VARIATIONS IN THE NUMBERS OF ANIMALS 133 



varying hare {Lepus americanus). The lemming has for 

 hundreds of years attracted attention by its periodic appearance 

 in vast numbers. A certain Ziegler, who wrote a treatise 

 which was published in Strasburg in the year 1532, was the 

 first to make any record on the subject. He said that when he 

 was in Rome in 1522 he heard from two bishops of Nideros a 

 story about a small animal called the *' leem " or '' lemmer," 

 which fell down from the sky in tremendous numbers during 

 showers of rain, whose bite was poisonous, and which died in 

 thousands when the grass sprouted in the spring. The story 

 about the downfall from the skies was " confirmed '* by 

 Claussen in 1599, who brought forward new evidence of 

 eye-witnesses who were *' reliable men of great probity." It 

 was not until the lemming had been described fairly accurately 

 by Olaus Wormius, in 1653, and by later writers, that the real 

 truth became known. 33 The Norwegian lemming lives nor- 

 mally on the mountains of Southern Norway and Sweden, 

 and on the arctic tundras at sea-level farther north. Every 

 few years it migrates down into the lowland in immense 

 numbers. The lemmings march chiefly at night, and may 

 traverse more than a hundred miles of country before reaching 

 the sea, into which they plunge unhesitatingly, and continue 

 to swim on until they die. Even then they float, so that their 

 dead bodies form drifts on the seashore. This migration, a 

 very remarkable performance for an animal the size of a small 

 rat — MacClure called it " a diamond edition of the guinea- 

 pig " — is caused primarily by over-population in their mountain 

 home, and the migrations are a symptom of the maximum in 

 numbers which is always terminated by a severe epidemic ; 

 and this reduces the population to a very few individuals. 

 After such a " lemming year " the mountains are almost 

 empty of lemmings. 



10. Owing to the striking nature of these lemming maxima 

 we possess records of nearly every maximum for a number of 

 years back, which enable us to find out the exact periodicity 

 of the pulsations in lemming numbers.^^' ^* In recent years 

 the maxima have occurred every four years, while in the middle 

 of the nineteenth century they sometimes occurred at rather 



