608 PROCEEDINGS OP THE AMERICAN ACADEMY. 



suggested for this phenomenon was that the moon is negatively electrified 

 like the earth, and, acting by induction on the earth's surface, diminishes 

 the electrical potential on the portion of the earth nearest the moon and 

 increases the potential on the opposite side of tlie earth. In 1896-97, 

 from observations made throughout the United States, I investigated the 

 relation of the frequency of auroras to the moon's position in declination, 

 and found the greatest number of auroras when the moon was in south 

 declination. (Paper i-ead before the Boston Sci. Soc, May 11, 1897. 

 See Boston Evening Herald of May 12, 1897, and Amer. Jour, of Science 

 of February, 1898.) 



In 1898, Dr. Ekholm and Mr. Arrhenius published an exhaustive and 

 most careful research on the relation of auroral frequency to the position 

 of the moon in declination, in which the various sources of error which 

 might influence the results were considered and eliminated. (Ueber den 

 Einfluss des Mondes auf die Polarlichter uiid Gewitter. von Nils Ekholm 

 and Svante Arrhenius, Bihang der k. sclnvedischen Akad. d. Wissen- 

 schaften, 1898.) They made use of all the available observations both 

 in the northern and southern hemispheres. Their results from all the 

 observations in each hen)isphere are plotted in the curves marked auroras 

 in Plate IV. The curves are nearly opposite in phase, and clearly 

 follow the same course as that of atmospheric electricity shown in the 

 upper part of the same plate. Ekholm and Arrhenius also determined 

 the frequency of thunderstorms in Sweden as related to the moon's posi- 

 tion in declination, and their results are plotted and marked thunder- 

 storms in Plate IV. The frequency follows the same coui-se as that of 

 the auroras and of the amount of atmospheric electricity except that there 

 is a secondary maximum about the time of the moon's gi-eatest northern 

 declination, a phenomenon which I found also for auroras in the United 

 States. From the observations in the lj"ited States published in the 

 Monthly Weather Review, Mr. A. E. Sweetland worked out for me the 

 frequency of thunderstorms as related to the moon's position in declina- 

 tion and obtained the results plotted in the broken curve marked United 

 States. The range is very small but the curve follows a similar course 

 to the one for Sweden showing a maximum a few days preceding the 

 greatest northern declination. 



My work in meteorology was instigated by the popular belief in the 

 influence of the moon on the weather. The general prevalence of this 

 belief led me when a youth to investigate whether the weather conditions 

 supposed to follow certain positions of the moon really did so. I found 

 the subject a complex one, and, taking up the study of modern mete- 



