954 



Transactions of the American Institute. 



line is not very long, or the insulation is good, the resistance will fre- 

 quently be above the range of the apparatus. This may also happen 

 in testing for conductivity on a very defective wire. In this case 

 another method is adopted. 



25. First arrange the wires as before; then unplug 10,000 ohms 

 resistance, using galvanometer coil No. 1, and an intense main bat- 

 tery. Note the deflection obtained through the whole 10,000 ohms 

 resistance, and call this the maximum of the galvanometer. Now 

 turn the switch to the right, directing the current through the line, 

 which is open, of course, at the distant end. Note the deflection as 

 before. The tangents of the deflections will each be inversely pro- 

 portional to the resistance under which it was produced. 



Suppose the deflection with the 10,000 ohms to be 30°, giving tan- 

 gent .5774, while that through the line is 10°, whose tangent is .1763. 

 Therefore, 

 Tan. 5774 : 10,000 ohms : : tan. .1763 inversely : 32,751 ohms. 



i. e., 5774 x 10,000 



= 32,751 ohms. 



.1763 

 This is the insulation resistance of the line, and this, multiplied by the 

 number of miles in length, gives the insulation resistance per mile. 



26. It is proper here to caution those using this apparatus against 

 directing the current from an intense battery, through rheostat coils 

 of low resistance, lest they should be spoiled by burning. Forty 

 cells of Grove's battery would be likely to greatly damage a fifty ohm 

 coil, or perhaps one of a hundred or two hundred ohms, if the cur- 

 rent were directed through it alone. Batteries of no greater strength 

 should be employed than is necessary to accomplish the work desired 

 — ten or twenty cells of any sulphate of copper battery are sufficient 

 for measuring great resistances. 



27. The daily testings of a line should be recorded in a form some- 

 thing like the following : 



