CARDED )746

NATIONAL ACADEMY OF SCIENCES.

Veo Trt.

SIXTEENTH MEMOIR.

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA OF NORTH AMERICA,

141

W. P. HAY COLLECTION

(XVL)

ON THE CARBONIFEROUS ALTPHOSUROUS FAUNA OF NORTH AMERICA.

READ NOVEMBER 18, 1885.
By A. S. PACKARD.

By the kindness of Messrs. R. D. Lacoe, of Pittston, Pa., and J. C. Carr, of Morris, Ill., I
have been able to examine a most valuable collection of rare Xiphosuran fossils from Mazon
Creek, Grundy County, Illinois, besides two specimens from the coal-beds of Pennsylvania. These

_have revealed the existence on this continent of two genera, hitherto confined to the European coal-
measures, viz, Cyclus and Belinurus. From the Pennsylvanian coal-measures a new species of
Prestwichia has been obtained, and it is probable that ultimately we shall find as many species
of this family as there are in Kuropean strata.

Of still more interest is the discovery of remnants of cephalic limbs in Cyelus and Prestwichia,
showing that in these animals the cephalic appendages were like those of the larval Limulus. It
also appears that the ontogenetic development of Limulus is an epitome of that of the Xiphosura
as a group. Furthermore, our studies have led us to restrict the Xiphosura to the three families
of Cyclide, Belinuride, and Limulide, while certain upper Silurian forms referred by Woodward
to the Eurypterida, and by Zittel placed among the Xiphosura, are, temporarily at least, referred to
a new suborder, the Synziphosura, a group combining with features of its own, characteristics of
the Xiphosura and some strong resemblances to the Trilobites.

e

Family CYCLID Packard.
CYCLUS AMERICANA Packard. Pl. V, figs. 1, la; VI, figs. 4, 4a.
Cyclus americana Pack., Amer. Naturalist, xix, 293, March, 1885.

In a nodule from Mazon Creek, Illinois, received from Mr. Lacoe, I recognize a species of this
rather obscure genus, which has not before occurred in North America, though in Europe nine
species have been described.

In form the animal is perfectly orbicular, the length being exactly equaled by the breadth,
The body is regularly disk-shaped, flattened hemispherical, with the edge of the body broadly and
regularly expanded, the margin being thin and flat, and apparently a little wider on the sides
than on the anterior or posterior end. The inner edge of the rim is separated by an impressed
line from the raised portions of the body-disk ; the surface of the rim is not plain and smooth, but
ornamented by a series of plate-like, squarish markings, apparently separated by a slight impressed
line, and with a slightly marked, raised tubercle on each plate or scale.

There are no indications of segments either of the head or abdomen, nor are the limits between
a head and abdominal region distinguishable, as is the case in Cyclus jonesianus Woodw.* There

* Contributions to British fossil Crustacea. By Henry Woodward, F.G.S., etc. Geol. Mag., vii, No. 12, pl. xxiii,

Dec., 1870.
143

144 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

are, however, indications of four, and perhaps five, pairs of short, thick, cephalic appendages on the
anterior third of the body. Unfortunately, they are not well preserved, the basal and distal por-
tions not present, and the indications of joints indistinct; they are directed outwards from near the
median line of the body, on each side of the intestine, the hindermost (6th) pair being directed
somewhat obliquely outwards and backwards. In their position and relative distance apart they
seem homologous with the cephalic limbs of the larval Limulus. The indications, slight as they
are, lead us to suppose that they approached in general shape and relative size those of Prest-
wichia, reaching near but not passing beyond the edge of the cephalic shield. The distal portion of
the limbs not being preserved, it is impossible to conjecture whether they were forcipicate or not.*

Through the middle of the body, from near the anterior to the posterior margin, passes the
cast of the digestive canal; it is swollen in front, the dilatation probably representing the pro-
ventriculus, and in outline the cast recalls that of the digestive canal of Limulus. Judging by
analogy, the mouth was probably, as in the larval Limulus, situated well in front between the
anterior pairs of appendages, and the csophagus curved forward and upward from the mouth,
while the vent was situated very near the hinder edge of the body.

There are no distinct traces of an abdominal region in the specimen, and it will be seen that
in some of Dr. Woodward’s figures there is also none. It is not probable that there was any spine
in the genus, none being indicated in any of the figures or descriptions published.

Length of body, 14°; breadth, 14"; width of the flattened rim or margin, 1". Locality,
Mazon Creek. No. 218a, b. Collection of Mr. Lacoe.

Judging by our specimens and Dr. Woodward’s figures, Cyelus if restored would have an orbic-
ular body, more or less disk-like or hemispherical, with a cephalic region composed of six seg-
ments, which are not, however, indicated externally; this region had a thin margin, as in Prest-
wichia and Limulus, <A pair of median ocelli were probably present, but no compound lateral eyes
have yet been discovered. An abdominal region was slightly differentiated, and it was composed
of three segments, the third representing that of the embryo Limulus, which in that form eventually
becomes the caudal spine. The Cyclus was provided with six pairs of cephalic appendages, which
were short, not reaching beyond the edge of the body. With these the animal could creep over the
bottom of the shallow, muddy portions of the carboniferous sea. It is not improbable that there
were two pairs of abdominal lamellated legs, adapted for respiration, short and broad, and not
unlike those of the empryo Limulus. In fact, our conception of the form of the living Cyclus is
that it was not much unlike the advanced embryo of Limulus, either in the stage represented in

Figs. 17 and 17a or 18, 18a, and perhaps 19 and 19a, of our memoir* of one of which

(19a) Fig. 8 is a copy. At this stage of development the body of Limulus is

“72 \ hemispherical ; Seen from beneath the outline of the body is nearly orbicular, the

SS CY) abdominal region completing the cirele. If Limulus were arrested at the stage of

SS development when only three abdominal segments had appeared, and the devel-

SS? opment of the feet and claws had beén accelerated and then hatched, it would be,
so to speak, a Cyclus.

We ice ie In our first memoir on the development of Limulus we adopted Dr. Wood-
Brae eo: ward’s view that Cyclus was a Xiphosuran. In 1868 Dr. Woodward stated: “ We

must differ from M. de Konink in referring tiis form to the Prilobites. If truly an adult, it must be
placed near to Apus, with the other shield-bearing Phyllopods; if a larval form, it may have been
the early stage of Prestwichia, or some other Limuloid of the coal-measures. Nor do we think it in
the least probable that the shield of Cyclus radialis was flexible or contractile, its original segments
being completely soldered together into one pire ”; and in 1870 he adds that, from the recent inves-

* The ee Siopment of Teimaiae apn GAS. 1872, Pl iv. momen Bost. Soe. _ Nat. Hist., Vol. We

Since this article was sent to the printer, I have reeeiyed, through the kindness of ime author, Mr. B. N.
Peach’s ‘‘ Further Researches among the Crustacea and Arachnida of the Carboniferous Rocks of the Scottish Border.
Trans. Roy. Soe. Edinburgh, 1882.” In this memoir Mr. Peach figures and briefly describes the limbs of Cyclus.
‘Prom the fact,” he says, “ that several of the Survey specimens exhibit limbs, the radiating lines of the sternum are
most probably the divisions between the cox:e.” Under Cyclus testudo Peach, he describes six triangular plates on
each side, divided from each other by deep sulci, and converging upon an oral sternum. He also refers to ‘‘ the jointed
cylindrical limbs, the tips of which have not been observed.”

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 145

tigations of Dr. Lockwood and myself, “these forms may indeed be the larval stages of Prestwi-
chia, Belinurus, ete., the antetypes in Carboniferous times of the modern king crab.” “ Were it
not for the large size of these fossils, some (C. Harknessi) measuring five lines in length, three
and one-half lines in breadth, and three lines in height, we should be disposed to agree with
Mr. Woodward; but, from what is known of the size and form of the freshly-hatched larva of Lim-
ulus and the Trilobites, I should infer that they were either the larve of some unknown genus of
Limulide, or adult but embryonic forms. The larve of Belinurus and its allies, Prestwichia and
Euprooéps, were, in all probability, closely allied in their form and size at the time of hatching to
the larva of Limulus. But on comparing the deep hemispherical form of Cyeclus, with the surface
of the body deeply lobed over a more or less extent, with the embryo of Limulus before it is
hatched (PI. iv, figs. 18, 18a), we find a striking similarity; indeed, we seem to be dealing with a
distinct embryonic type of Limulide. In Cyclus we have, in a tate larval or possibly adult condi-
tion, that state of Limulus in which the body is deeply hemispherical, and the abdomen has just
been differentiated from the rest of the body, while the deep transverse lobes of the yolk are not
yet absorbed, as seen in Pl. iv, figs. 18, 18a, in the embryo of Limulus; the cardiac or median lobe
being as distinctly marked in Cyclus as in the embryo of Limulus.” (Development of Limulus,
1872, p. 189.)

After again reviewing the characters of Cyclus, with the specimen of C. americanus before us,
we feel confirmed in the views above presented, and would regard Cyclus as the representative of
a family of Xiphosura, being an adult form, and embryotypic, to coin a word, of a Limulus, while
the Belinuridsz represent the larval Limulus.

Family DIPELTID4 Packard.
DIPELTIS DIPLODISCUS Packard. PL V, figs. 2, 2°.
Dipeltis diplodiscus Pack., Amer. Naturalist, xix, 293, March, 1885.

This name was proposed for a singular form which is not satisfactorily preserved, so that its
exact relations are not readily determinable. The body is suborbicular, flattened, disk-like, slop-
ing regularly and gradually from the median area to the edge; it is divided into two portions; the
larger one to be regarded as auterior or the cephalic shield, and the other as posterior, constituting
the abdomen (urosome). The edge of the body is very slightly marginate, not broadly so as in
Cyclus; nor is the body distinctly trilobate, as in the Belinuridee and Limulide, though unfortu-
nately the median area of the cephalic shield is wanting. The integument is rather thin, showing
no traces of segments; its surface may have had a few scattered small tubercles, at least there are
slight indications of them. The surface is smooth and shining.

The cephalic shield is nearly twice as broad as long; the posterior lateral angle is well-rounded,
with no sign of a lateral spine; in front the edge was probably obtusely rounded; the surface is
slightly convex, the disk being low and flat, with no traces of a glabella; the hind edge of the
shield is moderately concave, the limits between it and the urosome being clearly indicated by a
slight, but distinct, regular, curvilinear suture.

The urosome is about three-fourths as long as, but equal in width to the cephalic shield. The
front edge is somewhat arcuate, so that the projecting anterior-lateral angle is directed a little
forward, and is quite free from the lateral angle of the cephalic shield, which turns away anteriorly
from it, leaving a triangular space between the sides of the two regions. Posterior edge of the
urosome regularly rounded, and with a slight margin. No traces of a caudal lobe or spine. Total
length, 20""; total breadth, 20"; length of cephalic shield, 11°"; breadth, 26""; length of urosome,
9°"; breadth, 19.5"". Collection of R. D. Lacoe, 2017 *”* in a nodule from Mazon creek, Morris,
Illinois.

This remarkable animal was disk-like in shape, composed of two regions, the head and abdo-
men or urosome, which are more distinctly separated than in the Cyclide ; while there are no posi-
tive characters to separate it from this group, we would, for the present at least, refer it to an
allied family, as it is orbicular, tailless, and consists of a broad, large cephalic shield, with a shorter,
distinct, non-segmented urosome.

S. Mis. 154 19

146 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

Family BELINURIDA® Packard.

PRESTWICHIA DAN )Meek) PL. V, figs. 3, 38; VI, 1, 1%, 2, 2%.

Bellinurus dane, Meek and Worthen, Proc Acad. Nat. Sc., Phil., March 1865, Rt. Geol. Sury., Ill. ii, 395,
1866.

Prestwichia dane Meek, Amer. Journ. Sc., 2d ser., xliii, 257, 1867.
Euproéps dane, Meek, Amer. Journ. Sc., xliii, 394, 1867.
Meek and Worthen, Rt. Geol. Surv. Ill., iii, 547, 1868.
Packard, Amer. Naturalist, March, 1885,

Head and abdomen (urosome), in the largest specimens, of the same length; in younger speci-
mens the head is rather shorter than the abdomen; head about one-third as long as broad; genal
spine about two-thirds as long as the head, and turning at nearly a right angle with the straight
hinder edge of the cephalic shield; the spine as a whole is directed somewhat outward, nearly
reaching a point about opposite the hinder edge of the third abdominal segment. Median lobe
of the head or glabella, rather deeply excavated in front; at the bottom of the excavation are
situated traces of the simple eyes, which have the same situation and shape as in Limulus. The
small compound eyes are situated on the outer anterior angle made by the sides of the glabella
and are of nearly the same relative size and in the same general situation as in the larval Limulus,
though placed a little nearer the front margin. The eyes themselves are small, oval and promi-
nent. The sides of the glabella are produced behind into a sharp spine, projecting backwards over
the base of the abdomen.

The abdomen (or urosome) is from one-fourth to one-third broader than long, and is composed
of eight distinct segments, including the caudal spine; the body of the abdomen is full, convex,
and distinctly trilobate, the median or cardiac lobe being in general about a third narrower than
the lateral lobes or pleura, and contracting in width towards the fifth segment. The sutures be-
tween the segments on the lateral lobes are very distinct, being raised, narrow ridges, prolonged
into and forming the hinder edge of the long, sharp, slightly curved, lateral spines; of these lateral
spines those on the first and second segments are the narrowest and most acute, that on the seventh
the widest and most obtuse. In the cardiac lobe the third abdominal segment bears a high rounded
tubercle, and there is one about twice as large on the sixth segment; those on the other segments
are small, and in most of the specimens there are traces only of those on the third and sixth seg-
ments. The caudal spine (representing the eighth abdominal segment) is somewhat enlarged at
the base; it is three-cornered in section, much as in Limulus, the surface is smooth, and it is about
three-fourths as long as the abdomen.

Length of entire body (largest specimens), 60™™; breadth, 53™™,

Length of cephalic shield, 24™™"; breadth, 53™™,

Length of lateral cephalic spine, 15™™; breadth, near base, 3.5™™,

Length of abdomen (urosome) (not including the caudal spine), 23™™; breadth 35™™,

Length of longest Jateral abdominal spine, 6™™,
Length of caudal spine (telson), 15™™,

The smallest specimen is 10" in length, and 12™ in width, the caudal spine being less than
one-half as long as the abdomen.

4

Description of the cephalic appendages.

In a nodule from Mazon Creek received from Mr. J. C. Carr, containing the remains of a speci-
men 55™™ across the shield (Pl. VI, figs. 2, 2), the cephalic appendages are more or less distinctly
preserved. Of the first pair there are faint traces, the two small limbs lying parallel to each
other and in the same position as in the larval Limulus, and of nearly the same proportions. The
impressions of the succeeding limbs are distinct ; the second third, fourth, and fifth pairs are of
the same size, the fifth pair being perhaps a little longer, as the tips extend near the edge of the
cephalic shield. All four pairs, 7. e., second to fifth, are chelate, the forceps being well developed
and plainly visible in the third and fourth pairs, as these limbs are turned on their side; the fifth
pair are undoubtedly chelate, but lie so that the outline is a simple point. The sixth pair differs

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 147

from the others in ending abruptly, the penultimate joint being long and of the same width
throughout, and truncate at the distal end, where it gives rise to three small, sharp spines; there
are also traces of a terminal minute joint from which two spines arise,

Length of second, third, fourth, and fifth pairs of legs, 16™™,

Length of sixth pair, 17™™,

Length of penultimate joint, 6™™,

Thickness, 1™™,

The legs are nearly identical in shape and length with those of the larva of Limulus described
and figured in my Development of Limulus (PI. 1, figs. 24°, 25%, and 234), though perhaps a little
shorter, as they do not reach beyond the edge of the cephalic shield. It thus appears that in
respect to its limbs as well as the shape and proportions of the body the Prestwichia resembles
the larval Limulus. Thus Limulus in its development passes through a trilobitic, and afterwards
a Belinurid stage.

PRESTWICHIA LONGISPINA Packard. PI. V, fig. 4.
Euprodps longispina Pack., Amer. Naturalist, xix, 292. March, 1885.

The specimen upon which this species is founded is Mr. Lacoe’s Nos. 215*» (impression and
reverse), and was probably a molted skin (PI.-V, fig. 4). The body is considerably distorted by
pressure, but the specific distinctness from P. dane is marked. The species will he readily dis-
tinguished by the very long genal spines; they extend nearly or quite to a point opposite the base
of the caudal spine. The abdomen appears to be narrower in proportion to the cephalic shield
than in P. dane while the genal spines are longer and narrower. The caudal spine is not well
preserved.

Length of body (not including the caudal spine), 20™™,

Length of head, 10™™,

Length of abdomen, 10™™,

Breadth of cephalic shield, 24™™,

Breadth of abdomen, 13™™.

Length of lateral cephalic spine, 13™™.

Pittston, Pa., Butler mine, Nos. 215%», collection of Mr. Lacoe.

In another larger specimen (Lacoe’s No. 214, Pl. VI, fig. 3), the glabella, with the eyes, ocelli,
and a part of the left lateral spine are preserved. Whether this is the same species as P. longispina
I cannot tell with certainty, as the genal spines are not sufficiently well preserved, but provision-
ally it may be regarded as belonging to the species under consideration. The median lobe of the
head is larger in proportion to the entire cephalic shield than in P. danw, and the eyes are nearer
the lateral margin. The ocelli are situated on the median ridge of the lobe, somewhat behind the
indentation between the lobes. The individual is without doubt a Prestwichia having the same
number of abdominal segments as in P. dane.

Length of body (without the caudal spine), 30™™,

Breadth of cephalic shield (estimated), 37™™.

Length of cephalic shield, 17-18"™™.

Length of abdomen, 13™™,

Breadth, 23™™,

Estimated length of lateral cephalic spine, 15™™,

Distance between the compound eyes, 17™™.

Distance from ocelli to the front edge of body, 6™™.

Distance from ocelli to hinder edge, 21™™.

Oakwood Colliery, Wilkes Barre, Penn., collection of Mr. Lacoe, No, 214°.

Regarding the position of the Illinois and Pennsylvania beds containing these fossils, Mr. Lacoe
writes me: “The horizon of the Pennsylvania specimens of Euprodps is much higher than that of
Mazon Creek. The latter is at the very base of the productive coal-measures in shale over the
bottom seam of coal. The specimen from the Butler mine, Pittston, is from shale over coal ‘ H’
(Mammoth vein), at the top of the lower productive coal-measures, about 300 feet above, and that
from the Oakwood colliery is either from the same horizon or the bottom of the lower barren

148 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

measure next overlying it. The shaft from which it was taken, penetrating both, the exact posi-
tion of the rock containing it could not be ascertained when we. discovered it in the ‘dump’ or
rock pile.” Another specimen from Scotch Hill railroad cut, Pittston, Pa. Coal E. Lacoe’s No. C.
83-34.

Note on the validity of the Genus Euproops.

By referring to the synonymy of Prestwichia dane, it will be seen that in 1865 it was referred
by Messrs. Meek and Worthen to Belinurus for reasons given in Paleontology, vol. iii, of the
Geological Survey of [llinois, p. 547. After the appearance of Dr. H. Woodward’s paper read
before the Geological Society of London in 1866* in which the genus Prestwichia was separated
from Belinurus, the American form was referred to the new genus, Prestwichia, by Mr. Meek.

“At a later date (February, 1867), Mr Woodward published excellent figures in the Quart.
Jour. Geol. Soe., London, vol. xxiii, pl. 1, of the typical forms of both Preswtichia and Belinurus.
From these it became evident that the peculiarities of the ridges of the head of the form on which
he founded the genus Prestwichia, and which we had supposed probably due to some accident,
really exist. Consequently, our type was regarded as being generically distinct, and the name
Huprodps was proposed by one of us for it. Mr. Woodward, however, has since expressed the
opinion that these differences are probably of scarcely more than specific value. (See Geol. Mag.,
Jan, 1868, vol. v., p.2.) Without professing to have made an especial study of the fossil Orustacea,
on which Mr.~Woodward is well known to be an eminently reliable authority, we would state that
we can scarcely doubt that a comparison of specimens would lead him to the conclusion that the
American form is at least subgenerically, if not generically, distinct from Prestwichia.”

Finally the authors state that Euprodps differs from Prestwichia “‘not only in the position of
the eyes, and the form and size of the glabella, or central area of the cephalothorax, but in the
entire arrangement of the ridges and included areas of the same.” Fig. 9. is from an electrotype
of a cut published by Messrs. Meek & Worthen in illustration of their genus Euproops.

Fic. 9.—Euprodps dane. M.& W. Vie, 10.—Prestwichia rotundatus. Atter
After Meek. Woodward.

After repeated examinations of the series of about a dozen specimens from the collections of
Messrs. Lacoe and Carr, I am at a loss to find valid characters for the genus Euproéps. In one ex-
ample of P. dane, the glabella or middle lobe of the head, is distinctly divided into four sublobes,
as in Woodward’s figure of P. rotundatus; again the lack of lateral abdominal spines in his figure of
P. rotundatus appears to me to be due to the imperfect state of preservation of the specimen, as
some of the Illinois specimens do not show them; again the spines projecting from the sides of the
glabella over the base of the abdomen, and represented as wanting in Woodward’s figures, are
wanting in certain Illinois specimens. As to the position of the compound eyes in P. rotundatus
as represented in Woodward’s figure, I am inclined to believe that the author and artist have been
in error. I should not venture to make such a statement if in our Illinois and Pennsylvania
specimens of Prestwichia and Belinurus the position of the eye were not invariably on the outer

“On some points in the structure of the Xiphosura, ete., Quart. Journ. Geol. Soc, Feb. 1867.

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 149

angle of the glabella, in a position homologous with their situation in Limulus. I venture then to
give the opinion that the apparent differences between Prestwichia and Euprodéps, as stated by
Messrs. Meek and Worthen, did not exist in nature, and that the genus Prestwichia was common
to both Europe and North America during the Carboniferous Period. It is interesting in this
connection to observe that the descendants of the Belinuride in Europe, survive in the Solenhofen
Limuli until the Jurassic, and disappear during the Cretaceous period, not to arise again on the
western coasts of the old world, while in North America, so far as the record shows, the type became
extinct during the Mesozoic and Tertiary, to reappear in the Quaternary and present period.

As regards the differences between Belinurus and Prestwichia, the former genusis the higher
form, approximating Limulus in the consolidation of the eighth and ninth abdominal segments
(forming the “abdomen” so regarded by Dr. Woodward) and in the very long caudal spine. In
Prestwichia there is one abdominal segment less than in Belinurus, the short caudal spine forming
the eighth.

BELINURUS LACOEY Packard. PI. V, fig. 5.
Belinurus lacoei Pack., Amer. Naturalist, xix, 292, March, 1885.

Cephalic shield of the usual shape and length in proportion to the abdomen; the front margin
as usual; the genal spine long, acute, extending obliquely outwards to a point parailel with one
either a little behind the middle of the abdomen, or, in the older, larger specimens, nearly to a
point parallel with the base of the caudal spine. The median lobe is, as usual, divided by the
median line into two sublobes, so that the front edge of the entire lobe is indented in the middle;
each sublobe contracts in width posteriorly behind the ocular or lateral angle bearing the com-
pound eyes. The ocelli are not visible, but the compound eyes are partly- preserved; they are
small, and of the usual kidney shape. The abdomen is much more-rounded than in the European
B. regine, being twice as broad as long. It consists (including the caudal spine) of nine seg-
ments. The median lobe is as broad at the end as at the base next to the thorax; there isa
median tubercle on each segment, those on the third and last segment being larger than the
others. The margin of the abdomen is broad and thin, giving rise to broad, acute, lateral spines.
The caudal spine is very long and slender, a little swollen at the base; it is also triquetal, as in
Limulus; it is nearly one-half longer than the body, i. ¢., longer than the whole body by the
length of the head, and ending in a fine, slender, needle-like point.

Length of the best preserved specimen 33™™ (including the caudal spine).

Length of body, 15™™,

Length of caudal spine, 18™™.

Length of cephalic shield, 7™™; breadth at base of lateral spine, L6™™,

Length of lateral spine, 4-5™™.

Length of abdomen, 8™™; breadth (not including the spines), 12™™.

In nodules at Mazon Creek, Illinois; Nos. 210"!, 210%, 210*, 212»; 2138, collection of Mr. Lacoe.

While having the same number of abdominal segments, this species, the first representative
of the genus which has occurred in America, differs from B. regine chiefly in the more rounded,
less triangular outline of the abdomen, and in the smaller lateral abdominal spines. . It is prob-
able that in Dr. Woodward’s figure of B. regine the compound eyes are not correctly placed. In
our specimens of Belinurus they have the same relative situation as in Prestwichia dane and
longispina.

150 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

SYNOPSIS OF THE NORTH AMERICAN XIPHOSURA.
Suborder XIPHOSURA.
Family 1. CycLip@ Pack.

Body disk-like, orbicular; abdomen composed of three segments, scarcely if at all differen-
tiated from the cephalic shield; cephalic limbs nearly as in the larval Limulus; size small.
Genus Oyclus De Koninck, with the characters of the family.
Cyclus americanus Pack.

Family 2. DIPELTID2 Pack.

Body disk-like, elliptical ; abdomen differentiated from the cephalic shield, smooth, no seg”
ments indicated.
Genus Dipeltis Packard, with the characters of the family.
Dipeltis diplodiscus Pack.

Family 3. BELINURID Pack.

Body limuloid in general shape; cephalic limbs as in the larval Limulus; shield with long
slender genal spines ; abdomen with the segments distinct; caudal spine short or long.
Genus Prestwichia Woodward. Eight abdominal segments, including the short caudal spine.
Prestwichia dane Meek.
Prestwichia longispina Pack.
Genus Belinurus Konig. Nine abdominal segments, including the very long, slender caudal

spine ; segments 7 and 8 consolidated.
Belinurus lacoei Pack.

Family 4. LoruLip 2 Zittel.

Body longer than broad; abdomen with segments consolidated ; six pairs of abdominal limbs,
five pairs having over a hundred pairs of gill-leaves.

Genus Protolimulus Packard.* Seven abdominal segments, including the large thick caudal
spine.

Protolimulus eriensis (Williams).

Genus Limulus Miiller. Cephaliclimbs large ; body longer than broad ; abdomen with 9 seg-
ments; caudal spine longer than. the body.
Limulus polyphemus (Linn.).t

“In a notice of a new Limuloid Crustacean from the Devonian, Amer. Journ. Sc., July, 1885, p. 45, Prof. H. S.
Williams described an interesting Limulvid from the Devonian of Erie County, Pennsylvania (associated with typical
Chemung fossils). Itis described as Prestwichia eriensis, the author remarking that ‘“‘its identification with Prestwichia
must be regarded as provisional.” He then adds: ‘ The following characters exhibited in the specimen are regarded as
generic and as locating it with genus Prestwichia of Woodward: (1) the elliptical head shield; (2) the genal spines which
proceed backwards more directly than inany described species of the genus; (3) the thoracico-abdominal segments anchy-
losed to form a buekler, to which is attached (4) along telson. The general outline of the whole animal resembles that
of the modern Limulus.” We have ventured, without having seen the specimen, to regard this form as probably a mem-
ber of the family Limulida, and the forerunner of Limulus. It is certainly not a Prestwichia. The body is apparently
longer than broad, and in outline it leaves a strong resemblance to the young Limulus after its first moult. This is
seen in the shape of the abdomen and the caudal spine and in their relations to the rest of the body. It also seems
probable that the abdominal segments were not free; in this respect it differs from the Belinurida, especially Prest-
wichia. Judging by the number of lateral spines, the abdomen was composed of 6 segments exclusive of the caudal
spine, thus differing from Prestwichia, which has 7, also from Limulus, which has8 pairs of lateral spines. We therefore
venture to give it the generic name of Protolimulus, and to regard it as standing at the base of the family to which
Limulus belongs. Its occurrence in the Devonian makes it a connecting link between the Upper Silurian Neolimulus
and the Carboniferous and Jurassic Limuloids. We are indebted to Prof. Williams for the use of figures illustrating
his P. eriensis.

t Besides the American species, there are three others living, viz, L. moluccanus ; (East India) L. longispina Van
der Hoven, Japan; L. rotundicauda Latr., Molucca Is, and Malacca.

Fic, 12.—Diagram of Fig. 11. A, cephalic shield; Fic. 13.—Theoretical diagram of the upper side.
B, thypostoma; © genal spines; D, tho-
tanio aon buckle apse eA,
Fic. 11.—Protolimulus eriensis (Williams). A sandstone cast ongitudinal ridges of the buckler; KK,
of the under surface, Gata ne portions of the gnathopodes; L, ?folia-
: ceous terminations of the last gnatho-
podes; M, position of the mouth.

PROTOLIMULUS ERIENSIS. After Williams.

S. Mis. 154——To face page 150.

Z|
ie
S hs

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 151

The individual development of Limulus an epitome of that of the Xiphosura.

It is interesting to observe a clearly marked exemplification of the parallelism between the
embryonic or ontogenetic development of Limulus and the geological succession as well as evolu-
tion of the suborder of which it is a type. We have already compared the orbiculo-hemis-
pherical form of Cyeclus with that of Limulus in the early stages of its embryonic life. The par-
allelism is striking. Cyclus may therefore be called an embryonic form. Again, in Prestwichia
there is a close resemblance to Limulus shortly before it leaves the egg, in what we have called
the trilobitic stage, a stage antecedent to the true larval stage, in which the abdominal segments
become consolidated. Prestwichia may then be properly designated as a larval form, while Cyclus
was anembryonic form. The latter genus embraces eleven species (ten in Europe), which exist in beds
containing the species of Belinuride. One cannot regard it as a retrograde form however, but as an
embryonic Xiphosuran, whose development became accelerated, adapting it for active adult life.
While the specimens of Cyclus have not yet shown the presence of compound lateral eyes, it is
not impossible that the animal was provided with a pair of median simple eyes. This indicates
that these were the primitive visual organs, and that the compound lateral eyes of the Belinuridz
and Limulidz were secondary acquisitions, and that their simple eyes are legacies left by their
Cyclus-like ancestors.

Cyclus, and perhaps Dipeltis, appear to represent Agnostus among Trilobites, and the sim-
ilarity between all these simple types indicates a community of descent.

The Suborder SYNZIPHOSURA.

In the Upper Silurian beds of Europe have been revealed a number of exceedingly interesting
forms, which appear to be Merostomata, but not true Xiphosura. They serve, on the one hand,
to connect the Xiphosura with the Eurypterida, and also strongly suggest the community of origin
of the Merostomata and Trilobita. They have been associated by Dr. Woodward with the Euryp-
terida,* but it seems to us, in the light of our present knowledge of the latter suborder and of the
Xiphosura, that they are types of a third group or suborder. Perhaps the more aberrant form is
Bunodes of Hichwald. All the genera have a caudal spine or telson. They are, besides Bunodes,
Hemiaspis Woodward, Pseudoniscus Nieszkowski, Bxapinurus Nieszk., and perhaps Neolimulus
Woodward belongs with them, though the last form connects the Xiphosura and Synziphosura.
They possess nearly as high an antiquity as the Eurypterida, but did not persist so long, as none
have been discovered in the Devonian or Carboniferous rocks ; hence we would infer that they were
the forerunners of the Xiphosura rather than actual members of the group. In a word, the mero-
stomatous ordinal tree divided into three main branches—i. e., the Eurypterida; the forms under
consideration, which may be designated as the Synziphosura; and the genuine Xiphosura. In
the Synziphosura the head forms a solid plate, with a slightly marked glabella or median lobe.
Compound eyes are present in Pseudoniscus, and in Exapinurus the head is produced laterally
into large genal spines. All have free uromeres or abdominal segments, and in all except Bunodes,
in which the pleurum is shaped and marked as in Trilobites, the uromeres possess lateral projec-
tions or spines. None of them show traces of limbs or of simple eyes, and all are of moderate size.

The Synziphosura may be divided into three families, which may be diagnosed as follows
(these groups appear to be, on the whole, equivalent in rank to the families of Trilobites) :

1. Head rounded; no genal spine; abdomen divided into a ‘‘ thorax,” consisting of six trilobite-like segments,
with diagonal pleural lines; ‘‘abdomen” of four segments, besides the large telson (Bunodes and Exapinurus).

Bunodide@ Packard.

2. Head one-half broad as long, with several genal spines; abdomen triangular, with nine segments and a short
telson (Hemiaspis). Hemiaspide Zittel (restricted).

3. Body oval; head short; large compound eyes; nine abdominal segments besides a short telson (Psendoniscus).

Pseudoniscide Packard,

4. Head-shield short and broad; abdomen very broad, of nine segments besides the telson (Neolimulns).
Neolimulide Packard.

* Quart. Journ. Geol. Soc., Feb., 1867.

152 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

- Fia. 16.—
Fic. 14.— nodes. Mh arora Pseudoniscus.
After Nieszkow- After Niesz-
ski. Fic. 15.—Bnnodes. kowski.
After F. Schmidt.
Fe RSS

ea)

e

Fic. 17,—Exapinurus. Af
ter Nieszkowski.

Fic. 18.—Hemiaspis. Af-
ter Woodward.

Fic. 19.—Neolimulus. After Woodward.

After the foregoing classification was mostly written out, we found that Professor Zittel, in
his excellent Handbuch der Palontologie, Bd. i, 640, 1885, has divided the suborder of Xiphosura
into two families:

1. Hemiaspide, with the following genera: Bunodes (Exapinurus Nieszk.) subgenus Hemiaspis, Pseudoniscus,
Neolimulus, Belinurus, and Prestwichia; while Cyclus and ? Halycine are regarded as genera of uncertain position.

2. Limulide, Limulus.

It seems to us that this is scarcely a natural classification, and that it would be better to sep-
arate the Silurian forms mentioned above from the genuine Xiphosura, especially as we know noth-
ing of the nature of their appendages, and to assign them, at least provisionally, to a group dis-
tinct from the genuine Xiphosura, especially since we now know something definite as to the
nature of the cephalic appendages of Cyclus and Prestwichia, their resemblance to those of the
existing Limuli being remarkably close. Certainly Bunodes, in which there are, according to F.
Schmidt’s late researches,* as stated and figured by Zittel, besides a four-jointed abdomen, a “ tho-
rax” composed of “six trilobite-like, movable segments,” cannot well be allowed a position in the
genuine Xiphosura. Moreover, the pleura of the single segments show a diagonal longitudinal
ridge. This mark is a peculiarity of the pleura of some trilobites, and does not occur in any genu-
ine Xiphosura, and aids in lending to Bunodes a trilobitic appearance.

If we separate Bunodes from the true Xiphosura, Hemiaspis will have to go with it, since it
has a rounded cephalic shield, shaped somewhat as in Bunodes, but broader. We should not, with
Zittel, regard it as asubgenus of Bunodes, because the “thoracic” segments have on the free sides
no diagonal ridge, and the cephalic shield is ornamented with large spines, which perhaps indicate
the head segments of the embryo. In both genera no eyes have yet been discovered. For the
present we should, on the whole, regard the two genera as representing different families.

*F. Schmidt, Miscellanea Silurica III, Die Crustaceen fauna der Eurypteruschichten yon Rootzikiill auf Oesel.
Mém. de Acad. impér. de St. Péterbourg, 7: ser., xxxi, 1883.

Johnnes Nieszkowski, Zusiitze zur Monographie der Trilobiten der Ostserprovinzen nebst der Beschreibung
einiger neuen obersilurischen Crustaceen. Dorpat, 1859.

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 153

In Pseudoniscus we have another form which suggests a relationship to the Trilobites. Our
figure is copied from Woodward’s restoration. Nieszkonski, the original describer, remarked,
“On the inner side of the shield we notice a place cut out, with the convexity looking outward,
which should certainly be regarded as the outer edge of the eye.”

The foregoing remarks are suggested by a study of the figures and descriptions of these remark-
able forms, and as they are not based on a study of the specimens themselves, they will be taken
only for what they are worth. But the fact remains that we have, side by side with the Euryp-
teride in the upper Silurian strata, a group which does not apparently belong to either the Euryp-
terida or genuine Xiphosura of the Carboniferous and later periods, and to which it seems best to
assign, temporarily at least, an intermediate position. The group also is of great interest as
serving to bridge over the gap between the Merostomata and Trilobita.

The following view will express the relations of the three suborders :

Order MEROSTOMATA.

1. EHurypterida. 2. Synziphosura. 3. Xiphosura.
HISTORICAL REVIEW.
I.—History of the Xiphosura.

In 1764 Gronovius, in the second fasciculus of his Zoophylacium Gronovianum, p. 220 (according
to Van der Hoeven, for we have not seen this work), proposed the name Xiphosura. His work
appeared in three fasciculi, bearing date 1763 to 1781, the second fasciculus dated 1764.

The name Limulus was first proposed by O. F. Miiller (Entomostraca, 1785, p. 124), and adopted
by Fabricius (Ent. Syst., 487, 1893).

The name Limulus polyphemus (Linn.) was bestowed by Latreille in his Histoire Naturelle des
Crustacés et des Insectes, tom. 4, p. 96, 1802,

In 1798 Latreille, in Cuvier’s Tableau élémentaire de ’ Histoire Naturelle des Animaux, placed
the Limuli in the Crustacea, under the Monoculi.

-Previous to 1806, the exact year we have not been able to ascertain, Latreille (Suite 4 Buffon,
Sonnini, Paris, 1798-1807) assigned Limulus to the Entomostracan order 1 Xiphosura (fide Milne
Edwards).

In 1806 Latreille (Genera Crustaceorum et Insectorum, i, 10) placed Limulus in order 1 Xiphosura
of Legio 1 Entomostraca.

In the same year Duméril (Zool. Anal.) associated Limulus with Caligus, ete.

In 1809 W. Martin “ gave a figure and short description of a Limulus crustacean from the coal
measures, which he included with the Trilobita.”

In 1810 Latreille (Considérations générales, etc.) assigned Limulus a place under the Entomo-
straca in Family 1, Clypeaces, Aspidiota, associating it with Apus, Caligus, and Binoculus. The
term Xiphosura does not appear.

In 1835 Latreille (Familles naturelles du Régne Animal) places the Xiphosura between the
Phyllopods, the Trilobites, and the Siphonostoma.

In 1828 Straus Durckheim (Considérations générales sur Vanatomie comparée des Animaux
articulés) referred Limulus to a new order, Gnathopoda, forming the eighth order of Crustacea,
which he placed between the Decapoda and Arachnida.

After the publication of his “Considerations,” Straus-Durckheim removed the Gnathopoda from
the Crustacea to the Arachnida, as will be seen by the following extract from Lankester’s “ Lim-
ulus an Arachnid” (Quart. Journ. Mier. Se., 506, 1881) :

Straus Durckheim maintained that Limulus should be classified with the Arachnida, but the publication of his
views on the subject appears never to have taken a very definite or satisfactory form. In fact, the only record of
Straus Durekheim’s teaching on this subject which I can find isin the French translation of Meckel’s ‘‘ General Treatise
on Comparative Anatomy.” MM. Reister and Alph. Sanson carried out this translation and added many notes in the
form of appendices to each volume. At the end (p. 497) of the sixth yolume, which bears the date 1829-1830, there is
anote headed ‘‘Sur l’appareil locomoteur passif des Arachnides,” which appears to be an abstract of amemoir “‘On the

S. Mis. 15420

154 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

Comparative Anatomy of the Arachnida,” read to the Academy of Sciences June 1, 1829, but never, I believe, pub-
lished. M. Straus Durckheim communicated its contents to MM. Reister and Sanson. From this note I submit a few

extracts. The authors commence:
“La classe des Arachnides, dans laquelle M. Straus comprend le genre Limule, formant a lui-seul un ordre designé

sous le nom de Gnathopodes et dont il isole les Pyenogonides qu'il renvoie aux Crustacées, offre dans la disposition de
son squelette et des muscles qui en meurent les diverées pieces, des particularités tellement tranchées qu’onne peut,
y méconnaitre un type different. C’est de ce squelette que sont tirés les traits principaux propres & caracteriser la
classe des Arachnides en général, et qui consiste dans la disposition des pattes rayonnant sur wn sternum commun, dans la
présence dun sternum cartilagineux intérieur, dans Vabsence d’antennes.”

The Arachnida are then divided into three orders, ‘“‘les pulmonaires, les branchiféres, et les trachéens,” but it is
not explained whether the term ‘‘gnathopodes” is to be regarded as simply a synonym of the order ‘“ branchiféres.”

With regard to the internal sternum, the citation of the views of M. Straus runs as follows:

“Dans lintérieur du thorax de tous les Arachnides, 4 Vexception peut-étre des Acarides dont la plupart des espéces
sont trop petites pour qu’on puisse les dissequer et connaitre leur organisation, on trouye une piéce cartilagineuse
diversement configurée suivant les familles, et placée dans le thorax ou dessus du sternum, cette piece, & laquelle con-
vient le nom de sternum intérieur, est maintenue librement par le moyen de plusiers muscles qui ce conduit de diffé-
rents points de sa surface sur le bouclier, ou sur le sternum extérieur auquel ils se fixent. Elle sert en outre de point
@insertion 4 un certain nombre de muscles des pattes.”

In Cuvier’s Régne Animal, nouv. édit., 1829 (tom. iv), the group named by Latreille, Pacilopoda,
is characterized and described as the second order of Entomostraca. The order consists of two fami-
lies: Xiphosura (genus Limulus) and Siphonostoma (Caligus, Argulus, etc.). As the group Pecilo-
poda, by its founder, includes the parasitic Copepoda besides Limulus, it seems advisable to drop
it, retaining the term Xiphosura, which has never been applied to any other animal than Limulus
and its allies. On p. 46 he remarks: ‘De cet ordre de crustacés on arrive 4 la classe des ARACH-
NIDES, dont Vorganisation, en général, approche beaucoup de celle des Limulus.” f

In 1830 Milne-Edwards (Ann. des Se. Nat., xx. mars 1830) adopted the order Xiphosura,
placing it below the Siphonostomata.

In 1834 Milne-Edwards (Hist. Nat. des Crustacés) retained the order Xiphosura.

Straus-Diirckheim’s views were more explicit than supposed by Professor Lankester, as in
Straus’s work, published in 1842, entitled “‘Traité pratique et théoretique d’Anatomie compara-
tive,” etc., vol. 2, 169, we find the following statement :

J’ai formé Vordre des Gnathopodes avec le seul genre Limulus. Ces singuliers animaux ont été rangés parmi les
Crustacés par tous les naturalistes qui, ne connaissant pas leur organisation, les plagaient dans cette classe par cela
seul qu’ils ont des branchies, tandis qu’ils s’en distinguent essentiellement par le reste de leur organisation, en offrant
les plus grandes analogies avee les Ara-hnides{ et Vexistence des branchies ne saurait @ elle seule constituer un
caracteére suffisant pour les éloigner de ces derniers, vu que dans cette classe les organes de la respiration n’ont plus
cette grande prépondérance sur les autres appareils du corps, pour les tenir sous leur dépendance, comme cela a lieu
chez les vertébrés; ce qui est pronvé par Vanalogie qui existe entre les Arachnides pulmonaires et les trachéens, qu’on
ne saurait séparer.

Dana (1852) in his Crustacea of the U. S. Exploring Expedition, proposed the order Merosto-
mata for Limulus exclusively, which he places in the tribe Limuloidea. He makes no mention of
the Eurypteride. The Pecilopoda in Dana’s system forms the first suborder of Cormostomata, and
include the Ergasiloidea, Caligoidea, and Lernzoidea.

In 1866 Heckel (Generelle Morphologie der Organismen, ii, Ixxxix) regarded the Trilobita as
forming the third legion of Branchiopoda. They are in his system succeeded by the sixth subclass
of Crustacea, the Pecilopoda, which embraces the two legions of XYiphosura and Gigantostraca. The
latter name is proposed for the Pterygotide and Hurypteride alone. As Heckel’s Gigantostraca
appears to be exactly synonymous with Dana’s Merostomata as amended by Woodward, the awk-
ward, meaningless term, which has never been defined, should be discarded. It has, however,
been adopted by Dohrn in 1871 (Zur Embryologie und Morphologie des Limulus Polyphemus,
Jena. Zeits, vi, 1871), and by Claus, though in a greatly extended, and it seems to us an unwar-
rantable, sense. Dohrn remarks:

Limulus is nearest related to the Gigantostraca. Both appear to be related to the Trilobites, though this rela-
tionship cannot be established in all the details. The morphological and genealogical relations of these three fami-
lies to the Crustacea are not such as to be surely determined ; perhaps they will remain always doubtful. That they
are related to the Arachnida we are not, as the matter now stands, in a position to allow. So it only remains for us
to put these three groups under a common name, for which I might adopt Hwckel’s expression “Gigantostraca,” and -
let_them take their place in the system with (neben) the Crustacea,

ON THE CARBONIFEROUS XIPHOSUROUS FAUNA. 155

Il.— Affinities of the Eurypterida to the Xiphosura (Limulide) and the formation of the order Mero-
stomata as at present received.

Tn 1825 Dr. J. E. De Kay described and figured the first (an American) species of Eurypterus
known (£. remipes), and referred it to the class Crustacea and to the order Branchiopoda.

In 1844 L. Agassiz remarked of Pterygotus:

I am rather inclined to believe that this singular animal will become the type of a family intermediate between
the Trilobites and the Entomostracans in which perhaps, the Euryptert and the Hidothee will some day be included.

We have given on pp. 177, 178 of our essay on “ The Development of Limulus” (1872) a history
of the views of James Hall, Salter, and others, especially the first-named, who proved that the
Eurypterida belonged to the same order as Limulus.

Tn 1866 in his elaborate ‘“* Monograph of the British fossil Crustacea, belonging to the order
Mesostomata,” Dr. H. Woodward formally united the Eurypterida in the same order with Limu-
lus, remarking :

Having long been eonvinced of the propriety of expressing in some suitable manner the correctness of the conclu-
sions of Professors Agassiz and James Hall as to the close affinity existing between the Zurypterida and the Xiphosura,
and being fully persuaded at the same time that they naturally form two distinct although closely related groups, I
have ventured to unite them in the Order MEROSTOMATA—a name proposed by Dr. J. D. Dana for the recent king-
crabs only, retaining at the same time the names Lurypterida and Xiphosura as suborders.

In 1872 we adopted this classification, which seems eminently natural, and has since been
adopted by a number of leading zoologists.

In 1868 Claus (Grundziige der Zoologie) characterized the order Pecilopoda, but in the third
edition of this work (1876) the Pecilopoda (restricted to Limulus), though placed between the
fourth order, Phyllopoda, and fifth order, Arthrostaca, in the Crustacea, and at the end of the Phyl-
lopoda, are associated with the Trilobita in a special group to which no special rank is assigned.

Il1.— Transfer of the Merostomata (with the Trilobita) to an independent class.

In 1869 Huxley stated in the “ Academy” (November 18) :

The Xiphosura have such close morphological relations with the Arachnids, and especially with the oldest known
Arachnidan, Scorpio, that I cannot doubt the existence of a genetic connection between the two groups.

In 1871 Prof. E. Van Beneden (Comptes Rendus de la Soc. Ent. Belgique, October 14, 1871;
Annals and Mag. Nat. Hist., January, 1872) remarked :

The Limuli are not Crustacea; they have nothing in common with the Phyllopoda, and their embryonic devel-
opment presents the greatest analogy with that of the scorpions and other Arachnida, from which they cannot be
separated. * * * The Trilobites, as well as the Eurypterida and the Pecilopoda, must be separated from the class
Crustacea, and form with the Scorpionida and the other Arachnida a distinct branch, the origin of which has still to

be ascertained.

In 1872 A. Milne Edwards (Annales des Se. Nat.) published his important researches on the
internal anatomy of Limulus, which showed that Limulus essentially differs from the Crustacea.
In the same year we attempted to show the close affinities of Trilobites to Limulus.

In 1876, according to Claus’s own statement (Annals and Mag., July, 1886, p. 56), referring to
his change of views as to the position of Limulus, he remarks:

Even in the work entitled “‘ Untersuchungen iiber die genealogische Grundlage des Crustaceensystems” (Vienna,
1876) I adhered to the views of those who, like Straus-Diirckheim, regard Limulus and Branchiate Gigantostraca as
allied to the air-breathing Arachnoidea, and the latter as having proceeded from the former, although, having regard
to the possibility of a still undemonstrated Nauplius stage, I considered it probable that the common origin of the
true Crustacea was rather after than before the Nauplius period of the Stem-Crustacean. In the case of Limulus and
the Scorpions I also asserted the homology both of the six pairs of limbs of the cephalothorax, and, with reference to the
developmental history, of the six pairs of limbs of the preabdomen, of which the second pair represents the comb-like
organ of the Scorpions, while the following four pairs immediately undergo retrogression (p. 110). In the ‘‘ Grund-
ziige der Zoologie” of the year 1880 I went so much further as to divide the Branchiata, or Crustacea, sensu latiori,
into Encrustacea (with the Entomostraca and Malacostraca) and Gigantostraca (with no certain traces of the Nauplius
stage), and accordingly I affirmed expressly of the Tracheata that in opposition to the more ancient Branchiata they ‘were
not referable to a wnitary origin, since the Arachnoidea, which are derivable from the Gigantostraca, stand opposite to the
Myriapoda and Insecta, which are wnited by a closer affinity” (p. 519).

156 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES.

In 1885 and 1886 (Annals and Mag. Nat. Hist., July, 1886) Claus regarded the Gigantostraca
as a class intermediate between the Crustacea and Arachnida. He thinks that the Arachnida
descended from the Gigantostraca, adding, “I by no means affirm the Arachnoidal nature of
Limulus.”

In 1879, in our Text-book of Zoology, as the result of Milne Edwards’s researches, we divided
the Crustacea into two subclasses, the Neocarida and Paleocarida, the latter group comprising the
Merostomata and Trilobita. In a previous paper we had shown the close homologies of the eye of
Trilobites to the compound eyes of Limulus.

In April, 1881, Mr. C. D.Waleott (Bull. Mus. Comp. Zool., viii, No. 90, p. 209), under the class
Peecilopoda, places two subelasses, viz, Merostomata and Palade (Trilobita), giving definitions
of the groups.

In 1881, in his article “Limulus an Arachnid” (Quart. Journ. Mier. Se.) Prof. E. Ray Lan-
kester proposed the term Hematobranchia, which he regarded as the equivalent of Merostomata.
This group of the class Arachnida, as understood by Lankester, embraces the three orders: 1,
Trilobita; 2, Eurypterida; and 3, Xiphosura.

In 1885 (Embryology of Limulus Polyphemus, I1I, Proc. Am. Phil. Soc., January, 1885), we
referred Limulus, with the Eurypterida and Trilobita, to a class by themselves.

In 1885 Mr. J. S. Kingsley associated the Limulus with the Arachnids as a group by them-
selves, to which he gave the name Acera (Science News and Quart. Journ. Micr. Se.).

In 1886, in the 5th edition of our Text-book of Zoology, we suggested the term Podostomata
for the class comprising the two orders Merostomata and Trilobita.

TV.—The class Podostomata.

It thus appears that while at the present date (1886) A. Milne Edwards, E.Van Beneden, and
B. R. Laneaster regard Limulus and its allied forms as belonging to the Arachnida, and J. S.
Kingsley associates the Limulus and the Arachnida in a group by themselves under the name
Acera, the present writer and Professor Claus regard the Merostomata with the Trilobites as
forming a class intermediate between the Arachnida and Crustacea.

We have endeavored to show that the names Pecilopoda and Gigantostracea have been applied
in such different senses by different authors that they cannot well be retained for the Merostomata
and Trilobita taken together in the sense we advocate. We have therefore proposed the term
Podostomata for this class of Arthropoda. Itis derived from zovs, zod0's, foot, and oro a, mouth,
in allusion to the foot-like or ambulatory nature of the cephalic appendages which surround the
mouth in a manner characteristic of the group. ,

The class Podostomata may be defined as a group of Arthropods, in which the cephalic
(Limulus) or cephalo-thoracic (Trilobites) appendages are in the form of legs, %. ¢., ambulatory
appendages, usually ending in forceps, or large claws (chele), which in the sole living representa-
tive of the class are arranged in an incomplete circle around the mouth; the basal joint of each
leg is spiny, so as to aid in the retention and partial mastication of the food. No functional
antenne, mandibles, or maxilla. Eyes both compound and simple. Respiration by branchiz
attached to the abdominal appendages, which are broad and lamellate in Merostomata and probably
cylindrical with narrow gills in Trilobita. The brain (procerebrum) supplying nerves to the eyes
alone; the nerves to the cephalic or cephalo-thoracic appendages originating from an esophageal
ring; the ventral cord ensheathed by a ventral arterial system more pertectly developed than in
insects or scorpions; coxal glands highly developed, with no external opening in the adult. This
class differs from the Arachnida, among other characters, in having no functional cheliceres (*‘ man-
dibles”) or pedipalps (‘‘maxille”); in the cephalic appendages either ending in large claws or
forceps, or simple, the terminal joint not bearing a pair of minute claws or ungues like those of
Arachnida and Insecta, enabling their possessors to climb as well as walk. Podostomata have no
urinary tubes. Limulus undergoes a slight metamorphosis, while in Trilobites the adult differs
from the larva in having a greater number of thoracic segments.

OF THE CARBONIFEROUS XIPHOSUROUS FAUNA.

157

From the Crustacea the Podostomata differ in the lack of functional antennze and mouth
parts ; in the brain innervating the eyes (compound and simple) alone; in the shape of the head
and of the pygidium or abdominal shield, and in the arterial coat enveloping the ventral nervous

cord.

The Podostomata are divided into two orders :

I. Merostomata with three suborders, Xiphosura, Synziphosura, and Eurypterida.

Il. Trilobita.

Fig.
All the Cen on this plate drawn by Dr. J. S. Kingsley, with corrections by the author.

Explanation of Plate ¥.

. Cyclus americanus Pack. X#. la, lateral view restored. X?.

. Dipeltis diplodiscus Pack. Natural size; 2a, the same restored. X#.
. Prestwichia dane Meek. Natural size; restored; dorsal view.

. Prestwichia dane Meek. Natural size; partly restored; ventral view.
. Prestwichia longispina Pack. Partlyrestored. Xj.

. Belinurus lacoei Pack. Partly restored. X2.

Explanation of Plate VI.

Fig. 1. Prestwichia dane, showing the limbs; 1a, the reverse.

Fig.
Fig.
Fig.

2.
3.
4.

Prestwichia dane, showing the interior; 2a, the same, another specimen.
Prestwichia longispina, natural size.
Cyclus americana, natural size; 4a, reverse of the same.

From photographs taken by Mr. R. L. P. Mason, Providence, R. I.

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
All the figures of natural size and from photographs taken by Mr. F. O. Draper, Pawtucket, R. I.

1
2
3.
4
5

6.

Explanation of Plate VII.

. Paleocaris typus M. & W., natural size.
. Paleocaris typus M. & W., natural size.

Anthrapalemon gracilis M. & W., carapace laterally flattened.

. Anthrapalemon gracilis M. & W., from a small specimen without the carapace.
. Anthrapalemon gracilis M. & W., carapace wanting.

Antlerapalemon gracilis M. & W., carapace wanting.

ae ae

t

Aer

MEMOIRS NAT ACADFMY SC.VOL TI PLATE V

i
.

= = ae
‘ Julius Bien & Co.lith

J S.Kingsley, de!

CARBONIFEROUS XiPHOSURA.

MEMOIRS NAT, ACAD. SC. VOL. III PLATE VI

FIGS. 1. 2. PRESTWICHIA DANA; 3. P. LONGISPINA: 4. CYCLUS AMERICANUS

ale

—

EWISe MOI

ay)

ers
af

‘
om
J

4
‘SIME re ts

euMer

. are

MEMOIRS NAT. ACAD. SC. VOL. III. PLATE VII

FIGS. 1. 2. PALEZOCARIS TYPUS; 3 —6. ANTHRAPALAMON GRACILIS

FROM PHOTOGRAPHS BY F. D. DRAPER.

7

Ne

Rae

Ry.

we

pit hy
Pater