of the Natural Order Saxifrageae. 429 



although I possess Hooker and Taylor's Muscologia Britannica, 

 which affords a somewhat similar plan, as appears by their 

 preface, though I have not followed it. 



The difference between the method of Dumeril and my 

 binary one may be seen by the following remarks. His di- 

 visions are not invariably worked out through dichotomies, 

 but are sometimes trichotomous, tetrachotomous, or in one or 

 more instances forkless. 



His method is essentially descriptive through every branch, 

 and without names, until you arrive at a genus. Mine is 

 definitive ; and every division, subdivision, &c., is appropriately 

 designated by a single word, which is its name. 



His dichotomies bring you down to every individual genus, yet 

 the genei'a are oiten forced into dislocations, as their irregular 

 numerical figures confess ; but which the accompanying page- 

 work corrects. My tables, or rather their dichotomies, usually 

 end in two or more genera located natiirally, and which are 

 intended to be numbered cuiTcntly; and the page-work which 

 I contemplate, and its head-lines, will have corresponding 

 numbers, conducting the reader instantly to characters, de- 

 scriptions, and synonyms, of every published genus. 



The essence, however, of both Dumeril's and my own plan 

 is, that they equally work, or ought to work, ever dicliotomoiisly, 

 and by opposites. In a word, mine at least is intended to be 

 polar ic; and the very title of the book which I meditate is, 

 Organica Corpora, methodo polari digesta. Wherefore, if we 

 can admit that the dichotomies of the tables throw their com- 

 ponent subjects as distant as possible in point of affinity from 

 each other (being frequently merely analogies), and surely this 

 is possible, and will ultimately be practicable, — it clearly fol- 

 lows, that every thing which lies, in point of affinity, between 

 such groups, must absolutely, through the constantly I'epetite 

 spirit of the arrangement, fall between them ; and therefore 

 fall natui'ally, and truly, in the very place where it ought to 

 fall. And from this it results, as I have elsewhere observed, 

 that the apparent distribution by Nature of her organized 

 forms, is a continuous one that is unravelable in the way of a 

 straight line ; but arising dichotomously from one root (which 

 1 have called matter), and proceeding in a repetitely double, 

 and in point of magnitude or quantity, unequal series; resem- 

 bling, as it were, an inverted branching and exuberant tree. 

 And I have also elsewhere observed, that every dichotomy ex- 

 tending into, and ending in, various generji, may be imagined 

 (by bending its ends together) as forming a sort of circle of 

 affinity, either oj>en at lop, or closed by the root which pro- 

 duces it. These minor circles, moreover, are but pails of 



other 



