58 Professor CUMMING on the Developement 
tained with copper wires; if wires of silver or platina be used, 
the effects are similar, but the positions of the zero are changed. 
(Table VI.)* The proportions of these bars are between four and 
six of bismuth to one of antimony ; but the metals require so many 
repeated castings before they are thoroughly incorporated, and their 
union is so much affected by fusion, that it is nearly impossible, 
without destroying and analyzing them, to determine accurately 
their composition. 
If it be supposed that the formule, which express the relation 
between the increase of temperature and the magnetic effects, be 
different in the two metals, it is obvious that though in equilibrio 
at one temperature, they will not be so at any other. This would 
account for the phenomena of the first, third and fourth bars, 
supposing them to be rather mixtures than alloys; and perhaps 
the anomaly of the second bar may be resolved by supposing the 
metals in this to be partly mixed and partly forming an alloy ; 
in which case there would be three substances, each following 
a different law. With the same metal, when the differences of 
temperature at its extremities are not considerable, they correspond 
very nearly with the deviations: yet proceeding in an increasing 
ratio, which augments as it approaches the melting point; but not 
so rapidly with bismuth as with antimony. The deviations of two 
similar rods of antimony and bismuth, from 65° to 240° of heat, 
were 10° and 14° respectively, (Table VII.) Taking the deviations 
at intervals of 34° in the one, and 4° in the other, the corres- 
ponding differences of temperature will be 42, 47, 81, and 47, 
60, 68. If the melting point of bismuth be assumed 475°, and that 
of antimony 800°, the deviations corresponding to the last 235° of 
* The order of deviations with copper wires, as shewn by the small compass, was 
(Mable Vib): get; parser tienes a erttars,- aisle ciara 0.36 neg. 0.25 pos. O. 
With platina wires they were ................. 0.35 neg. 30 neg. 65 neg. 
