PALEONTOLOGY: S. W. WILLISTON 
653 
they are preserved throughout the column, and not a few of them have 
been mended more or less completely. They are relatively slender for 
the size of the animal, and but little curved. Their unusual length was 
unexpected. The rib of the fifteenth vertebrae, for instance, measures 
more than ten inches. Inasmuch as the depth of the body is fixed by 
the girdles these ribs show decisively a broad body. The width back of 
the fore-legs could not have been less than twenty inches, or nearly a 
half of the length of the body to the sacrum, and may have been more. 
The length of the twenty-second, or fifth presacral, rib is seven inches. 
In the cervical and dorsal region they articulate by two well separated 
heads, the tubercle with the diapophysis, which is prominent, the head 
with the intercentral space. Near the middle of the back the broad 
head is less markedly dichocephalous, and their movability was less. 
With the seventh or eighth presacral they begin to show sutural union 
with the diapophysis and a facet on the front end of the centrum. At 
the fifth presacral the sutural union is obliterated and the ribs were 
fixed and immovable. As in Dimetrodon, the ribs extend back quite to 
the sacrum, the first presacral pair having an expanse of not less than 
six inches. Perhaps this unusually broad body is another reason for the 
short tail. 
In a previous paper it was observed that the articular surface at the 
proximal end of the humerus in Ophiacodon has a spiral form, as is usual 
in the pelycosaurs, and that, in life the humerus could not have been 
depressed much below the horizontal without dislocation from the 
socket. Mr. Watson has called attention to this peculiarity of the 
American theromorphs and cotylosaurs, and has suggested that it ex- 
plains the unusual expansions of the ends of this bone in these animals. 
It has been assumed that the remarkably broad and stout humeri of the 
cotylosaurs and pelycosaurs meant burrowing habits, but I have shown 
the error of this assumption in the fact, that, for the most part, the 
front legs are short, the fingers scarcely reaching as far forward as the 
end of the beak. Certainly an animal that could not reach in front of 
its nose would have difficulty in excavating a hole for the body to enter, 
unless it temporarily removed its head! The legs were carried, for the 
most part, with the elbow bent at a right angle, and the movement of 
the foot backward and forward in this position would account for the 
spiral motion in the glenoid socket. Furthermore, such a position of the 
limbs would require, for the support of the body, strong and broad mus- 
cular attachments on the humeri, a sufficient explanation of their shape. 
Nearly every detail of the skull has been worked out and will be 
described and illustrated later. Suffice it to say here that the presence 
