444 



U. S. p. E. E. EXP. AND SURVEYS — ZOOLOGY — GENEEAL REPORT. 



List of specimens. 



* These specimens, in autumnal plumage, scarcely ailmit of an accurate determination. 



PASSEECULUS SANDWICHENSIS, B a i r d . 



Emberiza sandwichensis, Gm. I, 1788, 875. 



Emberiza arclica, Latham, Ind. Orn. I, 1790, 414. 



Fringilla arctica, Vigors, Zool. of Blossom, 1839, 20, (perliaps one of the smaller species.) — " Brandt, Icon. Ross. 



2,6." 

 Euspiza arctica, Bp. Conspectus, 1850, 469. 



Emberiza chrysops, Pallas, Zoog. Rosso-As. II, 1811, 45 ; tab. xlviii ; fig. 1, (Unalasclika ) 

 Sandicich Bunting, Lath. Syn. II, 1783, 202. 

 Unalascha Bunting, Pennant, Arctic Zool. II, 363, 320, No. 229. (Not of p. 364, No. 233.) 



Sp. Ch. Almost exactly like P. savanna, but half an inch larger, witli niucli larger bill. Length, 6 12 inches : wing, 3.00 ; 



tail, 2 55. 

 jjab. — Northwestern coast from the Columl)ia river to Russian America. 



This species is extremely similar to the F. savanna, and is only distinguishable by its greater 

 size, and more western locality. The tail feathers also are rather more acutely pointed. There 

 is also a greenish yellow shade on the top and sides of tlie head, brighter than that seen in P. 

 savanna. The bill is considerably larger and longer, measuring .51 of an inch above instead 

 of .44. 



The Sandwich Bunting of Latham (Synopsis of birds) and the Unalascha Bunting of 

 Pennant (page 303, No. 229) seem to belong unquestionably to this species, and as Gmelin 

 bases his sandwichensis upon these descriptions, it must be retained. The name has no refer- 

 ence to the Sandwich Islands but to Sandwich Sound, on the northwest coast. The Unalascha 

 Bunting of Pennant, page 304, No. 233, is a diU'eront species. 



The "temporibus atria" is not a very accurate expression in the species, but sufficiently near 

 not to be inconsistent with it. At any rate, as the descriptions of sandivichensis, arctica, and 

 chrysops, all seem to apply equally well, it will be best to take the oldest name as a provisional 

 one at least. 



