THE DIVISIBILITY OF CHARACTERS. 223 



are like the mother. But observation tells me not to be too confident in saying 

 that the "white" bird is the exact reproduction of the father. In the first place, it 

 is intermediate in size, and the white is the color that misleads us. When the bird 

 is examined a little more closely it will be found that the tail-feathers have some 

 dark pigment. The dark portion of the basal two-thirds of these feathers is the 

 part that is very dark — almost black — in the blond ring-dove. It is evident, there- 

 fore, that we have the old pattern of the blond ring, together with some of its color, 

 preserved here. I have not examined the other feathers of the bird microscopically 

 or otherwise to determine how much pigment there is, but certainly there is some, as 

 we can clearly see in the case of the tail. 



This case, then, does not come to Mendel's rule, but may suggest something 

 analogous to it. According to his rule the first hybrids are of one dominant color; 

 the offspring of these hybrids divide, so that one quarter are like the one pure 

 parent and another quarter like the other pure parent, and two quarters — or 

 one-half — are like the first hybrids. In the case which we have just described the 

 division into two colors is obtained in the first generation of hybrids, and the two 

 colors are found to be closely associated with sex. 



My main object in these remarks has been simply to present a few cases. The 

 number of such cases that I might place in evidence could be much increased. The 

 crosses I have made all tell the same story. 



One thing, however, I would add in conclusion. According to Galton's law 

 of ancestral inheritance, the two parents furnish, as is well known, only about 

 one-half of the offspring, while four grandparents furnish a quarter, and the great- 

 grandparents furnish one-eighth, and the sixteen great-great-grandparents would 

 furnish one-sixteenth, and so on, until the offspring is represented as the unit, 

 deriving these fractional parts from these different groups or grades of ancestors. 

 The necessity of such a mode of reckoning lies in the fact that parents represent 

 "mixtures." They are not "pure" to begin with. If a pure species is crossed it 

 can be said that each parent contributes one-half, with nothing to do with other 

 ancestors. That is, assuming that one has a pure species of birds — ^i3ure from one 

 generation to another — one knows what he is going to get, namely, one-half of 

 each of these species in each of the offspring. It is a very much simpler method" 

 than Mendel's or Galton's. 



" In tlie prefatory notes to this leoture Professor Whitman characterized his law, as elucidated in this chapter 

 as that of "Pure or direct inheritance." — Editor. 



