This page needs to be proofread.

OF PLANTS CALLED COMPOSITE. 275

much resembles ; or thirdly, that it is more analogous to m the outer calyx of Scabiosa, which M. Cassini seems to consider different in its nature from both these parts.

But the order of expansion in Lagasca, which is, though with some degree of irregularity, from centre to circum- ference, or that of the compound capitulum, seems to decide the question respecting the envelope of each flower, and to establish its identity with involucrum : nor does this involucrum differ materially from that of Echinops, except in the reduced number and confluence of its component parts.

The real structure of Casulia is perhaps less obvious.

This genus, which was first published by Dr. Roxburgh, 1 is referred by him to Polygamia segregata ; the tubular envelope or involucrum of each floret being described as distinct from the included ovarium.

Kcenig, on the other hand, by whom the genus was dis- covered, and whose account of it is given in the same work, describes the partial involucrum of Roxburgh as the sur- face of the ovarium itself ; its segments being, according to him, a pappus of two leaves. And lastly, Willdenow, regarding this involucrum as merely palea? of the recep- tacle, refers the genus to Polygamia sequalis ; in which order it is continued, both in Persoon's Synopsis, and in the second edition of Mr. Aiton's Hortus Kewensis.

This last view of the structure seems the most erroneous of any, and was probably adopted by Willdenow, in con- sequence of his having added to the genus a second species not really belonging to it, and which I shall have occasion to notice in another part of my subject.

An examination of the parts of fructification in different stages reconciles the opposite statements of Kcenig and [95 Roxburgh; for I find that at the time of flowering the envelope of each floret is, as Roxburgh has figured it, dis- tinct from the ovarium, with which, however, in a more advanced stage its tube becomes firmly united ; a fact that sufficiently accounts for Kcenig's description.

1 In Corom. Plants, i, p. 64, t, 93.

�� �