144 
65. (5.) Azara Gilliesii, (Hook. et Arn.); foliis ovalibus 
spinoso-dentatis glabris longe petiolatis, floribus axil- 
laribus simplicissime racemosis, pedicellis brevissimis 
ex excavationibus rachidis crassz oriuntibus, perian- 
thio 4—5-partito patente.—Chili, Dr. Gillies.—Nom. 
vernac. Navanjillo.—Closely allied to the last species, 
but very distinct in the inflorescence. If there be sti- 
pules at all, they must be deciduous, for there are none 
on our specimens. We have seen no stamens, and from 
the appearance of the fructiferous receptacle, we con- 
sider this, as well as A. celastrina, to be dicecious. The 
peduncle or rachis is thick, and furnished with a number 
of little excavations, out of each of which springs a short 
stout pedicel; at the lower edge of the excavation is a 
small deciduous bractea. 
66. (1.) Helianthemum Brasiliense, Pers.—Pampas of Buenos 
67. 
68. 
69. 
Ayres, Dr. Gillies. 
(1.) Viola maculata, Cav.—Conception, Messrs. Lay and 
Collie ; Cuming (N.795.)—var. 8. minor; foliis subrhom- 
boideis, petalis saturatioribus, calyce glabro.— Cordillera 
of Chili, Cuming (N. 194.) 
(2.) Viola rubella, Cav.— Conception, Mititei Messrs. 
Lay and Collie; Cuming (N. 130.) Valparaiso, Mathews 
(N. 248.); Bridges. Chiloe, Cuming (N. 25.)—Some 
of our specimens approach so near the V. stipularis, 
Cav. (V. setosa, Sm.), that we do not see how the latter 
can be distinguished. To V. rubella, Gingins has already 
united the V. chamedrifolia, R. and P., and V. tereti- - 
folia, Humb. and Kunth. 
(3.) Viola congesta, (Gill. mst. ); trunco crasso tortuoso, - 
foliis in apicem ramorum dense stellatim congestis 
rotundato-spathulatis ciliatis superne grosse serratis in - 
sinubus glandula alba supra subrugosis subtus glandulis 
oblongis fuscis obsitis, petiolis latis, stipulis 3 quorum 
intermedia bifida, pedunculisfoliorum longitudine, (TAB. 
XCVII.)—Nearthesummitof Cerro dela Polcura, Andes 
