42 REPORT—1840. 
temperature of October following should likewise be warmer than its 
average: on the contrary, if August be colder than average, October 
following should likewise be colder than average. The same method 
applies to the other months. 
If there were no other cause for variations in the mean tempe- 
rature of the winter months in different years, except differences 
in the amount of solar heat absorbed during the months of cor- 
responding temperature in the previous summer half of the year, the 
mean temperature of each winter month would always bear a strict 
relation to that of its corresponding previous summer month ; and, 
consequently, if the one were known, the other could with certainty be 
predicted. But there are other causes which diversify the mean tem- 
perature of the same months in different years; and these being inde- 
pendent of the one under consideration, may either be co-operating with 
it, or acting in opposition to it. For instance, the proportion of north- 
erly and southerly winds, and the amount of rain that falls during the 
same winter months of different years, varies greatly. Whether, there- 
fore, among so many other distinct causes of diversity of temperature, 
the difference in the amount of solar heat absorbed during the summer 
months does sensibly affect by its retrocession the mean atmospheric 
temperature of the corresponding months of temperature during the 
subsequent winter season, so as to afford any probable means of pro- 
gnostication, can only be determined by reference to statistical tables of 
the mean temperature of the different months in a succession of years. 
From tables then referred to, Mr. Hutchinson said, it appeared 
that in Scotland deviations in the mean temperature of the summer 
months have a visible influence in producing like deviations in their 
corresponding months of temperature in the subsequent winter half of 
the year. It appeared also, that in the generality of years, the other 
disturbing causes which diversify the temperature of the same winter 
months in different years, such as variations in the direction and 
force of the winds, &c. have less influence, when averaged for a 
month, than we would be apt, @ priori, to suppose. And when the same 
months, for a number of years, are grouped together, and compared as 
is done in the tables, the disturbing causes, which may occasion a great 
deviation from the mean temperature in any particular month in one 
year, seem partially to neutralize each other, and render the influence 
of unusual warmth or unusual coldness, in any summer month, in pro- 
ducing a similar degree of unusual warmth or coldness in its corre- 
sponding winter month, more apparent than could have been anti- 
cipated. 
The author concludes by some observations as to the most suitable 
hour of registering temperature for the purpose of testing and applying 
his speculations (preferring 11 a.m.), and notices as an inference from 
the simple consideration of the ratio of solar radiation in summer and 
winter, that the dependence of winter temperature on summer heat, 
according to the plan of examining it already exemplified, should be 
greater and more obvious the higher the latitude of the place of obser- 
vation. 
