45 
The means for each decade vary from the above given value 
(29-905) by + 0*045—0-015 —0 025,—0*025, and + 0-025. 
The last decade was in execess for eight montlis, all except 
April, May, July, and November, of the annual mean (29*905) 
for the 50 years. 
It is interesting to note the variation in mean for 10, 20, 30, 
40, and 50 years. 
Mean, 1841—'50, 29*95 inches. 
„ 1841—60, 29*92 „ 
„ 1841—70, 29*905 „ 
,, 1841—80, 29*90 „ 
„ 1841—90, 29*905 „ 
As the variation is only 0*07 between any two decades, we 
ma}^, apparently", regard the 50 years mean as correct within 
0*007 of an inch. For we may properly assume that the error 
will not exceed J of the extreme variation, in any decade, from 
the mean of the five decades. 
The following table is based upon the more salient points of 
pressure. 
EELATIVE PEESSUEE IN DECADES. 
1 
Subjects for Compabisox. 
’40’s. 
’50’s. 
’60’s. 
’70’s. 
’80*8. 
Highest Max. for Morah. 
4 
3 
3 
0 
2 
,, Min. .. 
91 
-2 
5 
2 
2 
0 
,, Average „ .... 
5 
2 
0 
2 
3 
No. out 15 High-Pressure 
Years . 
No. of Monthly Maxima above 
30-75 . 
8 
2 
0 
2 
3 
3 
9 
3 
2 
2 
Totals, A. 
22i 
to 
tCiM 
8 
8 
10 
Lowest Max. for Month .... 
0 
2 
1 
0 
4 
,, Min. ,, - 
1 
1 
3 
3 
3 
,, Average ,, 
1 
H 
4 
0 
No. out of 147 Low-Pressure 
Years . 
1 
4 
4 
5 
0 
No. of Monthly Minima below 
28-156 . 
4 
4 
5 
7 
4 
Totals, B. 
7 
I4i 
24 
11 
Resutls: —A -4- (24—B) .... 
39| 
31| 
8 
23 
jMean Annual Pressure .. . . 
29*95 
29-89 
29-88 
29 88 
29-93 
This table, and the similar ones for warmth and rainfall, are 
intended to compare such points as usually impress the mind as 
Decades. 
Successive 
means. 
Probable limit 
of error. 
