64 PROCEEDINGS OF THE ACADEMY OP 



bido marginatis, in pogoniis rectricum interioribus colore rufo 

 latissimo, fusco angustissimo, linea, recta segregates ; rostro fusco, 

 infra ad basin pallido. Long. 7.50-8.00, aire et caifdae 3.80-4.20, 

 tarsi .75 .85 (nunquam .90 ?), digiti 3tii .65-. 75, rostri .70-.80, 

 rostri lat. .33-.40. 



Habitat sestiv. partibns orient. Reip. Amer. Sept., hyeme Amer. 

 Centrali, prsesertim Guatemala. (Costa Rica, Lawr., Ann. Lye. 

 1868, 115.) Cuba? (Gundl., 239.) An Amer. Merid. ? 



Obs. These references and diagnosis apply exclusively to the 

 bird that breeds in the United States, entirely withdrawing in the 

 fall to winter in Central America. This is pure crinitus ; the birds 

 that summer in Mexico and elsewhere south of the United States 

 have developed other varieties (infra). During the extensive 

 migrations, its passage seems rapid and its path narrow ; for we 

 have no Antillean (except as above) nor West Mexican quotations 

 of the bona fide crinitus, and few Mexican skins are certainly re- 

 ferable to it. In passing from its winter headquarters, either it flies 

 across the Gulf, or else it hugs the eastern coast of Mexico. I have 

 yet to see typical crinitus from South America. 



Diagnostic points to be remembered are these : bill never quite 

 black ; stout and comparatively short, hardly or not equalling 

 the tarsus, which latter never (?) touches .90 ; back pure olive, 

 throat pure ash, belly, etc., pure yellow; inner secondaries and 

 upper wing coverts and outer rectrix edged with yellowish-white 

 (never rufous cf.validus), in marked contrast with rufous edgings 

 of primaries and inner webs of rectrices ; all rectrices but the 

 middle pair so nearly completely rufous on the inner webs that a 

 mere line of fuscous persists next the shaft (cf. irritabilis et 

 cooperi) ; this runs of equal breadth the whole length of the 

 feathers (cf. cinerascens) ; it is sometimes inappreciable on some 

 feathers (then about as in validus) ; none of the rectrices ever with 

 more than a trace of rufous on the outer web. 



About sixt}' specimens examined. (Mus. S. I., etc. etc.) Xearly 

 all these not of the United States are Guatemalan, are positive 

 duplicates of Pennsylvania skins, for instance, and were doubtless 

 hatched in the United States. Other Guatemalan examples and 

 man3 r Mexican skins of birds that never saw the United States, 

 represent different varieties, as follows. 



[July 2, 



