22 
PRELIMINARY TREATISE. 
teguments of the seed, which are variable in one family, 
may assume a more certain character in others. I do not 
presume to give an opinion concerning those which I have 
not examined, but we should admit with caution in one race 
that which has utterly deceived us in another. The form of 
the leaves might be expected to furnish some distinguishing 
feature, but it is so variable, even in particular genera, that 
little reliance can be placed upon it, and it must be admitted 
with great circumspection. For instance, Professor Lindley 
characterized Hypoxideae as having plicate leaves, whereas, 
if the fact were absolutely true with respect to the particular 
plants he meant to designate, we find plain and plicate 
leaves so intermixed in the nearly related order of Iridese, 
and even in the several genera thereof, that it cannot be 
taken here as a natural distinguishing character, without the 
most striking inconsistency. I wish it were true, that in the 
leaves of exogenous or bicotyledonous plants, the veins are 
concurrent and form a kind of network, and in endogenous 
or monocotyledonous are parallel and cross-barred ; and that 
scitamineous plants are distinguishable amongst the latter 
by being feather-veined, that is, having oblique veins pro- 
ceeding from the midrib. I have before me the leaf of a 
Nepal Arum, closely allied to Arum Dracontium, in which the 
veins are as concurrent as in the leaf of a lime tree ; and of 
Caladium, in which they are confluent, though with less of net- 
work ; of a Dioscorea, in which the venation is very like that 
of a poplar ; of Costus among the Scitamineae, of which the 
leaf is not more feather-veined than of Crinum spectabile, 
and other petiolated wide-leaved Amaryllideee ; and on the 
other hand, of a lofty Dracaena or Cordyline, which I raised 
by seed from Norfolk Island (perhaps # C. stricta) amongst 
Asphodeleae, which is completely feather-veined. Dr. Brown 
describes parallel veins even as distinguishing an Austra- 
lian species of Dioscorea from others of that genus, and some 
such variability appears according to the representations of 
different species of Peperoma. These peculiarities require 
more careful inspection than they have yet received. 
Another feature presents itself in the nature of the cap- 
sule or fruit, some opening by valves, some indehiscent; 
and in the integuments of the seed, some hard, some fragile, 
others pulpy and soft. To shew that such peculiarities are 
* The Botanical Magazine gives parallel veins to that plant, I believe in- 
correctly. 
