+ 
ge weight 
f 
fi 
Avera 
300 200 
458 BOTANICAL GAZETTE [DECEMBER 
question, as the average percentage of sugar, average size of root, 
and yield of sugar per acre have been increased greatly. Moreover, 
there is no valid reason for assuming that these characters have 
reached their maximum development. The differences exhibited 
by the varieties Morrison’s Kleinwanzleben, Kleinwanzleben’s 
Original, and Madison Original Selections, as illustrated in fig. 1, 
show that sugar-beet varieties may still be altered by breeding. 
For the sake of discussion it will be assumed that beets are improved 
of 
Row numbers 
in grams 
3 
30 
40 
50 
60 
70 
80 
50 
/00 
/10 
/20 
/30 
/40 
50 
/60 
wit VA 
0 
ry 
By 
ff 
VAM AL 
‘el Ve een “Is, fy 20 
- 
> 
° 
3 
w 
° 
3 
5 
eo 
o toad 
—— 
“e. 
> 
eo 
i 
200 j000 : | 
Fic. 49.—Effect of soil irregularities on relationship between average root weight 
and yield of sugar in consecutive check rows, Madison, 1914; solid line indicates 
average weight of root per row; broken line, yield of sugar. 
either (1) by a gradual accumulation of small variations through 
the process of continuous selection, or (2) by the isolation and multi- 
plication of an occasional mutation. 
Continuous selection is based upon the theory that “like 
begets like,” or the tendency of superior individuals to transmit 
their qualities to their progeny. As shown in the foregoing tables 
and graphs, the best roots transmit no better qualities than do 
mediocre roots. Their differences are mere fluctuations, therefore, 
and have no influence on beet improvement. This has no bearing, 
however, upon the value of continuous selection when applied to 
real differences, as the possibilities of improvement through a 
