302 Dr. Smith's Botanical Charallers of four Netv-Holland Planis, 



this cbara61er, flight as it is, connedlecl with any peculiarity of habit 

 by which a Melaleuca can be known from 2^ Metrofideros\ nor, I 

 believe, would any bo' anhl venture to guefs at a Melaleuca without 

 feeing the flamina, in which the only peculiarity of the genus refides. 

 What then is to be done, when even this peculiarity feems eluding 

 our grafp? We can only retain the genus as an artificial one, along 

 with many other fuch, till the fcience be arrived at a greater degree 

 of perfection; keeping, m the mean limQ, natural orders in view as the 

 grand obje6l of our fyftematic inquiries, and cherifliing every truly 

 natural genus as a fixed point, on which we may found the principles 

 of future difcoveries. 



I. * Eucalyptus Wi^r_g-w^A7, operculo conico magnitudine caly- 

 cis, umbellis lateralibus, foliis ovatis margine incralTatis. 

 E. marginata. Tionn. Hort. Cant. ed. 2. 10 1.? 



Mr. Aiton favoured me with fpecimens of this plant three years 

 ago from Kew Garden. The feeds were brought from Port Jackfon. 

 Its leaves agree very much in form with thofeofE. robujla, (next to 

 which ir ought to be placed,) but the footftalks are fhorter, veins 

 more prominent, and the margin more thickened, fomewhat carti- 

 laginous, and redditli. The umbels arefolitary, axillary, and fimple. 

 Flowers fcarcely one-third of the fize of the robufta^ and their covers 

 are neither broader than the calyx, nor longer; neither are they 

 contraded in their middle. The flowers much refemble thofe of my 

 E. pilularh^ but the leaves are totally different. 



XXL Addl- 



