366 
the tidal currents in bringing up the cold bottom waters 
of the ocean is perhaps a cause sufficient to produce 
most of the coldness of the water in this region. 
ON DINOCERAS MIRABILIS (MARSH) 
AX: SHORT time ago we gave a note respecting one of 
the recently-discovered gigantic fossil mammals 
from the Eocene of Wyoming in the region of the Rocky 
Mountains; the accompanying woodcut, copied from a. 
paper by Prof. Marsh, on this extraordinary extinct 
animal, named by him Dinoceras mirabilis, will further 
assist in making its peculiarities easily understood. 
The animal must have been nearly as large as the 
elephant, to which its limb-bones were very similar. The 
only teeth it possessed in the upper jaw, were a pair of 
well-developed canine tusks, and six pairs of small molars, 
whose crowns were formed of two transverse ridges, 
separated externally, but meeting at their inner extremi- 
Dixoceras MrraBitis 
ties. The frontal region of the skull was concave, on 
account of the lateral projection upwards of a bony ridge 
or crest on each side, which posteriorly developed into a 
large osseous process that may have been a horn core 
but perhaps was only covered with thick skin, and acted 
like the fibrous pads on the cheeks of the wart-hog, to 
shield the thinner skull from direct blows. Behind these 
the crest extended back beyond the level of the occipital 
condyles. The maxillaries each bore a conical process, 
which in a profile view is evidently seen to be directly 
above the root of the canine tusk, and supported it ; it 
probably carried ahorn. At the anterior extremities of 
the nasals were also two smaller horn cores. The horns 
must have been of a character very different from those 
in the rhinoceros, in which animal, however long they 
may be, they are only supported on a roughened surface 
of bone; if they resembled those of the cavicora ungu- 
lata, from analogy we must suppose that they were small, 
for in those animals there is a close relation between the 
size of the core and thut of the horn which it carried. 
There were no postorbital processes to the frontal 
bones. The zygoma was completedin front by the malar, 
the lachrymal was large, and formed the anterior border 
NATURE 
> a Ia es. 
Aa eae eet 
[ Mar. 13, 1873 
of the orbit; its foramen was exserted. The infraorbital 
foramen must have been behind the zygomatic ridge, as 
it does not appear in any of the drawings. The premax- 
illaries did not carry teeth; they sent forward two 
branches, which partially enclosed the sides of the ex- 
ternal nares ; the upper branch «joined the nasal, and 
the lower, as in the Ruminants, continued free, and pro- 
bably carried a pad. Prof. Marsh gives no illustration 
of the mandible, and only remarks of it that “the lower 
jaw was slender and the tusks small.” The limbs were 
short, the fore limbs shorter than those behind. The 
radius did not cross the ulna so obliquely as in the ele- 
phant. In the head of the femur there was not any pit 
for the insertion of the round ligament. The great tro- 
chanter was flattened and recurved ; the third trochanter 
was absent. The tail was short and slender. The ribs 
had rudimentary uncinate processes. 
Prof. Marsh feels justified in placing Dinoceras in an 
order Dinocerata, distinct from the Proboscidia, on 
account of the absence of upper incisors ; the presence of 
canines and horns; the absence of large cranial air 
cavities ; the malar forming the anterior portion of the 
zygoma ; the absence of a proboscis, which could not 
have been necessary in an animal that could easily touch 
the ground with its nose, and other less important 
differences. 
This Dinoceras of Marsh is the Zodasileus of Cope and 
the Uintatherium of Leidy. The shortness of the pub- 
lished descriptions prevents.us saying more about it at 
present. a 
THE TROGLODYTES OF THE VEZERE* 
Ill. 
Our Troglodytes of the latest epoch had, in fishing, an- 
other resource unknown to their predecessors. Their 
different stations contain a-large number of fish bones ; 
but it is remarkable that all these fish were salmon. Now 
the salmon in these days neither frequent the Véztre 
nor the part of Dordogne where that river joins the sea. 
At some leagues below the confluence, not far from La- 
linde, in the centre of Dordogne, there is a bank of rocks, 
which, at high water, forms a rapid, and at low water a 
regular fall, called, The Leap of the Gratusse. The salmon 
do not pass this boundary, and, as it did not stop them 
at the epoch of the Troglodytes, we must conclude that, 
since that time, the level of the Dordogne has fallen, either 
by hollowing out its bed so as to lay bare the bank of rocks, 
or by losing part of its volume of water. We are led to 
believe that the fishermen of that time did not use nets, 
for with a net could be caught fish of all sizes. We thus 
understand why they could only catch large fish, and why 
they chose, among these, the kind they preferred. Had 
they any fishing boats? We have as yet found no 
proof of such. And besides, the Véztre is sufficiently 
enclosed for the large fish to swim along the banks within 
reach of the harpoons. . 
The harpoon of our Troglodytes was a small dart of deer- 
horn, very similar to the large barbed arrows, except that 
it was only barbed on one side. A little notch at the 
base enabled the fisherman to secure the cord which he 
held in his hand (sez above, Fig. 10). The barbs are in- 
tended to secure the fish which it has struck. Why are 
these barbs all placed on the same side? Is it to dimi- 
nish the width of the dart and make it more penetrating ? 
This I cannot venture to affirm. 
* Continued from p. 325 
+ One of my colleagues of the French Association, M. Lecoq de Bois- 
beaudrau, who did me the honour of being present at this lecture, communi- 
cated, the following day, to the Section of Anthropology, a very interesting 
note on the mode of action of theunilateral barbs of the harpoon. While the 
harpoon is traversing the air, these barbs cannot make it deviate sensibly ; 
but directly it enters the water, the unequal resistance it meets there must 
necessarily change its direction. Itseems, then, that the fisherman who 
aims straight ought the most frequently to miss his aim. But M. Lecoq 
de Boisbeaudrau reminds us of the well-kaown experiment of the straight 
