9*2 Tnniwcthnx of the Soutlt African Philosophical Society. 



dis^iiiviji frOtQ vf ihe'teiirth, but there is some evidence of its having 

 been distinct. In my previous description of this carpus I mistook 

 the first carpale for the first metacarpal. 



It is unfortunate that in the Theriodontia the group from which 

 the mammals appear to have directly sprung, only one carpus is at 

 present known, and that an imperfect one. This is the carpus of 

 Microgomphodon eumerus (13) described and figured by Seeley. 

 Seeley considers there are three bones in the proximal row, but as it 

 is very probable that the large element with which the ulna articulates 

 is not the pisiform as Seeley believes, but the ulnare, and that the 

 pisiform is missing, it seems likely that the proximal row in the 

 Theriodonts has four bones, as in the more primitive groups and the 

 majority of mammals. There appears to be only one centrale, and 

 only four carpalia. It is possible, however, that the element between 

 the radiale and the ulnare is not the intermedium, but a second 

 centrale, and that the intermedium is not seen in the specimen. 

 The arrangement of the bones in the Endothiodont carpus suggests 

 this possibility. It is impossible to decide the point by the figure. 



The examination of the series of carpi shows that there has been 

 very little evolution in the carpus from the higher Labyrinthodonts 

 to the Theriodonts. The fifth carpale becomes lost, but otherwise 

 any of the carpi from the Cotylosaurians to the Theriodonts might 

 be the one from which the mammalian type has been derived. 



The stages in the evolution of the tarsus, though they are less 

 fully known than the stages of the carpus, throw much more light 

 on the line of descent of the mammals. 



The Labyrinthodont tarsus is practically unknown. The tarsus of 

 the primitive Stegocephalian Archegosaurus is known, but there is 

 some difference of opinion as to its interpretation. Its most interest- 

 ing feature is that it consists of at least nine distinct elements, and 

 certain of these may represent the fusion of two. But while the 

 Stegocephalian tarsus contains so many distinct elements, the 

 number is found to be very considerably reduced in all the 

 descendants. In some the reduction appears to be due to a number 

 of the elements uniting together ; in others the reduction is evidently 

 due to some of the original elements being lost. 



In the Cotylosauria the tarsus is very imperfectly known. In 

 Pareiasaurus and Sclerosaurus there is a large tarsal bone probably 

 made up of the united tibiale, intermedium, and fibulare. The other 

 elements have possibly been cartilaginous. Case has figured a 

 fairly satisfactory tarsus of Pariotichus (5). It shows that the 

 proximal tarsal row consists of two distinct bones — either tibiale and 

 fibulare, or with possibly an intermedium united to one of the other 



