There is no reason whatever for separating the American species from those of the Old World, and the subgenus Pœcilornis, established upon these by Kaup, is not tenable.
Since the publication of my first paper upon the American forms of Tinnunculus,[59] a large amount of additional material has fallen under my observation; the total number of examples critically examined and compared together amounting to over three hundred and fifty skins of which I have kept a record, besides many others which have come casually to my notice. This abundant material merely confirms the views I first expressed, in the paper alluded to, regarding the number and definition of the forms; their comparative relation to each other being the only respect in which I have reason to modify my arrangement.
In my first paper on the American Tinnunculi, three distinct species were recognized; one (sparverius) belonging to the whole of Continental America and the Lesser Antilles, one (leucophrys, Ridgway) to Cuba and Hayti, and one (sparveroides, Vig.) peculiar to Cuba. The first is one modified in different climatic regions into several geographical races, as follows: Var. sparverius, L., North and Middle America, exclusive of the gulf and Caribbean coast region
; var. isabellinus, Swains., the eastern coast region of Tropical America, from Guiana to Florida; var. dominicensis, Gmel. (Lesser Antilles); var. australis, Ridgw. (South America in general); and var. cinnamominus, Swains. (Chile and Western Brazil). That each of these races is well characterized, the evidence of a series abundantly sufficient to determine this point enables me to assert without reserve; for I find in each instance that the characters diagnosed in my synopsis hold good as well with a large series as with a few specimens.
The following synopsis, essentially the same as that before published, may, to most persons, explain satisfactorily my reasons for recognizing so many races of T. sparverius,—a proceeding which, I am sorry to say, does not meet with favor with all ornithologists.[60] Though there are at the present time three well-characterized or permanently differentiated species of Tinnunculus on the American continent, yet it is, to my mind, certain that these have all descended from a common ancestral stock, for evidence in proof of this is found in many specimens which I consider at least strongly “suggestive” of this fact; some specimens of var. isabellinus from Florida having blue feathers interspersed over the rump, thereby showing an approximation toward the uniformly blue upper surface of the adult male of T. sparveroides of the neighboring island of Cuba; while in the latter bird the embryonic plumage of the male is very similar to the permanent condition of the male of sparverius.
Synopsis of the American Species.
A. Back always entirely rufous (with or without black bars.) Lower parts white, or only tinged with ochraceous; front and auriculars distinctly whitish.
a. Inner webs of primaries barred entirely across, with white and dusky; “mustache” across the cheeks conspicuous; no conspicuous superciliary stripe of white.
1. T. sparverius.[61] Crown bluish, with or without a patch of rufous. ♂. Wings and upper part of head slaty, or ashy-blue; scapulars, back, rump, and tail reddish-rufous; primaries, basal half of the secondaries, and a broad subterminal zone across the tail, black. ♀. The bluish, except that of the head, replaced by rufous, which is everywhere barred with blackish, and of a less reddish cast. Hab. Entire continent of America, also Lesser Antilles, north to St. Thomas.