231 
obviously the attributes of ferns, which have no flowers, as of phx- 
nogams, which have. These two objects or purposes are always 
closely connected, often so elaborately interwoven as to be con- 
founded in our short-sighted investigations ; yet are in every instance 
found to be distinct and independent, if the inquiry be carefully and 
skilfully conducted. The frond or leaf, which in ferns normally bears 
the fruit on its back, does not, on this very account, seem, at first 
sight, so distinct an organ as the leaf of a phenogam, for it appears 
to occupy a somewhat neutral ground between the preservation of the 
individual and the preservation of the species; this difficulty, how- 
ever, is dispelled by the consideration that those ferns which have 
fronds entirely covered with fructification have other fronds also on 
which no fructification whatever appears. There can be no doubt 
that such unfruitful fronds perform the office of leaves; and hence we 
may fairly assume that those leaves which contain the receptacle of 
the fruit also perform the same office. 
Having thus indicated the existence of two classes of functions, 
and the provision of appareils or systems of organs adapted to each, 
it would become necessary to weigh the claims of both, were it not 
obvious.that the system of generation or preservation of the species 
must have the greater claim on our consideration; but no sooner do 
we seek guidance from this than we find the physiological characters 
of generation, as far as we know them, exactly alike in all ferns ; and 
we feel disinclined to regard the slight differences in external form 
and condition of the clusters of capsules as of sufficient importance 
for the establishment of divisions.* We are therefore compelled to 
* In cryptogams we have a succession of individuals somewhat different to that 
recognised in phenogams; for each individual appears to produce its kind without the 
intervention of a second individual; and the definition given to a species in the ani- 
mal world, and which is partially true also in phenogams, is not true, as far as we 
have yet ascertained, in cryptogams, where each individual appears as the parent of a 
race. This, perhaps, accounts for the extreme difficulty in fixing the limits of a spe- 
cies in ferns, and is likely to lead to the unnecessary multiplication of species; for it 
is quite certain that a homogeneity of appearance will, and does, constantly occur on 
mountain ranges, another homogeneity of appearance in the individuals found in a 
wide range of bog at a lower elevation, and a third homogeneity in individuals occur- 
ring under the shade of a forest; and in every instance this appearance may be due 
to the isolation, for many centuries, of an individual and its descendants, which con- 
tinue to reproduce their kind under peculiar conditions of soil, altitude, temperature, 
exposure, and so forth. This difficulty appears very often to have been unnecessarily 
magnified; for, except as a question for the consideration of speculative theorists, it 
really matters very little whether two individuals grown under different natural con- 
ditions are representatives of races, local varieties or closely allied species. 
