- 1897] THE LIFE HISTORY OF SALIX 153 
into two cells, a smaller one nearer the micropyle, and the larger 
one which becomes the fertile macrospore. The smaller cell 
either undergoes one transverse division, thus giving rise to two 
potential macrospores, or it does not divide at all ( figs. 1g and 
77). Ina case like fig. 77 there is a possibility that the two 
| smaller cells may have been cut off in succession from the larger 
| cell, but as no mitotic figures were found in this stage this 
question could not be settled. 
Sometimes the macrospore mother cell does not divide but 
develops directly into the macrospore (fig. 23). If any potential 
macrospores have been cut off, they are crowded and absorbed 
by the growing fertile macrospore until nothing remains of them 
but a refractive cap, and even this soon disappears. These varia- 
tions are noteworthy. In Gamopetale it is said (21) to be the 
rule that the macrospore mother cell becomes the macrospore 
directly; in monocotyls and in the Archichlamydee among 
dicotyls the macrospore mother cell gives rise to four potential 
macrospores, one of which becomes the fertile macrospore. In 
some plants there are three potential macrospores; in others 
there are two; in still others the macrospore mother cell 
becomes the fertile macrospore without any division. Standard 
texts, as well as the original papers from which their informa- 
tion is obtained, leave the impression that for a given species 
there is little or no variation in the mode of origin of the macro- 
spore, and I must confess that as far as the number of poten- 
tial macrospores is concerned, I have noted the same uniformity 
in my study of Composite. But serial sections of about three 
: hundred ovules of S. glaucophylla showed all of the above men- 
— tioned variations. S. discolor ane other Teves indicated a —: 
2 dar variation. It is possible b 
_ sometimes upon a few sections, and even these taken f | the : 
oh a > plant. ‘Such results are Spear oii ertain 
oy 
