192 
Allen. — Nuclear Division in the 
uneven thickness, is retained, as will be seen, until the time of the formation 
of the spirem. There is no such stage in the history of the pollen mother- 
cells of this species as is found by Miss Sargant (’97) and Mottier (’97) in 
other species of Liliwn , in which there appears a fine network of linin 
fibres, containing imbedded chromatin granules. 
The two differently staining substances that I have described are 
certainly not to be considered respectively as the chromatin and linin, 
which may be sharply differentiated later in the spirem ; it is more probable 
that the larger masses, which may correspond to the ‘ net knots ’ described 
by Flemming (’78, ’80, ’82) in the resting nucleus, contain both chromatin 
and linin, remaining for the present in these clumps derived from the 
chromosomes of the preceding division, and destined later to be distributed 
more uniformly along the spirem. This notion is confirmed by the fact 
that in iron-haematoxylin preparations of nuclei at the stage shown 
in Fig. 8, the knots are sometimes seen to be composed of dark granules 
imbedded in a more lightly stained substance. I have seen nothing at 
all comparable to Miss Sargant’s (’97) ‘ amorphous chromatin.’ 
An apparently similar construction of the nuclear reticulum is described 
and figured by K. and A. E. Schreiner (’04) in the primary spermatocytes 
of Myxine. In this case, however, each knot of the reticulum is formed by 
one of the fifty-two chromosomes, which is connected with its neighbours 
by linin fibres. This condition in Myxine becomes transformed, by the 
spreading out of the chromatin along the linin fibres, into a fine reticulum. 
The amount of nuclear sap is still (at the stage of Fig. i) relatively 
small. Masses of nucleolar matter are already present, varying in number 
and shape. In the figure two of these are shown, and they are quite 
rounded, but in many cases they are more irregular in outline. In this case, 
too, the nucleoles appear on the polar side of the nucleus, as though 
the material from which they have been re-formed had entered the nucleus 
from the side opposite that to which the spindle-fibres are attached. This 
is the position usually occupied by the nucleolar masses at a very early 
stage in the reconstruction of these nuclei. The nucleoles are readily 
distinguishable from the chromatic knots, not only by their size, but 
also by their staining properties. In preparations which have been exposed 
to the hydrogen peroxide for but a brief period, the nucleoles show little 
affinity for the safranin, but remain orange or yellow. If they have been 
bleached somewhat longer, they take on a bright red colour, which is 
still quite different from the dull red of the knots of the reticulum. The 
staining properties of the nucleoles seem also to be affected by the degree 
of penetration of the fixing fluid. 
The nucleus soon rounds itself up, as the old spindle-fibres disappear, 
and increases in size, apparently by the taking in of additional sap, since 
the amount of chromatic material increases slowly if at all, while its consti- 
