376 ANNALS OF THE CaRNEGIE MuSEUM. 



surface, a vertical line is raised where the clavicles are joined below. 

 The free extremities above are drawn out to some extent and pointed. 

 Each has on its external face an antero-posterior elongated facet for 

 articulation with the corresponding coracoid. The apices of the free 

 extremities are 2 cms. apart; and from either apex of a free end to 

 the median lowermost point of the arch equals 3.4 centimeters, more 

 or less. 



For its entire length a scapula is very much compressed vertically, 

 and about a centimeter or less of its free posterior extremity is 

 markedly dilated. It is uniformly curved in the vertical plane, from 

 one end to the other, while its outer border is gently concave through- 

 out its length, its inner one presenting a similar convexity — either 

 border being rounded off rather than sharp. Distally, the bone is 

 blunt-pointed, while the vertically flattened head, supporting the usual 

 articular facets, presents anteriorly a slightly thickened transverse 

 edge of margin for articulation with the coracoid of the same side. 

 The extreme length of a scapula averages about four centimeters, 

 with a width of three millimeters.^*^ 



The Pectoral Limb (Plates XXI V-XX VI). —There is an unusually 

 large number of the bones of this extremity in the combined collec- 

 tions before me at this writing — that is, when we come to consider 

 the fact that they represent an extinct species, and the additional fact 

 that our National Museum has not a single skeleton of any Petrel of 

 the genus ^strelata — extinct or existing — in its entire collection of 

 skeletons. Here there are, to represent ^^. vocifcrans alone, nearly 

 50 humeri, with all the other long bones of the arm in proportion. 



It is very interesting to find that the humerus of Ai. vocifcrans 

 agrees very closely, in all of its characters, with the humerus in any 

 average Shearwater of the genus Puffinus; at the same time it departs, 

 in many respects, from the humerus of any of the smaller Petrels of 

 the genus Occanodroma with which I have compared it, and perhaps 

 from the humeri of the allied genera Buhveria, Halocyptesia, and 

 Thalassidroma, which have not been seen by me. These differences 



10 In examining these scapulae for the last time, I came across the anterior 

 half of one that evidently belonged to a much smaller bird than ALstrelata 

 vocifcrans ; there is a great probability that it may have belonged to a speci- 

 men of Puffinus parvus, a distinct species described above. It is to be found 

 in the collection belonging to the .American Museum of Natural History. 



