Chase.l ■^If' [Jan. 16, 



cloudy, and on 23 days no special evidence of stormy disturbance was re- 

 corded. This represents 81 per cent, of verification by storm or stormy 

 tendency, and 19 per cent, of failure. 



V. Of the 120 days on which the lunar tidal tendencies were more fair 

 than stormy, 74 were fair; on 21 other days there were winds from fair 

 directions, and on 25 days there was rain or snow with no special record of 

 fair influence. This represents 62 per cent, of complete verification, 17 per 

 cent, of partial verification, and 21 per cent, of failure. 



^. Making allowance for one day's possible shifting of fair and stormy 

 tendencies, by the acceleration of equatorial cyclonic currents, or the retar- 

 dation of polar anticyclonic currents, 85 per cent, of the fair, 71 per cent, of 

 the stormy, and 79 per cent, of all the indications were completely verified. 

 In this test no stormy verification was admitted in which there was not an 

 actual measurable amount of rain or snow. 



o. Of the winds from stormy quarters during the year, 03 per cent, were 

 followed within 24 hours by measurable amounts of rain or snow. Of 

 the winds from fair quarters, 72 per cent, were followed by fair weather 

 for 24 hours, with no measurable amount of rain or snow. The mean 

 verification of all the wind indications was 68 per cent. 



t:. The comparative value of forecasts, from lunar indications which 

 might have been foretold years in advance, and from wind indications 

 which are good only for a day in advance, was 45 -^ 63 = 71.4 per cent, 

 for storm ; 62 -h 72 = 86.1 per cent, for fair ; 53 -h 68 = 77.9 per cent, for 

 all. 



p. The percentage of verification for stormy indications was greatest 

 in winter and least in summer. 



<7. The percentage of verification for fair indications was greatest in 

 autumn, and least in winter. There were marked indications, however, 

 of a tendency toward general maximum verification in summer. 



T. The percentage of total verification was greatest in winter, when the 

 thermal disturbance of Moon's tidal action is least, and least in summer, 

 when the thermal disturbance is greatest. 



u. The percentage of verification, both for the fair and for the stormy 

 indications, was greater in the equinoctial semester than in the solstitial. 



(p. The conflict of solar daily and monsoon influences with lunar 

 monthly tidal influences was shown in numerous cases of stormy anti- 

 cyclonism and fair cyclonism which had been overlooked in the daily 

 forecasts from Washington. Predominating solar influence accounted for 

 32 of the 46 abnormal days during the fair lunar tendencies, and 51 of the 

 66 abnormal days during the stormy lunar tendencies, or 74 per cent, of 

 the whole. 



It would be unwise to draw any positive conclusions from the results of 

 a single experiment, or from observations for a single year at a single 

 station, but there is certainly encouragement for continuing the line of 

 investigation which is here indicated. 



