1918] DUDGEON—RUMEX CRISPUS 401 
tetrad of cells results. Further, the polar nuclei fuse early, a 
process characteristic of normal haploid embryo sacs; and a fair 
number of embryos is formed. Many embryo sacs have the 
appearance of having waited a long time for fertilization, before 
finally degenerating (fig. 47). On the other hand, there is equally 
convincing evidence that fertilization does not occur. I cannot be 
sure that I have seen normal pollen in any instance; that which 
might be considered normal is so scarce that there seems little 
chance for adequate pollination. Some pollen grains have been 
observed on the stigmas, but they all had the appearance that is 
believed to indicate degeneration, and certainly none of them was 
germinating. At the same time, the stigmas wither early, and 
most of them would seem to be incapable of supporting pollen 
tube growth. No pollen tubes were ever observed, either in the 
styles or in the micropyles, nor any densely staining remains that 
might indicate the position of former tubes. The dead impervious 
layer over the outer integument would seem to preclude the entrance 
of pollen tubes through any other point than the micropyle and the 
tip of the nucellus. No fusion of gametes, which after all is the 
final proof of fertilization, was seen; indeed, in all the prepara- 
tions I have made and examined I have seen just one embryo 
sac that contained a normal appearing egg apparatus (fig. 46), and 
it was in a still unopened flower. In the many flowers I have 
examined in section, only a relatively small number contained 
embryos. 
It is possible that Rumex crispus behaves as OVERTON (7) 
found for Thalictrum purpurascens, where some megaspore mother 
cells undergo normal reduction, producing haploid embryo sacs and 
reduced eggs that require fertilization for further development, 
while other megaspore mother cells fail to reduce, and give rise to 
diploid sacs and eggs, the latter developing embryos apogamously. 
If this should prove to be the situation in Rumex crispus, then it 
would be expected that the embryos arise apogamously from unre- 
duced eggs. I regret that there has been no opportunity to collect 
new material to settle this important question. 
SUBSEQUENT DEVELOPMENT OF FLOWER.—It is beside the pur- 
pose of this paper to discuss the details of fruit development, and 
