402 Tricotylous Races. 
sisted of the 20 best tricotyls only, and 16 of these ripened 
their seeds. In this generation, however, there was a 
considerable step back, for the individual parents varied 
between 0.5% and 3.7% ; and one parent had not a single 
aberrant form amongst 200 seedlings. The mean was 
1.8%. The seeds of the five best plants were again sown 
and 1000 seedlings from each examined. The values 
were now 2.6 2.8 3.2 3.2 and 3.6%. In the following- 
years I endeavored to determine the cause of this re- 
turn by a series of collateral experiments, but without 
success. The variability of the hereditary value in such 
races obviously depends in great part on causes which 
we do not yet understand. 
I now planted out the tricotylous offspring of two 
plants with 3.2 and 3.6%, but in two separate groups 
in order to be able to confine the selection to the offspring 
of one of them later, if desired. These cultures consisted 
of 12 and 13 tricotyls, no essential difference between 
them being manifested. The values calculated from the 
batches of 300 seedlings from each seed-parent were : 
SEED- VALUES CALCULATED FROM THE OFFSPRING 
PARENT (Spring, 1896) 
with 3.2 % 1.3 1.7 1.7 2.0 2.3 2.3 2.7 2.7 2.7 3.0 4.0 5.5 
with 3. 6% 0.7 0.7 1.0 1.3 1.7 2.0 2.7 2.7 3.0 3.0 3.7 4.7 5.3 
The mean for both cases was therefore about 2.5% 
(actually 2.5% and 2.7% respectively). This experi- 
ment proves how little effect an enlargement of the ex- 
tent of the cultures has in such an experiment in selection, 
for if I had only dealt with the offspring of one of the 
two parents, the result would obviously not have been 
essentially different. 
In the summer of 1896 I did not go on with this cul- 
ture, but tried to find out whether by planting out a con- 
