422 I'ricotxlous Races. 



ISOLATION OF TRICOTVLOrS RACES. 

 HEREDITARY VALUES OF THE SECOND GENERATION. 

 FIRST 



GENER- 1 5 10 15 20 25 30 35 40 45 50 55 60 65 



Phaceliatanaceti folia 14%033113 8 6 4 2 2 2 



It is easily seen that in such cultures the choice of 

 specimens which belong undoubtedly to the intermediate 

 race is made possible. All plants with a low hereditary 

 \alue are simply rejected, since the possibility exists that 

 among them may be hidden In-ljrids between the two 

 races. The high values are free from this doubt and 

 indicate the pure race we are searching for. 



Thoroughbred races. Under this term those races 

 are included in agriculture which have been considerably 

 improved by stringent selection. In the first part of the 

 first volume we have seen that they remain de]:)endent 

 on continued selection, and do not really become constant. 

 For studying the features of such races, the tricotylous 

 intermediate races afford most suitable material, for, 

 after reaching the mean value of about 50%, we may 

 select the extreme variants, and, by their culture, improve 

 the race much farther. As instances I choose Mercurialis 

 annua, a large agricultural crop, and Silenc infJafa, which 

 I derived from seed introduced with cereals. Botli cul- 

 tures were started in 1892, and reached a hereditary 

 value of 55% in the harvest of 1894. From ])lants with 

 this value I raised a fourth generation, which oft'ered me 

 the following figures : 



