Mr Roscok, who first bestowed on C. indica, in the 8th vo- 
lume of the Linnean Transactions, a character by which it 
may be distinguished from its congeners; and I have the sa- 
tisfaction of being able to say, that on Mr Roscor’s inspec- 
tion of the accompanying figure, he pronounced it to be a spot- 
ted flowered variety of his Canna indica, the original colour of 
which is wholly scarlet. The rarity of the individual in que- 
stion must plead my apology for introducing i in this work a va- 
riation from the common type of a species,—a practice that will 
be avoided, in futute, as much as possible. 
_ What plant was originally intended by Linnavs to bear 
the name of Canna indica, will now, perhaps, be never pre- 
cisely determined ; or rather, if one may judge from his refe- 
rences, that author had, more probably, no one particular species 
in view. Mr Arron, in the first edition of the Hortus Kew- 
ensis, included four species, according to Mr Roscor, under 
that name. In the second edition, Mr Roscor’s C. indica is 
taken up, and his specific character adopted ; but there is a re- 
ference to a figure in REpourr’s Plantes Liliacées (t. 201.), 
which, having but two segments to the inner linib of the co- 
rolla, cannot belong to this plant, and which, though not quo- 
ted by Mr Srms in the Botanical Magazine, seems to me 
exactly to correspond with C. speciosa (t. 2317.) of that work. 
The Messrs SHerHern have lately informed me, that not 
even in their rich collection of Scitamineous plants at Liver- 
pool, are they sure of possessing, at this time, the true C. i- 
_ dica ; and that the present variety was quite unknown to them, 
as well asto Mr Roscoe. 
ge 
Fig. EE cone dhe ge Fig. 2. Germen and calyx. Fig. 3. Outer limb 
the corolla, nck 4. Flower deprived of the ey and outer limb.— 
Me the ntl oe 
