NO. 13 SHORT PERIODIC SOLAR VARIATIONS ABBOT 3 



then noticed that maxima and minima in the seven-column means 

 occurred progressively later as the years went on. 



CORRECTION OF THE PERIOD 



By graphic methods applied to both maxima and minima, and to 

 several different months of tabulation, it appeared that the secular 

 displacement of features just referred to amounted to 3 days in 18 



Table i. — Nezu York City temperature departures. May 1929, arranged in the 

 6.6436-day period 



years. As there are 55 recurrences of the period each year, this dis- 



placement corresponds to -~ =0.0030 days per cycle. It is more 



X 55 

 convenient to use numbers ending in 5 than in 6, and the accuracy 



of the determination does not justify the inconvenience, so, instead 



of 0.0030, a correction of 0.0029 was applied, making the corrected 



period 6.6485 days. 



By subtracting and adding this number of days many times to 



January 17.0000, 1946, a tabulation was prepared giving dates of 



all expected minima of temperatures at Washington from 1928 



to 1948. Applying this table also at New York it would be found 



whether minima, occurring there by the effect of the periodic solar 



variation, fall on the same dates as at Washington. 



PREPARATION OF THE DATA 



Using Weather Bureau forms 1030, the departures from normal 

 temperature were tabulated, exactly as in table i, for both Wash- 

 ington and New York, from January 1928 to December 1948. • In 

 these tabulations the newly corrected period 6.6485 days is used. As 

 stated above, the tabulation is based on January 17.0000, 1946. 



As the reader will note in table i, this arrangement results in tables 

 of seven columns, and these were labeled i to 7. Each month — 



