I9i6] 



de vries— dimorphic m u ta n ts 



251 



on a sufficiently large scale, using the seeds produced by my pure 

 strain of O. la la fertilized by 0. Lamarckiana. The seeds of 1909 

 gave 564 seedlings, with 18 per cent lala and 2 per cent cana. Those 

 of 1908 gave 1550 seedlings, 8 per cent of which were lata and 9 per 



Fig. I. — Oenothera Lamarckiana mut. cana: a young plant showing the narrow 

 leaves by which it is easily distinguished from rosettes of the Lamarckiana type in 

 the same sowings; June 11, igiS- 



cent were cana. Other mutants appeared in these cultures in 

 different proportions, as usual. 



Among the seeds of pure 0. Lamarckiana, 0. cana is much the 

 rarer. In 1913 I fertilized, on 5 strong biennial specimens, almost 

 all the flowers during two months and got sufficient seed to have 

 20,000 seedlings in 1914. Of these only 6 were cana, giving a per- 

 centage of 0.03 per cent. In the same boxes 7 rubrinervis and 5 



