Pollen Mother -cells of Lilium canadense . 195 
nuclear membrane ; and this flattening of the nucleole or nucleoles becomes 
more pronounced as the aggregation of the other nuclear materials becomes 
closer. The aggregation is at first more commonly in the central portion 
of the nuclear cavity, and has no apparent relation to the eccentric position 
of the nucleole ; but later the whole mass moves over to one side of the 
nucleus (Figs. 12, 14), usually to that side already occupied by the nucleole, 
and then the appearance is as if the nucleole were pressed against the 
membrane by the synaptic mass. 
As the material of the reticulum becomes more and more densely 
aggregated, the fibres grow in length at the expense of the knots, which 
finally disappear entirely (Fig. 12) ; the fibres become quite uniform in 
thickness. The nuclei represented in Figs. 10 and 11 are from sections 
respectively twenty and thirty micra in thickness ; the nuclei in question were 
uncut, and the free ends of fibres which appear in them are not due to the 
cutting of strands. At this time, therefore, a continuous spirem has not 
been formed. But at the stage shown in Fig. 12, no such free ends are to 
be observed in an uncut nucleus. It is impossible to say with certainty that 
no free ends are present in such a dense mass as is shown in this figure ; 
but the fact that they never appear in any of the strands that pass out from 
the peripheral portion of the mass makes it extremely probable that there 
is now present a continuous spirem. 
The most important fact to be noted in connexion with the spirem at 
this time (Figs. 12-14) is that it is plainly composed of two slender threads 
lying side by side. This double nature of the spirem is best distinguished 
in relatively thin tangential sections of the nucleus, such as that shown in 
Fig. 13, rather lightly stained with iron-haematoxylin. Often the two 
threads run closely together for a considerable distance, sometimes loosely 
twisted about each other ; at other times they are in contact and appear 
to be fused; in some places they diverge more or less widely. Fig. 12 
represents a tangential view of a nucleus whose chromatic materials are 
massed at the surface turned toward the observer. Here on account of the 
density of the mass the construction of the spirem-thread is not so easily 
determined ; but all the portions which are disengaged from the main 
mass have the same double structure as appears in Fig. 13; and the same 
structure is occasionally visible even in the interior of the synaptic mass. 
For the most part the paired threads within the mass are in close contact, 
if not actually already fused, giving the appearance of a single strand of 
double thickness. In a thin median section, such as that shown in Fig. 14, 
less evidence appears of the double nature of the spirem ; though even in 
such a section the loops of the spirem, which occasionally run out from the 
synaptic mass into other portions of the nuclear cavity and back again 
(as the one shown at a\ are almost invariably seen to be composed of two 
parallel or twisted threads. From what has been said of the process of 
