ME. E. COLLETT ON MTODES LEMMU8 IN NOBWAT. 329 



and there, and is shown in the sporadic appearance of individuals 

 down into the nearest valle}- s ; and between these smaller and 

 the true migrations every gradation is to be found. 



This is a point that ought not be overlooked. 



In some years the reproduction is still stronger, and the sudden 

 collection of individuals of an animal that under ordinary circum- 

 stances seems to be nowhere very numerously distributed, but 

 always keep themselves scattered on the larger plateaux, will 

 naturally cause a movement of the masses towards the sides, so 

 as again to make the balance even ; and by the natural desire to 

 wander possessed by this species (which is also to a certain 

 extent shown by the allied species of Lemming, M. schisticolor, 

 Lilljeb., and also in some of the other species of Arvicolinse), 

 these migrations very soon reach the borders of the plateaux, 

 and subsequently spread over an area that is considerably larger 

 than obtains in any other of the species under similar circum- 

 stances. 



In cases where in two succeeding years the production of young 

 has been excessive, the masses are incessantly pushed towards the 

 sides of the fells ; and the migration becomes an overrunning of 

 the lower and far remote portions of the country, as the indivi- 

 duals gradually penetrate further in search of localities suitable 

 to their habits (and which are capable of giving them a perma- 

 nent subsistence), until they are stopped by the sea or destroyed 

 in some other manner. 



In the years of the great migrations the first families reared in 

 the spring reproduce in the autumn of the same year, causing an 

 over-population. This drives the individuals to Avander down to 

 the plains, often before they are full-grown. The next year the 

 masses may increase in number to such an extent, that myriads 

 continually move from the central plateaux and overrun the low- 

 lands, where they join those that have previously arrived. Many 

 pairs breed during the journey ; they are therefore more numerous 

 in the late summer, become less so in the autumn, die in immense 

 masses in the winter, are seen, though sparingly, the next 

 spring, and disappear gradually during the summer (the second 

 year of their wandering). This stage of increasing and disappear- 

 ing is the rule in many instances ; but, as previously said, the 

 migrations may be shorter as well as longer, and then not be 

 finished before the third year from their starting from the 

 fells. 



