IIQ White. 



lot, but were distinguishable by some character difference, and this 

 population was ]»y far the most variable of all those which I grew. In 

 Table 25, some idea of the striking differences in height may be gained. 

 The range is: aa segregates, 12-2 — 20*6 dem.; AA and Aa segregates, 

 10*7 — 24*4 dem. In leaf count, the aa class varied from 10 to 17; 

 the Aa between 10 and 33, and the AA, from 14 to 85 leaves per plant. 

 Fasciate-stemmed segregates are not confined to any particular type, 

 but are present in a large number of the different character combinations, 

 among them being dwarf individuals (9*15 dem.) with only 15 leaves, 

 and extremely tall plants with either few or many leaves. The 

 correlation between height and number of leaves per plant was not 

 marked. Branched fasciated stems appeared for the first time. The 

 branching type had come from the 396 grandparent and the fasciated 

 stem character of the 300 — 309 race had spread itself out over the 

 three or four main branches, expressing itself even in the little twigs. 

 This was a distinctly new type, and not a bifurcate or multiradiate 

 fasciatiou. Other well defined types with fasciated stems occurred, and 

 one of these, especially distinctive, gametically A ABB or AaBB, had 

 a slender, unbranched, flattened axis bearing only 10 or 15 leaves. 

 All types as described in Table 25, w^ere selfed and much more light 

 will be thrown on the subject by the Fs generation. Pending the 

 growing of this, the explanation given for changes in dominance of A 

 in the various Fi crosses, is presented to account for the unexpected 

 distortion in the Fg ratios. 



Discussion. 



From the results of these three varietal crosses, together with 

 the data from 9 Fi hybrid families, it would appear that other factors 

 must markedly affect the somatic expression of the factor A. And this 

 may be so much modified that the intermediate expression of dominance 

 in the Aa segregate may be changed to complete dominance of the normal 

 (aa) condition, provided certain other unrelated, but interacting factors 

 were present in the zygote. If this occurred, more segregates somatically 

 normal would be expected in Fa, because heterozygotes of this kind 

 could not be distinguished from aa plants, except through the breeding 

 test. The failure of the factor A to alter the normal appearance of 

 the stem can be accounted for in the same manner. It is very evident 

 from the F2 results of the cross 304 X 402, that when two homozygous 

 pure lines differing in a single factor are crossed, the ¥> individuals 



