1885.] 209 [Chase. 



■wliere the verification was 100 per cent., tlie weather at every station, on 

 each of the days, being such as was foretold. 



II. Tests of Solar Influence. 



^. The special report of the Board of Indications, showed a verifica- 

 tion of 70.7 per cent, for the fair forecasts and a rainfall, in amounts suffi- 

 cient to be measured, in 25 per cent, of the regions for which stormy indi- 

 cations were foretold. The full significance of this test cannot be satisfac- 

 torily determined, because the normal proportion of stormy winds 

 which bring actual rainfall has never been published. 



rj. The forecasts which were authorized by the Signal Service tables 

 showed a verification of 1275 -=- 19 ^ 67.1 per cent, on Christmas day, and 

 of 575 -r- 15 = 38.5 per cent, on New Year's Day, or a general mean of 

 1850 -7- 34 = 54.4 per cent. In three of the regions at the former date, and 

 in seven at the latter, no forecasts were prescribed by the tables, the wind- 

 deflecting tendencies being from doubtful azimuths. This test, like the 

 foregoing, is affected by the uncertainty as to what constitutes a satisfac- 

 tory verification of storm-forecasts which cover winds from one-half of 

 the azimuths. 



0. As nearly as could be ascertained from the tri-daily reports, the 

 stormy indications in the Signal Service tables were verified in 65 per cent, 

 for the fair winds, and there was measurable precipitation after 43 per 

 cent, of the stormy winds, the general mean verification being 60 per cent. 

 This would indicate a verification, according to the preceding test, of 

 54.4 -=- 60 = 90.7 per cent. 



i. The ratio of lunar to solar monsoon influence, which was shown in 

 tests a and t] (59.1 ^- 54.4 = 1.086) agrees very nearly with the ratio which 

 was shown by the winds at Haverford during the past year (545 -^ 514 = 

 1.06). Of 1059 observations, 545 were nearer the azimuth which repre- 

 sented the lunar tidal tendency and 514 nearer that which represented 

 the solar monsoon influence, as given by Coffin (Winds of the Globe, 

 p. 431). This degree of accordance seems to justify the belief that 

 the subsidiary value of lunar normals would be found as great elsewhere 

 as it is at Haverford, for detecting and coordinating the abnormal in- 

 fluences of equatorial or polar, cyclonic or anticyclonic, local or general 

 currents. 



III. Tests at Haverford College for ISS4. 



K. Tidal acceleration of atmospheric currents was accompanied by low 

 barometer ; tidal retardation by high barometer. 



/.The percentage, both of stormy winds and of cloudiness, was greater 

 in the lunar stormy cycles than in the fair cycles. 



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

 stormy than fair, 54 were accompanied by measurable precipitation ; on 

 32 other days there were winds from stormy directions ; 11 others were 



TROC. AMER. PHILOS. SOC. XXII. 118. 2a. PRINTED MARCH 14, 1885. 



