354 Journal of Agricultural Research voi.xxii.no. 



higher classification. The few mdividuals found in these speltoid classes 

 at present may be regarded as exceptions. Their significance will be 

 considered later. 



The distributions of the families in the series segregating in the ratio 

 of 15 to I (13 126a) (Table V) are entirely different. They do not take the 

 general course described above. Some of them have very low frequencies 

 at classes i and 2. Families 5, 7, and 12 have no individual in class i, the 

 population of family 5 being composed of class 9 individuals only. Family 

 25 produced its spelts in classes 6 to 9, inclusive. Of the 12 families 

 being considered 6 produced class 9 individuals, while among the families 

 segregating in the 3 to i ratio there is but one instance (13255a, family i) 

 where class 9 individuals have been produced. 



The explanation of the increased variability of the constant spelt- 

 producing families of series 13126a, as compared with series 13255a and 

 13260a, will be found in the factorial explanations given for these two 

 groups of crosses. 



Families 13255a and 13260a segregated in the simple monohybrid 

 3 to I ratio. By hypothesis, all the spelts producing only spelts are sup- 

 posed to have the genotypic composition SS. 



As to the cross 13126a which segregated in the 15 to i ratio, it was 

 shown that there were five constant spelt forms, namely : 



SjvSiSaSg SiSiSjSs S1S1S2S2 SiSjSzSg S1S1S2S2 



Although these forms would keep on producing only spelts, they are 

 not genotypically identical. 



The fact that in the families segregating in the 3 to i ratio there was 

 only one genotypic spelt form and in the family segregating in the 15 to i 

 ratio five such forms were present may account for the increased varia- 

 bility among the pure-breeding spelts of the latter cross. 



The F3 progenies can not be separated into the five theoretical geno- 

 typic groups just referred to because, among other things, there is positive 

 evidence that modifiers also are concerned which have the tendency to 

 shift the classes one way or the other. This phase of the subject will 

 next be discussed. 



MODIFICATION OF THE DEGREE OF INHERITANCE OF SPELT CHAR- 

 ACTERS, DUE TO THE PRESENCE OF MODIFYING FACTORS 



For the consideration of the subject of modifiers the analyses will be 

 confined primarily to the spelt classes (i to 9, inclusive) of the progeny 

 of the heterozygous F.^ individuals of series 13260a, shown in Table X. 

 This family has been chosen because it represents a simple mode of 

 segregation. Whatever is said about modifiers for this family will be 

 found to apply just as well to the other families. 



It has been shown that only one spelt factor was concerned in the cross 

 under consideration. All the Fj heterozygous plants had the formula 



