96 
LIEUT.-COLONEL SABINE ON TERRESTRIAL MAGNETISM. 
the disturbance in the same ship was sometimes greater on the eastern, and some- 
times on the western points ; these small irregularities, having no uniform character, 
are regarded as included amongst those varying accidents which are classed gene- 
rally under the name of observation error. It is proper, however, in consequence 
of this occasional irregularity, that the data from which constants are to be derived 
for general corrections should consist of the mean of observations on corresponding 
points on the east and west sides of the compass ; in this view we have as available 
observations in the preceding table those on the following points of the compass. 
North ......... 
N.W 
N.E. .... 
W.|S. . . . .' 
E. ^ S. (from E. and E. ^ S.) . . 
W. by S . . . 
E. by S . 
W.S.W. 
E.S.E. . 
s.w. 
S.E. (from E.S.E. and S.S.E.) . 
South 
-81 19-5 
| —80 58-5 
j -79 58-5 
| —79 49 
j -79 29 
j -79 04-5 
. —78 30 
We have here 2° 49 ,# 5 for the difference between the inclinations observed with 
the ship’s head north and south ; the value of a which will give that amount for the 
sum of the corrections at north and south when the dip is between —79° and —80°, 
(neglecting c as too small in such case to require consideration), is about +' 026 . 
The observations at north were four in number, — those at south six, and on different 
days, — they were as follows : — 
North. South. 
January 8, —81 19 
8, —81 20 
8, —81 18 
13, —81 21 
Mean .... 81 19 - 5 
January 7, —78 28 
8,-78 31 
11, —78 28 
13, —78 25 
13, —78 33 
14, -78 34 
Mean 78 30 
From the accord which these observations respectively exhibit, it is clear that we 
should not be justified in taking a value of a which should differ much from +-026. 
If we now refer to the observations which were made in the Terror soon after her 
arrival at the Falkland Islands, when the ship’s head was placed on the principal points 
of the compass for the purpose of determining the values of the constants, we shall 
