504 
Farmer . — On Spore- Formation and 
General Conclusions. 
In endeavouring to obtain a general survey over the various 
processes which have just been detailed, one is struck at the 
outset by the difference existing between the behaviour 
exhibited by the cytoplasm in the Jungermannia and Mar- 
chantia series respectively. The (probably) constant appear- 
ance of four attraction-spheres in the four-lobed cells of the 
former, coupled with the absence of this peculiarity in the 
spherical or regular biscuit-shaped cells of the latter, can 
hardly be devoid of significance, and it seems to me it may 
have some bearing on the nature and function of these much- 
discussed structures. 
The facts may be briefly summarized as follows. In the 
ordinary cells, whether of the archesporium, as in Fossom- 
bronia , or in the gametophyte, as in Fegatella and in Pellia , 
amongst the first signs of nuclear division is the well-marked 
appearance, at the two opposite poles, of a fan- or brush-like 
series of radiations. These can often be traced to a minute 
dense particle situated on the outside of the nuclear wall. 
Until the radiations become visible, there is nothing to distin- 
guish the particular particle just mentioned from the many 
other granules of various sorts and sizes which coexist with it 
in the cell. As soon, however, as the radiation-systems are 
formed, they are seen in these cells to be definitely referable 
to certain focal spots, in the centre of each of which the 
‘ centrosome ’ is usually to be discovered. But I did not 
always observe one body only, even during the early stages. 
Sometimes it was possible to speak of a microcentrum in 
Heidenhain’s 3 sense of the term, for as many as three or four 
of these point-like granules might occur in the centre of each 
of the two systems of radiations. Just around the centrosome, 
or centrosomes, a clear hyaline space can be discerned. I 
believe this to be due mainly to diffraction, but partly, perhaps, 
also to the fact that a large number of hyaline radii are 
1 Heidenhain, Neue Unters. liber d. Centralkorper und ihre Beziehungen z. Kern 
tind Protoplasma. Arch. f. Mikr. Anat., Bd. XLII1. 
