152 BOTANICAL GAZETTE [FEBRUARY 
excelsa (POLLOCK) is similar in this respect. The division of the 
stalk cell cannot be so regarded; it may represent ancestral condi- 
tions. If so, it indicates a survival in Abies of a more ancient 
type of gametophyte than has been reported in any other of the 
Abietineae. Further, as the presence or absence of prothallial 
cells is a rather constant character of the different groups of Conif- 
erae, and thus may have a phylogenetic significance, the presence 
of several prothallial cells may suggest a connection of the abietin- 
Fics.'9-15.—A. balsamea ten days ae shedding: figs. 9-13, ees) stages 
in telophase of mitosis of central cell; the relative positions are indicated by that t of 
the prothallial cell; fig. 14, polar view of stage shown in fig. 13; fig. 15, asi of 
generative cell below ee line during stage shown in fig. 13, or just after formation 
of cell membrane; all X92 
ean and araucarian lines; which connection is also indicated by 
evidence accumulating from other sources. 
Certain peculiarities in the division of the central cell to form 
the generative cell and tube nucleus have been mentioned. Atten- 
tion was drawn to these by the very conspicuous radiating strands 
which surround the generative nucleus, as seen in a polar vieW 
(fig. 14). Further investigation shows these to be derived from 
the spindle, and that they take part in the formation of the ¢ 
membrane. 
ell 
