96 Rood — Experiments on High Electrical Resistance. 



more, according to the dryness of the atmosphere. Various 

 expedients were tried to measure and allow for this loss, but it 

 was finally decided to compare together only resistances that 

 were nearly equal, which is not a matter of difficulty after one 

 has accumulated a stock of resistances, those which had been 

 accounted failures now becoming useful. It is hardly neces- 

 sary to add that if a unit, of about a given resistance, is 

 desired, it is best, after due consideration of the case, to make 

 five or six. 



Manipulation. — One pole of the battery is connected with a 

 contact maker which must be carefully insulated ; from thence 

 the current flows to the unit or resistance, and finally to the 

 binding post, B. The electrometer having been discharged 

 and the gold leaf set free, the wire R is allowed to remain for 

 ten seconds in contact with post B, the current meanwhile 

 flowing through the unit or resistance into the ground. As R 

 and Y are turned to the left the time is noted with a stop 

 watch, and the leaf watched with a lens for the stroke, after 

 which contact is immediately broken. An attempt is always 

 made to arrange matters so that the time noted shall not be 

 less than fifty nor more than eighty seconds, and in the case of 

 two resistances which are being compared together, care is 

 taken to make their times as nearly equal as possible, so as to 

 avoid errors due to leakage. This is accomplished by the use 

 of small subsidiary units of known resistance, which are added 

 to either of the resistances under consideration as occasion may 

 require. When a string of units is being compared with 

 another string, the actual number of pieces in both cases 

 should be the same, any deficiency being made up by the use 

 of dummies which in all respects imitate real units, except 

 that they have no resistance at all. 



The second pole of the battery and the plate P are perma- 

 nently connected with ground. The striking distance usually 

 employed has been one millimeter throughout whole sets of 

 experiments. Striking distances as large as seven millimeters 

 can be used for regulating the time of the stroke, but it is 

 better to manage this with the aid of condensers. In any case, 

 the striking distance and the electromotive force of the battery 

 should remain constant in a set of experiments, so that com- 

 parable results can be directly obtained without the use of 

 coefficients. Not only can the units be used "in series," but 

 also "in parallel," to measure smaller units. 



Building iip a set of resistances. — Mr. H. C. Parker meas- 

 ured for me a resistance which turned out to be 32,000 

 megohms, the electromotive force being 50 volts, and this was 

 used as a base in building up a set which reached as high 

 as fourteen million megohms. The plan for building up 



