100 DISCOVERY CH. 



barometer. Though this amount is well within the 

 limits of accurate measurement, it is too small to be 

 distinguishable with an ordinary barometer. A rise or 

 fall of one-fiftieth of an inch is scarcely worth considering 

 as a sign of change of weather, yet that is the greatest 

 deviation for which the moon can be held directly 

 responsible. While, therefore, air-tides can be detected, 

 they are of no value whatever as a means of forecasting 

 weather. 



The only definite association that can be regarded as 

 established between changes of the moon and weather 

 relates to the occurrence of thunderstorms, and it is 

 noteworthy that this is overlooked completely in pro- 

 verbial philosophy. Thunderstorms are found to be 

 slightly more frequent near New Moon and the First 

 Quarter than near Full Moon and the Last Quarter. 

 This is clear from the adjoining table, which is given as 

 much to show that the lunar influence has been the 

 subject of many investigations in different countries 

 and extending over long periods of years as to illustrate 

 the kind of result obtained. 



Referring to this table, Prof. W. H. Pickering, who 

 brought the results together, remarks : 



The number of observations here collected seems to be large 

 enough to enable us to draw definite conclusions, without fear 

 that further records will revise or neutralise them. From these 

 observations we conclude that there really is a greater number 

 of thunderstorms during the first half of the lunar month than 

 during the last half ; also, that the liability to storms is greatest 

 between New Moon and the First Quarter, and least between 

 Full and Last Quarter. Also we may add that while theoretically 

 very interesting, the difference is not large enough to be of any 

 practical importance. 



So far, then, as earnest investigations have been 

 carried out with the purpose of associating the weather 



