i56 Mr. J. P. Harrison on LuiKrr Influence over 



for forty-five years. On tlie 20tli day there was a slight rise, 

 reducing the deficiency to lOi degrees, which was still further 

 reduced on the 21st (the day of full moon) to 8 degrees defi- 

 ciency. On the three following days the temperature fell again, 

 reaching its minimum on the 24th; and then, on the 25th 

 day — two days before last quarter — the frost broke up. These 

 variations in temperature were found to follow the line of the 

 lunar curves of annual mean temperature, even in the rise which, 

 as it has been stated, occurs so frequently at the time of full 

 moon. It should be mentioned that the wind continued steadily 

 in the north or north-east during the short thaw at first quarter, 

 and the rise at full moon. 



5. So striking was the aspect of the month, that it led Mr. 

 Glaisher (from whose description of tlie weather in the daily 

 papers the above figures are abstracted) to search for similar 

 periods of continued cold in former Novembers. During the 

 last forty-five years, it appeared from the investigation that he 

 then instituted, there had been no period analogous to it since 

 the year 1815. There were, however, eight periods which were 

 noticeable ; and on referring to the Nautical Almanac, I found 

 that they occurred on the following days of the lunation : — viz., 

 in 1816, between new moon and first quarter; in 1827, also 

 between new moon and first quarter ; in 1829, from the day 

 before last quarter to the day before new moon ; in 1838, from 

 the same day as the last to the third day after last quarter ; in 

 1841, between uew moon and first quarter; in 1849, from the 

 third day after first quarter to the day before full moon ; in 

 1851, from the day before last qiuu'ter to tlie third day after last 

 quarter; in 1856, between new moon and first quartei', — in all 

 the above instances at the three periods of supposed depression 

 in the mean temperature of the lunation. 



6. A table was now formed of the mean temperature of each 

 day of the moon's age at Greenwich for forty consecutive No- 

 vembers ; in the following manner : — The mean temperature of 

 the days on which the moon entered upon her four principal 

 phases being first set down in columns arranged vertically at 

 equal distances, at whatever hour of the civil day the changes 

 may have occurred, the remaining observations were entered in 

 intermediate columns. Thus, if the day of the new moon fell on 

 the fourth day of the month, the mean temperature of that day 

 having been first set down, the means of the third and fifth days 

 of the month were entered in the columns immediately adjoining 

 on either side; and the observations in these columns would 

 then be considered as the mean temperatures of the first day 

 before and the first day after the day of new moon*. And so 



* It is hardly necessary to point out that tlic means of the observations 



