1913] Goodspeed: Nicotiana Hyhrids 173 



range of variation in corolla diameters but in this hybrid the 

 upper limit of variation of the 1912 variety II is never exceeded 

 in the flower of any individual though the largest sized flower 

 is 4 mm. larger than the largest flower on variety II in 1910 or 

 1911. It is also to be noted that in the Fg of this cross there 

 are no individuals which exhibit a small range of variation such 

 as was found in certain plants in F^ of the cross I^XHd^- Every 

 plant bears flowers smaller than the smallest flowers ever meas- 

 ured on the small flowered parental variety III and every plant 

 also bears flowers larger, with possibly one exception, than the 

 smallest flower measured on Variety II in 1910 or 1911. In 

 other words the Fo plants from this cross exhibit identical ranges 

 of variation with respect to the corolla diameters of the flowers 

 which they bear. 



Averages of corolla diameter have, in none of the cases, been 

 calculated. The results at present available and as outlined above 

 seem, however, to preclude the possibility of a simple Mendelian 

 interpretation of the inheritance of the flower-size characters 

 peculiar to varieties of ^V. acuminata. Some difference of opinion 

 is at present apparent with reference to the interpretation of 

 the mode of inheritance of morphological characters — size and 

 form characters. Certain eases of blending inheritance in F^ are 

 seemingly not followed by the occurrence of segregation in F,. 

 On the other hand the method of interpretation of certain other- 

 wise obscure experimental results as given by Nillson-Ehle 

 (1909), East (19]0), Hayes (1912) and others seems to some 

 to open the path for the explanation of an almost limitless range 

 of experimental results, derived from a study of the inheritance 

 of morphological characters, on what seems to be a modified IVIen- 

 delian basis (cf. Hayes, 1912, p. 31). Thus the occurrence, as 

 outlined above, of a seemingly greater degree of variation in F, 

 than in F^ — and the truth of this situation is open to question 

 lintil the true meaning of fluctuating variability in our experi- 

 mental material is more thoroughly determined — appears to make 

 it plain that segregation does occur (Castle, 1911, p. 137) or, as 

 one writer has so aptly put it "the segregation of potential 

 f italics mine) characters in tlic germ cells of hybrids and their 

 chance recombination in later generations" (Hayes, 1912) has 



