THE FERNS 127 
are usually sterile, and seem to be the result of exces- 
sive vegetative activity. They not infrequently multi- 
ply by means of special buds, or gemma, by which the 
number of the gametophytes may be rapidly increased 
exactly as in the liverworts. 
In some species of Trichomanes (Fig. 35, E), a genus 
of the filmy-ferns, the gametophyte may have the form 
of an extensively branched filament, closely resembling 
an alga; and it has been suggested that this may be the 
primitive type of the gametophyte. However, as many 
closely allied species produce the usual flat thallus, and 
all of the forms, when exposed to excessive moisture, 
show a tendency to assume a filamentous stage, it 
is quite as likely that this is an adaptation to a moist 
environment, rather than being the primary condition. 
Another group of ferns, the so-called Eusporangiate, 
which includes the adder-tongue (Ophioglossum) (Fig. 
34, A) and its allies, as well as certain interesting trop- 
ical forms, the Marattiacee (Figs. 31, 84), show a long- 
lived gametophyte of a somewhat different type. In all 
of these, so far as they are known, the gametophyte is 
massive and quite different from the thin, delicate 
thallus of the filmy-ferns and Vittaria, but like these the 
gametophyte may live for a long time, often for several 
years, and not infrequently remains alive long after the 
young sporophyte is quite independent. The gameto- 
phyte in the Marattiacee, especially (Fig. 32, A, B), 
is extraordinarily like a simple thallose liverwort, both 
as regards the thallus itself and the sexual organs devel- 
oped upon it. In the adder-tongues the gametophyte, 
so far as at present known, is subterranean and quite 
destitute of chlorophyll; but whether this is originally 
