ts,
nearly round and of great size in cobs and Spanish nuts, oblong and
longitudinally striated in Cosford's, and obtusely four-sided in the
Downton Square nut.

_Cucurbitaceous plants._--These plants have been for a long period the
opprobrium of botanists; numerous varieties have been ranked as
species, and, what happens more rarely, forms which now must be
considered as species have been classed as varieties. Owing to the
admirable experimental researches of a distinguished botanist, M.
Naudin,[750] a flood of light has recently been thrown on this group of
plants. M. Naudin, during many years, observed and experimented on
above 1200 living specimens, collected from all quarters of the world.
Six species are now recognised in the genus Cucurbita; but three alone
have been cultivated and concern us, namely, _C. maxima_ and _pepo_,
which include all pumpkins, gourds, squashes, and vegetable marrow, and
_C. moschata_, the water-melon. These three species are not known in a
wild state; but Asa Gray[751] gives good reason for believing that some
pumpkins are natives of N. America.

These three species are closely allied, and have the same general
habit, but their innumerable varieties can always be distinguished,
according to Naudin, by certain almost fixed characters; and what is
still more important, when crossed they yield no seed, or only sterile
seed; whilst the varieties spontaneously intercross with the utmost
freedom. Naudin insists strongly (p. 15), that, though these three
species have varied greatly in many characters, yet it has been in so
closely an analogous manner that the varieties can be arranged in
almost parallel series, as we have seen with the forms of wheat, with
the two main races of the peach, and in other cases. Though some of the
varieties are inconstant in character, yet others, when grown
separately under uniform conditions of life, are, as Naudin repeatedly
(pp. 6, 16, 35) urges, "douees d'une stabilite {358} presque comparable
a celle des especes les mieux caracterisees." One variety, l'Orangin
(pp. 43, 63), has such prepotency in transmitting its character that
when crossed with other varieties a vast majority of the seedlings come
true. Naudin, referring (p. 47) to _C. pepo_, says that its races "ne
different des especes veritables qu'en ce qu'elles peuv