COLLECTING AND CURATING FRUITS AND SEEDS. 49
for the tiny seeds and picking them out with the forceps, one by
one, from a very much more bulky collection of grit and dirt,
from which they are indistinguishable without the aid of a lens
or powerful hand glass. As a matter of fact a very little material
is enough, and my advice is to harvest that little as exclusively
and as cleanly as possible.
A word must be said about collecting seeds that are shot
away by the parent plant. My own plan is to put a few sprays
bearing nearly ripe fruits into a paper bag, in which they can be
left in a dry place or exposed to sunshine, and then the first—
instance dispersal will take place in the fulness of time. Few
people, who have not tried, would believe how difficult it is to
collect half a dozen seeds even from so large a plant as Geranium
pratense without some help in catching them as they are shot off.
There are, of course, more ways of collecting than from the
parent. I have already mentioned animals, alive and dead,
as well as dead leaves ; there are also, among other things, the
pellets of birds, while the margin of a stream or of a sheet of
water will often provide specimens which could not otherwise be
secured without a boat or a wetting. All such methods, however,
have the drawback that they involve naming the seeds without
knowing the plants that bore them, and it is sometimes very
difficult to be sure of the identity when there is no correctly named
material for reference and comparison. I do not put into my own
collection any seed so acquired which I cannot match in detail
with others taken from a known parent. A wrong name must
be avoided at all costs ; rather than make a mistake it is far
better to call a seed of uncertain identity "Oenanthe," for ex-
ample, or even "Umbelliferae," or, to be quite frank and label
it "incog."
It is, however, always interesting to sow such seeds in order
to get the desired information, and so acquire incidentally some
knowledge of the seedling and of the juvenile stages of the
plants. Identification can often be helped by examining the
embryo and endosperm, if there is any, but seeds so dissected
should not be admitted to the collection, because such nutritious
things as exposed endosperm are certain to be attacked by fungi
and mites. I do not, however, wish to be misunderstood ; the
wind dispersal of the Samphire cannot be appreciated without
cutting into the pericarp and seeing the inter-cellular spaces,