Nipher—Gravitational Repulsion. 183 
ciably change the result. It was finally decided to check 
the results above described by placing the large masses 
on alternate sides of the suspended masses and obsery- 
ing the deflection due to gravitational attraction. 
In order to maintain equality of stresses in the two 
threads, which had been hung around a hook attached to 
the torsion head, they were hung around a pulley, having 
a diameter of one inch. Below were two adjustable 
pulleys. The threads approached each other around 
these two pulleys, and their distance apart could be so 
varied as to make the threads parallel within the long 
suspension cylinder. The metal shield before described 
was replaced by one having equal width from end to end. 
The top, bottom and ends were of wood % inch in thick- 
ness. A sketch of this wood frame is shown in vertical 
section in Fig. 1, 
Fig. 1. 
This wood frame was varnished with shellac and all 
points of contact of its parts were closed. The sides of 
this frame were each closed by two layers of heavy card- 
board outside of which was a sheet of flexible tinned iron. 
They were clamped to the wooden frame by means of 
wood bars screwed to the bottom, top and ends. The 
edges of the cardboard were then sealed by means of 
beeswax applied by means of a hot iron. The entire de- 
vice as thus described was then enclosed by a metal shield. 
