100 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 5 



figure 4. A similarly abrupt inversion is shown at the bottom of 

 figure 5 relating to the 21-month periodicity. The third curve 

 represents the 44 months ending with June 1864, and the lowest 

 curve represents the 44 months beginning July 1864. The inversion, 

 even in details, is marked, and the similarity to the corresponding 

 mean curves covering lli/^ years each in figure 4 is obvious. 



As another illustration, figure 6, I give the departures from the 

 normal of 110 years in the 12-month periodicity in the temperature 

 of Berlin. As the 12-month periodicity is primarily due to the 

 yearly revolution of the earth in its orbit around the sun, no one 

 would have anticipated that the departures from its normal course 

 would be governed by the 23-year cycle. Yet observe in figure 6 

 how the curves of 12-month periodic departures from normal tem- 

 perature at Berlin go in pairs, each pair covering 23 years beginning 

 1819, and how the phases change sharply at the conclusion of each 

 23-year interval. Finally in figure 7 I give a complete analysis of 

 the temperature departures at Cape Town. The phase changes are 

 very obvious. 



Having discovered that the periodicities of solar variation were 

 also found in weather, and that all were integrally related to 23 

 years, both as to period and as to change of form or phase, it 

 seemed to us appropriate to search for the 23-year cycle itself in 

 weather and in phenomena closely related to the weather. Figures 

 8 to 13 show the effects of this cycle in the level of the Nile, in the 

 levels of the Great Lakes, in the catches of mackerel and cod in the 

 Atlantic, in the width of tree rings, and even in the thickness of 

 varves or layers formed by the yearly settling of sediment in glacial 

 lakes in Pleistocene geologic age. 



But perhaps most interesting of all is the 23-year cycle in ordinary 

 weather. Figure 14 shows a plot of the smoothed monthly mean 

 percentage precipitation at Peoria, 111., and figure 15 a similar plot 

 of monthly mean percentage precipitation at Nagpur in central 

 India. In figure 14 various features have been marked with letters 

 to show their approximate repetition at successive intervals of 23 

 years. Besides these small details the reader may trace some prin- 

 cipal trends of the successive curves which show much similarity 23 

 years apart. In figure 15 note that in 1865, 1868, and 1870 there are 

 three pillarlike features of high percentage precipitation bounding 

 two features of subnormal precipitation. Thus there stand out 

 two intervals of 3 and 2 years, respectively, as if guarded by these 

 sentinal features, but embracing besides nearly a score of subordinate 

 features. The reader's attention is now invited to similar features, 

 1888-93, and 1912-17, in which nearly all the details are recognizable. 



