Intelligence and Miscellaneous Articles. oil 
LVII. Intelligence and Miscellaneous Articles. 
INFLUENCE OF MAGNETISM UPON THEEMAL CONDUCTIVITY. 
BY JOHN TROWBRIDGE AND CHARLES BINGHAM PENROSE. 
'"PHE following experiments were made in order to test Maggi's 
-*- results * in regard to the effect of magnetism upon the thermal 
conductivity of iron. Maggi's conclusions have never been con- 
firmed, and have been much doubted by other observers. The 
experiments of Sir W. Thomson f, in which he found that longitu- 
dinal magnetization diminished, while transverse magnetization 
increased, the electrical conductivity of iron, afford — from the fact 
that electrical and thermal conductivities are in general proportional 
— the chief confirmation of Maggi's results. The experiments of 
Thomson have, however, been questioned. 
In the method employed by Maggi, a circular plate of soft iron 
was placed horizontally upon the poles of a vertical horseshoe 
magnet. Through the centre of the plate passed a lead tube 
conveying steam. The surface of the plate was covered with a 
mixture of oil and wax. When the magnet was made, the melted 
wax was bounded by an ellipse. If the conductivity had been 
equal in all directions, it would have been bounded by a circle. 
The long axis of the ellipse was perpendicular to the line joining the 
two poles ; the short axis was parallel to the line. The ratio of 
the axes was 6:5. The two poles were separated from the iron 
plate by paper. In order to compensate for the direct effect of 
the poles upon the flow of heat, two bars of soft iron were placed 
symmetrically beneath the plate, at the extremities of the diameter, 
perpendicular to the line joining the poies. No effort seems to 
have been made to avoid the effect due to the frictional generation 
of heat in the magnetic coil. Several other complicating causes 
are apparent in the arrangement which Maggi used. 
A year ago we made some experiments by a rather rough method, 
to find the effect of magnetism upon thermal conductivity, and 
obtained decidedly negative results. The same results have been 
obtained in the present experiments, though a much more sensitive 
arrangement was employed. The following method was used : — A 
bar of soft Norway iron, 95 cm. long, 1-3 cm. wide, and 0*2 cm. 
in thickness, was placed horizontally through the sides of a wooden 
box 6 cm. wide and 25 cm. high. The top and one side of the box 
were removed. At 17 cm. from each end of the bar was soldered 
a thick German-silver wire. Each projecting arm of the bar was 
enclosed in a glass tube 1*4 cm. in diameter. The ends of the 
tubes were closed with cotton. The ends of the iron bar projected 
slightly beyond the ends of the tubes, and were exposed to the air 
of the room. One arm of the bar was placed between the poles of 
a large electro-magnet, with its flat surface perpendicular to the 
axis of the magnet. The axes of the poles were in the same 
* Bibl. Univ. Archiv. 1850. 
t Phil. Trans, vol. cxlvi. 
