622 ; '.,*' COMPARATIVE ELECTRO-PHYSIOLOGY 



If we take a cyclic curve, recording the effects under 

 increasing, followed by those under diminishing, force, it will 

 be found that the forward and return portions of the curve do 

 not in general coincide (fig. 380 a). Thus, when an increasing 

 magnetising force starting from zero acts on an iron rod, and 

 is brought back to zero, the condition of the rod at the end 

 is not exactly the same as at the beginning. A certain 

 amount of molecular work, which is not reversible, has been 

 done during the cycle. A certain molecular distortion persists 

 as an after-effect in residual magnetisation. Similarly, when 



FIG. 380. Cyclic Curves of Magnetisation (a) and of Conductivity (b) 



metallic particles are subjected to cyclic electro-motive varia- 

 tion, an after-effect is found to persist in a change of con- 

 ductivity. In substances belonging to the negative class 

 the after-effect is one of enhanced conductivity l (fig. 380 b). 



Referring again to that molecular condition of the 

 substance which is represented by b in fig. 378, we find 

 that a new increment or accession of force will raise its 

 condition to b'. In this case the acting force has been 

 continuously operative and continuously increasing. On 

 the cessation of the acting force, a substance possessing 

 marked self-recovery will fall back from b' to b. But 

 if there be a certain persistence of after-effect, then a 



1 Bose, ' On the Change of Conductivity of Metallic Particles under Cyclic 

 Electro-motive Variation.' The Electrician^ September 1901. 



