114 
the sum of the amounts is 4*387, which is less than that of 
any nine ot the remaining fifteen hours ; and in fact eight 
out of the nine amounts are the lowest eight in the whole 
twenty-four. It appears therefore from the means of all 
the years that the amount of rainfall is least at those times 
when the declination magnet is at and near its mean posi- 
£ 
tion for the twenty hours, and it will he seen from the 
following table that this relation also holds good for each 
separate year. 
Sums of Hourly Amounts of Rainfall. 
-- .. 
1861 
1862 
1863 
1864 
1865 
1866 
h. h. 
20 to 1 
5 to 9 
in. 
2-904 
1-727 
in. 
2-938 
2-471 
in. 
1-806 
1-750 
in. 
1-821 
1164 
in. 
2171 
2-930 
in. 
2-962 
1-677 
Sums 
4-631 
5*409 
3-556 
2-985 
5-101 
4-639 
Hourly means . 
•514 
•601 
•395 
•331 
•566 
■515 
Sums of remain- 
ing 15 hours. . . 
8-316 
11-444 
8-547 
6-906 
11-769 
12-201 
Hourly means.. 
"554 
•762 
•569 
•460 
•784 
•813 
In every year, then, notwithstanding the great irregu- 
larities presented by Mr. Glashier s curves, the mean hourly 
amount of rainfall was less during those periods of the day 
when the declination magnet was in and near its mean 
position than during the periods when it was at and near 
the extremes of its daily oscillations. 
A valuable series of bi-horary observations of rainfall 
was made at the Lisbon Observatory during the nine years 
1856-64, the results of which are given in the first and 
second volumes of the Annals of the Observatory. The 
mean annual bi-horary amounts of rainfall were as follows • 
h. 
h. 
mm. 
h. 
h. 
mm. 
0 
to 2 
55-9 
12 
to 
14 
65-9 
2 
„ 4 
68-1 
14 
16 
81-5 
4 
„ 6 
57*3 
16 
18 
75-3 
6 
„ 8 
62*4 
18 
20 
77*5 
8 
„ 10 
61*3 
20 
>> 
22 
68-4 
10 
,,12 
62*8 
22 
24 
52*3 
