THE DEVELOPMENT OF THE MARSUPIAL SHOULDER GIRDLE. 753 



The Shoulder Girdle op a Newly-born Embryo of Trichosurus vulpecula 

 (14*8 mm. Greatest Length), (figs. 3, 5, 10, and 20). 



It was in the shoulder girdle of the newly-born Trichosurus that the interesting 

 development of the marsupial coracoid was first noticed, and it is remarkable that it is 

 here that the development is most perfect, for whereas before birth chondrification is 

 incomplete, almost immediately after birth degeneration begins. 



The scapula is an irregular triangular plate lying in a plane parallel to median plane, 

 and with its long axis directed upwards and forwards. Its upper border is markedly 

 convex, the posterior concave, and the anterior moderately straight. The outer surface 

 of the plate is almost perfectly flat, there being no trace of a cartilaginous spine. From 

 the anterior border of the scapula, near the union of the middle with the lower third, 

 arises the well-developed cartilaginous acromion. It passes outwards as a moderately 

 round process, and then forms a more flattened plate, which passes downwards and 

 slightly inwards to articulate with the clavicle. Where the outward part of the 

 acromion meets the downward part, a sort of cartilaginous knuckle is formed. From 

 this point upwards and forwards, along the outer surface of the scapula in its anterior 

 part, passes a distinct membranous structure, which divides the supra-spinatus and infra- 

 spinatus muscles. It may be compared to a membranous flap whose posterior surface is 

 somewhat concave, owing to the outer border partly overlapping the infra-spinatus 

 muscle. It is this membranous structure which ossifies to form the spine of the scapula 

 along with the acromion. Though it is not as yet ossified, it is on the point of ossifying, 

 and osteoblasts and osteogenic fibres can be distinctly made out. Though the projecting 

 part is very thin, it rests on a well-formed base, which on section is somewhat triangular, 

 thus giving the membranous spine a firm attachment to the scapula. In connection 

 with the occurrence in Pareiasaurus of a distinct ' epiclavicle,' or, as it has been termed 

 by Gegenbaur (2), a ' cleithrum,' Seeley (3) has already pointed out the probability of 

 the mammalian scapular spine representing this structure, and of its thus being morpho- 

 logically a distinct element. The fact that in the Marsupial, and most probably all other 

 mammals, the spine, apart from the acromion, is not developed from cartilage, affords 

 very strong confirmation of the correctness of Seeley's opinion. In the acromion being 

 itself a cartilaginous structure, it is manifestly not a part of the cleithrum, but a develop- 

 ment of the scapula proper. 



The coracoid is very well developed, and may be regarded as formed of an upper part 

 which articulates with the lower end of the scapula and forms part of the glenoid cavity, 

 and a lower part which articulates with the sternum and the first rib ; the two parts 

 being united by a slightly constricted intermediate portion. The outer surface of the 

 upper part is mostly hollowed out as if forming part of the glenoid cavity, but probably 

 only a part of the hollowed surface forms an articular surface for the humerus. To a 

 little ridge on the lower edge of this part of the coracoid, and a little behind the plane 

 of the front of the glenoid cavity, is attached the long head of the biceps. The lower 



