88 Reduplication Series in Siveet Peas 



support the idea that we are here dealing with a 9:3:3:1 

 ratio. As already mentioned it is not impossible that in 

 Fam. '12, 71 (Table IX) this type of family has been encoun- 

 tered once before. 



(3) The relation hetiueen N and D in plants homozygous for F. 



Hitherto the opportunity of studying this particular case has been 

 limited to a single family. This family ('13, 119) was from a normal 

 fertile belonging to Fam. '12, 88 (Table IX) and proved to be hetero- 

 zygous in N and D but homozygous for F. It contained 100 plants 

 and was composed of 



Normal Normal Cretin Cretin 



dark light dark light 



48 22 27 3 



Expect, on 1 : 3 : 3 : 1 system ... 51-6 23-4 23-4 1-6 



The figures are in consonance with the idea that the reduplication 

 for N and D is on a 1 : 3 basis, and this is borne out by the behaviour 

 of these two factors in relation to one another among plants which are 

 heterozygous for N, D, and F (see p. 89). 



(4) The relations between D, F, and N in plants heterozygous 

 for all three factors. 



As in the case of the B, E, L series there are four ways theoretically 

 possible by which plants heterozygous for three factors can be pro- 

 duced, viz. : 



(a) DFN X dfn. 



(13) DfNxdFn. 



(7) DFn X dfN. 



(B) dFN X Dfn. 



Of these four ways I have hitherto been able to study two, viz. (/8) 

 and (7). 



(/3) Nature of mating, DfN x dFn. 



The details of four such families raised from the same parents and 

 comprising in all 442 plants are set out in Table XI, and may be 

 considered together. The proportion of dark axils to light axils, of 

 fertiles to steriles, and of normals to cretins is closely 3 : 1 in each case. 



