1915] HUTCHINSON—MALE GAMETOPHYTE OF PICEA 297 
Relationships 
No one character is sufficient to establish relationships of plant 
groups. Since similarity of male gametophytes gives only one- 
sided evidence for the relationship of species or genera, the dis- 
cussion will be limited to comparison of types. In Taxodineae and 
Cupressineae the gametophyte consists of an antheridial cell, 
which may divide before or after shedding, and a tube nucleus; 
no evanescent cell is present. Similar gametophytes are found in 
Picea canadensis (fig. 30). In cycads two cells are cut off from the 
primary cell, one of which is antheridial (cf. fig. 38). The shedding 
stage characteristic of the abietinean gametophyte (also of Ginkgo 
and Ephedra) contains two more or less evanescent cells and an 
antheridial cell, which may or may not divide, beside the tube 
nucleus (cf. figs. 50, 51, 52). The podocarp type is similar, but 
the polar cells are not evanescent and frequently divide (cf. figs. 39, 
40, 52). A massive polar tissue containing free nuclei, which 
LopriorE (18) regards as antheridial, is characteristic of the 
araucarian type. A similar gametophyte is shown in fig. 33. The 
male gametophyte of Picea canadensis is in a state of unstable 
- equilibrium. Slight differences in conditions are sufficient to shift 
the balance in one of several possible directions. The resulting 
forms correspond to the various types of gametophytes found in 
gymnosperms. 
Summary 
In the male gametophyte of Picea canadensis, one, two, or three 
potentially antheridial cells are cut off from the primary cell; one 
of these divides to form a spermatogenous and _a sterile cell; the 
others, when formed, are more or less evanescent. Occasion- 
ally there are two functioning antheridial cells, resulting in a bi- 
antheridial gametophyte. 
The writer is indebted to Professor Joun M. Covutrer, 
Dr. CHartes J. CHAMBERLAIN, and Dr. W. J. G. Lanp for many 
Suggestions and criticisms. 
UNIVERSITY oF CHICAGO 
