350 BOTANICAL GAZETTE [NOVEMBER 
making them entirely over again. It seems quite possible that in the 
diffuse vegetative reticulum the chromosomes are as completely fused 
as the molten metal, and until we know more than we do now con- 
cerning the nature of the reticulum, this alternative possibility should 
be kept before us. The analogy of the molten metal is more suggest- 
ive in cases like Synchytrium, where the chromatin is partially or 
completely concentrated in a karyosome from which part or all of 
the spirem is directly derived. Although cytologists have as a rule 
paid but little attention to their behavior in mitosis (cf. WAGER 19), 
such structures are common occurrences in many plants and animals. 
If we may infer safely, as I believe we may, that the chromosomes 
of Synchytrium decipiens number constantly four, it becomes a matter 
of primary interest to determine how that constant number is main- 
tained in all of the amitoses through which the nuclei pass. It has 
long been assumed that the principal if not indeed the sole function 
of the complicated process of mitosis was to insure exactly equal 
division of the chromatin between the daughter nuclei. In view of the 
fact, however, that the majority of the nuclei of Synchytrium are 
derived by some form of amitosis either directly or through their 
ancestors, and yet maintain the number of their chromosomes con- 
stant, it is evident that mitosis is not necessary to maintain this 
number. 
The possibilities of an exact mechanical division of the chromatin 
are somewhat different in the different varieties of amitosis. In 
heteroschizis there occurs a metamorphosis of the nucleus suggestive 
of that of mitosis. In the loss of the nuclear membrane and the 
apparent cessation of interaction between the nucleus and the cyto- 
plasm there may be a pause during which the chromatin is divided 
granule by granule in such a way that the karyosomes of the daughter 
nuclei are furnished with exactly equal chromatin content, just as is 
visibly accomplished by the fission of the chromatin granules or the 
chromosomes in mitosis. 
Likewise it is not inconceivable that the karyosome may be divided 
in a similar manner when it is broken up preparatory to nuclear 
gemmation, though in this variety of amitosis there is no loss of the 
nuclear membrane or other indication of a pause in metabolism. 
An exact equational division of the chromatin under such circum- 
