124 White. 



(second count) or 15 "2 per cent on the basis of the third count. In 

 each succeeding cross, the total number of Fa progeny grown was 

 larger and the discrepancies between the actual and the theoretical 

 ratios should have been less in the case of 396 X 301 than greater. 

 In addition to the discrepancies between the actual and theoretical 

 ratios, the abnormal character was modified in its somatic expression 

 to such a degree that it was very difficult to separate the three classes, 

 AA, Aa and aa. Since going over my records, I think it would be 

 impossible to classify them accurately except through breeding tests. 

 Owing to the variable nature of dominance and recessiveness in these 

 crosses, even the selection of the aa (normal) segregates became 

 somewhat difficult, and no doubt the deficient number of abnormal 

 plants (AA and Aa) in the actual ratio obtained, may be explained by 

 my inability to distinguish properly between normal (aa) and Aa 

 segregates. Variation in the expression of the abnormal character was 

 extremely noticeable in all three of these crosses. The proportion of 

 fasciate - stemmed F^ segregates to those with normal stems was 

 respectively 1 — 5, 1 — 9 and 1 — 8-5. Many abnormal plants (AA) then 

 were not characterized by fasciated stems. In the case of the cross 

 304 X 402, all the Fa AA segregates possessed the fasciated stem con- 

 dition in varying d-egrees, but in some of the AA segregates of these 

 other crosses, the stems were as normal as any normal tobacco plant's 

 main axis well could be (Fig. 27). The explanation for this difference in 

 the expression of the factor A appears to lie in the different nature of the 

 gene complexes or genotypes. In 304X402, both parents appear to 

 be genotypically alike except for the factor A, while in aU the other 

 crosses, it is evident that this was not the case. Factor A expresses 

 itself as described under materials when in genotypical environment 304 

 and 402, but very differently under genotypical environments 373, 353, 

 396, etc., because it is modified in its somatic expression in all these 

 environments. Taking 304 or 402 as the standard genotypical environ- 

 ment by which to compare the remainder and calling it Xi, the other 

 environments may be referred to as Xi, X3, or Xi, etc. Under Xi, 

 factor or gene A always gives a certain typical somatic expression, 

 while under any other X, that somatic expression may or may not 

 remain the same. Under Xi environment, no branched fasciations were 

 produced, the ribbon -like linear expansion being characteristic only of 

 the main axis, but in the cross 396 X 301, under Xn environment, 

 segregate plants appeared expressing this anomalous condition in several 



