CAUSES OF ANIMAL CYCLES 195 



received for 1934-35 and most of these have been mapped. They 

 cover most of Canada (including Anticosti) and sample areas of 

 Alaska and the United States. In the Northwest, rabbits were still 

 increasing and abundant. Over much of the rest of Canada a steep 

 decline had set in, partly associated with epidemics. Rabbits in 

 Northeast United States had also declined in numbers. Those in the 

 Rocky Mountain areas were stable or showed no consistent regional 

 trend, except for an increase in Montana. The figures for previous 

 years, summarized with trends in furbearer catches in Canada, show 

 a well-marked cycle of about ten years. 



"The cycle of about ten years in lynx and red fox is clearly shown, 

 also the lack of any cycle in the marten, which has gone steadily 

 down to about a third of the catch for 1919-20. The marten used 

 to be one of the furbearers showing the most conspicuous ten-year 

 cycle. After about 1900 the cycle gradually broke down and this is 

 probably connected with overtrapping. There is a marked parallelism 

 between lynx, red fox, and snowshoe rabbit, and it is apparent that 

 the snowshoe rabbit inquiry should provide a firm basis for forecast- 

 ing the general trends of some of the furbearing animals. The peak 

 in lynx and red fox ought to occur in the winter of 1935-36 or 1936-37. 

 Of course, it is realized that we are not comparing exactly homogene- 

 ous data, since the richest areas for furbearing animals are not neces- 

 sarily, at the present day, the most important snowshoe rabbit dis- 

 tricts, and there are important regional lags in rise and decline." 



Elton (1933) has recently given an informative account of the 

 unique conference on animal cycles sponsored in 1931 at Matamek, 

 Canadian Labrador, by Mr. Copley Amory, as has Huntington like- 

 wise (1932). The minutes of the conference have been collated and 

 unfortunately not published in full. (See Huntington, 1932.) 



In Britain, IMiddleton (1930) has indicated a cycle of 4 years, 

 with a range from 3 to 5 years, in two species of vole, which resembles 

 the cycle in the related lemming of northern Europe. He also finds 

 marked fluctuations in the number of rabbits and hares as compiled 

 from old game records on estates (1934), but these seem to bear little 

 if any relation to the sunspot cycle or its extremes. 



MacLulich has recently published the results of a comprehensive 

 and detailed investigation of numbers in the varying hare (1937), 

 which support some of the conclusions reached in the above account. 

 He states that the year of fur production is the second year before the 

 year of sale, a fact that results in a corresponding displacement of 

 agreements previously noted between sunspot minima and peaks of 

 abundance. For the whole period of lynx returns from 1751 to 1925, 



