On the Genetics of Tetraploid Plants in P. sinensis. 



491 



To turn to the crosses of green x red stigma (Table II) : it should first be 

 pointed out that the GT race sprang from diploid races pure for green stigma 

 and no red stigma has ever appeared in this race bred in the direct line. 

 Plants of this race may, therefore, be written GGGG. It is entirely in 

 accordance with this that the Fi's from crosses of this race with red stigma 

 have all proved to be of the type GGgg, giving the ratios 15 : 1 when selfed, 

 and 3 : 1 when crossed with the recessive. Heterozygotes of the type Gggg 

 have, however, been found by selfing green-stigma plants chosen from 

 families in which some of the plants had red stigmas. Ten such plants 

 have given altogether 99 green stigma, 34 red stigma. 



It will be noticed that the Fi's which appear in the green x red-stigma 

 crosses also appear in the thrum x pin crosses. The I\ QQ/IS is giving the 

 same ratios in respect of each pair of characters, namely, 15:1 when selfed, 

 and 3 : 1 when crossed by the recessive. But the Fi's 72/13 are giving 15 : 1 

 and 3 : 1 for green and red stigma, and 3 : 1 and 1:1 for thrum and pin. 

 Taking the two pairs of characters together, and assuming for the moment 

 that there are no special inter-relations between the factors, these would 

 give the curious ratios of 45 TG : 15 tG : 3 Tg : 1 tg when the Fi is selfed, 

 and 3:3:1:1 when the Fi is crossed by the double recessive (tg). The 

 actual numbers obtained are 28 TG : 4 tG : Tg : tg in the former case, 

 and 22 TG : 25 tG : 2 Tg : 10 tg in the latter. 



In the foregoing results the different kinds of heterozygote stand out 

 clearly identified by their progeny, and, although there are considerable 

 discrepancies in individual cases, yet the general trend of these results 

 clearly shows, I think, that the tetraploid plants are endowed with a double 

 set of factors, as compared with the diploid races. 



As has been remarked, the results of experiments with the intermediate 

 types are in general accord with the idea that my existing intermediates are 

 heterozygotes of the type Aaaa, in which one " dose " of the factor is not 

 sufficient for the full development of the " dominant " character. The 

 variations among the intermediates themselves, which are probably of the 

 same nature as the variations exhibited among heterozygotes in cases where 

 dominance is imperfect, are, of course, still to be explained. It is curious to 

 notice that when there is any marked variation between the organs of the 

 same plant it appears generally to take the form of a gradual retrogression 

 towards the recessive character in the successively younger and younger 

 organs, the effect of the positive factor being a little less pronounced in each 

 new organ formed. 



In conclusion, it must be remarked that the results so far obtained do 

 not of themselves throw any direct light on the problem of the possible 

 vol. lxxxvii. — b. 2 p 



