Mr. Roscor’s Remarks on Dr. Roxburgh’s Description, &c. TN 
fromthe more im meminit parts of fructification. That PRINTER 
founded on the corolla may occasionally be of use, even in deter- 
mining the genus, I shall not deny ; but that any distinctions 
which can be derived from a corolla, which is strictly speaking 
-monopetalous, can be so described as to characterize the many 
genera. of which this order is composed, I greatly doubt; and 
the ineffectual attempts that have been heretofore made for that 
purpose may be allowed to justify such distrust. On the other 
‘hand, the distinctions founded on the anthera and its filament, 
are not only characteristic and permanent, but are sufficiently 
various-and ‘distinct to extend throughout the whole order, and 
to assign to each genus its proper situation. That these distinc- 
tions are confirmed by many others, as well from the sections and 
form of the corolla, as from the general growth and habit of the 
plant, is certain ; but as these peculiarities have not been found 
sufficient to lay the foundation of an intelligible and entire ar- 
rangement, they must always be considered in a subordinate 
light ; in which, however, they may occasionally be found of con- 
siderable use. ! 
With these preliminary remarks, I shall now proceed briefly to 
point out such parts of Dr. Roxburgh’s valuable Paper as seem 
to me to require observation ; being well convinced that, from 
the interest you take in this subject, any apolegxs for the trouble 
i may give you will be unnecessary. SE 
Of Canna, it appears the:garden at Calcutta possesses but one 
species, the Indica, of which, we are informed, the red and yellow 
varieties are-common in every garden in India. shall, however, 
take this opportunity of contributing, as far as in my power, 
towards the elucidation of this genus, which, as you have ob- 
served in Exotic Botany, (page 83,) wants a thorough investiga- 
tion. The species may be divided into two sections; 1. C. with 
the 
