102 THE UNIVERSITY SCIENCE BULLETIN. 



cycles. In investigating individual cycles it is necessary to 

 eliminate the seasonal effect from each individual month. This 

 has, therefore, been done for each month and each state by 

 dividing the actual rainfall of each state for each month by 

 the normal of that state and month. As stated in the first 

 paper, this method is as reliable as the former one, except on 

 the extreme western coast of the country where normals are 

 practically zero for certain months, and where these zero 

 months are thus given an equal weight with months of heavy 

 normal rainfall. The results for these twenty-four consecutive 

 cycles are tabulated as table 1. The attention of the reader is 

 called to the fact that in twenty-two cycles there are only two 

 in which the percentage of rainfall, for months when the 

 cycle calls for a minimum, has actually been above normal. 

 Each of these cycles is strictly independent of any other and 

 their lengths are dependent only upon extra-terrestial causes. 

 For the maximum phase it is to be noted that sixteen are 

 above normal, seven below and one exactly normal. The 

 author believes that this table establishes the probability 

 much more strongly than the previous treatment, so strongly 

 in fact that only very strong definite negative evidence can 

 combat it. 



California, western Washington and western Oregon are, 

 as shown in the preceding paragraph, not available for treat- 

 ment by individual cycles unless the summer months are en- 

 tirely disregarded. It has been felt best, therefore, to treat, 

 instead of the whole Pacific group of the first paper, the states 

 of eastern Washington and Oregon, Idaho, Montana, Utah and 

 Nevada as a unit. For these states there are available eighteen 

 consecutives cycles. The results are shown as table 2. For 

 the minimum phase fifteen of the eighteen are found to be 

 below normal and for the maximum phase thirteen out of the 

 eighteen are above normal. 



As shown in the first paper, it is impossible to continue the 

 varying period beyond the last date which is followed by both 

 a sun-spot maximum and sun-spot minimum. This is 1913. 

 The tables previously referred to are based on Wolfer's esti- 

 mate of May, 1913, as epoch of minimum. This has been re- 

 vised by him, placing the minimum nearly three months later.^ 



1. Prof. A. Wolfer. Monthhj Weather Review. July, 1915, p. 314; August, 1920, 

 pp. 459-461. 



