ON THE THREE REPORTS OF THE LIVERPOOL COMPASS COMMITTEE. 95 
were built was probably unknown to Mr. Airy. He suggested that the 
particular character of the semicircular deviation in these vessels might be 
due to the direction of rolling of the plates of which the ship was composed. 
Subsequent experiments, made by the same eminent philosopher, on iron 
rolled in different directions, lately communicated to the Royal Society, but 
not yet published, show, as we understand, that the effect of direction in 
rolling, though appreciable in each separate plate, is not great, and probably 
has little, if any, appreciable effect in a ship. In concluding our observations 
on the paper, we must not omit to say that one of the most valuable parts of 
Mr. Airy’s paper, viz. the mechanical correction of the deviation, does not, as 
we consider, come within the scope of this Report, and that, in passing it over, 
we must not be considered as underrating its importance. 
Mr. Airy’s second paper has not that value which is given to the first by 
careful observations made by himself on selected ships. It contains a dis- 
cussion equivalent to the determination of 33, €, and BD of the magnetism of 
various wood-built and iron-built ships from observations made in various 
latitudes, and an endeavour to deduce from such observations the two parts 
of which 33 is composed; but Mr. Airy had the disadvantage which is still 
met with by those who attempt the discussion, viz. the want of sufficient 
determinations of the deviations of the same iron vessel in different magnetic 
latitudes, and he was consequently unable to obtain any very precise evidence 
of the amount of the subpermanent magnetism in iron ships, or its change on 
a change of Jatitude. 
The work of Captain Johnson, to which we have referred, is a great store- 
house of the results of observations of deviation made on board ships of war. 
There are, however, several reasons why it does not require very detailed 
mention here. The deviations are chiefly those of wood-built ships. They are, 
therefore, generally small and regular. They are not compared with theory, 
and do not in all cases furnish sufficient data for the comparison. Such 
comparison as can be made will, as regards iron-built vessels, be found in 
My. Evans’s paper in the Phil. Trans. of 1860, referred to above. 
It is to Dr. Scoresby that we are indebted for the observation that the 
semicircular deviation of iron ships is chiefly due to their position when 
building. 
In considering this subject, there are one or two points which must be 
borne in mind. Supposing, as we may no doubt do, that the iron is, as 
regards position and quality, symmetrically placed on each side of the midship 
line, we may consider separately the permanent or subpermanent magnetism 
caused by fixing, first, the magnetism induced by the horizontal force, and 
secondly, that induced by the vertical force. As regards €, the same reasoning 
which shows that it cannot arise from transient induced magnetism also shows 
that it cannot be caused by the fixing any vertically induced magnetism, but 
must arise either from independent permanent magnetism in the iron, or 
from fixing the horizontally induced magnetism. 
On the other hand, as regards %§ the case is different. It may be 
caused not only by the subpermanent magnetism originally induced by the 
horizontal force, and fixed in building, but by transient vertically induced 
magnetism, and also by the subpermanent magnetism arising from fixing, in 
the process of building, the transient vertically induced magnetism. Between 
approximated to that of the line of no deviation in the ‘Garry Owen,’ which was about 
N.W. by W. 3 W., and that in his subsequent works he did not revert to the subject ; and 
that the hint here given was not pursued by subsequent investigators. 
