Nipher — Gravitation and Electrical Action. 173 



The large masses were then wholly enclosed in two 

 metal boxes, from which the masses were insulated. The 

 masses rested upon rubber blocks placed upon the bot- 

 tom of the boxes, and each box was mounted upon rubber 

 blocks which rested upon plates of glass. The masses 

 were in the same position as in the work above described. 

 The boxes were separated from the shield, enclosing the 

 suspended masses, by the plates of glass and the asbes- 

 tos paper. 



The two metal boxes and the metal shield were then 

 electrified precisely as has been done before. It was then 

 found impossible to increase the arc of vibration of the 

 suspended masses, although the operation was continued 

 for several hours. If the former results had been due to 

 heat effects, they should have produced the same results 

 in this case. The large masses were removed from the 

 enclosing boxes. The empty boxes and screen were then 

 electrified as before. 



No change in the amplitude of vibration could be pro- 

 duced. 



These results seem to indicate clearly that gravita- 

 tional attraction between masses of matter depends upon 

 their electrical potential due to electrical charges upon 

 them. To assume a special case, such as exists when the 

 gravitation constant is being determined, Newton's law 

 holds only when the common potential of the two bodies 

 is such that their gravitational attraction for each other 

 is a maximum. If the two masses are not separated by 

 a metal shield and their common potential is that of the 

 earth, these masses will repel each other by a force rep- 

 resented by the final term in Eq. (1) and the value of K 

 will also be diminished, if the above conclusions are cor- 

 rect. If the two masses are separated by a metal shield, 

 this final term will be eliminated, but the value of K will 

 be diminished and may seem to be variable, if there are 

 local variations in the potential of the earth. By adjust- 

 ing the potential of the attracting masses by well known 

 means we may hope that the real value of K and the 



