^. merely to the one that precedes and that which follows it, but approaches 
^ bol 
a 
(0 XXvill | PREFACE. 
characters, we are now to proceed to subdivide these into orders; but 
in doing this must observe, that though it be quite possible to group 
genera into orders, and to place these orders into one of the great divi- 
sions above given, it is almost impracticable to place them satisfactorily 
one after the other so as to constitute a book. Each order is not allied 
more or less to several others. In arranging them, then, in a linear se- 
ries, all that we can attempt is to place those next each other that have 
common characters of most value or of greatest number; and as the - 
value and number of these are added to or diminished by the discovery - 
. of new plants and modern analyses, we must expect daily changes to : 
take place. In this consists the difficulty of the natural method, when : 
` compared with a purely artificial one ; but these difficulties are from — 
time to time smoothed down, so that ultimately it will be viewed as the — 
only useful one, while the artificial system will be almost entirely ne- - 
glected, except by those whose desires extend no farther than to make 
out the name of a vegetable. 
Jussieu considered the whole floral envelopes of the Monocotyledones 
as a calyx, whether coloured or green ; and observing that a portion of the 
Dicotyledones were also destitute of a corolla, he laid hold of this t 
divide them into two parts, as had been already done by Tournefort : at 
the same time he separated those with a corolla into one and many- - 
petaled. These sections, however, being artificial, cannot be entirel 
depended on in practice, many species belonging to a polypetalous or- 
der having the corolla abortive ; while again, as in the Pentapetaloidex of 
Ray, several of those with a polypetalous corolla have the ungues or claws 
united, so that it is in reality a monopetalous one, although arranged 
with the other. Linnæus had already introduced the mode of insertion . 
of the stamens to distinguish his class Icosandria from Polyandria ; and. 
such bringing together plants tolerably nearly allied in other respects, 
Jussieu extended the idea, and having observed that stamens not hypo- 
gynous were either perigynous or epigynous, he made use of these three 
characters to subdivide both the Monocotyledones and each of the sec- 
tions of the Dicotyledonous plants. Only one of these he found neces- 
sary further to divide, in order to separate the class Syngenesia of Lin- 
næus from those approximated to it in several other characters, but of a 
different habit or appearance, and had here recourse to the union or sepa- 
ration of the anthers. These ultimate subdivisions he termed classes, and 
to the end of them he appended one consisting of such Dicotyledones 
. 8s had the flowers unisexual, and were therefore principally contained in 
the 21st, 22d, and 23d classes of Linneus. We may here remark, that 
the stamens and corolla have always the same insertion, and that when 
a corolla is gamopetalous, or, in common language, monopetalous, the 
i 3 i 
