160 Miscellaneous Intelligence. 
the cold part of it be placed under the needle. The following experi- 
ments were made with cast bars six or seven inches long, one inch 
in thickness, and formed either as cylinders, or as prisms with three, 
four, or six sides; solid and hollow balls were also employed :— 
1. If one extremity of a bar of bismuth be heated, the bar be 
placed in the direction of the magnetic needle, with its cold end to 
the north, and the hot end be brought under the needle, the point 
of the needle will turn towards the east. 2. If the direction of the 
bar being preserved, it be moved towards the south until its cold 
end is under the needle, the needle will turn towards the west. 
3. The inverse effects are obtained when the hot end of the bar is 
towards the north. 4. When the bar is heated in the middle, and 
the ends preserved cold, the same effects are obtained for each half 
of the bar. 5. The magnetic effects are sensible when one part of 
the bar is heated merely by the hand and the other cooled by snow. 
III. The magnetic action of metals unequally heated depends on 
the form given them in casting, and in this it differs from the 
action of Qérsted’s connecting wire. 1. If an equilateral tri- 
angular prism of bismuth be used as in the former experiments (1), 
and its faces be turned upwards successively, one of its faces will 
make the needle deviate to the east, the next face (that towards 
the east) brought into the place of the first, will make the needle 
deviate to the west; the third face has so uncertain an effect that it 
may be considered as null. 2. If a square or four-sided prism of 
bismuth, antimony, or zinc, be used in a similar manner, it will be 
found that two contiguous faces when turned upwards will make 
the needle move eastward, whilst the other two faces will move it 
westward, so that the prism may be considered as composed of two 
triangular prisms of which the un-magnetic faces are in contact. 
3. With a regular hexagonal prism three of the faces move the 
“needle eastward, and three move it westward. 4. Cylinders pre- 
sent peculiar effects; a cylinder of bismuth had been thrown with 
its mould into cold water immediately after being cast, another was 
suffered to cool slowly ; when these cylinders were used in place of 
the prisms, the ends which were uppermost in the moulds being 
placed under the needle, one part of the curved surface urged the 
needle to the east, and the other part to the west; these parts were 
equal in the first cylinder, but unequal in the second. When the 
other extremities of the cylinders were placed under the needle, 
then the curved surface of the first cylinder presented four nearly 
equal portions which successively turned the needle to the east and 
west: the second bar presented six similar portions. 
The differences remarked between the extremities of the cylin- 
der, and also between the cylinders themselves, when cooled slow- 
ly or rapidly, induces M. d’ Yelin to conclude there is some rela- 
tion between the crystallization of metals and their magnetic pro- 
perties.— Bibliotheque Universelle, xxiv. 253. 
