148 BALFOUR—NEW SPECIES OF PRIMULA. 
“Yunnan. Mountains of the Chungtien plateau. Open 
moist alpine pasture. Lat. 27° 55’ N. Alt. 11,000 ft. Plant 
of 20-30 inches. All parts when fresh strongly aromatic with 
an odour resembling aniseed. Flowers deep plum-purple, 
almost black, drying a lighter shade.” G. Forrest. No. 10,617. 
July 1913. 
“Yunnan. Chungtien plateau. Alt. 10,000 ft. Lat. 
27° 30’ N. Plant of 2-3 ft. Flowers deep black-purple, drying 
lighter. Strongly fragrant with an odour like aniseed. Open 
moist meadows.” G. Forrest. No. 12,730. July ror4. 
This species has the darkest coloured flowers of all Candel- 
abra Primulas—perhaps of all wild Primulas. It belongs to the 
series of the efarinose purple-flowered Candelabras in which 
the redolent gland secretion is profuse. Of that series we now 
know seven species :—P. Poissonii, Franch., P. Wilsoni, Dunn, 
P. glycosma, Petitm., P. Miyabeana, Ito et Kawakami, P. 
oblanceolata, Balf. fil., P. canthina, Balf. fil et Cave, and P. aniso- 
dora, Balf. fil. et Forrest. The odour is least marked in P. 
Poissonit, where it is sometimes hardly perceptible ; and 
in P. oblanceolata it is, though evident, not very conspicuous. 
But in all the others the perfume is strong even in dried 
material. 
P. anisodora is not far removed from P. glycosma. The 
two species are marked out amongst the efarinose purple- 
flowered Candelabras by the dark, nearly black, purple 
colour of their corollas—paler in P. glycosma than in 
P. anisodora—and they are about the most aromatic of 
the series. In dried specimens the difference between 
the species is not difficult to make out:—the broader 
shorter leaves of P. anisodora and their more glaucous 
under surface; the long petiole in P. glycosma; the very 
From seed obtained by Mr. Forrest plants of P. anisodora 
have been raised and flowered at Edinburgh. The flower 
colour recalls that of some of the brown auriculas with 
yellow eye. Mr. Forrest tells me seeds of P. glycosma, 
are in his collection also, and we may hope, therefore, for 
a crop of living plants which will enable us to appraise more 
accurately the relationships of the two forms. They ought to 
