PARING. 85 



lo range across Europe and Southern Siberia as far as the Stanovoi 

 Mountains on the shores of the Sea of Okhotsk, with little or no 

 variation in eolour. South of the Amoor the yellow suddenly dis- 

 appears from the underparts, and in South China it gradually dis- 

 appears from the mantle, leaving the Indian or tropieal form white, 

 black, and slate-grey (the usual characteristics of an Arctic race), to 

 be suddenly represented in Persia by the species found in the British 

 Islands. Neither in the east nor in the west does the Common Great 

 Tit intergrade with the Indian Great Tit; and although the Japanese 

 birds are intermediate in the colour of the upper parts, they are not 

 in the least so as regards the colour of the underparts. The Loo- 

 Choo Islands appear to have received their Great Tits from South 

 China. The Japanese Great Tits may have come from the Coreau 

 Peninsula, since so far as is known there are no Great Tits in 

 Sakhalien or in the valley of the Amoor north of its junction with 

 the Ussuri. I can see no difference between examples from Yezzo 

 and those from Yokohama. 



The probable explanation of this anomalous variation is that the 

 Japanese birds are the modified descendants of Parus atriceps hoc- 

 char tnsis, which was differentiated as a desert form in Mongolia, and 

 that the true ti'opical representative of Parus major is Parus monti- 

 cola, which ranges from the Himalayas across Southern China to 

 Formosa. 



The Manchuriau Great Tit is described as the commonest Tit in 

 Japan, abundant everywhere on the mountains in summer and very 

 common in the plains in winter (Jouy, Proc. United States Nat. 

 Mus. 1883, p. 286). 



53. PARUS VARIUS*. 

 (JAPANESE TIT.) 



Varus variu^, Temminck and Schlegel, Fauna Japouica, Ave.-^, p. 71 (1847). 



* The name of Parus varius having been applied to the Parus americanus of 

 Linneud (now known as Parula americana) as long ago as 1791 (Bartram, Trav. 

 Florida, p. :292), will probably be rejected by the devotees of the Stricklandian 

 code, who maj^, if they like, substitute for it the name of Parus sieboldi ; but 1 can 

 see no reason whatever for abandoning the name already in use. 



