130 
MR. GLAISHER ON THE CORRECTIONS TO BE 
Table III. — Showing the corrections to be applied subtractively to the simple arith- 
metical mean of the maximum and minimum thermometers, to deduce from their 
readings the true temperature of the air. 
January . . . 
0 
. . 0-2 
July. . . 
0 
.... 1*9 
February . . . 
. . 0*4 
August. . 
.... 17 
March .... 
. . 1*0 
September 
.... 1-3 
April .... 
. . 1*5 
October . 
.... 1-0 
May .... 
. . 17 
November 
.... 0-4 
June .... 
. . 1-8 
December . 
.... 0-0 
We have thus two easy methods of finding the true mean temperature ; first, by 
taking observations several times a day, and applying corrections to their means 
from Table II. ; and, secondly, by taking the half of the maximum and minimum 
readings and correcting it by the numbers in Table III. 
At all places the form of the diurnal variation is a single progression, having one 
ascending branch and one descending branch, the maximum occurring early in the 
afternoon and the minimum occurring at about sunrise ; but the amount of the dif- 
ference of these extremes is variable, depending upon latitude, elevation, locality and 
geological formation of the country. 
If we compare the mean temperatures of places that differ considerably from each 
other in latitude, we shall find that the mean values are lower as we proceed north. 
If we compare the mean temperatures of places having the same latitude, we shall 
find that the mean value of those situated at the higher level will be less than those 
at the lower level. 
If we compare places having the same latitude, we shall find that the mean tem- 
peratures of those places situated inland will be higher in the summer months, and 
lower in the winter months than those situated in the vicinity of the sea. 
If we compare places differing only in their geological formations, we shall find 
that those places situated upon an arid, dry soil, will have a greater range of tempe- 
rature than those situated upon a clayey, wet soil. 
It is therefore possible that the corrections in Table II. may not be of universal 
application, but, as the form of the curve described by the daily march is similar at 
all places, with the exception of being more or less bold, the turning points occurring 
at nearly the same local time, it is most probable that the amount of the correction 
applicable to any hour at any place is the same part of the whole monthly mean 
daily range at that place, as the correction at Greenwich is of the monthly mean 
daily range at Greenwich ; I have therefore computed the following table upon this 
assumption, to be used at those places where the daily range of the temperature of 
the air is remarkably small or remarkably large. 
