372 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



June 20. Plate 3. 



The following set differs from all the other measurements in this paper 

 in the method of reading the galvanometer. Instead of attempting to 

 pick out times when the galvanometer needle was not affected by out- 

 side disturbances, the experimenter determined each position of the 

 index by taking the mean of several (usually six) instantaneous obser- 

 vations made at regular intervals, about 8 seconds, as measured by the 

 metronome. Five such determinations of the position of the index 

 were made in a single test, three with the magnetic field in one direc- 

 tion and two with the field in the opposite direction. The following 

 results were obtained. 



JoNB 23. Plate 3. 



