112 Mr. 0. Salvin on the Genus Cinclus. 



districts ; 4. Cinclus mexicanus, wiiich is represented by two 

 forms ; and 5. Cinclus leucocephalus, also represented in different 

 localities by two members, — the total number of recognizable 

 species being thirteen. 



As all Dippers frequent subalpine streams, their general dis- 

 tribution is necessarily divided into isolated districts correspond- 

 ing to the position of mountain-chains where such streams exist. 

 Theii iion-migratory habit, in the wide sense of the term, 

 maintains this isolation ; and if the scanty materials at my 

 disposal indicate any facts concerning their geographical dis- 

 tribution, they show that each race or representative species is 

 isolated and restricted to its subarctic or subalpine region, ac- 

 cording as each belongs to a northern latitude or to a corre- 

 sponding climate found in a southern mountain-chain. The 

 distribution of the several members of groups inhabiting the 

 northern hemisphere seems fairly traceable to the same cause 

 which so satisfactorily accounts for the distribution of arctic 

 plants — namely, the glacial period, when the supervening cold 

 drove the then existing species before it into wove southern 

 latitudes, and afterwards, as it receded, left individuals stranded, 

 as it were, on every available mountainous region, where isola- 

 tion under slightly different external influences brought about 

 the various modifications that are now found. 



The British Dipper is only a partial migrant; it leaves the 

 higher streams in winter, and seems to follow them down, 

 keeping to the same watershed. In the lower and stiller 

 waters of rivers having mountainous sources fewer suitable 

 feeding-grounds for the Dipper are found; hence the wider 

 dispersal of individuals. In the case of the northern species, 

 it would seem that cold winters drive individuals beyond the 

 southern extremity of the Scandinavian peninsula ; hence the 

 occasional occurrence of C. melanogaster in Norfolk, Holland, 

 and the southern shores of the Baltic*. 



* If, as I suspect, C. peregrinus, Brehm, is the same as C. melanogaster, 

 it shows that this race perfonns a more lengthened migration than any 

 other, and that when once beyond the southern extremity of the Scan- 

 dinavian peninsula it passes the low lands bordering the Baltic, and 

 reaches the mountain-streams of Central Germany. 



