No. 384.] CHELONIAN CARAPACE AND PLASTROA. 937 
third neural. It has, as it should have, an epidermal sutural im- 
pression across it. On the anterior end of this bone there 
was present a prominent tubercle, which in form and position 
resembled the distinct ossicles further behind; but there was no 
suture at its base. The conclusion drawn was that it was once 
distinct but had become coéssified with the underlying neural. 
Fig. 3 represents, of actual size, another bone that was found 
in the same collection of Toxochelys materials. Its upper por- 
tion resembled closely one of the ossicles described above ; but 
below this there was a thinner portion, 
which had evidently been buried in the 
flesh. It was regarded as equivalent to 
one of the rows of bones which are to be 
found along the upper edge of the tail of 
Chelydra. 
It seems to me evident, therefore, that 
that row of tubercular ossicles along the back of Toxochelys 
was simply a continuation forward of the row that, like those 
of Chelydra, must have been present on the tail. These last 
must be reckoned as of purely dermal origin ; so, too, must 
those on the carapace. Moreover, the neuralia on which they 
were reposing must belong to a deeper stratum of bone. 
A median keel along the back is a not uncommon feature of 
turtles. Nearly all species possess it at an early stage of life; 
although it may become obsolete as growth proceeds. In 
some tortoises this keel is elevated at intervals into prominent 
tubercles. These occur near the hinder border of each of the 
median bony scutes. And that ts just where those tubercle-like 
ossicles of Toxochelys were found. I conclude, therefore, that 
these dorsal tubercles of our existing turtles have originated 
from a median dorsal row of dermal bones, distinct in the 
earlier forms, but now ossifying continuously with the under- 
- lying neurals. In the young of the diamond-back terrapin 
(Malaclemys terrapin) these dorsal tubercles are greatly devel- 
oped and consist of four or five globular masses like small peas. 
It would be interesting to have their development studied, in 
order to ascertain if possibly they may yet possess distinct 
centres of ossification. 
FiG. 3: Xi 
