230 



THE AMEBIC AN NATURALIST [Vol. XLIV 



found in some two of the species with which they grow. 

 These odd forms, of which I have noted nearly a hundred, 

 are, in my opinion, largely the result of hybridism in the 

 remote or recent past. The detailed evidence of this can 

 not be presented in a twenty-minute paper ; it demands a 

 printed essay with abundant illustration. I shall here 

 attempt only a brief summary of methods and results. 



I would first call attention to the transitory existence of 

 a hybrid, especially when self-fertilized, as is usually the 

 case in Viola. This follows from the laws of Mendel as a 

 mathematical corollary, though I am not aware of any 

 paper in which these deductions have been clearly set 

 forth. Mendel, indeed, showed that in the case of the 

 monohybrid (where the parents differ in only one char- 

 acter) the offspring in the first generation are one half 

 stable, in the second three fourths stable, in the third 

 seven eighths, and so on; until in the tenth generation 

 there is on the average only one hybrid in 1,024 plants. 

 But it can be proved that where the parents differ in 

 two characters there will be on the average in the tenth 

 generation only two hybrids in 1,024 plants; where the 

 parents differ in three characters, there will be less than 

 three hybrids in 1,000 plants: where they differ in four 

 characters, less than four hybrids in 1,000 ; and so on. In 

 the ordinary hybrid over 50 per cent of the offspring in 

 the fourth generation will be stable ; in the sixth genera- 

 tion over 85 per cent will be stable ; in the eighth genera- 

 tion over 95 per cent; in the tenth generation over 99 

 per cent. The life of the self-fertilized hybrid is there- 

 fore always precarious and relatively brief. The organ- 

 ism is in unstable equilibrium, and is rapidly shifting into 

 conditions of stability. 



But the stable forms that emerge are not, as a rule, the 

 forms found in the parents of the original hybrid. They 

 are exclusively such only in the case of the monohybrid ; 

 in hybrids of a higher grade the stable forms consist of all 

 possible recombinations of the opposed characters of 

 parent species. Two species that differ in 



of the 



