Page:Illustrations of Indian Botany, Vol. 1.djvu/335

ILLUSTRATIONS OF INDIAN BOTANY.

167 draw useful characters from it. Of the genus Zanthoxylon for example, in all the species I have examined, 4 in number, the cotyledons are large with the albumen, if indeed such it be, reduced to a mere membranous covering. In Tuddalia bilocularis it is altogether wanting. True these may, and perhaps ought to be removed from the genera to which they are referred on that very account, but the fact of its absence in some and presence in others, shows how value- less it is an ordinal character in this tribe, and is still further shown by what is observed in Rutece and Diosmece, which, though so nearly related, are yet distinguished by the one tribe having albumen, the other being exalbuminous. Having premised these few remarks on this point of structure, I shall proceed with the extract.

"This is one of the families which comprehend genera with both distinct and concrete carpella; the latter are often entirely distinct, even in the ovarium, but more frequently there is a union, or at least a cohesion, of the styles, by which their tendency to concretion may be recognized. In a few instances the carpella are absolutely solitary.

"The place originally assigned, and for a long time preserved, for most of the genera of Zanthoxyleae, proves sufficiently how near the affinity is between them and Terebinthaceae. If, with Messrs. Brown and Kunth, the latter are divided into several orders, Zanthoxyleae will be most immediately allied to Burssraceae and Connaraceae, agreeing with the former in the genera with a simple fruit, and with the latter in those with a compound one. Notwithstanding the distance which usually intervenes in classifications between Aurantiace.ae and Terebinthaceae, there are nevertheless many points of resemblance between them ; Correa has pointed out a passage from one to the other through Cookia, Kunth, in new-modelling the genus Amyris, and in considering it the type of a distinct order, suspects its near affinity with Aurantiaceae ; we cannot therefore, be surprised at the existence also of relations between the latter and Zanthoxyleae. A mixture of bitter and aromitic principles, the presence of receptacles of oil that are scattered over every part, which give a pellucid dotted appearance to the leaves, and which cover the rind of the fruit with opaque spaces, — all these characters give the two families a considerable degree of analogy. This has already been indicated by M. de Jussieu in speaking of Toddalia, and in his remarks upon the families of Aurantiaceae and Terebinthaceae; and it is confirmed by the continual mixture, in all large herbaria, of unexamined plants of Terebinthaceae, Zanthoxyleae, and Aurantiaceae. The fruit of the latter is, however, extremely different; their seeds resembling, as they do, Terebinthaceae, are on that very account at variance with Zanthoxyleae, but at. the same time establish a further point of affinity between them and some Rutaceous plants which are destitute of albumen. Unisexual flowers, fruit separating into distinct cocci, seeds solitary or twin in these cocci, enclosing a usually smooth and blackish integument, which is even sometimes hollowed out on its inner edge, a fleshy albumen surrounding an embryo the radicle of which is superior, are all points of analogy between Zanthoxyleae and Euphorbiaceae, particularly between those which have in their male flowers from 4 to 8 stamens inserted round the rudiment of a pistil, and in the female flowers cells with 2 suspended, usually collateral, ovules. Finally, several Zanthox yleous plants have in their habit, and especially in their foliage, a marked resemblance to the ash. The dioecious flowers of Fraximm, its ovarium, the two cells of which are compressed, having a single style, 2 ovules in the inside, and scales on the outside, and which finally changes into a samara which is 1 -celled and I -seeded by abortion, all establish certain points of contact between Ptelea and Fraximus." Adde Juss.

The greatest number of the plants of this order are found in tropical America, a few are natives of Africa: two of which, Zanthoxylon Senegalense and Z. Leprieurii resembling our Z. Rhetsa, are from Senegambia. On continental India the number hitherto discovered is not great : they appear more numerous in the Islands to the eastward. Blume, including Rutea, has from Java 14 species, while Roxburgh has for India only seven, Wallich's list has about 28 for all India. The Peninsular flora at the time of our publication only presented a catalogue of 10 species for the whole order Rutacece, but for Zanthoxyhce excluding Rutece and AHanthece only 5, one of which is since excluded, and the genus of another still doubtful. In addition to the species described in the Prodromus, I have since found Zanthoxylon (Fagara, Roxb.) triphyllum and two new species referable to the subgenus Langadorfia, also Toddalia Jloribunda, Wallich, thus adding four species to our list which now exceeds Roxburgh's.