46 BRUSH— KINETIC THEORY OF GRAVITATION. 



ference between the gravitation field of a comparatively large mass 

 of the metal bismuth, which is the most diamagnetic substance 

 known, and the gravitation field of a similar mass of lead and of 

 zinc, which are very much less diamagnetic than bismuth, and also 

 of tin which is slightly paramagnetic. To this end it was proposed 

 to measure the minute gravitational attraction between each of the 

 above masses and a very much smaller nearby mass of some metal, 

 the same small attracted mass to be used in all cases. In such a 

 scheme the large masses would do nearly all the attracting, and their 

 several gravitational pulls per unit of mass would be comparable. 



To carry out this scheme Professor Dayton C. Miller very kindly 

 provided, from his large collection of physical apparatus, a beautiful 

 instrument designed for class-room demonstration of gravitational 

 attraction between two small silver balls and two large lead spheres 

 in the usual manner of such apparatus. It is a modification of the 

 apparatus designed and used by Professor C. V. Boys for determin- 

 ing the gravitation constant and the mean density of the earth. 

 Each small silver ball weighs three fourths of a gramme, and the 

 pair are mounted at the ends of a horizontal small straight metal 

 rod, with their centers 2-^ cm. apart. Rising from the center of 

 this connecting rod is a small vertical rod carrying, at a distance of 

 6 cm. above the silver balls, the usual small mirror for scale reading, 

 set at an angle of 45° with the ball-carrying rod. These parts con- 

 stitute the oscillating system, and are suspended by a long quartz 

 filament in a brass tube, the balls only projecting below the tube 

 into a narrow glass-walled chamber, made shallow in order to mini- 

 mize convection currents inside. ]\Ieans are provided for leveling 

 the whole apparatus, for orienting the free-hanging system and for 

 clamping the balls when not in use. The apparatus is permanently 

 grounded through one of its leveling screws. 



The large lead spheres and their carriages were discarded and 

 replaced by a light reversible wooden carriage. 



The photograph, Plate V., shows the apparatus as set up in my 

 basement laboratory. The delicate part first described is mounted on 

 a heavy marble slab firmly bracketed in the angle of two twenty-inch 

 brick walls. These are inside walls, and hence not liable to sudden 



