1885.] MISS B. LINDSAY ON TBE AVIAN STERNUM. 705 



former view rather than the latter. Even in Camarius it is only 

 occasional {cf. tig. 8). 



The same uncertainty must exist with regard to the precoracoid of 

 Diomedea (fig. 1, Plate XLIV.) ; although there can be little doubt, 

 from comparison with the early stages of the various embryos 

 figured, that we must agree with Sabatier in regarding this region as 

 the Avian precoracoid, rather than the precoracoid of Parker, which 

 Sabatier calls an epiphysis {cor. ep., fig. 1). 



Under these circumstances it would be hazardous, notwithstanding 

 the Struthionic aspect of the coracoid in the Gull, to put forih with 

 any certainty a view that the pre-Avian kind of precoracoid is to be 

 identified with the region suggested as precoracoid in the Chick ; 

 but at the same time the development of this marginal region in the 

 types figured renders it extremely likely that the region already 

 described in the shoulder-girdle of the late five days' Chick is 

 equivalent to a precoracoid of some kind, though possibly of a 

 secondary character, developed in the more immediate ancestors of 

 the Gallinaceae. 



(v.) The Gannet. 



(9 specimens, from the bird just hatched to a stage comparable 

 with a 5 days' Chick.^ 



The sternum of the adult Gannet, it may be observed, since it 

 possesses a furcula continuous with the keel, and behind it a 

 median paired apophysis for the support of ligaments, exhibits at 

 once the iuterclavicle of Gotte in front, and the inlerclavicle of 

 Harting at a more dorsal level. 



In this bird, if in any, we might expect to trace a genetic con- 

 nection between clavicle and keel, since in the adult these parts are 

 fused. Such, however, is not the case. 



The embryo is particularly easy to examine, because it is so large 

 that the earliest aggregations in the mesoblast can be clearly traced. 

 The following were the stages observed, recorded, for the sake of 

 convenience, in inverse order, youngest first. 



1«; Stage, corresponding loith a five days' Chick. — There are 8 

 ribs, which end freely, and no trace of sternum. The three parts of 

 the shoulder-girdle are separate and almost parallel, overlying a 

 remarkably thick mass of tissue representing the pectoral muscle. 



2nd Stage, corresponding with a later five days' Chicle. — The 

 sternum is indicated by thick opaque aggregations in a layer of 

 mesoblast, shown in fig. 14, Plate XL V. A large mass indicates the 

 sternum, a smaller one the keel, while three smaller masses, which 

 in the next stage have disappeared, may probably be compared with 

 the three anterior muscle-bands previously described in the Chick. 

 The end of the clavicle, which is of course superficial, lying next to 

 the skin, is completel)' shut off from the deeper layer, in which lie the 

 differentiating sternum and keel, by the whole thickness of the 

 pectoral muscle, a depth nearly half the length of the clavicle 

 itself. 



^rd Stage, corresponding with a six days' Chick. — This shows 



