48 
Change in 
original labels in 
1865, from mean 
of max. andmin. 
to mean of 8 a.m. 
and 8 p.m. 
Result of the 
latter. 
71, 79, ’81 ; February of 1853, ’55; December of ’46, ’74, 
’78, ’79. Only in 18 79 were the means of two months below 
32'^, but the same is true of the consecutive months, Dec. ’78 
and Jan. ’79. 
An instructive table is given by adding to these results the 
decades in which extremes of monthly records occurred. 
EELATIYE WAEMTH IN DECADES. 
SUBJFX’TS FOR CoMPARISUX. 
^ ’40s. 
50s. 
’60s. 
'70s. 
’80s. 
Plighest Max. for Month .. 
5 ^ 
2 
3 
0 
2 
„ Min. ,, 
0 
1 
5 
0 
1 
* 
,, Av. 1 , for Month.. 
2 
3 
5 
0 
2 
No out of 17 Warmest Years.. 
1 
2 
7 
5 
2 
No. of Monthly Maxima above 
63° . 
4 
1 
5 
1 
2 
Totals, A. 
12 
9 
25 
11 
9 
Lowest Max. for Month 
0 
4 
2 
1 
1 
5 
„ Min. ,, 
0 
3 
0 
1 
3 
,, Av. 1 for Month.. 
3 
6 
1 
D 
No. out of 16 Coldest Years .. 
4 
7 
0 
1 
4 
No. of Monthly Minima below 
32°. 
3 
2 
1 
0 
1 
Totals, B. 
15 1 
22 
H 
9 
Resflts a (22 — B) . . ..; 
19 
’ 
9 
43i 
24 
16* 
Mean Annual Temperature ..! 
' 
47*5 
46-7 
48-8 
47-9 
47-7 
The line “ Result, A (lo—B) ” is useful for comparison 'vvith the consideration, 
later on, of the mean annual temperature of the several decades. 
In the earlier tables, besides the mean of the max. and min., 
the average of the two is given. An inspection shows that, 
without any record of tlie change being made, this was replaced 
in 1866 by the means of the observations at 9 a.m. and 9 p.m. 
By a fortunate accident the change occurred exactly at the 
close of the 25 years. Whereas the average for the second 
half is 48'00, of the maxima and minima, the mean of 
morning and evening observations is 47'60. Hence we may 
accept the mean for the whole period to be 47'32. 
The divergence of the a.m. and p.m. mean for the ditferent 
months, from that of the max. and min., during the 25 years, 
ranges from — IT in June to + 0 4 in December, the two 
coinciding in early January and November. It '^dll be noted 
