THE SOLAR HYPOTHESIS. 



239 



Table 15. 



1. The year of maximum spots. 



2. The year after maximum. 



3. Two years after maximum. 



4. An intermediate period of decreasing 



spots having an average length of 

 about 1.5 years. 



5. Two years before minimum. 



6. One year before minimum. 



7. The year of minimum spots. 



8. One year after minimum. 



9. An intermediate period of increasing spots 



having an average lengtli of about 1.5 j-ears. 



10. Two years before maximum. 



11. One year before maximum. 



In order to have two separate sets of data for comparison, the period since 1610 is 

 divided into two parts, 1610-1754, and 1755-1900, each consisting of 13 complete sun-spot 

 cycles. The two parts differ in respect to the length of time from maximum to minimum. 

 In the earlier period the average lapse of time from maximum to minimum was approxi- 

 mately the same as from minimum back to maximum, or about 5.5 years in each case. 



Siinspots 

 (unsinouthed) 



Fig. 75. — The Relation of Sun-.spots and Tree Growth in the 11-year Cycle. 



In the later period the time from maximum to minimum was not far from 6.5 years, while 

 that from minimum back to maximum was 4.5. Hence in the first period it was necessary 

 to omit No. 3 in table 15, and in the second period to omit No. 10, thus in each 

 case leaving the sun-spot cycle divided into ten parts, eight of which have a length of a 

 year, and two of 1.5 j'ears. For each of these ten parts I have computed the average 

 growth of eleven sequoia trees, the only ones whose yearly growth it has yet been possible 

 to measure back to 1610 a. d. with sufficient accuracy. The results are shown in figure 76. 



In order to test the matter in a somewhat similar way in Europe I have computed the 

 average price of wheat in England according to the tables of Thorold Rogers, using the 

 same ten divisions of each sun-spot cycle. The use of these tables was suggested by 

 A. H. Swinton, Esq., of Totnes, England, who has devoted much time to a careful study 

 of the relation of sun-spots and weather. During the peiicd ficm 1610 to 1754 England 

 produced her own food supply, economic conditions did not change greatly, and there were 

 no protracted or highly devastating wars. Accordingly the main factor in determining 

 variations in the price of wheat was the amount actually raised in the country itself; and 

 that, of course, was dependent chiefly upon variations in the weather from year to year, 

 dry years being in general favorable so that prices were low, and wet years unfavorable so 

 that prices were high. During the later period, 1755 to 1900, the Napoleonic wars, the 

 growth of manufactures, the importation of food from America, and changes in the tariff 

 have completely altered the conditions of British agriculture, so that the price of wheat no 

 longer depends upon the amount raised locally. Hence it is not possible to use this later 

 period. 



Thus in figure 76 we have three curves, H and I on the left for the period from 1610 

 to 1754, and I' on the right for 1754-1900. Above these the sun-spot curves, G and G', 

 have been placed, G being merely an estimated curve, since exact data are not available, 

 while G' is the mean of the actual observations as given by Wolf. In using data derived 

 from sources such as the growth of trees and the price of wheat it is obvious that there is a 

 large opportunity for error, partly from mistakes in actual observation, but chiefly because 



