132 BRIDGMAN. 



Historical Survey. 



A few attempts were made to detect the existence of the effect 

 before Maxwell, but we may pass these over as not approaching in 

 sensitiveness the method of Maxwell.^ Maxwell was the chairman 

 of a committee, the other members of which were J. ^. Everett and 

 A. Schuster, appointed by the British Association in 1876, to investi- 

 gate the accuracy of Ohm's law. Two experimental methods were 

 used, both apparently proposed by Maxwell; the experiments were 

 performed by Chry^al. 



The first was a substitution method, by which the resistances of 

 five similar coils were compared in various combinations of two in 

 parallel and two in series against a single one. The current density 

 in the single coil was thus douWe that in the multiple arrangement, 

 and if there is an effect of the kind sought, the resistances should be 

 different. A small positive result was found, but was ascribed to 

 errors in view of the fact that the second method, which was much 

 more delicate, gave negative results. 



The current densities employed in the first method were so low as 

 not to cause appreciable heating of the wires. The second method 

 was such that currents large enough to heat the wires to incandescence 

 could be used. A heavy and weaker current were passed alternately 

 in rapid succession through a fine wire, which was made one of the 

 arms of an ordinary bridge. The period of alternation was so high 

 that there were no appreciable cooling effects. Observations were 

 made with a galvanometer of period long compared with that of the 

 alternations. The apparent resistance was read first with the large 

 and the small currents flowing in the same direction, and then with the 

 small current reversed. Let us suppose that there is an effect of the 

 kind sought, which means that the resistance for the large and the 

 small current is not the same. If the galvanometer indicates balance, 

 it must be because it is really off balance for both currents, to the one 

 side for the large current, and to the other side for the small current. 

 If now the small current is reversed the galvanometer will be off balance 

 to the same side for both currents, and there will be a steady deflection. 

 Hence if there is an effect, the balance will be altered by changing the 

 direction of the small current. There were difficulties in the method. 

 The period of alternation had to be chosen as high as 60 per second in 

 order to avoid appreciable cooling effects in the short intervals of time 

 when the small current replaces the larger one. There were consid- 

 erable mechanical difficulties in designing an alternator of the requisite 



