302 Beer, Development of tlie pollen grain and. antlier of some Onagraceae. 
The interstitial bodies are shut off from the cavitv of the 
kJ 
pollen grain by a well developed closing-disc which consists of 
two parts; an outer dense and homogeneons layer and an inner 
less dense, stratified lamina which is cap-shaped in form with 
the con cavity directed towards the cell cavity. Lying on the 
outer lamina of the closing-disc is an aggregation of granulär 
material which extends some little way np the sides of the 
interstitial body. The cavity of the interstitial body is no longer 
occupied by the mucilaginous deeply staining substance which 
iilled it at an earlier period, bnt it now contains a watery fluid 
which does not readily stain. I believe that the mucilaginous 
material of the young interstitial body has been, to a great 
extent, used up in forming the closing-disc and tliat the granu¬ 
lär substance which lies upon the outer portion of the disc is a 
remnant of the stainable material. (See Fig. 23.) 
In pollen grains which measure between 85 y and 95 y in 
both 0. hiennis and 0. longiflora the protoplast lias become re- 
duced to a liollow sphere or vesicle which has expanded again 
until it is nearly or quite in contact with the cell-wall (Fig. 251. 
At one point upon the protoplasmic vesicle a flattened, ratlier 
dense nucleus can be seen which encloses a nucleolus. A little 
finely granulär cytoplasm surrounds the nucleus but in its other 
parts tlie protoplast appears to be reduced to a plasmoderma 
(Hautschiclit) which surrounds the enormous central vacuole. 
Very soon the nucleus enlarges, becomes round er and less dense 
and passes into the propliases of mitotic division (Fig. 26). 
I liave not followed the details of this division which leads 
to the formation of two distinct cells within the pollen grain: 
the large vegetative cell and tlie small generative cell. The 
latter is limited by a well marked plasmoderma. (Hautschicht) 
(Fig. 27). 
The tapetum now breaks down and its contents cleariy 
furnish the material for the renewed growth of the pollen¬ 
protoplast s. 
In order to understand the nature of tliis material it is 
necessary to consider the changes which take place in the tape¬ 
tum during its earlier development. 
In the very young anthers, before the full number of pri- 
rnary sporogenous cells is established, the tapetal cells contain a 
not very dense cytoplasm which encloses a single nucleus. 
This nucleus, besides small, scattered cliromatin granules, 
contains one to four nucleoli. 
The nuclear membrane colours deeply with iron-haematoxylin 
or with methylene blue-fuchsin mixture. Very rapidly the cyto¬ 
plasm increases in density and the originally single nucleus 
divides into several, as many as eiglit nuclei being not uncommonly 
met with in a cell (Fig. 30 and 31). 
Until ab out the pollen - mother-cell stage the tapetal nuclei 
multiply exclusively by mitotic division but at the mother-cell 
stage nuclear figures occur which are strongly suggestive of 
