Nuclear Division in the Pollen Mother-cells 
of Lilium canadense. 
BY 
CHARLES E. ALLEN, 
Assistant Professor of Botany in the University of Wisconsin. 
With Plates VI, VII, VIII, and IX. 
HE investigations to be described in the present paper were carried on 
X upon the pollen mother-cells of Lilium canadense , L., which is 
abundant in this vicinity. Some study, for purposes of comparison, 
was made of the pollen mother-cells of L . tigrinum , Andr., and L. longi- 
flornm y Thunb. So far as they went, my results with these species were 
entirely in harmony with those obtained from the first-named species, 
in which alone, however, a complete series of stages was studied. All 
the figures and descriptions have reference to L. canadense. 
I shall speak, for the most part, only of the behaviour of the nuclear 
substances, with especial reference to the history of the chromosomes. 
Some of the more important results have already (Allen, ’04) been briefly 
announced. The process of spindle-formation, which I hope to describe in 
a future paper, agrees in general, though with some interesting variations 
in detail, especially in the second division, with the course of events as 
I have (’03) observed it in the pollen mother-cells of Larix. 
Methods. 
A number of fixing fluids were tried ; the best results were obtained 
from Flemming’s stronger chrom-osmic-acetic acid solution, and from 
the slight modification of this formula used by Mottier (’97). There 
seemed to be little choice as between the two. Some good preparations 
were obtained from material fixed in Flemming’s weaker solution, but 
as a rule this fluid produced more or less plasmolysis and distortion of 
the cell. 
Material was fixed in the field from day to day during the season 
of development and division of the pollen mother-cells. Anthers containing 
[Annals of Botany, Vol. XIX. No. LXXIV. April, 1905.] 
