260 Mr. S. BidwelPs Experiments illustrating 



Upon first gripping the metals, especially the harder ones, 

 it frequently happens that before beginning to pull a thermo- 

 electric current is set up between different parts owing 

 to local differences in physical condition, and these thermo- 

 currents are sometimes so strong as to completely mask the 

 effect produced by stretching. Their effect may be practically 

 eliminated by giving a series of short pulls at regular inter- 

 vals, corresponding with the periodic swing of the galvano- 

 meter. The results can thus be observed with perfect accu- 

 racy, but the method is not well suited for use before an 

 audience on account of the difficulty of rendering evident the 

 correspondence between the pulls and the consequent galva- 

 nometer-deflections. The hand and eye should, in fact, work 

 together*. 



The directions of the currents through the hot junctions 

 are given in the following table, which is taken from my 

 former paper, with the addition of the results for wires of tin 

 and lead: — 



S = Stretched. U = Unstretched. 



Metals. 



Forms used. 



Direction 

 of current. 



Hall's 

 effect. 



Copper 



Wire and foil ; pure 



StoU 

 Uto S 

 S toU 

 Uto S 

 S toU 

 S toU 

 S toU 

 Uto S 

 S toU 

 S toU 

 S toU 

 Uto S 

 Uto S 

 S toU 

 S toU 

 No current 



+ 

 + 



-? 



4- 

 Nil. 



Iron 



Wire and sheet ; annealed „. 



Brass . . 



Wire, commercial 



Zinc 



Wire and foil 



Nickel 



Wire 



Platinum 



Gold j 



Silver 



Wire and foil 



Foil, purity 99*9 per cent 



Wire, commercially pure , 



Jewellers' 18 ct. wire and sheet ... 

 Jewellers' 15 ct. sheet 



Wire and foil 



Aluminium . . . 

 Cobalt 



Wire and foil, pure 



Rod ; 8 mm. diameter 



Magnesium . . . 

 Tin 



Ribbon 



Foil and wire 



Uead 



Foil (assay) and wire 







It will be seen that, in every case excepting that of alumi- 

 nium and one out of five specimens of gold, there is perfect 



* The experiment was, however, successfully exhibited at the meeting. 

 All Hall's metals were used except cobalt, which could not be obtained 

 in the form of wire, and the only doubtful case was that of tin, the wire 

 being ruptured before any marked deflection was produced. It should be 

 noticed that the temperature of the hot junction should not exceed about 

 100° C, otherwise, at least in the case of iron and nickel, different 

 results may be obtained. 



