148 University of California Publications in Botany [Vol. 5 



lings are also interesting in this connection (see also, Gregory, 

 1909, and Gard, 1911). The flnctuation in the length of 

 the parent cotyledons was great in each cross, — never 

 less than 5 mm. and nsnally greater than 10 mm. — and 

 the degree of fluctuation of cotyledon length in the F^ hybrids, 

 considerably' diminished {loc. cit., p. 7), for many of the crosses. 

 A rather strange predominance throughout of the female influ- 

 ence in each cross is reported (see Hagedoorn, 1908 ; Loeb, King 

 and Moore, 1910). The general result of the experiments seems 

 to show that the "absolute size .... tends to be larger in 

 the cross than the mean between the corresponding characters 

 of the parents" (Groth, loc. cit.. p. 33). The fluctuation in coty- 

 ledon length almost always equalled and surpassed the higher 

 extreme, but in no case reached the lower extreme of the parental 

 fluctuations. It is in this connection especially that Groth 's 

 results differ from those reported in this paper. In the case 

 of tomato seedlings it might possibly be held, since the D in- 

 fluence was practically always strongly evident; since the R 

 influence was likewise absent, and finally since the average length 

 of the hybrid cotyledons was nearly always greater than the 

 mean between the average length of the parent cotyledons, that 

 in respect to length of cotyledons, "longness" is dominant over 

 both "shortness" and " intermediateness. " It will again be 

 repeated, for comparison, that in reference to corolla diameter 

 in N. acuminata hybrids both D and B influences were always 

 present, and the mean between the average corolla diameters 

 of the parents was practically identical, in every cross, with 

 the average corolla diameter of the hybrid in each case. In a 

 second report upon the mature F^ tomato hybrids Groth finds 

 that in the seedlings ' ' opposing factors .... are still struggling 

 for mastery .... an equilibrium is reached when the plant 

 becomes older, and that equilibrium is generally the same for 

 the same combination of factors." This equilibrium is often 

 shifted, however, "so that even in grown plants reciprocal and 

 duplicate crosses may differ in characters of size, shape and 

 number" (1911, part 2, p. 11). The fact that the F^ generation 

 of N. acuminata hybrids is far more variable in respect to floral 



