9 2 



THE AMERICAN NATURALIST. [Vol. XXXV II. 



shown by Cope, figure 4, there are several bones, the quadrato- 

 jugal is broken and crowded under the lower end of the suspen- 

 sorium, and that there was a distinct process is indicated by a 

 broken stump, which is undoubtedly the posterior end of the pro- 

 squamosal arch, figure 5. The position of this arch shows that 

 the upper vacuity was large and rounded and the lower narrow 

 and smaller, perhaps nearly closed, but there were two distinct 



Clepsy drops natalis. — The posterior portion of the skull in 

 the type specimen is badly crushed, and the whole specimen is 

 covered with a thin layer of matrix, so that there is much uncer- 



tainty regarding the exact limits of each bone. The anterior 

 portion of the skull is strikingly similar to the same region in 

 Dimetrodon and Kmbolophorus in all the characters of extreme 

 specialization such as the high and narrow facial region, the 

 deep diastema and the enlarged canine and incisor teeth so that 

 there is strong presumptive evidence that the posterior arches 

 had an equal similarity. This is further borne out by the very 

 fact of the fractured condition of the posterior angles of the 



because of the weak arches. Certain elements of the temporal 

 region can be made out beneath the thin coating of matrix, but 

 these do not seem to me to be what Cope thought them. He 



