1905.] THERIODOXT REPTILE DIADEMODOX. 99 



largely on carrion — possibly the carcases of the large Dicynodons 

 that had been killed by Cynognathi. 



It seems to me unnecessary to discuss the supposed relationships 

 between Diademodon and Trit)/lodo7i, n.s I have recently elsewhere 

 endeavoured to show that Tritylodoii is, as beheved by Owen, 

 Lydekker, Cope, and others, a true Mammal, and probably not at 

 all nearly related to the Theriodonts. 



The following are the principal measurements of thp jaw and 

 teeth :— « 



miUini. 



Leng-tli from tip of coronoid process to angle ... 61 



Depth of jaw at m 6 22 



Depth of jaw at m 3 20 



Length of seven molars 35 



Width of m6 6-7 



Length of m 6 6 



Both ilia are well presei'ved, but unfortunately only the inner 

 sides of each are displayed, and owing to the softness of the bone 

 it does not seem advisable to remove the matrix from the outer 

 side of either. The most striking character of the ilium is the 

 great antero-posterior development of the crest. From the 

 constricted part above the acetabulum the ilium extends upwards 

 and forwards to end in a rounded anterior expansion, and also 

 extends backwards to form a sharp posterior portion. The 

 anterior pai't has its anterior edge turned considerably outwards, 

 so that a deep concavity is apparently formed on the front part of 

 the outer side of the iliac expansion. Below the constriction the 

 ilium expands again to form the upper part of the large 

 acetabulum. In Cynog)iatha% the anterior part of the ilium is 

 missing, but the parts preserved are very like those of Diademodon, 

 and it is thus probable that Seeley's restoration of the anterior 

 part is too small. 



The most uuportant diflerence between the INIammalian ilium 

 and that of the Theriodont is that, owing to the pelvis in the 

 mammal lying more antero-posteriorly. the anterior part of the 

 ilium is itself sufficient for the attachment of the sacrum, and 

 hence the posterior part becomes usually greatly reduced or lost 

 completely. In Orycteropus the posterior part of the ilium is 

 unusually well developed, and not unlike the Theriodont condition, 

 but in most Marsupials it is pi-actically absent. In the Monotrenies 

 and in the Wombat there is a small but distinct posterioi' 

 portion. 



The pubis is a little imperfect, as a portion of the anterior part 

 is missing. In general structure it is very mammal-like. Imme- 

 diately below the acetabular portion, the bone is somewhat 

 constricted, and at this point there passes inwards and forwards 

 a plate which is probably of a similar nature to the anterior part 

 of the pubis in Cynognathus ; but in Diademodon it seems to be 

 directed more inwards than in Cynognathus. Whether it is a 



