276 
Wisconsin Academy of Sciences, Arts and Letters. 
With reference to the theory of these changes Prof. Wiedemann says: 
44 Besides the permanent effect dne to an alteration in temperature there 
is a temporary change. Each heating diminishes the permanent mo¬ 
ment of the molecules. Moreover, for the time being, it loosens the 
particles of the body and lessens the strain in which they have been 
placed by the action of external forces, therefore they return a little 
toward their first position of equilibrium, in which they were held by 
the forces acting between them before the external forces came into 
play. Heating thus diminishes the magnetization temporarily, but on 
cooling the molecules return to their former position and the lost mag¬ 
netization is regained. 
“We can produce entirely analogous phenomena if we change the 
temperature of bodies which have suffered a change of form (tortion) as 
a result of the mechanical forces, and observe the increase and decrease 
of this on heating and cooling.” 
*Barus and Strouhal carefully distinguished the mechanical effect of 
heating from the purely magnetic effect. They found that a temperature 
20° or 30 e C. above that of the water in which a glass hard steel rod was 
dipped in hardening produced quite perceptible annealing effects. This 
change in the hardness of steel would naturally affect the magnetiza¬ 
tion. 
According to their experiments, if a glass hard steel rod is thoroughly 
annealed by being kept at the temperature of boiling water for a day or 
two and then magnetized to saturation at the temperature of the room, 
the loss in magnetism on being heated to the boiling point is rela¬ 
tively small and is nearly independent of the time it is kept there. 
Nearly the whole change takes place during the first ten minutes. On 
the other hand, if the bar is not first annealed, the change is much 
larger and is not complete after twenty-two hours heating. 
We pass now to investigations on changes in distribution due to tem¬ 
perature changes. This field has not been extensively worked. These 
two investigations are all that could be found. 
■fKupffer determined at two different temperatures the period of 
vibration of a short needle placed opposite different parts of a long 
magnet, and found the proportional change in distribution greater at 
the ends than in the middle of the bar. All his measurements were 
made before the bar had reached the permanent state. 
X Poloni measured the distribution at various temperatures, by slip¬ 
ping a coil from different parts of the magnet to such a distance that 
the magnet exerted practically no effect, and measuring the quantity 
of electricity thus induced. He worked between temperatures 0° and 
* Bull U. S. Geol. Sur. No. 14, p. 151. 
tKuppfer, Pogg. Ann. 12, p. 133. 
X Poloni, Beibl. 5, p. 802. Atti della R. Acad.dei Lincei 5, p. 262, 1881. 
