Shufeldt: Birds from Bermuda. 373 



easily detached from it. Along the sides of the latter we are to note 

 that, in the case of the leading seven (7) vertebrae, the exit for the 

 sacral ncrz'cs consists, in any case, of small, twin foramina, one being 

 placed immediately above the other. We also find this in the Shear- 

 waters and many other birds, both land and water species. The crest 

 of the sacrum of the bone here being considered is well developed and 

 extends from the leading vertebra, backwards, to the middle of the 

 bone, where it is lost upon the general surface. Its superior margin 

 is thickened by the edge being capped by an osseous dilation of nearly 

 a millimeter in width. This projects well beyond the crest anteriorly, 

 overhanging the prezygapophyses of the first sacral vertebra. (Com- 

 pare the figures of sacra upon Plates XXVI and XXVII.) 



Passing to the ///;/;n of an " os innominatum," I find the preacetabular 

 part to have an average length of about 2 cms. Its edges are sharp 

 and round ; the smooth surfaces of its dorsal and ventral aspects 

 being concave in the first and correspondingly convex in the latter 

 instance. Dorsad, the preacetabular part of the ilium is of limited 

 extent, and the aforesaid concave and convex surfaces are reversed. 

 The rounded antitrochanier faces directly downwards and forwards, 

 occupying its usual site. Hardly any osseous base is to be found in 

 the circular cotyloid cavity, and it is divided from the large elliptical 

 ischiadic foramen by a very narrow isthmus of bone. There is no 

 evidence of any prepubic process, while the rather small obdurator 

 foramen is almost entirely merged into the " obturator space," which 

 is here very extensive. On its lateral aspect, the ischium is quite flat 

 and entirely smooth. Posteriorly it is produced backwards and down- 

 wards as a delicate and curved process of considerable length, which 

 finally terminates in a little foot-like expansion that articulates by its 

 entire lower edge with the superior margin of the narrow, slender, 

 and elongate postpubis at just a short distance anterior to its free 

 posterior extremity (fig. 6t,, PI. XXVI). There is no true " ilio- 

 ischiadic notch " present in this pelvis, as seen in many other birds, 

 for example, in Pnffinus boreal is and other Shearwaters. This 

 ■' notch " is also to be seen in some of the Fulmars, as Fulmaris 

 glacialis, but it is more open than it is in the Shearwaters. (Com- 

 pare figs. 17, 21, and 25.) 



But four or five caudal vcrtcbrcc were discovered of ^strelata 

 vocifcrans, and they appear to be from the middle of the series of 



