4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I39 



T. punctulatus, and especially difficult to demonstrate any real affini- 

 ties with Eosauravus copei. 



Description. — The specimen consists of the posterior two-thirds of 

 a postcranial skeleton preserved belly-down on a coal stratum. Neither 

 the opposing slab, probably containing a dozen thoracic vertebrae and 

 caudal neural arches, nor adjoining blocks of matrix containing the 

 tip of the tail, some terminal phalanges of the left pes, and the anterior 

 end of the skeleton, were collected. The remaining parts of the skele- 

 ton have undergone very little deterioration since Cope's time, judging 

 from the excellent photograph presented by Williston (1908) and 

 republished (with inaccurate retouchings) by Moodie (1909). 



The presacral, sacral, and anterior caudal vertebrae lie on their right 

 sides (as observed by Cope, 1897) in such a manner as to cover the 

 proximal tips of the right ribs while the proximal ends of the left pre- 

 sacral ribs are pressed against the upper (left) surfaces of their cor- 

 responding centra. The outline of successive neural spines is clearly 

 visible on the right side between successive ribs. The caudal vertebrae 

 posterior to the rib-bearing caudals are preserved with ventral side 

 down and have lost their neural arches, thus exposing the neural canal 

 as a longitudinal groove in the dorsal surface of the centra. Un- 

 fortunately, Moodie (1909, pi. 5) illustrated the entire column as 

 though it were oriented with the dorsal side uppermost (figure repro- 

 duced by Case, 191 1, fig. 8). The result is an erroneous picture of the 

 vertebrae from anterior caudals forward. Cope's illustration (1897, 

 pi. 3) shows the correct orientation, but is only slightly suggestive of 

 the true form of the vertebrae. 



The true form of the presacral and anterior caudal vertebrae may 

 be reconstructed with reasonable accuracy from a composite of details 

 exhibited along the column. Specifically the impression of the anterior 

 presacrals clearly shows the contour of the centrum ; the first 5 pre- 

 sacrals and anterior caudals preserve details of swollen neural arches 

 as well as of the centra and intercentra. The position of intervertebral 

 foramina is clearly indicated by a series of circular pits. Figure I is 

 presented as a reconstruction based on composite detail. 



There seems to be little doubt that the neural arch is low and broad 

 as mentioned by Romer (1947, p. 300), has a low spine, and has a 

 perceptible swelling above the posterior zygapophysis ; also that small 

 intercentra are present. The latter are indicated between the first 

 several presacral centra, between the 1st and 2d caudal centra, and 

 by a haemal wedge between the 3d and 4th caudal centra. In the pre- 

 sacral series the left ribs appear to have been crushed precisely against 



