130 s. iKENo : 



The agreement between the theoretical and the actual numbei-s is not very 

 bad in lx)th cases in \ie\f of the small number of plants, except in respect 

 to the orange where the number of individuals is much smaller than might 

 be expected theoretically ; this may be due possibly to the fact that a 

 compai'atively large proportion of seeds of this class of zygotes failed to 

 germinate. ITie further behaviom* of one of the two reds segregated out in 

 i^, was ascertained, becausa I could get seeds on it by selfing. It has 

 produced in 1919 the i^-oflfepring containing 1 magenta, 28 reds, 12 oranges 

 and 17 whites, thus proving itseK to be heterozygous. We may consider the 

 red examined to have been of the constitution CCHrhh,^ and that it has 

 changed into CcHrlih before the formation of gametes ; the male and the 

 female gametes CHh, cRh, Crh, crh are formed, and their mating has 

 given rise to 29 reds: 12 oranges: 17 whites, theoi-eticaUy 32-6±3-8 : 10-9± 

 3"0:14-5±3'8 on 9:3:4 basis, if we vnW. count 1 magenta as 1 red; 1 

 magenta was formed because at least one gamete CJtB was produced by 

 reversion, so that the whole process consists, exactly as in No. 29 (Table 

 VIII, discussion p. 127), in the " loss-mutation " combined with reversional 

 change. 



In the Table VI, A (p. 109) we have seen that the i^'^-oflfepring ex 

 white-II X orange No. 2 contains 81 magentas, 20 oranges, 34 wliites and 1 

 red, 136 in all. They are the total of all the offspring derived from 12 

 i^-plants by selfing. Of these 12 families of the i^^-offepring 11 conhiin no 

 red at all, so that the one red under discussion belongs to one family 

 composed of only 7 plants, viz. 5 magentas, 1 red and 1 orange.^ The 

 occurrence of 1 red in this family might parhaps be due to the partial linkage 

 similai- to that in the case just stated above, but no sure conclusion cau be 

 drawn here on account of the small number of indi^aduals. 



Bud-m-utaKom have been observed, though not fi-equently. They are 

 below simply described, but no discussion -svill be made, this being impossible 



' Smce the factors K antl B were linked according to the series 5:1:1:5 or 6:1:1:6 

 in the preceding generation, we shoiild have in the present case the reds of the composition 

 CClirbb iind Ci'ltltbh in the ratio 10 : 1 or 12 : 1 respectively, and it is fpiite natural that 

 we have now VCIit'hh, be<'.'mse only two nils were cxaraineii. 



2 Whites are wanting in this small family, perhaps because all seeds which will give 

 rise to whites failed to germinate. 



