156 
FIFTH REPORT— 1836, 
ing chart for differences of *01 in the value of the intensity, the 
corresponding intervals of distance being 24*4 geographical 
miles. 
On a comparison of the separate determinations with the re¬ 
sulting lines, it will be observed that the intensities in the northern 
group are greater than those due to their position, those of 
the western group less, and those of the south-western, again, 
greater. These deviations may, in part, arise from the inex¬ 
actness of the assumption with which we set out in the computa¬ 
tion of the lines, and from the sensible deviation of those lines 
from parallelism. But they are probably owing in a much greater 
degree to the disturbing causes to which we have already alluded. 
The separate results composing each of these groups were for 
the most part obtained about the same time, and they are there¬ 
fore probably affected in the same manner, and nearly in the 
same amount, by the irregular fluctuations in the direction and 
intensity of the resultant magnetic force. Of these, the changes 
in the direction of the force are by far the most influential. 
The relation between the corresponding changes in the dip and 
in the horizontal intensity is expressed by the formula 
dh 
h “ 
— tan 3 sin V d l; 
d 5 being expressed in minutes. Hence when <5 = /1°0 ? , the 
change of the horizontal intensity, ~, corresponding to a change 
of one minute of dip is — *00084 ; and for a variation of 12' in 
the dip, the corresponding variation of the horizontal force 
is *01. 
In deducing the lines of dip from observation, it seems ad¬ 
visable to separate the results of the two years. For the weights 
we shall assume 
Limerick (1834) . . . weight = 5, 
Armagh-. . . - = 2 ; 
the weights of each of the other determinations being unity. 
Making the computations for the year 1834, we obtain the 
following results \ 
