THE RUSSIAN F UK-SEAL ISLANDS. 25 



and determine its character. I regard them as arcto-tertiary species, of which many, at least, liave 

 formerly had a wider distribution than at present. 



The other element consists of species which by their present distribution are indicated as arctic- 

 aljiino. Several of these are to be regarded as among the characteristic i)lauts of the present Arctic 

 regions. 



The Commander Islands, with the other Aleutian Islands, compose a lloral district which forms a 

 transition chiefly betweeu three other districts, viz, the Manchu-Japauese, the Americo-Pacific, and 

 the Arctic district, although less closely related to the latter than to the other two, the northern 

 outpost of which it may be regarded to represent. 



Dr. Erust Almquist lias investigated the liclieus of Bering Island and lias pub- 

 lished a very interesting account of bis studies (Vega Exp. Vet. lakt., iv, 1887, pp. 

 518-511), 521, 521-531), in wliicli lie gives an ingenious explanation of the curiously 

 sculptured surface of the heath-like plant covering of the lower plateaus as due to a 

 natural rotation of the i)lants composing It. 



The general character of the Hora is very much like that of the treeless regions 

 of Northern P^urope, the most discrepant features being the splendid rhododendrons 

 {R. himfschaticiim and chrysanthum) and the beautiful dark- maroon-colored Saranna-lily 

 (Fritillaria cnmtftchatceiisis), the bulbs of which the natives gather for food in late 

 summer. These plants indicate the close relationships to the flora of Kamchatka and 

 the other Aleutian Islands. The plants of both islands are in most cases identical, 

 but the manner of their immigration very likely has caused the occurrence of some 

 species in one island which are absent in the other. Thus I have from Copper Island 

 the conspicuous yellow flowering Viola hiflora (also found by me at Petroiiaulski), 

 which I failed entirely to find on Bering Island, and which I could scarcely have 

 overlooked. 



The islands are completely destitute of trees, the few species of Sali.i; ryriin, and 

 Betula hardly ever rising above 6 to 8 feet, though I have a section of Bctnla 

 eversmaniii from Bering Island, with a diameter of 2 inches at the root. The Pyrus, 

 in many places, forms extensive, nearly impenetrable thickets. 



There are two tolerably well defined belts of vegetation on the island, one a very 

 luxuriant growth of higher plants in the lower valleys and plains, the other a heath- 

 like formation above the former. 



The luxuriance of the vegetation in the lower belt, due to a rich soil and extreme 

 moisture, is marvelous. Some species familiar to me from boyhood I could hardly rec- 

 ognize in the enormous specimens before me. 8uch plants as Anemone narciHsiflora 

 and Geranium crianthum sometimes reach a height of 3 feet, while in some particuliirly 

 favored localities many acres of ground may be found covered with an almost impene- 

 trable jungle of Archangel ica, Heracleum lanatum, Artemisia Hlesii, Picris ja-ponica, 

 Spira'a kamtuchafiea, Aconitum, Veratriim album, etc., often reaching a height of 5 to 

 6 feet. The exuberance of the umbellifers, particularly near the coast, is very striking, 

 as shown in the accompanying j)hotograi)li of Heracleum lanatum (pi. 15a). Near the 

 beach this belt shows the usual influence of the neighborhood of salt water in the 

 presence of such plants as Lathyrus maritimiis, Mertennia maritima and Ligusticum 

 scoticiim. 



The heath commences often quite abruptly above this belt, covering the surface 

 of the beach terraces and the lower plateaus. Its presence does not depend so much 

 upon the altitude as the character of the ground, for where the coast escarpment is 

 low the heath formation commences even at an altitude of 20 to 30 feet. The fuuda- 



