THE DIVISIBILITY OF CHARACTERS. 



217 



male of this latter species is shown in pi. 21, a juvenal male also in pi. 21. This 

 species is considerably smaller than the ring-dove. It has a quite dark reddish 

 color. The characters of the hybrid of this last-named cross, consisting of four 

 species, are shown in plate 26. It will be observed that the suratensis streaks on 

 the wings are still preserved. Indeed, they seem to be increased, but this is 

 only indirectly so, due to the strengthening of the pigmentation as a whole which 

 results from the admixture of the very dark humilis. I have two hybrids of this 

 class, and both have the streaks on the wing-feathers quite as shown in the 

 illustration. 1" 



Text-figure 8. 



Neck-mark of an adult female hybrid 

 (C) between a male Columha guinea and 

 a chequered C. lii'ia. Hatched May 19, 

 1901. Natural size. Hayashi del., Apr. 

 1903. 



This hybrid resembles both parents in 

 its form. The neck-feathers lie quite 

 smoothly and the bifurcation is not ap- 

 parent except on close cxiiiiiinaluin. The 

 feathers in the male paii-nl li:nifi loosely 

 and present a rough, li.irklcd iip|>c:i ranee. 



Plate 33, figs. 7 to I 1, luiihcr assists 

 in showing that the divided teal hers do 

 not cover so large an area as in I he male 

 parent; the divisions aic iml s.. ,\rrp, and 

 are strongest on the sides and liai'k of the 

 neck, weakest in fnml . 



In the hybrid it is possible to see 

 plainly the rows of feathers, which we 

 could scarcely discern in the male parent. 

 The skin around the eye is le.ss in extent 

 and smoother than in pure C. guinea. 

 This hybrid was mated during two sea- 

 sons, but produced no eggs. 



I next took the offspring of this last cross (i.e., the young of the male trispecific 

 hybrid x the humilis female and crossed it back witli the white ring-dove. The 

 result was the bird in this cage— the only one I have thus far obtained. It is rather 

 light in color^i and we can not yet see what the neck-mark is going to be, but one 

 can see that the bird is of generally lighter color than its father (the mother was 

 white; see table 154). As we go on with such studies, therefore, we find very 

 definitely placed before us the fact that we can readily calculate what the result 

 of crossing is going to be. We know definitely what the parents are and we can 

 calculate mathematically— although I have never done such things myself— what 

 the hybrid will be. 



'" It should be said, however, that not all of this progeny were of this sort. The records (see table 154) show 

 that 3 or 4 young (in a total of 12) died very early and that all of these were recorded as of "white" color.— Editor. 

 " This bird seems not to have been figured.— Editor. 



