338 BOTANICAL GAZETTE [November 



instance, therefore, affords a means for the experimental study of 

 the origin of such a form by mutation. The main result of this 

 study is the proof of the occurrence of mutant MendeHan hybrids 

 besides the pure dwarfs. 



In my book on the mutation theory I have pointed out that the 

 origin of 0. brevistylis from O. Lamarckiana may have been induced 

 by the mutation of a single sexual cell. If this combined in fertih- 

 zation with a normal gamete, a hybrid would be produced which 

 would not be distinct from the parent species in its external features. 

 This hybrid would then, in its self-fertilized seeds, follow the law of 

 Mendel and produce, besides constant Lamarckiana plants, partly 

 hybrids of the same type and partly specimens of the type of 

 0. brevistylis. From this origin and the subsequent free intercross- 

 ing in the field, the yearly appearance of O. brevistylis would receive 

 a sufiicient explanation (DeVries, op. cit., p. 506). 



If the process of mutation into this type were more often 

 repeated, it should be possible to discover the original hybrids. 

 They would, it is true, not be discernible from their normal sisters 

 by external marks, but would yield, after artificial self-fertiHzation, 

 about 25 per cent of brevistylis. And since mutants are produced 

 ordinarily in a proportion of 1-2 per cent or less, the difference 

 would be large enough to be noticed. Until now, however, such 

 cases have not been observed. 



I have, therefore, been looking for another example in which a 

 MendeKan behavior of the mutants might be associated with a 

 normal coefficient of mutation from the parent species. Such 

 cases would betray themselves by exceptionally high coefficients 

 in single parent plants. Instances of such individual deviations 

 are very rare, partly on account of the necessarily limited number 

 of mother plants from which the seeds of our cultures are taken. 

 But ScHOUTEN^ has observed that Oenothera gigas, which ordinarily 

 produces 1-2 per cent dwarfs, may be seen to throw them off in as 

 large a number as 15 per cent. The same phenomenon has been 

 described by Gates {pp. cit., p. 137), who counted 9 per cent and 11 

 per cent of dwarfs among the offspring of two self -fertilized plants 

 of 0. gigas. 



3 ScHOUTEN, A. R., Mutabiliteit en Variabiliteit. 1908. 



