Paleontological Bulletin, No. 40. 
A CONTRIBUTION TO THE 
VERTEBRATE PALEONTOLOGY 
OF 
BRAZIL. 
{Read before the American Philosophical Society, April 17, 1885.) 
SECOND CONTINUATION 
OF 
RESEARCHES AMONG THE BATRACH1A OF THE COAL 
MEASURES OF OHIO. 
(Read before the American Philosophical Society, June 19, 1885.) 
BY E. D. COPE. 
FOR SALE BY 
JL. ZE. FOOTE, 
1223 BELMONT AVENUE, 
PHILADELPHIA. 1 
April 17, 1885.1 
[Cope. 
A Contribution to the Vertebrate Paleontology of Brazil. By E. D. Cope. 
{Read before the American Philosophical Society, April 17, 1885.) 
Professor Orville A. Derby, Director of the Geological Section of the 
National Museum of Brazil, has desired me to furnish to the museum an 
account of the extinct Vertebrata from the various strata found within 
the limits of the Empire, which are preserved in the Museo Nacional at 
Rio de Janeiro. I have also received a considerable collection made in the 
neighborhood of Bahia by Mr. Joseph Mawson, of London, England, 
which has aided me much in the determination of the extinct fauna of 
that region. 
The horizons from which the species now enumerated have been derived 
are the Pliocene, the Cretaceous and the Carboniferous. The work hitherto 
done in this field is small in amount. The researches of Lund and Rein- 
hardt into the fossils of the caves of Brazil are well known. Professor 
Owen has determined the existence of Crocodilia and Dinosauria in the 
Cretaceous beds near Bahia ; and Professor Marsh has described a gavial 
from the same horizon and locality. Professor J. S. Newberry has iden- 
tified some fishes from Ceara in Eastern Brazil as of Jurassic age ; and I 
have described a genus and species from the same locality. In more 
detail, the localities and horizons from which the specimens of vertebrate 
fossils of the Museo Nacional have been derived, are the following, so far 
as determined : 
Pliocene, Pampean. 
North-eastern pt. of Province of Bahia ; Toxodon expansidens, sp. nov. 
Cretaceous ? Laramie. Near Bahia. Diplomystus longicostatus, sp. 
nov. Chiromystus mawsoni, sp. nov. 
Fox Hills. Province of Pernambuco. Hyposaurus derbianus, sp. nov. ; Cope.] 
2 
[April 17, 
Enchodus subaquilateralis, sp. nov. ; Galeocerdo pristodontus Agass. ; 
Apocopodon sericeus, sp. nov. 
?- . Province of Sergipe del Rey. Pycnodus flabellatus, sp. nov. 
Jurassic. Province of Ceara. Ancedopogon tenuidens Cope. Aspido- 
rhynchus, sp. 
Carboniferous. Province of San Paolo. Stereosternum tumidum, 
sp. nov. 
The following pages contain the detailed descriptions of the new species, 
and the determination of their affinities. Others yet remain to be deter- 
mined. 
PISCES. 
Apocopodon sericeus, gen. et sp. nov. Myliobatidorum. 
Char. gen. Founded on teeth which formed a pavement like that of 
Myliobatis, but which are mostly separated in the specimen. These con- 
sist of longer ones of a median series, and smaller ones of the lateral series. 
The teeth of the median series are shorter than in the typical forms referred 
to Myliobatis, having rather the proportions characteristic of Zygobatis. 
They differ from the corresponding teeth in both genera in being exactly 
parallelogrammic in outline ; that is, the extremities are truncated instead 
of angulated as in those genera.* The lateral teeth display the usual an- 
gulation among themselves, though doubtless joined by a straight suture 
to the middle row. The roots are well distinguished from the crowns, and 
are short. Their grooves are very shallow, or merely indicated. The 
triturating surface is covered by a dense layer which is wrinkled like the 
sides of the crown, and is continuous with it. 
Char, specif. The teeth are robust and indicate a species of considerable 
size. The crowns are considerably more elevated than the roots, and have 
perpendicular sides. The sutural surfaces are straight, and marked by fine 
grooving which runs at right angles to the grinding face, and is continu- 
ous with the wrinkling of the latter on the long sides of the crown. From 
this it follows that the wrinkling crosses the grinding face at right angles 
to its long diameter. There are in the wrinkling six ridges to a millimeter. 
The roots are constricted from the crowns by a groove, which is itself 
divided by a narrow collar-like rib, resembling cement, which is ex- 
pressed on the junction of two pieces by pressure, grown cold. The 
sizes of the teeth diminish externally. The roots of those of the antepe- 
nultimate are crossed by four shallow grooves, and those of the penulti- 
mate by two. External row lost. Six grooves cross the root of one of the 
larger teeth. 
Measurements. 
M. 
r transverse 
024 
Diameters of larger tooth-j anteroposterior.' 
013 
1 vertical 
021 
Vertical diameter of root of do 
00$ 
•One endqf one of the large teeth has the usual two faces. 1885.] 
3 
[Cope. 
Measurements. 
M. 
Diameters of root of antepenult tooth 
( transverse 
.015 
C anteroposterior 
.010 
Diam. of root of penultimate of row 
transverse 
.007 
anteroposterior.. 
.010 
From Maria Farinha, Province of Pernambuco. Probably of Fox Hills 
or Maestrichtian Cretaceous age. Coll., No. 306. 
Enchodus sub.equilateralis, sp. nov. 
This species is represented by a premaxillary bone bearing the long 
laniary tooth characteristic of the genus, and by another osseous fragment 
bearing a similar tooth, which may perhaps belong to the distal part of the 
dentary bone. I describe the first-named specimen. The fragment of the 
premaxillary is so small that little can be said of it, except that its surface 
is smooth, and but slightly convex, and that it projects but little beyond 
the long tooth. The tooth is long and slender, and has a very slight sig- 
moid fore and aft curvature. It has two opposite cutting edges, the ante • 
rior of which reaches to its base, and the posterior for half of its length. 
The inner face of the tooth begins to be more convex than the external 
at about the middle of its length, but this convexity is not much marked 
beyond the basal fourth. The surface of the tooth is smooth everywhere. 
This species is readily distinguished from such species as E. mortoni, 
where the edges are not opposite. From the E. carinatus and E. gladiolus, 
where the edges are opposite, the smooth surface separates it. In the allied 
E. dolichus* the posterior cutting edge only extends one-quarter the length 
of the tooth. 
Length of crown M. .023 ; diameters at middle, long, .004 ; short, .0023. 
Diplomystus longicostatus, sp. nov. 
This herring is represented by numerous specimens, and possesses well- 
marked characters. These may be stated in general thus: The caudal 
part of the vertebral column is very short. The abdomen is very deep 
and the ribs are long. The caudal fin is deeply forked, and has long acute 
lobes. The other fins are very small. 
The scales are so attenuated as not to be countable. The scutes of the 
median dorsal line are longer than wide, and are emarginate behind, and 
hence cordate. The superior surface of two of them is roughened with 
radiating ridges. The inferior surfaces are smooth. None behind the 
dorsal fin. , 
The dorsal outline rises gradually to the dorsal fin, and then gradually 
descends to the caudal peduncle. The general convexity is slight. On 
the other hand the abdominal convexity is very great, and is especially 
protuberant below the dorsal fin. The depth at this point enters the total 
length, minus the caudal fin, one and five-sixth times. The length of 
the head enters the same three times. The superior surface of the head 
*See Report on Cretaceous Vertebrata of the West, E. D. Cope, p.300. Cope.] 
4 
[April 17, 
slopes gently from the dorsal line ; hence the pectoral outline is very 
steep. The head is a good deal injured in the typical specimen, but it is 
somewhat longer than deep. 
The middle of the base of the pectoral fin is half-way between the ver- 
tebral column and pectoral border inclusive. The dorsal fin begins along 
the anterior border of the fourteenth vertebra. It is elevated in front, and, 
having a short base, has a rapidly descending posterior outline. The anal 
fin originates much behind the posterior border of the dorsal. It is also 
short and weak. Formula, D. 10 ; C. -f-18 +; A. 8. Vertebrae, Abd. 
24 ; C. 10 ; only one included between the external caudal rays. Neural 
and haemal spines weak and rather short. Ribs long and robust. Abdom- 
inal scuta rather large, and with a free posterior accumination. The sup- 
plementary ribs, if they ever existed, are not preserved. Ventral fins lost 
from the typical specimen. 
Measurements. 
M. 
Total length (axial) 
126 
Length to basis of caudal fin 
096 
Depth at free edge of operculum 
044 
" " " " " dorsal 1st ray 
055 
" " " " " anal " " 
020 
Length of dorsal fin J in front 
t on base 
016 
014 
Length of anal fin {*n 
. .007 
I on base 
010 
Length of a caudal lobe from base 
035 
" " abdominal vertebrae 
056 
" f' caudal " 
020 
The specimens are from the coast near Bahia. The type comes from 
near Itacaranha, where it was found by Mr. Joseph Mawson. Other 
specimens are from the same locality, while others are from Plataforma 
and Agua Comprida. In none but the type do I find the dorsal scuta pre- 
served. 
This genus has hitherto been only known from the Lower lacustrine 
Eocene of North America. Its occurrence in this supposed marine forma- 
tion indicates that, like its close ally Clupea, Diplomystus has considerable 
range in time and space. The D. longicostatus falls into the section of the 
genus represented by D. humilis Leidy. From this and the allied D. altus, 
it differs in the more numerous abdominal and less numerous caudal ver- 
tebrae, and the longer lobed more deeply furcate caudal fin. 
Chiromystus mawsoni, gen. et sp. nov. 
This new genus and species are indicated by a single large specimen 
from the same horizon as the Diplomystus longicostatus. It is nearly com- 
plete, with the important exception that the head and a few anterior dor- 
sal vertebrae are wanting. Tlje impression of the scapular arch, however, 
gives the position of tl)e skull, and the anterior fibs give a clue to the 1885.] 
5 
[Cope. 
character of the anterior dorsal vertebne. From these it appears that the 
genus is Isospondylous and not Plectospondylous. 
Char. gen. Dorsal fin small, above the anal, which is moderate. Pecto- 
ral fin with several superior rays thickened and robust. Caudal fin fur- 
cate. Ventrals small. No ventral or dorsal scuta. Scales much attenua- 
ted. No basilar interneurals or hmmals. 
This genus may belong either to the Hyodontidoe or Chirocentrida so far 
as the characters given by authors are concerned, since the only distinc- 
tions given are found in the soft parts. I have pointed out* that the parie- 
tals are in contact, and the caudal fin embraces two vertebrae in the Hyo- 
dontidce, while in the Chirocentridce the parietals are separated by the supra- 
occipital, and there is but one caudal-fin vertebra as in the Clupeidce. I can 
only observe the caudal fin in Chiromystus, and find that it includes two 
vertebrae, as in the Hyodontida. 
Char, specif. The form is rather elongate. The depth of the longest 
ribs, and vertebra corresponding, enter the length, exclusive of the head 
and caudal fin, four and a third times. Vertebrae, Abd. 28 ; C. 22. The 
anterior dorsals are obtained by counting the ribs, and three are added to 
the caudals visible, in order to fill up an interruption caused by fracture. 
The centra are longer than deep, and have two lateral longitudinal fossae, 
bounded above and below by a narrow rib, and separated by a flattened 
rib. 
The posterior part and apex of the dorsal fin are wanting, so that its 
characters cannot be given, except by stating that the rays are slender 
and weak. The anal fin is injured at its posterior extremity, but by 
counting the interhaemal bones I find the rays to number sixteen. The 
four superior pectoral rays are very robust, the inferior most so. The 
three upper are preserved, and it can be seen that they are compressed 
and smooth, and not segmented. The caudal fin is very deeply forked, 
and the lobes are long. Each one consists of six strong external rays, 
besides the fulcral rays, and a number of very fine rays on the inner side 
of these, giving each lobe a narrow form. The scales are extremely 
attenuated, and cannot be counted. The ventral fins are quite small, and 
the rays may not all be preserved, although those that are visible are in 
place. They number only four. 
Measurements. 
Length of vertebral column 
M. 
310 
" " a lobe of the caudal fin , 
.100 
" from base of ventral to base anal 
.081 
" " " " anal to base caudal 
.085 
" " " " dorsal to base caudal 
.063 
" superior spine pectoral fin...., 
.065 
" of ventral fin 
.027 
Diameters of last abdominal vertebra ■ 
j longitudinal... 
.008 
! vertical 
.0075 
* Proceedings Amer. Assoc. Adv. Sei., Vol. xx, p. 333. 6 
Cope.] 
[April 17, 
The specimen was obtained near Agua Comprida, near Bahia, by Mr. 
Joseph Mawson. I dedicate the species to him with much pleasure, in 
recognition of the valuable service rendered by his collection in the pres- 
ent investigation. 
Pycnodus flabellatus, sp. nov. 
A slab of limestone contains a skeleton of this fish, but the latter is in 
several points imperfect. The head anterior to the orbit is wanting, and 
the superior half of the anterior dorsal region is broken away. The ven- 
tral fins are lost. Some isolated teeth are of the proper size to belong to 
this species, and will be mentioned later. 
The outline of the profile of the body is discoidal, and the axis of the 
skull (vomer, etc.) is directed obliquely downwards at an obtuse angle 
with the vertebral column. This requires an extensive production of tbe 
operculum above and posterior to its articulation with the hyomandibular. 
The clavicle is slender, while the coracoid is produced backwards be- 
low the pectoral fin, its superior border being concave to the first rib, 
which is overlapped by the posterior edge. The coracoid also extends 
downwards and forwards as usual. The four basilar bones of the pectoral 
fin are rod-like, and are moderately expanded distally. The determina- 
tion of this point is of much importance in fixing the position of the Pyc- 
nodontidce in the system. The basis of the dorsal and ventral fins de- 
scend steeply downwards to a narrow and very short caudal peduncle. 
The caudal fin is of characteristic form. Its lobes are long, but they di- 
verge so widely that the posterior edge of the fin is slightly convex from 
tip to tip. Radii, D. 53 ; C. 3 + 40 + 2 ; A. 24. 
The constitution of the vertebral column is not easy to make out. Only 
the anterior half is preserved. This displays the usual superior and inferior 
plates. In the present species the edges of these are in contact, so that 
the condition of the centrum, if there be any, as to ossification, is not posi- 
tively determinable. The neural spines above their basal expansions 
are connected by a series of longitudinal teeth which interlock closely so 
as to resemble a series of ribs. On examination it is found that half of 
these originate from one neural arch, and half from the other, there being 
Six or seven in all. A slight protuberance, probably for the rib-head, ap- 
pears 3 mm. below them. The true ribs are broadly alate, so as to form a 
continuous wall. The rhabdopleurs agree in number with the vertebrae, 
and are present to the end of the vertical fins. On the caudal region they 
extend downwards .66 the length of the haemal spines. The latter ex- 
tend to the superior apices of the interhaemals. The rhabdopleurs are 
not segmented as is represented in some species of this family. The 
caudal fin includes one or two vertebrae. There are two short, widely ex- 
panded hypurals, much as in Physoclystous fishes where they are distinct. 
Vertebrae, D. 19 ; C. 15 or 16. 
Measurements. 
M. 
Total length to anterior edge of orbit 
172 
Longitudinal diameter of orbit 
01G 1885-1 
7 
[Cope. 
Measurements. 
M. 
Distance from orbit to free edge of operculum 
022 
Length of vertebral column 
116 
Diameters of caudal fin J anteroposterior 
027 
093 
Depth above rib-heads at front of dorsal fin 
055 
" below haemal plates, front of anal fin 
044 
The teeth preserved are loose medians, and perhaps laterals, but the 
reference of the latter is uncertain. The crowns of the former are a little 
more than twice as wide as long, and have the extremities a little oblique. 
The summit is a little flattened, and the sides project a little beyond the 
base. The surface smooth. Length, M. .010 ; width, .0045. 
The peculiar form of the caudal fin distinguishes this species from most 
of the known members of the family Pycnodontidce. The feeble dorsal 
and anal fins distinguish it from others, and the discoidal form from still 
others. 
The structural characters observed in the specimen described have 
been instructive, especially those of the pectoral fin. These confirm alto- 
gether my reference of the family of the Pycnodontidce to the Isospondyli 
as distinguished from the Ilalecomorphi* 
The typical and only specimen of this species in the collection is from 
the southern centre of the Province of Sergipe del Rey. It is on a slab of 
cream-colored calcareous rock which has a coarse slaty cleavage, and 
probably belongs to the Cretaceous formation, and is of marine origin. 
? BATRACHIA. 
Stereosternum tumidum, gen. et sp. nov. 
Char, gen., etc. This genus is known from numerous vertebrae and ribs, 
sometimes forming consecutive series, but more frequently isolated ; but 
especially from two slabs, which exhibit the posterior part of a skeleton; 
i. e., dorsal vertebrae and ribs, pelvis and posterior limbs, and caudal ver- 
tebrae. 
The dorsal vertebrae present some of the general characters of the rep- 
tiles and batrachians of the Permian period. One of these is the existence 
of a notochordal canal. The small size of the vertebral centrum as com- 
pared with the arch and its appendages constitutes a resemblance to the 
batrachian class ; as also do the horizontal position and weak development 
of the zygapophyses. On the other hand the simple articulation of the 
ribs resembles that of the Lacertilia in general, though not of any known 
group of that order ; and has no resemblance to any known reptile of the 
Carboniferous period. 
The vertebral articular surfaces are both funnel-shaped, the anterior 
deeply, the posterior shallowly excavated. The dorsal centra are undi- 
* On the classification of the Extinct Fishes of the Lower Types. Proceeds. 
Amer. Assoc. Adv. Science, 1878, p. 292. Cope.] 
8 
[April 17, 
vided, and the notochordal canal is small. The caudal vertebrae have a 
groove, more or less obliterated by codsiflcation, surrounding the middle 
of the centrum, and cutting off a part of the base of the neural spine above. 
This looks as though the genus possesses intercentra, which were primi- 
tively separated by the protovertebral fissure. The posterior part of the cen- 
trum carries chevron bones, which are distinct from it. Besides the zyga- 
pophyses, there is, in the dorsal vertebrae, a modified form of zygosphen, 
though there is no zygantrum. The former consists of a roof-like projec- 
tion of the neural arch above each prezygapophysis, which is applied to the 
superior surface of the postzygapophyses. In some of the vertebrae, this 
zygosphenal roof is horizontal; in others it is slightly oblique, rising out- 
wards on each side, in the manner of a true zygosphen. It differs further 
from a true zygosphen in being fissured vertically, above the neural arch, 
but there is no corresponding process of the adjacent vertebra to occupy 
it. On the contrary there is a corresponding fossa of the posterior side of 
the vertebra in front. These fossae may be points of insertion of ligaments 
which strengthen an articulation otherwise weak. 
The ribs appear to be coossified with the centra, so that it is difficult to 
say whether they are truly ribs or diapophyses. In one specimen, the 
proximal ends of the ribs are seen to be expanded, and applied to the cen- 
trum so as to embrace it. These expanded extremities are simple and are 
separated on the median line of the centrum by a narrow space. Others are 
not so expanded proximally, but contract to their connection with the cen- 
trum. In some of the centra each side is produced into a depressed coni- 
cal apex in the position of a diapophysis. The position of these vertebrae 
is uncertain. The ribs are long, cylindric, curved and remarkably robust, 
having characters like those of the genus Ischyrosaurus of the Laramie 
formation, or of Mesosaurus of Gervais. They could not have had any 
movement on the vertebrae. 
The scapular arch is represented by ft coracoid bone, which though isola- 
ted, is lying on a slab with numerous remains of this genus. As no other 
form is represented on the slab, I suppose the coracoid to belong to Stereo- 
sternum. It is expanded fore and aft, most so posteriorly, and possesses a 
supracoracoid foramen. Its internal border presents a deep notch oppo- 
site the glenoid cavity. 
Portions of several humeri are preserved. They demonstrate either that 
the head is subround, or that if expanded it is at right angles to the distal 
end. The latter is perforated near one of its borders by an epicondylar 
foramen, but whether entepicondylar or ectepicondylar, I cannot ascer- 
tain. The opposite foramen is represented by a shallow groove at the dis- 
tal end of the opposite side. There are no well marked condyles of the 
humerus. 
The head of the femur is truncate and subround, and without trochan- 
ter. The shaft is subround and is of considerable length. There are no 
distinct condyles, but the articular surface is convex anteroposteriorly. 
The tibia is a stouter bone than the fibula, and its distal extremity is ex- 1885.] 
9 
[Cope. 
panded cutwards. Its tarsal articular suface forms an acute angle with 
the long axis of the shaft, presenting outwards. It has besides a slight 
distal transverse truncation. The fibula has a robust head and is slender 
distally. The tarsus consists of seven bones. These are a tibiale, an in- 
termedio-centralo-fibulare, and a tarsale corresponding to each of the five 
metatarsals. There is a foraminal notch on the internal edge of the inter- 
mediocentralo-flbulare, next to the tibiale. The bones of the foot beyond 
the tarsus are well distinguished from each other. The metatarsals are 
rather slender, and are considerably longer than the phalanges of the first 
row. The phalanges are not much shortened, but diminish in length 
regularly to the end. The ungual phalanges are not preserved in a per- 
fect condition on any of them. The proximal portion remains on the 
second digit, and it is depressed, offering no indication of a claw. The 
first toe is not shortened, and appears to be longer than the second. Its 
distal segments are lost. Neither the metatarsals nor the phalanges have 
distinct condyles, but are truncate in the vertical direction. 
Abdominal protective armature is present in the form of osseous rods. 
Several of these rods form a single girdle. They are not connected with 
the ribs. 
The pelvis is partially preserved in the specimen on the slab. Both 
pubes and ischia are well developed, and if there is any obturator foramen 
it is very small and median in position. It probably does not exist, but I 
am precluded from certainty by the condition of the specimen at the point 
of crossing of the median and transverse sutures. The pubis is not so 
large as the ischium, and has a foramen near its posterior border. The 
ilia have less transverse, and greater longitudinal expanse than the pubes, 
and are in contact on the middle line throughout most of their length. 
Affinities. It is not easy to decide as to the position of this genus. 
While many of its characters are reptilian, some of them are batra- 
chian. Of especial interest in this connection is the structure of the pel- 
vis. Its characters are only like those of some of the Urodele Batrachia, 
and the Theromorphous Reptilia. It is, however, quite certain that it does 
not belong to any known family of either class. The vertebrae might be 
those of a Theromorph reptile, and the pelvis also agrees with that of those 
animals. The abdominal rods are found in species of that order referred 
to the genus Theropleura. The ribs and tarsus are however of an en- 
tirely different type. The former would refer the genus to the Rhyncho- 
cephalia or the Sauropterygia, and there is nothing known in its structure 
which positively forbids either reference, unless it be the character of the 
pelvis. It differs from the types of the Batrachia which it most resem- 
bles, the Protonopsidae, in the replacement of the cartilaginous plate 
which represents the pubis by two osseous plates. It presents a near 
resemblance in important characters to the genus Ichthycanthus* which 
*1 refer to the IcMhycanthus ohiensis from the description and from memory, 
as the specimen is not at present accessible. The I. platypus is one of the Rhach- 
itomi, and has in the tarsus, astragalus, calcaneum, navicular, and five dis- 
tinct tarsals of the second row. Cope.] 
10 
[April 17, 
I described from specimens procured by Professor Newberry, in the 
coal measures of Linton, Ohio.* The peculiar structure of the tarsus 
is identical as to the number of its elements, and the other characters 
agree in general. There are important differences also, which would 
refer Ichthycanthus to another family. Thus the dorsal vertebrae have the 
centra deeper than long, and the ribs are free. In the absence of the 
skull, it is not possible to be sure as to which of the classes, Reptilia and 
Batrachia, these genera represent. 
Another form presents some important points of resemblance ; that is 
the genus Mesosaurus of Gervais.f The M. tenuidens Gerv. was brought 
by Verreaux from an undetermined formation of Griqualand, South 
Africa. The specimen, like that of the Stereosternum tumidum, is exposed 
on a slab, and embraces only the head, neck, thorax and anterior limbs. 
As the dorsal vertebrae are obscured by matrix the only point in which actual 
comparison can be made is the ribs. These are quite identical in the two 
types, but the articulations with the vertebral centra are invisible in the 
Mesosaurus. There are apparently impressions of abdominal dermal rib- 
lets, but they are suspected by Gervais to be the tracks of Annelids. Ger- 
vais thinks the skull has but a single condyle. The scapular arch consists 
of coossified scapula and coracoid, but clavicle, prsesternum and sternum 
are not visible. The coracoid is different in form from that of Stereoster- 
num. The humerus is, on the other hand, almost identical, and the carpus 
is nearly what one would expect to find in the Brazilian form. There are 
in the first carpal row, two large bones, and in the second, four small 
ones. 
Habits.-The structure of the limb articulations and those of the ele- 
ments of the posterior foot show that this was a genus of aquatic habits. 
The firm attachment of the ribs shows further that this type had no inter- 
costal respiration, but used its sublingual or its abdominal muscles, or 
both, in the act of inhaling air. We may suppose that in its aquatic habi- 
tat it retained air in the lungs for considerable periods, and only respired 
on reaching the surface of the water; or later investigation may show 
that it is branchiate. 
Geological position.-The peculiar characters of this form and the diffi- 
culty of determining its true position in the system, present an obstacle 
to the interpretation of its probable geological age. It has a good many 
resemblances to the suborder Choristodera of the order Rhynchoce- 
phalia (represented by the Champsosauridae). This type first appears in 
the Laramie or latest Cretaceous, and continues only to the top of the 
lower Eocene. The order Rhynchocephalia is an unsatisfactory one for 
geological purposes. It still exists in one genus, the Hatteria of New Zea- 
land, and may have existed in the Trias ; although this is not certain. 
Prof. Derby informs me that some specimens of Schizodus have been 
found in the same beds, and he therefore infers that their age may belong 
♦Proceedings Amer. Philosoph. Society, 1887, p. 573. 
tGeneral Zoology and Paleontology. 1885.] 
11 
[Cope. 
to the Coal Measures or to the Permian. There is nothing in the charac- 
ters of the genus Stereosternum to contradict such a supposition. The 
primitive characters of various parts of the skeleton and the obvious re- 
semblances to Ichthycanthus, add probability to such a view. 
Specific characters.-These may be first drawn from the specimen of the 
slab already alluded to. 
The relative length of the body is not certainly known, as it is only par- 
tially preserved in the specimens sent by Prof. Derby. To judge from 
the one above referred to, it has the ordinary proportions of a lacertilian. 
The hind legs are well developed, as for example in an Iguana. The tail 
is well developed, but its length is not determinable as the distal por- 
tions are lost. 
In the slab specimen the dorsal vertebrae are split or otherwise damaged, 
so that I describe them preferably from other specimens. The few that are 
well preserved show characters identical with the latter. I derive the fol- 
lowing however from the slab specimen. The dorsal vertebrae have the 
neural spines well developed but not much elevated. In profile their ver- 
tical diameter is about equal to their anteroposterior, and the superior bor- 
der is squarely truncate. They diminish in height posteriorly. The 
spines are present at the lumbar vertebrae. The shafts of the ribs have 
a round section. The proximal portions are for a short distance abruptly 
incurved to the vertebral body. The distal extremity is pointed. The tissue 
is dense, and there is no medullary cavity. In the lumbar vertebrae the rib is 
much more slender, and is shorter. It is coossified with the centrum. The 
caudal vertebrae have strong diapophyses, which are acuminate and de- 
pressed. In the anterior caudals they are recurved at the extremity, but 
shorten rapidly posteriorly and are transverse. In the specimen they are all 
separated from the centrum by a fissure which appears to be too constant 
and too regular to be regarded as a fracture. I suspect therefore that the 
diapophyses are free, and are joined to the centrum by a simple truncate 
head, which has an outline nearly round. This view is confirmed by the 
presence on each side of the centrum of the median caudal vertebrae, of a 
bone which resembles the sesamoids of the feet of Mammalia, which is quite 
free from the centrum, and is applied longitudinally to its anterior half. 
It is probably the rudimental diapophysis. The posterior caudals have no 
diapophyses. The latter are the only ones in the specimen which are well 
preserved. 
The bodies of the caudal vertebrae have a low ridge in the place of the 
fissure which is seen in some other specimens to divide them into equal an- 
terior and posterior halves. The entire centrum is longer than deep or 
wide, and is a little deeper than wide in section. The neural arch is 
divided into two parts by the characters of the surface. The anterior 
half is swollen and roughened by minute pits, and is separated from the 
less prominent posterior half by a pair of small fossae, one above the other. 
The neural spine stands entirely on the posterior half, and is thus widely 
removed from the prezygapophysis, which is above the anterior border of Cope.] 
12 
[April 17, 
the centrum. The neural spine is slender and rather elevated, and is sub- 
cylindric at the base, and has a narrow compressed apex, with rounded 
extremity. The chevron bones are quite slender. 
The shaft of the femur is nearly straight, and its distal half is moderate- 
ly compressed from before backwards. The tibia is generally flattened. 
Its interosseous border is shorter than its internal border, and is strongly 
concave. The internal border is gently convex. The shaft is narrower than 
the proximal end, which is narrower than the distal end. The fibula has 
an enlarged subtriangular head. The shaft is gently curved, the concav- 
ity being, as in the case of the tibia, on the interosseous side. The inter- 
medio-calcaneum, or, according to Baur's view, the astragalocalcancum, is 
much the largest bone of the foot. It has a truncate side in contact with 
the tibia, and a concave interosseous border. The rest of the outline is con- 
vex, with a slight truncation for the fibula, and one between the tibial 
border and the posterior notch. The greatest extent of the bone is trans- 
verse, and the greatest longitudinal diameter is in line with the fibula. The 
tibiale has a T-shaped outline, but the spaces below the transverse extrem- 
ity and the shaft are filled to the truncate narrower extremity of the shaft. 
The wide end also has the angles rounded off. The tarsals of the second 
row are longitudinal wide ovals, excepting the first, or internal, which is 
round. 
The extremities of the metatarsals are depressed ovals, and are wider 
than the middle of the shafts. The phalanges are more depressed. The 
metatarsals and phalanges of the fifth digit are the shortest, and the lengths 
of these elements steadily increase to the first. The phalanges of the first 
digit are lost excepting the first; and the ungual phalanges of the third, 
fourth and fifth are wanting. Adding the latter, we have the following 
number of phalanges for the digits from the second to the fifth conclusive, 
4-4-4-S. 
The anterior border of the pubes is concave, leaving a lateral convex 
border in front of the acetabulum. The pubes of opposite sides meet at an 
entrant right-angle. The external posterior angles of the ischia are rounded 
and prominent, since the posterior borders are oblique and meet each 
other at a deep entrant right-angle. 
Measurements of slab specimen. 
M. 
Length of a series of five consecutive dorsal vertebrae. 
.041 
Length of second of this series 
.007 
Elevation " " " 
.015 
Length of neural spine of do 
.0005 
Elevation " " " to neural canal 
.009 
( vertical 
Diameters of a vertebra without spine < 
.010 
I transverse 
.0105 
f vertical 
Diameters of separate centrum of do. < 
.003 
I transverse 
.004 
Length of chord of a rib ; apex restored 
.040 
Width of shaft of do. at middle 
.0035 1885.] 
13 
[Cope. 
Measurements of slab specimen. 
M. 
Elevation of a lumbar vertebra with spine 
.. .017 
" " spine of do 
.. .006 
Width of centrum at base of ribs 
.. 0075 
Length of rib 
.. .0152 
Diameters of a pubis J anteroposterior 
.. .014 
( transverse 
.. .0176 
Diameters of an ischium $ anter°Posteri°r 
.. .019 
< transverse 
.. .012 
Length of femur 
.. .038 
Diameters of shaft of femur at middle 
.. .004 
Length of tibia 
.. .025 
r proximal 
.. .005 
Diameters of tibia at middle 
.. .003 
t distal 
.. .0055 
Width of sole, including tibia and fibula 
.. .0125 
" " intermediocalcaneum 
.. .009 
Length of " at middle 
.. .006 
" " tibiale 
.. .007 
Width " " 
.. .005 
Length " tarsale I 
.. .0038 
" " " II 
.. .0040 
" " metatarsale I 
.. .0175 
" " " II 
.. .016 
" " " III 
.. .014 
" " " IV 
.. .012 
" " " V 
.. .009 
" " second digit, minus end of unguis 
.. .036 
" " " " first phalange 
.. .0085 
" " " " second phalange 
.. .004 
" " " " third phalange 
.. .0035 
" " ten proximal caudal vertebrae 
.. .075 
Transverse extent of diapophyses of second of do... 
.. .040 
Length of six distal caudal centra 
.. .047 
Depth of one centrum of do 
.. .0037 
Elevation of neural arch with spine 
.. .0115 
" " " " without spine 
.. .003 
A number of vertebrae are preserved on fragments of a softer rock of 
darker color than the specimen above described. It is possible that they 
belong to another species of the genus, as I observe some peculiarity in the 
caudal vertebra. The base of the neural spine is so robust as to cover the 
anterior section of the centrum, and does not therefore present the appear- 
ance of coming off from the posterior section alone, as is the case in the 
typical specimen. I have, however, not seen the arches of the anterior 
caudals of the latter. 
A marked character of the dorsal vertebrae, is the appearance of hyperos- Cope.] 
14 
[April 17, 
tosis presented by the neural arch and its parts, and in some degree by 
the centrum. The outline of the latter viewed from below is barrel-shaped, 
and the space between the inferior surface of the centrum and the extrem- 
ity of the diapophysis is filled with osseous tissue, so as to be bounded by 
a nearly straight line connecting the points in question. The diapophyses, 
where not continued into ribs, are somewhat flattened cones. The neura- 
pophyses are greatly thickened, having more than twice the transverse 
diameter of the small neural canal. The zygapophyses are mere ledges ; 
the prezygapophyses of the neurapophyses ; the postzygapophyses of the 
neural roof. The latter is expanded and thickened, an anterior thickening 
on each side, constituting the zygosphen. The neural spine is moderately 
a little elevated, and is compressed ; its base extending the length of the 
neural arch. The prezygapophyses are opposite the middle of the neural 
canal. The postzygapophyses are connected by a thin prolongation of 
the roof of the neural canal, which is not interrupted in any of the ver- 
tebrae at my disposal. 
The anterior caudal vertebra is flattened below, and has a median shal- 
low fossa. A large basis for a rib marks the upper part of the anterior 
half of the centrum, and below it is a low tuberosity. Between the latter, 
on the ? intercentral half, is a short accuminate tubercle directed forwards. 
The posterior articular face is supplemented by two facets below, as if for 
separate chevron bones. 
A more posterior caudal vertebra has a longer, and compressed centrum, 
without transverse processes or tubercles. The inferior surface has a ridge 
on each side, which are interrupted by the constriction already mentioned. 
Those of the posterior half are continued into coossified chevron bones. 
The postzygapophyses are more elevated on the dorsal vertebrae, and the 
neural spine is robust and is directed strongly backwards. 
The surfaces of the dorsal vertebrae are smooth ; that of the anterior 
caudal is minutely punctate, and at some points wrinkled. 
Measurements of Vertebrae. 
No. 1 (with rib). 
M. 
Total elevation 
0125 
Elevation of centrum anteriorly 
0040 
" to prezygapophysis 
0055 
" " zygosphen 
0070 
" " highest base of neural spine 
0090 
Width of centrum anteriorly 
0035 
" " prezygapophyses 
0090 
No. 2 (without ribs). 
Total elevation 
0160 
Length of centrum 
0080 
Elevation to neural canal posteriorly 
0038 
" " postzygapophyses 
0058 
" " neural spine 
0100 1885.] 
15 
[Cope. 
Measurements of Vertebrae. 
No. 2 (without ribs). 
M. 
Width of centrum posteriorly 
0040 
" at diapophyses inclusive 
0180 
" " postzygapophyses 
0100 
No. 3 (without rib). 
f longitudinal 
0072 
Diameters centrum j vertical anteriorly 
0035 
[transverse anteriorly 
0035 
Width at diapophyses inclusive 
0170 
' ' of postzygapophyses 
0094 
No. 3 ; posterior caudal. 
Length centrum 
0070 
Width at middle 
0035 
Diameters centrum in front / 
0040 
t transverse 
0040 
Elevation to postzygapophysis 
0065 
Width of neural spine at postzygapophysis 
0065 
This species was probably of elongate form. Prof. Derby informs me 
that he has seen considerable series of consecutive vertebrae. The speci- 
mens sent me indicate that the size of the body is about equal to that of 
the fully grown Tejus lizards now inhabiting Brazil. 
The specimens are from four localities in the province of Sao Paolo ; 
viz : Rio Claro, Limeria, Itapetininga and Tiete. These localities are a 
considerable distance apart, and represent the considerable extent of the 
formation from which the bones have been procured. As a Lepidoden- 
dron and a Schizodus have been obtained from the same beds, they are 
probably of Carboniferous or Permian age. 
The specimen preserved on the slab belongs to the private collection of 
Madam Ribeira de Andrada, to whom science owes a debt of thanks for the 
opportunity of determining its characters which she has given by lend- 
ing it to the Museo National. 
REPTILIA. 
Hyposaurus derbianus, sp. nov. 
The genus Hyposaurus has been hitherto represented by but one well 
known species, the H. rodgersi Owen, of the green sand of Cretaceous 
No. 5, of New Jersey. Specimens in my possession demonstrate that 
the genus Hyposaurus belongs to the Teleosauridae, and that its nearest 
ally is the Steneosaurus of St. Hilaire. It differs from Metriorhynchus 
Meyer, in the presence of distinct lachrymal bones, and in the relatively 
small size of the prefontals. From Teleosaurus proper it differs in the 
robust size and vertical directions of the teeth. The orbits are vertical, and 
the sagittal region is a keel. In the IL rodgersi the frontal bone is nar- 
rower than in any of the species of Teleosauridae figured or described by Cope.] 
16 
[April 17, 
Deslongchamps. The palatal foramina extend forwards to the line of the 
posterior maxillary teeth, and the anterior border is rounded, not acute as in 
most of the speqies of the family.* The specimens are not sufficiently com- 
plete to enable me to state postively the generic distinction from Steneo- 
saurus. In Teleosaurus the vertebral hypapophyses only appear on the 
first and second dorsal vertebrae, while, as Owen observes,f they are pres- 
ent on many of the dorsals in Hyposaurus. This peculiarity, and the great 
contraction of the frontal bone, render it very probable that the genus is 
distinct from Steneosaurus, but the diagnostic character yet remains to be 
discovered. 
The Brazilian Hyposaurus is represented in the collection of the Museo 
National, by the left malar and quadratojugal bones ; by a nearly entire 
lower jaw ; by several vertebrae from the middle and posterior parts of the 
column ; by a humerus ; a coracoid bone ; and by several dermal bones, 
all belonging to one individual. There are several isolated teeth of the 
same animal, and others which probably belong to the same species, as they 
closely resemble those which are contained in the lower jaw mentioned. 
The mandibular rami early unite into a long slender symphyseal por- 
tion. There are twenty alveoli in each, and only five of these are in the 
portion of the ramus which is posterior to the symphysis. The free por- 
tion of the ramus is compressed ; both of them are broken off from the coro- 
noid region, inclusive, posteriorly. The symphyseal region has a semi- 
circular section, which is a little angulate ; that is, is flattened laterally and 
below. The splenial bones appear on the inferior surface as far anteriorly 
as opposite to the fourth tooth from the beginning of the symphysis. The 
teeth have a lenticular section in the posterior part of the series, and the 
section becomes rounder, that of the first pair being entirely round. All 
display a more or less distinct cutting edge in front, and one opposite to 
it on the posterior face of the crown. The enamel surface is marked with 
rather close, straight, longitudinal ridges on the internal side of the crown. 
The middle of the external side is quite smooth. The crowns are acute at 
the apex and slightly recurved. Those of the more posterior teeth are 
shorter, becoming little higher than wide anteroposteriorly. 
Measurements of Ramus and Teeth. 
M. 
Length of symphysis 
... .336 
Width at posterior end of symphysis 
... .075 
Depth " " " 
... .037 
(transverse 
... .037 
Diameters symphysis at middle j vertjcai 
... .030 
(transverse 
... .042 
Diameters at second pair of teeth vertjra] 
... .021 
* These comparisons are rendered possible by the admirable monograph of 
these reptiles by M. Endes Deslongchamps in Vol. x, Bulletin Soc. LinnSenne 
de Normandie, 1866. 
t Quarterly Journal, Geol. Society, London, 1849, p. 383. Proc- Am. Philos . Soo Vol XXIII N°121. 
T. Sinclair & Son,Phila 
STEREOSTERNUM TUMIDUM COPE 1885.] 
17 
(Cope. 
Measurements of Ramus and Teeth. 
M. 
.Diameters of base of seventh 
f anteroposterior 
... .011 
tooth from end 
(transverse 
... .0085 
Length of crown of a loose tooth (same animal)..., 
... .0225 
Diameters middle crown of a I 
[ anteroposterior 
... .080 
loose tooth 1 
[ transverse 
... .050 
From these measurements it is evident that the anterior extremity of 
the lower jaw is not expanded. The teeth of the anterior pair are directed 
rather more anteriorly than exteriorly. Ai the symphysis a horizontal 
figure oo shaped fossa marks the junction of the splenial and dentary 
bones, and the inferior side of the former is grooved on the middle line for 
15 mm. in front of the symphysis. 
The malar bone is elongate and strongly compressed, showing the great 
obliquity of the os quadratum. It sends upwards a postorbital branch, 
which is external as in other Teleosauridae, and not internal as in Croco- 
dilidae. The surface is marked with shallow longitudinal fossae like those 
of the lower jaw. Length from postorbital branch to quadratojugal, 
upper edge, .120 ; lower edge, .165; depth at middle, .024; thickness, .010. 
In the most anterior dorsal preserved, the diapophyses are entirely on 
the neurapophyses. The articular faces of the centrum are shallowly 
concave, and the sides between them are flattened but not very concave. 
The hypapophysis has a long compressed base, which ceases 10 mm. an- 
terior to the posterior extremity of the centrum. The neurapophysial 
suture is very little decurved in the middle. The diapophysis displays a 
capitular articular process, with small facet, -which originates just above 
the suture with the centrum. The tubercular facet is at the extremity of 
a robust process, whose posterior edge originates near the posterior edge 
of the neurapophysis, and is wide at the base, enclosing a fossa. A sec- 
tion of the base of the diapophysis is subquadrate, with the superior or 
anterior angle rounded, and the inferior anterior produced downwards 
and forwards for the base of the capitular portion, like the tail of a comma. 
The general form of the tubercular part of the diapophysis is subconical. 
A convexity proceeds from its anterior base, its continuation forming the 
lateral convex face of the prezygapophysis. The latter is small, and its 
superior or articular face is on a level with the roof of the neural arch, 
thus having a rather low position. The arch rises steeply to the neural 
spine. The latter is moderately elevated, and is much compressed and 
thin, having a narrow anterior edge, and a posterior edge not quite so nar- 
row. The summit is not thickened, as is the case in Teleosaurus cadomen- 
sis, according to Deslongchamps, and is wide anteroposteriorly. Both 
anterior and posterior edges of the spine are a little thickened, and are 
medially grooved for a short distance above the neural canal. The neural 
canal is ample, and is a little wider than high at its anterior extremity. 
In a dorsal vertebra near that last described in the series, the capitular 
part of the diapophysis is carried nearer to the tubercular portion, and 
the base of the two combined is less robust, the section having an elon- Cope.] 
18 
[April 17, 
gate triangular outline, the base anterior. The capitular portion is still 
decurved so as to present below the tubercular, and is narrow. The pre- 
zygapophyses are small. The postzygapophyses are close together, and 
are separated by a deep groove. The articular faces are shallowly and 
equally concave, and are vertical to the long axis of the centrum. The 
hypapophysis occupies the anterior three-quarters of the middle line of 
the centrum. 
In a dorsal posterior to the one last described, the diapophysis is still 
more depressed at the base, which is oblique to the long axis by about 25°. 
The postzygapophyses are concave on their articular faces, the concavity 
extending as a shallow groove to the posterior base of the diapophysis. 
They are separated by a vertical groove of the base of the neural spine 
still deeper than in the vertebra last described. The centrum is less com- 
pressed than in those more anterior, and there is not even a keel to repre- 
sent the hypapophysis. The neural spine is less elevated than in the 
other dorsals described, and its summit is rounded oil in front, and is 
compressed. The dorsal which precedes this one in the series is repre- 
sented by a centrum only. This has an inferior median angle represent- 
ing the hypapophysis. 
An anterior caudal has a diapophysis of medium length, depressed, and 
when viewed from aHove, displaying an outline of an elongate cone with 
truncate apex. The zygapophyses are fairly well developed, and the 
neural spine is large, especially anteroposteriorly. The chevron facets 
are large and close together. The median line of the inferior face of the 
centrum is concave. The articular faces of the centrum are slightly con- 
cave, and the anterior is deeper than wide. 
Measurements of Vertebrae. 
M. 
Total elevation of No. 1 
.114 
f vpi'tic^l 
Diameters of centrum posteriorly < 
.035 
I transverse 
.039 
f VPT'tlP'11 
Diameters neural canal posteriorly 1 
.027 
( transverse 
.023 
From centrum to face of postzygapopliysis 
.031 
Length centrum at base neural canal 
.048 
Anteroposterior width neural spine above postzyga- 
pophyses 
.040 
/ anteroposterior 
.049 
Diameters centrum dorsal No. 2 < vertical (front) 
.037 
v transverse (front).... 
.038 
Length diapophysis from base, below 
.038 
Width at postzygapophyses, inclusive 
.030 
r anteroposterior 
.048 
Diameters of dorsal No. 4 7 vertical (behind) 
.040 
1 transverse (behind) 
.040 
Diameters neural canal posteriorly •* 
I transverse 
.015 
.018 1885.] 
19 
[Cope. 
Measurements of Vertebrae. 
M. 
Elevation neural spine from canal 
. .054 
Length diapophysis below 
. .038 
Width at postzygapophyses 
. .030 
f anteroposterior 
. .045 
Diameters of a caudal vertebra vertical (front) 
, .037 
1 transverse (front)... 
. .035 
Length diapophysis below 
. .033 
Width at postzygaphyses ..... 
. .023 
Length of base neural spine above postzygapophyses.. 
. .030 
The coracoid bone has an expanded proximal extremity, which contracts 
on the external side abruptly, into a slender shaft which continues to the 
distal end, which is but little expanded. The coracoid foramen is well 
within the external border, and is small. The distal end is flattened below, 
and has a convex margin. The shaft has an oval section. This element 
is much more slender than in the Alligator mississippiensis, and even more 
so than in the Teleosaurus cadomensis, according to Deslongchamps. 
Measurements of Coracoid. 
M. 
Total length 
165 
Long diameter of proximal end 
065 
Thickness of proximal end at glenoid facet 
023 
Diameter of shaft $ 
010 
ctransverse 
015 
Width of distal end 
0295 
The humerus is rather elongate, and is but little curved. The head is 
directed a little inwards and forwards, and the condyles (which are lost) 
a little backwards. The section of the shaft is nearly round from below 
the deltoid crest to near the condyles. The head is flattened and its artic- 
ular extremity is convex and narrow. Near the internal border of the 
anterior side is a shallow fossa. The deltoid crest is elongate, and lies on 
the external edge of the posterior face. Its elevation increases distal, 
i. e., to a point nearly two-fifths the length from the head. 
Measurements of Humerus. 
M. 
Length of part preserved 
. .220 
Diameters of head S anteroposterior 
. .019 
(transverse 
. .051 
Diameters shaft 3 c.m. below crest {anteroposterior.. 
. .025 
t transverse 
. .027 
General Remarks.-The characters of this species are much like those 
of H. rodgersi, so far as they are known. I observe the following differ- 
ences on comparison with several individuals of that species. The artic- 
ular faces of the vertebral centra, are less concave than in the Northern 
species. The symphyseal part of the mandible is a part of a cylinder in 
the H. rodgersi, while it is flattened below and at the sides in the Brazilian Cope.] 
20 
LApril 17, 
species. The bones of the limbs are relatively less robust in the H. derbi- 
anus* The differences, especially in the humerus, are well marked. 
I name this species in honor of Prof. Orville A. Derby, in charge of the 
department of Geology in the Museo National of Brazil. 
Toxodon expansidens sp. nov. 
The incisors of the first and second places of the upper jaw, represent 
this species. Comparison with the corresponding teeth of the known 
species, reveals well-marked distinctive characters. 
The incisor of the median pair has greater transverse, and less antero- 
posterior, diameter than in any of the known species. Its diameters are 
uniform. The cutting edge is five and a half times as long transversely 
as it is anteroposteriorly. The anterior enameled face has two planes, a 
wide exterior one which is concave, and a narrower inner one which re- 
treats inwards, and is plane to the convex inner (median) edge. The 
enamel extends round the narrow external edge, but disappears at the 
middle of the inner beveled faces. The angle between the two faces 
forms a rib, parallel with the borders of the tooth. No enamel on the 
internal face. Enamel surface with rather coarse obsolete longitudinal 
grooves. 
The external incisor is a robust, prismatic, rodent-like tooth, strongly 
curved. Its section is triangular, the posterior (enameled) face being 
convex. The external face is flat, and its plane forms less than a right 
angle with the anterior face, from which it is separated by a convex inter- 
mediate surface. The prominence of the latter causes the anterior face to 
be slightly concave. The angle is the most prominent portion of the cut- 
ting edge. The enamel ceases a little short of the narrow internal edge 
of the tooth ; its surface is marked with obsolete longitudinal grooves. 
MAMMALIA. 
Measurements of Teeth. 
M. 
First incisor. 
'vertical 
transverse 
... .080 
Diameters of crown 
... .059 
anteroposterior / at ' ' 
... .011 
L 1 at angle 
... .015 
Width of internal level 
... .022 
Second incisor, 
vertical on curve 
... .150 
Diameters of tooth 
transverse 
... 033 
anteroposterior (externally) . .. 
... .021 
This species is as large as the Toxodon platensis Owen. As compared 
with that animal, the median incisors have much greater transverse ex- 
tent, and relatively smaller anteroposterior diameter. These teeth are 
♦For figures of humerus and femur of H. rodgersi, see Transae. Amer. Philos. 
Soc. xiv, Pl. Iv, figs. 10-11, 1879. 1885.] 
21 
fCope. 
still more different from those of 7. burmeisteri. The external incisors 
are, on the other hand, more like those of the latter species in their trian- 
gular form, though their inner angle is not produced as in that species. 
Explanation of Plate. 
The figures represent the Stereosterum tumidum in various pieces ; all of 
the natural size excepting fig. 1, which is three-fourths natural size. 
Fig. 1. The typical specimen on a slab of calcareous shale of the car- 
boniferous formation ; the anterior part of the skeleton wanting ; viewed 
from below, nc, notochordal canal exposed by the splitting of the verte- 
bral centrum. 
Fig. 2. Vertebrae in a piece of weathered rock of darker color than the 
slab. 
Fig. 3. A lumbar vertebra from the piece of matrix represented in fig. 
2, anterior view ; a, inferior view. 
Fig. 4. A caudal vertebra from the same piece of stone, left side ; a, in- 
ferior side. 
Fig. 5. A dorsal vertebra with proximal portions of ribs embracing the 
centrum ; from a different piece of matrix. 
Fig. 6. A vertebra of uncertain position, with descending processes, an- 
terior view ; a, the same lateral view. 
Fig. 7. Humerus, the proximal portion represented by a mould ; from a 
separate piece. 
Fig. 8. Coracoid bone from a separate piece. 
All the specimens are preserved in Museo Nacional of Rio Janeiro, ex- 
cepting that represented in fig. 1, which is in the collection of Madame 
Ribeira de Andrada. 
Printed October 12, 1885. 405 
1885.] 
[Cope. 
Second Continuation of Researches among the Batrachia of the Coal 
Measures of Ohio. By E. D. Cope.* 
{Read before the American Philosophical Society, June 19, 1885.} 
Cercariomorphus parvisquamis, gen. et sp. nov. 
Char. gen. Represented by a fusiform body which terminates in a long 
slender cylindric tail, and which is covered with small subquadrate scales 
quincuncially arranged. No fins or limbs are preserved, and the form of 
the head cannot be made out. There are some scattered bodies in the body 
portion, which look like deeply concave vertebrae with the zygapophyses, 
of batrachians. There are some linear impressions at one point, which re- 
semble the bristle-like rods of many Stegocephali. These are so few as to 
be of little importance. The scales are like those ot fishes. There are 
traces of segmentation in the axis of the long tail. 
The position of this curious form is quite uncertain. It is quite different 
from anything observed hitherto in the American coal measures. 
* The first continuation of these researches, subsequent to the publication of 
the Report of the Geological Survey of Ohio, appeared in these Proceedings for 
February, 1877. Cope.] 
406 
[■Tune 19) 
Ghar. specific. Scales in their present condition with entirely smooth 
surface. At a distance of .20 m.m. from the base of the tail they are in 
twenty longitudinal series. At that point the transverse diameter is 140 
m.m. The outline contracts rather abruptly to the tail, of which 750 
m.m. are preserved. The surface of the tail is obscured by a thin layer of 
carbonaceous matter not sufficiently thick to obscure scales; but no scales 
appear. The body, or such part of it as is preserved, measures 900 m.m. 
in length. The cephalad half of the body is depressed and somewhat dis- 
ordered ; the caudad half is well preserved. Where the undisturbed por- 
tion commences the width is 150 m.m. The diameter of the cup of a sup- 
posed vertebral centrum is 1.6. m.m. 
A tooth of Pleuracanthus compressus Newb. lies close to one edge of the 
body. 
Discovered by Samuel Huston. 
Anisodexis enchodus, sp. nov. 
The generic characters are apparent in the very unequally sized teeth 
with round section. The portion upon which the species is based is a 
part of the right ramus of the mandible, which is in the specimen viewed 
from the inner side. The jaw is obliqely and smoothly truncate from be- 
low, for the symphysis and the surface of the bone is smooth. There is 
a very large tooth near the extremity of the dentary bone. Behind it is 
an interval equal to three times the diameter of its base, which is followed 
by a tooth of about one-third the length of the first tooth. Posterior to 
this one are two teeth of the same size as the second, all being separated 
from each other by about a tooth's diameter. These are followed by three 
sub-equal teeth of about two-thirds the length of the first tooth, and sep- 
arated by about their own diameter from each other. They are all per- 
fectly straight, very acute, and without trace of a cutting edge. The 
inflection-grooves extend to or a little beyond the middle of the length. 
Measurements. 
M. 
Length of jaw including seven teeth 
031 
Depth of ramus at second tooth 
Oil 
Length of first tooth 
0105 
" " third tooth 
004 
" " sixth " 
0075 
The type of the genus is Anisodexis imbricarius from the Permian beds. 
The present species is very much smaller, and the apices of the teeth do 
not display the opposite cutting edges seen in the A. imbricarius. 
Discovered by Samuel Huston. 
Ceraterpeton divaricatum, sp. nov. 
This species is represented by a skull whose superior surface is visible 
on a block of coal shale. In size it exceeds that of the other two Ameri- 
can species, the G. punctolineatum, and G. tenuicorne. Its epiotic horns 
are of the straight acute type of the latter species, and its sculpture is also 
different from that of the G. punctolineatum. 1885.] 
407 
[Cope. 
The cranium is distinguished by its elongation, the outline of the muz- 
zle being a regular oval. The orbits are situated at about the middle of 
the length of the skull. A transverse line dividing the skull equally 
marks the anterior two-fifths of the orbit. This proportion distinguishes 
the species from the 0. tenuicorne,* where the orbit is entirely within the 
anterior half of the skull. The horns are rather short, and are straight 
and acut» The lateral border of the skull contracts somewhat to their 
base on each side. They diverge at an angle of about 45° from the 
median axis of the skull. The orbits are oval, the transverse being .75 
the longitudinal diameter, which is in turn about equal to the interorbi- 
tal width. The nasal bones are very large, forming the upper surface of 
the muzzle, and are not distinguishable from the prefrontals, so as appar- 
ently to enter by their postero-external angle, into the orbit. Their sculp- 
ture consist of sparse thin radiating ridges, which originate near the cen- 
ter of each bone. The frontal bones are rather wide. Their sculpture 
consists of radiating ridges, which are not very close together, and which 
originate at a point near the supraorbital border, a little in front of a line 
which connects the middles of the orbits. On the postfrontal bone the 
radii run posteriorly. On the parietal they inosculate so as to form coarse 
fossae. The teeth are rather small, and their sections at the middle and 
upwards, are round. 
Measurements. 
M. 
Total length skull on middle line 
053 
Length to line of anterior orbital border 
023 
Width at middle of orbit (with lower jaw) 
044 
Width at base of horns 
040 
Interorbital width 
017 
Length of horn from base 
0075 
Should it turn out that the Tuditanus obtusus of the same locality and 
horizon is founded on a Ceraterpeton which has lost its horns, it may still 
be distinguished from the present animal by the more anterior position of 
its orbits. These are so placed that their posterior border is crossed by the 
transverse median line. 
Discovered by Samuel Huston. 
Claspers of Batrachia. Dr. Anton Fritsch has obtained in the Gas- 
kohle of Bohemia, in connection with specimens of two species of Ophi- 
derpeton, bodies which he believes to belong to the external genitalia. 
They consist of a curved rod terminating in a secund expansion, whose 
projecting edge is divided into fine teeth like a comb. Mr. Samuel Huston 
has obtained at Linton, in Eastern Ohio, at the locality which has furnished 
the species here described, a similar body, It differs from those described 
by Fritsch, in the greater curvature of the shaft in the direction to which 
the teeth present. Its axis is nearly at night angles to that of the body of 
* Report of the Geological Survey of Ohio, Palaeontology il, p. 372, Pl. xlii, 
ilg. 2. Cope.] 
408 
[June 19, 
the bone. The latter is oval and convex, and its thin edge is divided by 
fine grooves more closely placed than in the species described by Fritsch, 
which terminate in fissures separating delicate teeth. See Fauna der Gas- 
kohle und der Kalksteine der Permformation Boehmens, p. 122, Pl. 20. 
Similar bodies were found by myself in the fresh-water beds of the 
Laramie formation of Montana, and described under the name of Arotus 
hierogylphicus. (Bulletin U. S. Geol. Survey Terrs., F. V. Hayden, iii, 
1877, p. 574.) The shaft of this body is not curved, and the body is flat- 
tened. As specimens of the batrachian genus Scapherpeton are abundant 
in this formation and locality, it is not unlikely that these comb-like bones 
are their claspers. 
Printed, July 30, 1885.