84 BOTANICAL GAZETTE [AuGcusT 
METHODS. 
These were essentially the same as those used in my work on 
Cobaea and Gladiolus. As full details are recorded in these 
papers, it will be unnecessary to repeat them. It is only neces- 
sary to add that the material was all fixed in the field, and by 
far the most satisfactory results were obtained by using Flem- 
ming’s weak solution of chromic-osmic-acetic acid as a killing 
agent, and the triple stain, safranin, gentian, and orange G. 
THE POLLEN MOTHER-CELLS OF IRIS FLORENTINA. 
The young anthers of the common garden Iris furnish very 
good material for the study of spindle formation. If conditions 
are favorable and the material is fixed in the field, immediately 
after being detached from the plant, all the stages in the forma- 
tion of the spindle of the first division of the pollen mother- 
cells may be observed in a single anther. As the anthers at this 
stage are very large, many sections may be obtained from one 
of them. Previous to the formation of the first spindle, the 
nucleus of the mother-cell is very large, containing one or two 
nucleoli and the chromatin thread. The cytoplasm appears to 
be a reticulum, the threads of which are more or less granular. 
In the immediate neighborhood of the nuclear wall, the cyto- 
plasm is more dense than at the periphery, but this dense region 
is not as sharply differentiated as that described by the writer 
for this stage in the pollen mother-cells of Cobaea and Gladiolus. 
Numerous small spherical bodies, probably oil globules, were 
found distributed irregularly through the cytoplasm. 
While the chromatin is yet in the spirem stage, the cytoplasm 
immediately in contact with nuclear membrane becomes differ- 
entiated into a distinct weft or felted zone of fibrils. This weft 
appears at first to consist of a few delicate but very distinct fibrils, 
which stain blue with the gentian violet, in contrast to the slightly 
orange color of the rest of the cytoplasm. They do not radi- 
ate from the nuclear membrane, but lie more or less parallel to 
it. When followed along their course, they were found to lose 
gradually their affinity for gentian violet, and to terminate in the 
regular orange-colored threads of the cytoplasm. This indicates. 
