138 LUTHER BURBANK 



This Chilean race has been crossed with the com- 

 mon CEnothera acaulis, or Tarcucacifolia, and 

 produced a large number of intermediates, from 

 the best of which selection has been made. These 

 hybrids seem to come absolutely true in the sec- 

 ond generation, so far as foliage is concerned, 

 being in all cases intermediate between the two 

 species. This is perhaps what would have been 

 expected in a member of this race, in view of the 

 observations of Professor DeVries. The plant 

 seems to have an exceptional propensity to form 

 new fixed types. 



This, of course, is precisely the characteristic 

 that gives the plant interest from the standpoint 

 of the amateur experimenter. So a plot may 

 very well be set aside in the flower garden for 

 some evening primroses of two or three species. 

 Crossing will sometimes be effected by the in- 

 sects, if the experimenter does not care to take 

 the trouble to hand-pollenize the plants, and the 

 production of some interesting new forms may 

 be expected. 



Only two other common plants from among 

 the almost numberless ones that might be selected 

 are named as offering advantageous material for 

 selection by the amateur experimenter. But 

 these are about the commonest of all, and in some 

 respects among the most beautiful and interest- 



