TWENTY-FOURTH ANNUAL MEETING. 



19 



three places, and also for Fort Scott. It shows also the division into seven-year dry 

 and seven-year wet periods. 



By a dry period we do not mean that each year of it is a dry year. The rainfall 

 of some years, in a dry period, may be greater than that in some years of a wet pe- 

 riod. One period is dry compared with another when the mean rainfall of the first 

 period is less than that of the second. 



Beginning with the Fort Leavenworth line, we see from it that from 1836 to 1842 

 there is a seven-year dry period, the mean rainfall for this period being 30.43 inches. 

 From 1843 to 1849 there is a wet period, the mean rainfall for this period being 32.03 

 inches. Table No. 2 gives the results in tabular form. 



TABLE No. 2. 



Examining the means in column three, table No. 2, we see that those of the fifth, 

 sixth and seventh periods are larger than those preceding them, indicating a large 

 increase in the rainfall. We will subtract from the mean of the fifth jseriod 1 inch, 

 from the mean of the sixth 9 inches, and from the mean of the seventh 4] inches, in 

 order partly to eliminate this large increase. Column four gives the means when 

 thus reduced. Examining the means in this column, we see that they are alternately 

 •wet and dry, and that each dry mean is wetter than the preceding dry mean, and 

 each wet mean is wetter than the preceding wet mean. Hence, after taking away 

 100 inches of rainfall from this record, there is left a gradual increase during the 

 whole 56 years covered by the record. 



Consider next the Manhattan line: The record of rainfall at this place begins in 

 1854. This is not the beginning of a period; the first full period of this record (the 

 fourth period of the Fort Leavenworth record) begins in 1857. Column six, table 

 No. 2, gives the means for this place. We subtract li inches from the sixth mean? 

 and 64 inches from the seventh mean, giving the means in column seven. Examin- 

 ing the means in this column, we find that they are alternately wet and dry, and each 

 dry mean is wetter than the preceding dry mean, and each wet mean is wetter than, 

 the preceding wet mean. Hence, after taking away 54J^ inches of rainfall from this 

 record, there is left a gradual increase during the whole 35 years covered by the 

 record. 



Examining the three Lawrence means, in column nine, table No. 2, we see that 

 the eighth is larger than the seventh, the seventh larger than the sixth. Hence, there 

 has been a gradual increase in rainfall at this place dulring the 21 years covered by; 

 the record. 



It should be remembered that the writer does not claim that the rainfall has al- 

 ways been increasing in Kansas, or that it will continue to increase in the future. 

 His conclusions apply only to these three places for the time covered by the record.. 



