430 MURIDAE—AGRESTIS 
only be distinguished by a peculiarity of the enamel-pattern, occurring 
as a normal feature, instead of, as in other members of the group, as a 
rather rare anomaly. If, however, a Scottish origin for the Skandinavian 
Microtus agrestis be accepted, the sub-species exsu/ is then geographically 
the more primitive, and it is quite natural to find the newer forms 
exhibiting a simpler and less primitive dentition. It is also natural to 
find exsw/ more closely related to the true agrestis of Skandinavia 
than are either macgi/livrait or neglectus. None of these large forms 
are especially closely related to Azrtws of England. 
The habits are not known to differ from those of other Grass Mice. 
In Arran Mr R. W. Sheppard found the mice living mostly in scattered 
colonies on the edges of sheltered woods and young plantations of 
conifers on the north side of Brodick Bay; he met with no trace of 
them in the country around Blackwater foot. In Jura they frequented 
open pastures and had also runs through clumps of coarse grass at 
the edges of woods. In Mull they had scattered colonies on the cliffs 
above Tobermory on the western sides, at Croggan in open glades on 
hill-sides in natural woods, and on a hill at Buncosan, but were very 
scarce in all three localities. 
In North and South Uist they are extremely abundant all over the 
hills, and in the sandy meadows or machars; in some localities they 
make a network of runs in the grass. Mr N. B. Kinnear has supplied 
sketches (Fig. 68) of the sleeping-places as found in small mounds in 
North Uist. These blind, somewhat circular, terminals to the burrows 
were always of the same shape and had their sides smooth, as if much 
used ; they contained no grass or moss. 
They seem to perform a seasonal migration, being most abundant 
in autumn on the lower grounds and near farms, while in summer they 
are distributed all over the uninhabited districts. 
In North Uist Mr J. G. Millais has seen the collie dogs hunting them 
all day long, when not actually engaged in work, and a collie once 
stuffed himself so full of them that he was sick. 
They are hardy and do not hibernate [I have seen Skandinavian 
specimens of agrestis taken in a forest with 2 feet of snow on the ground]. 
The habits of the Norwegian form are given with much interesting 
detail by Robert Collett, but cannot be quoted here. It is evident that 
in northern countries of heavy snowfall, the habits are quite distinct 
in many respects from those prevalent in more southern districts. 
Nowhere, however, is there any hibernation, the mice being, like 
lemmings, active under the snow in winter. 
1 See p. 432 supra. Cocks (in Norway) had a setter very nearly die from bolting 
lemmings wholesale. 
