276 Zoologica : N. Y. Zoological Society. [I ; 15 



To careless examination and description is due much of the 

 confusion existing among the birds of this genus. Instead of 

 having twenty tail feathers, the majority of the leucurum-\\k.e. 

 individuals have twenty-two. The male type possesses twenty 

 at present, but a glance at the skin surface over the pygostyle 

 shows two gaping holes from which the other pair has been 

 pulled, while the female type still has twenty-one, the outer left 

 rectrice having been lost. 



In the latter bird the rectrice white is present only as a 

 small spot on the outer tail-feathers, but increases rapidly until 

 it dominates the next to the central pair. In another specimen 

 the reverse arrangement is found. In two leucurum-\ike birds 

 with very little white, there are only twenty as the full comple- 

 ment of tail feathers. And so, to be logical we must stretch 

 the description of harmoni and of leucui^um to include a multi- 

 tude of variations, or make new species of almost every indi- 

 vidual, or, to follow the only reasonable course of action, both 

 must be sunk as synonyms of one or the other species. Hybrid- 

 ism seems the only explanation of such variation and asymmetry. 



The variation in number of tail feathers in Crossoptilon is 

 of interest. In tibetanum there are twenty, and all with vanes 

 quite solid and normal in structure. The whiter leuciiTumA\ke 

 birds show no change, but in most of those individuals which 

 approach harmoni and auritwni, an extra central pair of highly 

 disintegrated feathers appears, above the others, suggesting, 

 from their position, derivation from the upper tail-coverts. In 

 auritum still another pair is present, making twenty-four in all, 

 this additional pair also being central, superior, and much dis- 

 integrated. So the specialization is definite as to position. In 

 more than one bird, which in color from beak to tail is typical 

 auritum, I have found after careful macro- and microscopical 

 examination only twenty-two rectrices ; none having been lost 

 accidentally, but one of the central pairs being congenitally ab- 

 sent. These birds were unquestionably hybrids with the les- 

 sened number of rectrices as the sole indication of their mixed 

 blood. 



In the rather isolated, more generalized, brown mantchuri- 

 cum, twenty-two is the normal number, and here we find but a 



