THE DEVELOPMENT OF THE MARSUPIAL SHOULDER GIRDLE. 761 



saurus and of other forms turns the long axis of the scapula almost directly backwards. 

 As may be seen in the almost perfect skeleton of Pareiasaurus in the British Museum, 

 if this be the case the interclavicle has to lie almost at right angles to the axis of the 

 body — a rather improbable condition. Moreover, if the inner borders of the precoracoid 

 and coracoid of Dicynodon be placed in contact with an imaginary antero-posteriorly 

 directed interclavicle, the scapula will be found to incline forwards. The imperfect 

 remains of two specimens of Kistecephalus, in the British Museum, show that in that 

 genus at least, the scapula almost certainly was directed somewhat forwards. The 

 early developmental condition of the scapula in the marsupial also seems to confirm the 

 view that in the early types this was the direction of the scapular axis. In Pareiasaurus 

 the anterior border of the scapula is supported by a distinct cleithrum or epiclavicle, as 

 is most probably also the case in Kistecephalus and Dicynodon. In the early develop- 

 ment of the monotremes it seems probable that the anterior part of the scapula will be 

 found to develop like the spine in the marsupial, from a membranous basis, and if this 

 be the case it will mean that the monotremes, like the lower Theromorpha, have the 

 anterior border of true scapula extended by the addition of a membrane-bone element, 

 the cleithrum, but with the cleithrum becoming co-ossified with the scapula, as is pro- 

 bably the case in the higher Theromorphine genus Cynognathus (14). In the marsupial 

 and eutherian type the change which has taken place is, that the cleithrum has become 

 shifted from the anterior border of the cartilaginous scapula to its outer side. The 

 marsupial scapula in its early condition scarcely differs from that of the adult monotreme, 

 except in the greater development of the acromion. There is an acromion process 

 situated on the anterior border, and no spine on the outer surface of the scapula, and 

 it is probable that if the scapula of the foetal marsupial were compared with that of a 

 very young monotreme before the development of the supposed cleithral portion, the 

 resemblance would be greater. If this comparison be correct, it will follow, that 

 though the spine in the marsupial and the eutheria is the homologue of the actual 

 anterior scapular border in the monotreme, the anterior border of the marsupial scapula 

 is the homologue of the anterior cartilaginous portion of the scapula in the monotreme. 

 Whether the study of the early development will prove that the anterior part of the 

 cartilaginous scapula is coincident with the ridge on the inner side of the scapula in 

 Ornithorhynchus, which is believed by Owen and Wilson and M'Kay to be the 

 morphological anterior border, remains to be seen. 



How the muscles may have arranged themselves in the Theromorpha it is difficult 

 to tell, but it may be regarded as fairly certain, that no muscle could have passed from 

 the inner side of the scapula to the humerus in front; and the peculiar development of 

 the supra-spinatus muscle in the monotremes is probably due to the attitude of the 

 scapula, for if this muscle existed in Cynognathus, it must have been extremely 

 minute. 



The question as to whether the posterior border of the monotreme scapula is really 

 the homologue of the posterior border in the higher forms, or, as held by Wilson and 



