$ 74 NATURAL CLASSES. SECT: XX, 6.

{ 1]

4

with the pifullum. This has alfo siven names to three clafes of the

Linnæan fyftem, monœcia, diœcia, polygamia

f

Ninthly, the connexion of the anthers, which has given the name 1

l fully extenfive and natural clafs.

To thefe varieties of fituation, proportion, and adhefion, of the filaments, may be added thofe of the anthers on their fummits: which to an attentive obferver may perhaps be as numerous as thofe of the filaments, and to thefe may again be added the various forms of the filaments, as capillary, flat, wedgeform, fpiral, feathered,&c. and alfo the various forms of the anthers, as oblong, globular, ar- rowy, angular, horned. Al which are defcribed in the Philofophia Botanica. And by an adoption of fome of thefe feparately or in conjunétion for claffical characters, I fhouid hope that new claffifica- tions might be difcovered inftead of thofe, which are fimply numeri- cal. Which might be more natural ones, lefs fubjeét to variation, eafer to be diftinguifhed from each other, and more fimilar in their good or bad qualities; and might thus add to the great beauty and utility of the prefent wonderful arrangement of fo many thoufand vegetables in the Linnæan fyftem.

6. The fame obfervations and mode of reafoning are applicable to the various orders of the fexual fyftem. Which if the great Lin- næus had fortunately deduced them from the proportions, fituations, or forms of the ftyles and ftigmas, the characteriftic figns might have been lefs liable to change by foil or climate, and many of the orders have been more natural collections of vegetables, than thofe are, which he has derived fimply from their number.

The uncertainty of the number of piftilla, and the confufon, which might be occafioned by a reliance on itj was mentioned in No. 2 of this feétion; there is a nigella pentagyna, and a nigella decagyna; there is an hypericum floribus pentagynis, trigynis, and digynis; and in the whole order of fruftraneous polygamy in

the