21 
characteristic, and that it is only by using the whole 
range or suite of the mineral, that we are enabled to 
employ this character with advantage. Thus, it is 
not sufficient to say that epidote’is green, that beryl 
is green, or that topaz is yellow ; we must mention 
every variety of colour which these minerals possess, 
aoe each species of mineral is expressed by a par- 
ticular suite or group of colours. . | 
“ Although colours are frequently applied by bo- 
tanists for distinguishing species of plants, particu- 
larly in the class cryptogamia, still they in general 
hesitate in employing them in the discrimination of 
plants in the higher divisions of the system. It is 
alleged that the colours of plants change very readily, 
_ particularly when cultivated in our gardens, and that, 
therefore, so variable a character should not be at- 
tended to. It is not denied, that the colours of plants 
frequently undergo very considerable changes when 
cultivated in our gardens; but these domesticated 
plants are no longer the natural unaltered species, 
and therefore are not objects of the attention of the 
systematic botanist. It is also known, that plants, 
even in their natural situations, owing to disease, ex~ 
perience great changes in their colours; but these 


