208 BOTANICAL GAZETTE [SEPTEMBER 
species of Taxus, Cupressus, and Juniperus the pollen grain divides 
only after reaching the nucellus; while in Thuja, Cephalotaxus, 
Podocarpus, Chamaecyparis Lawsoniana, Cryptomeria, Sequoia, 
Araucaria, and the Abietineae it divides before shedding. 
In examining so many pollen cells it was not surprising that a 
good many abnormal forms were met with. In Cupressus it happened 
not infrequently that the genem 
tive nucleus was not cut off int 
separate cell of its own, bit 
remained free in the general cyto 
plasm, side by side with the pollen 
tube nucleus. Insuch cases (jig. 4) 
the two nuclei were so much alike 
as to be hardly distinguishable 
In fig. 5 is illustrated a poll 
grain of peculiar shape and about 
twice the normal size. The genet 
tive cell, however, is of the usual 
appearance. It will be remembert! 
that the tree from which ts 
pollen was taken was not growils 
in perfectly congenial surrount 
ings, and this may account for the 
10 +ather unusual number of abnor 
Fics. 6-10. Larix europaea; figs. malities. 
6-9, abnormal pollen grains; fig. ro, In Larix europaea a number of 
normal grain. X 315. Fics. 11, 12. 
Encephalartos sp.; fig. 11, shed pollen 
grain; jig. 12, sprouting grain. x 315. 
cases were found in which 
mother-cells had divided but "% 
so that only two instead of - 
pollen grains were formed (fig. 6). In some of these double grains 
the divisions proceeded about as usual. Fig. 7 gives such @ mn 
but here only one prothallial cell is evident. Abnormal pao 
two large free nuclei were also found (fig. 8). In fig. 9 4 8% § coe 
cell (g) of very unusual appearance is shown. For the sake : 10) 
parison a drawing is given of the normal pollen grain of Lang (is: ot 
In Encephalartos sp. ? the mature pollen grain contained on No 
thallial cell, p, a generative cell, g, and a tube cell, / (fig. 11) 
