A POCKET GALVANOMETEB. 



279 



This gives («) + (^) = (c) + {d) within the errors of experi- 

 ment, an arrangement which was accordingly adopted. We may notice 

 here, that if the determination of any one of them were in error of 1 

 in the unit figure, its uncompensated eifect would be 1/^M = 1/1250 

 of the whole. 



In other details of construction this form of apparatus is exactly 

 the same as the previous one, except that the jaw is side-wise instead 

 of vertical. This gives a greater precision to the instrument as 

 will he seen immediately, and still more facilitates the introduction 



of the coils of fine wire for 



Fig. 4. 



measuring large differences of 

 potential and small currents in 

 the way to be described fur- 

 ther on. The following sketch 

 shows one of the working 

 instruments. 



In calibrating the instru- 

 ment, it is best to use a small 

 current repeated many time 

 instead of a single strong cur- 

 rent which is so difficult to 

 keep steady. For this, 200 

 turns of a fine insulated Avire 

 were wound round in the form 

 of a circular ribbon having a 

 circumference of about 1 metre. 

 The wires were tied together 

 with a string so that the ex- 

 ternal appearance was like one 

 thick insulated wire of circular 



