GABRIELLE L. C. HOWARD. 61 



those of the cultures obtained from 738 and indicate the occur- 

 rence of a number of intermediates. This would point to 

 factors of small value. From the fact that no plants shorter than 

 Type 9 have been found in all the generations, it would appear 

 as if Type 51 must contain the factors which go to build up 

 the height of Type 9. 



In the cross between Type 23 and Type 38 the matter is 

 complicated by the fact that the F^ is taller than either parent. 

 The Fo again shows a greater range of variation than the com- 

 bined range of the parents and the large number of plants 

 shorter than Type 23, the shortest parent, is especially notice- 

 able. In the Fy cultures the limits of variation are very different, 

 cf. cultures 204 and 6. The occurrence of plants shorter than 

 Type 23 is again marked. No cultures uniform in height have 

 yet been obtained from this cross. Distinct segregation is, 

 however, apparent, and there are indications that Type 23 and 

 Type 38 contain certain factors in common which give a height 

 somewhere about 80-85 cm. and that the other factors involved 

 in the height of both parents are different, giving rise by 

 combination to plants taller than either parent. 



3. THE NUMBER OF LEAVES PER PLANT. 



This character is somewhat unsatisfactory on account of 

 the amount of variation in each pure type, in proportion to the 

 total variation possible between the types. The smallest number 

 of leaves found in any kind is nineteen, the greatest thirty-four. 

 Table VII shows the approximate number of leaves in each 

 type — the average number of leaves in ten plants. Want 

 of time prevented the measurement of a larger number of 

 plants. It will be seen that every possible number of leaves 

 present and all occur with approximately equal frequency. 



The total number of leaves on the main stem was measured 

 in the following manner. The plants were uprooted, the large 

 roots cut off short, and the base of the plant well washed. It was 

 then possible to count all the leaf scars as well as the living 



