1887.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 77 



^strelata sand'wichensis Eidgw. 



?Uan. 



1869.—? ProciUnria alba Dole, Proc. Boston See. N. H., XII, 1869, p. 308, Extr. p. 15 



(nee Gmel. ?). — Id., Hawaiian Almanac, 1879, p. 55. 

 1884. — GEstrelata sandwichensis Ridgway, Water B. N. Am., II, p. 395. — ^strelata e. 



Id., Proc. U. S. Nat. Mus., IX, 1886, p. 95. 



In the great work on the Water Birds of North America Mr. E. Eidg- 

 \ray writes as follows (II, pp. 394-395) : 



"A specimen from the Sandwich Islands (Xo. 61259, Y. Kmidsen, 

 coll.), labeled ' Pufflmis rneridionaUs,^ differs from the above diagnosis 

 [of ^. luesitata] in several particulars, and may possibly be distinct. 

 The entire upper parts, except forehead, are continuously uniform dusky, 

 nearly black on the head, the nape, the back and scapulars more gray- 

 ish brown ; this dark color even covers uniformly the entire side of 

 the head and neck, except that portion of the former before the eye, 

 and thence downward and backward across the malar region. The 

 feathers of the nape and side of the neck, however, are white imme- 

 diately beneath the surface, this color showing conspicuously wher- 

 ever the feathers may be disturbed. There is likewise no exposed white 

 on the upper tail-coverts or base of the tail 5 the former are, however, 

 very abruptly white beneath the surface, but the latter is white only 

 at the extreme base, and the outer rectrices have a considerable amount 

 of white on their inner webs. The lower parts are almost entirely white, 

 there being merely a few plumbeous irregular bars on the flanks. The 

 measurements are as follows : Wing, 11.80 inches (less than the aver- 

 age of (E. hwsitata as given by Dr. Coues) ; tail, 5.75 ; its graduation, 2.40 ; 

 culmen, 1.22; depth of bill at base, .99; tarsus, 1.40; middle toe (without 

 claw), 1.55. In view of the difierences of coloration, much more gradu- 

 ated tail, and smaller dimensions — and especially in view of its different 

 habitat, no specimens of (E. hcesitata having to our knowledge been re- 

 ported from any part of the Pacific Ocean — the specimen in question 

 may be really distinct. Should such prove to be the case, the name 

 (E. sandicichensis is proposed as a suitable designation." And in a foot- 

 note he adds: ''In pattern of coloration this specimen agrees exactly 

 with an example of Q^. cooM, but has the back, scapulars, rump, and 

 tail decidedly less ashy." 



Aft«r having had an opportunity to compare Kuudsen's bird with ex. 

 amples of true JE. h(esitata, and also with the type of Lawrence's 

 JE. meridionalis, the same author afterwards (Pr. U. S. Kat. Mus., IX, 

 1886, p. 96) pronounces the opinion that they are entirely distinct from 

 JE. sanduiichensis, but has " a suspicion that the latter is the same as 

 JE. pha^opygia Salv. (Trans. Zool. Soc. Lond., Vol. IX, part ix, May, 

 1876, p. 507, pi. 88, fig. 1), from the Galapagos." 



This point, however, can only be determined by direct comparison of 

 the types, and until then we prefer to retain the name which belongs 

 strictly to the Hawaiian specimens. 



Latham's "White-breasted Petrel " (Gen. Syn., Ill, ii, p. 400) "from 

 Turtle and Christmas Islands," upon which Gmelin based his Procellaria 



