The Isolation of Tricofylous Inicnncdiatc Races. 443 



A. Tricotylous offspring 



B. Dicotylous offspring -I 21 22 24 25 25 



[ 18 20 



The group is therefore dimorphic, the intermediate 

 race is not yet pure. I found the same conditions to 

 obtain in other cases. 



In Clarkia pidchella also, the atavists reverted when 

 sown in 1897 alongside the tricotylous culture already 

 mentioned and from seed of the same parent. This 

 parent already had a hereditary value of 64%, but was 

 the first in this race with a high figure, and therefore 

 still belonged to the transitional period. The values are 

 arranged in groups of 5 to 15, 15 to 25 etc., and the 

 number of individuals which fall into each group are 

 given under the mean percentage number, i. e., 10, 20 etc. 



Proportion of tricotyls in percentage 10 20 30 40 50 60 70 80 

 Number of tricotylous individuals 02588 12 3 1 

 Number of atavistic individuals 4 13 911 2 — — 



The mean for the tricotylous individuals is 49%, for 

 the atavists 30%. 



The race once isolated, the atavists obviously still 

 have smaller values than the tricotyls ; but the difference 

 is never more than a small one, and the curve for the 

 two races together will have only one apex. In Oeno- 

 thera hirtella I examined a series of values of the atavistic 

 offspring as well as the tricotylous offspring of the same 

 parent of 1896 which had a value of 56%. This parent 

 had been self-fertilized, the visits of insects havine been 

 excluded : and so were all its offspring. I shall group the 

 values in tlie same wav as above. 



