54 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 94 



and ninth, and the tenth with the eleventh correspond, respectively, 

 closely to V-VI, VII-VIII, and IX-X of figure 15 in time intervals. 

 Hence, results were to be expected in these cases nearly parallel to 

 those previously obtained. 



But as to the other six sets of curves in figure 22, there is hardly a 

 vestige of indication supporting the periods chosen, excepting for the 

 last three curves among the 29-month group. The similarity of these 

 three curves is indeed curious. In the lo-month group there is the 

 nearest semblance to continued periodicity. Here it may be that a 

 case could be argued for a periodicity of 10 months plus 3 days. But 

 this would be 1/27 of 273 months, and would but add one more to 

 the group of nearly integral submultiples of 23 years, already discussed. 



15. Other Analyses 



Besides the temperature of Berlin, both temperature and generally 

 ])recipitation also have been analyzed with equal thoroughness at 

 Helsingfors, Copenhagen, Greenwich, Cape Town, and Adelaide. The 

 results were very similar to those already discussed for Berlin. The 

 dominating importance of the 23-year period displays itself quite as 

 conspicuously in these other analyses as in the case of Berlin tempera- 

 tures. That is to say, all the periodicities which seemed to be indicated 

 were nearly integral submultiples of 23 years. Also if the date Janu- 

 ary I, 1819, was selected as a point of departure, changes of phase and 

 amplitude occurred abruptly at multiples of 11^ or of 23 years 

 thereafter. 



As it is felt that the united evidence from these widely separated 

 stations is of great importance, excerpts from the results from various 

 stations are given in graphic form in figure 19. In addition, the com- 

 plete analysis of the temperature of Cape Town is shown in figure 23. 



16. Conclusions Derived from Analyses of Berlin and Other 

 Temperatures and Precipitations 



a. It is shown that 14 apparent periodicities may be found in the 

 smoothed temperature departures of Berlin and other stations since 

 1819. 



b. Summing these periodicities and subtracting their sum from the 

 original smoothed departures, the residual departures at Berlin have 

 an average range of two-thirds of the originals. Similar results occur 

 in the other analyses. 



c. Thirteen of the supposed periods are primarily attributed to solar 

 changes, and are approximately aliquot parts of 23 years, being, re- 



