HtRDS OF NOKTH AND MIDDLE AMERICA. 711 



the niiixilla iiiiiuilcly miciiuitc; niaxilliirv toinimii faintly concave or 

 nearly stniiylit, distinctly notched sul)torminally; jifonys straight or 

 very faintly convi'x, about as lonj:;- as distance from nostril to tip of 

 niaxilhi. Nostril wholly exposed, ratlu>r laro-e, lontjiitudinal, distinctlv 

 opercuhite. Rictal bristles well developed; latero-f rental plumules 

 short, erect, not bristly pointed. Winj,^ moderate, rather rounded at 

 tip; eighth, seventh and sixth primai-ies longest and nearly e(pial, 

 nintli about ecpial to fourth, tenth less than half as long as ninth. 

 Tail about as long as wing (sometimes a little shorter or somewhat 

 longer), much rounded, the rectrices rather narrow, with broadly 

 rounded tip. Tarsus long and slender, about one-third as long as 

 wing, twice as long as middle toe without claw, the acrotarsium dis- 

 tinctly scutellate; lateral toes about equal, their claws reaching to base 

 of middle claw; hallux (without claw) about as long as lateral toes 

 (without claws), its claw dec^idedh' shorter than the digit. 



(7o/(>r(ffio9i.— Above plain gray or bluish gray, the pileum some- 

 times black, or j)artly black, in adult males; tail black with more or 

 less of white on outermost rectrices; under parts white or pale gray. 

 Female duller in color than male (never with black on ])ileum), the 

 young essentially like adult females. 



Xidijicatlon. — Nest attached to branches of trees, deeph' cup-shaped, 

 compact, composed of plant fibers, spiders' webs, mosses, and other 

 soft materials the outside often "stuccoed" or ornamented with small 

 l)riglit-colored lichens. Eggs pale greenish blue or bluish whit(» 

 speckled with reddish brown. 



liatujc. — Tropical and warm-temperate portions of America, except 

 Galapagos Archipelago and greater part of West Indies; north to and 

 including Lower Austral life-zone of United States; Bahamas; Cuba. 

 (About twenty species and subspecies.) 



The following ke}' to the species and subspecies of this genus 

 includes several South American forms Avhich are closely' related to 

 one of the Mexican species, though in reality not so nuich so as has 

 very generally been supposed. 1 must frankly admit the uncertainty 

 which I feel concerning the Central American forms, not only with 

 regard to the number which should be recognized, but also as to their 

 relationship toward one another. At least five times as many speci- 

 mens as I have l)een al)le to examine ^vould be necessary to ena))le one 

 to feel any considerable confidence in the results i'ea( lied after careful 

 study; ten times as man}' specimens would, of course, be better. The 

 present treatment of these Central American forms is, therefore, to 

 be considered as tentative. Of only one thing do I feel (piite certain: 

 That is, that I believe, more strongly than ever, that those authors 

 are mistaken who contend that /■*. >il</r!(i'j)s Baird occurs in any part 

 of South America. 1 have been unable to trace it southward of the 

 State of Oaxaea. in southern Mexico,* (>xcept foi" a single spt>cimen in 



