[bronson] construction OF VERY HIGH RESISTANCES 16S 



resistances larger than 3 megohms. However, he states that resist- 

 ances of 40,000 megohms could be made from a film 4 meters long 

 and 0.2 mm. wide, and he assumes that it would behave just as satis- 

 factorily as the ones with comparatively small resistance. This is 

 probably not the case, as we shall see later. 



The writer spent considerable time trying to make a satisfactory 

 high resistance by means of carbon lines on ebonite or ground glass, 

 as well as out of various liquids in capillary tubes. It was found 

 possible in nearly all cases to get resistances that were large enough 

 (about 10^^ ohms), but there was always considerable variation 

 and uncertainty. The carbon line seemed to have its entire resistance 

 at one point and to be merely a case of bad contact, which was bound 

 to be very uncertain and subject to all kinds of external condi- 

 tions. The liquid resistances have very large temperature coeffi- 

 cients, and seem to be subject to a variable polarization. It is 

 possible that this variation was not due to polarization at all, but to 

 the electrical absorption caused by the relatively large amount of 

 insulating material, which was necessarily present. The fact that 

 the carbon resistances exhibited the same effect to a less extent 

 would also point to this conclusion. 



The effect of this electrical absorption is quite noticeable in most 

 condensers and is very marked in the case of insulation testing. It 

 decreases with the time, but may continue for some hours. It is 

 evident that this absorption will be relatively greater, the greater 

 the resistance, that is, the smaller the conduction current. It will 

 also be increased by an increase in the amount of the insulating 

 material in contact with the conductor. It is therefore not at all 

 certain that a satisfactory resistance of 40,000 megohms could be 

 made as suggested by Stewart, for, as we have seen above, we should 

 expect that the disturbance due to absorption would be from 10^ to 10^° 

 times as great in a resistance of 40,000 megohms as in a resistance of 

 3 megohms. This difficulty seems to be inherent in the very nature 

 of all this class of resistances. In order to avoid it, it would be neces- 

 sary to have a conductor of large cross section and short length, so 

 that a relatively small amount of surface would be in contact with 

 the insulating material. 



This condition has been fulfilled in a large measure by the use of 

 an ionized gas as a high resistance standard. Professor Rutherford 

 suggested the possibility of doing this, as he had previously shown that 

 the current, through a gas, subject to a constant source of ionization, 

 was approximately proportional to the difference in potential, as long 

 as this difference was small. The writer has been using this principle 

 successfully for the piist three years in making radio-active measure- 



