186 DE. R. BEOOM ON THE STEUCTUEE AND [June 4, 



is even more greatly expanded antero-posteriorly than in Udeno- 

 don, and the ischium is produced backwards to a greater extent, 

 while the obturator foramen is considerably larger. The pelves in 

 the two forms are, however, manifestly closely related, though the 

 Theriodont pelvis comes nearer to the Mammalian type than does 

 the Anomodont ; while the latter retains more of the primitive 

 characters. 



The femur does not greatly resemble the femora of those few 

 Theriodonts in which that bone is known. It shows hardly any 

 of the remarkable features of the femur of Ictidosuchus, There 

 is, however, a distinct depression immediately inside the great 

 trochanter corresponding to the deep pit in the femur of that 

 early Theriodont. It is probable, however, that in man}'' Therio- 

 donts the femur was as little specialized as in Udenodon. In 

 Cynodraco the distal end of the femur is fairly like that in the 

 Anomodont. On the whole it must be admitted that the Anomodont 

 femur resembles the Mammalian type more than that of any other 

 known Reptilian group. 



The fibula agrees with that in the Monotremes and Marsupials, 

 in having the upper end expanded and passing beyond the head 

 of the tibia. A similar condition is seen in Aristodesmus. 



The tarsus is especially interesting, in that while it shows close 

 affinities with the primitive types, it also foreshadows the tarsus 

 of the mammal. Assuming that my interpretation of the 

 elements is correct, the tarsus would show some resemblance to 

 that in the Pelycosauria, where, according to Cope (16), a distinct 

 intermedium and centrale are present. In the large majority of 

 Eeptiles the intermedium and the centrale become early united 

 with the tibiale. In the mammal it would appear that the astra- 

 gulus is the tibiale, the calcaneum the fibulai-e, the navicular the 

 centrale, and that the intermedium is missing. The condition of the 

 elements in Udenodon would seem to favour this view. By Gegen- 

 baur, Flower, and others the mammalian astragalus has been held 

 to represent the united tibiale and intermedium ; while according 

 to Emery the astragalus represents a fusion of the intermedium 

 and a paracentrale. When, in any form, a single cartilage or bone 

 occupies the situation held by two elements in an ancestral type, 

 many morphologists incline to regard the single element as morpho- 

 logically equivalent to the two ; but in the very large majority of 

 cases there is good reason to believe that the single element is 

 only functionally equivalent to the two, and that it is the homo- 

 logue of only one, the other being completely lost. Thus, if we 

 compare the human carpus with that of the Baboon, we might 

 readily conclude that the scaphoid or radiale in Man was equivalent 

 to the small scaphoid and centrale together of the loM^er form ; 

 but we know from development that a rudiment of the centrale 

 is present and quite distinct in man in early foetal life, and that 

 as development advances it completely disappears. Very many 

 similar instances might readily be given. Of course this rule does 



