THE DEVELOPMENT OF THE MARSUPIAL SHOULDER GIRDLE. 767 



and the monotremes is, that there is believed to be evidence of the amphibian pre- 

 coracoid being absorbed by the clavicle. This view is held by the majority of present 

 comparative anatomists, though opposed by Seeley. It has apparently originated in 

 the belief that the mammalian clavicle is developed from a cartilaginous basis. In 

 1864, Gegenbaur (18) discovered in connection with the early developing human 

 clavicle, cells which he believed to be cartilage cells. Gotte (19), in 1877, and 

 Hoffmann (20), in 1879, confirmed Gegenbaur's view. From these observations it 

 has been assumed that the clavicle has a cartilaginous basis, and further, that this basis 

 represents the amphibian precoracoid. It seems rather remarkable that these con- 

 clusions should have been so readily accepted, seeing that none of the observers have 

 clearly demonstrated the presence of true cartilage in the clavicular region at a stage 

 antecedent to the development of bone. In the development of the marsupial clavicle 

 I have shown that not only is there no cartilaginous basis, but that what cartilage 

 appears at the ends of the clavicle is only developed at a late stage in development, and 

 is probably of no more morphological significance than is the cartilage found in some 

 mammals in connection with the premaxillary bone. It is very probable that in many 

 Eutheria the secondary cartilage at the ends of the clavicle develops at a much earlier 

 stage than in the marsupial, but I very much doubt whether in any mammal cartilage 

 is present till a considerable time after the bony clavicle is well formed. In a human 

 embryo of seven weeks, though the clavicle is well ossified, I can find no cartilage at 

 either end ; and even in the mole, where cartilage is so well developed in the later 

 stages as to lead Parker to believe that the short clavicle represented the clavicle + 

 coracoid, the clavicle develops as a pure membrane bone. In the lizard (Mabuia 

 sulcata) the clavicle is also a pure membrane bone, and is developed in almost exactly 

 the same manner as in the marsupial. 



It will thus be seen that there is good reason for believing that the precoracoid has 

 not been absorbed by the clavicle ; and if this be so, there is left no reason for doubting 

 the homology of the amphibian precoracoid with the anterior coracoidal element in the 

 theromorphs and monotremes. 



In most reptilian orders the precoracoid is either quite rudimentary or lost. It is 

 found as a rudiment in some birds, e.g., Apteryx, Struthio, Casuarius, and possibly 

 also in some lizards. In both the Chelonia and the Sauropterygia a well-developed 

 osseous bar passes from the scapula forwards and inwards, and which has generally been 

 regarded as the precoracoid. Andrews (21) has shown, however, that in the Plesio- 

 saurians the structure is a true portion of the scapula, and there is no satisfactory 

 evidence that the anterior bar in the turtles is anything more. 



Another element in connection with the shoulder girdle that has given rise to a 

 difference of opinion is the little cartilage situated at the inner end of the clavicle. By 

 Parker it was named the ' omo-sternum,' and was held to be the homologue of the 

 median element situated in front of the precoracoid in the frog. By Gegenbaur it 

 was formerly held to be a ' lateral epi-sternum,' but recently he has referred to it as 



