150
ILLUSTRATIONS OF INDIAN BOTANY.
Affinities. These appear to be very imperfectly understood at least if we may judge from
the different positions in which different authors place the order, which however is not a good
criterion. Jussieu placed it between Meliacece and Geraniaceae — DeCandolle retains it in the
same situation, Bartling (Ord. Naturalis) forms a class of this and Meliaceae, but still retains
it beside Geraniaceae, Lindley in the first edition of his Introduction places it between Mal-
pighiaceae and Meliaceae, but far removed from Geraniaceae and Pittosporeae, and still fur-
ther from Berberideae, while in the second edition it is placed between Berberideae on the one
side and Pittospoieae on the other, but still far from Geraniaceae. Meisner retains it nearly
in the old place between Meliaceae and his class Malpighinae, (see page 137) which last is fol-
lowed by his class Geraniodeae. To decide between such authorities is more than I dare at-
tempt, but upon the whole feel disposed to adopt in preference the Jussieuian arrangement,
though it places the order in a situation where, so far as I can see, it has no very close relation-
ship with those on either side, nor indeed do I think its affinities well made out any where. In
the artificial arrangement of the orders adopted by Jussieu it is very conveniently placed at the
end of a series of orders, having some well marked affinities, and separates another set, forming
the Gijnobasious group of Lindley, at the head of which the Geraniodeae may be, and is by most
authors, placed ; though in Lindley's arrangement, they form the second of the four Alliances
into which he divides that group. As a strictly natural arrangement Lindley's perhaps
approaches the nearest to perfection, but with this evident disadvantage that the group in
which he places the order (Albumenosae) is distinguished by a character not always easily made
out, and subject to some striking exceptions, such as excluding many genera in which albumen
abounds, and including at least one order (JVelumbiaceae) in which it is wanting. One half of
Meliaceae and of Cedrelaceae, have each, with many others, albuminous seed, but in smaller
proportion, yet do not find a place in this group. From this, and innumerable similar instances
which may be cited of irregularities in this part of the structure of seed I do not see, even taking pro-
portion into consideration, how any arrangement made to depend on it, can be good in practice, as
applied to orders until a higher value is assigned to it, though perhaps it may occasionally be em-
ployed as a useful generic character, though even that is doubtful. The genus Cassia, for example, as
now constituted has at least two species C. fistula and C. Roxburgii, having very copious albumen
while most of the others are exalbuminous. In such cases the presence of albumen may be
advantageously employed to aid in removing badly associated species from an otherwise natural
genus, and in this instance confirms the judgment of those who had previously separated the
genus Cathartocarpus, on account of the irregular structure of its legumes. But while we
meet with similar irregularities of structure, in so far as this organ is concerned, in almost
every family, it is surely a questionable arrangement which brings together a series of upwards
of twenty orders, many of them, so far as I can see, having scarcely any other mark of relation-
ship, merely because they agree generally, but not universally, in having seed with a copious
albumen and small embryo. For these reasons I am not yet prepared to adopt Dr. Lindley's
arrangement in that particular group, though far from thinking t hat we ought to lose sight of the
idea which led to its formation, since, if thoroughly investigated some important relations may
be found to exist between albuminous seed and vegetable structure generally, which has not yet
been discovered, but which may ultimately tend to modify our whole system, much in the same
way as the discovery of Exogenous and Endogynous structure, being connected with the struc-
ture of the seed, has given stability to the classification of the whole vegetable kingdom accord-
ing to that structure. Upon the whoie then I think we may safely conclude that our knowledge
of vegetable structure is not yet sufficiently advanced to admit of our constructing a system on
such principles, and that therefore, for the present the safe course to pursue is to arrange our
natural orders according to some convenient artificial system suited to f-tcilitate the investiga-
tion of new plants, though we ought never to forget for a moment, that such is not the object
of our studies, but the discovery of a truly natural system throughout. Here I leave the sub-
ject merely observing that Araiiaceae is the order to which this seems most nearly to approxi-
mate, though abundantly distinct, and in the structure of the seed, even to Rabiaceae.
Geographical Distribution. Principally confined to the tropics and warmer countries of the temperate zones, very numerous in India, less so in America and Africa, but found in both. According to Wallich's list there are 53 species of Vitis including Cissus, in India, while ( in the Peninsula we enumerate 25. He again has 13 species of Leea, while we have only three.
