128 BOTANICAL GAZETTE [february 



cluster (fig. jj) or in a looser group (fig. 18) . The origin and fate of 

 these small nuclei is the subject of the present paper. Although such 

 variations in the size of the nuclei are sometimes found in almost any 

 stage of the period of nuclear division, they are most conspicuous 

 immediately after the division of the primary nucleus and continue 

 prominent until there are 200-300 nuclei in the cyst. It is at this 

 same stage that the other peculiarities in the cytology are most pro- 

 nounced. While this period of irregularities is not sharply marked 

 off from the succeeding phases of the life-history, yet as the nudei 

 become more and more numerous there seems to be a tendency for 

 them to settle down, so to speak, and to conform more nearly to the 

 usual habits of dividing nuclei in growing tissue. 



The isolation of these groups suggests that their constituent nuclei 

 have a common origin. Because of the absence of any pairing, and 

 because of their great differences in size, one is inclined to suspect 

 that they have been derived by some process other than mitosis. Since 

 mitosis in this plant is always simultaneous, involving all the nuclei in 

 a cyst, the differences could not be due to the failure of some nuclei to 



division. 



became smaller 



■ated 



some process of mito 



num. 



from 



which the products were unequal, as in the reduction division of an 

 animal egg. But all the mitoses observed gave rise to equal daughter 

 nuclei. Further, mitoses in cysts of this age are uncommon. Th 

 led Stevens (12) to suggest the possibility of an amitotic origin for the 

 nuclei of this stage. 



There are several processes of direct nuclear division in Synchyt- 



the commonly observed 



division by an amoeboid constriction of the parent nucleus. ^ nli 

 they may be considered under the general term amitosis, which la 

 come to include several forms of non-mitotic division, they require dis- 

 tinctive terms for their designation. Indeed, there is considerable 

 need for a classification of the different forms of direct division. 

 especially in view of the increased importance amitosis is likely to 

 assume in future cytological discussion. The first process, wbj* 

 cons,sts of a budding-out of a small nucleus from a larger, may be 

 designated nuclear gemmation. The second differs from ordinal 

 amitosis in that ihe nucleus loses its membrane and vacuole of W* 



