918 Gates.—Pollen Formation in Oenothera gigas . 
to the left) comes to appear strikingly like a threadwork or the ‘ spireme ’ of 
the nucleus from which it was extruded. The membrane surrounding this 
threadwork is at this time quite as definite as any nuclear membrane, so it 
is difficult to see how this differs in any particular, except its origin, from 
a real nucleus. It seems impossible to believe that any ‘structure ’ persists 
through this series of transformations. This will be discussed later, for, 
without laying too much stress on this series of observations, it is obvious 
that they may have an important bearing on our present views of chromo¬ 
some individuality and nuclear structure. 
After the stage of the pseudo-nuclei shown in PI. LXVII, Fig. u, they 
become more indefinite, the membrane disappearing so that soon all that 
remains of the pseudo-nucleus is a densely-staining portion of the cytoplasm, 
as shown in Fig. 12. This becomes looser and more indefinite, taking many 
forms, and finally is probably entirely incorporated in the cytoplasm of 
the cell into which it was extruded. 
This process of extrusion of chromatin from the nucleus of one 
pollen mother-cell into the cytoplasm of the adjacent mother-cell I call 
cytomyxis . Whether it always occurs at this stage of synapsis is not certain, 
but there are some observations which lead me to believe that it does not 
always occur. In some flowers one finds the nuclei of the mother-cells 
occupying positions almost or quite in contact with the cell-wall, but without 
any chromatin extrusion. As already stated, evidence that extrusion of 
chromatin has taken place remains for some time in the cytoplasm of the 
adjacent cell, so that in examining an anther it is easy to determine whether 
such extrusion has occurred. If the cytoplasm is entirely colourless (in iron- 
haematoxylin stain) it is evident that there has been no extrusion, no matter 
what the position of the nuclei. On the other hand, dark-staining cytoplasm 
does not always indicate that this process has taken place, for this is 
characteristic of degenerating cells. The position of these dark areas on the 
periphery and next to an adjacent mother-cell, together with their relation 
to the cytoplasmic connexions, serves to identify them. 
It might be supposed that this process of chromatin extrusion was 
followed by degeneration of such mother-cells, but there are several reasons 
why this does not appear to be the case. In the first place, the nuclei appear 
perfectly normal after the extrusion/has taken place, and later (Fig. 12) 
begin to move back towards the centre of the cell. The cytoplasm of such 
cells also appears ^wholly normal, and differs in no way from those in which 
no extrusion has occurred, except in having the dark areas on their periphery 
near certain of the'cytoplasmic connexions. 
At the time the movement of the nucleus to the cell-wall occurs, the 
spireme appears to be always on the side of the nucleus next th,e cell-wall, 
whether extrusion takes place or not. This is also true when the movement 
is to the wall in contact with the tapetal cells, with which there are no 
