i fe) BOTANICAL GAZETTE [JULY 
twelve, thus confirming the observations made on the microspore 
series. 
Lang (’97) has shown that in Stangeria the mother-cell forms a 
row of three megaspores; Treub (’81) reports the same for Cerato- 
zamia; and three are reported for Ginkgo. Among the Coniferales 
four are frequent. In Pinus Laricio I have observed that either three 
or four are formed indifferently. Strasburger (’79) gives three as 
the usual number in Taxus, although four frequently occur; in his 
later work on Taxus baccata (04’) he says that four cells are formed 
from the megaspore mother-cell. Juel (’oo) finds four in Abies 
sibirica and Larix sibirica; Shaw (’96) reports four in Sequoia; 
Lawson (’o4) studying the same species of Sequoia finds three; 
Coker (’o3) finds three in Taxodium distichum and (’04) four in 
Thuja orientalis, where they are not arranged in a row, but in nearly 
‘regular tetrad form. Strasburger finds three in Ephedra campylopoda, 
and Jaccard three in E. helvetica; in E. trijurca three or four are 
formed indifferently, dependent on the rapidity with which the func- 
tioning megaspore encroaches. This encroachment is probably the 
reason for the differences reported in the forms mentioned above. 
THE FEMALE GAMETOPHYTE. 
The two nuclei resulting from the division of the megaspore 
seem invariably to take the position with reference to the major 
axis of the ovule shown in fig. 37, for in no observed instance did the 
nuclear plate vary from this position. Before the spindle fibers and 
cell-plate have disappeared, a ring-like vacuole appears, entirely sur- 
rounding the cell-plate. The rapid increase in size of this vacuole 
is one of the chief factors concerned in the parietal placing of the 
free nuclei. These two nuclei divide simultaneously, and the result- 
ing four take equidistant positions at the periphery of the embryo 
sac (fig. 32). Successive simultaneous divisions (figs. 33, 35) rapidly 
follow each other until the maximum number of nuclei is reached, 
which in the present instance apparently does not exceed 256. It 
may be of interest to note that at only one time—immediately after 
the division of the megaspore—is the vacuole free from cytoplasm. 
Careful staining shows that at all later stages (figs. 32-35) it is filled 
with a delicate cytoplasmic structure, which gradually increases in 
density until free nuclear division ceases, which was about April 1, 
