26 Mr. J. Joly on a 



therefore, desirable to ascertain if this sudden loss of per- 

 manent molecular strain would, like permanent magnetization, 

 take place at a lower temperature than the loss of temporary 

 susceptibility. The iron wire in the last experiment was., 

 therefore, permanently twisted through twenty revolutions 

 and then released from torsional stress. When, on heating, a 

 certain temperature had been reached there was a sudden 

 untwisting of the wire, followed after an interval of several 

 seconds by the usual deflection of the Thomson's galvanometer, 

 which showed that the temporary magnetic susceptibility was 

 beginning to disappear. This experiment was repeated at 

 least six times with different pieces of the same kind of wire, 

 and, invariably, with the same result — the permanent torsion, 

 like the permanent magnetism, began to disappear suddenly at 

 a temperature considerably below the temperature at which 

 the temporary magnetic susceptibility begau to disappear. 



The contrary, however, was the case with a specimen of 

 annealed nickel wire* of the same length as the iron but of 

 rather less diameter. When this wire had received a few 

 turns of permanent torsion f, and was afterwards heated, the 

 loss of temporary magnetic susceptibility occurred at a much 

 lower temperature than the rapid loss of permanent torsion ; 

 whilst Berson has shown that the permanent magnetization 

 of nickel resembles the permanent magnetization of iron in 

 disappearing before the temporary susceptibility. 



This experiment, therefore, shows that the temperature at 

 which permanent magnetism begins to suddenly disappear is 

 not the temperature at which permanent torsion begins to 

 suddenly disappear. 



III. On a Diffusion Photometer. By J. Joly, M.A.B.E.% 



A PECULIAR appearance presented by a translucent 

 substance, when rendered discontinuous by a crack or 

 break, is probably a matter of common observation. When 

 such an object — it may be the familiar one of a cracked 

 paraffin candle — is placed in an unequally illuminated field, 

 the presence of the discontinuity shows out, in general, boldly, 

 as a plane of separation between a bright and a dark portion 

 of the substance. The light, in fact, diffused through the 

 substance, passes only in a feeble degree across the crack, so 

 that that part of the body near the source of light appears 



* Containing nearly 98 per cent, of nickel 



t Only two or three turns can be given before the wire breaks. 



% Communicated by the Author. 



