branched receptacle, and the whole umbel was considered as one aggregate flower. It necessarily followed that a Cyme, see p. 237, must be considered in the same light, nor did the sagacity of Linnæus overlook the arguments in favour of this hypothesis. Many of the umbelliferous tribe, as Heracleum, t. 939, Caucalis, Coriandrum, &c., have their marginal flowers dilated, radiant, and more or less inclined to be imperfect or abortive, thus evincing an analogy with real compound flowers like the Sunflower, which analogy is still more striking between Oenanthe, t. 363, 347, 348, and the Marigold, Calendula. So the cymose plants, as Viburnum Opulus, t. 332, bear dilated and abortive marginal flowers, and Hydrangea hortensis, Sm. Ic. Pict. t. 12, has scarcely any others. Cornus sanguinea, Engl. Bot, t. 249, has a naked cyme, C. Suecica, t. 310, an umbel accompanied by coloured bracteas, or, as Linnæus judged, a coloured involucrum, proving the close affinity between these two modes of inflorescence.
Notwithstanding all this, I presume to dissent from the above hypothesis, as offering
