in Lilium Martagon: II. Sperma togen esis. 213 
the dividing figures are more or less distorted and obscure. 
The division of the generative nucleus generally takes place 
soon after the formation of the pollen-tube 1 . It is already 
complete in one of my preparations which shows a germinating 
pollen-grain with a tube of very moderate length (Fig. 34). 
In this instance the vegetative nucleus looks as though it 
were in the act of dividing by the direct method, but I have 
never identified two nuclei at the end of a tube. 
Longitudinal fission of the chromosomes is apparent at 
a comparatively early period in the history of the generative 
nucleus. It is very clear in the preparation from which 
Fig- 33 was drawn, and I have seen it unmistakably in six 
other cases. One of these shows the chromosomes much 
twisted round each other, and evidently only just formed from 
the spirem-ribbon. Stages intermediate between those shown 
in Figs. 33 and 34 are, however, wanting in my preparations. 
One section which contained a nucleus rather older than that 
of Fig. 33, the chromosomes being shorter and broader, and 
the fission in each even clearer, was lost in an attempt to 
restain it. The absence of such later stages is of the less 
importance, as M. Guignard has figured two in the memoir 
from which I have quoted so often already (Plate XI, Figs. 33, 
34). The stages I possess are conclusive as to the fact of 
longitudinal fission of the chromosomes. 
As we have seen, the pollen-grains are usually brushed on 
to ridges in the neighbourhood of the clefts in the stigma. 
The pollen-tubes have therefore no great distance to travel in 
search of them. By the time a tube has reached the nearest 
cleft, it usually contains three nuclei. The vegetative nucleus 
is more or less disorganized at the end of the tube. The two 
generative nuclei are exactly alike, each enclosed in its own 
envelope of protoplasm (Fig. 34). Neither nucleus possesses 
a nucleolus or a membrane ; they are both in the form of 
a naked spirem, and they continue thus throughout their 
journey down the style. M. Guignard has pointed out that 
1 Guignard, 1 . c., p. 178. 
Q 2 
