342 TAXONOMY. 



of DeCandolle's Thalamiflora? with some of his Caiycifloroe) ; 

 and under these the orders are arranged in cohorts, fifteen 

 cohorts in the Poly pet ahe, and ten under three "series" in the 

 Gamopetahe. The remainder of this particular classification 

 has not yet appeared in print, although partly sketched by its 

 authors. It will generally lie adopted in this country, with some 

 occasional minor modifications. 



r.'.ii;. Various modifications have been from time to time pro- 

 posed. One of the best of them in principle is that initiated by 

 Adolphe Brongniart and adopted by many European botanists, 

 which, recognizing that most apetalous dowers are reductions or 

 degradations of polypetalous types, intercalates the Apctalae or 

 Monochlaniyde;e among the Polypetalae. But this has never yet 

 been done in a satisfactory manner, or without sundering orders 

 which should stand in contiguity. 



C>'.7. It should be borne in mind that the natural system of 

 botany is natural only in the constitution of its genera, tribes, 

 orders, &c., and in its grand divisions; that its cohorts and the 

 like are as yet only tentative groupings ; and that the putting 

 together of any or all these parts in a sj-stem, and especially 

 in a lineal order, necessary as a lineal arrangement is, inu-t 

 needs be largely artificial. So that even the best perfected 

 arrangements must always fail to give of themselves more than 

 an imperfect and considerably distorted reflection of the plan of 

 the vegetable kingdom, or even of our knowledge of it. 1 



1 In the first place, tin- relationships of any group cannot always be 

 rii^litlv estimated before all its members are known and their whole struct- 

 ure understood ; so that the views of botanists are liable to be modified with 

 the discoveries of every year. The discovery of a single plant, or of a point 

 of structure before misunderstood, has sometimes changed materially the 

 position of a considerable group in the system, and minor alterations are 

 continually made by our increasing knowledge. Then the uroups which we 

 recognize, and distinguish as genera, tribes, orders, &c., are not always, and 

 perhaps not generally, completely circumscribed in nature, as we are obliged 

 to assume them to be in our classilication. This might be expected from 

 the nature of the case. For the naturalist's groups, of whatever grade, are 

 not /'<///'//'(.<, but iiltnx. Their consideration involves questions, not of things, 

 between which absolute distinctions might be drawn, but of </"//><.< of resent- 

 liliiiKT, which may In' expected to present infinite gradations. Mesides, al- 

 though the grades of allinity among species arc most various, if not wholly 

 indefinite, the naturalist reduces them all to a few, and treats his genera, 

 tribes, &c., as equal units, or as distinguished by characters of about equal 

 value throughout, which is far from being the case. And in Ids works 

 he is obliged to arrange the groups he recognizes in a lineal series; but 

 each genus or order, >xc . is very often about equally related to three or four 

 others : so that only a part of the relationship of plants can in any way be 

 indicated by a lineal arrangement. 



