102 
It appears, therefore, that at 9 a.m. the mean amount 
of cloud is slightly greater on days of new moon than on 
days of full moon ; at 1 p.m. the amounts are sensibly equal, 
but at 9 p.m. the amount on days of full moon is greater 
than on days of new moon. It will also be seen that while 
on days of full moon the mean diminution in the amount of 
cloud between 9 a.m. and 9 p.m. is (bo 4 ; on days of new 
moon it amounts to 1*72. 
It has been supposed by Sir John Herschel and other 
meteorologists that the full moon had the effect of partially 
dispersing the clouds, but the above results show very clearly 
that any effect which the moon may have is of an opposite 
character, as in every winter the mean amount of cloud was 
greater at the time of full than of new moon. 
The observations of mean daily temperature at the times 
of full and new moon have yielded the following results 
Mean Daily Temperature 
on days of 
Full Moon. New Moon. 
1871-2 
44-3 
43-4 
+ 0-9 
1872-3 
40-3 
43-0 
-2-7 
1843-4 
45-4 
46-0 
-0-6 
1874-5 
41-6 
41-9 
-0-3 
1875-6 
38-2 
43-5 
-5-3 
1876-7 
43*7 
45-4 
-1-7 
1877-8 
45-9 
44*5 
+ 1-4 
Means . . . 
42-77 
43-96 
-1-19 
The days of the full moon were therefore on the average 
1T9 colder than the days of new moon. This result led 
me to infer that at the times of full moon northerly winds 
had been more frequent than at the times of new moon. 
Observations of the direction of the wind are made at 9 a.m., 
1 p.m., and 9 p.m. daily ; and as the velocity of the wind is 
generally greatest about 1 p.m., its general direction can be 
more accurately determined at that time, and therefore a 
table was formed, showing its direction at 1 p.m. on the 
selected days at the times of full and new moon during the 
