254 _ JLLUSTRATIONS OF THE BOTANY OF - [Goodenoviec. 
‘‘ Ona close examination I do not find in either of these plants (which are veryclosely 
allied to each other in habit and character), those peculiarities which have been 
described by Dr.Wallich and Mr. Don, and which induced M. A. De Candolle, in his : 
excellent and elaborate monograph of Campanulee, to exclude the C. thalictrifolia not 
only from the genus, but from the order altogether. The leaves of the slender sterile 
shoots, which are in fact:simple and opposite, have been mistaken for the leaflets of a 
compound leaf, as mentioned by Mr. Don, and the supposed free ovarium proves also to 
have been an illusion. The whole of the part which contains the ovula is perfectly 
adherent, and the apparently free part is but a conical apex, dehiscent indeed at the 
maturity of the fruit in three septiferous valves (as in other Codonopsides), but con- 
taining no placente, and therefore forming more properly a part of the style, than of the 
ovarium. The irregularity of the calyx and corolla is but very slight and often scarcely 
perceptible, and the character derived from the hairy and mucronate anthers is of no 
importance, and moreover exists in one only of the above two species, which certainly 
cannot be separated from each other.” Es | 
The following are the other Campanulacee figured in this work, of which, as well as of the other new 
species, specimens have been sent to M. A. De Candolle. 
Codonopsis rotundifolia ; caule ascendente volubili, foliis alternis quandoque ciaoeia membranaceis 
leviter pilosis subtus glaucescentibus rotundatis crenatis, corolla campanulata pallide ecerulea.—Tab.62. 
fig. 2. (a). corolla spread out ; (b). flower with the corolla removed ; (ce). transverse section of capsule ; 
(d). a flower in which the calyx and corolla are divided into six lobes. 
Hab. This new and very distinct species of Codonopsis belongs to M. A. De Candolle’s first section 
and is allied to both C. viridis and purpurea. The stamens alternate with the lobes of the corolla, and 
the filaments are only a little broader at the base. The flower, with the senary division of its parts, is 
interesting, as ai - alliance of this genus with Canarina, near which it has been placed by M. A. 
De Candolle. 
Campanula Cashmeriana ; caule ascendente alétiacaistteeias ¢ foliis cmulinis scacilibus hide 
ovatis subcrenatis integrisve, ramis axillaribus floribus erectis solitariis terminatis, calycibus velutino- 
pilosis, corolla campanulata lobis calycinis triplo longioribus. —Tab. 62 fig. 1. (a). a flower with the 
corolla removed. 
Hab. Kioonthul in Cashmere. 
99. STYLIDIEZ. 
This order forms a portion of the characteristic flora of New Holland, but two species 
of Stylidium very similar to certain intratropical ones of that country were found by 
Keenig (S. uliginosum) in Ceylon, and (S. tenellum) in Malacca. Two additional species 
are enumerated in Dr. Wallich’s catalogue, S. Wightianuwm from the Peninsula, and 5S. 
Kunthii from Silhet, showing, as in many other instances, the spreading of species into 
congenial climates beyond what at first appears the natural limits of an order. 
100. GO ODENOVIEZ. 
This is another of the families of plants almost peculiar to New Holland, and of 
which a representative is found in the Indian Flora in Scevola of the tribe Scevolee of 
Mr. Brown, but formed into a distinct order by Dr. Lindley. The species S. Kenigii and 
Taccada are natives of the coasts of the Bay of Bengal and of the Peninsula. The 
former 
