20 
PRELIMINARY TREATISE. 
to make such an alteration, on account of the extreme incon- 
venience of such an extensive change. We are not certain 
that there is an exact analogy as to the diversification of 
individuals in their generation, between animals and vege- 
tables, the latter having branched more extensively into 
different grades of variety, some of which can re-unite with 
facility, while others seem so constitutionally altered, as to 
be almost incapable of now mixing with the other grades ; so 
that the arrangement and nomenclature is not objectionable, 
if we only bear in mind that the botanic genera define the 
individuality of kind. If any botanist should tell me, that 
in framing a generic character he was not seeking signs 
whereby to come at the knowledge of a fact, namely, the in- 
dividuality of the thing described, I should answer, that if 
such is not his object, whatever name he may assign to the 
character, he has nothing in his science real or tangible; 
that it is the vague offspring of variable opinion ; that he is 
amusing himself with child’s play, and might as well build 
castles with the sand. If, instead of proceeding by single 
steps from the first point of general agreement to the last 
important point of generic identity, we throw together a num- 
ber of kinds, and include jthem under the character of an 
order framed to cover their diversities, we are arranging 
them in groups to which no such limits were assigned by the 
Creator ; and our classifications, by whatever name we may 
call them, are artificial ; and we must be liable to find our 
labours, on further investigation, defeated by the reality. If 
we divide all vegetables, according to their agreement, by 
single points of difference, till we reach that which separates 
them from their nearest kindred, every plant, when rightly 
known, will have its proper place, and the whole system will 
be built on truth, and not the offspring of opinion. 
To make a perfect arrangement of the monocotyledonous 
plants, to which division the plants of which I am about to 
treat belong, it would be requisite that I should have a tho- 
rough knowledge of all the orders it contains ; and even in 
framing the character of Amaryllidaceae, I must be liable to 
lapses, from the want of perfect acquaintance with those from 
which the}^ are to be separated; and I must apologize for any 
such errors, if they should occur, which will be easily recti- 
fied by those who are more intimately acquainted with the 
other portions of the previous divisions. If any botanist 
should, therefore, perceive that my limitation does not effec- 
