1904] GARBER: RICCIOCARPUS NATANS 165 
be produced by the enlarging of the sporophyte, but it occurs 
independently of this and is not present in purely vegetative 
plants. As growth proceeds, the plants become broader and 
thicker, and at the same time the older posterior parts decay 
(fig. 1). The decay sets free the ripened spores and provides 
for vegetative multiplication by separating the branches. No 
vegetative multiplication occurs during the spore-producing 
season. 
THE SEX ORGANS. 
It is a common statement that Ricciocarpus fruits only when 
in contact with the soil. In this vicinity no stranded fruiting 
plants have ever been collected except such as showed clearly 
that the fruiting process was already far advanced before the - 
plant came to rest. A careful examination of abundant material 
in numerous localities shows that the fruiting plants normally 
float free on the surface of the water from germination of the 
spores until the spores of the next generation are ripened and 
discharged. The untimely withdrawal of the water often 
leaves mature plants stranded, and although they develop rhi- 
zoids very quickly, no change of form occurs until the spores 
have been discharged. 
Ricciocarpus is described by Schiffner, Leitgeb, and Camp- 
bell as being strictly dioecious. Our investigations prove con- 
clusively that it is strictly monoecious. The error of previous 
observers doubtless arose from the exclusive examination of 
mature plants, which normally bear only sporophytes because 
the antheridia have been lost by the decay of the older portions 
of the thallus. A very few plants usually produce antheridia 
only, and retain these until the plants are full grown, while still 
other plants may be quite sterile. Young plants and those 
formed from broken-off floating tips of soil plants very soon 
begin to produce antheridia in from three to five rows. The 
antheridia are at first superficial on the floor of the dorsal furrow 
just behind the apical cells. They gradually become sunken in 
cavities by the upgrowth of the vegetative tissue surrounding 
them (figs. g, 5). The series may be interrupted at intervals so 
that the antheridia may occur in more or less definite groups, 
