

1865.] Mr. J, P. Harrison Lunar Influence on Temperature. 225 



on the scale-paper were united by straight lines, and thus formed what 

 are usually termed " curves " of temperature. 



The results of the tabulation of the Greenwich observations. The 

 tabulation of the mean temperatures of the 520 lunations between 1814 

 and 1856, resulted in the complete confirmation of the phenomenon origi- 

 nally observed ; that is to say, the maximum mean temperature showed 

 itself, as before, in the first half of the lunation, and the minimum mean 

 temperature in the second half of the lunation. The difference between 

 the maximum and minimum temperatures for the 520 lunations was 1 

 Fahr. (see PI. V. fig. 1). 



In the autumn of 1860 M. Faye communicated the above results to 

 the French Academy*. 



Additional Results in 1856-65. The author has now the honour of 

 laying before the Royal Society additional confirmatory evidence derived 

 from a tabulation of mean temperatures at Greenwich for the eight years, 

 or 99 lunations, which have elapsed since the year 1856f. 



Upon examining the lunar curve of temperature derived from these 

 means (see PI. IV. fig. 2), the maximum mean temperature will be again 

 found in the first half of the lunation, at the moon's first quarter, and the 

 minimum mean temperature in the second half of the lunation. The 

 difference is 3'5 ; the maximum is 51'7 ; and the minimum 48 0< 2. The 

 mean of the period is 49'56. 



And on adding the sums of mean temperature of this period to the 

 sums of the mean temperatures in the Table of 520 lunations, and divid^ 

 ing the sums of the several columns by 619 (the number of lunations 

 which occur in fifty years), the maximum is still found to occur at the first 

 quarter, and the minimum shortly after last quarter. The difference be- 

 ween the maximum and minimum mean temperatures is 1'33. A curve 

 of the mean temperatures for the 619 lunations will be found in PI. IV. 

 fig. U. 



Explanation of the Phenomenon. Although the recurrence of higher 

 temperatures in the first half of the lunation, and more particularly at the 

 moon's first quarter as a meteorological fact is not affected by the 

 correctness or incorrectness of any explanation which may be given of the 

 phenomenon, yet it will be well to state that a probable cause for the 



* Comptes Rendus, December 1860. 



t The Tables were laid before the Society, and are available for reference. 



t As regards the annual sums of temperature of the two days of maximum and minimum, 

 the sums on the former day are higher than the sums on the latter day in 34 years out of 50. 

 And the sum of the differences, in the years in which the mean temperature of the day 

 before first quarter is higher than the mean temperature of the second day after last 

 quarter is 783 6, whilst the sum of the differences, in the years in which the mean 

 temperature of the former day is lower, is 220 0> 0. For several years together, however, 

 the day of maximum temperature presents itself, not on the clay before first quarter, 

 but a day or two later (see PI. V. figs. 2, 3, and note oil fin.). 



