FLORA OF THE ISLAND OF SHIKOTAN. 445 
were taken into account. Miyabe arrived at the conclusion that “at the time 
of the last great southerly migration of the rich polar flora, Japan received 
few, if any, plants through the then uncompleted chain of the Kurile 
Islands”? *. If so, how could one explain the occurrence of those plants T 
which have not been found in Sakhalien, but are distributed over north- 
astern Asia, the Aleutian Islands, ete., and the Kuriles as well as Yezo, 
and even farther south in Honté (on high mountains)? At the same 
time, certain plants which have their head-quarters in North Japan, but 
are not known from Sakhalien, are nevertheless met with in the Kuriles. 
The examples given below are taken only from the Shikotan Flora, but if I 
were allowed to take other islands of the Kurile group into consideration, 
I could certainly give many more instances. It is also of great interest, as 
well as of much importance, to compare the flora of the alpine region of the 
high mountains in Central and Northern Japan with that of the Kuriles and 
of Sakhalien. There is no room at present to discuss this point in detail ; 
but, so far as I can say with certainty, the alpine flora of Japan is very 
much of an arctic character, and the plants representing it are by no 
means always met with in Sakhalien, Confining ourselves to the island 
of Shikotan, we have there a good many plants common to it and Yezo or 
to Yezo and Hontó, but not found in Sakhalien. Miyabe’s theory would be 
sufficient to explain the presence of the plants common to Japan and 
Sakhalien, but not the occurrence of the elements which are found in Japan 
and the Kuriles, yet absent from Sakhalien. Those which are not found in 
Sakhalien must have come through the Kuriles, or, at any rate, must have 
been equally distributed over these particular districts, either when they were 
compelled to migrate towards south, or when they returned north after the 
amelioration of the climate. Taking examples from Shikotan alone, we know 
that there are 42 species occurring in Yezo and Hontó as well, but not in 
Sakhalien, and some of them even belong to a comparatively warmer climate. 
These facts show how intimately the Kurile Islands are connected with 
Hontó through Yezo, and much less so with Sakhalien. I may also mention 
here that the flora of Yezo has a closer relationship to that of Honto than 
to that of Sakhalien. It was put forward by Blakiston $ that Yezo and 
* Miyabe, in Mem. Boston Soc, Nat. Hist. iv. (1890) p. 212. 
+ For example: Stellaria ruscifolia, Geum calthifolium, Erigeron salsuginosus, Arnica 
unalaschkensis, Swertia tetrapetala, Polygonum viviparum, P. polymorphum, Euphrasia 
mollis, Platanthera chlorantha, P. Chorisiana, Cnicus kamtschaticus, Carex scita, and 
Clematis fusca. 
I E. g., Anemone Taraoi, Maackia amurensis var. Buergeri, Saxifraga cortusifolia, S. fusca, 
Leucothoé Grayana, Primula modesta, Syringa amurensis var. japonica, Pedicularis japonica, 
Salix Reinii, Carex Wrightit, Taxus cuspidata, etc. 
$ T. W. Blakiston, “ Zoological Indications of Ancient Connection of the Japan Islands 
with the Continent,” in Trans. Asiat. Soc. Japan, xi. (1883) pp. 126-140. 
