948 ~ THE AMERICAN NATURALIST. 
there is not much tendency toward codssification. This view 
is borne out by the suppression of the entoplastron in Dermo- 
chelys and the Cinosternidz. For the same reason, and because 
the Ichthyosauria and the Plesiosauria seem to have possessed 
only one set of gastralia for each pair of ribs, I am inclined to 
believe that such was the condition in turtles. It appears 
that from the Stegocephali upward there has been a tendency 
toward a reduction in the number of sets of gastralia belonging 
to each body segment. If later on in the history of turtles 
certain plastral elements were excluded from the hinder portion 
of the plastron, as the mesoplastra have in most turtles been 
excluded, we have no record of the fact. If, then, our surmises 
are correct, the plastron of most of our turtles consists of the 
interclavicle, the clavicles, and elements derived from the gas- 
tralia corresponding to three pairs of ribs. 
In turtles, it will be recalled, the plastral elements of the right 
and left sides are always distinct. In all for a varying period 
of life there is in the centre of the plastron a large fontanelle. 
In some, as Chelydra and the sea turtles, the fontanelle persists 
through life, or at least until a late period. We are safe in 
assuming that its existence is a primitive condition. On the 
other hand, in Sphenodon and in many of the extinct reptiles 
which possess gastralia, the median element is continuous 
across the mid-line of the abdomen. It seems therefore prob- 
able that the plastral bones of turtles, except clavicles and 
interclavicle, have been derived from the lateral elements of 
the primitive gastralia, while the median element has become 
aborted. I do not overlook the fact that in the crocodiles two 
distinct bones represent the median element found in the gas- 
tralia of Sphenodon. Even in the crocodiles, however, these 
bones are smaller than are those which lie farther from the 
mid-line. 
U. S. Nat. MUSEUM, 
October 10, 1898. 
