312 University of California Publications in Botany [Vol. 5 



DISCUSSION OF EESULTS 



The results which have been presented demonstrate that in crosses 

 involving sylvestris and varieties of Tabacum the F^ hybrids through- 

 out display the characters of the particular Tabacum variety used 

 in the cross, but usually on a greatly enlarged scale. The consequences 

 of the increased growth and vigor of such hybrids must not be lost 

 sight of in judging as to the completeness of correspondence between 

 them and their Tabacum parents. This is true because such stimula- 

 tion may conceivably give rise to variations in proportional effect in 

 different characters, in such cases disturbing somewhat the ratio re- 

 lations usual for the character expressions, although fundamentally 

 the whole series of characters owes its expression entirely to the 

 directive effect of the Tabacum reaction system. Specifically we have 

 shown elsewhere (Goodspeed and Clausen, 1915) that under green- 

 house conditions it is possible to modily somewhat such a compara- 

 tively constant character complex as flower size, the spread of corolla 

 decreasing significantly when compared with that found under field 

 conditions, the length of corolla, on the other hand, increasing 

 somewhat. It is therefore possible that whatever slight deviations 

 may be found in these species hybrids of Tabacum with sylvestris, 

 when compared with tlieir parent Tabacum varieties under similar 

 conditions, may be largely dependent upon differences in the specific 

 reactions of certain character complexes to the new set of conditions 

 obtaining in the F, hybrid. 



The results lierein set forth are particularly striking when com- 

 pared with the type of behavior exhibited by Tabacum in varietal 

 crosses within the grouj). The Tabacum varieties differ strikingly in 

 a large number of characters and when crossed give in Fj forms which 

 are intermediate in their characters. Tliis intermediate condition is in 

 part undoubtedly due to the fact that each parent contributes a scat- 

 tered set of factors which display more or less complete dominance, so 

 that a hybrid is produced the general ty])e of which occupies a position 

 intermediate between the two parents.^ Much of this intermediacy, 

 however, is due to a blending in F^ which may be followed by definite, 

 although often very complex, segregation in subsequent generations. 

 Very often this segregation is so complex as to indicate the existence 



1 A number of character contrasts in Tabacum are known which do display 

 nearly or quite complete dominance: viz., the petioled condition of the leaf 

 as opposed to the non-petioled or sessile condition, the normal type of flower 

 as opposed to the abnormal, calycine type, and the normal fruiting condition 

 as opposed to the parthenogenetic tendency which obtains in the variety 

 ' ' Cuba. ' ' 



