A. B. Stout 91 



the self-fertile plants in the F^ series appeared sporadically. In the 

 columns giving fertility of ancestry, the individual percentages are given 

 for the parents in line of descent and the performance of the series 

 to which these belonged is, of course, to be found from the pedigree. 

 A general summary is given in bold face type for each family as a 

 whole. 



It is a most noticeable fact that at least some self-sterile plants 

 appeared in every series but one, and this was a small series of only 

 two plants. The proportion of self-fertile plants varies considerably. 

 Considering the families as a whole, it is highest in the family (£";? xA)-^.-. 

 In the three sub-families of J. and E22 parentage, the percentages of the 

 proportions are quite the same (40, 49, and 40). The distribution of 

 the self-fertilities is quite similar in all families with the larger number 

 of plants of low fertility. The range extends into somewhat higher 

 values in the family {A x E22)-9-. On the basis of the average 

 fertility of the self-fertile plants, there is a range from 0"165 to 0"223. 



Considering all these data, there appear to be no very decided family 

 differences in regard to the heredity and variability of self-compatibilities 

 and incompatibilities. 



It is to be recognized that the data are not sufficient to give an 

 adequate judgment of the performance of a family or a line of descent 

 constituting a considerable progeny and having an ancestral record of 

 feeble fertility. As I have conducted the experiments, to select con- 

 tinually for very feeble fertility is to greatly limit the number of the 

 progeny. When the pollinations are made, there is no way of knowing 

 with any certainty what the degree of fertility is. When this becomes 

 known, it is usually too late to make in that year the large number of 

 pollinations necessary for the production of considerable seed by feebly 

 self-fertile plants. It would be quite possible, however, to keep feebly 

 fertile plants, and by making large numbers of pollinations in succeeding 

 years to obtain considerable seed. 



Of the families gi-own thus far (A x E22)-4.- has an ancestry of the 

 lowest fertility. The data for the first self-fertile ancestor (Stout, 1916, 

 Table III) are quite adequate to establish its low fertility. Of 26 

 different heads pollinated on eight different days, 20 set no seed ; in the 

 six heads the number of seed were 1, 3, 3, 4, 4, 4, giving a percentage of 

 4. Of the ten plants grown to maturity from such seed, 7 were self- 

 fertile, but the highest individual fertility was 23. In the next genera- 

 tion from four parents, 28 plants were grown. Ten plants of one series 

 were all self-sterile, and in another series of 8 plants, 6 were self-fertile, 



