GEOGRAPHICAL DISTRIBUTION. 35 



Geocichla sibirica, Accentor aJpinus, Pericrocotus cinereus, Lanius 

 magnirostris , Corvus pastinator, Corvus dauricus, Hirundo daurica, 

 Grus cinerea, Grus leucogeranus, Grus monachus, Ciconia boyciana, 

 Numenius minutus, Tringa canutus, Totanus terekius, Totanus calidris, 

 Stercorarius pomarinus,Anas strepera, Anasformosa, Tadorna cornuta, 

 Anser hyperboreus, Cygnus bewicki, and possibly one or two other 

 species have a somewhat similar distribution. 



The explanation of these at first sight rather startling facts is not 

 difficult to find. In the first place, about half of the species enume- 

 rated above are winter visitors to Japan, and migrate every spring 

 and autumn along the coasts of Yezzo to and from their winter- 

 quarters. It is not surprising that they have escaped detection in 

 Yezzo, because they only pass through on migration and do not 

 winter there ; nor is it surprising that they do not winter there, 

 because the mean winter temperature of Yezzo is so much lower 

 than that of Southern Japan. According to the ' Physikalische 

 Atlas' of Berghaus the mean temperature of Hakodadi during 

 January is 4 degrees (Cent., or 7j degrees Fahr.) below freezing, 

 whilst at Yokohama it is as much above it. In the second place, the 

 remaining half of these species breed in Southern Japan, and many 

 of them may not breed in Yezzo because of the difference in the 

 mean summer temperature. According to the same authority, the 

 mean temperature of the valley of the Amoor and its tributaries 

 during July ranges from about 63° (Fahr.) in the north to about 

 73° (Fahr.) in the south. The mean temperature of Hakodadi for 

 the same period is below the lowest of these figures, whilst that of 

 Yokohama is above the highest. The mean temperature appears to 

 be a much more potent factor in the distribution of Japanese birds 

 than the distance from the land or the depth of the intervening 

 ocean. The reason why the Tsugaru Straits, or Blakiston's Line, is 

 an important one in the distribution of birds is not because it 

 represents deep sea as Wallace's Line does, but because it happens to 

 coincide with certain Isothermal Lines which bound both the breedingr- 

 grounds and the winter-quarters of so many species. 



Besides the 30 species that have been recorded from Yezzo but not 

 from Hondo, there are at least 50 species of birds which have been 

 recorded from Hondo but not from Yezzo, and there are very many 

 more that have been recorded from Yezzo but not from the Kuril e 

 Islands. Most of the former are species which breed in the Arctic 

 regions and seldom migrate so far south as Japan ; but many of 

 them are species that migrate further south than Japan, and it is 



d2 



