in general, easily be traced upon the interior of the carapace. But the most marked results of this defective ossification is the entire separation of the eighth pair of ribs from those anterior, they being held in position only by dense fibrous investment. This is entirely different from what we are led to expect; for Cuvier, in remarking upon the carapace of this genus ('Ossemens Fossiles,' tome ix., p. 398), observes that "There are eight pairs of ribs, united by sutures, which in these, as well as in the Marine Turtles, are not dilated to their outer end, but of which the dilation extends with the age of the individual."
The relation of the alimentary canal to the size of the reptile, too, varies considerably from the proportions taught. In a specimen' two feet in extreme length, the entire length of the digestive apparatus was exactly four feet, divided a,s follows :—Œsophagus, nine inches; stomach, six ; small intestine, twenty-seven ; and colon, including the cloaca, six inches only— this absolute and relative length of the tract indicating a purely carnivorous diet on the part of its possessor. Cuvier gives the relative length of the alimentary tract to the body, as observed in the Emydidæ, as five to one ; and Rymer Jones, in contrasting the carnivorous with the herbivorous Testudinata, remarks that the "Emydidæ are more carnivorous in their habits, and in the Trionychidæ the alimentary canal is shorter — at least the larger intestine, which is not longer than the small." With Aspidonoctes, the large and small intes- tines, as we have seen, vary materially in length ; but they are almost continuous with each other, no angle being formed at the junction, and the separation marked only by a slight valvular constriction : the former, too, though possessed of a trifle greater diameter, is much less muscular than the small intestine, and consequently more distensible.
Again, Cuvier remarks that in those genera of the order other than salt-water Turtles, "the œsophagus presents only longitudinal folds and numerous orifices of mucous crypts." But in the species under consideration we find the greater portion of the gullet covered with bifid fringe-like processes; and upon the fauces, the upper portion of the larynx (rima glottidis), and the arches of the hyoid bone, presenting both the character and arrangement of those of the tongue, before noted as resembling the gills of the Menobranchus, or more properly, perhaps, the internal ciliated respiratory apparatus of the Tadpole.
