430 



Tricotylous Races. 



tween 16 and 79% and gave a mean of 49%. The race 

 could therefore be regarded now as perfectly pure. 



Hclichrysiuii hractcatuni conipositum, Fig. 86. From 

 the same set of sowings as that which contained Clarkia, 

 I planted out some tricotylous seedlings of HclicJirysiun. 

 There was a relatively large number of them, and alto- 

 gether 19 set seed. For each one the value was calcu- 



Fig. 85. Clarkia piilchcUa alba. A flowering sprig. 



lated se])arately. In the case of 15 plants these were 

 distributed between 2 and 8% and exhibited a mean of 

 4% ; but there were higher figures besides, viz., 12 — 12 — 

 16 and 41%. The latter plant was obviously a repre- 

 sentative of the intermediate race sought for; the re- 

 mainder were thrown away as doubtful, although if 

 there had not happened to be such a faA^orable plant 



