382 THE MOON AND THE WEATHER. 



from three to five days ; for the changes recorded at the quarters are to be un- \> 

 derstood those which occurred within the space of two or three days ; and for I 

 those not coinciding with the phases the changes which occurred on a particu- > 

 lar day. It will not, we presume, require much mathematical sagacity to per- ' 

 ceive that the results of such an inquiry must have been just what Toaldo found 

 them to be ; and that if instead of taking the epochs of the lunar phases he had 

 taken any other periods whatsoever, and tried them by the same test, he would 

 have arrived at the same results. Five days at the new and full moon would \ 

 include rather more than a third of the entire lunar month ; and thus a third of 

 all the changes of weather which occurred in that period were ascribed by To- 

 aldo to the lunar influence at these epochs. 



Professor Pilgrim has examined a series of observations on the lunar phases j 

 as connected with the changes of weather, made at Vienna, and continued from ( 

 1763 to 1787 a period of 25 years and he has found that, of every him- ( 

 dred cases of the phases, the proportion of the occurrence of changes to that 

 of the settled state of the weather was as follows : 



Changes. Settled Weather. 



New moon 58 42 



Full moon 63 37 



Quarter 63 37 



Perigee 72 28 



Apogee 64 36 



New moon at perigee 80 20 



New moon at apogee 64 36 



Full moon at perigee 81 19 



Full moon at apogee 68 32 



Admitting these results, it would follow, contrary to popular belief and to the 

 observations of Toaldo, that the new moon is the least active of the phases ; and 

 that the full moon and quarters are equally active ; also that the influence of 

 perigee, or the nearest position of the moon, is greater than than that of any of 

 the phases, while the influence of apogee, or its greatest distance, is equal to 

 that of the quarters and full moon, and greater than that of the new moon. 



But Pilgrim's calculations are liable to objections similar to those to which 

 Toaldo's are obnoxious. Like Toaldo, he included in his enumerations of 

 changes, corresponding to the phases, changes which occurred the days pre- 

 ceding and following the phases : this being the case, the only wonder is that 

 the proportion which he has found, especially for the new moon, is not more 

 favorable to his hypothesis. But independently of this, Pilgrim's results are 

 not entitled to any confidence : they bear internal evidence of their inaccuracy; 

 and besides, the observations were not continued for a sufficient length of time 

 to give a safe and certain conclusion. 



In the years 1774 and 1775, Dr. Horsley directed his attention to the 

 question, and published two papers in the Phi/.osophical Transactions (to which 

 we have already adverted), with a view to dispel the popular prejudice on the 

 subject of lunar influences. Horsley's observations, however, were confined to so 

 short a period of time (two years) that they could not be expected to afibrd any 

 satisfactory results. He. found that in the year 1774 there were only two 

 changes of weather which corresponded with the new moon, and none with 

 the full moon ; and that in the year 1775 there were only four changes which 

 corresponded with the new moon, and three with the full moori. 



Dismissing, then, this popular notion of the correspondence of changes of 

 the weather with the lunar phases, let us consider the question of lunar influ- 

 ences in a more general point of view, and see whether observation has sup- 

 plied any ground for the supposition of any relation of periodicity between the 

 moon and the weather. M. Schubler examined this question with considera- 



