A. n. Stout . 73 



The group of sister plants grown from the same plant or from the 

 same cross in any one season is called a ser-ies, and in these reports such 

 a group bears the number of the plant that was the immediate parent, 

 together with the numbers in serial order designating the previous 

 jjarentagc. Thus, for example, series (A x E'£2) is a generation of 

 plants derived by using pollen of plant ESii^ on pistils of plant A ; 

 series (A x E"22)-4.- is a group of sister plants grown from self- 

 fertilized seed of the F^ plant {A x E2S) no. 4- Thus the series 

 {A K E22)-.i-o-lI-, for which data are given in Table II, has had three 

 generations of self-fertile ancestry,- of the different series and numbers 

 as designated, which indicate the line of descent from the original cross 

 between A and E22. Although somewhat cumbersome, this treatment 

 presents a complete record of pedigree, from which the performance of 

 individuals, of lines of descent as a whole, and of families may be ascer- 

 tained. All the plants of a sub-family will be given the designation of 

 the common ancestor, thus all the descendants of plant (A x E23) no. 4- 

 may be considered as family (A x E'22)-.!f.-, and all the descendants of 

 plant {A X E22) no. 10 will be designated as family {A xE22)-10~, both, 

 however, being sub-families in the main family (A x E22) but descend- 

 ing from two different sister plants. 



The data in detail for any parents, or for any series referred to, but 

 which were grown previously to 1916, are given in a report already 

 published (Stout, 1916). 



Results of the Experimental Studies. 



1. Performance of a family of Barbe de Capucin x tuUd 

 white-fiowered : the family (E3 x A)-^-. 



In 1916 two series fi'om two generations of self- fertile ancestry were 

 grown in the family (£'r? x A)-4~. The results obtained from the self- 

 pollinations of these 49 plants are compiled in Table I. Of the 

 23 plants of the series first presented in this table, all but five were 

 self-fertile with percentages of fertility ranging from 4 to 48, and with 

 an average fertility for the self-fertile plants of 20 °/^. The percentage of 

 fertility, frequently referred to in this paper, is determined on the basis 

 of the proportion of seeds produced by the flowers involved in the con- 

 trolled pollinations. Of the other series, 16 plants were self- fertile, and 

 10 were self-sterile; the fertilities of the self-fertile plants ranged from 

 2 to 60 °/g with an average of 15 °l^. 



6—2 



