﻿606 Dr. R. S. Willows on the 



the other sign would make p come outside the limits 0-1,. 

 and this value it is readily seen corresponds to a maximum E. 



It appeared possible to put this false resistance in evidence 

 in two ways, one direct, the other not so. It was only after 

 working for a considerable time that I became acquainted 

 with Liebenow's work, which is entirely concerned with the 

 indirect method. 



This paper is a description of the attempts I have made to 

 separate the true and false resistances directly. Suppose a 

 current runs through an alloy and that it sets up a back 

 E.M.F. : if it is now quickly reversed, this inverse E.M.F. 

 will at first assist its passage, and more will flow in the 

 second direction than in the first, or, what comes to the same 

 thing, the resistance will appear to be less for the quickly 

 reversed current than for the steady direct one. As the 

 temperature of the junctions will be equalized rapidly on 

 account of their small distance apart, the current reversals 

 must be rapid. I therefore used an alternating current. 



The alloy to be tested formed one of the arms of a 

 Wheatstone's bridge, the adjacent arm being a simple metal 

 such as copper or lead, generally the latter on account of its 

 greater resistivity. The resistances were first balanced for 

 alternating and then for direct current. The pure metal 

 possesses no spurious resistance, and hence the apparent 

 resistance of the alloy should decrease when the alternating 

 current is used. 



The figure (p. 607) shows the arrangement of apparatus. 



A is the resistance-coil of alloy ; B the copper or lead 

 resistance made from specially purified material obtained 

 from Johnson and Matthey ; C, D, are two coils of manganin 

 or eureka wire wound on the same bobbin ; G is the 

 secondary of a small induction-coil ; H, a Leclanche cell ; 

 P, a galvanometer ; Q, a telephone specially wound for bridge 

 work, or a vibration galvanometer ; Si, S 2 , mercury-paraffin 

 switches by means of which the source of current or the 

 detector could be changed. 



There is, in addition to the false resistance of the alloy, 

 the skin effect, arising from the concentration of current in 

 the outer layers, to be taken into account. If R is the true 

 resistance of a ware of length I to steady currents, then the 

 resistance to currents of moderate frequency and of simple 

 periodic form is 



where //, is the permeability and p = 2rr (frequency). The 



