214 



The Journal of Heredity 



1920 I9ZI 1922 



year 



RESULTS OF FIVE YEARS OF SELECTION 



Figure 10. On the left is shown the yearly variation in percentage of central stem in 

 four selections of broom corn, compared for the last three years with two commercial 

 varieties, Longbrush Evergreen (L. E. ) and California Longbrush (C. L. ) On the right is 

 shown the variation in length of stem in the same varieties. Yearly variation in seasonal 

 conditions generally affects all varieties similarly, but in 1922 the amount of central stem 

 increased considerably in the commercial varieties, and decreased in the selections. The 

 reasons for this are discussed on page 215. 



The degree of uniformity within a 

 head-row, on the whole was decidedly 

 greater than usually occurs in ear-rows 

 of maize, except after inbreeding, indi- 

 cating that a considerable amount of 

 self pollination occurs naturally in 

 broom corn as no bagging of heads was 

 done the first year. Beginning with 

 1 91 9, some heads have been bagged 

 each year and a full setting of seed 

 has always resulted, showing the exis- 

 tence of complete self-fertility. In 

 most cases selfed seed from bagged 

 heads has been planted, but some open 

 pollinated heads were used when a sat- 

 isfactory number of good quality 

 l)agged heads were not available. 



Since 191 8 selection has been con- 

 tinued within a few of the better 

 strains, first for freedom from central 

 stem and after that for good length and 

 fineness of fiber. In 1920, and each 

 year since then, Longbush Evergreen 

 and another variety, California Long- 

 brush, both secured from a commercial 

 seedsman, have been grown in compari- 

 son with the selections. The two 

 varieties have been ])ropagated bv com- 

 positing seed from a random sample of 

 the marketable brush, rejecting the bad- 

 Iv twisted and worthless stemmv heads. 



but including the better grade of stem- 

 my ones. 



Of the original 105 heads planted in 

 1 91 8 the number represented by prog- 

 eny has rapidly decreased in successive 

 years, due to the dropping out of in- 

 ferior strains, until in 1921 only three 

 remained of the good qualityi strains. 

 The performance of four strains 

 through five generations is shown 

 graphically in Figure 10. 



The values in the graphs are yearly 

 averages of several head rows repre- 

 senting four of the original strains, 

 each average involving from 100 to 

 1,000 plants. 



Environmental Influence 



That seasonal conditions have a pro- 

 noimced efifect on the amount of central 

 stem and U])on length of brush is evi- 

 dent from a study of the graphs. The 

 values fluctuate widely from one year 

 to another Ijut the relative positions of 

 the strains are fairly well maintained 

 and all the strains and varieties show 

 about the same kind and degree of 

 restwnse to seasonal influences. 



The year 1920 was a very favorable 

 cro]:) season at Columbus, yields of 

 small grains and of corn being high. 



