12 
FIFTH REPORT— 1835. 
of the phenomena , although there did not exist till lately so 
correct and pertinent a collection of measures in support of the 
magnetic as of the electric theory. Indeed, the peculiar fea¬ 
ture in the case of magnetism, that the fluid cannot be trans¬ 
ferred from one body to another, prevented measures of the same 
kind, or of the same accuracy, being employed. And the same 
circumstance affected the theory more directly, in that it left the 
theorist without any steady conviction of the real existence of 
the fluids which were the subject of his calculations. There was 
no magnetic discharge analogous to the electric discharge; and 
thus the magnetic fluid was a hypothesis resting for its evidence 
only upon the one class of facts for which it accounted. More 
recently, indeed, we have had the magnetic fluid proved to be real, 
and connected with the electric fluid by most curious and un¬ 
foreseen relations, but of these it is not our business now to 
speak. The Coulombian theory of magnetism is still an impor¬ 
tant portion of science, as claiming to give the laws of the sta¬ 
tical phenomena: in what manner these laws result from the 
best views we can obtain of the cause of the facts is a matter 
for subsequent consideration. 
The next important event in the history of the theory was a 
series of good observations, made at first with no reference to 
the theory. In the year 1819 Mr. Barlow undertook a course 
of magnetic experiments with a view to enable himself to cor¬ 
rect the local attraction exerted by the iron which ships contain 
upon the compass. In the course of these researches 44 he dis¬ 
covered,” as he says*, 44 certain magnetic laws which seemed to 
him likely to pave the way to a mathematical theory of magne¬ 
tism a mode of expression which seems to show that Cou¬ 
lomb’s theory had obtained little currency. Mr. Barlow’s ex¬ 
periments were made at first by measuring the deviation pro¬ 
duced in a compass needle by an iron sphere. The result, was, 
that, for such a sphere, there exists a plane of no attraction 
coincident with the magnetic equator; and if we measure the 
magnetic latitude of the compass from this plane, and the mag¬ 
netic longitude from the east and west points, the tangent of the 
deviation produced by the attraction of the sphere is as the sine 
of the double latitude, and as the cosine of the longitude : it is 
also as the cube of the diameter of the sphere directly, and as the 
cube of the distance of the compass from the centre inversely. 
These rules were discovered empirically, and published in 
1820; and shortly afterwards Mr. C. Bonnycastle undertook to 
deduce these laws from a theory analogous to Coulomb’s the- 
* Magnetic Attractions , Preface, p. 1. 
