136 BOTANICAL GAZETTE [February 



Knowledge that in certain instances the reproductive cells of a 

 species are independent of mitosis for their origin must affect current 

 theories of heredity, which, since the renaissance of Mendel's law, 

 have leaned very heavily on the individuality of the chromosomes and 

 their separation in the reduction division. Child rejects the chromo- 



some 



theory in any universal application. He believes (p. 290) that 

 " these processes appear to consist essentially in the production of new- 

 nuclear material like that already present and without the periodical 



ery 



recurrence of metamorphosis. The act of division is ■ 



a mere incident of the increasing volume of substance." Accordingly 



he is inclined to doubt the constancy of the chromosome number in 



orm 



admit of certainty. In Synchytrium, likewise, the 



determination 



dogmatize. But in all the 



chromosomes could be counted 



the number seemed to be constantly four (cf. fig. 36b). The same 

 number was given provisionally by Stevens (13) in his first paper 

 and is shown by the drawings of his second paper (12). This matter 



mitoses 



been worked out in detail. But whether the chromosome number is 

 found to be constant or variable, it is obvious that our theories of 

 heredity will require considerable revkiW 



SUMMARY OF RESULTS 



The numerous peculiarities in the cytology of Synchytrium occur 

 stly in a somewhat definite ****** *f 2~»<~A~mi*i»* immediately 



■period of 



primary 



mitosis 



more 



This takes place by at least two processes : 

 Nuclear gemmation.~The karyosome of the parent nucleus 



mi 



brane, forms a vacuole and a membrane about itself, and becomes an 

 independent small nucleus, the whole looking like a budding yeast 

 plant This process is repeated until the parent nucleus is con- 

 verted into small nuclei, often forming a definite group. 



2. Heter 0S chizis.~The membrane of the parent nucleus dissolves, 



