Page:Illustrations of Indian Botany, Vol. 2.djvu/114

ILLUSTRATIONS OF INDIAN BOTANY. {{c|{{sc|Class 3d Calyciflorae Epipetalae.}}

" Torus lining the tube of the calyx and forming at its extremity a small disk (epigynous) on the summit of the ovary which coheres with the calyx tube. Petal* distinct and with the stamens inserted on 'he outside of the disk."

In our fl >ra two orders only are referred to this class, UmheUi ferae and Araliaceae ; a third comae, the only other, is also found in the northern parts of India, Wallich having discovered two handsome species of corn us in Nepaul, and Blume 5 in Java.

{{c|LXXVIII.—UMBELLIFERAE.}}

A large and most natural order, hut at the same time by far the most difficult we have yet had to encounter, whether our object be to dispose of its genera into suborders and tribes, or its species into genera and suh-genera. Various indeed are the efforts which, from the days of Linnaeus to the present, have been made systematically to distribute the species of this or- der, but each in its turn has been found wanting in some particulars, in proportion as our know- ledge extended, and at this moment, two arrangements contend for the palm of adoption by future Botanists.

Linnaeus viewed the umbel as a sort of compound flower with a branched in place of solid receptacle and considered the involucrum as a sort of exterior calyx. Observing that most species had a double umbel, these, he designated univeml and partial umbels. The first consisting of several rays springing from the apex of the branch, each of which again terminat- ed in a smaller umbel of flowers forming the second. Further observing, that some species bad involucres at both, some at the second only, and some at neither, he, with his usual tact, availed himself of these simple and easily observed distinctions to distribute the 44 genera which he knew into three principal groups. His first division including all those with both universal and partial involucres his second all those with partial only, and his third those having neither. His generic characters were then taken from the flowers, the involucrum and form of the seed. Spiengel, whose system r.ext claims attention as being that which has met with most support, forms his sections on the seed, 1 compressed flat — 2 rather solid winged — 6 fruit blad- dery — 4 fruit coated — .5 fruit armed — 6 fruit solid, naked — A variety of secondary characters are afforded by the involucra the costae and the intermediate furrows of the seed, whether with or without vi.ttae, &c. Professor Hoffman, of Moscow, proposed another arrangement the sec- tions of which are founded on the seed and petals — seed furnished with, or without, vittae — whether the vittae are epicarpous or epispermous — whether the vittae are on both the dorsum and commissure or on one or other, &c. The generic characters are taken from the involucrum, petals, seed, and the number of vittae bet ween the costae, &c.

These have successively given way to the arrangement of Koch and DeCandolle, which seems now almost universally adopted — though not unquestioned, as a more recent arrangement has been proposed, the peculiarities of which will be noticed in the proper place.

Enough has here been said, though but a fraction of what might have been, to show how much this order has engaged the at tention of Botanists, aqd to prove how difficult it must be to chalk out a just arrangement, since all that has yet been attempted must to a greater or less extent be considered failures, except, perhaps, the last and that remains to be ascertained, as no qualified Botanist seems yet to have examined it. critically.

The plants of this order, which exceed 1000 species are nearly all herbaceous, some are suffruticose but none either shrubs or trees. The stems of the larger ones are frequently hol- low others are filled with a large Medulla or Pith, with thick fibres scattered through it like the stems of monocotyledohous plants. The leaves are almost always alternate, variously divided, and sheathing at the base. The flowers are umbelle>l,t be umbels either perfect, that is.having both general and partial umbels, or imperfect that is simple, or capitate, or dichotomous, or irregularly divided. The involucra are equilly variable hence they have been employed to supply section- al, generic and specific characters, which the fl>wers, owing to their uniformity of form through- out the order were supposed less fitted to supply.

" Calyx 4-tooihed, or entire. Petals 5, inserted on the outsile of a fleshy disc, around the top of the ovary, alternate with the teeth of the calyx, often inflexed at the point, the inflex-