The Isolation of Tricotylous Intermediate Races. 441 
were already present in the original sample of seed, or 
because between the intermediate and the half races hy- 
brids were met with, by the subsequent segregation of 
which the race was produced. 
In the first experiments the isolation of the tricotyl- 
ous intermediate race took from three to four genera- 
tions. Later when I started the experiments on a larger 
scale, the number was reduced to two years. Further 
selection brought it up, in one year as a rule, to 70- 
80% and sometimes even to 90%, either immediately 
in the course of another year, or after two generations. 
Unfavorable conditions of life led to exceptions, or even 
to retrogression; but only Helichrysum and Silene ex- 
hibited this feature. Cannabis, Mercurialis and Antir- 
rhinum maintained a value of 80 90% under continued 
selection, and it is quite probable that even 100% might 
have been attained in occasional individuals. The mean 
figures of the whole group correspond, as a rule, to these 
maximum values. They maintained themselves at about 
55%, but are liable to be increased by selection or dimin- 
ished by unfavorable conditions. 
Lastly, it should be noted that the ngures for these 
intermediate races are so far removed from those of the 
half races (p. 392), that all suspicion as to the possible 
effect of occasional errors in the choice of the samples 
is excluded; in other words, that hereditary values from 
1 to 5% or even of 5 to 20%, if they are maintained in 
sf>itc of selection, may be regarded as characteristic of 
half races, whilst values of 40 to 60%, when found in 
separate crops, may be taken to indicate the presence of 
intermediate races. 
When an intermediate race is isolated from an orig- 
inal sample of seed, it is separated from the half race 
