142 PRINCIPLES OF BOTANY, ETC. 
17. Polypetali. 
18. Flores incompleti.—Impertecti. 
This system is very easily understood, and the se- 
lected character is to be found without any trouble. 
But the regularity of the corolla, which often varies 
in the different species of a genus, and the number 
of petals, which likewise not unfrequently vary, 
make it difficult in practice. The orders are taken. 
from the fruit according as it is naked, (fructus 
nudus), or contained in a pericarp; and this last is 
distinguished according as it 1s dry (pericarpium sic- 
cum), or fleshy (pericarpium carnosum). 
§ 184. 
CHRISTIAN KnauT has adopted Rivinus’s me- 
thod almost unchanged, but in some degree re- 
versed. ‘The classes he forms from the number of 
the petals, and his subdivisions he takes from their 
regularity or irregularity. But he denied that there 
were any flowers without a corolla, or that there 
was such a thing as naked seeds. 
§ 135. 
The System of TouRNEFORT was for a consi- 
derable time the favourite system of all botanists, 
and it deserves particular attention. 
Herbe et suffrutices. 
1. Floribus monopetalis campaniformibus. 
----——~ infundibuliformibus et 

Pa (ote area msinin rd 
rotatis. 
yn a anomalis. 
4, Flor- 
