1918] DUDGEON—RUMEX CRISPUS 407 
The explanation of these tubes seems to be that the high osmotic 
pressure set up in the pollen grain literally blows the distensible 
intine out through the germ pores, and continues elongation of the 
tube as long as the pressure is maintained. Presently the pressure 
goes down and the process ceases. It may be that the retention of 
some semblance of organization by the nuclei indicates that the cell 
is not dead, and that tube formation is a weak abnormal growth 
process. Finally, the contents of the grain begin to shrivel (fig. 74), 
the walls become wrinkled and collapse, and the cell dries up. 
By the time the anthers open, little or no trace of the pollen tubes. 
remains, and the nuclear material is no longer distinguishable. I 
would estimate that 99 per cent of the pollen grains undergo some 
such degeneration process. Very infrequently one can find a 
grain of normal size with the nuclei clearly outlined, lying among 
the hypertrophied grains, and occasionally one or two of the 
loculi of an anther will contain seemingly normal grains, while the 
remaining loculi are filled with the large cytolyzed grains. 
FUNGUS INVASIONS.—Such a mass of disorganizing cells would 
seem to be particularly favorable material for the growth of sapro- 
phytes. In every dehisced anther sectioned there was an abundant ~ 
growth of an unidentified fungus. The septate hyphae ramify 
everywhere through the pollen chambers among the pollen grains, 
and in a very few instances were seen penetrating pollen grains 
through the germ pores (fig. 70), but with the appearance of a 
chance entrance, rather than a definite exploitation of the contents 
of the grains. It may be stated confidently that the fungus does 
not attack unopened anthers, for it never was present until after 
the anthers had opened. In those anthers where the fungus had 
been developing longest and had formed a felt of hyphae, it had 
produced a great abundance of minute spores. 
DEGENERATIONS IN OVARY.—Degenerations just as widespread 
and devastating occur in the ovary also. They do not begin so 
early as in the anthers, never having been observed with certainty 
before reduction is completed (figs. 31, 32), although there are 
faint suggestions that even during the second reduction division 
the cells may not be entirely healthy. The non-functioning mega- 
spores seem to degenerate prematurely, although this is by no 
