476 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



This variation extends to all parts of the external anatomy, and is by 

 no means equally correlated, as specimens having the wing or tail of 

 average length, or even unusually lengthened, may have the bill or the 

 tarsus unusually small, and vice versa. The tarsus, in forty-eight speci- 

 mens, varies from .60 to .80 of an inch, — a variation amounting to nearly 

 one-third of the mean length. The bifurcation of the end of the tail varies 

 even more remarkably, the depth of the fork ranging from .20 to .85 of 

 an inch in specimens having the feathers of this member fully developed 

 and otherwise normal ! 



In the series under examination there are a few specimens more or 

 less noticeable on account of deviations from the usual coloration in one 

 respect or another. Xo. 1G710, from the Amazon (Lieut. Herndon), has 

 the crown-patch clear yellow instead of orange-red, while the wings are 

 almost devoid of the usual light edgings. The plumage, however, of this 

 specimen is much abraded. Specimen No. 39900, from the headwaters 

 of the Huallaga Eiver, Eastern Peru (W. S. Church), is one of the 

 darkest in the entire series. It agrees almost exactly in colors with 

 ~No. 55701 from Conchitas, Buenos Ayres ; but the sides of the breast are 

 dark greenish slate, quite as dark as the color of the back, in very 

 marked contrast to the much paler yellowish olive of the central portion 

 of the breast. No. 3795G, from Merida, Yucatan, an adult male, has the 

 orange-red crown-patch surrounded by a strong suffusion of olive green, 

 like the color of the back. A very highly colored specimen from Costa 

 Itica (Xo. 33392, J. Carmiol) has the two longer lower tail-coverts chiefly 

 dusky, with wide borders of pale yellow. 



Autumnal specimens of couchi have the conspicuous paler edgings to 

 the wing-feathers strongly suffused with sulphur-yellow, and the back 

 more decidedly green than in summer examples. The single young 

 example of this Northern race (X<>. 58849, <? , Tehuantepec, June, 1809; 

 Prof. Sumichrast) differs conspicuously from four individuals of the 

 same age from Labia, and one from Costa Rica (the latter being exactly 

 like the former), in the borders of the wing-coverts being pale sulphur- 

 yellowish instead of light cinnamon-rusty, and in the more creamy yellow 

 of the lower parts. It is perhaps doubtful, however, whether other 

 examples from Mexico would not agree more closely with Southern 

 ones. 



The dimensions vary in this species not only with the individual, but 

 also to a very considerable extent with the locality. Thus, dividing the 

 large series before me into groups representing the several zoo-geograph- 

 ical provinces into which Tropical America is divisible, and taking the 

 average of the several measurements of each, the following is found to 

 be the result : 



