382 Gates and Rees . — A Cytological Study of 
there are indications that the uninucleate condition may be restored by the 
degeneration of one of the nuclei. 
This binucleate condition in meristematic plant cells is of too frequent 
occurrence to regard it longer as a mere abnormality. Its chief significance 
appears to be an indication of an incomplete co-ordination between nuclear 
division and cell-wall formation. Possibly some unfavourable condition, such 
as a lowering of temperature at the moment of cell-plate formation, might 
lead to the development of a binucleate phragmosphere. It appears that 
the condition is a more or less temporary one, with subsequent restoration 
of a uninucleate condition in older parts by the stem. There is no 
indication of a phragmosphere in any of the binucleate mother-cells of j 
lettuce, and, as indicated above, we believe this condition to have originated 
probably in another way, through the break-down of an incompletely formed 
cell membrane. 
Related phenomena of failure in wall formation after the tetrad 
divisions have been described in the pollen development of various forms. 
In Oenothera gigas (Gates, 1911 , PL LXX, Fig. 85), in certain anthers of the 
flower, the pollen mother-cells failed to round off or free themselves from 
each other and the tapetum. Such cells not infrequently become quadri- j 
nucleate owing to the failure of walls to appear after the reduction divisions, I 
the four nuclei moving together into the centre of the mother-cell. But 
since these cells are not set free they cannot be functional. Holmgren 
( 1919 , p. 13) has described quadrinucleate pollen grains arising in a similar 
way in Erigeron eriocephalus and E. unalaschkensis (Fig. 26, p. 14). He 
finds them only in certain anthers in the flowers which are transitional between 
hermaphrodite disc-florets and female ray-florets, but believes they might ; 
be functional although they develop a thick wall. A fusion of nuclei in 
such cells might conceivably give rise to pollen grains which were 
functionally diploid or even tetraploid. 
■ 
B. Cytomyxis . 
The condition known as cytomyxis was found to be of relatively 
frequent occurrence in pollen mother-cells of all species of Lactuca under 
observation (Fig. 13). It is found to occur frequently during synizesis in 
otherwise perfectly normal mother-cells where the nucleus of one was so j 
eccentric in position as to come in contact with the cell- wall of an adjoining ! 
mother-cell. It was not found to occur necessarily in all cells of one 
loculus, nor in the same direction, but merely haphazard according to the 
position of the nucleus with relation to an adjacent mother-cell. 
This condition was found to be even more marked in loculi which 
showed traces of abnormality, for example those in which all or most of 
the nuclei appeared hyperchromatic. In extreme cases there was complete 
transference of the chromatin material from one cell into the cytoplasm of an 
