536 
The nucleus of the upper one undergoes 
repeated division without, however, form- 
ing the embryo-sac, which develops from 
the lower cell. Fischer (’80) reports a 
similar case in WMelica nutans with an axile 
row of three cells ; the two upper ones each 
contain two nuclei, and yet the embryo-sac 
is developed from the lower. Similar cases 
Guinard (’80) found in Cercis, Phaseolus, 
Erythrina, Miss McKenney (’98) in Scilla, 
and Wiegand in Convallaria (1900). The 
cells of the axile row with several nuclei, 
Guinard points out, are undeveloped em- 
bryo-sacs. Their nuclei divide several times 
in adjacent cells, but the intervening walls 
do not dissolve and permit the fusion of the 
two cells of the axile row to form the em- 
bryo-sac, which is always developed from a 
single cell. Other examples are known, 
like that shown by Strasburger in Rosa li- 
vida, Benson in Fagus and Carpinus, where 
several embryo-sacs side by side, or one above the 
other, begin to develop and attain considerable 
size, but do not fuse. 
The weight of evidence then goes to show 
that the embryo-sac is developed from a 
single cell of the axile row, though this row 
may consist of but a single subepidermal 
cell, as in Lilium, Tulipa, etc., or of two, 
three, or four cells, as in other types. 
If this single cell, the mother cell of the 
embryo-sac, or macrospore, is the homologue 
of a spore in a strict morphological sense it 
would be necessary to show that it is de- 
rived by the same, or similar, processes of 
development. In those plants where the 
axile row consists of four cells, the cells 
might be regarded as tetrads, or spores, one 
of which develops the embryo-sac, while 
the others degenerate. But their axile ar- 
rangement, so constant in all spermaphytes, 
is against that supposition, and indicates 
that the course of development of these 
cells is more in keeping with the develop- 
ment of adjacent nucellar tissue. The ax- 
ile arrangement itself, however, would not 
SCIENCE. 
[N. 8S. Vou. XIII. No. 327. 
constitute a bar to their homology as spores. 
But in plants where the axile row consists 
of but two cells, or one cell, not even such 
a slight suggestion of their homology with 
spores is presented. 
80 far as investigation has been made, 
there is evidence that the development of 
spores in the bryophytes and pteridophytes, 
and in the development of the pollen in the 
spermaphytes, from a mother cell, is accom- 
panied by nuclear phenomena known under 
the head of reduction of the chromosomes. It 
has been suggested that the reduction of the 
chromosomes in the formation of the em- 
bryo-sac might be employed as a criterion to 
determine what constitutes the spores. The 
facts, however, which have been obtained in 
the few cases investigated do not offer any 
more hopeful evidence as to the identity of 
the spores. In Lilium, and other observed 
cases, where the single cell develops directly 
into the embryo-sac, the reduction takes 
place in the first division. In other cases 
where an axile row of two, three or four cells 
is formed, the reduction of the chromosomes, 
80 far as I know, always takes place in the 
primary mother cell of these, several cell 
divisions removed from the beginning of 
the embryo-sac. This is what we should 
expect, since this cell undergoes a matura- 
tion period prior to the formation of the 
axile row. So, whatever criterion we em- 
ploy to determine the identity of the spores, 
we are led to irreconcilable confusion ; 
either that the embryo-sac just prior to fer- 
tilization consists of 1, 2, 4 or 8 spores, or 
that it is developed. in some plants from a 
spore representing one-fourth of a mother 
cell or sporocyte; in other plants from one- 
half of a mother cell, and in still others 
from the entire mother cell. 
It is clear then that there do not exist 
here spores in the sense in which they are 
represented in the pteridophytes, or in the 
microspores of the spermaphytes ; neither 
in actual form, nor according to processes 
