494 THE NATURE OE THE CLAVICLE. 



views have recently been challenged by Grotte (No. 466) and Hoff- 

 mann (No. 467), on tlie ground of a series of careful enibryological 

 observations ; and until the whole subject has been worked over 

 by other observers it does not seem possible to decide satisfactorily 

 between the conflicting views. It is on all hands admitted that the 

 scapulo-coracoid elements of the shoulder girdle are formed as a pair 

 of cartilaginous plates, one on each side of the body. The dorsal half 

 of each plate becomes the scapula, which may subsequently become 

 divided into a supra-scapula and scapula proper; while the ventral half 

 forms the coracoid, which is not always separated from the scapula, 

 and is usually divided into a coracoid proper, a praecoracoid, and an 

 epicoracoid. By the conversion of parts of the primitive cartilaginous 

 plates into membranous tissue various fenestrse may be formed in 

 the cartilage, and the bars bounding these fenestras both in the 

 scapula and <;oracoid regions have received special names ; the ante- 

 rior bar of the coracoid region, forming the prascoracoid, being espe- 

 cially important. At the boundary between the scapula and the 

 coracoid, on the hinder border of the plate, is placed the glenoid 

 articular cavity to carry the head of the humerus. 



The grounds of difference between Gotte and Hoffmann and other 

 anatomists concern especially the clavicle and inter-clavicle. The 

 clavicle is usually regarded as a membrane bone which may become 

 to some extent cartilaginous. By the above anatomists, and by 

 Rathke also, it is held to be at first united with the coraco-scapular 

 plate, of which it forms the anterior limb, free ventrally, but united 

 dorsally with the main part of the plate ; and Gotte and Hoffmann 

 hold that it is essentially a cartilage hone, which however in the 

 majority of the Reptilia ossifies directly without passing through the 

 condition of cartilage. 



The interclavicle (episternum) is held by Gotte to be developed 

 from a paired formation at the free ventral ends of the clavicles, but 

 he holds views which are in many respects original as to its homo- 

 logies in Mammalia and Amphibia. Even if Gotte's facts are ad- 

 mitted, it does not appear to me necessarily to follow that his 

 deductions are correct. The most important of these is to the effect 

 that the dermal clavicle of Pisces has no homologue in the higher 

 types. Granting that the clavicle in these groups is in its first stage 

 continuous with the coraco-scapular plate, and that it may become 

 in some forms cartilaginous before ossifying, yet it seems to me all 

 the same quite possible that it is genetically derived from the clavicle 

 of Pisces, but that it has to a great extent lost even in development 

 its primitive characters, though these characters are still partially 

 indicated in the fact that it usually ossifies very early and partially 

 at least as a membrane bone^. 



^ The fact of the clavicle going out of its way, so to speak, to become cartilaginous 

 before being ossified, may perhaps be explained by supposing that its close connection 

 with the other parts of the shoulder girdle has caused, by a kind of infection, a change 

 in its histological characters. 



