THE DEVELOPMENT OF THE MARSUPIAL SHOULDER GIRDLE. 757 



of the biceps. From the posterior part of the inner end of the main part of the 

 coracoid an irregular quadrangular process passes backwards and downwards to meet 

 the lateral process of the pre-sternum. Though, the tip of this coracoid process rests on 

 the lateral process of the sternum, the two structures are quite distinct and separated 

 by a few fibres of connective tissue. It will be noticed that this mode of attachment 

 of the coracoid to the sternum differs considerably from that of Trichosurus. Here, one 

 would almost be inclined to think that there never had been any structural continuity. 

 If in an earlier stage the cartilage of the coracoid is continuous with that of the ster- 

 num, the separation of the two elements can hardly have been by the degeneration of 

 a part of the coracoid, as in Trichosurus, and the appearances would rather suggest that 

 the structures had split apart without loss to either of the elements. 



The clavicle is shorter and stouter than in Trichosurus embryos. The outer half 

 is considerably flattened, being oval on section. The inner half is much stouter, and 

 towards its inner end becomes hollowed on the under and inner side to accommodate 

 the small omo-sternum. At both ends of the clavicle there is a distinct cap of large 

 irregular cells which are becoming cartilage cells, though true cartilage can scarcely, as 

 yet, be said to exist. 



The omo-sternum is comparatively small, and its anterior end is ill defined. It is 

 specially interesting in that it is structurally continuous with the true sternum. 



In fig. 6 an anterior view is shown of the right shoulder girdle with the clavicle 

 removed. In fig. 12 is shown a ventral view of the same arch, with the sternum, 

 omo-sternum, and base of first rib. 



The Shoulder Girdle in a Newly-born Embryo of Petrogale penicillata 

 (21 mm. Greatest Length), (figs. 9 and 13). 



The Rock-wallaby differs so little in its anatomical structure from the true Kan- 

 garoos that this embryo may safely be taken as the type of the Macropodidse ; and 

 though some differences will likely be found in the girdle of the Rat-kangaroos, it is 

 probable that the condition in true Kangaroos will be found to agree closely with that 

 in the present embryo. 



The axis of the scapula is directed upwards and well backwards, and the blade of 

 the scapula is larger in comparison with the size of the glenoid cavity than in phal- 

 angers. The acromion arises from the outer side of the anterior border a little distance 

 above the glenoid cavity. It passes forwards and slightly outwards as a stout process 

 which is oval on section. After passing forwards for some distance it passes downwards 

 and inwards to meet the clavicle — this lower part being somewhat flattened. There is a 

 tract of ossification representing the beginning of the spine, and presumably homo- 

 logous with the cleithrum of the primitive reptiles. With the exception of this-low 



VOL. XXXIX. PART III. (NO. 29). 6 B 



