CAUSES OF ANIMAL CYCLES 193 



The last fact may serve to explain in part the differences in duration, 

 as well as a variation of a year in either direction, in the Swedish 

 cycle. As this indicates, cycles are not necessarily synchronous, even 

 in adjacent regions, but this is likewise true of the climatic effects 

 related to sunspot numbers. 



As the only material in which the actual numbers are known, as 

 opposed to guesses or estimates, the fur returns of the Hudson's Bay 

 Company furnish the best, though imperfect, scientific basis so far 

 available for the study of cycles in mammal populations. These 

 records were first made known by MacFarlane (1905), who had been 

 chief factor of the Company, and they have since been utilized by 

 Seton (1911, 1929), Hewitt (1921), and Elton (1924). MacFarlane 

 recognized that the lynx passed through a cycle of increase and de- 

 crease about every decade, and that the numbers of this and the 

 marten in particular fluctuated with abundance and scarcity in the 

 rabbit. His statement that the Company did not trade rabbit skins 

 in the interior, but only at the posts situated on the shores of Hudson 

 Bay, helps to explain some of the discrepancies between the curves for 

 these skins and for their predators, which were bought throughout the 

 vast extent of the North. Seton (1911) has given graphic expression 

 to the numbers for the various species and finds a close agreement 

 between the basic curve of the rabbit and those of lynx, marten, 

 skunk, fox, and mink; the basic curve for the wolf is discordant, 

 while the bear, badger, and wolverine fluctuate but little. Hewitt's 

 conclusions are essentially in accord, the cyclic animals yielding a 

 period close to 9.5 years, with the exception of 4 years for the arctic 

 fox. The basic cycle of the rabbit is given as 8.5, but it is actually 

 10 years if the subcycle of 1854-57 is not included. 



Elton (1924:136) points out that the take of skins for any year 

 includes some from the year before and hence it is not possible to de- 

 termine the actual date of the maximum. INIoreover, there may be a 

 variation of a few years between different regions. He also emphasizes 

 the fact that there is nothing in physiology to support the view of a 

 mysterious and obscure rhythm of reproduction. He finds "that there 

 is a rabbit maximum just before or on, or just after each sunspot 

 minimum, except in 1905, when there was a small maximum near the 

 sunspot maximum." However, when the years of maxima for the 

 rabbit and lynx are compared with the year of each minimum, as in 

 Table 5, the agreement is seen to be less satisfactory. Thus, 4 of the 

 8 rabbit maxima are 2 or 4 years away from the year of sunspot 

 minimum, while 5 of the 9 lynx maxima depart as much as 3-5 years. 

 Nevertheless, the table of intervals brings out a clear cycle of num- 



