CACOPS ASPIDEPHORUS 273 



seen in a small but preminent liooklike process on the inner posterior 

 side near the head. The proximal articular surface is, it is seen, quite 

 at the extremity, suggesting a limited range of motion as compared with 

 other forms, and it is also the case with articular surfaces for the radius 

 and ulna. Here, as in other cases, the absence of distinct condylar sur- 

 faces may be of generic value or merely ontogenetic. While, as is well 

 known, many amphibians, both extinct and living, do not have a com- 

 plete chondral ossification, leaving the articular surfaces throughout life 

 composed of cartilage, yet it is also quite possible that such a condition 

 found in fossilized bones may be merely indicative of juvenility. In 

 Trimerorliacliis, however, it seems certain that the character is one of 

 maturity, since more than a dozen humeri and a score or two of other 

 limb bones in the collection all present the same unossified condition. 

 From all of which facts I think the conclusion is obvious that Trimero- 

 rhachis was an aquatic form, notwithstanding the slenderness of the 

 humeri, which simply is an evidence of a short tail and limb propelling 

 habits. 



Three other humeri figured in plate 15 are certainly of new genera. 

 Two of them, figures 4 and 5, are of the usual temnospondylous type, 

 but a third (figure 1) may represent a new type of amphibian from 

 Texas. This humerus, it is seen, is an unusually slender bone, with its 

 articulations well developed. The lateral process is small and the con- 

 dyles are feebly developed. The bone, as preserved, shows but a slight 

 twisting of its proximal from the distal plane, which may be in part due 

 to the conditions of fossilization, though I think not. The genus when 

 it is better known will, I believe, be found to be of. a slender terrestrial 



type. _ ^^ - 



The humerus shown in figures 4a, h, and c'is remarkable for its mass- 

 iveness and shortness, in both respects excelling any other which I know. 

 Here also the chondral ossification is deficient, the articular condyles, 

 both proximal and distal, being roughened, more or less concave surfaces, 

 doubtless thickly covered by cartilage in life. The humerus shown in 

 figure 5 was found associated with that illustrated in figure 4. It is more 

 of the Dissorophus type, though distinctly different, as were also its 

 femora found associated with it. It also has an imperfect chondral ossifi- 

 cation. 



All of the humeri so far described lack the peculiar process just above 

 and to the outer side of the capitellar convexity, so characteristic of 

 Eryops, Euchirosaurus, and apparently also Trematops. It may be called 

 the ectepicondylar process, and seems to be homologous with a similar 

 process found in the Pelycosauria, but hitherto unknown in the Cotylo- 

 sauria. 



