G58 PROCEEDINGS OP THE AMERICAN ACADEMY. 



Electrical Apparatus. 



In determining the strength of the electric field, since the double 

 refraction varies as the square of the field-strength, it is necessary to 

 measure the difference of potential with a percentage error less than 

 one half that of the optical measurements. With differences of potential 

 as high as ten thousand volts and higher, the readings of electrometers 

 depend to an uncertain extent on the condition of the air, especially on 

 the humidity, and it seemed preferable to use a high-resistance poten- 

 tiometer and Clarks' standard cell. 



Potentiometer. — To extend the range of the Crompton potentiometer 

 to fifteen thousand volts, and still permit readings to an accuracy of one 

 tenth of one per cent, required one million ohms of extension coils with 

 a negligible temperature coefficient and a carrying capacity of fifteen 

 milliamperes continuously. 



Special Resistance. — The resistance was therefore built up of ninety 

 coils, each of ten thousand ohms, wound with a single strand of number 

 37 (B. & S.) double silk-covered, paraffined special resistance wire, on 

 seasoned maple spools. Single winding secures the least possible differ- 

 ence of potential between adjacent turns, and successive layers are sepa- 

 rated by paraffined paper. The spools before winding were tested, and 

 none found to have less resistance between opposite sides of the core than 

 100 megohms. The wound spools were gradually heated in an electric 

 oven to 140° C, and maintained so for about three hours, when they 

 were immediately immersed in melted paraffine also at 140°, and kept 

 hot for three hours more. The spools finally ceased to give off air 

 bubbles, and they were left in the bath while it was gradually allowed 

 to solidify, so that as the cooling paraffine in the spools shrank, air 

 might not penetrate. The bath was then barely melted and the spools 

 withdrawn and dipped once more. In this way dry coils, well insulated 

 and well protected from atmospheric influences, were obtained. 



They were mounted in groups of nine on bright hard-rubber and the 

 terminals of each coil were soldered without the use of acid to brass 

 binding posts on hard-rubber strips on opposite sides of the ventilated 

 containing boxes. 



The resistance of each coil thus mounted can be measured with a 

 standard bridge. The resistance of the sum can be taken as the sum of 

 these resistances, since the coils and their terminals are so highly 

 insulated from one another. 



But since the fall of potential may reach as much as one hundred and 



