278 Notice of a Flora of North America. 



stance before us, as sections of one large family, than in endeav- 

 oring to circumscribe them as separate orders, where no sufficient 

 natural characters appear to exist. We feel convinced, indeed, 

 that the onward progress of the science will have a tendency 

 rather to diminish than increase the number of orders, which 

 have been so largely multiplied since the days of Jussieu. Nor 

 is it surprising that such should have been the case, since the 

 introduction of new plants irreconcilable with the characters of 

 any of the families indicated by that great physiologist, required, 

 from time to time, the establishment of such additional orders as 

 appeared to limit and define their respective properties and pecu- 

 liarities. But with our increased, and daily increasing, knowl- 

 edge of vegetable forms, we become acquainted with genera so 

 exactly intermediate in character between what have been 

 regarded as separate groups, or orders, existing in nature, that 

 we are compelled to view the dividing lines as of secondary, or 

 subordinate value ; and hence, we conceive, will arise the neces- 

 sity of reuniting some, perhaps ultimately many, orders now 

 considered as distinct. It is in this way only, that families of 

 plants can be preserved as natural alliances ; and any approach 

 towards artificial arrangement will be deprecated by every true 

 botanist, who feels that his province is limited to the observation 

 of kindred tribes, and the elucidation of such characteristics as 

 they are severally found to present. No doubt the types of new 

 and very distinct orders are yet to be discovered, but we repeat 

 our conviction, that for the present, at least, the number will, 

 upon the whole, be rather reduced than augmented. 



Under the sub-order LoganiecB, is described a second species 

 of the very interesting genus Coslostylis, T. and Gr., collected 

 in Texas by the late Mr. Drummond. In Valerianacece, we 

 have a new species of Valeriana from the West, and two new 

 species of Fedia, collected on the plains of Arkansas, by Mr. 

 Nuttall. These last are particularly interesting, forming a sepa- 

 rate section of the genus, {Siphotiella,) which " is almost exactly 

 intermediate between Fedia and Valerianella of Moench and 

 De Candolle, having the corolla of the former, with the fruit, 

 stamens and stigmas of the latter." The genus Valerianella, 

 of Moench, is accordingly reduced to a section of Fedia. Pass- 

 ing over the order Dipsacece, in which we have .only a single 

 (naturalized) species, we come to the immense family of Cofii- 



