That the emargination was overlooked by the author seems strange, as in the same paper in which this figure is given occurs the description of Holosaurus, founded upon that very character. If the emargination is sufficiently important to base a genus in the one case, then it should be in the other, and the character could not be applied to Edestosaurus, based upon characters which it hardly seems possible that the author himself could seriously consider, for E. dispar was the type of Edestosaurus.

It will be observed, further, that the figured coracoids differ very materially in size, those with the emargination pertaining to a small species, while C. dispar is one of the largest. In our Museum there are three specimens with the emarginate coracoid, all of them small or very small, the described specimen of C. velox being the largest.

The point of chief interest in this relation is the value that can be given to this character. Is it individual, specific or generic? Marsh has called it generic, but we think an examination of the two very complete specimens of C. tortor and C. velox in our Museum will convince any unprejudiced student that he is in error.

A comparison of the figures herewith given of the paddles will show their great resemblance, and these two forms of paddles have been figured because the species are the most unlike of any that we know in the genus. As all the small specimens seem to possess this character, and as they cannot be called immature specimens, we believe the character is a specific one. As Marsh says, typically both Clidastes and Edestosaurus have a non-emarginate coracoid, so that neither name could apply to the emarginate form, were it generically distinct.

Our Museum also contains both forms of the coracoid pertaining to the genus Platecarpus, of which Holosaurus is a synonym.

While studying the specimen above described, a striking similarity was observed to several other specimens already determined with confidence as C. pumilus Marsh. A more careful comparison failed to bring out any real differences beyond size, and even this was shown to be very inconstant.

The following comparison of the descriptions given by Marsh will be of interest.

   C. pumilus.    C. velox.
Teeth. Nearly round at base somewhat curved and with smooth enamel.Premaxillary and maxillary teeth smooth and subcompressed.
Quadrate. The rugose knob near the distal end of the quadrate is similar to that in C. Wymani (just below the posterior superior process is a prominent rugose knob with a deep pit under it), but has no articular pit under it. The hook is comparatively short and has a free compressed extremity. The articular margin is not deflected toward the meatus.The great ala less curved than in E. dispar, concave transversely on both surfaces. The alar process has its articular process very narrow in its extension over the great ala. No notch in posterior margin of external angle. On the ridge below the angle and nearly opposite the meatal pit is a strong rugosity which is rudimentary or wanting in C. dispar. The posterior margin of the hook is only a narrow tongue projecting towards the meatal pit, instead of a broad articular surface.
Cervical Vertebrae. Articular face nearly vertical, and having a broad transverse outline with faint superior emargination. The hypapophysis stout and transversely triangular.Articular face transverse.