1897 ] THE LIFE HISTORY OF SALIX 151 
in the same anther. In S. glaucophylla and S. tristis the number 
of these intervening layers is usually two. Some anthers of S. 
cordata with the pollen grains nearly mature showed no layer at 
all between the endothecium and the tapetum (fig. 8). The 
cells of the mature tapetum often have two nuclei. The 
strengthening of the endothecium does not commence until the 
tapetum has begun to disorganize. All of the layers between 
the endothecium and the spores disintegrate, and the spores 
float in a granular fluid ( fg. 9) 
The sporogenous cells, as shown in figs. 7 and 2,are the mother 
cells of the microspores. This is proved by the fact that the num- 
ber of sporogenous cells in a transverse section of an autumn or 
winter microsporangium is approximately the same as the number 
found in spring microsporangia whose sporogenous cells are 
beginning to show by their spherical form that they are undoubted 
mother cells. The large size of the nuclei also favors this 
interpretation. An examination of several species indicates that 
most staminate buds pass the winter in the spore mother cell 
stage. In buds of S. s#istis, collected jJate in March, the spore 
mother cells had not yet assumed the spherical form. 5. cordata 
and S. glaucophylla, collected at the same time, had already 
passed the tetrad stage. 
The nucleus of the microspore divides some time before the 
spores are shed. The division of the nucleus is not followed 
by the formation of a cell wall. 
Salicacee, a wall is formed 
separating a smaller lenticular 
cell from the larger one. In 
Salix the generative nucleus 
Soon organizes a part of the 
Surrounding cytoplasm and 
_ becomes a fusiform cell. Since 
Spores already upon the stigmas — 
_ Showed no further differentia— ——— hee 
