I9i6] 



DeVRIES— DIMORPHIC MUTANTS 



273 



described in that book as 0. cruciata Nutt. The other names are 

 still the same as in my book. 



With the same purpose crosses were made with O. Lamar ckiana 

 and some of its derivatives. 



TABLE X 

 Crosses or 0. cana with 0. Lamarckiana and its mutants 



The main result is the same. In all the crosses of both tables 

 the characters of O. cana are handed down through the ovules to a large 

 part of the progeny, but not through the pollen. The behavior is 

 exactly the same as in O. scintillans and 0. lata. The two speci- 

 mens of 0. cana from the crosses of 0. lata and 0. nanella must 

 evidently be considered as mutants, that is, as having arisen from 

 the fertilization of mutated sexual cells, since we have seen that 

 such mutations occur from time to time, especially among the 

 seeds of 0. lata. 



Apart from the appearance of plants of the cana type, the 

 results of the crosses are, in every case, such as would be expected 

 if 0. Lamarckiana had been used instead of 0. cana. In this 

 respect they simply confirm the conclusions given in my book. 



Let us now consider in its details the analogy of 0. cana with 

 the aUied forms of 0. scintillans and 0. lata. Two cases offer 

 themselves for this consideration. The first one is afforded by 

 the crosses with 0. biennis. Apart from stray mutants, these 

 produce two types, one of which combines the visible marks of 

 both parents, whereas the other wholly lacks the characters of the 

 mother, but is simply hke the hybrid of 0. Lamarckiana and 

 0. biennis, as described in my book. In the first group the com- 

 bination is such as to make the characters of the mother the most 



