464 BOTANICAL GAZETTE [JUNE 
because of the inconstancy in amount of salt in a salt-marsh during the 
growing season.— M. A. CurysLerR, Zhe University of Chicago. 
CHROMOSOME REDUCTION IN LILIUM CANADENSE. 
THE investigations, a brief résumé of whose results is given below, 
were carried on upon the dividing pollen mother-cells of zim 
canadense \.., collected in the vicinity of Madison during the past five 
summers. A fuller account, with figures, is practically completed, but 
as it must be some time before it appears in print, and as the results are in 
some points quite different from those obtained ‘by any previous 
observer, it seems advisable at this time to publish a brief statement of 
my observations. 
After the completion of the division which forms the pollen mother- 
cell, there is a long period during which the cell and nucleus increase 
greatly in size. The nucleus contains during this period, in addition 
to the nucleoles, numerous irreguiar masses of considerable size, con- 
nected by narrow strands or fibers, the whole forming an extremely 
irregular network. In preparations stained with Flemming’s triple 
stain, the larger masses show an affinity for the safranin, the fibers for 
the violet. There are also numerous short, fine, blue-staining fibers 
attached to the larger bodies, giving the network a ragged appearance. 
This general arrangement persists until just before the passage of 
the nucleus into the condition of synapsis, when the blue-staining 
fibers begin to grow longer and to become more uniform in thickness; 
at the same time the larger masses or knots decrease in size. While 
these changes are going on, it is evident in many portions of the net- 
work that two fibers lie side by side and parallel; sometimes such 
parallel strands are attached at their corresponding ends to the same 
red-staining mass. While this rearrangement or pulling out of the 
nuclear material into threads is going on, all of the chromatic nuclear 
substances become massed against one side of the nuclear membrane, 
resulting in the synaptic figure so often described. The formation of 
the spirem is not fully completed until after the occurrence of this 
eccentric massing; there is, therefore, in this case no “dolichonema” 
stage preceding synapsis. 
As has been said, while the spirem is being formed, it is seen in 
many places to consist of two parallel threads; and this continues to 
be the case until all of the staining substance within the nucleus, 
excepting the nucleoles, has been distributed along the spirem. In 
