192 BOTANICAL GAZETTE [MARCH 
equal distribution of substance to the two daughter nuclei, to 
assume that this arrangement of granules into an apparently 
double strand results from the splitting of a single thread or 
series of granules, but the very minuteness of the granules in the 
stages preceding the formation of the spireme precludes any 
direct observation of such a division even if it takes place. 
There are two ways in which one might explain the forma- 
tion of quadripartite threads out of the network of the quiescent 
nucleus. We might imagine that in a row of self-propagating 
units each divides into two, the result being two parallel rows, 
and that then each of the resultants again divides into two, thus 
giving rise to the groups of four which are joined into strands 
by the linin. Judging, however, from what it is possible to 
observe in the quiescent nucleus and in the later stages of divi- 
sion, there is some evidence in favor of another explanation. 
The grouping of the granules into pairs, which are at least double 
and probably quadruple, can be seen in early stages back to the 
point where the granules are so fine and so closely massed 
together that they cannot be said to have any definite arrange- 
ment. It is conceivable, therefore, that such a tetrad condition 
may be permanent, the granules having an inherent tendency to 
group themselves into fours. 
An examination of the large cells of the plerome, which by 
their extreme size and presence of large vacuoles indicates the 
loss of the power of division, yields evidence that the tetrad 
condition may be permanent. In fig. 50, taken from a large ple- 
rome cell which is later to break down to form a central vessel, 
and fig. 52, of cells elongating to form wood fibers, fixed in 
Flemming and stained by the iron alum haematoxylin method, 
the chromatic network has the appearance of strings of vacuoles, 
the borders of the vacuoles consisting of the chromatin sub- 
stance. A comparison of these figures with those of the early 
prophase, such as figs. 7 and 8, shows similarity of structure. 
The condition seen in the plerome cells may be interpreted as 
the result of the linin connecting the successive groups etrads 
having become more stretched, the tetrad granules less conspicu- 
ous; the space in the center of each group of granules acquiring 
