1893.] OF THE TEEEESTBIAL VEETEBEATA. 587 



portion of the coracoidal skeleton not involved in the clavicular 

 apparatus, and to its homologue the term coracoid is applied in all 

 the lower Yertebrata, whether it be ossified or not. What, then, 

 are we to term this, if Mr. Lydekker's system is to endure ? The 

 context of his paper suggests coracoido-metacoracoid as a likely 

 terra ; but before that could be introduced it ought to be shown 

 that the single ' coracoid ' of living Lizards, which is coincident 

 in area with the conjoint coracoidal elements of Anomodonts and 

 Mammals, is the product of fusion of these. No one has yet 

 demonstrated the remotest trace of more than a single centre of 

 ossification in the Lacertilian coracoid ; while, on the other hand, 

 its double ossification in the Mammalia, in its non-abbreviated 

 form (OrnithorhifncJms), is preceded by its segmentation while still 

 cartilaginous. Mr. Lydekker's proposals might perhaps be accepted 

 were the Mammaha and AnomodontLa alone concerned. Anatomical 

 terminology, however, unlike nomenclature in systematics, must 

 needs be applicable to all classes of this or that sub-kingdom ; and 

 it has therefore to cover a very Ande range of structural variation. 

 These considerations, together with those which I have already 

 raised, appear to me fatal to the acceptation of Mr. Lydekker's 

 terms, which seem no more tenable than the application of the 

 human anatomists' term 'scapula' to the coraco-scapular 'blade- 

 bone,' into which he lapses in his final footnote on p. 174, and 

 which, on grounds of sheer priority, should be adhered to. "We are 

 dealing \vith a common {coracokUil) cartilage, which is in some 

 animals replaced by a couple of osseous elements, and in others 

 by but one. Setting aside the precoracoid and clavicle, the 

 ultimate homologies of which are by no means yet fully worked 

 out, our present requirements may be met by the retention of the 

 universal term coracoid for the entire set of structures (i.e. the 

 cartilaginous 'coracoid' bar and its derivatives), with the introduc- 

 tion of, say, the terms unicoracoidal and bicoracoidal for its diversely 

 modified types, and the retention of Cuvier's epicoracoid for its 

 anterior and Lydekker's metacoracoid for its posterior segment in 

 the latter one. Upon this basis, the sum of our knowledge of the 

 coracoid of Amphibia and Amuiota may be formulated as follows : — 



Coracoid. 



i. iinicoracoidal. Amphibia, all living Eeptilia, Aves. 

 ii. bicoracoidal. Some Anomodontia, Mammalia, Ich- 

 thyosauria and Nothosauria (?). 



A simple alternative would be the description of the common 

 coracoid as either uni- or bi-segmented ; but this, for obvious 

 reasons, would be insufficient. The arrangement which I here 

 propose admits of the retention of the human anatomists' term 

 ' coracoid process ' as all-sufficient for the requirements of the 

 systematic mammalogist, ^^-ho, except for his concern with the 

 Mouotremes, deals \vith the vexed element only in its most abbre- 

 viated and vestigial conditions. 



