the Br'itiJ}: Species cf Mentha. 1 77 



fpecies is found capitate and verticillate in the fame meadow, though, 

 I acknowledge, no intermediate varieties, by which we could trace 

 their connexion, are in general obfervable ; but I have feen one or 

 two fuch fpecimens. I have ventured to mention this novel opinion 

 in the Engl'ip Botany, p. 4.^8, antl h^n'e converfed with fcveral ac- 

 curate botanifts upon it. Moft of them arc fo averfe to it, that I 

 cannot venture abfolutely to infifl: upon it. The following arrange- 

 ment of BritilTi Mints will therefore be founded on nearly the efta- 

 bli(hed principles of inflorefcence, but I fliall mention wherever thofe 

 principles appear to me to divide a real fpecies. 



II. Having found fo much difficulty in determining the fjTecies of 

 Mentha by all the modes of difcrimination hitherto contrived, I was 

 foon convinced of the neceffity of difcovering fome other principle, or 

 of leaving the genus altogether a chaos as I found it. The expe- 

 rienced botanift well knows how peculiarly difficult it is to meet 

 with certain difcriminative marks between genera In very natural 

 orders ; happily it is not always equally difficult in very natural ge- 

 nera to trace out fpecific characters. On the contrary. Nature ge- 

 nerally makes fome one peculiar part, either in the herbage or 

 flower, fo various in the diffeicnt fpecies, and fo conftant in the 

 fame, as to afford, to a careful invcfligator of the fubjccf, a veiy 

 certain clue. Thus, the various hairincfs of the ftamina in Oro- 

 banche, the ribs of the calyx in Arcnana^ the form of the fligma in 

 Crocus, the abfcnce or prefenceof a neclarium in Cujhita, the fhape 

 of the capfulc in Fumaria and fome fpecies of Juncus^ and its fituation 

 in different Saxifraga, the various numbers of flamina and flyles in 

 Polygonum, Pljytolacca, Cerafiium, and feveral other genera ; all thefe 

 circumftances, fome of which in other inftances afford generic dif- 

 tinftions, in the above natural genera conflitute the beft and moft 

 /VoL.V. ' A a important 



