. 
1907] COKER—CEPHALOTAXUS FORTUNEI 3 
all other conifers. These observations I cannot confirm, as all the 
pollen tubes examined had the usual two vegetative nuclei as in 
Taxus, Torreya, and other Coniferae.? 
In the spring of its second season the pollen tube still grows very 
slowly, remaining broad and sac-like until about two or three weeks 
before fertilization, when it rapidly penetrates the short distance to 
the female prothallium and spreads its tip over the neck of an arche- 
gonium, The body cell which has been growing slowly all the time 
now shows the structure indicated in fig. 2.2. Its protoplasm is very 
dense and shows well-marked radiations from a denser point just 
below the nucleus, The nucleus occupies a very eccentric position 
near the upper side of the cell, thus indicating the unequal division 
that is to follow. 
When the tube becomes flattened out on the neck of the arche- 
gonium, the stalk and tube nuclei may be touching or slightly pressed 
into the body cell; usually, however, they are at a greater distance 
om it. ; 
A few days before fertilization the body cell divides into two 
sperm cells of unequal size, the lower being in every case the larger. 
The difference in size is not so great as in Torreya taxifolia (COULTER 
and Lanp, 6) or in Taxus (BELAJEFF, 3) but it is nevertheless 
decidedly constant. In the seventeen pollen tubes that appeared in 
this stage in my preparations, the difference in size of the two sperm 
cells could in every case be easily made out, In two of three super- 
humerary tubes that had been left over after fertilization, the nucleus 
of the smaller cell had grown larger than the other, thus obscuring 
to some extent the unequal distribution of the protoplasm. In jigs. 
3, 4, 5 are shown three pairs of sperm cells of unequal size, It will 
be seen that each sperm cell is quite distinct from its fellow and from 
the protoplasm of the pollen tube. Soon after their formation the 
smaller cell tends to round itself more quickly than the larger (fig. 5). 
Another indication that the two sperm cells are not of equivalent 
value is that the nucleus of the smaller may not be so dense as that 
of the larger (fig. 4). 
*It is probable that ARNotp1’s material varied at times from the normal, as I 
: also to have been the case in my material of Podocarpus, where three vegeta- 
tive nuclei were found in the pollen tube. 
* For figs. 2-17, see Prate I. 
