 88, Mp. 3) is long, slender, six-jointed, with a rudimentary
endopodite.

The other three limbs (fig. 88, T 1, T 2, T 3) are six-jointed and they end in enlarged
chela?. The first is the shortest ; the second is about as long as the third maxilliped,
and the third is still longer. The fourth is now represented only by a small rudiment
and a ganglion ; and the fifth is entirely absent, as it has been at all stages.

A STUDY IN MORPHOLOGY. 107

The three pairs of pleopods are alike in structure, and each consists of two joints
about equal in length (fig. 89). The outer half of the terminal joint is toothed and
carries six pairs of long slender non-plumose hairs, so arranged as to form a paddle.

The rostrum (fig. 85) is long and curved, and it has a single secondary spine in its
upper surface. A pair of very small spines have also appeared at its base.

This specimen had nearly completed its moult into the stage shown in fig. 90 on
the last day of my stay at the seashore — five days after fig. 85 was drawn.

Fig. 90 was drawn from another specimen, -^Q inch long, which was captured at
the surface on September 25th.

The eye-stalks are long and very movable, the flagelluui of the second antenna is
considerably longer than the body, the five pairs of thoracic limbs have developed gills,
and the fourth and fifth pleopods have appeared ; but in other respects the structure is
nearly as it was in the preceding stage. The endopodites of the maxillipeds are
pointed, but those of the three pereiopods end in rudimentary chela?. The endo-
podite of the third pereiopod is much longer and thicker than the others, and its tip
reaches nearly to the rostrum.

The buds for the first three pairs of pleopods are long, obscurely jointed, and they
meet each other on the median line. There are as yet no traces of the fourth and
fifth pairs. The spines on the abdominal somites are long and sharp. Those on the
first three somites point outwards and forwards, those on the fourth point almost
directly outwards, and those on the fifth outwards and backwards. The sixth abdo-
minal somite has lengthened, and is now about as long as those of the others. The
telson is short and shield-shaped, with two pairs of long and one pair of very short
spines, and the swimmerets are perfectly formed and fringed with long plumose
swimming hairs.

The exopodite is long, narrow, with a smooth outer edge which ends in a tooth, and
a rounded point. It carries fifteen hairs : ten on the inner edge, two on the tip, and
three between the tip and the tooth. The endopodite is nearly as long and wide as
the exopodite, and it carries nineteen hairs : two at the end, eight on the outer, and
nine on the inner edge.

The ocellus is still present, and the pigment-spots have nearly the same arrangement
as before, but some of them are now yellow or green instead of red.

The eye-stalk is about as it was before, and the ocellus is still present and double.
As regards the more minute structure of the appendages the first antennae are now
about as long as the carapace, and most of the increased length is in the flagellum,
which now consists of seven joints. The secondary flagellum is still quite short. The
shaft of the antenna is three-jointed, as before, but the basal joint is much lengthened,
and now makes more than hah0 the total length. The auditory organ at its base is now-
very conspicuous, and the inner edge of the shaft carries eleven hairs instead of six ;
five of these are on the basal joint, three on the second joint, and three on the third.

The scale of the second antenna has lengthened and is now more than half as long as

- P 2

108 MR. W. K. BROOKS OX LUCIFER:

the first antenna, including the flagellum. Its inner end carries eight, and its tip
three hairs, and the outer edge of the tip is toothed. The swollen base of the flagellum
of the second antenna carries a large red pigment-spot, and the flagellum, which is
considerably longer than the body of the animal, is also marked by bright-red pigment
throughout the greater part of its length.

The second maxilliped has become completely bent upon itself, and it bears a close
resemblance to that of the adult Lucifer, although it carries a gill, as do the second
maxillipeds and the three pairs of thoracic limbs. All traces of the exopodites have
disappeared from all these appendages, but their structure and comparative length are
about as before.

The first, second, and third pairs of abdominal feet have increased in length, and
the first is now almost as long as the last thoracic limb. The second is a little shorter ;
the third is still shorter, and has acquired a second terminal branch, which is as yet
rudimentary.

The fourth and fifth pleopods, which have now made their appearance, are much
smaller than the others, and each has one large and one small terminal branch.

The swimmerets and telson are very similar to those of the immature Lucifer,
although the telson is shorter and wider. The exopodite has fifteen hairs on its
inner edge, two on its rounded tip, and four between the tip and the tooth. The
endopodite has nineteen hairs. The surface of the carapace is finely punctated, and
the rostrum has no secondary spine. The spine has disappeared from the first
abdominal somite, and the one on the third somite is longer than any of the others.
The dorsal surface of the third somite is bent, so that the abdomen is no longer per-
fectly straight. Large conspicuous red pigment-spots have appeared on the lower
edges of the second, third, fourth, and fifth abdominal somites.

As the series of drawings which I have given was made from such a small number
of specimens, I am unable to contribute much information as to the changes of the
mouth parts, and must leave this, as well as the exact determination of the adult form
and systematic position of the species, to future research.

In his ' Facts for Darwin ' FR. MULLER has figured a larva (fig. 33) which is
extremely like, if not identical with the one shown in fig. 90, and he regards it as the
young of a Prawn, closely related to Penceus. GLAUS has suggested (' Crustacean-
System,' p. 35) that it is much more likely to prove to be a young Sergestid than a
Prawn, and the facts regarding its metamorphosis which T have given above, certainly
seem to point in the same direction. An earlier stage of development is given in
FR. MULLER'S fig. 32, and a comparison with my fig. 84 will show that the same larva
at an earlier stage might, when crushed by a cover glass, present very much the same
appearance as this larva. If they are the same FR, MULLER is certainly mistaken in
his statement that fig. 33 follows directly after fig. 32, without the intervention of a
Schizopocl stage, for the metamorphosis is really quite complicated, and a true Schizopod
stage exists, although it is of extremely short duration.

A STUDY IN MORPHOLOGY. 109

VI. — RELATION BETWEEN THE LARV^ OF LUCIFER, ACETES, SERGESTES, PEN.-EUS,
AND EUPHAUSIA, AND THE SIGNIFICANCE OF THE DECAPOD ZoEA AND THE
CRUSTACEAN NAUPLIUS.

The general significance of the peculiar type of Decapod metamorphosis, of which
Lucifer is now the most thoroughly known illustration, has been discussed with the
greatest ability and knowledge of the facts by GLAUS in his ' Untersuchungen zur
Erforschung der Genealogischen Grundlage des Crustacean-Systems.' My own ac-
quaintance with the phenomena of Crustacean morphology in general is very far from
being sufficiently extended and minute to qualify me for a critical discussion of this
work ; but while the facts in the life-history of Lucifer seem to tend to a similar con-
clusion, and even to place it upon a much firmer basis than before, they also indicate
that CLAUS'S views cannot receive unqualified acceptance in their present shape.

I shall not venture at present upon the broader aspects of the question, but I wish
to draw attention to the resemblances and differences between the various larval
stages of Lucifer and those of a few closely-related forms. The materials which are at
present available for a comparison of this kind are extremely scanty, for there is no
other closely-related form in which all stages, from the egg to the adult, have been
actually traced in a single species by rearing captive specimens.

Comparison of Lucifer and Acetes.

The genus which shows the closest similarity to Lucifer is Acetes, but in this case
we are ignorant of both the early and the later stages. During the last "Zoc'a " stage
the resemblance between the two forms is well marked, and is shown in such features
as the similarity in the shape of the carapace and hind body ; in the length and struc-
ture of the two pairs of antennae ; in the mode of locomotion ; by rowing with the
antenna3 ; in the presence of an ocellus ; the presence of a spine on the labrum ; the
close similarity of the mouth parts and maxillipeds ; the rudimentary structure of
the thoracic limbs and swirnmerets ; the total absence of the fifth thoracic somite ;
and the absence of the first five pairs of pereiopods. Notwithstanding these resem-
blances the differences are quite conspicuous. The eye is sessile in Lucifer, stalked in
Acetes. The shaft of the first antenna is one-jointed in Lucifer, two-jointed in Acetes.
The endopodite of the second antenna has two basal rings in Lucifer, four in Acetes.
The two lobes of the metastoma are conspicuous in Acetes, and could not be made
out at ah1 in Lucifer. The abdominal somites are rounded in Lucifer, and spiny in
Acetes; and the telson is deeply forked in the latter, slightly notched in the former.
In a word, the resemblances between the two are general rather than detailed, and
the differences are specific differences of the same character as those between closely
related adult animals. A comparison of column 2 of Table V. with column 1 will
show these resemblances and differences in tabular form.

HO ME. W. K. BROOKS ON LUCIFER:

If we assume the correctness of the extremely probable assumption that DOHRN'S
and CLAUS'S unknown larva is the earliest Protozoca of Acetes, the resemblances
between it and the corresponding larva of Lucifer (compare fig. 27 with GLAUS, fig. 2,
taf. iv.) are much greater than they are at a later stage. The chief differences are the
presence in Acetes of rudimentary compound eyes ; the great length of the carapace ;
the absence of a rostrum and spines ; the great number of joints in the first and second
antenna, and the difference in the length of these two appendages ; the deep notch in
the telson. The close similarity between the two larvae at this stage will be seen by
comparing column 1 of Table IV. with column 2.

After the moult which ends the Zoea series the differences between the Acetes
larva (fig. 89) and the Lucifer larva (fig. 53) become much greater, although they do
not obscure the fundamental similarity between the two forms. In each of them the
carapace makes less than one-third the total length of the body, and it has a rostrum
and two antero -lateral, but no postero-lateral or dorsal spines. The first antenna has
lost its swimming hairs, and has developed one flagellum in each form and two in
Acetes. In both forms a series of long plumose hairs has appeared on the inner edge
of the shaft of the appendage. In both forms the second antenna has lost its
locomotor function and assumed the adult form, but it is rudimentary in Lucifer and
well developed in Acetes.

The ocellus is present and the eye stalked and movable in both.

The fifth thoracic somite and its appendages are entirely wanting in both forms.
' The fourth is biramous in Lucifer, and similar to the ones before it, but in Acetes the
limb proper has disappeared and the appendage is represented only by an exopodite.
The second and third pairs of maxillipeds, and the first, second, and third pairs of
pereiopods are essentially alike in structure in both forms, but in Acetes the endo-
podites are rudimentary, covered by a cuticle, and functionless. The swimmerets are
present and very similar in the two forms, but the other abdominal appendages are
absent in Lucifer, while the first, second, and third pairs are developed, but rudimen-
tary in Acetes. The abdominal somites have acquired ventral spines in both forms, but
these are very small in Lucifer and long and prominent in Acetes. The telson is long
and narrow in Lucifer and short and wide in Acetes. The relation between the two
forms at this stage of development will be seen by a comparison of columns 1 and 2 of
Table VI.

The later history of the two genera can hai'dly be divided into parallel stages.
Lucifer keeps all its Schizopod limbs for at least two more moults, and as shown in.
fig. 54, acquires the rudiments of all the abdominal feet at one time, and before the
fourth pair of thoracic limbs and the exopodites of the others and of the maxillipeds dis-
appear, while Acetes (fig. 85) loses its exopodites at once, and the maxillipeds, thoracic
limbs, and antennae become like those of an adult Sergestid some time before the appear-
ance of the five pairs of pleopods ; and these do not appear together, but in two sets.

It is interesting to note that although the changes which the two forms undergo

A STUDY IN MORPHOLOGY. Ill

at successive moults do not admit of exact comparison with each other ; the outcome,
after a few moults, is almost exactly the same, as will be seen by a comparison of
fig, GO with fig. 90.

The number and character of the somites and appendages is now the same, and
while the two forms differ greatly in outline and proportion, the young Acetcs is
essentially like the young Lucifer, except in the length of the flagellurn of the second
antenna, the presence of chelre on the thoracic limbs, the presence of gills, and the
absence of a "neck." The outcome of the process of development is alike, but the
paths followed diverge from each other to converge again at this stage.

Comparison of Lucifer and Sergestes.

The metamorphosis of Sergestes is more like that of Lucifer than is the case with any
other known Crustacean except Acetes, but our knowledge of the development of
Sergestes is incomplete, and we have no assurance that the various stages which have
been described belong to the same species.

In 1870, DOHRN described a remarkable larva (" Untersuchungen liber Ban und
Entwickelung der Decapoden, No. 10, Beitrage zur Kenntniss der Malacostraken und
ilirer Larven, Part 4, Beschreibung einer neuen Decapoden-Larve," Zeit. f. Wiss.
Zool., xx., p. 607) which he collected at the surface at Messina, and which he was
unable to refer to any adult form. He proposed for this larva the provisional name
Elaplwcans. Elapliocaris is a Zoea which so far as its appendages are concerned
does not differ much from the last Zoca of Lucifer, but its abdomen is very spiny,
and the spines on the carapace are drawn out so that each one of them is nearly half
as long as the body, and they are fringed with rows of long secondary spines which
are hooked at their tips, and so arranged as to give to the body a very grotesque
appearance, a.nd the larva does not, at first sight, show any similarity to the simple
Erichtliina larva of Lucifer.

GLAUS had several years before described (" Ueber einige Schizopoden und niedere
Malacostraken Messinas," Zeit. f. Wiss. Zool., xiii., 1863) a larval Crustacean with
swimrnerets, biramous thoracic limbs, and a very spiny body, which he calls an
Acanthosoma. This same larva, or a very closely related form, had been figured and
described nearly twenty-five years before by DANA (' Crustacea/ p. 664, plate 44,
fig. 5) as Sceletina armata.

In the same paper GLAUS gives a figure of a young Crustacean, which had previously
been described by LEUCKART under the name of Mastiyopus, and shows that it is in
all probability a young Sergestes.

In his ' Untersuchungen zur Erforschung,' &c., he describes an Elapliocaris at a
much younger stage than DOHRN'S figure, and shows that this larva, DOHRN'S
Elapliocaris, his own Acanthosoma, and LEUCKART'S Mastigopus are successive stages
in the development of Sergestes.

112 AIR. w. K. BROOKS ON LUCIFER:

From independent researches in the South Pacific, WILLEMOES-SUHM also ascertained
(Proc. Royal Soc., Dec. 9, 1875, p. 133) that Elapliocaris is the larva of Sergestes, and
he traced its development through the AcantJiosoma stage, which, from its resemblance
to Amphion, he calls the Amphion stage.

These various observations, and especially those by GLAUS, give us a pretty complete
acquaintance with the metamorphosis of Sergestes from the first Protozoea stage to
maturity.

The first Protozoea (GLAUS, ' Untersuchungen,' taf. v., fig. l) has, like the Protozoea
of Lucifer, locomotor antennae, a spine on the labrum, a partially segmented hind
body, and a very spiny telson. The mandibles, first and second niaxilla?, and first
and second maxillipeds are like those of the corresponding Lucifer larva. In addition
to the spiny carapace it presents the following conspicuous differences from the Lucifer
larva. The eyes are stalked, movable, and compound. The first antenna has seven
joints. The endopodite of the second antenna has no small rings at its base. There
is a third pair of maxillipeds. Five thoracic somites are represented in the figure.
The telson is very deeply cleft. The relation between the larva and the first Protozoca
of Lucifer will be seen by a comparison of columns 1 and 3 of Table IV.

The next stage which GLAUS describes (taf. vi., fig. 1) is no doubt separated from
the first by one or more intermediate stages. The rostrum has developed a pair of
long secondary compound spines at its base, which do not correspond to anything in
the corresponding Iarva3 of Acetes and Lucifer.

The thoracic limbs are represented \>j five pairs of rudimentary bilobed buds. There
are five free abdominal somites without appendages, and the sixth and telson are
represented by an unsegmented region, which carries a pair of long bilobed pouches,
the rudimentary swimmerets.

The relation between Elaphocaris and the corresponding larvae of Acetes and Lucifer
may be understood by a comparison of column 3 of Table V. with columns 1 and 2.

In the next or Acantliosoma stage ('Untersuchungen,' taf. v., fig. 6) the two pairs of
antennae assume the adult form, and the thoracic limbs and swimmerets become
developed as they do in Lucifer, and the carapace loses its posterior spines, although
there are three in place of one pair of anterior spines. The telson is distinct from the
last abdominal somite, and all the abdominal somites have projecting spines.

The eye-stalks are much longer than they are in the other two forms. The first
antenna has a secondary flagellum, as in Acetes, and the scale and flageUum of the
second antenna are well developed.

The exopodites of the maxillipeds and pereiopods are very long, many-jointed, except
in the first rnaxilliped, and they are longer than the endopodites in all the pereiopods.

The fifth pair of pereiopods are present and like the others. The swimmerets are
very long and slender, and the telson very short and forked.

This stage, like the corresponding stage of Lucifer, and unlike that of Acetes,
persists for more than one moult, and the five pleopods make their appearance

A STUDY IN MORPHOLOGY. 113

together, as rudimentary buds, before the exopodites of the pereiopods and niaxillipeds
disappear.

The third column of Table VI. shows the resemblances to Lucifer and Acetes at
this stage.

In the immature or Mast'ujopus stage (see CLAUS'S ' Ueber einige Schizopoden und
niedere Malacostraken, Messinas ' ) the three forms are almost exactly alike, except
as far as the generic distinctions are concerned, and the young Seryestes scarcely
differs from the young Lucifer except in the absence of a neck, the length of the
flagellum of the second antenna, and the presence of rudiments of the fourth and fifth
pairs of pereiopods.

Comparing the whole course of development of the three forms, as far as it is known,
we notice that while the larval stages of Seryestes are much more different than those
of Acetes from the corresponding stages of Lucifer, the character of the change at
each moult is much more like what we have in Lucifer than what we have in Acetes.

We cannot fail to notice, in the second place, that the attempt to express the facts
of the metamorphosis of these forms, so far as we know them, in a tree-like diagram,
would result in a tree placed upside down, with the branches which represent the
three Protozoeas much more divergent than those which represent the three young
Sergestids. A similarity of type runs through the whole metamorphosis, but it is
no more marked at the early stages than it is in the late stages, while the secondary
differences are much more conspicuous during the Zoea and Aeanthosoma stages
than they are as we approach the adult form.

While this is true it is also true that if we imagine a metamorphosis which
shall agree with these three in all their common features, but shall have none of the
features which they do not all share, we shall have something much more like the
metamorphosis of Lucifer than that of Acetes or Sergestes, and we must therefore
regard the life-histories of these three forms as somewhat divergent modifications of a
form of development which is at present more closely adhered to by Lucifer than by
the other two, and in this metamorphosis we must recognise a Protozoea stage when
the two pairs of antennse are locomotor, the ocellus present, the labrum furnished with
a spine, the carapace armed with posterior dorsal and lateral spines and a rostrum ;
the two pairs of maxillse, and two pairs of maxillipeds present, and the thoracic and
abdominal segments without appendages. This stage persists, with slight modifi-
cation, through several moults in all of them, and is followed by an Aeanthosoma stage,
in which the carapace has a rostrum and antero-lateral spines, and a smooth posterior
edge ; the eyes are stalked ; the two pairs of antenna? have their adult character ;
there are at least four pairs of pereiopods with swimming exopodites ; the swimmerets
are large and have their adult form, and the other abdominal appendages are absent.
The duration of this stage and the mode of transition to the next varies in the three
forms, but it is followed in all by what may be called a Jfdstigopus stage, characterised
by the general features of the family.

MDCCCLXXXII. y

114 MR. W. K. BROOKS ON LUCIFER:

In all three forms the somites, and with the exception of the swimmerets the
appendages also, develop in serial order from in front backwards.

The interesting question whether we are to attribute to this typical form of develop-
ment a fifth thoracic somite and appendages must, I think, be left in doubt. A
comparison of the Sergestid larvae seems to indicate its absence, but wider comparison
with Penceus and the Schizopods seems to lead to the opposite view.

Comparison of Penseus with the Sergcstidce.

In order to render the comparative tables as complete as possible, I have added
columns showing the corresponding stages of Penceus and Euphausia.

FRITZ MULLER has described a number of stages in the development of a species of
Penceus (" Verwandlung der Garneelen," Arch. f. Naturgeschichte, 1863, pp. 8-23,
taf. ii.). The series commences with a Nauplius which may belong to the same species,
although we have no certainty of this. In a second paper (" Ueber die Naupliusbrut
der Garneelen," Zeit. f. Wiss. ZooL, xxx., 163-166) he gives, in reply to doubts which
had been expressed to him by SPENOE BATE, ALEX. AGASSIZ, PAUL MAYER, and
others, the following reasons for believing in the specific identity of all the forms in
his series : — 1st, the peculiar mode of locomotion ; 2nd, the resemblance in colour ;
3rd, the great length of both pairs of antennae ; 4th, the character of the mandible ;
5th, the presence of four pairs of buds in the Nauplius, and four corresponding pairs
of limbs in the Zoea ; 6th, the similarity in the structure of the heart, digestive tract,
and liver in the Nauplius and the youngest Zoea ; 7th, the presence of frontal organs
in both stages. As all the points except the colour would apply to any Crustacean
which passes through a Protozoea stage, there is certainly nothing more than a pre-
sumption that the whole of his series represents a single species ; but as there is no
doubt that the Nauplius belongs to Penceus or to some closely -related form, I have
included it in the table.

FR. MULLER'S account of the later stages is supplemented by a few additional
observations of other species by GLAUS (" Untersuchungen," &c., pp. 11 and 41, taf. ii.
and iii.), and I have compiled the columns in the tables from both sources.

The first Nauplius stage (Table II., column 3) appears to be more simple than that
of Lucifer, as MULLER failed to observe any buds to represent appendages posterior to
the mandibles.

The Nauplius stage is followed by a meia^-NaupUiis stage (Table III., column 2),
which is distinguished from that of Lucifer by the great size of the blade of the
mandible, by the presence of frontal organs, and by the shortness of the carapace.

The next stage is a Protozoea (Table IV., column 5), with a rounded carapace
without spines or rostrum, four basal rings and six terminal joints in the endopodite
of the second antenna, a spine on the labrum, two pairs of maxilke, two pairs of
maxillipeds, and a long hind body which, according to GLAUS, is divided into six

A STUDY IN MORPHOLOGY. 115

thoracic and five abdominal somites, and terminates in a deeply-forked telson with
seven pairs of spines.

This stage persists with slight change for several moults, and at the last the buds
for the thoracic limbs and swimmerets appear. According to GLAUS, the rudiments of
all the abdominal appendages can be seen at an earlier stage.

The passage from the last of the Protozoea series to the first Schizopod stage is
attended by a complete change in the structure of the antennas, and these now assume
the adult form. The carapace also acquires two antero-lateral spines and two more
at the base of the rostrum. At this time it is much like Lucifer, as shown in column 4
of Table VI., but the endopodites of the third pair of maxillipeds and of the pereiopods
are rudimentary, and shorter than the very long-jointed exopodites.

The significance of the various stages in the metamorphosis of the higher Crustacea
is one of the most interesting questions in the whole field of morphological science,
and it has given rise to at least its due share of speculation, but it will not be out of
place to examine the relation between the facts which have been described and the
various theoretical views which have been expressed upon the subject. In the case
of the Sergestida? it is obvious, in the first place, that the adult Lucifer and Acetes
also, if Acetes be an adult, are little more than mature representations of the Masti-
gopits stage, complicated in the case of Lucifer by the formation of a neck, and in the
case of Acetes by the presence of gills, and chela) on the pereiopods. There can also
be little doubt that the Schizopod stage of development in the Sergestidse and Pcnceus
bears a similar relation to the adult Schizopods, especially to Amphion, the adult
character of which seems to be established by WiLLEMOES-SuHM's observations (Proc.
Roy. Soc., Dec. 9, 1875).

The significance of the Zoea stage in the higher Decapods is one of the most vexed
points in Crustacean morphology. We have shown that in the Sergestidse and in
Penceus the so-called Zoea stage is nothing but a preparation in the Protozoea for the
next or Schizopod stage ; that it involves no changes of structure except those which
are related to the form which it is to assume after the next moult, and that the Zoea,
as a distinct stage, is absent. The life-history of these forms would therefore lead us
to suspect that the Brachyuran Zoea is a secondary modification of the more primitive
Protozoea, and we may perhaps see in the larval skin which many Crdb-Zoi:as shed
soon after or even before they leave the egg, and which usually has a conspicuously
forked and very spiny telson — a remnant of the unmodified Protozoea stage.

DOHRN (' Geschichte des Krebsstammes, Jenaische Zeitschr.,' 1871) and FRITZ
MULLER ('Ftir Darwin') have held that the typical Zoea, with segmented abdomen
and suppressed thorax, is the ontogenetic recapitulation of an ancestral form which
has formerly existed as an adult, and DOHRN even goes so far as to recognise the still
more remote ancestor of this Zoea type in an embryo (" Untersuchungen tiber Ban
und Entwickelung der Arthropoden; eine neue Nauplius-form : Archizcea gigas," Zeit.
f. Wiss. Zool., xx., 597), which WlLLEMOES-SuHM has recently shown ("On the

116 MR, W. K. BROOKS ON LUCIFER:

Development of Lepas fascicularis and the 'Archizcea' of Cirripedia," by R. von
WILLEMOES-SUHM, Ph.D., Proc. Eoy. Soc., Dec. 9, 1875, pp. 129-130) to be the
Nauplius of a Barnacle, in all probability Lepas australis.

GLAUS, on the other hand, believes that the Zoea has no such ancestral significance
(" Untersuchungen," &c., p. 31). That it has been formed by secondary modification
of the Protozoea, and that the views of MULLER and others, that the Zoea presents a
picture of the remote ancestor of the Malacostraca, is fundamentally erroneous ; and
not only this, but that the Proto'.oca itself is the result of the extreme secondary
modification of an ancestral form which GLAUS proposes to call an Urophyllopod, and
which he believes to have had the following characteristics (" Untersuchungen," p. 23) :
A greatly developed shield-like carapace, produced by a fold of the integument in the
region of the maxilke, and probably armed with median and unpaired spines ; two
maxillary segments and appendages, eight somites of the mid-body with appendages,
and six abdominal somites with swimmerets and telson ; a many-chambered heart ;
compound eyes, probably stalked ; a first antenna with sensory hairs ; locomotor
second antenna?, in which the exopodite was probably a scale ; the mandible probably
lacked a palpus ; the metastoma was represented by a pair of paragnathi ; the ruaxillre
had their basal joints modified for mastication, their endopoclites reduced to a jointed
palp, and the exopodite modified to form a scoop or scaphognathite for regulating the
flow of the respiratory current under the carapace.

The following eight pairs of appendages were more like Schizopod feet, and each of
them carried a basal gill-plate ; the six pairs of abdominal appendages had large basal
joints with two branches and gill- plates.

GLAUS believes that we may recognise in NeLaUa, which has stalked eyes, a scale
on the first antenna ; only one long flagellum on the second antenna ; a mandibular
palp ; a highly specialised, long jointed endopodite on the first maxilla ; two long
limb-like rami on the second maxilla ; eight pairs of phyllopod-like thoracic limbs with
jointed endopodite, flat, spiny exopodite and gill ; six pairs of pleopods, the last two
rudimentary ; and a seventh somite between the sixth abdominal somite and the
deeply-forked telson (" Ueber den Bau und die Systematische Stellung von Nebalia,"
Zeit. f. Wiss. Zool., xxii. p. 323-330), a very slight modification of this ancestral
Urophyllopod.

He gives on pages G9-71 of his "Untersuchungen," &c., a long, minute, and
extremely ingenious explanation of the way in which this Urophyllopod stage of
development became converted by secondary modification into the Malacostracan
Protozoea, and afterwards, by still greater modification in the same direction, into
the typical Zoea of the higher Decapods.

The facts which have been detailed and tabulated with reference to the metamor-
phosis of the Sergestidre and Penceus seem to substantiate at least a portion of this
view, and to show that the typical Zoea is a secondary modification of the Protozoea ;

A STUDY IN MORPHOLOGY. 117

but a comparison of these forms with the metamorphosis of Euphausia, upon which
GLAUS lays especial emphasis, seems to demand a directly opposite interpretation.

If the Zoea has been produced by a secondary modification of the Protozoea we
should expect to find the characteristics of the Protozoea better preserved in the
Schizopods than in the lower Decapods, and if we find in the Schizopods certain
features of the typical Zoea, which are absent in the Protozoea of the lower Macroura,
we can hardly accept without question the interpretation which sees, in secondary
modification of the latter, the origin of the Zoea. In Euphausia the somites appear
in regular succession, from in front backwards, but the somites of the abdomen
acquire appendages before the pereiopods appear, and there is a stage when the
abdomen is fully developed and the thorax almost absent ; a stage which, therefore,
resembles the Brachyuran Zoea more perfectly than any stage in the development
of Lucifer, Acetes, Seryestes, or Penceus.

We have no complete history of any one species of Euphausia, but the observations
of METSCHNICKOFF (Zeit. f. Wiss. Zool., xix., pp. 479-481, and xxi., pp. 390-401),
and GLAUS (Zeit. f. Wiss. Zool., xiii., pp. 442-454, and " Untersuchungen," &c.,
pp. 9 and 33) give us a tolerably complete account of the metamorphosis of the
genus.

METSCHNICKOFF'S larva is extremely like that of Lucifer, although there are many
differences. It is interesting to note that it leaves the egg in a much more rudi-
mentary form, passes through a greater number of moults, and attains to much greater
structural complexity than Lucifer during the Nauplius stage. We can select
three stages which agree pretty closely with the egg Nauplius, the first free Nauplius
and the last, or mein-Nauplius, of Lucifer, but between, after, and before these stages
there are others which are not found in Lucifer.

The youngest Nauplius (Zeit. Zool. xxi., fig. 2) is so much less advanced than the
egg Nauplius of Lucifer six hours before hatching, that it does not seem probable that
it normally leaves the egg in this condition.

It has an oval body, without ocellus, mouth, or labrum, and there is no trace of
more than three pairs of appendages or of the carapace. At the next stage the
swimming hairs of the first three pairs of appendages are fully developed, and the
anus, notch, and two spines of the telson are present. In these respects it is more
advanced, but in the rudimentary condition of the labrum and metastoma less
advanced, than the first free Nauplius of Lucifer. The buds for the first and second
maxillae and the first pair of maxillipeds are present, but continuous across the median
line of the body. According to METSCHNICKOFF, the larva shown in fig. 3 of his first
paper is in the next stage of development ; but I can scarcely believe that it belongs
to the same species, for the ocellus is absent, and the hairs on the three pairs of
locomotor appendages are much more rudimentary than they are in fig. 3 of the second
piipa.

The next stage (fig. 4) of the second paper agrees with the first free Nauplius of

IIS MR. W. K. 15BOOKR ON LUCTFER :

Lucifer, so far as the form and number of the appendages is concerned ; but the last
pair of buds are biramous, and the carapace and telson are well developed. The next
stage (fig. 5) of the second paper is more advanced in nearly every respect than the
second free Nauplius or rciQizi-NaiipHus of Lucifer. The mandible is rudimentary, but
still bilobed, with no trace of a blade. The outline of the carapace is free from the
body, and its anterior and posterior edges are spiny. It has frontal organs, and the
basal joint of the second antenna carries five recurved hooks.

According to the author, figs. 2 and 3 of the first paper show the next stage ; but
the structure of the hairs on the antennae, the fact that they are plumose, and the
very deep notch in the telson, seern to indicate that this is another species. However
this may be, the structural complexity at this and the next (first paper, fig. 6) stage
is much greater than we find it in Lucifer at the end of the Nauplius series.

It will be observed that, while METSCHNICKOFF'S larva and the Nauplins of Lucifer
are essentially alike, there is at no time an actual agreement, since certain structures,
as the carapace, become developed earlier, and others, as the labrum, later than they
do in Lucifer ; and certain structures, as the frontal organs and the hairs on the base
of the antennae, are entirely absent in Lucifer.

In column 4 of Table II. I have compared fig. 4 of METSCHNICKOFF'S second paper
with the first free Nauplius of Lucifer, and in column 3 of Table III. his fig. 5 with
the last Nauplius stage of Lucifer.

The various Protozoca stages are shown by GLAUS in plate 1 of the " Unter-
suchungen," &c. The early Protozoca (Table IV., column 5) is much like that of
Lucifer, but the carapace is serrated, there is only one pair of maxillipeds, and,
according to GLAUS there is a fifth thoracic somite. In the last Zoca stage (Table V.,
column 5) all the abdominal somites and the rudimentary swimmerets are present, but
there is no trace of the second and third pairs of maxillipeds or of the pereiopods.

Up to this point the course of development has followed essentially the same line
as in the Sergcstidce, but, as we should expect, the Protozoca series is not followed by
a larval Schizopod stage, but by a series of moults during which the adult characteris-
tics are gradually acquired. In the loss of the posterior spine of the carapace, the
acquisition of antero-lateral spines, and the change in the antennae from the Nauplius
form to the adult form, the moult is like that of PCIKBUS and the Sergestida? ; but the
second and third maxillipeds and the pereiopods appear one at a time in succession
from in front backwards, and the abdominal feet appear before the pereiopods. There
is no Zoca stage it is true, but the course of development differs from that of Penaus
and the Sergestidas in the very feature in which the larvaB of these forms differ from
a typical Zoca — the irregular manner in which the pereiopods appear.

I am therefore unable to give CLAUS'S interpretation of the significance of these
larva? unqualified acceptance at present, and feel that our groundwork in this depart-
ment of knowledge can be made sure only by new observations. Every naturalist
who can trace the whole life-history of a single species of any of the genera of lower

A STUDY IN MORPHOLOGY. 11'.)

Malacostraca by actual moults, will not only help us to a sound and thorough appre-
ciation of the significance of Crustacean embryology, but will also contribute to a
better knowledge of the relation between ontogeny and phytogeny in the whole
province of biology.

The phylogenetic significance of the Xauplius stage of development seems to me to
rest upon a much firmer basis, and there are many reasons for believing that this is
really an ancestral form. Its occurrence in so many widely-separated groups of
Crustacea shows its great antiquity, and if it does not represent the adult form of the
ancestral Crustacea, but a later larval form which has been produced by secondary
modification of the original course of development, this secondary modification must
have taken place very early in the history of the group, at a time when the adult
forms were very primitive and unspecialised. A sufficient difference between the
habits and surroundings of a young animal and those of the adult to favour secondary
modification of the young is much less probable in an early unspecialised form, with
simple habits, than it is in later and higher forms ; and the older a larval form can
be shown to be, the more probable does it become that it at one time existed as an
adult.

The great age of the Naiqrfius stage and its definite structure therefore indicate
that it is ancestral, and nothing except the supposed necessity for believing that the
primitive Crustacean had a great number of somites and appendages seems to oppose
this view.

I shall try to show further on that the serial homology shown by the parts of the
body of one of the higher Crustacea cannot be fully accounted for by assuming, with
BALFOUR ('Comparative Embryology,' p. 418), that the primitive Crustacean had, in
addition to its three pairs of appendages similar to those of existing Nauplii, a long
segmented body with simple biramous appendages ; and I shall also try to show that
this homology can be accounted for without any such supposition, so that the
peculiarities which BALFOUR points out — 1st, that the mandibles have the form of
biramous swimming feet ; 2nd, that the second pair of antenna? are biramous swim-
ming feet ; 3rd, that the body shows no traces of segmentation ; 4th, that the heart is
absent ; 5th, that the ocellus is the sole organ of vision — must be allowed their full
weight, and must not be opposed by any d priori assumption of the theoretical need
for a greater number of somites and appendages.

120

UK. W. K. BROOKS ON LUCIFER:

8

•§

8

I
1

03

of

S

^

e"

'%

1

^

I

<xT
S

«
4

o

PH

I

0

•3

SH

o

«rf

PH
o

3
*s

Ol

1

O)

03

o-

I

.

d o> -

OJ co

§
a

I

-

2 CD

§
s

1

-s §
•«„ », a

. . H

'S 1

.2 a
g S

12

o -e

be •:
o

A.

BA

S).

pus

Free Nauplius of A
fig. 208 from GLA

O •£

a I • ' z

z I • :!;:::

X ^o w '

a ^ cu o

* S a . 4=

^

1 'I3"! s -i

% 3 .s « • g

w _.-r3 £, • S

3

a ^

O2 O2 <1 M O2 PQ t> O2 S

. . . . _r t-t cT • • • • • • • • • • • • '• • • "

^ O fe rrt

' ' • -HI | •', |

• • • -&° I «• ' • --9 1

.... ^J § o . . g-g «,

O ^ •— H ^- -Q i— ^

f-i rr* T ^T? (-)

• • • • P Is T3 o" • • • -13 -g o

. . . ill ? ^ •§

•^ 03 «S C3 ,** ."

O OQ Q i— i t^i ^^ P% t>^ n^

1 • • • a" s "7 >> l«!«^i« sS

p^aTO'r^ -M-U-*^-M -a

• •• otr§'T3^ . ccaa.. s-S

•*•* El n S *3 ffl CDCDOJCD 3 M*

sssll-^2^ ° JJJ.l'I'islsslisssssJ"

llil'aS.SS'o i p|'|'erT3'»SSStDSSiSS»Sai<B^s

ja2O202PHh3 PP S O: |"

i . . . ; '. : . : ^ ~

'£

-ts* or

... . , . ^i 4_1' ^ ^j _t^ ^ ^3 ^ ^J ^i ^^ -4^ ^i 4^ »^ ^1 <— ^

QJQJQJQCJ 03 3J ^-^"^~"KJ^"WHHH3*-CMHj«RQ)pJ

!! r4 S A a rt i-^ Qcl3C3G3fl3OOCLJ0303QJ030303QjOWg3 .—

^ — — — — H • C tnXXCQt/JtOXWKCCKXODCOWt/ltQQ Q '

sSc«jsc3 s s ^^-^^s^^S^^^^^^^^^ii;31

'C IT? O m T3

S • J'S a J

a a — '&, fl

'3 -5- m ;s

B. 3 H. s ^ b b b b

^s; fe"S csscsos

s ^ »r3^ £ $ -s -s -g -g '

* * § ^ § 1 1 JL^^^^,,^^ *

& |3||||sgssgggggggg|g g

rtl W tT'<rH """ t- *^ t!r ^ C ^^ ^^ ""^ '^ ^Jt/;^;cOWICy3McDC/}tBC/J(KC/l tC

j52rq^fi^.;3F°;3 .SaSPpS

15^ .^

a

'O 5

o K

i I §

° ,c 'a

ft, 02 (in 0! &H fc 02 EH

2 EH fe

A STUDY IN MORPHOLOGY.

121

o

02

-p

s

OJ

O

M

MDCCCLXXXII.

Second free Naupl'us of Ajmn (BALFOCR,
after CLAUS).

Same
Same
Same
Elongated backwards
Same, but shorter than second

Biramous, locomotor, obscurely
jointed, has blade
Very large
Not mentioned
Same
1

O

CD

3

S
*.

| segments

DO

09

1

^

Present

Meta-ATrt«/>7i«s of Euphaus'a
(METSOHNICKOFP, Zeit. Zool.. xx., taf.
xxxiv., tig. 5).

03

Present

"S

'3

o

•i-S

o

CO

0
cc

' 'Eb

O O Qi G7 O

a a s a a

£ ce t. cS 03
O3 O3 PH O3 O3

Biramous, locomol
blades

a

;|

. -*^>

a

<D

"1
^Pn

EQ

OS
A

. -+3 -^ -^ -43 O>
§ § § g g g g-g

1 1 i^j^^^

»coco<l<l<1^s

Meta- Nauplius of Penceux
(MiiLLER, Zeit. Zool., xix.,
fig. 2).

spines
Present

0

,g

Biramous, locomot

Moderately large

*

0> 0)

a a

C3 03

Same .
Well marked

\
£

CV.

-a

• • • - g. • •

03

. . • • S -M -«

m m m S £ rt
• If fill

x. O3 02 O3 PH << •<

' ^

•3

1

Ceo

^
11

t!
ij

CO

11

g-g
£

ted, locomotor

irs
;ed, locomotor

03

. -a

O

O-.

' OJ

. 3

^

^

Absent
Present .

Not observed .
Indicated by fold

Uniramous, unjoin

long, sensory ha
Biramons, unjoin
no blade

<c

'ft

M
5

3

S
-

T3

I

c

<£

^

2

Rudimentary
Rudimentary

Rudimentary

Absent ? .

Absent .
Absent

' S
"5 ^_

go
P

•5

Third luaxilliped
Thoracic somites

Abdominal somites
Telson ....

t • a ' C3

o S "3 § i
£••5 -S S-Jj

IH 2 11
o o fa o S

Second antenna

Labrum

Metastoma

| a'§ a
1 1.1 1

fa CO fa O3

K

MR. AV. K. BROOKS ON LUCIFER:

e

5.
Protozofa of Euphau
(GLAUS, ' Untersucliung
&c., plate i., fig. 2).

1? s?

be n r

1 s s

OJQ

r-|« c/; 02

CL. ~ ^

" I Ij

2 ;e &

J re1 o

r§ a K~-2"CS«caortac€cScicSc2O -°™

CO O2 K2 P3 OQ COCOCOCOCOCOCOCOCO92CO PH

a; . . ... .... *•• & o -^ £

w o 3 ?- >> i -^ K o

.... g 60 g |C . . . . ^ * g >

....§« a | £ 1 .a3

.tAcu ^ . 'S PT1 . . « ^ ' — i^^'TScc

rrt C3 (U J3 CC O -O m CLI

O •"-• ^>i v" "tf ^ U © _y G

L5 • • • -° F S s?2 cf'g « -° s 'a

^ . 10 3 § * K'

§-^/30r32 §"5 -i«3 fe" to" « . °

a'°'fcS"° ' oj ' ' cuoai^§'':aJaJcj"ci^!::'"3^'"'£

l|i-iJ|ai | | .• | lii'iliillli-alJii*!

^ co PH • H<M~tJ "co co co. co cococofScococO'COcococo' O

. - , .CM^K. . ... ^ ^

Snfl DQtDOOtn '^

' •=• S 'S, ' ' ij; S: fco bo be =

• • • • bt O _ • • ... ~3 rQ rg rQ r^ — , - ^

scxos acCdo-i" C£H

cD^co n CQ CD CD O ijt ' r-

r s° ' ' § '§:&§:&§:§ 1 &

• • • • "••' C°x Q,?3clGGc:~2 ^'o

S"coc"c S-£°°°22cB Pco

- : : :i?.|i : t : ^JliiJJJit it

_M Ci r. _b* m ^ OJ ^ ^i flJ ® 03 fl3 ffL.B «"S «"S -"S ^H <"rH ®*_fl

^ e • • § o -§ g a s^a aaa^^aaaasa^ s .s

JP S ^2 -° — £ S tS^S Sssi^SSooSSo vs as

S co co MP5co cococo^JpHCOcococococo co

I CO . f^ • • r- CO

CO M Q *^ ^

y . :3 &D '3

&c-e csaaaaS ^

.- S a .9 • M .2.2.2.2.2.2"^ «

T3 g r2 .°- 4; sc'So tc &o so Si,-1-' -53

u . . .... o - a/ a> cu cu a; j.-

I

8"

O

8

so

O

&H

I
D

-t-j

O

to

H->

S

S
-P

4w

O

Qj

J
_>

'-3

I

S

6

3.
zofa of Sergestes
'Untersuchungen,
plate v., fig. 1).

P
A

D
&

2.
CLADS'S larva
Untersuchungen,' &c.
plate ir., fig. 2).

i>

g '

02

6C"

s ra

c- B

fi rr-;

„- S •

111

02 P3 CO

a> o>

Q

u P

aaaaaa — 's

So 6c be be be bo aT.S

p, CO

o ««_.

0°

-£=!;

"- CUD

. e B 5

:i!= i

n ^i T"

ccacaosoi

<; a, . PH HO,

g .2

*

pq

11111

i

Notched, four pairs large
spines, one pair small ones

-

OOE^CBCJ

s

0

02

iil

a

o

A STUDY IN MORPHOLOGY.

123

eo"
«3

so

S?

8"

zoea of Euphausia
tersuchungen,' taf.
fig. 4).

1

OQ GO DQ OQ m

CJ Q; OJ 03 G3

iC bD be 60

« OS CD CB

be

Go

§^

'S "Sflfl (ddflflrt

*pi ^~ * QJ o o o o oi o

ET* r^ p i _ p t p, p l pj p^

gi S CL, £l< P- CL, CH PH P-4

.^3 =3c3c3c3ric3c3c3

.r§ § _§O2O2222

"g g '§ § £ fe ^ ^ ^ ^ ^ ^ ^ ^

d ~on KJ «S fl .-"•oraToro.)*^^^ ^

. 2 oj o c> O- o oT -t> aj • i ."t^ -"S ."S -t^ ."t- -"t^ ."S o QJ as o a> a.) . ££

^|S J | S | II 1 111 I i I 1 I i 1 1 I oi H I

PH co % h3 co l^coco co P3 co co cococococo cocococococo co

• ^^t«. ... .... .

PH g.SS

^.QpPcS.

J .| .^.~ -g

-^ 'Lilt I I

03 l— 1 r- H C^ ,,03

M33 OS, 0!a)03 0303 J> rj 033)03330) 0)0)03333333 33

3 1 e '&£ -1 .8 §aa 1 1 g5! aaaaa aaaaaa a

-^c^^a!1^-?'^ 2"^^ dcecS-^-H c^rfrtc-pcs c3rfc3dcec« c3

M«2^i hJ h-5 WCCC/.) CQ CQ (.....I W t.Orcc.O^C.Q WmwitfiWT/} &

•13 ^ -^ o ... . . . , .

-*j S ^

1 ' 5 -g -S . -2 ' ' ' ' I 6

P< O -PH CO •- a r-,

be • a '^ f M « • • • ^ -n •

aoaao -« g a S

O*O ^ "*^ *" '"^ ... ..«3 »...^HQJ ' ^

f O 1 I ^"3 ^1 33 O B««pf'a««««/^«««««^ ^

III | iiioli IlifiJIiii'liiiilii §"

COCOCO CO CO IZnCOCO COCOi-^Co'^COCOCOCOCO'^COCOCOCOCOCC Q

1^1 : :T: :::: :::::" : : : : : : :

CQ

-_T ' "Ti

« ^

M 0 o> oT "S o 0)0503 C3<u®o> OJQJOQJO; ajojoajfljo.! ?2

311 I !® I 111 1111 I I I § § llllll ^

^MC...! OQ '^/2 OQ02O2 OQOi^C/2 CQ GQ OQ O2 QQ OQo5^C.OC:Oc2 fe

<rrf '"A •>.. WOIMCG" '

So) x Qjocta. O

. . S ^ - - • • o .... bo ic &c be ^> g^ .

6®—; §SojQ;.-Q

' £ ^ ® " ••'w .... PnO-i&HC-i^ g jj .

-4J J* -4-2 O — - — ' — —i /

* " O -J" *" " **fiflj*'"" " fcn Q Qi QJ i) QJ '—* ^2 "

M . _?Tl.;i.'— fl3 . , . ?H tn tj SH r^ Cj ffl C€ C6 Cj t-^ .

?•§"•« .-3 ^ 'g-:°^M|^ Illlrf ^^ff^lf

1

•

&H
-P

J
pS

O

S

HP

I

s

o
O

P3
<)

H

Is i^

2I.J

ill

j:
a

SPSS'
1-4

b" M
S"3
S»

p^ 03

otozor
long

oa ~ ^
"*•* x -S

9-H B^

Wfflcopq

aaaaa

OOOOO

coco co co co

aaaaa a Jc^

OOOOOOr3°

cocococococo s

-2

• a

13.1

a-s g

0 0 c3
OOO

1

CQ

^ A^ SH^
£..2 ^.2 PH

a

ll

a S a s

~ rr-

8l8i

0) .K O) pM

CO fe COH

124

II R,

K. BROOKS ON LUCIFER:

03

9P

- -

P°

"

0) OJ 0>

w

&, 0,

o

PH

ooooo

MMKMX
CD 0> 0> 0) 0)

03
i-i

--G
03

a

Scl
' Ve

.

•

-

111?

02

3 QQ DQ CC QD

S 3
-

be

a g «

a S S

t- -M a

X r^ £1,

•s •§

•s

-3

o
S

^

r-=H

nd

o

OH

O

N

O

?j

O
OQ

03

O
03

"s 8)

Afantho
' Uuters

o
• '

i'lili i

03 H» M 02

co x x

sJg

cc to

S) •

bo 0)

a p

'

<D

be

o
a

js
>>

be

n

o

be
c
c

be
s
o

I

I

H

' ' -

a a a

"8

. -^

•~S. "S -^ "X •"£ "o i§ !S 5 3 'a

&l

*H ^ pM ,

=* o <*
f> 9 s.

K

3 3 a • • • g

£ h h s

in vi w • • * aT

O 0} <D r^i

be bo be .S

ce ce cs • • ' >

rO 'o T3

O Q3 O 0 ^ "Ir*

g pL,p,p,g g | g

co <1 <^ <j x co co x

I •§

11
2

^ a o3
•a 'o .-S S

a= ^- S

-

§-

1-

§•

§•

•

•

rT - -u T3 -^ T3 -^ nS -" T3 ."S

5aa

- .,

^3 -+3 -*-- -t^ J-3

S 2 C S d
0 (D OJ O ^

__

'

H

i
*— i

63
J

pq

-
PnCO-ln

,

D-i K

^=^:j3^3^:^o ooooonSo
CO CO CO CO COCOCOCOCOh^J

,

111

OOO

CO

'saa

" '

rS: (if CO |5q CO

1

CO

o

-s

I

fe CO

ll

H 5n

CO EH

A STUDY IN MORPHOLOGY. 125

VII. SERIAL HOMOLOGY AND BILATERAL SYMMETRY IN THE CRUSTACEA.

The Phyllopods and the highest Brachyura are connected with each other by a
tolerably complete series of intermediate forms, and as we pass this series in review
we cannot fail to notice that, as has been so frequently pointed ont by morphologists,
each successively higher form is a little in advance of the one next below it in the
degree to which the functions and structure of the somites and appendages are
subordinated to the individuality of the organism as a whole.

In the lower forms the body is made up of a series of nearly similar somites, and
the appendages, with the exception of those at the anterior end of the body, are
essentially alike in structure and their functions are indentical throughout the series.
The greater part of the body of such a Crustacean as Artemia consists of a series of
similar somites, and in Aptts we find more than sixty pairs of limbs which agree with
each other so perfectly in function as well as in structure that any one of them might
be substituted for any other without involving any essential change in the structure
of the animal as a whole.

At the other end of the series we have Crabs with the primitive distinctness of the
somites so obscured by the centralised individuality of the whole organism that it
cannot be traced at all without careful study and comparison of various stages in the
life of a number of forms.

Comparing the various appendages of a Crab with each other we find that their
functions are not at all alike. The mandibles are nothing but masticating organs, and
the power which they once had, and which they still retain in the Nauplius of Lucifer
to aid in locomotion, has entirely disappeared.

Other appendages have become organs for procuring food, or weapons of offence or
defence ; others have become walking legs ; others long oars or paddles ; others again
have lost all limb-like functions, and are changed into accessory reproductive organs ;
whilst others again have entirely disappeared.

In accordance with this specialisation of each appendage to a particular function, a
corresponding structural change has been brought about, and it is only after careful
study of the younger stages that we perceive the mandibles, maxilla?, foot-jaws,
walking and swimming legs, and copulatory organs of an adult Crab to be as strictly
homologous with each other as are the unspecialised appendages of Apus.

The integration of the somites into a centralised whole has been accompanied by a
differentiation of each appendage from the others, and a specialisation to a restricted
function.

An adult Crab resembles and differs from one of the higher Macroura in about the
same way that it resembles and differs from its own Meyalops larva, and the transition
from the larval form to the adult form is accompanied, like the transition from an
adult low Crustacean to a high one, by increased dependence of the various parts on
each other, by the increased prominence of the general individuality over the indivi-

126 MR. W. K. BROOKS ON LUCIFER:

dualities of the somites or metameres, and by the increased structural and functional
specialisation and differentiation of each appendage as compared with the others.

This series of changes is so well exemplified by the study of adult and larval
Crustacea ; it is so remarkable and interesting ; so very conspicuous and unquestion-
able, that it has long attracted the attention and called forth the speculation of
morphologists. It is natural to suppose that the process of change which is open to
our observation through study and comparison of living Crustacea, is a continuation
of a, similar process which went on in the remote past. There seems then at first
sight to be reason for believing that, if we could go far enough back, we should find
the individuality of the whole organism gradually disappearing and giving place to
the separate individualities of the component somites ; that we should find the
specialisation of the appendages gradually disappearing, until we should at last find,
as the remote ancestor of the Crustacea, a series or community of independent
organisms, each one essentially like the others, and able to provide for its own wants
and to lead an independent existence when accidentally or naturally detached.

This view has been advocated at length by H^ECKEL (' Generelle Morphologie,'
1866) and by SPENCER ('Principles of Biology,' vol. xi., 1867), and used by both
these writers as an explanation of the origin of all segmented or compound animals
and plants. It has been accepted, with more or less qualification, by many other
writers, although HUXLEY (' Oceanic Hydrozoa :) and METSCHNICKOFF (Zeit. f. Wiss.
Zool., xxiv.) have pointed out that, even in the Siphonophorte, where the individualities
of the units in the compound are extremely well marked, the view that the organism
has been evolved by the gradual integration and specialisation of originally inde-
pendent Zooids is attended with serious difficulties.

So far as we can see there is no reason why the Crustacea might not have originated
in this way, by the gradual integration and differentiation of a community of inde-
pendent metameres, but the evidence which is attainable seems to directly oppose the
belief that this has actually happened. We are able to trace the higher Decapods
back, very satisfactorily, to a Phyllopod-like ancestor with a long series of undif-
ferentiated somites and appendages, but even here the somites are simply parts of
the body, and they furnish no more evidence than those of a Crab to show that they
ever were the independent organisms of a community.

When we attempt to go still further back we find that the facts of embryology, if
they show any thing whatever about the phyllogeny of the Crustacea, lead us back to a
Nauplius with three inter-dependent somites and three pairs of specialised appendages,
rather than to a form with a great number of unspecialised somites and similar
appendages.

Turning now to a somewhat different aspect of the subject, we notice that, if we
confine ourselves to structure, and leave out of sight the question of origin, there is
the closest similarity between serial homology and the homology between the corre-

A STUDY IN MORPHOLOGY. 127

spending organs of allied animals which OWEN has proposed to distinguish by the
term " special homology."

The structural relation between one appendage of Lucifer, say the first pereiopod,
and another, such as the swimmeret, is identical with the relation between the pereio-
pod of Lucifer and that of Squilla, or a Crab or Lobster. In both cases we have a
fundamental similarity of plan, which is independent of external conditions; and joined
to this essential similarity, we have a more superficial diversity of structure which is
plainly due to difference in the functions of the appendages, and their relations to the
external world. The resemblance between the two kinds of homology does not stop
here. Tracing the ontogeny of the appendages we find that there is much less differ-
ence between the larval pereiopods of Lucifer and those of the Lobster than there is
between the appendages of the adults, and we find exactly the same thing when we
compare the pereiopod and swimmeret of the same individual at earlier and earlier
stages of development.

There is precisely the same resemblance between symmetry and special homology.
The right and left claws of the Common Crab (Callinectes) are not exactly alike, since
the cutting edge of one claw is sharp and set with pointed teeth, while the edge of
the other is thick, with thick blunt crushing tubercles. The two appendages are alike
in plan or homologous, but each is fitted for a specialised function by a slight structural
peculiarity. In this case, as in the others, the differences are less marked, and the
common plan more closely followed, in the larva than in the adult.

Serial homology and bilateral symmetry are thus seen to be like special homology
in all purely structural features. In each case the homology is a resemblance which
is independent of external conditions, but which may be obscured by secondary modifi-
cations whenever external conditions render it necessary.

In each case, too, the secondary modifications become less marked, and the underlying-
plan more evident as we pass back from the adult to earlier and earlier stages of deve-
lopment. We must therefore include ah1 three kinds of homology in a single class or
category, and the employment of one term to denote the phenomena of special homo-
logy, of another for serial hornologies, and a third for bilateral homologies, and others for
other sorts of general homology must not be allowed to obscure the fact that they are
all different forms of the same thing, essential similarity joined to superficial diversity.
The terminology which has been employed by BRONN, H^ECKEL, LANKESTER, and
others for the different kinds of homology is valuable, and the only reason why I
have not made use of it is that the more familiar terms, "serial homology" and "bi-lateral
symmetry" answer every purpose equally well in treating of the Arthropods. H.-ECKEL'S
subdivisions are natural, but they are simply subdivisions of a great class of similar
phenomena, which must still be included under the general term " homology."

Special homology may be defined in two ways, morphologically and phylogenetically.
From the morphological point of view an homology is a similarity in essential plan of
structure, which may be obscured by differences due to diversity of function. From

128 MB, W. K. BROOKS ON LUCIFER:

the phylogenetic point of view it is a resemblance which is due to community of origin
or heredity from a common ancestor, while the differences between homologous
organs are due to the divergence of allied forms, and to the selection and perpetuation,
through natural selection, of variations which are in accordance with changed conditions
of life.

Now are the phenomena of serial and lateral hornology like those of special homology
in this second or phylogenetic sense, as well as in a morphological sense ?

On the assumption that the remote ancestor of the Crustacea was a community of
independent organisms, all of which had inherited their organisation from the same
parent, we might answer that serial homology is like special homology when viewed
from a phylogenetic stand-point, and if we assume that this series was at first double,
and that the progress of centralisation suppressed one side of each metamere as the
community became gradually fused into a bilateral organism, we may make the same
statement regarding symmetry.

A process of evolution of this sort is not impossible, and in some cases there seems
to be evidence that it has actually occurred. Pyrosoma is clearly a community of
independent Ascidians, which has been brought by natural selection into a form which
has a certain degree of individuality of its own, independent of that of the component
units ; although in this case the peculiar form of the community has called for little
differentiation, and the polymorphism is therefore very slight.

The salpa-chain is a bilateral community, and in Dolioluin we have a similar
community which exhibits considerable polymorphism. If this process were carried
a little further we might ultimately have a bilaterally symmetrical organism in which
corresponding parts in the series or on opposite sides should be strictly homologous by
descent ; but we are not therefore justified in assuming that all instances of serial and
lateral homology have originated in this way, and even if we were a more careful
analysis will show that the assumption does not remove all the difficulties.

If we grant, for the sake of argument, that the Crustacea are not the descendants
of a Nawplius, but of a remote ancestor which consisted of a community of independent
metameres, we shall still be forced to recognise a bond of relationship between the limbs
of a Decapod, which is very much more recent than that which they owe to common
descent from the parent of the group of Zooids which formed the ancestral community.

A reference to the figures will show that the first, second, and third thoracic limbs
of the adult Lucifer agree with each other, or are homologous, in certain features which
are not present in a Schizopod. The exopodite is absent and the endopodite is long
and slender in all of them, and it carries short hairs along its entire length, while, in
the Schizopod -larva, the exopodite is present, and the long hairs are restricted to the
tip of the stout endopodite. We must therefore recognise a bond of union or homo-
logy between these three appendages which has determined that they shall be like
each other in the adult Lucifer, and the assumption that this similarity is due to
heredity from the parent of the imaginary metameres which joined together to form

A STUDY IN MORPHOLOGY. 129

the primitive Crustacean is out of the question, for we know that no further back
than the Schizopods these appendages had quite a different structure.

The study of serial or lateral homology in other groups of animals forces us to the
same conclusion, and compels us to recognise a persistent bond of union between
them which cannot be due to what we usually understand by heredity.

On the assumption that the Vertebrates are the descendants of a community of
metameres, the genetic relationship between a man's arm and a bird's wing must be
almost infinitely closer than that between a man's arm and his leg, and this again
much more recent than that between his right and his left arm. The arm and wing
inherit their homology from the anterior limb of the common ancestor of man and the
birds, but man's arm and leg have no common ancestor more recent than the limb of
the parent of the imaginary metameres which gave origin, by their union, to the ancestor
of the Vertebrates, and the common ancestor of the right and left arms must have been
still more remote.

When we compare man's arm and leg we find that they have homologous features
which are not only more recent than the time when man's ancestors diverged from the
ancestors of the birds, but more recent than the separation of the anthropoid and
simian stems. They resemble each other in the texture of the skin and in the shape
of the nails, and these resemblances are strictly homological, that is, they are not
due to external conditions, but in spite of them ; and we meet with countless similar
resemblances all through the animal kingdom. They are not accounted for by the
" metamere " theory, even if this is fully accepted, for in many cases they are not old,
but are of recent acquisition.

In the case of the Crustacea the assumption that the remote ancestor of the group
had a many-jointed body does not account for them ; and as the supposed necessity
for an explanation of serial homology is the only reason for believing that this remote
ancestor had a great number of body-segments, it is clearly illogical to reject the
embryological evidence that this ancestor was a three -jointed Nauplius, in order to
hold an hypothesis which fails to account for the facts which are supposed to render it
necessary.

EXPLANATION OF THE PLATES.

All the figures where the magnifying power is not stated were drawn with a power
of 160 diameters (ZEiss, Oc. 1, Obj. D) ; but the actual amplification of the drawings
is not uniform. In copying the original sketches it has been convenient to reduce
the size of some of them, and no inference as to relative size should be drawn from
any of them except where measurements are given.

In order to render the figures as truthful and lifelike as possible, the animals were
subjected to very little confinement while under examination, and as their incessant

MDCCOLXXXII. S

130 MR. W. it. BROOKS ON LUCIFER:

and violent movements rendered the use of a camera impossible, they are not drawn
to a fixed scale.

In all the Plates the capital reference letters are used to denote the same parts, as
follows :—

A. First antenna.

A. I to A. G. The series of abdominal somites.

An. Second antenna.

C. Carapace.

E. Compound eye.

L. Labrum.

M. Mandible.

Mp. 1, Mp. 2, Mp. 3. The first, second, and third maxillipeds.

MX. 1, MX. 2. The first and second maxillse.

Oc. Ocellus.

PL l-Pl. 6. The six pairs of abdominal appendages.

Pr. 1-Pr. 4. The four pairs of thoracic limbs.

R. Rostrum,

T. l-T. 4. The four thoracic somites.

T. Telson. (27. in Plate 3, fig. 26.)

PLATE 1.

The letter a in figs. 1 to 8, which were all drawn from the same egg, marks the

same point in all.

Fig. 1. An egg during the period of rest which follows the first period of segmenting
. activity.

Fig. 2. The same egg at the beginning of the second period of segmenting activity,
five minutes after the stage shown in fig. 1.

Fig. 3. The same egg five minutes later. One of the primary segments is beginning to
divide into two.

Fig. 4. The same egg ten minutes later. One of the primary spherules is perfectly
divided into two, and the division of the other is less advanced.

Fig. 5. The same egg five minutes later, and completely divided into four equal seg-
ments. This stage ends the second period of segmenting activity.

Fig. 6. The same egg, fifteen minutes later, during the second period of rest.

Fig. 7. The same egg, ten minutes later, entering upon the third period of activity.

Fig. 8. The same egg, twenty-five minutes later, divided into eight equal spherules.
This stage ends the third period of activity.

Fig. 9. Another egg during the third period of rest.

Fig. 10. Another egg near the end of the fourth period of activity, and divided into
sixteen equal spherules, arranged around a central segmentation cavity.

A STUDY IN MORPHOLOGY. 131

Fig. 11. Optical section along the principal axis of a somewhat older egg, showing the

yolk spherule, c, and the segmentation cavity, b.
Fig. 12. Optical section of the same egg at right angles to the one shown in fig. 11.

b. The segmentation cavity.

c. The yolk spherule.

PLATE 2.

Fig. 13. An optical section of an egg somewhat older than the one shown in fig. 1 1.

Fig. 14. Optical section at right angles to that of fig. 13.

Fig. 15. Optical section of an egg a little older than the one shown in fig. 13.

d. Orifice of invagination.

Fig. 16. Optical section of a still older egg in the same position as figure 15.
Fig. 17. Optical section of a still older egg in the same position.
Fig. 18. Surface view of the formative pole of the egg shown in fig. 17.
Fig. 19. Optical section along the principal axis of a still older egg.
Fig. 20. A similar section of a still older egg.

Fig. 21. Ventral view of an embryo in the egg-shell 24 hours after oviposition.
c. Anterior end of body.

f. Large spherules in the region of the digestive tract.

g. Metastoma.

Fig. 22. Dorsal view of the same embryo.

(e and /as in fig. 21.)

h. Cerebral ganglia.

i. Pigment spots.

m. Muscles.

Fig. 23. Similar aspect of the same embryo artificially removed from the egg shell.
(Letters as in fig. 22.)

PLATE 3.

Fig. 24. Ventral view of the same embryo, seen from a point of view a little anterior
to that of fig. 2 1 .

ex. Exopodite.

en. Endopodite.

(The other letters as in fig. 21.)

Fig. 25. Side view of the Nauplius, r^foth inch long, as it leaves the egg 36 hours
after oviposition.

g. Metastoma.

p. Pigment spots.

ex. Exopodite.

en. Endopodite.

132 MR, W. K. BROOKS ON LUCIFER:

Fig. 26. Side view of a Naupliu-s, ymroth inch long, and a little older than the one
shown in fig. 25. The animal is a little flattened by pressure. In a
ventral view of the same larva the anus was visible at the point mai-ked fi.
in the figure.

a>-. (Esophagus.

*. Stomach.

/. Large cells around stomach.

i. Intestine.

</a. Cerebral ganglia.

n. Ventral nerve-chain.

p. Pigment spots.

Fig. 27. Dorsal view of a ProtozoM, To§rrtu mcn l°ng. which moulted from the
Nauplius shown in fig. 26, about 96 hours after oviposition.

ff a. Cerebral ganglia.

s. Stomach.

h. Heart.

i. Intestine.

p. Pigment spots.

a. Anus.

/». Muscles of (Esophagus.

h. Postero-lateral spines of carapace.

ds. Posterior dorsal spine of carapace.

abd. Unsegmented portion of abdomen.

ex. Exopodite.

en. Endopodite.

Fig. 28. Right mandible of the same specimen seen from below.
Fig. 29. Right mandible seen from behind.
Fig. 30. Back or outer surface of first maxilla of the same specimen.

sc. Scaphognathite.
Fig. 31. Posterior surface of left-second maxilla of the same specimen.

«'. Scaphognathite.

b. Basal joint.
en. Endopodite.

Fig. 32. Left-first maxilliped of the same specimen ; posterior surface.

e.r. Exopodite.

en. Endopodite.

b. Basal joint.
Fig. 33. Right-second maxilliped of the same specimen.

e.r. Exopodite.

en. Endopodite.

b. Basal joint.

A STUDY IN MORPHOLOGY. 133

Fig. 34. A Protozoea, yg-^oth inch long, and a little older than the one shown in fig. 27.
os. (Esophagus.
pp. Annl pigment spots.
(Other letters as in fig. 27.)

PLATE 4.

Fig. 35. Side view of another larva at the stage shown in fig. 34.
ga. Cerebral ganglia.

s. Stomach.

sg. Shell gland.

h. Heart,
Fig. 3 5 A. Surface view of the area of attachment of the oesophageal muscles of the

same larva.
Fig. 36. Second antenna of the same specimen.

I. Basal joint.

ex. Exopodite.

en. Endopodite.

Fig. 37. Mandible, at same stage, seen from below.
Fig. 38. First maxilla of left side at same stage.

1 and 2. The two cutting joints of the basal portion.

ex. Exopodite.

en. Endopodite.
Fig. 39. Left-second maxilla at same stage.

b. Basal portion.

en. Endopodite.

sc. Scaphognathite.
Fig. 40. Left-first rnaxilliped at same stage.

ex. Exopodite.

en. Endopodite.
Fig. 41. Left-second maxilliped at same stage.

ex. Exopodite.

en. Endopodite.

Fig. 42. Ventral view of the same larva after the next moult and TooTJ^1 'inc^ l°ng-
Fig. 42A. Outline of the anterior end of fig. 42, more enlarged.

b. Basal joints of appendages.

ex. Exopodites.

en. Endopodites.

ftc. Scaphognathite.

134 MR. W. K. BROOKS ON LUCIFER:

PLATE 5.

Fig. 43. Ventral view of the last Protozoea (DANA'S Erichthina demissa), Toucjth mcn
long.

ex. Exopodite of second antenna.

en. Endopodite of second antenna,.
Fig. 44. Dorsal view of the same larva (ZEiss, A. 2).

ds. Median dorsal spine of carapace.

Is. Postero-lateral spines of carapace.
Fig. 45. Side view of the same larva at the same stage.

<is\ Median dorsal spine of carapace.

tg. Thoracic ganglia.

?'. Intestine.
Fig. 46. Posterior surface of left first maxilla of the same larva.

1 and 2. Cutting joints of basal portion.

en. Endopodite.

so. Scaphognathite.
Fig. 47. Left second maxilla of the same larva.

1>. Basal joint.

sc. Scaphognathite.

en. Endopodite.
Fig 48. Left first maxilliped of same larva.

b. Basal joint.

ex Exopodite.

en. Endopodite.
Fig. 49. Left second maxilliped of same larva.

(Letters as in fig. 48.)

PLATE 6.

Fig. 50. First Schizopod stage (this stage is the equivalent of DANA'S genus Sceletina;
of CLAUS'S Acanthosoma stage ; and of WILLEMOES-SUHM'S Amphion stage).
The larva passes, by a single moult, which was frequently observed in iso-
lated specimens, from the stage shown in fig. 43 to the one which is shown
in this figure.

Fig1. 51. Mandible at same stage.

D O

Fig. 52. First maxilla at same stage.

Fig. 53. Second maxilla at same stage.

Fig. 54. Sceletina larva, two moults later (ZEISS, A. 2).

s. Antero-lateral spine.

p. Pigment spot on fourth abdominal somite.

A sTtrot IN MORPHOLOGY. 135

Fig. 5 4 A. Outline of anterior end of carapace.

Fig. 55. Dorsal view of the anterior end of the same larva.

Fig. 56. First antenna of the same larva.

Fig. 57. Second antenna of the same larva.

Fig. 58. Mandible at same stage.

Fig. 59. Locomotor appendages of the left side, at same stage, seen from above.

PLATE 7.

Fig. 60. Side view of a young Lucifer, about i^th inch long, which was produced by
the moulting of a larva like that shown in fig. 54 (this stage agrees pretty
exactly with the Mastiyopus stage of Seryestes).

g. Antennal gland.

n. "Neck."

c. Carapace.
Fig. 61. Half-grown Lucifer, about ^ an inch long.

g. Antennal gland.

eg. Cerebral ganglia.

co. Commissures to ventral ganglia.

s. Cephalic pouch of stomach.

h. Heart.

c. Carapace.

Fig. 62. Inner surface of mandible of adult.
Fig. 63. Outer surface of mandible of adult.
Fig. 64. Outer surface of first maxilla of adult.
Fig. 65. Inner surface of same.

PLATE 8.

Fig. 66. Head of a small female about f rds of an inch long, seen from below.

Fig. 67. Second maxilla of same.

Fig. 68. First maxilliped of right side of adult.

Fig. 69. Inner surface of same.

Fig. 70. Second maxilliped of adult.

PLATE 9.

Fig. 71. Dorsal view of tip of telson of adult female.

Fig. 72. Side view of last abdominal somite and swimmerets of adult female.

Fig. 73. The same parts of an adult male.

136 MR. W. K. BROOKS ON LUCIFER:

Fig. 74. Side view of posterior half of carapace and first abdominal somite of a mature
female, to show the reproductive organs. In order to reduce the number
of figures this specimen is represented with its ovaries full of ripe eggs,
while a large bunch of developing eggs are attached to the basal joints of
the last pair of thoracic appendages, but these two features are never
exhibited completely at the same time in a single individual.

co. Carapace.

h. Heart.

i. Intestine.

trj. Thoracic ganglia.

or. Ovary.

od. Oviduct.

AT. Seminal receptacle.
Fig. 75. The corresponding part of the body of an adult male.

c. Carapace.

h. Heart.

i. Intestine.

tg. Thoracic ganglia.

a. First abdominal somite.

t. Testis.

vd. First division of vas deferens.

sp. Second division.

sv. Third division.
Fig. 76. First pleopod of young male.

Fig. 78. Side view of the larva shown in Plate 10, fig. 77 (Zfiiss, A. 2).
Fig. 79. Dorsal view of the same larva.
Fig. 80. Mandible of the same larva,
Fig. 81. First maxilla of the same larva.

en. Endopodite.

so. Scaphognathite.
Fig. 82. Second maxilla of the same larva.

so. Scaphognathite.

en. Endopodite.

PLATE 10.

Fig. 77. Ventral view of the last Protozoca stage of Acetes Toootn inch long.

ex. Exopodite of second antenna.

en. Endopodite of second antenna.

Fig. 85. The specimen shown in fig. 84, after another moult and TcuMjth mcn
Fig. 86. First antenna of the same larva.
Fig. 87. Second antenna of the same larva.

A STUDY IN MORPHOLOGY.

137

PLATE 11.

Fig. 84. Ventral view of the specimen shown in tig. 81, after the next moult and
T-JSoth inch long.
ex. Exopodite.
en. Endopodite.

Fig. 83. First maxilliped of the larva shown in Plate 10, fig. 77.
Fig. 88. Second and third maxillipeds and periopods of the larva shown in Plate 10,

fig. 85.

Fig. 89. First pleopod of same larva.
Fig. 90. An older specimen y^-th inch long, which was captured at the surface.

MDCCCLXXXTI,

•

•

•

Brooks.

Jlu'l. Trams. L88U. 7Y<//<- 1.
a.

Fiq 1.

x.

Fiq. 2.

Fiq. 3

Fig. 6,

a

,

Fie, :

Fi

.1

Fi a. 9.

S.

"

Fig. 10.

W.K.B. a J. .1

Fi a 12

-Brooks

Phil. Trans, 18 H'J. />/«/*•

c

. 13.

14,.

Fiq. 16.

Fin 17

Fxg. 18.

Fi a 18.

An.

23.

Fiq. 20.

. 22.

IV. K. II .. /

•

il. Trans. 188'J. /»/,//,• 3.

•

AM..

W..H.B cut mit Ari.

Phil. Trans. 1882. Plate

s? Mar 1

Fie,. 35 a.

W, K- B axt naU- <leL.

Phil. Trans. 1882. Male 5.

W,K B. a^™**. del

i' •

Brooks

J'/nl Trans. ISS'J. I 'I (tic &

i . •

Brooks .

Phil. Trims. 1S82. Plate 1.

W. K B ouA. nai d,l

Wtstli,.-.-

Jl rooks

Phil. Tniim. 1881'. 7V//A-8.

Pi a. £8

Ficr. 69.

Fie,. 67

Fiq. 70.

W K b ,...' .,,f Ji,l.

Li

Phil Trans. l882.Plate$.

Fiq.79

IV A P .,J ,.,,1 .1,1

, .<

Brooks

Pli.il. Tntns. 1 8 B2. />/«/<> 1O

86.

87.

Pi g 85.

W.K B. aul n«i. del.

,,;. .: I '

/%//. Trwvs.lSQZ. Plate LI

r , ,, 9i

W K. B. jut >...( .