400 Clianye of Lenyth hy Magnetization. 



iron, tungsten-steel, cast and annealed cobalts, we notice the 

 remarkable fact that the changes o£ length of these metals at 

 ordinary temperature, so very different from each other, 

 assume, at sufficiently high temperatures, an extraordinarily 

 simple character ; they tend to become proportional to mag- 

 netic force, a fact which has no doubt an important bearing 

 on the theory of molecular magnetism. 



It is also to be observed that the change of length of the 

 ferromagnetic substances at their critical temperature nearly 

 disappears, and even in cases for which w^e actually observe 

 it the amount of the change is only a small fraction of the 

 change at ordinary temperature. 



In conclusion, let us give a short account of the permanent 

 change with regard to the magnetic change of length. Cool- 

 ing the specimen in liquid air has almost no permanent effect 

 on the change of length at the ordinary temperature ; but 

 heating it to a very high temperature generally produces a con- 

 siderable permanent change. As seen from fig. 15 (PL XV.) 

 the heating of soft iron up to 746° C. does not sensibly affect 

 the length-change at ordinary temperature. Here the crosses 

 (x ) denote the points obtained after heating to 746° C. In 

 tungsten-steel the effect is very large, tending to reduce the 

 elongation in high fields. 



In cast cobalt the effect is still greater, changing totally 

 the course of the curve, as seen from fig. 16. If the spe- 

 cimen is once annealed at a high temperature subsequent 

 heating and cooling between the same limits of tempe- 

 rature produce a slight effect on the change of length at 

 ordinary temperature. But if the upper range of temjDe- 

 rature be further increased the change of length is slightlv 

 affected. This will be seen from the example of annealed 

 cobalt. 



In the experiments at temperatures higher than 700° C. 

 the suspended weight was reduced to 1 or 0*4 kilogram, 

 according to the circumstances. This was found necessary 

 to avoid the gradual elongation of our specimens caused by 

 the yielding at high temperatures. 



It is hoped that the present investigation may be completed 

 by studying the change of length in every stage of rising 

 and falling temperatures, and also by measuring the magne- 

 tization at the corresponding temperatures. The further 

 extension to other ferromagnetics, such as nickel-steels of 

 different percentages, will also be undertaken in the near 

 future. 



