6 PROCEEDINGS OF THE NATIONAL MUSEUM vol 7 J 



plainly colored species luteoviridis, signatus, flaveolus, and richard- 

 soni in Ba^'tileutenis. The forms with a short and bicolored tail, as 

 well as those with a long slender tail, must both be removed before 

 any pertinent diagnosis of the genus can be framed. Even with 

 these changes made, there is still much diversity within the group. 

 When we consider the size of this particular group, and remember 

 that after all a genus is a more or less arbitrary assemblage of forms, 

 and is not necessarily sharply cut off from other such groups, to 

 several of which it may approximate, this is what might be expected. 



PHYLOGENY 



Basilenterus hivltfatus may be taken as the central or most gener- 

 alized form of the genus. It is related on the one hand to such species 

 as B. -flaveohis, B. luteoviridis^ etc., which have the under parts yellow, 

 but the pileum plain, and which lead off in the direction of Mi/io- 

 thlyjns. On the other hand it seems to be allied to B. grlseiceps, 

 which stands as a connectant between the plainly colored forms typi- 

 fied by B. rivularis and those with olive green back, yellow under 

 parts, and heavily streaked pileum of which B. coranatus is the 

 best-known example. B. rivularis leads off through B. holivianus 

 and B. mesoleucus directly to the group which we are here character- 

 izing as Phaeothlypis^ while B. coronatus is not distantly related to 

 B. fraseH, which in its turn approximates Euthlypis. The smaller 

 species with streaked pileum {B. auricapiUiis, etc.) are obviously 

 inter-related, and run in an unbroken chain, from which B. trisfri- 

 atus is possibly an offshoot. B. hasilicus and B. belli appear to be 

 more or less isolated forms, whose relationships are possibly with 

 B. tristriatiis. B. inelanogemjs, B. ignofas, and B. hensoni constitute 

 a compact group of uncertain affinities. B. delattni also stands by 

 itself, and leads off toward Idiotes. 



To indicate these diverse and complex relationships in a " key " or 

 linear sequence is decidedly out of the question, but they may be 

 partially represented by using two dimensions. The accompanj'^ing 

 figure (fig. 1) is an attempt at such a diagrammatic representation, 

 the species alone being considered. It at once suggests how a genus, 

 in the course of its development in time, may have given rise to off- 

 shoots so distinct as to require generic separation. In other words, 

 we have here a clue to the origin of genera as such. Virtually all 

 the color characters of Myiothlypis nigrocristata are foreshadowed, 

 so to speak, in Basil^uterus luteoviridis (both species of the Tem- 

 perate Zone), and this in its turn is similarly related to B. signafu-^ 

 of the Subtropical and to B. fmyeolus of the Tropical Zone. But 

 Myiothlypis has developed structural characters of its own. by virtue 

 of which we can separate it generically from Basileuterus luteovin- 



