CAUSES OF ANIMAL CYCLES 199 



ceded the corresponding sunspot minimum by 1 to 4 years, contrary to 

 Koeppen's assumption. 



Uvarov (1928) reports the attempt of Krasilshchik (1893) to evolve 

 a law of locust periodicity in accordance with which Locusta migra- 

 toria should appear at the mouth of the Danube every 12 years and 

 in the Caucasus every 5-6 years, but dismisses this on the basis of the 

 unreliability of many old records. He also discusses briefly (1931. 'ISSj 

 the views of Simroth (1908, 1909), his supporters and critics, and 

 concludes that it is hardly to be disputed that the fluctuations of cer- 

 tain insects coincide to some extent with the 11-year cycle. As an 

 example, he cites the widespread appearance of the desert locust from 

 Morocco to India in 1915-16 and again in 1927. However, these were 

 respectively 2^ years after the minimum and 1-2 years before the 

 maximum, so that their cyclic relations are confused. 



The most striking and definite instance of a relation between the 

 sunspot maximum, drought, and grasshopper outbreaks has been fur- 

 nislied by the studies of Parker (1930). The sunspot maxima of 1870 

 and 1917 were associated with 4 years of drought in the Montana 

 region studied (cf. Clements, 1921, a), and were immediately followed 

 by 3-year outbreaks of grasshoppers. However, it is difficult to under- 

 stand why the equally severe drought of 1893-95 produced no like 

 effect, unless the local region was exempt from its action. The inves- 

 tigations of Criddle (1932) deal with fluctuations in Manitoba, and 

 lead to the conclusion that there is a correlation between sunspot 

 minima and high numbers of grasshoppers, though this is often ob- 

 scured by other factors. This qualification is reinforced by the fact 

 that, out of 11 single or initial years of grasshopper abundance, 6 fall 

 nearer the maximum than the minimum, 2 being the maximum year 

 itself. 



Cycles among Fish. Phelps and Belding (1931, 1933) have in- 

 vestigated the salmon catch in the Restigouche and Grand Cascapedia 

 rivers in Canada over a period of 50 years and recognize a conspicu- 

 ous cycle of about 10 years. The dates of the maximum period of 

 catch were about 1885, 1895, 1905, 1915, and 1925. It is interesting 

 to note that the first two are 2 years after the respective maxima, the 

 third is a maximum year, and the last two are 2 years after the mini- 

 mum (cf. also Wing, 1935). The cod and other species are also known 

 to exhibit remarkable fluctuations from year to year, but it is doubt- 

 ful whether these show intervals that may be related to the sunspot 

 cycle (Kyle, 1933). 



