1918] DUDGEON—RUMEX CRISPUS 415 
cell, and may account for the long wait in the prophase of the 
heterotypic mitosis, and the subsequent completion of the division 
as a typic mitosis, as has been repeatedly reported for well marked 
apogamous plants. . 
The strong tendency to failure in the sexual process may also 
contribute to the development of highly successful methods of 
vegetative propagation. Nearly all the species of section Lapa- 
THUM perennate by strong storage tap roots crowned by a short 
stem region, or by storage rhizomes, and all propagate very freely 
by detached fragments of these underground stems. R. Acetosella 
propagates by long lateral roots which produce new plants at 
intervals. It is a striking fact that patches of the plant are dense 
growths of almost exclusively staminate or ovulate plants. This is 
what would be expected to result from such vegetative propagation, 
and would be a curious segregation to result from plants produced 
to any great extent from seeds. All the evidence I have seen 
points to apogamy in R. Acetosella; seed production is very scanty 
in proportion to the number of flower buds, and a large percentage 
of the fruits are empty. A considerable number of seedlings 
scattered about in patches of ovulate plants indicates that many of 
the seeds are viable. 
From the great number of diclinous angiosperm flowers that 
contain remnants of the other organs, it seems very probable 
that the degeneration processes here described are of wide- 
spread occurrence, and are scattered throughout the group from 
the lowest to the highest forms. It is planned to make a 
study of dicliny in the future, in the effort to substantiate or 
disprove this theory of the origin of dicliny as the result of 
degenerations. 
The cause of these degenerations isnot known. The few authors 
who have discussed the problem all agree that faulty nutrition is 
important, if not as the direct cause, at least in producing condi- 
tions that call the phenomenon into activity. HorrMann (5) 
supposed that the embryos of R. Acedosella and other dioecious 
plants are sexless, and that sex is determined in the early stages 
of the seedling by the conditions under which they germinate. 
GARTNER (4) used the evidence in the reverse order, and thought 
