56 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 94 



multiples of ii^ or 23 years from January 1819. It is therefore hard 

 to attribute these similarities of behavior to causes not extra-terrestrial. 



17. Summary of Preceding Studies and Their Guidance 

 Toward Those Which Follow 



It has been shown that the sun is variable. Its variations comprise 

 numerous periodicities. These periodicities are so definite as to justify 

 synthetic forecasts of solar variation. Apparently, all the periodicities 

 in solar variation are integral subdivisions of 23 years. 



With this background it seemed reasonable to attempt to trace the 

 effects of solar periodicities in weather. Analyses have been presented 

 of temperatures and precipitation at several stations widely separated. 

 The 23-year period is thereby found to exercise a dominating influence 

 in weather. Numerous periodicities which are integral submultiples 

 of 23 years seem to exist in weather. Nevertheless, changes of phase 

 and amplitude complicate these relations. But it has been shown that 

 these changes of phase and amplitude are apt to occur abruptly at 

 times which are integral multiples of 11^ or of 23 years after Janu- 

 ary I, 1819. 



These studies lead us to expect that many of the features in weather 

 which occur apparently unordered are really produced by the summa- 

 tion of periodic changes integrally related to 23 years. Hence they 

 will be apt to be found, though doubtless with considerable modifica- 

 tion, in successive 23-year cycles. There is ground to expect that 

 the similarity of such features will be greater after 46 than after 23 

 years. As these periodic changes seem to be of solar origin they should 

 be observable throughout the world. 



We may also expect that phenomena which depend intimately on 

 the sunshine or the weather, such as the growth of vegetation, the 

 numbers of creatures that feed on vegetation, the flow of rivers, the 

 level of lakes, the thickness of varves, whether produced by the flow 

 of glacial rivers or by the summer dessication of lakes, all such phe- 

 nomena may display the influence of the 23-year cycle. In the re- 

 mainder of this paper it will be shown in how far it has been found 

 that these expectations are realized. 



18. A Test of the 23-YEAR Hypothesis in the Precipitation 

 OF Southern New England 



In 1934, C. M. Saville" published a table of annual precipitation 

 over southern New England given as percentages of base values from 

 1750 to 1932. The values depend on reports from i to 10 stations. 



^ Quart. Journ. Roy. Meteor. Soc, vol. 60, p. 324, 1934. 



