Nipher—Gravitational Repulsion. 187 
The possibility that one terminal of the machine pro- 
duced greater heat effects than the other, the deflecting 
effect being decreased when the terminal producing the 
lesser heating effect was applied is also eliminated. It 
matters not which terminal is first applied. The result 
is the same, and has been obtained many times. 
During the afternoon of May 4 the operation repre- 
sented in the upper curve of Plate XLV was repeated, 
the positive terminal being first applied. <A result of pre- 
cisely the same kind was obtained. 
The upper curve of Plate XLV also means that the gra- 
vitational attraction between the masses at 9:35 a. m. had 
been decreased by about 110% at 10:10 a. m. Gravita- 
tional attraction had been decreased to zero, and had then 
been converted into a repulsion. An hour later it had 
regained its initial value. On the afternoon of June 1, at 
2:30 p. m., it was decided to change the electrical condi- 
tion of the suspended masses. One of the end caps form- 
ing the outer metal shield of Fig. 2 was removed. A hole 
which had been bored through the wood frame, and which 
had been closed by a rubber stopper, was opened. A glass 
tube was passed through the opening, about two centi- 
meters beyond the inner surface of the wood frame. A 
copper wire to the end of which the head of a pin had 
been soldered, was passed through the tube, the point of 
the pin projecting slightly beyond the end of the tube. 
The wire was about six inches in length. The apparatus 
was arranged as shown in Fig. 3. The positive terminal 
of the machine was used in draining negative corpuscles 
from the air within the shield around the suspended 
masses. A slow change of 1.2 em. or 12 scale divisions 
in the scale reading was produced. The terminal was 
disconnected when this deflection had been produced, and 
the machine was stopped. The tube was removed, the 
end cap of metal was replaced and the large masses and 
shield and the pin-point conductors were grounded until 
the next day. The mean reading had not been appre- 
Ciably changed. 
