ON THE PHYLOGENY OF THE CARAPACE, ETC. 797 
variable and very slender at the anterior end, often not quite reaching 
the vomer, and, to judge from its form and relations, seems to be 
undergoing a process of reduction.1! From Hay’s point of view, it 
remains quite obscure why the palatine should send such a slender 
process forwards in the Leatherback; the development of this process 
is best explained by Dollo’s hypothesis,’? that Dermochelys descends 
from a form with a secondary palate of the chelonoid type (for instance 
like Toxochelys). 
5. The nasal cavities are built upon the same specialised type,’ 
being divided by partition-walls into several independent chambers and 
showing several other minor details common to both forms. This is 
an important indication of relationship of Dermochelys and the Chelonii- 
de, as the nasal cavity of the Leatherback is exceedingly short and its 
proportions are therefore quite different from those in the Cheloniide ; 
these differences in proportions are best understood on the assumption 
that the nasal cavity was secondarily shortened in Dermochelys, and 
this points to a secondary forward movement of the internal nasal 
opening, as assumed by Dollo (here referred to under 4). 
6. There is an intertrabecula in the skull of the ripe embryo.'* 
The intertrabecula is a short median bar of cartilage, lying between 
the trabeculz, reaching from the dorsum selle behind to the inter- 
orbital septum in front. It is not known in other Testudinata, and 
can only be interpreted as a special character of marine turtles, not 
as a primitive character inherited from the Prochelonia. That its 
presence in the Cheloniide and Dermochelys is the result of con- 
vergence is very improbable, as its later development in both forms is 
absolutely different, and there is no evidence for connecting it with 
marine life. Consequently the intertrabecula strongly favours near 
relationship of the Leatherback with other marine turtles. Nick 
gives 1* numerous examples of other less important points of common 
structure in the skull of the Leatherback and Chelonia. 
7. The cesophagus is covered with long, pointed, horncapped 
papille, directed towards the stomach.’” Such papille are not known— 
at least not in the same completeness—in other Testudinata.1® 
Though one or other of these special resemblances may, in the 
light of more complete knowledge, lose its value as a proof of the 
relationship of Dermochelys with the Cheloniide, taken altogether 
they leave, to my mind, no other possibility than to accept a rather 
close relationship of these forms. Even the important differences in 
the shell, in the skull, &c. (compare p. 794), cannot make this con- 
clusion invalid. But these differences are important, as they prove that 
Dermochelys is not a direct descendant of typical Cheloniide, as Baur 
1 Nick, 1912, p. 62-65. 
P Dollo, 1903, p. 30; compare v. Bemmelen, 1896 ; Nick, 1912, p. 63. 
13 Nick, 1912, p. 180-145 and p. 191-195. 
14 Nick, 1912, p. 106-114. 
15 Compare also Nick, 1912, p. 197. 
16 p. 169. 
17 Hoffmann, 1890, p. 243 ; Burne, 1905, p. 315. ; 
18 According to Oppel, Lehrbuch vergl. mikrosk. Anat. d. Wirbeltiere, vol. 2, 1897, 
p. 82, Nuhn (Lehrbuch vergl. Anat., Heidelberg, 1878) found dispersed pointed papille 
in some fresh-water tortoises. 
