28 K. TSTJRUTA. 



apparent lagging of the change of the thermo-electric E. M. F. behind 

 the change of the stress. When a piece of iron after being annealed 

 is stretched beyond its limit of elasticity and by being alternately 

 ioaded and unloaded is brought to show a cyclic state, the thermo- 

 electric E. M. F. changes, during the application of the load, from a 

 value generally positive towards a value generally negative and, after 

 passing through a minimum value, goes on towards, or tends to take, 

 a positive value again, while, during the removal of the load, it under- 

 goes a similar succession of changes in the reverse order, and in a way 

 différent from that during the application. Thus arises a great dif- 

 ference between the "on" and "off" branches of the curve 

 showing the relation of the thermo-electric E. M. F. to load. For 

 the sake of convenience and brevity in referring to this conspicuous 

 feature, Prof. Ewing proposed to apply the term Hysteresis and to 

 call the phenomenon the Thermo-electric Hysteresis (with respect 

 to longitudinal tension). 



§ 3. To this subject I have since 1891 myself paid attention, 

 and I now propose to describe my own observations about these 

 thermo-electric effects of longitudinal tension in iron. They are by 

 no means altogether novel ; in fact, they contain little, if any, more 

 than general confirmation and illustrations of the results obtained by 

 others, especially by Prof. Ewing. 



§ 4. I will begin by giving sume details of my arrangement, 

 which may be said to stand between those of Dr. Cohn and 

 Prof. Ewing. 



A piece of annealed soft iron wire about 1 metre long (call it wire 

 A) is stretched over a very stout long block of wood mounted horizont- 

 ally on a stone pier. At one end of the block there is fixed a strong 

 cylindrical heating apparatus, which has an axial hollow space. This 

 space is shut in at one end by a small brass piece with a hole sufficient to 



