THE MOLECULAR THEORY OF EXCITATION 593 



But before proceeding further with the magnetic responses, 

 we must call to mind two different responsive manifestations 

 of living tissues. We have observed these under the polar 

 action of electric currents, one being the K- and the other the 

 A-effect. Similarly in magnetic substances also, under the 

 action of the magnetising forces, we observe two different 

 effects brought about by opposite polar changes. One of 

 these is the result of north and the other of south polar in- 

 duction ; and of these, for the sake of convenience, we shall 

 fix our attention on the effect induced by the north pole as 

 the normal negative or K-effect. 



The fundamental molecular change induced may here, 

 as in the case of living tissues, be recorded in various ways. 

 In the present case, of the response of magnetic substances 

 under magnetic stimulation, the methods of record may be 

 classified as mechanical, snagnetometric, and electro-motive. 

 Joule discovered that a rod of soft iron undergoes a change 

 of length on magnetisation. Though this variation is very 

 small, I find it comparatively easy to demonstrate and record 

 the responsive change concerned by means of the following 

 device. One end of the iron rod is fixed, and the free end, 

 carrying a wooden disc, rests on a tambour covered with 

 stretched indiarubber. The tambour chamber is closed 

 except at the point where a capillary tubing of glass enters 

 it. This tube contains a short index. On now suddenly 

 inducing magnetisation by a magnetising coil, the rod under- 

 goes instantaneous elongation, and the resulting expulsion 

 of air from the tambour causes a corresponding movement 

 of the index outwards. Cessation of the magnetising current 

 is attended by immediate recovery. It need only be men- 

 tioned here that by making the diameter of the tambour 

 sufficiently large, and that of the capillary tube sufficiently 

 small, and by optically magnifying the movement of the 

 index, it is easy to obtain for this mode of experiment a very 

 high degree of sensitiveness. 



It is much easier, however, to record responsive molecular 

 changes by the usual magnetometric, or by the induction or 



QQ 



