146 DR. R. BROOM ON THT?; 



that we have this tavsus well preserved, as it shows us how the 

 reptilian tarsus has been derived from the primitive amphibian 

 type. As figured by Jaekel, it has three proximal elements and 

 four distal. The elements articulating with the fibula are 

 manifestly those we have identified in the more pi-imitive type 

 as fibulare and intermedium, and they are so identified by Jaekel. 

 The third proximal element articula,tes with the tibia and is 

 manifestly the tibiale. It is much smaller than the other 

 proximal elements. The distal elements are the first four distal 

 tarsalia. Between the fibulare and the intermedium is the 

 tarsal foramen. 



The great difi^erence between the Mici'osaurian tarsus and that 

 of the Temnospondylous types is that in the former all the 

 central elements have disappeared or become generally reduced 

 and cartilaginous. 



In the earliest undoubted reptiles known — the Cotylosauria — 

 the tarsus has onl}^ been preserved for us in a very few forms. 



In LinDioscelis pcducUs Williston (text-fig. 4), though the tai-sus 

 is not perfectly preserved and was probably largely cartilaginous, 

 we have the two principal elements preserved in position. Willis- 

 ton identifies them as the fibulare and the vuiited tibiale and 

 intermedium, but it seems much more probable that the}^ are 

 the fibulare and the intermedium, and that the tibiale is either 

 lost or was cartilaginous, and this latter view is admitted by 

 Williston as not impossible. 



Another very primitive type of which we know the tarsus is 

 Eosauravus copei Williston (text-fig. 5), from the Middle Penn- 

 sylvanian of North America. Unfortunately, the head of this 

 animal is unknown, and we are thus in doubt whether or not it is a 

 Cotylosaur. The tarsus has the elements preserved in only slightly 

 disturbed relations. There are two large proximal elements, which 

 a comparison with other early types leads us to consider as the 

 fibulare and intermedium. Tliere are five distal tarsals, and a 

 small element on the tibial side of the tarsus which is 

 probably the reduced tibiale. The metatarsals are somewhat 

 displaced, and it is not improbable that the distal tarsals and 

 the tibiale are also a little displaced. 



In Seymouria haylorensis Broili (text-fig. 6) the neai-ly perfect 

 tarsus has been discovered by Williston. It consists of two large 

 proximal elements and a third small one, and apparently five distal 

 tarsals. The two large tarsals are regarded by Williston as the 

 fibulare and tibiale, and the small proximal element the centrale. 

 I interpret them as in Scincosaurus — -the fibulare, intermedium, 

 and tibiale. 



The only other Cotylosaur in which the tarsus is satisfactorily 

 known is Procolophon trigonicejys Owen (text-tig. 7), and though 

 Froeolophon in having a roofed temporal region is usually placed 

 with the Cotylosaurs, yet, being a late Triassic fo:m, it has 

 advanced in many respects so far from the Cotylosaurs of the 

 Permian of North America that \% ought, perhaps, really to be 



