48 
twenty years, and apparently chiefly owing to the lower 
mean temperature of the month of July: July only exceed- 
ing 60° in two years, 1865 and 1868, whilst in the earlier 
period July in six years exceed 60 J, 0 , viz., in 1850, 1852 
(67°'9), 1854, 1855, 1857, 1859. 
On examination of the mean variations of temperature 
for each month (Table II.), we find that the greatest amount 
of variation of the mean monthly temperature may be ex- 
pected in February, and the least in October, or, arranging 
the months in their order of amount, we have — ■ 
o o 
October . . . . 
.. 1-50 
March 
... 2-17 
April 
.. 1-67 
J My 
... 2-35 
September . 
.. D89 
November ... 
... 2-37 
May 
.. 1-91 
August 
... 2-61 
January .... 
.. 2-02 
December . . . 
... 3-01 
June 
.. 2-06 
February ... 
... 3-26 
One point in this seems somewhat curious, and that is, 
that the wettest and driest months should be liable to about 
the same changes of temperature. The distribution seems 
very irregular, months widely apart coining next one another 
as regards this element of temperature. 
At the bottom of Table 2, 1 have given the values of the 
Probable Yariation of Mean Temperature for each month, 
computed from my own observations for the twenty years, 
and, as the late Manuel J. Johnson gave similar values com- 
puted from Dalton’s Observations, 1794 to 1818 (Radcliffe 
Observations, vol. xv., 1854), I annex them for comparison 
25 Years. 20 Years. 
Dalton, Old Trafford, 
1794 to 1818. 1849 to 1868. 
January ±27 ± D8Q 
February T8 3 ‘50 
March P6 1*93 
April 2*1 T49 
May T9 1*70 
June T6 T84 
July 1-8 2-09 
August 1‘2 2 ’48 
September 1'6 P68 
October 1-7 D33 
November 2*0 2'11 
December ±1*8 ± 2-68 
