CHAP. XLII. ROSA CEA. PY‘RUS. 879 
Genus XVIII. 
‘| A NG 
y 
ne 
PY‘RUS Lindl. Tur Pear Tree. Lin. Syst. Icosandria Di-Pentagynia. 
Identification. Lindl. Lin. Soc. Tr., 13. p. 97.3; Dec. Prod., 2. p. 633.; Don’s Mill., 2. p. 605. 
Synonymes. Pyrus, Malus, and Sérbus, Tourn.; Pyrus and Sérbus, Lin.; Pyréphorum and 
Apyr6éphorum, Neck. 

Description. Low trees, and some shrubs; almost all deciduous ; natives 
of Europe, Asia, and North America. Some of them in great estimation 
throughout the world for their fruit; and others cultivated chiefly for their 
flowers. Under the genus Pyrus, botanists have lately united the Linnzean 
genera Pyrus and Soérbus, together with several species formerly included 
under Méspilus, Cratz‘gus, and other genera. Taking the generic character 
from the fruit, this union appears strictly in accordance with the canons laid 
down by botanists: but we cannot help stating our opinion, that it would be 
much more convenient, in a practical point of view, in establishing genera, to 
take into consideration the leaves, the character of the vegetation, the phy- 
siology, and even the habit, of the plant, than merely to draw the distinctive 
characteristics from the parts of fructification. In consequence of attending 
only to these parts of plants, the genus Pyrus, as at present constituted, con- 
tains species, such as the apple and pear, which will not graft on each other ; 
a circumstance which clearly shows that the union of these two kinds of plants 
in one genus is not a natural one. We not only think that no plants should 
be comprehended in the same genus which will not graft reciprocally on each 
other, but that plants of different habits or constitutions should not be united ; 
and, consequently, that twining plants should not be united with trees and 
upright shrubs; nor deciduous trees and shrubs with evergreens. In short, 
as we have stated in p. 812., we would form genera on a kind of na- 
tural system, from all the circumstances of the plant taken together, 
and not from any particular part, or circumstance, or class of circumstances, 
belonging to it. We think we may refer, in confirmation of the propriety of 
this doctrine, to the excellent observations that have been quoted from Dr. 
Lindley, under the head of Lowea; not without a hope, as it was in the 
commencement of Dr. Lindley’s botanical career that he brought so many 
species, dissimilar in habits, together into the genus Pyrus, that he will, in ac- 
cordance with what he has stated in the passage referred to, be at some future 
time induced to separate them, and to restore the genera Malus, Sorbus, 
A‘ria,-and Aronia. We request our readers to observe that here, as in other 
similar cases, we merely state our opinion; and that we by no means consider 
ourselves entitled to separate assemblages of species, or to alter established 
names, in any manner whatever. No one ought to do this who has not 
attained a degree of rank in the botanical world to which we have no preten- 
sion: and hence, in all those cases in which we have assumed a species to be 
a variety, we have only indicated our opinion in parentheses, leaving the reader 
to adopt it, or not, as he chooses. We may be allowed, however, to throw out 
suggestions for the consideration of botanists; and, as these are always made 
with the most perfect good feeling, and are merely submitted as speculative, 
with a view to do good, we hope our readers will receive them in the same spirit 
as that in which they are made. When a more perfect knowledge is obtained of 
all the vegetable productions of the earth, we have no doubt that it will be 
found necessary to remodel the whole of the genera, as well as to give new 
and characteristically composed names to all the species ; a labour which, great 
as it may appear at present, will be diminished to a degree scarcely credible, 
when the present chaos of names, and, apparently, of species, is reduced by 
simplification. 
To return to the genera Pyrus, we believe we may assert that some 
of the species it contains are, and have been for ages, the most universally 
3N 
