Gates, Tetraploid Mutants and Chromosome Mechanisms. J47 



that his material presented more irregularities in chromosome 

 behaviour, which is probably due to his material having been col- 

 lected vjry late in the season, while mine was collected in the 

 height of flowering. 



8. The various types of chromosome change in Oenothera, 

 based on my own observations together with those of Geerts 

 and Davis, are 1) in diploid races an irregularity in the distri- 

 bution of the heterotypic chromosomes, resulting from the weak 

 attraction between homologous chromosomes, 2) a similar irregula- 

 rity (certainly much less frequent if it ever occurs) in the distri- 

 bution of the halves of the homotypic chromosomes. The first of 

 these is sufficient to account for the occurrence of 15-chromosome 

 mutants, such as lata, or 16-chromosome forms. 3) The omission 

 of chromosomes from interkinesis. 4) The union of a diploid egg 

 with a haploid pollen grain. This is sufficient to account for the 

 occurrence of all triploid mutants. 5) Possibly the latter may also 

 arise by the union of a haploid egg with a diploid pollen grain, 

 though there is at present no direct evidence for this. 6) Doubling 

 of the chromosome number through a suspended mitosis in the 

 megaspore mother cell, followed by apogamous development. This 

 is sufficient to account for the origin of all tetraploid mutants, but 

 they may also arise through 7) a suspended mitosis in the nor- 

 mally fertilized egg, or 8) the union of a diploid egg with a diploid 

 pollen grain, though there are no direct observations in support 

 of 7) or 8). 



Chromosome numbers intermediate between the tetraploid, tri- 

 ploid and diploid may arise by crossing, and also 9) by irregular- 

 ities in the heterotypic chromosome distribution in the tetraploid 

 or triploid races, 10) by chromosomes being left behind in the 

 heterotypic mitosis, 11) by chromosomes being omitted from the 

 homotypic nuclei or fragmenting during meiosis. 12) By the failure 

 of certain chromosomes to divide in the homotypic mitosis. To 

 these we may add perhaps 13) the loss of chromosomes in the 

 nuclear divisions of the male and female gametophytes, and 14) the 

 loss of chromosomes during the divisions of the fertilized egg. All 

 these processes except 2), 5), 7), 8), 13), and 14) are based on 

 observations. 



9. In the valuable paper of Nilsson he assumes that 0. yigas 

 has arisen through the accumulation of many independent quanti- 

 tative factors for the size and form of various organs, an hypothesis 

 which is inadmissible because it is contrary to the cytological facts 

 as well as the facts regarding the sudden discontinous origin of 

 the giant types and their subsequent wide variation. This varia- 

 tion is doubtless due in part to loss of chromosomes and in part 

 to the fact that gigns contains the capacity of producing the same 



10* 



