66 BRIDGMAN. 



The effect of parasitic electromotive forces in the circuit was 

 eliminated by taking the mean of readings with direct and reversed 

 potentiometer current. A reversing switch was supplied for this 

 purpose, as also in the circuit of the balancing variable e.m.f. The 

 only reciuirement in the potentiometer current is that it shall remain 

 constant during the substitution of the pressure coil for the slider. 

 The substitution was made rapidly by a double throw switch, so that 

 an ordinary dry cell was entirely good enough as the source of the 

 potentiometer current. 



The function of the resistance R^ (Figure 1) was merely to provide 

 an additional adjustment by which the deflection of the galvanometer 

 might be easily made null. An ordinary decade box was good enough 

 for this. 



The usual procedure was as follows. A preliminary examination 

 indicated the order of the pressure efi'ect. i?3 was then so chosen that 

 the motion of the slider would be of the order of 40 cm. for the entire 

 pressure range, the total length of the slide wire being 60 cm. jRi and i?2 

 were then so chosen, depending on the actual resistance of the pressure 

 coil, that the gahanometer deflection was unaltered on substiiuting 

 the pressure coil for the slide wire. The following coils for i?^ and R3 

 usually gave sufficient range of choice; two 0.5 ohm coils, two I's, 

 and one each of 2, 3, 4, 10, 20, and 30 ohms. For i?2 a plug box from 

 0.1 to 1000 ohms was used. R^ was then set at about 2000 ohms, or 

 less if the parasitic e.m.f.'s. were troublesome, and the balancing e.m.f. 

 adjusted to gi"\'e no deflection. These adjustments were preliminaiy 

 to the application of pressure. The adjustments after each change of 

 pressure were, first, an adjustment of R^ to give again approximately 

 null deflection when the pressure coil is in circuit, second, setting of 

 the slider to give no change in deflection on substitution, third, re- 

 versal of the potentiometer current and the balancing e.m.f. and re- 

 setting of the slider for no change of deflection, and fourth, change of 

 the potentiometer current and the balancing e.m.f. back to their 

 original directions and again setting the slider. By taking an odd 

 number of readings the efl'ect of any slow change of the parasitic e.m.f. 

 due to dissipation of the heat of compression was eliminated. The 

 potentiometer current was usually so chosen that the difference of 

 slider setting for the two directions of the current occasioned by the 

 parasitic e.m.f. was less than 1 cm. 



The coils were all compared with standards and proper corrections 

 applied. The slide wire was calibrated and corrections applied for 

 lack of uniformity. In general, all the precautions of manipulation 



