DEVIATIONS OF THE COMPASS. 
437 
absolutely small, may be large relatively to their value. The value of D, 
which in iron vessels, as might be expected, is generally considerable, is 
remarkably permanent in different geographical positions. 
In wooden ships, P' and Q' are generally very small; 15 and C, which are 
ernerally the largest coefficients, vary as the tangent of the dip, and the devi¬ 
ation changes its direction when the ship changes from a northern to a 
southern magnetic latitude. 
In iron ships, P' and Q' seem to be considerable. 
If 15' and C' were small, the part of the deviation depending on 15 and C 
*ould vary inversely as the horizontal force. It would therefore he a niini- 
'■iinn at the magnetic equator, and would increase as the ship went to a di- 
•'.lrice from it on either side. But even in iron ships the values of 13 and C 
•" in to be considerable, except for places in the ship where by accident 
II " rc is a compensation of the effects of the soft iron. And it wo consider 
'* lp dillercnt combinations which may exist, we shall see how very uncertain 
at| V prediction as to the change in the deviations on a change ot magnetic 
hlitudo must be. 
11ms, if in Eugland we have selected a place for the standard compass in 
"'m li the deviations are small, this may arise from the magnetism of the soft 
lr, ’ 1 " counteracting that of the hard iron ; but if so, in southern latitudes they 
, conspire, and cause a large deviation. . ... 
Again, a place may happen to have been selected for the compass in wmen 
'[' ,ro }* « Compensation of the magnetism of the soft iron, arid in which 
*" rc-forc the deviation is caused solely by the hard iron. In lids ease, m a 
WUthern Intitlldt* the ilcvinlinn irnnlrl tinvn llm 0.1 nlC SHTU Slid BUlOUnt. J 
C. 
D. 
E. 
A. 
—« » • 1 w 
‘ SkV /*/ 10 ’ 1852 
n •> .. . 
o / o l 
04 10 11 
-3 2 G 25 
-2 48 G 52 
1 40 7 4 
-3 6 3 55 
0 7 7 17 
-0 12 7 42 
1 17 9 18 
