186 Trans. Acad. Set. of St. Louis. 



enclosed in a glass case containing drying material, and 

 it was driven by a single-phase motor. Placing the large 

 masses in the position shown in Fig. 3 at the time when 

 the work to be described began, increased the scale read- 

 ing by 0.40 cm. 



An illustration of the results obtained is shown in 

 Plate XLV. In each case the masses had been grounded 

 during the preceding night. The scale was displaced in 

 order that the two diagrams might be shown on the same 

 plate, without interference. On the lower diagram the 

 arrangement of apparatus was as shown in Fig. 3. The 

 hour of the day is laid off upon the horizontal axis. The 

 scale reading in cm. is along the vertical axis. One mm. 

 represents an angle of 2.6 minutes of arc. The negative 

 terminal was applied at 9 :38 a. m. On the diagram the 

 arrow indicates the time. At 10:12 a. m. the terminals 

 were reversed, an operation which required a few sec- 

 onds of time. At 11:30 the machine was disconnected. 

 There was evidence shown in the drop in the reading at 

 11 :20 that the reading would begin to decrease. The up- 

 per diagram of this plate shows results obtained on May 

 4. Here the positive terminal was first applied at 9:35 

 a. m. The terminals were reversed at 10 :24 and the ma- 

 chine was stopped at 11:20, the readings being con- 

 tinued to 12 :05 p. m. In this work the conductors upon 

 which the 800 pins were supported were directly con- 

 nected with the large masses and shield. 



When direct contact was made between the pin-con- 

 ductors and the large masses, the changes took place more 

 rapidly than when the air surrounding the masses was 

 supercharged with negative corpuscles emitted from the 

 pin points, or when the reverse action took place, this 

 alone being depended upon to change the potential of the 

 masses. The most interesting feature of this work is the 

 complete elimination of the possibility that the apparent 

 decrease in the attraction between these masses was due 

 to the convection currents of air resulting from heat ef- 

 fects. Eeversing the terminals would not reverse these 

 heat effects. 



