608 DR A. ANSTRUTHER LAWSON ON 
shortness of the threads, they seem to have more freedom to move and separate from 
one another. So that the cubical space now occupied by the chromatin (fig. 7), instead 
of having decreased, has considerably increased. This continues, as shown in figs. 
8, 9, and 10, until the spaces between the spiremes are fairly equalised and the 
chromatin area completely fills the nuclear vacuole. 
We might regard fig. 7 as a stage representing the close of the growth period, 
2.e. when there appears to be no further increase in the amount of nuclear sap. The 
nuclear vacuole has reached its maximum size. A close study of the chromatin up 
to this time (from figs. 1 to 7) failed to reveal any real evidence in support of the 
view held by some cytologists that the chromatin consists of a continuous spireme 
thread. I was unable to find any indication that the chromatin filament ‘‘ segmented 
transversely so as to form definitive chromosomes.” My present observations in 
every way confirm the opinion expressed elsewhere (Lawson, 1911) that the chromatin 
thread—in Smilacina—never forms a continuous spireme, but that there are as many 
spiremes as there are chromosomes. In this connection I am quite in agreement with — 
the conclusions of Stomes (1910) in his work on reduction in Spinacea. I can also 
confirm this author's observations that the free ends of the chromatin threads may 
be distinguished as far back as the reticulum stage represented in fig. 2. 
When the spiremes have become distributed throughout the nuclear cavity, the 
individual threads may be studied to better advantage in sections that have not 
been cut in a median plane through the nucleus. Such sections are shown in figs. 
8,9, 10,and 14. From these figures it will be seen that although the spiremes are 
now comparatively very much condensed, they are still of great length. Their double 
nature may be easily seen. If we follow the series of stages back, it will become 
perfectly evident that this double nature is not due to the lateral fusion of filaments 
during the growth period, but is due to the same longitudinal fission described for 
the earlier stages and represented in figs. 1, 2, 3, 4, and 5. More highly magnified 
details of the longitudinal fission in older spiremes are shown in figs. 11, 12, and 13. 
Here it will be seen that the fission by no means completely separates the two 
longitudinal halves, for the latter are in frequent contact (figs. 11 and 18). 
The beaded nature of the spireme is very noticeable in the stages represented 
in figs. 8,9, and 10. The disks or short segments which give this beaded appearance 
are no doubt the structures described as chromomeres by ALLEN (1905). ‘This feature 
has been noted by several other writers, and I have observed it in many plants. It 
seems to be of fairly general occurrence, but up to the present no facts have been 
revealed to indicate the significance, if any, these chromomeres have in the mitotic 
process. In the material under observation it was quite clear that this feature was 
not of long duration. It gradually fades out as shown in figs. 14 and 15, and in 
fig. 16 no trace of the chromomeres could be distinguished. 
In fig. 15 we have represented a section that has been cut in a median plane— 
* Farmer, J. B. (1905), 
