GABRIELLE L. C. HOWARD. 53 



to the nearest centimetre, but in order to reduce the number of 

 data and to remove accidental inequalities, the heights are given 

 in classes differing by five centimetres in Tables IV, V, and 

 VI. 



Height is one of the few characters which in the F^ genera- 

 tion is not strictly intermediate between the two parents, and 

 the results in the five crosses were not the same. The F^ h^^brid 

 between Type 9 (a dwarf form) and Type 51 (a very tall form) 

 was nearly as tall as the taller parent. The actual mean 

 measurements were, in 1910, Type 9 — 68 cm., Fi — 140 cm., 

 Type51— 165 cm., and in 1912, Type 9—79.9 cm., F^— 158.0 cm. 

 and Type 51 — 195.4 cm. In the cross Type 16 x Type 35, in 

 which the respective heights were. Type 16—106.3 cm., 

 Fi — 93.7 cm. and Type 35 — 86.6 cm., the height is inter- 

 mediate. In the other three crosses the Fj was taller than 

 either parent. This increase in height was slight in the case of 

 Type 23 x Type 38, but very marked in the other two crosses. 

 The actual measurements were Type 2, average height, 155 cm.. 

 Type 3, average height, 152 cm., F^ 200 cm., Type 51, 

 178 cm., Fi Type 2 x Type 51, 195 cm. It has been suggested, 

 in the case of maize, ^ where the F^ is also not strictly inter- 

 mediate, that this is due to the increased vigour of the hybrid 

 plants. This may be the explanation here, but it is difficult 

 to see why this increase should be so much more marked in 

 some cases than in others, unless this increase in vigour be 

 correlated with the number of factors in which the parents 

 differ. 



The F2 generation in the cross Type 9 x Type 51 presented 

 no striking features. A continuous series was formed within 

 the limits of the parents. 



The F2 generation of the cross between Type 16 and 

 Type 35 was absolutely different. These two parents do 

 not differ very greatly in height, the average height of 

 Type 35 in 1912 was 94.1 cm., with a range of 60 to 



1 East, I.e. 



