550 



THE PURE LINE HYPOTHESIS. 



Owing to the fact that the original variegated plant was the 

 paternal parent of all the plants, I. -XXL, and the families given 

 in the above table were raised by self-fertiHsation, the factor for 

 variegation would be identical in all the individuals of the 

 families according to the strict Mendelian standpoint. 



In accordance with the pure-line theory there might be 

 considerable variation in the expression of the character in the 

 different individuals, but there should be no tendency for the 

 averages or means of the families to differ from one another. 

 There is, however, a very clearly-marked difference in the means 

 of the families, e.g., in family XX, there was a mean variegation 

 of 51, and in family I a mean of 85. This result can scarcely be 

 explained except by supposing that there had occurred a modifica- 

 tion of the factor in the parents oi these families. 



3. Size of Leaves and its Relationship to Habit and Variecfation. 



From a casual inspection of the plants it was clear that the 

 general size of the leaves varied considerably in different indivi- 

 duals. The maximum widths of some ten full-grown leaves of a 

 plant were measured, and the mean was taken as the character- 

 istic leaf-width for the plant. 



The mean leaf -width of the 9 parent with uniform green 

 leaves was 55 mm., and of the S parent with variegated leaves 

 was 44 mm., while the mean width of 18 hybrids raised by 

 crossing these plants was 69 mm. The distribution was as 

 follows : 



Thus the hybrids possessed large leaves, and, speaking in 

 Mendelian terms, large leaves may be said to be dominant over 

 small leaves. It must be remembered, however, that the size of 

 the leaf depends very largely on the general environment and the 

 age of the plant, and consequently the fact that the mean leaf- 

 width of the hybrids was greater than the leaf-width of the 9 

 parent is not surprising. It is clear, however, that there was no 

 tendency for the leaf-wadths of the hybrids to cluster around the 

 leaf-width of the male parent, or to approach an intermediate 

 condition between the widths of the two parents. 



From 19 hybrids families (containing in all 144 plants) were 

 raised by self-fertilisation, and all the plants were grown under 

 similar conditions of light, soil and moisture. The means of the 

 different families varied considerably (see table), but owing to 

 the fact that the leaves from the parents were not collected when 

 the plants were at the same age, or at the same time of year, as 

 in the case of the offspring, it is not oossible to ascertain in any 

 reliable manner the strength of inheritance between parent and 

 offspring. 



