It is customary to divide the Eusuchia, most of which are extinct, into a longirostral and a brevirostral section. In the former the snout is much elongated and narrow, and the nasal bones, although they are sometimes very long, do not reach the nasal groove. The mandibular symphysis is very long, and is formed not only by the dentary but also by the splenial bones. In the brevirostral section the snout is shorter, sometimes broad and rounded off, and the nasal bones are supposed to reach the nasal groove, or at least to approach it very nearly; the mandibular symphysis is formed by the dentaries only. But these distinctions are quite arbitrary, and there exist all kinds of intermediate forms. For instance, in Goniopholis and Diplocynodon, which are both undoubtedly near allies of the recent Crocodiles and Alligators, the nasal bones are considerably removed from the nasal groove; and in Crocodilus cataphractus they are separated even from the premaxilla by the medio-dorsal suture of the maxillaries. Again, in Goniopholis the mandibular symphysis is so long that it comprises part of the splenial bones. Both typically long- and short-snouted forms occur already in the Upper Oolite, but in the Lower Jurassic age only long-snouted kinds seem to have existed. The latter cannot easily be connected with Belodon, one of the Parasuchia, on account of the position of the nostrils; the mere shortening of the long premaxillaries of Belodon would not transfer its distinctly paired nostrils to the anterior end of the premaxilla. To account for the position of the nasal groove in the Eusuchia, we have to go back to a primitive condition, such as that of the Pseudosuchian Aëtosaurus, and this consideration shows that the Parasuchia and Eusuchia are collateral branches.
The Eusuchia have been split into many families. Zittel, for instance, divides them into ten, some of them on insufficient grounds, since there are too many intermediate forms; and more, sometimes quite unexpected, modifications are still being found. Several of the accepted families represent collateral or convergent lines of development.
Fig. 105.–Group of Crocodiles. A long-snouted Gharial or Gavial (Gavialis gangeticus) on the top of Crocodilus acutus; a Nile Crocodile (C. vulgaris) in the foreground; C. palustris, a "Mugger," in the right upper corner. Observe the peculiar floating attitude of the young specimen.
There is the same tendency to transfer the choanae further back, owing to the formation of a solid secondary roofing in of the mouth, to transform the amphicoelous into procoelous vertebrae, to reduce the supratemporal foramina, and to obtain a better development of the dorsal armour, whilst that on the ventral side is gradually reduced. Lastly, there is a tendency towards a shortening and broadening of the snout, a condition which has reached its culmination in the Alligators, while the Gavials are survivals of another branch. The notches in the premaxilla, for the reception of some of the lower teeth, have also been acquired independently. Although the recent Crocodilia cannot now, as has been pointed out by Boulenger, be separated into different families, no valid diagnoses being possible owing to the existence of Tomistoma, their phylogeny shows them to belong to at least two heterogeneous groups.
Key to the Genera of recent Crocodilia.
I. Snout very long and slender. The mandibular symphysis extends at least to the fifteenth tooth, and is partly formed by the splenial bones.
a. Nasal bones very small, and widely separated from the premaxillaries .......... Gavialis gangeticus, p. [451].
b. Nasal bones long, in contact with the premaxillaries .......... Tomistoma schlegeli, p. [453].
II. Snout not slender, but triangular or rounded off. The mandibular symphysis does not reach beyond the eighth tooth, and does not reach the splenial bones.