No"MKN('i.ArT'i;i: ok ini: Ampmimiav Siiotldku ( iiiiOM:. ]6r> 



cepliiilians tlirougli tlie various reptilian groups and on to tli(; inaininals 

 ami birds, hut we always find it in front of the pectoralis. Fuither, 

 we know the evolution of the preeoracoid \ ery fully from tlie Labyrin- 

 thodonts up to the Monotrenies, but in no case do we ever find the 

 clavicle developing in conjiection with the preeoracoid, and even when 

 the preeoracoid disappeai's, as in most leptilian groups, the clavicle is 

 quite unaffected and continues to perform its own function. And fiom 

 what we know of the Stegocephalian clavicle, we see that it was in 

 no way intimateh^ connected with the preeoracoid. In the frog the 

 ossification usually calletl clavicle is so intimately co)niect(Ml with the 

 preeoracoid that it cannot be regardevl as a superficial ossification, and 

 if we begin to look on superficial ossifications as distinct elements we 

 will find ourselves in hopeless confusion. For example, the human 

 humerus and other long bones ossify first in the inner part of the 

 shaft, but in marsupials the same bones connnence to ossify super- 

 ficially. In juarsnpials exen the arches of the vertebrae commence 

 to ossify as exostoses. All the evidence we have at present seems to 

 me but the more fully to confirm Parker's view that the whole element, 

 cartilaginous and bony, should be regarded as preeoracoid. If it be 

 asked why does the eoracoid ossify as an endostosis and the precciracoid 

 as an exostosis, one might suggest that the Urodele ancestor of the 

 frog had onl}^ the eoracoid ossified, and that the precoi/aeoid only 

 became ossified in connection with the peculiar Anuran speeialisrtti<m, 

 and as it came to have a somewhat similar function to that of the lost 

 clavicle the supei-ficial ossification was more serviceable than an endos- 

 tosis would have been. 



The only text-book I have seen in which the views here advocated 

 are maintained is the recent edition oi Zittel's Pahfonto/oy)/, where 

 the eminent American paheontologist Case is responsible for the chapter 

 on Amphibia. Case calls the whole of the anterior median element 

 omos(t^r)iHtn, and the wdiole anterior bar, preeoracoid. Moreover, in 

 the figure he gives of the frog's shoulder girdle he specially calls atten- 

 tion to the fact that the cartilaginous inner part of the eoracoifl is also 

 coracoifl, in case any one might be misled into calling it epieoi'acoid. 



Another element in the frog's arch whose alfinit}' is not <iuite 

 settled is the so-called sternum and xiphisternum. The bony part is 

 usually spoken of as sternum and the expanded cartilaginous portion 

 as xiphisternum. Here again one might discuss the advisability of 

 giving two names to what may be only one element, but, leaving this 

 for the moment, there is a much more serious point. In mannnals, 

 reptiles and birds the sternum seems to <levelop in connection with the 

 tips of the ribs of the two sides, and as the two halves come together 

 in front of the heart we get the median sternum. Now in the frog 

 and other Anura the ribs are absent or rudimentary, and they do not 

 appear to have anything to do with the supposed sternum. So that if 

 the manniialian sternum is really a de\el<^pment from the ribs, the 

 amphibian " sternum " must be something different. Gadow suggests 

 that the term metasternnm should be used for the frog's sternum : but 

 until we know for certain that the element is )iot the homoliigue of 



