ABOUT HYBRIDS. 



227 



the ratio of one white, two hybrids, one blue. 

 . This ratio is a constant for all cases that 

 have so far been put to the test. 



Mendel chose twoj)airs of dififcrentiating: 

 characters for a second experiment. One 

 made by DeVries may be cited by way of 

 illustration. DeV^ries crossed a blue thorny 

 form, Datura Tatula, with the white thorn- 

 less Datura Stramonium Inermis. The 

 members oi the first generation were blue 

 and thorny, so that these were evidently 

 dominant characters. The second genera- 

 tion was broken up into the following 

 classes : ( i ) 56.25 per cent, blue with 

 thorns; (2) 18.75 P^"^ cent, blue without 

 thorns; (3) 18.75 P^^ cent, white with 

 thorns ; (4) 6.25 per cent, white without 

 thorns. This is a proportion of 9:3:3:1. 

 Out of the 56.25 per cent, constituting class 

 I, 6.25 per cent, bred true, and 55 per cent, 

 ■were unstable hybrids. Of the 18.75 P^r 

 cent, of class 2, 6.25 per cent, continued to 

 breed 'blue and thornless plants, and the rest 

 were unstable hybrids. In class 3, 6.25 per 

 cent, bred true, and in class 4 all bred true. 

 It will be seen that two groups reverted to 

 the original types, two new types appeared, 

 and the rest were hybrids, the offspring of 

 which again split in a definite manner. 



With a greater number of pairs of differ- 

 entiating characters the difificulties of the 

 experiment increase and the statistics be- 

 come much more complicated, though the 

 law applies to these complex cases just as it 

 does to the simpler. 



The explanation of this law is embodied 

 in the hypothesis of germinal purity, that is 

 that every germ is pure with regard to any 

 particular character. Thus the hybrid be- 

 tween the blue flowered and the white flow- 

 ered types of the first example is supposed 

 to produce two kinds of pollen grains and 

 two kinds of eggs, namely, those that are of 

 the blue-flowered type and thise that are of 

 the white. These may be illustrated thus : 

 Pollen (B), Egg (B); Pollen (W), Egg (W). 



It will be seen at once that Pollen (B). 

 may fertilize Egg (B) or Egg (W), giving 

 rise to the pure blue-flowered strain in the 

 one instance and the hybrid in the other. 

 Likewise Pollen (W) may fertilize Egg 

 (W) or Egg (B), giving rise to the white- 

 flowered variety or to the hybrid. Thus 

 in a large number of fertilizations the pro- 

 portions of the blue-flowered offspring to 

 the hybrid and to the white-flowered variety 

 will be 1 :2:i. 



In the hybrids formed from the blue-flow- 

 ered thorny Datura and the white-flowered 

 smooth variety (a cross involving two dif- 

 ferentiating characters), there is a larger 

 num'ber of possible combinations. The dis- 

 tinctive characters of pollen grains and &%^ 

 cells borne by them may be represented 

 thus : Pollen (B and T), Egg (B and T) ; 

 Pollen (B and S), Egg (B and S) ; Pollen 

 (W and T), Egg (W and T) ; Pollen (W 

 and S). Egg (W and S). 



it will ibe seen that several classes may 

 arise and that two new types will occur in 

 the offspring, namely, a blue and smooth 

 Datura, and a white and thorny Datura, and 

 that the number in every class can be calcu- 

 lated by an application of the theory of pro- 

 bability, since the fertilization takes place 

 with equal likelihood in all directions. The 

 characters of the progeny can be foretold 

 with equal certainty. 



From a practical point of view the dis- 

 covery of this law has not been directly of 

 as great value to the plant breeder as was 

 at first predicted. Obviously this is not sur- 

 prising when it is considered that the law is 

 of limited application, that a large propor- 

 tion of the hybrids revert, that the numbers 

 of any new type produced are comparatively 

 small, that the difficulties of preventing sub- 

 sequent crosses are very great, that the 

 number of differentiating qualities between 

 desirable crosses is usually large and so en- 

 tails a most cumbersome experiment, that 

 no new characters can be produced by 



