COMMENTART. 49 



purpureo-emruleum. In another point of view the motive 

 which prompted this decision is an excellent one : too much 

 cannot be done to oblige authors to, be accurate; now, to 

 assert that an offspring is of such or such a parentage, when 

 no evidence can be produced, is anything but accuracy. 



40. The system we recommend for cultivated plants (Art. 

 14 and 40) may be recapitulated as follows : — 



(1.) Adopt for the principal modifications of species the 

 names and forms in use for uncultivated species, that is, 

 class subspecies, varieties and subvarieties according to im- 

 portance ; say, where possible, which are habitually heredi- 

 tary (races, comparable to subspecies), which are less con- 

 stantly so (subraces, varieties), which are rarely so (sub- 

 varieties) : employ for all these degrees, as well as for their 

 half-breeds, Latin adjectives, as in the case of ordinary 

 species. 



(2.) For modifications of a lower kind, the number of which 

 is unlimited (seedlings, half-breeds of low degree, sports), 

 take names from modern tongues, entirely different from 

 Latin ones, such as horticulturists are in the habit of apply- 

 ing. 



By means of this double combination the chief modifica- 

 tions, interesting to general natural history, are brought into 

 connection with scientific forms, whilst at the same time the 

 innumerable unimportant modifications produced in gardens 

 bear distinctive names. In books there will be no longer 

 a possibility of confounding them with botanical species. 

 This is a necessary precaution ; horticulturists, for the sake 

 of abridging, being wont to drop the names intervening be- 

 tween the generic name and that of the seedling or sport. 

 Instead of saying, Brassica oleracea, aeephala, vulgaris, 

 viridiis, Cavalier, expressing thus completely the relations of 

 the cavalier-cabbage with other species of Brassiaa, they 

 must needs say Cavalier cabbage. If, instead of CavaUer, 

 there was such a name as grandis, they would infallibly call 

 it Brassioa grandds, and it might then very well be taken for 

 a spontaneous species. 



This source of ambiguity must be avoided henceforth. 



D 



