Beer, Development of tlie pollen grain and anther of some Onagraceae. 309 
gularities in tlie divisions similar to those described by Juel in 
Hemerocallis fulva. 
4. The first pollen membrane is formed by the direct 
activity of the protoplast and is deposited as a delicate layer of 
pectic material upon the inner face of the special-mother-cell 
wall. Although it originales in the most intimate contact with 
the callose wall it is chemically distinct from this from the very first. 
5. The interstitial bodies originate as specialised areas on 
the first pollen wall. These spots are at first characterised by 
their greater thinness; later a homogeneous mucilage is develo- 
ped at these places. In older pollen grains a portion of this 
mucilage is deposited as a dense closing disc whilst, in Oenothera, 
the rest of the interstitial body is filled with a thin fluid. In 
Gaura more or less solid laminae are deposited throughout the 
interstitial body. 
6. A secondary thickening layer is laid down by the proto¬ 
plast within the first pollen membrane. This layer gives most of 
the pectic reactions but also a very distinct violet colour with a 
strong solution of lodine in potassium iodide. It gives none of 
the usual cellulose reactions. 
7. Both the first pollen wall and the secondary thickening 
layer are firmly attached to the protoplast when they are first 
developed. 
In pollen grains wTiich have reached 40 g in size the proto¬ 
plast is no longer fixed to the wall at any place although it still 
completely fills the cell cavity. 
The pollen grain continues to grow and its walls increase 
both in thickness and in extent. Whilst the pollen grain doubles 
its diameter the cell-cavity increases in size from about 26 g 
to ab out 46 g. The protoplast, liowever, grows far less rapidly 
during this time and its diameter only enlarges from 26 g to 
ab out 34 g. 
In consequence of this inequality in growth the protoplast 
becomes separated from the pollen wall by a space wliicli is 
filled with liquid. 
The conditions seem to be quite similar to those which 
Fitting and others have described in the megaspores of Isoetes 
and Selagin ella. 
In Oenothera also, as in the megaspores, the growth of the 
layers of the wall is not equally rapid and the first pollen wall 
becomes separated from' the secondary thickening layer and is 
thrown into folds upon its surface. 
• These observations show tliat during the period of most 
active growth of the membrane (both in surface and in thick¬ 
ness) the protoplast is completely separated from it and we 
must conclude either that the growth of a cell-wall is a purely 
physical process or that the living protoplast can exert its in- 
fluence across a space filled with liquid. 
I may add here that, although the growth of a membrane 
whilst this is separated from the protoplast by an actual space 
Beihefte Bot. Centralbh Bd. XIX. Abt. I. Heft 2. 21 
