224 Clayton — Six and Seven Day Weather Periods. 



lived long, that at a remembered period some particular day 

 in the week, for several weeks, was stormy ; sometimes with, 

 and at other times without an intervening storm, or showery 

 condition, on some day in the week." 



In Yennor's Weather Almanac for 1883, p. 13, it is stated : 

 "We find closely corresponding weather periods have fre- 

 quently occurred in seven, fourteen, and twenty-one year 

 divisions of time, and most of us are familiar with the every 

 seventh storm day of our winter and summer months. Only 

 as recently as last autumn (1881) the general remark was that 

 every Saturday stormed ; and it will be of further interest to 

 the public to learn that these stormy Saturdays lasted through 

 a period of just about seven weeks." 



The fact that this period was independently discovered by 

 the present writer, without any knowledge of its previous 

 observation, may perhaps be taken as additional proof of its 

 existence. 



The widespread and tenacious belief in a connection between 

 the changes of the moon and the changes in the weather 

 probably arose from the popular observation of the seven day 

 weather cycle which agrees so approximately in length with 

 the mean interval between the phases of the moon. 



In the American Meteorological Journal for August, 1885, 

 p. 162, the writer gave a marked example of this seven day 

 period in which the temperature and weather at several widely 

 separated stations in the United States oscillated with almost 

 perfect regularity for six or seven weeks in succession. More- 

 over, the oscillations were not simultaneous at all the stations, 

 but due to barometric minima which moved from west to east 

 at the normal rate of American storms. It was stated by the 

 writer that he had followed this seven day period for several 

 years, and found that at intervals it rose to considerable promi- 

 nence, then decreased in intensity and apparently died out. 

 He also stated that " there seems to be an oscillation of about 

 three days which acts in a similar manner." Later this was 

 called a six day period on account of its similarity to the seven 

 day period, as will be seen from what follows. 



In 1887 the trustees of the Elizabeth Thompson fund 

 granted the writer a part of the fund for further research on 

 this subject ; but he was unable to undertake it at that time. 

 In January, 1893, however, his attention was again called to 

 this subject by noticing a marked regularity in the oscillations 

 of the temperature at Blue Hill Observatory ; and in the 

 American Meteorological Journal for May, 1893, he showed 

 that all the temperature maxima at the Observatory during 

 1891 and 1892 could be arranged so as to follow each other at 

 approximate intervals of six or seven days, or half these inter- 

 vals, and the regularity of the sequences was much greater 



