Pollen Mother -cells of Certain Plants. 21 
I was unable previously to study. Flemming’s mixtures of chromic, osmic, 
and acetic acids, although very favourable for the early prophases, have 
proved almost worthless for the post-synaptic stages of this plant. Good 
results were, however, obtained by first wetting the anthers in Carnoy’s 
solution, and then fixing in the usual manner in Flemming’s strong solution. 
This solution was employed with success in fixing the other plants studied. 
A one-per-cent, aqueous solution of platinum chloride was also used to fix 
some stages of Thalictrum. The sections, when necessary, were cut thick 
enough to contain entire nuclei. Being first mordanted in a two -per-cent, 
solution of potassium permanganate, the sections were stained with Flemming’s 
triple stain or Heidenhain’s iron-alum-haematoxylin combination. 
In common with Gregoire (’07), we may adopt, provisionally at least, 
the following scheme of phases for convenience of clearness in description. 
The prophases of division naturally fall into two periods, the pre-synaptic 
and the post-synaptic phases. The resting nucleus, according to some 
authors, is transformed into a mass of fine filaments. To this transformation 
period, Winiwarter (’01) gives the name ‘ leptonema ’. The fine filamentous 
stage described by some writers, in which the elements conjugate in pairs, 
Gregoire (’07) designates as ‘ zygonema ’. Synaptic contraction is charac- 
teristic of this period. The period of the thick, unsplit spirem is called 
‘pachynema’. During the stage of the thick spirem, which is of long 
duration, the so-called longitudinal splitting becomes evident. In other 
words the spirem distinctly shows its bivalent character at this time. Dixon 
(’96) has designated this phase as ‘ strepsinema ’. Whenever a second con- 
traction occurs it is during this period. The phase when the chromosomes 
are definitely separated by the segmentation of the spirem, and lie freely 
distributed in the nuclear cavity, usually on the periphery next the nuclear 
membrane, Haecker (’97) has called ‘ diakinesis’. We may also still retain 
the terms ‘ multipolar ’, ‘ polyarch,’ and ‘ multipolar diarch ’ as stages in 
spindle formation. 
DESCRIPTION OF OBSERVATIONS. 
I. Thalictrum purpurascens. 
In Thalictrum purpurascens , shortly after the last pre-meiotic divisions, 
the cells of the sporogeneous tissue and their nuclei assume the appearances 
characteristic of spore mother-cells. These cells undergo a growth-period 
in which they increase in size. Their nuclei also enlarge proportionately. 
The cells in one loculus of an anther show a remarkable uniformity in 
development, and usually all the mother-cells in all four loculi exhibit 
a striking uniformity as to their stages of development. 
I have attempted to trace the processes of reconstruction of the nuclei 
of the pollen mother-cells from the last pre-meiotic division, and to compare 
