128 Hayes. 
That two out of the five F, generations of this cross bred true to 
the parental leaf shape is a further reason for a belief in a Mendelian 
inheritance of quantitative characters. 
Fig. 8. From left to right middle leaves of Plants shown in Figures 4—7. At the 
left, (402 X 405),—1I—3, next 402—1—1, Havana, then 405—-1—1, Cuban and at the 
right, 402 405)—1—2. Grown under cloth shade ad Blooınfield, 1912. 
Conclusion, 
Our results show the F, generation of a cross between pure lines 
of tobacco to be of a uniform nature and no more variable than the 
parents. Different F, individuals gave similar results in F,, showing 
the variability in Fi to be of no germinal value. In F, there was an 
increase of variability and, where a sufficient number of individuals 
were grown, there was a range of variation equal to the combined 
range of variation in the parents. Certain F, individuals bred true in 
F,, giving no greater variability than the parent forms, while others 
gave variabilities ranging from the parents to that the F, generation. 
These results give further confirmation of the hypothesis that 
the inheritance of quantitative characters, such as size, shape and 
number of various plant organs may be due to the interaction of a 
multiplicity of factors, each inherited separately and capable of adding 
to the character, the heterozygous condition being half the homo- 
zygous. The difficulty of determining the number of factors involved 
is very great, owing to the obscuring effect of fluctuating variability. 
Whatever may be the ultimate accepted hypothesis concerning 
inheritance of quantitative characters, from the plant breeding stand- 
