Page:Transactions of the Linnean Society of London, Volume 10.djvu/445

 {| width="100%" exterior perianthium in Dipsaceæ may perhaps most properly be deemed inferior, only embracing the seed closely, being enlarged and hardened in the fruit; witness Scabiosa. Now this is precisely the case with what I have above described as the inner perianthium of Brunonia, the outer one, of four leaves, not being analogous to any thing in Scabiosa, except the solitary scales or leaves in many species accompanying each flower. Can it be possible, therefore, that what I have taken for the inner is really the only perianthium in Brunonia, and exactly analogous to the outer one in Scabiosa? They both alike, in an indurated state, envelop and crown the ripe seed.
 * width="10%" align="left" |
 * width="80%" align="center" | Dr. &apos;s Account of Brunonia.
 * width="10%" align="right" | 369
 * }
 * }

If habit were to be much insisted on, nothing can be stronger in my favour; for, besides the inflorescence, when I lay the dried specimens of the two Brunoniæ by the side of Scabiosa cretica and graminifolia, nothing can be more striking than the exact agreement of the foliage of B. australis with the former, both in shape and colour; while the same circumstances, including the silky pubescence, no less agree in B. sericea and S. graminifolia. I am, however, aware how treacherous these analogies are in the productions, whether vegetable or animal, of New Holland, but their technical characters are no less so. If it would lead us widely astray to make the wonderful Ornithorinchus a bird, on account of its beak, it would be equally dangerous, were any botanist to refer Brunonia to the Campanulaceæ, for the sake of its stigma alone. "Upon the whole," as Mr. Brown very candidly remarks, "instead of our being able to determine the order to which this genus belongs, Brunonia seems to afford no small proof of the limits of these groups being purely artificial; for does it not break down the barrier between Syngenesiæ and Campanulaceæ, Dipsaceæ and Globulariæ?" To this I most heartily subscribe; but if it leads to the overthrow of artificial definitions,