140 
BOTANICAL GAZETTE 
[FEBRUARY 
animal egg has undergone some development so that it contains eight 
nuclei. 
This is no serious objection, however, when we remember 
that in the pines the cell directly homologous with the larger cell in 
D contains 
oO 
& co ro io) 
. 1.—A-D, successive 
ee 
3 
reduced; the shaded nucleus 
in D is the egg nucleus. 
thousands of nuclei, and in Gnetum contains too large 
a number to be counted accurately, and 
within the angiosperms shows a series 
ranging from more than a hundred nuclei 
down to sixteen and then to eight. Com- 
paratively few plants have been investi- 
gated, and it is not improbable that plants 
will yet be found which will complete the 
reduction series by showing only four nuclei, 
or only two, or even only one. The latter 
condition, with the megaspore functioning 
directly as the egg, would be illustrated by 
C of the diagram. 
To me the comparison seems so obvious 
that I can explain the previous absence of 
a theory of alternation of generations in anl- 
mals only by the fact that the gamete 
bearing generation is’ extremely reduced 
and is not approached by any gradual 
series as in plants. Had observations in 
plants been confined to angiosperms, there 
would doubtless be no theory of alternation 
of generations in plants. 
(2) The male gametophyte—It is not 
necessary to trace in detail the reduction 
of the male gametophyte. Originally 4 
conspicuous, independent plant, it becomes 
reduced, loses its independence, acquires’ 
the parasitic habit, and then undergoes 
a progressive reduction until it becomes 4 
simple, microscopic, parasitic structure 
which no one would think of homologizing 
with the gametophyte of a liverwort, were 
it not for the close series leading to the 
