PICIDzi: — THE WOODPECKERS — PICOIDES. 357 



Upper tail coverts sometimes tipped with white. Top of tlie head spotted witli white ; 

 the crown of the male with a yellow patch. Nasal bristles black, mixed with <;ray. 



Female, with the whole top ofliead spotted with white ; very rarely entirely black. 



Hab, Arctic regions of North America. 



Var. dorsulis. The black transverse bands on the emls of the central dorsal feathers 

 wanting, partly or entirely, leaving the white more or less continuous as in Picas vlllusns. 

 Ilah. Rocky Mountains of the United States and British America. 



This species varies considerably in its markings, especially in tlie amonnt 

 of white above. The head is sometimes niore coarsely spotted with white 

 than in the average ; very rarely are the white spots wanting, leaving mere- 

 ly the broad malar- and interrupted post-ocnlar stripe. The rictal black 

 stripe is sometimes mnch obscured by Avhite. In typical specimens i'rum 

 the Hudson Bay and Labrador provinces, which seem to Ijc darkest, the 

 featliers of the centre of the back have three transverse bars of white (one of 

 them terminal), and rather narrower tlian the intermediate Ijlack Ijars ; the 

 basal white one disaj^pearing both anteriorly and posteriorly, leaving but two. 

 In specimens from the Mackenzie Iliver district, there is a greater develop- 

 ment of white, the white bands being broader than the black, and some- 

 times extending along the shafts so as to reduce the black bars to pairs of 

 spots. Tlie next step is the disappearance of these spots on one side or the 

 other of the feather, or on both, leaving the end entirely white, especially 

 anteriorly, where tlie back may have a longitudinal stripe of white as in 

 ricus villosus. Usually, however, in this extreme the npper tail coverts re- 

 main lianded transversely. 



In all the specimens from the Eocky Mountains of the United States, 

 especially Lai-araie Peak, tliis \\hite back unbarred, except on the rump, is a 

 constant cliaracter, and added to it we have a broad nuchal patch of white 

 running into that of the Ijack, and connected with the white post ocular 

 stripe. The bands, too, on the sides are less distinct. It was to this state 

 of plumage that I applied the name of P. dorsalis, in 1858, and although, in 

 view of the connecting links, it may not be entitled to consideration as a 

 distinct species, it yet ajspears to be a well-marked geographical variety. 



This same character prevails in all the Rocky Mountain specimens (from 

 more northern regions) before me, including those from Fort Liard, and in 

 only one not found in that region, named No. 49,905, collected at Nulato liy 

 Mr. Dall. Here the middle of the Ijack is very white, although the nitchal 

 band is less distinct. Other specimens from that locality and the Yukon 

 River generally, as also from Kodiak, show the transverse bars distinctly. 



In one specimen, 29,ll'(), from tlie ^Mackenzie River, all the upper tail 

 coverts are banded decidedly \vith white, and the wing coverts spotted with 

 the same. Even the central tail feathers show white scollops. The back is, 

 however, very distinctly banded transversely, not longitudinally. 



