D. W. Shea — Calibration of an Electrometer. 211 



Charge of Needle, 20 volts. 

 Temperature 17° 0. 



Resistance 



between point 



to ground and 



point to the 



quadrants in 



ohms. 



Charge of 



the 



quadrants in 



volts. 



Scale : 



readings in cm. 



Mean 



of 



two scale 



readings. 



Deflection 



be 



3 



of 



needle 



in degrees. 



1000- 



3-2 



1-19 



1-22 



1-205 



0°16'-5 



2000- 



6-4 



2-55 



2-55 



2-55 



35 



3000- 



9-6 



3-95 



393 



3-94 



53 



5000- 



16-0 



7-07 



7-04 



7-055 



1 36 



7000- 



22-4 



1050 



10-53 



10-515 



2 23-5 



8000- 



25-6 



1.2-44 



12-45 



12-445 



2 46-5 



900G- 



28-8 



14-42 



14-41 



14-415 



3 16 



10000- 



32-0 



16-61 



1663 



16-62 



3 44-5 



In making the calibrations, examinations for leakage were 

 frequently made by charging the quadrants and breaking their 

 connection with the battery circuit. The constancy of the 

 gravity and water batteries was determined by means of a con- 

 stant cell, devised by Dr. Willson.* The electro-motive force 

 of this cell, which was taken as the standard in these observa- 

 tions, is 1*085 to 1'088 volts. This variation is so small that it 

 is not observable with the electrometer, which is not capable 

 of measuring less than *01 of a volt, when set up in the manner 

 described, and the needle having a charge of 200 volts. 



It will be noticed that the variation in the sensibility decreases 

 as the charge of the needle becomes great relatively to that of 

 the quadrants. The mathematical theory supposes that the 

 charge of the needle is high when compared with that of the 

 quadrants, but it gives no idea of what the order of the charges 

 should be. These observations seem to show that in making 

 use of an electrometer for electrical measurements, we should 

 ascertain by experiment what charge the needle must have, in 

 order that the sensibility may remain constant for the range of 

 charges to be given the quadrants. 



Much trouble was experienced at first through the electrom- 

 eter being set up in a room in which several students were 

 at work upon various experiments in electricity. This 

 trouble seemed to be due to induction effects on the quadrants, 

 which the electrometer case did not very completely shield, for 

 on enclosing the electrometer in a box coated with tin foil, 

 and put into connection with the ground, the trouble was re- 

 moved. After removing another difficulty, i. e., leakage, due 

 to use of glass rods in the construction of a commutator, by 

 substituting paraffine for the glass, it was found that the elec- 

 trical zero suffered a displacement to the side, to which the 

 spot of light was deflected. f This displacement increased with 



* Uber Daniell'sche Normal-Elemente, Inaugural Dissertation von Robert W. 

 Willson, aus Cambridge, U. S. Wtirzburg, 1886. Pamphlet, 

 f Thomas Gray, Phil. Mag., V, vol. xxiii, p. 46. 



