60 KANSAS Academy of science. 



more than .88° above or below the normal). Of these four bad failures, three are in 

 predicting a hot summer, which was cool. There has been only one summer in the 

 23 years in which one would have predicted a cool summer which proved to be a hot 

 one. 



SEVEN-YEAR PERIODICITY l5f RAINFALL. 



BY E. C. MUEPHY, LAWBENCE. 



In my paper answering the question, "Is the rainfall in Kansas increasing'::'" I 

 have divided the rainfall record of Fort Leavenworth and Manhattan into seven- 

 year wet and seven-year dry periods. It is the object of this investigation to deter- 

 mine to what extent the rainfall records of other places in the United States and 

 Canada may be thus divided. 



For this purpose, I have used the 11 longest and most nearly continuous rainfall 

 records in United States and Canada. These records each cover a period of from 

 40 to 76 years, and from 6 to 10 seven-year periods. 



The records down to 1875 are taken from vol. 24 of the "Smithsonian Contribu- 

 tions to Knowledge." The records from 187.5 to 1890 are signal-service records, sent 

 me by the chief signal officer. 



The method of investigation is: (1) Plat the record or curve of annual rainfall; 

 (2) study the curve, and from its appearance divide it into seven-year periods; (3) 

 find the mean of each period. In nearly every case, it is easy to see from the curve 

 how to divide the record, or where to begin the division of the record, so that when 

 the means of the periods are found they will be, approximately, alternately wet and 

 dry. 



Each curve has been studied by itself, without any reference to other curves. I 

 have tried, in each case, to find the most natural division into seven-year periods, 

 without regard to whether the periods at the different places begin with the same 

 year. The accompanying chart shows these curves platted to scale in the usual way, 

 and the division into periods, and the means of the periods. 



Table No. 1 shows the years in the periods, the means of the periods, and whether 

 wet or dry. 



