198 AGGREGATION, COMPETITION AND CYCLES 



past decade or so, especially in the matter of phases (Uvarov, 1928, 

 1931), the paramount problem of numbers and migration in terms of 

 causes and processes still demands much more experimental and quan- 

 titative investigation. For this reason, as well as that of brevity, the 

 following discussion is limited to a consideration of cycles in locust or 

 grasshopper populations and their relation to solar phenomena. 



The observation that locust outbreaks often coincided with or 

 immediately followed drought, occurs here and there in the older 

 chronicles, and was perhaps most definitely expressed by Purchas 

 (1657; Thomas, 1880), who said that "great droughts produce them, 

 at least cause a prodigious increase of them; in 1553, after five years' 

 drought, there were great armies of them." This relation was empha- 

 sized by Koeppen (1870), and the view has received the support 

 among others of Riley, Packard and Thomas (1880), Filipjev (1928), 

 Parker (1930), and Uvarov (1931). Koeppen also appears to have 

 been the first to consider the question of a connection between the 

 frequency of outbreaks and the sunspot cycle. He decided that "the 

 greatest number of locust years falls in the first five years after the 

 minimum of sunspots, particularly in the third and fourth year (16 

 times) ; in the fifth and sixth year (i.e. during the probable maximum 

 of sunspots), there were 12 and 13 years of the appearance of locusts." 

 For years 7, 8, 9, and 10, the respective numbers are 10, 12, 5, and 7. 

 Since the correlation between the minimum and increased temperature 

 characterizes the year of the minimum, it is difficult to explain the 

 occurrence of the greatest number of outbreaks the third and fourth 

 year afterward. However, the major criticism of Koeppen's table 

 is that it is impossible to know the exact course of the sunspot cycle 

 before definite records began in 1749, and the dates from 592 to this 

 year are without adequate foundation. 



Swinton (1883) returned to the task of correlation a decade later, 

 but the careful scrutiny of his figures affords little warrant for his 

 conclusion that "Properly generalized observations show almost invari- 

 ably an exact concordance between the sim changes and these effects." 

 A few years earlier, as a result of the comprehensive studies of the 

 U. S. Entomological Commission during the "grasshopper years" of 

 the seventies, Thomas expressed the opinion that "locust migrations" 

 are not governed by any law of regular periodicity, in spite of the fact 

 that the average interval between 173 invasions in China and 30 in 

 Germany was a little over 11 years. However, he thought it but fair 

 to state that for the "noted locust years in our own western country, 

 to wit, 1820, 1855, 1866, and 1874-76, the interim in each case is very 

 nearly a multiple of 11 years." It is worth noting that all these pre- 



