1907.] OF MAMMAL-LIKE REPTILES. 1055 



clavicle passes up along the front of the scapula, taking the place 

 of the lost cleithrum. The interclavicle is large and T-shaped. 

 In MesosaibritjS and Heleosaurus {11) the scapula, coracoid, and 

 precoracoid are completely anchylosed. 



In Galechirus the scapula is long and narrow, except at the 

 lower end, where it broadens out for articulation with the pre- 

 coracoid and coracoid. There is no distinct acromion, though the 

 anterior border of the scapula is somewhat twisted. The coracoid 

 and precoracoid are well developed, but much smaller than in 

 Procolojihon and not anchylosed. The interclavicle has a large 

 and rather broad posterior portion. The clavicles are well 

 developed and appear to pass a considerable distance up the front 

 of the scapula, but there does not appear to be a cleithrum. 



In the Therocephalia the shoulder-gu-dle is not very fully 

 known. In Ictidosuchus (15) only the cartilage-bone elements 

 are preserved. The scapula is long and slender in its upper part 

 and broad at the lower end. There is no distinct acromion and 

 no twisting of the anterior border of the bone. The precoracoid 

 is a large flat, somewhat square-shaped bone with the foramen 

 completely surrounded by the bone. The coi-acoid is smaller and 

 of the usual shape. It is not known whether there is a cleithrum 

 or not. 



In the Anomodonts the shoulder-girdle is well known. The 

 scapula is long and well developed and has a well-marked 

 acromion. The precoracoid has a large foramen, which is in part 

 formed by the scapula. One of the most noteworthy characters 

 in this type is the presence, at least in Dicynodon and Oudenodon, 

 of a distinct cleithrum. The interclavicle in some forms is short, 

 e. g. Lystrosaurus (16), while in others, e. g. Dicynodon, it is 

 elongated. 



In the Cynodontia the shouldei'-girdle is not fully known. 

 Seeley (17) has figured the scapula with portions of the coracoid 

 and precoracoid of Cynognathits, and I have recently figured an 

 imperfect shoulder-girdle of J^lurosuchus (18). The scapula is 

 well developed and somewhat like that of the Anomodonts, having 

 a well-formed acromion. The coracoid and precoracoid, so far as 

 known, are also Anomodont-like. There is no evidence of a 

 cleithrum. There is a well-formed clavicle in Diademodon, but 

 the interclavicle is not yet known in any Cynodont, but probably 

 occurs in all species. 



In the closely allied Monotremes the only essential difierence 

 in the shoulder-girdle from the Cynodonts is in the precoracoid, 

 which has become reduced and no longer articulates with the 

 scapula. 



In Marsupials and Eutheria the precoracoid is lost as a skeletal 

 element and the coracoid rudimentary, but, as I discovered some 

 years ago (19), the Marsupial at birth still has a large coracoid 

 whicli articulates with the sternum as in the Monotremes. 



The examination of the humerus, radius, and ulna in the various 



