The Isolation of Tricotylous Intermediate Races. 421 
capacity are rarely found at first. I came across an 
instance of one only once, at the beginning of my experi- 
ment with the syncotylous Helianthus animus (see the 
following chapter). If individuals with low hereditary ca- 
pacities are found, those with mean values can as a rule 
be easily raised from them, for they are to be regarded 
simply as minus variants of the race sought for; and 
will therefore, in conformity with the law of regression, 
revert to this value even if selection be only suspended 
( see above, p. 5 ) . Experience shows, and the table al- 
ready referred to will demonstrate, that one or two gen- 
erations are, as a rule, sufficient for the attainment of 
values of 50 to 60%. 
Before I proceed to a detailed description of my ex- 
periments I will give a few instances to show the course 
which these experiments in the isolation of tricotylous 
intermediate races follow. 
The attainment of the mean value. On page 380 I 
have given, in a small table, the numbers of tricotyls 
which I found in some samples of seed in the spring of 
1895. For some of these species these tricotyls were 
planted out and their seeds saved separately and sown. 
I obtained values from 12 to 19% in the best individuals 
(spring 1896). and on the rest, as a rule, much fewer. 
These were regarded as belonging to the half race or as 
of doubtful significance. The tricotyls from the best 
parents were now (1896) planted out, and the hereditary 
value for each was calculated in the following spring. 
Below I give a resume of these figures in groups of 2, 
3 7, 8--129f, with means of l--5--10% and so forth, 
for convenience of comparison, and indicate, for each 
such reduced value, the number of individuals which 
exhibited it. 
