neurocostal shield plus dermal marginals (see Fig. 3); (3) The Protostegids (and Toxo chelys) have the most complex bithecal cara pace known, retaining the subdermal elements with not only the marginals but supramarginals and a median keel of expanded epineurals; (4) Dermochelys in contrast with Trionyx has an osteodermal (mosaic) carapace, plus only the subdermal nuchal; (5) The Testudinate plas tron is far less variant than the carapace. It remains complex in recent form only in Der mochelys, with both subdermal and dermal ele ments.
Taking both carapace and plastron, Der mochelys and the Protostegids are thus closely apposed, and other forms variant. But as in dicated by large nutrition canals of the epineu ral keel, Archelon had a heavy leathery hide instead of an outer hornshield. It is, therefore, possible and probable that there were also dermal elements in the plastron. In this case the Protostegids were completely bithecal and fall still closer to the "leatherback° than is in ferred from the resemblances cited.
The deeply interesting paleontologic problem here briefly outlined, is in a measure solved. But many details and doubts yet remain which only discovery of further evidence can resolve.
It is then virtually proved that in the primi tive carapace there are two overlapping bony layers—the neurals and pleurals (costals) un derneath, with a neural, pleural, supramarginal and marginal keel above. In the great majority of turtles the underlying neuropleural layer early interlocked with the marginal dermal keel, and all traces of the other dermal keels were gradually lost, or the hornshields were formed instead. But the Protostegas with the related form Toxachelys (see Fig. 2) indicates that an epithecum of median and lateral rows of dermal elements could also reach prominence, finally replacing the endothecum as in Der mochelys. The latter could not always have had so prominent an osteodermal mosaic as now, and it is difficult to believe that weak and thin nenrals like those of Archelon were not once present in sequence with the nuchal. It is also significant that the nuchal, the pelvis and the epiplastron of Archelon strongly resemble these several elements in Dermochelys.
In recapitulation it is seen that: (1) The Trionyx carapace is now completely endothecal or subdermal; (2) in the majority of Cryptodi ran and Pleurodiran Testudinates the carapace is diplogene,— that is, equals the subdermal The general biologic problem has been outlined for just 50 years and naturalists who have suc cessively given their attention to it are E. D.
Cope, L. Dollo, G. Baur, 0. P. Hay and G. R. Wieland; while especially J. Versluys and his students have considered the osteological evidence.
As a corollary the features of Dinosaurian armor are also brought into unity. In the Dinosauria as in the Testudinata there are the two armor forming strata. But while in the Testudines, Miolansa affords the sole example of cranial armature, this occurs frequently in Dinosaurs and becomes extraordinary in the Ceratopsids, in which both the nether and upper layers take part in armor formation. Con versely in Polaeanthus the deep layer forms a lumbar hip carapace. In the Stegosaurs the simple dermal armor takes the most singular expression seen in reptilians existent or ex tinct. Wieland, who was the first to point out this significant parallelism says: °At first sight all development of dermal armor may appear to be mainly a senile feature due even to inertia — the general life move ment of the individual and the race. But it is also evident that the development of dermal ossicles in series resulting finally in a protective sub, and ostcodermal armature or carapace is a profound change co-ordinated with striking endoskeletal alteration. The Dinosauria de veloped their varied armor types late and when already highly specialized, apparently with only varying or far less success than the Testudin ates. In them an early development of simpler patterns appears to have resulted in an exceed ingly long lease of life.a
Baur, G., (Systematische Stellung von Dermochelys Blainville> (1889); Cope, E. D.,