32 THE UNIVERSITY SCIENCE BULLETIN. 



phases. If this periodicity is generally accepted, the recent sun- 

 spot epochs can be revised to give the best rainfall results, since the 

 short period and the great amount of data will locate them more 

 accurately than the sun-spot counts themselves. 



(c) The curve probably actually undergoes changes, similar in 

 shape and magnitude to those of the sun spots, one maximum of 

 which will be several times higher than another. This is indicated 

 directly by the persistency with which a phase for quite a number 

 of consecutive cycles will often differ from its mean by fairly large 

 amounts. 



(d) If the rainfall is not a pure continental or pure marine type, 

 we will have one type often prevailing, although in the long run the 

 other dominates. 



Although I have examined this period as though it varied in 

 length, I do not desire to stand in the least committed to an actual 

 variation. This period, the eleven-year period and the Bruckner 

 are all harmonics. When examined by itself each is found to be 

 variable. However, it is quite possible that their variations and 

 that of the sun-spot period are only apparent, being caused by the 

 superposition of a number of constant periodicities. Regardless of 

 this constancy, I believe these three periods not to be separate, but 

 merely terms in an irregular, long-period rainfall variation. It is 

 very important that a search be made very carefully to determine 

 what other terms there may be of such large magnitude as these. 



If the relationship between sun spots and rainfall were a direct 

 one, the eleven-year period would certainly far overshadow both 

 this and the Bruckner. Instead, its magnitude seems usually to be 

 less than either. The search for a thirty-three-year period in sun 

 spots has been inconclusive, although analysis shows a very strong 

 sun-spot variation of twice this length. The relationship of the 

 Bruckner cycle to the sun-spot period stands out vividly, however, 

 if we look for its epochs in long, homogeneous records from which 

 the eleven-year period has been eliminated by averaging between 

 Consecutive sun-spot maxima or minima. In concluding, I desire to 

 quote from Pickering's statement, at the close of his article men- 

 tioned above, as most nearly expressing my own opinion on this 

 relationship : 



"I do not believe that the sun spots themselves, or their absence, cause the 

 droughts. The spots are merely a surface indication of an overturn of ma- 

 terial and temperature occurring beneath the solar surface in connection with 

 magnetic storms. ... I have only to derive statistics from observed rain- 

 fall data to show the coincidence." 



I wish to acknowledge the assistance of the research committee 



