556 DOUBLED CHROMOSOMES OF OENOTHERA 



semigigas into the discussion concerning the stage at which the 

 doubling of the chromosomes takes place. He counted its chromo- 

 somes and found them to be 21. From this he concluded that the 

 doubling must occur in the production of the sexual cells before 

 fecundation. The semigigas would then originate by the fertilization 

 of a doubled gamete by a normal one, whereas the conjugation 

 of two doubled ones would yield a gigas. This conception has since 

 been generally accepted, although it was in flagrant contradiction 

 with the older view of a doubling after fertilization, as had been 

 proposed by Strasburger. Ernst (9) has formulated the theory of 

 Stomps clearly, stating that in the production of the mother cells 

 of the pollen and the embryo sac, the reduction division is sometimes 

 omitted. This, of course, must be a rare phenomenon, and may be 

 considered the consequence of some internal mutation. 



On the basis of this view, the term doubling is rather a misleading 

 one. In the process of synapsis the chromosomes conjugate, and 

 at the time of the reduction division they quit each other. If this 

 latter event is omitted, they must remain in the combined condition 

 at least for some time, and thus cause the image of a doubling; 

 and if this doubling is considered to be a phenomenon of mutation, 

 we must assume that the separation of the paired chromosomes in 

 reduction is determined by some special factor, which may become 

 inactive in rare instances and from an unknown cause, thus pro- 

 ducing the mutation. This view is a direct consequence from the 

 theory of Stomps, and is interesting for our present discussion, inas- 

 much as it throws light on the relation between the process of 

 doubling and the external mutations which ordinarily accompany it. 

 This relation cannot be considered a causal one, but only one of 

 binding or linkage. In the description of our experiment we shall 

 find some direct proofs for this conception (0. Lamarckiana with 

 15 and 20 chromosomes). 



In the synapsis of semigigas, the doubled chromosomes are paired 

 with the single ones, and during reduction they separate, and one 

 of each pair goes to one pole and the other to the second pole. Stras- 

 burger pointed out that in this process the distribution of the chro- 

 mosomes depends upon chance. One nucleus will receive a larger 

 number of rods from the maternal group, and another more from 

 the paternal, and many may have equal groups of both. Ziegler 

 (13) then applied to this conception the laws of probability, showing 

 that the distribution must proceed according to these. Applying 

 this principle to the case of semigigas, we should expect its doubled 



