690 MUTANT RACES DERIVED FROM OENOTHERA 



vestigated. Besides them, I succeeded in getting a sufficient harvest 

 from four specimens with 16 or 17 chromosomes, derived from the 

 cross Oe. semigigas x {biennis x Lamarckiana) velutina and from two 

 plants of the cana type, from the cross Oe. semigigas x tardescens, 

 which behaved like them in their progeny, but whose chromosomes, 

 unfortunately, have not been counted. 



If the doubling of the chromosomes may be considered as an 

 indication of a specific mutation, we should conclude that in the 

 16- and 17-chromosome mutants, two or three chromosomes, respec- 

 tively, were in this condition. And if the application of the law of 

 probability to the distribution of the mutations among the egg- 

 cells of Oe. semigigas, as proposed by Boedijn and myself (1924 b), 

 holds good in these instances, we must expect that among the 

 progeny of such plants the two or three mutated types, respectively, 

 might separate and be seen each for itself. This is, however, not 

 always the case. The dominant type is repeated in about the expected 

 percentage, but the second form is either absent, or concealed in 

 a few specimens among the other mutants, or shows in too small 

 figures. Only in one instance it was plentiful, as will be seen in table 

 8. The third type, to be expected from the 17-chromosome mutants, 

 has not shown itself clearly in the experiments here to be recorded. 

 It seems probable that the second and third type suffer in a larger 

 degree from the mortality of the germs within the seeds of Oe. 

 semigigas, which is different for different categories, (See Opera VII, 

 p. 611). But why this should be so, remains an open question. 



In table 8 the type of the parent plants is introduced as "main 

 type", while of the other types, aside from Lamarckiana, that form 

 is considered to be a "second type", which occurred in the largest 

 number of individuals. The remainder are given under the head 

 of "mutants." In this table the mutants are given for each of the 

 cultures, since they might include a third type in one or more of 

 the four last-named cultures. The number of chromosomes was 16 

 in the first two and 17 in the following two parent plants. The 

 nanella and rubrinervis are mutants of the central group, having 

 only 14 chromosomes; therefore, they cannot be included in this 

 discussion. The other mutants were albida, auricula, pulht, pallescens, 

 oblonga, spathulata and semigigas. The chance of finding a "third 

 type" among the progeny of No. 3 and No. 4 is, however, very small, 

 since three occurred only one trisomic mutation, namely, auricula 

 in one of them, and none in the other. It is to be noted that the 

 second types of our table are secondary forms of the class of scintil- 



