HALL. — THERMO-ELECTRIC HETEROGENEITY IN ALLOYS. 559 



by 4 the algebraic sum of these deflections and comparing them with the 

 effect obtained with a german-silver and copper couple having a like 

 interval of temperature, we found that, on the average, the spiral- 

 straight german-silver couples had a thermo-electric force about ^^ 

 as great 2 as that of the german-silver and copper couple, the spiral wire 

 being, apparently, a little farther removed, in thermo-electric quality, 

 from copper than the straight german-silver is. 



This seems to show that moderately sharp bending (with a radius of 

 0.5 cm.) is not very injurious in the ordinary thermo-electric use of 

 german-silver wire. It is to be noted, however, that, if a german-silver 

 and copper couple is used to measure some very small interval of tem- 

 perature, 0°.l, for example, and if in some bent part of the german 

 silver there is a very much greater interval of temperature, 15°, let us 

 say, such a difference of quality in this bent part as we have inferred 

 from our experiments on the spirals would make an error of 10 per cent 

 in the estimation of the 0°.l interval. But if in such a case it is prac- 

 ticable so to vary the conditions of the experiment as to reverse the 

 small temperature-interval to be measured, without at the same time 

 changing materially the gradient of temperature in the german-silver 

 wire, error caused by the non-homogeneity of this wire will be practi- 

 cally eliminated from the result of the combined observations. 



It seems unlikely that the bends in the upper parts of the german- 

 silver wires of the couples A 2 , B 2 , C 2 , D 2 , or A 3 , B 3 , C 3 , and D a , had 

 any important effect in the performance of these couples. 



Tlier mo -electric Heterogeneity in Straight German- Silver Wire. 



We now return to the question whether in a straight piece of german- 

 silver wire there is any discoverable thermo-electric heterogeneity. 



After some tentative experimenting the arrangement indicated, as seen 

 from above, by Figure 5 was adopted: G G is the german-silver wire, 

 held nearly straight but with no great tension by the screw-heads S and 

 S ; C and C are copper wires soldered to the german-silver, the junc- 

 tions being immersed in glass vessels containing oil at the temperature 

 of the room ; M M is a meter-rod. 



The copper wires being connected with the terminals of a galva- 

 nometer, the sharp edge of a piece of ice was placed in contact with the 

 under side of the german-silver wire, as in Figure 6, opposite the 25 cm. 



2 This electromotive force is too small to be with certainty attributed to the 

 bending. 



