"cervico-dorsals" of the Sauropoda and those of the emeu (Dromæus), as follows: "The long neck, similar in structure and almost as flexible as that of an emeu (Dromæus), could thus pass through a prodigious arc in the search for food, either under or above water. The neck motion partly involved the anterior non-spine-bearing dorsals (vertebræ with free ribs, equivalent to the cervico-dorsals' of birds), as in Dromæus, behind which the comparatively inflexible, large, spine-bearing dorsals rose to maximum height in the sacrum for the inser

FIG. 6. Neck of bird (Dromæus). Cervicals 13-14 and cervico-dorsals 1-2 entirely lacking median spines; dorsal 3 with a large blunt median spine. After Osborn.

tion of the ligamentum nucha and elevator muscles." The importance of such an hypothesis of function will appear in the following description and discussion, and it applies to all the Cetiosauria, namely, to the Morosaurus and Diplodocus types as well, which, so far as known, are uniform with the camarasaur type in the peculiar bird-like arrangement of the posterior cervicals and anterior dorsals. (See Fig. 7.)

Again, in the description of Diplodocus, Osborn ('99 (1), p. 200) pointed out the resemblances in the relations of the posterior ribs to those of Apteryx. Two features were brought out, namely: two of the ribs actually underlie the anterior

border of the ilium, both in Diplodocus and Apteryx; the last dorsal vertebra of Diplodocus coalesces with the superior border of the ilium by a bar, which may be considered either a metamorphosed rib or an expansion of the metapophysial lamina if this is a rib, Diplodocus presents a condition analogous to that in Struthio, in which the last vertebra and rib, technically known as a "pelvic vertebra," is all but united with the ilium. (See Figs. 8 and 9.)

Dorsal Vertebral Formula. - The latest contribution to this subject results from the explorations of 1899 in the dinosaur beds of Wyoming. Holland (1900,

p. 817) shows from the explorations of

FIG. 7.

the Carnegie Museum in the Jurassic of Wyoming that the number of dorsal vertebræ in Diplodocus has been overestimated hitherto by Osborn; that this animal possesses in fact only ten dorsal vertebræ, the entire vertebral formula being estimated as follows:

Contemporaneous with this discovery is that of the American Museum party, that the dorsal vertebræ of Morosaurus also number ten.

Thus, in two of the largest Cetiosauria or Sauropoda we have an extremely short back, resembling the short back of birds, also an extremely long flexible neck, a very rigid attachment between the sacrum and ilium, correlated with the power of temporarily raising the entire presacral portion of the body.

The significance of these avian resemblances in the neck and trunk of these gigantic dinosaurs is rather homoplastic than genetic, for the peculiar paired cervical and cervico-dorsal spines, the posterior abdominal ribs, the lengthened preacetabular iliac bar, correlated with certain feeding motions, are bird-like structures mingled with other non-bird-like structures too numerous to mention. So also with the resemblances among the bipedal dinosaurs, in which the presacral portion of the body is permanently raised, bird-like and non-bird-like structures appear in close propinquity.

The main avian character pervading all Dinosauria is the one originally observed by Gegenbaur, namely, the close junction of the astragalus with the tibia or tendency to form a tibiotarsus. However, where there is so much smoke there may be some fire, and we may now proceed to look into the probability of the existence of a primitive bipedal dinosaur-avian

ancestor.

V. THE CLAWED QUADRUPEDAL ANCESTRY OF BIRDS.

Pycraft ('96, p. 261) has recently discussed with care the osteology of Archæopteryx. In opposition to the view of Hurst, that the manus retains five digits, two of which were used in climbing trees, Pycraft supports the older view, that digits I, II, III are the only ones represented, and that digit III, as in the Archosauria1 generally, and in the Dinosauria in particular, had four phalanges, the terminal of which was armed with a claw. In addition to these reptilian characters are the thecodont, or socketed teeth, the flat, or amphiplatyan,

1 "Archosauria" is a term employed by Cope for reptiles with two cranial arches at the back of the skull, namely, Rhynchocephalia, Crocodilia, and Dinosauria. In the writer's opinion this group should be extended to include the Lacertilia, in which one arch has been lost.

FIG. 8. Diplodocus. Ninth dorsal vertebra with 9th free rib behind ilium, roth dorsal, or "pelvic " vertebra, with rib element coalesced with ilium. After Osborn.

FIG. 9.- Apteryx. Ilium overlapping the two posterior ribs 9th and roth, as in Diplodocus. Decided ileo-pectineal process. Uncinate processes as in Rhynchocephalia.

Struthio. On right side, P.S.2, 7th dorsal; on left side, P.S.1, 8th dorsal, overlapped by ilium; 9th and 10th dorsals absorbed in

sacrum.