HALL. THERMAL AND ELECTRICAL EFFECTS IN SOFT IRON. 53 



Observations on the ratio g c -f- g h were made on four days, with the 

 following results : 



The marked difference between the first two and the last two values 

 found for g c 4- g h is due to the reversal of the direction of How, there 

 being, no doubt, small differences in the placing of the thermo-electric 

 couples at the ends of the bars. See p. 27. The reversal should elimi- 

 nate error due to this source. 



The value 87.3°, found for t h on February 7, was obtained from obser- 

 vations made with thermo-electric couple H, the outer junction of this 

 couple being placed, on this day, in oil heated to a temperature not far 

 from that of the inner junction, so that uncertainty of calibration of the 

 couple had little effect on the result. The other values of t h , which 

 are given in brackets, were estimated from this value, with use of the 

 barometer observations of the various days as an indication of the various 

 temperatures of the boiling water. 



Estimation of the loss of heat by lateral flow required observations on 

 the difference of temperature between points on the main bars, with no 

 electric current on, and corresponding points on the guard-ring bars. 

 Such observations were obstructed, and in some measure restricted, by 

 the breaking of one of the thermo-electric wires and by the inadequate 

 length of several others, the full scope of their possible use not having 

 been foreseen when the couples were made. The results were as 

 follows : 



On a and £. On guard-ring. 



Feb. 1 13.0° [E-H.] 

 " 4 49.3° [/] 48.6° [Z] 47.8° [Z] 87.1° [A''] 



" 7 12.0°[<7] 48.7° [/] 87.3° [//] 11.3° [X] 47.4° \z\ 48.0° [Z] 86.7° [A*] 



" 9 12 7°[G] 1 2.1° [X'] 



12.0° 49.0° 87.3° 11.7° 48.0° 8i5.9° 



Taking account now of the fact that the junctions X and X were 'not 

 placed just opposite H and G (see p. 50), and estimating for points 

 which are opposite //and G, we get 



