198 . IMIKKII AM K. FKKTI1.ITY. AM) SKX IN I'KiKONS. 



color' (Nonioncl. Colors, pi. Ill, No. 23), while the Indian bird has the back duller and 

 darker, or like his 'drab' (pi. Ill, No. 18). The latter, which is the true T. douraca, seems 

 also to be somewhat smaller." 



This difference in color seems to me of doubtful value as a basis for distinguishing 

 species, since in the ring-doves, as in the turtle-doves proper, and indeed in most if 

 not all species of pigeons, we meet with a similar difference the "lighter" and the 

 "darker" shades often quite marked. 



The difference in "size," based on a single Indian bird skin, is of no value, unless 

 it be shown to be general. Salvador! (loc. cit., p. 432) gives the following measure- 

 ments of T. douraca (including Chinese and Japanese forms) : Total length 1 1.9 in. ; 

 wing 6.8 in. ; tail 5; bill 1.6; tarsus 0.9. The wing measurement is 172 mm. (6.8 in.) 

 Stejneger's specimen of T. douraca measured only 160 mm. evidently an unusually 

 small bird, or a poorly preserved skin. 36 



Stejneger points out as a striking constant distinction between T. risorius and 

 T. douraca, that the outer web of the outer tail-feathers in T. risorius is "entirely 

 white, " while in T. douraca it is "blackish. " Further, the tail is longer in the latter. 



I find this color distinction is overdrawn. Examining two female St. risoria, 

 I find that in neither case is the web "entirely white." 



Comparing these with a female Japanese ring-dove, I find that in the latter the 

 outer web starts at the base as a light pearl gray, lighter (more whitish) at the outer 

 edge, becomes gradually darker gray until, at about the middle, the blackish appears 

 (the edge continuing whitish) and continues for about 32 mm., or for about the 

 third quarter of the entire feather-length, lightening up into paler and paler gray 

 towards the tip. 



In the first St. risoria female I find as follows : The outer web starts at the base 

 as a light pearl-gray lighter at the outer edge and darkens into a clear gray, 

 which for the first third of the length is not distinguishable from that of the above 

 Japanese bird; but towards the middle the gray becomes paler (just where the 

 blackish prevails in the Japanese) and becomes pure white only in the terminal fifth. 



The second St. risoria differs from the first in having the gray a shade paler and 

 the light edge wider. But the edge is not pure white, and it shades into the gray of 

 the mesial half. The real distinction, then, is that the Japanese ring has black in a 

 portion of the outer web i.e., has a darker web as a whole 36 but the inner web 

 also has a larger and deeper extent of black. This greater amount of black in the 

 tail is correlated with the darker color of the Japanese bird as a whole. 



The midrib, curiously, is decidedly darker in St. risoria than in the Japanese 

 bird. The tail-feathers are from 12 mm. to 18 mm. longer in the Japanese birds 

 than in the blond rings. 



In all three of the birds just compared the black is stronger and more sharply 

 limited in the under side of the feathers than above. In the Japanese ring-dove it 

 reaches nearer to the tip than in the blond rings, and is continued on the outer web 

 beyond its limit on the inner web by at least 12 mm. The wing of the above- 

 mentioned Japanese ring-dove measured 162 mm. 



"Schickel (Inc. cit.) points nut (lint. T . rixoriux differs from T ilnnriicn of India in having: (1) a shorter tail; 

 (2) the outer wel> of tin- outermost tail feathers white ; (3) a very decidedly different voice. 



" Salvador! (p. 432) says of T. douraca: "Lateral tail-feathers leaden grey, fading gradually into white towards 

 the ti|is." Notice that he says nothing of "blackish." 



