624 Wolfe . — Cytological Studies on Nemalion . 
prophase is far in excess of the chromosome-number characteristic of that 
generation, sporophyte, or gametophyte, respectively, in which the division 
occurs. Since the diameter of the nucleus does not exceed 3 n, it is 
evident that an exact enumeration is a very difficult matter, and, in fact, 
in the earlier divisions of the gonimoblasts estimates of the number of these 
masses varied from twenty to thirty, and appeared to be correspondingly 
less in the gametophyte. Inasmuch as the number of chromosomes 
eventually appearing is less than the number of granules by about half, 
it is not impossible that these granules may be chromatic units which 
may fuse two and two at the time of their organization into morpho- 
logical chromosomes, although the details of this process could not be 
determined. 
The chromosomes are rounded bodies which stain intensely, and are 
thus, notwithstanding their small size, fairly distinct. From a number of 
estimates very carefully made and repeatedly confirmed, the writer feels 
entirely safe in stating that in the divisions of vegetative nuclei they are 
about eight in number ; certainly not less than seven, nor more than nine 
(Figs. 29 and 75). 
In the cystocarp the details of mitosis could not be followed for the 
first two or three divisions of the zygote, owing to the fact, previously 
mentioned, that the cytoplasm at this time stains so intensely. In all 
divisions subsequent to this period, however, mitotic figures were observed 
in a variety of stages and in great abundance. In the earlier divisions in 
which the mitotic figures could be clearly differentiated, the chromosomes, 
owing to their crowded position upon the nuclear plate, form at times an 
almost continuous band (Fig. 65), and in general their number could not be 
ascertained with the same degree of certainty as in the vegetative phase. 
Owing to the fact that these preliminary aggregations of chromatin, illus- 
trated in Fig. 56 in the prophase stages, might possibly be mistaken for 
polar views of the spindle at metaphase, polar views were taken as cor- 
roborative only, and such approximations alone relied upon as could be 
obtained from longitudinal and oblique views in which, since both centro- 
somes were present and the spindle clearly differentiated, it could be 
definitely determined that the bodies in question were morphologically 
chromosomes. A great number of estimates were made, based upon such 
stages, and certainly the number is greatly in excess of that characteristic 
of the vegetative phase, approximations varying from twelve to sixteen, 
Avith the weight of probability in favour of the latter number (Figs. 
61-65). The decided increase in breadth of the spindle, which has pre- 
viously been noted as characteristic of the earlier mitoses in the cystocarp, 
further indicates the presence of a greater number of chromosomes in these 
divisions. In this connexion, perhaps, it should be stated that the 
phenomenon of ‘splitting’ presents characteristic features and cannot be 
