CROCUS VALLICOLA. 81 



Capsule pale buff, about three-quarters of an inch (0.019 metre) long, the extremities of the valves 

 produced into fine points about three-sixteenths of an inch (0.005 metre) long. 



Seed one-tenth of an inch (0.0025 metre) high, one-sixteenth of an inch (0.0016 metre) broad, the 

 surface papillose, and bearing a few hairs, dull buff, the chalaza, raphe, and caruncle not very 

 prominent, of the same colour as the body of the seed. 



Var. 1. lilacinus, Figs. 6, 10, and 11, from Stauros near Trebizond, a single corm of which was 

 obtained for me by Mr. A. Biliotti, Her Majesty's Consul at Trebizond, is much smaller than the 

 type, and the perianth segments are delicately coloured throughout with a network of lilac veins. 



Var. 2. Suwarrowianus, Figs. 2, 3, 5, 8, 12, and 15, from the Pallan ducken mountains near Erzeroum, 

 differs from the type in so many important characters that I have had some hesitation about uniting 

 it with this species. I described it in the Gardeners' Chronicle as a variety of C. vallicola, under 

 the name of Zohrabii ; but Monsieur Boissier has pointed out to me that the type specimens of K. 

 Koch's C. Suwarrowianus are evidently identical with the Erzeroum plant, and agrees with me in 

 the necessity of separating it from Herbert's type in some way. Notwithstanding its several points 

 of difference, viz, its beardless throat, the short double proper spathe, the orange stigmata, the 

 less acute segments, and the later flowering time, I feel better satisfied to place it as a well-marked 

 variety of Herbert's C. vallicola, than as a distinct species. 



Crocus vallicola is a native of the high mountains of Armenia, Lazistan, 

 Kurdistan, Georgia, and the Caucasus, at elevations ranging from six thousand to 

 eight thousand feet; occurring between 38 and 45 east longitude, and 38° and 

 44° north latitude. Tchihatcheff, in his Asic Mincur, refers to its occurrence (as 

 C. Suivarroivianiis,) in the neighbourhood of Broussa in Western Bithynia, but I 

 think the record must be erroneous. Broussa is ten degrees west of the most 



o 



westerly point at which the species has been elsewhere found. Mount Olympus is 

 the only sufficiently high land near to Broussa; but on this mountain, which I have 

 twice ascended, I failed to find it, and the species collected for me there on several 

 occasions by Mr. Gilbertson, Her Majesty's Vice Consul, have never included it. 



Ruprecht, in Regel's Gartcnflora, records its occurrence, in company with C. 

 Scharojaui, on the north slope of the western Caucasus, near the high Mount 

 Ostchen, in the district of Abadsechen. Dr. Radde, of Tiflis, informs me that he 

 has found it near the Col de Mammisson, midway between Poti and Tiflis, near 

 the source of the river Rion (Phasis), which falls into the Black Sea at Poti. Mr. 

 Ball's herbarium contains a specimen collected by Sir A. Henry Layard in 

 Kurdistan. The type form of the species was known in 1846 to Dean Herbert, 

 who described it from specimens obtained from elevated mountain hollows of the 

 Alps of Trebizond, on the mountain Koulak Dagh, near the village of Stauros. 

 Balansa, in 1866, again collected the plant from the same locality, and also in the 



M 



