A. B. Stout 89 



Other races are shown in Plate VI. In the right foreground, 

 extending to field label 63, are plants of (A x E23)-Ji^6-3-, which 

 were of a very decided dwarf habit with few brittle branches, and a 

 very marked susceptibility to a stem-rot due to fungous or bacterial 

 infection in the tips of the branches. The field no. 63 designates the 

 15 plants of {E2i2 x A)-10 Ser. II. 10-, all but one of which were self- 

 sterile ; these, it will be noted, are of a rather tall and much-branched 

 habit. Field no. 57 is for series {A x E3;2)-9-4-10-, which exhibited 

 the rather incongruous combination of large vigorous erect rosette 

 leaves with a weakly developed and sparsely branched main stem. 

 To the left of field no. 57 is a row of series {A x E2:i)-^-3-ll-. This 

 series was somewhat like the dwarf series already noted, hut was of 

 more vigorous growth. The branches were brittle and very susceptible 

 to the same disease. The dying tips of the branches are (juito will 

 shown in the reproduction of plants in the foreground. 



The above rather brief description will give some idea of the various 

 vegetative forms that have appeared in the various lines of descent. 

 It is to be noted that the plants of each series of this Fi generation and 

 of most of the series of the F^ are very uniform among themselves in 

 general vigour and habit of growth, but that various vegetative types 

 are very different firom each other. The general data for flower colour are 

 given in the various tables, which show that for some of the series 

 the flower colour was quite uniform, while for others white-flowered and 

 blue-flowered plants were both in evidence. 



The data in detail given in Tables I — VI, and the summary of Table 

 VII, show that self-incompatibility develops in these various races in quite 

 the same degree. The most dwarf and the most robust races or lines are 

 quite alike in performance. Plants that are widely dissimilar may be 

 self-fertile or self-sterile. Also the performance within the various 

 series indicates that plants that are quite identical in all vegetative 

 characters may be either self-fertile or self-sterile. 



4. General Smnmary of Results. 



In Table VII, the results obtained in 1916 are compiled together 

 with those obtained in previous years, thus presenting a summary for 

 the various series and generations of each family. Here the summary 

 for the Fi generation derived by crossing the self-sterile plants A and 

 E3 or A and E32 is presented in italics for comparison, but is not 

 included in the summaries of families of self-fertile lines of descent, for 



7—2 



