170 CANNON: STUDIES IN PLANT Hysrips 
place. In all of the material, however, which was gathered later 
in the season, and which had nuclei in the process of division at 
the time of putting up amitosis occurred. I have not been able 
satisfactorily to determine whether these abnormal divisions were 
due to the cultural conditions, to the fact that the plant was a 
hybrid, or to both these factors with the added one that the flowers 
were the last to form on the plants. 
A sharp distinction was made in the cotton between the nuclei 
of the abnormal cells which accomplished division, and those that 
were abnormal but degenerate and never underwent division either 
by the direct or by a modification of the indirect method. De- 
generation of the sex cells, either of the archesporium or in later 
stages, could be found to some extent in practically all of the ma- 
terial examined, but this was always so well marked structurally 
that no confusion arose by which the degeneration of the cells was 
questioned. 
When the degeneration took place in the archesporial cells it 
was very noticeable both in the nucleus and the cytoplasm. The 
cytoplasm assumed various conditions and structures. It was a 
coarse reticulum of rather delicate meshes which had a uniform 
appearance as if the microsomes were wanting, or a large vacuole 
occupied the center of the cell, pushing the nucleus to one side 
and causing the cytoplasm to gather in various parts of the cell. 
The cytoplasm of the degenerate cells takes stains with great 
avidity. This is especially to be noticed in the anthers that have 
both normal and degenerate nuclei. When in such anthers the 
cytoplasm of the normal cells is properly stained that of the 
degenerate cells will always be badly overstained. This is very 
different from the staining capacity of cells whose nuclei divide by 
the direct method, since these cells stain in a perfectly normal 
manner. 
The nuclei of degenerate archesporial cells are often smaller 
than those of the normal cells in the same anther, the nuclear 
membrane is frequently very pronounced, and the chromosomes 
are of the size and number of the somatic ones. I found a few 
cells which were preparing for the first division and which had the 
two-spindle condition observed by others in hybrids ; but I did not 
notice any such nuclei in later stages, since these cells degenerate 
