1897 ] THE LIFE HISTORY OF SALIX 171 
1. Organogeny of the flower.—Pistillate buds, collected in 
August show the carpels outlined but no trace of ovules. October 
buds of S. glaucophylla and S. cordata show the nucellus, but the 
integument as a rule does not appear until spring. Staminate 
buds collected in October show the stamens well outlined. The 
nectaries in both staminate and pistillate buds can be seen in 
October. A diligent search failed to reveal the slightest trace 
of rudimentary floral organs, which those who regard Salix as a 
reduced type might expect to find. 
2. Development of the microspores A comparison of autumn, 
winter, and early spring buds shows that most stamens pass the 
winter in the spore mother cell stage. The division into gener- 
ative nucleus and tube nucleus takes place before the tapetum 
breaks down. The generative nucleus soon organizes a part of 
the surrounding cytoplasm and becomes a fusiform cell. No wall 
is formed between the nuclei. Populus monilifera differs in this 
respect, a definite wall separating the two cells. The cells of the 
tapetum are often binucleate. 
3. Origin of the macrospore—The macrospore has its origin 
in a hypodermal cell at the apex of the nucellus. Sometimes 
there are two or three archesporial cells, but it is very seldom 
that more than one develops. The primary tapetal cell usually 
gives rise toa tierof three or four cells, but sometimes does 
not divide. The macrospore mother cell usually cuts off one or 
two potential macrospores, but sometimes germinates without 
cutting off any such cells. This variation is prevalent in the 
genus. | 
4. Germination of the macrospore—The first division of the 
primary nucleus of the macrospore is transverse. In the second 
and third divisions the spindles at the micropylar end are trans- 
verse, while the spindles at the antipodal end are longitudinal. 
After the first division, development proceeds with increasing 
rapidity until the female gametophyte has reached the fertiliza- 
tion period. Great difficulty was experienced in demonstrating 
the presence of antipodal cells, several hundred macrospores, just 
before the fertilization period, yielding only six cases of 
