HALL. — THERMO-ELECTRIC HETEROGENEITY IN ALLOYS. 555 



B 2 , etc. Going into details, and looking all the time for the causes 

 of descrepancies, we found that couple C 3 , like couple C 2 in the quartet of 

 couples previously studied, was the worst of its set, giving the lowest 

 electromotive force and showing the greatest difference between the 

 " parallel" and the " crossed " condition. 



As C 3 occupied in the pots the same place which C 2 had before 

 occupied, we naturally suspected that there might be something peculiar 

 in the C part of one or the other of the pots. Moreover, moisture was 

 sometimes seen in the upper part of the glass tube containing the ger- 

 man-silver of C 3 , which suggested that there might be a small leak of 

 steam into the tube containing one junction of this couple. Accordingly, 

 on September 25 we changed all the couples about, putting A 3 where 

 D 3 had been, and vice versa, and B 3 where C 3 had been, and vice versa. 

 The effects, such as they were, of this change may be inferred from the 

 following comparative set of numbers, those for September 23 being 

 obtained before, and those for September 25 after, the transposition just 

 described : 



A 3 (P-C) B 3 (P-C) C 3 (P-C) D 8 (P-C) 



Sept. 23 20.30 (+0.27) 20.17 (+o.i6) 20.02 (-0.19) 20.23 (-0.04) 



" 25 20.30 (+0.36) 20.14 (+0.36) 20.08 (+0.52) 20.14 (-0.25) 



The deflections here given for the two days are reduced to like con- 

 ditions of galvanometer sensitiveness and temperature interval, so that 

 they are strictly comparable. Under (p — c) is given the "parallel" 

 deflection minus the " crossed" deflection. The mean deflection is 20.180 

 for September 23 and 20.165 for September 25, an insignificant differ- 

 ence. Couple C 3 gives the smallest deflection on each day; its (p — c) 

 was larger on the second day than on the first, and of the opposite sign, 

 but its (p — c) for September 23 was exceptionally small and unusual in 

 sign. Couple B 3 , which had been put into the place of C 3 , showed no 

 change that appeared very significant. 



On September 26, also, the couples were used in their interchanged 

 positions, but with an interval of temperature different from that used 

 September 23 and 25, and on this day also C 3 gave as usual a deflection 

 about £ per cent below the mean given by the four couples, while B 3 

 gave, as usual, a deflection very near this mean. 



On the whole, then, it seems probable that the low electromotive 

 force of the two C couples was an accidental coincidence, not due 

 in any discoverable measure to their position in the apparatus. Never- 



