BIRDS OB' NORTH AND MIDDLE AMERICA. 499 



KEY TO THE SPECIES AND .SUBSI'Et^IES OF DENDKOICA. 



a. Inner webs of rectrices (except middle pair) partly yellow. {^^ (Inlden Warblers.^ ^^) 

 1). No orange, rufons, or chestnut on sides of head or throat, or else the pileum also 

 rufous or t-hestnut and chest and sides streaked with the same. 

 c. Tnder parts bright yellow, the chest and sides streaked with chestnut or 

 rufous. (Adult males.) 

 (/. Chin, throat, and sides of head yellow. 

 e. Wing more pointed, the outermost (ninth) primary equal to or longer 

 than sixth, often longest; tarsus shorter, never more than 20, averaging 

 about 18.8; tail relatively shorter, averaging less than 48.5, or else wing 

 averaging 68; continental. (Dendrolca irstiva.) 



g. Larger (wing averaging more than 62, tail averaging more than 44) 



and brighter colored; pileum, in fully adult plumage, decidedly 



yellowish, often pure yellow, sometimes tinged with tawny orange; 



wing-coverts and tertials broadly edged with yellow; back, etc., 



lighter, more yellowish, olive-green. 



//. Back, etc., more decidedly olive-green, the upper tail-coverts with 



less yellow; chestnut streaks on chest and sides much broader.'^ 



(United States in general, except southern border from western 



Texas to Arizona; more southern British Provinces.) 



Dendroica aestiva aestiva, adult male (j). 508) 

 ///(. Back, etc., more yellowish olive-green, the upper tail-coverts 

 with more yellow; chestnut streaks on chest and fiank much 

 narrower. 



' The so-called " Golden Warblers " (see Baird, Keview of American Birds, p. 193) 

 end)race numerous forms which at first sight seem to be easily arrangeable into three 

 groups according to the color of the head in adult males: (1 ) Those without a sharply 

 defined orange, rufous, or chestnut patch on the crown; (2) those with a sharply 

 defined crown-patch of rufous or chestnut, and (3) those with the whole head, 

 including the throat, rufous or chestnut. A careful examination, however, shows 

 that no sharp line can be drawn between su])iiosed groups 1 and 2, one form 

 (D. aureola, of the Galapagos Archipelago and Cocos Island) being so clearly inter- 

 mediate that different specimens would fall into either group. In short, there is, in 

 various West Indian forms, including that from the island of Cozumel, every inter- 

 mediate condition between the yellow or olive-green crown of D. wstiva (in which, 

 itself, the crown is sometimes slightly tinged with orange) and the very sharply 

 defined dark chestnut cap of I), capitalin. Furthermore, I fail to find other color 

 characters which will serve to segregate the various forms into smaller groups, not 

 a single one of those which have hitherto been used for that purpose and which at 

 first sight give promise of utility in that way standing the test of careful examina- 

 tion of even a moderately large series of specimens. With adult females the case is 

 still more difficult, there being in that sex a far greater range of individual variation 

 and the color characters altogether less pronounced. The preparation of an infallil >le 

 ' ' key ' ' to the various forms of these ' ' Golden Warblers ' ' the author therefore acknowl- 

 edges to be beyond his ability, and he wishes it to l>e understood that the one here 

 given, while rendering identification easy in the case of most specimens, may fail with 

 others. 



It is clearly evident that all these "Golden Warblers," with the exception, jier- 

 haps, of the little-known D. eoa, are of common origin, and that many of them rep- 

 resent merely local forms or slightly differentiated subspecies; but where todraw^the 

 line between those which seem to be now S[)ecifically distinct and those which do not 

 isa very difficult matter, regarding which i)r()bably no two authors would entirely agree. 



'^ Except in some western specimens, which otherwise are like eastern ones. 



