488 



Mr. R. P. Gregory. 



Phenomena of the kind just described, taken together with the fact that 

 the tetraploid giants have produced intermediates peculiar to themselves, 

 suggest some considerations as to the factorial constitution of the tetraploid 

 plants. Both classes of phenomena can, I think, be explained by means of 

 the hypothesis that, as compared with diploid plants, the tetraploid plants 

 possess a double set of factors. Since in the zygote of a diploid pure race 

 each factor is to be regarded as represented twice, AA, it follows that the 

 tetraploid plant, according to this hypothesis, will be AAAA, and the 

 gametes from which such a zygote is formed must be A A, that is to say, 

 the factor will be represented twice in the gamete, instead of once, as it is in 

 the gametes of the ordinary diploid race. 



Heterozygous tetraploid plants may, then, be any one of three possible 

 kinds, AAAa, AAaa, Aaaa. Since each gamete will contain two of the four 

 units (" presences " or " absences ") which make up the tetraploid group, the 

 gametes produced by the three kinds of heterozygote, and the resulting 

 progeny in F 2 , will be as follows : — 



Case I. — Heterozygote, AAAa ; gametes, AA, Aa ; 

 F 2 , 1 AAAA : 2 AAAa : 1 AAaa. 



No pure recessives in F 2 , but, of every four plants, one will give pure 

 reeessives in F3 in the proportion of one recessive in every 16 plants (see 

 Case 2). 



Case II. — Heterozygote, AAaa ; gametes, AA, Aa, Aa, aa ; 



F 2 , 1AAAA : 4 AAAa : 6 AAaa : 4 Aaaa : 1 



F 2 contains one pure recessive in every 16 plants. 



Case III. — Heterozygote, Aaaa ; gametes, Aa, aa ; 

 F 2 , 1 AAaa ; 2 Aaaa : 1 aaaa. 



F 2 contains one pure recessive in every three plants ; no pure dominants, 

 but one plant in every four will give pure dominants in F 3 . 



Of the various kinds of heterozygote shown in the foregoing scheme, one, 

 namely AAaa, has the same proportion of positive and negative elements 

 (" presences " and " absences ") as the ordinary diploid heterozygote. With 

 regard to the characters in respect of which the tetraploid giants have 

 produced peculiar intermediates, it is suggested that the intermediates may be 

 either AAAa or Aaaa. The former would presumably show the cumulative 

 effect of the three factors, like that which Mlsson-Ehle and East have 

 recognised in some of their cases, by giving a type more cfosely resembling 

 the pure dominant than does the ordinary diploid heterozygote, but in the 

 Primulas such types have not yet been definitely recognised by inspection. 



