Pollen Mother -cells of Lilium canadense. 19 r 
until we reach the second point of greatest progress. The variation within 
one sac is, however, extremely small. The same general statement is true 
as- to a uniformity between different sacs of the same anther, and different 
anthers of the same flower, but in these cases there is a somewhat greater 
range of variation as to the stages represented. 
I have attempted to follow carefully the processes concerned in the 
reconstruction of the daughter-nucleus of the last pre-heterotypic division, 
and its growth to the size characteristic of the nucleus of the pollen mother- 
cell. The daughter-chromosomes of the mitosis just mentioned aggregate 
into a dense mass at either pole of the spindle. After the formation of the 
new nuclear membrane, nuclear sap appears within it, and the mass of 
chromatin becomes less dense. The chromosomes are now seen to have 
lost their regular outline and to have become reticulated, so that it is 
impossible to trace in the newly-formed nucleus the outline of any 
individual chromosome. Fig. 1, PI. VI, shows the nucleus of a young pollen 
mother-cell at the time when cell-division has just been completed by the 
splitting of the cell-plate and the formation of a thin, orange-staining wall. 
The old spindle-fibres (not shown in the figure) are still attached to the 
nucleus, and on the side to which they are attached (the lower side in 
the figure) the outline of the nucleus is flattened and still quite irregular, 
while on the opposite (polar) side it is convex and more evenly rounded. 
The nucleus, as a whole, is still small and considerably flattened, its 
shortest diameter coinciding with the axis of the spindle. The chromatic 
material is massed on the equatorial side of the nucleus. It is broken 
up into irregular masses, the shape of some of which suggests that of 
the chromosomes which were their source. The larger masses show an 
affinity for the safranin stain, and are connected by blue-staining fibres of 
varying thickness, down to that of very fine threads. The blue-staining 
fibrous material also forms short threads attached to various portions 
of the red-staining masses, the whole effect being that of a very ragged and 
irregular reticulum occupying the greater part of the nuclear cavity, 
especially on the equatorial side. The colours of the constituents of this 
network vary, of course, with the time of exposure to the different stains, 
also to some extent with the time of exposure to the bleaching reagent and 
with the degree of penetration of the fixing fluid. Preparations may be 
obtained in which the whole reticulum is violet, others in which it is 
all stained orange-red. Even in these cases, however, a certain difference 
is observable between the affinity of the substance of the larger masses and 
that of the fibres for the different stains, and by careful regulation of 
the times of exposure a fairly distinct differentiation may be secured. 
The substance of the larger masses also shows a stronger affinity for the 
haematoxylin than does that of the fibres. This construction of the nuclear 
reticulum, out of rather red-staining masses and blue-staining fibres of 
