244 The Botanical Gazette. [June, 
this plant the author finds that the initial cell gives rise to a 
row of only three cells, the lower one of the series developing 
from two successive cell divisions of the initial cell, of which 
the lower one gives rise directly to the embryo-sac. It is not 
improbable, however, that occasionally only three cells are 
formed here as in Myosurus, a phenomenon that was observed 
in other genera to be mentioned later. 
The most remarkable phenomenon met with here is the 
presence of two and sometimes more initial cells and their 
development into mature embryo-sacs. In figs. 1 and 3 may 
be seen two initial cells of about equal’size. In a number of 
cases one of the cells was somewhat larger than the other. 
The larger would in all probability take the upper hand of the 
smaller, thus crowding it out when only one embryo-sac would 
be the result. Frequently two cells were observed in later 
stages of development (fig. 7), both of which would, without 
doubt, produce mature embryo-sacs. As a rule one cell is a 
little in advance of its neighbor in development (fig. 7, the 
cell on the left). This state of things is not exceptional, but 
in some carpels a large number of ovules presented two, an 
occasionally three, initial cells. It is certain that every 
initial cell does not ultimately result in a mature embryo-sac, 
for it seems that there is unquestionably a ‘‘struggle of the 
parts” here as elsewhere. No more than two embryo-sacs 
were observed in one ovule. In fig. 9 are shown two fully 
antipodal cells were visible. The two embryo-sacs were con- 
tiguous for their whole length, separated only by a very thin» 
and indistinct membrane. For lack of time a very large 
number of ovules was not examined to ascertain the frequency 
of the phenomenon, nor is it known to the writer whether 
more than one embryo-sac is fertilized, and in that event 
whether more than one embryo ever reaches maturity in 4 
single seed of Delphinium. 
We do not have to do here with an isolated phenomenon, 
for the presence of more than one embryo-sac in one 0 
has been observed several years ago by Hofmeister, * Tulasne, 
Schacht and others, 
*Strasburger, Ueber Befruchtung und Zelltheilung 36 and foot note. 
