MB. M. A. C. HINTON ON HEBRIDEAN BIAMMALS. 839 



the M. arvalis group in a modern guise does not appear until the 

 late pleistocene era. Secondly, we have to face the remarkable 

 fact that both E. alstoni and M. agrestis exsid, with its ally 

 macgillivraii from Islay, are most closely related to the Skandi- 

 navian species. Moreover, M. agrestis macgillivraii appears to be 

 a little more primitive in its organization than exsul ; the latter 

 form in turn a little more primitive than true agrestis. The 

 interpretation of these peculiar facts which we suggest is that 

 the Hebrides have fonned a refuge for some very old mammals ; 

 that from the former continuous land area, of which they are 

 now the disconnected remnants, certain of these old mammals 

 were dispersed along former land bridges to Scandinavia. The 

 proximity of the Atlantic would ensure the existence of much 

 milder conditions in the Hebridean area during the glacial 

 period, at whatever moment in pleistocene time that remarkable 

 event transpired, than would be present elsewhere to the east. 

 It will be seen therefore that we are practically in full accord 

 with the views put forward by Mr. L. Stejneger in his very able 

 and suggestive paper " On the origin of the so-called Atlantic 

 Animals and Plants of Western Norway " (Smithsonian Misc. 

 Coll. vol. xlviii. p. 458, 1907). Just as the presence of S. grantii 

 implies that Islay was separated from the mainland at an earlier 

 date than were the other islands of the Outer Hebrides, so the 

 existence upon it of M. agrestis macgillivraii and of a distinct 

 subspecies of Stoat, Mttstela erminea ricino) (found in Jura as 

 well), may be taken as evidence that it was detached from the 

 common Hebridean land surface at a relatively early moment. 



The early detachment of Islay and its peculiar mammals has 

 an impox'tant bearing on the origin of the Irish Fauna. It 

 indicates that Islay and Ireland have not been connected since 

 the genus Ilicrotus and Sorex araneus reached the Inner Hebrides. 

 And if that were so, the existence of the deep North Channel 

 between Ireland and South Scotland must at the same time have 

 interrupted communications in that direction also. We ai'e 

 thus unable to accept E. R. Alston's suggestion (Fauna of 

 Scotland, 1880, p. 5) that Ireland received its mammals from 

 Southern Scotland, and are forced to derive the Irish Faiuia from 

 England and Wales (see Barrett-Hamilton, Clare Island Survey, 

 Mammalia, Proc. Roy. Irish Acad., March 1912). 



EXPLANATION OF PLATE LXXXIV. 



Fig. 1. Sorex grantii. Natural size. 



2. Sorex araneus castanetis. Natural size. 



