Potentiometer for Thermoelectric Work. 45 



be altered as required, instead of a new slide-wire being- 

 necessary. 



The blocks at the ends o£ the slide-wire, in addition to 

 beino- firmly screwed, were connected to the bus-bars by 

 mercury-cups and U-pieces, as it was not thought advisable 

 to trust to the screws alone for a contact whose resistance 

 must be negligible compared to '02 ohm. 



One great advantage of this form of instrument is that all 

 the coils could be arranged so as to have their resistance 

 adjusted, when in position in the box, by a method which could 

 be made to include such of the contacts as were actually in 

 circuit. The finest size of wire used will easily stand 1 ampere 

 continuously through it ; so adjustment even of the low coils 

 to one part in a hundred thousand is not difficult. 



The three sets of '01 and *! ohm were made each of one 

 continuous wire by selecting appropriate specimens of man- 

 ga nin, whose resistance from end to end was fairly uniform, 

 and soldering to these by silver solder at equal distances short 

 copper wires 1'5 mm. diam. 



The whole was then bent to approximately the shape 

 required, varnished with several coats of good shellac, and 

 annealed by repeated heatings for some time to about 140° C, 



The copper wires were then soft-soldered into place, 

 the rough adjustment of the coils to about 1 part in 1000 

 beino- previously made by an appropriate method. After a 

 week or two the fine adjustment was made by successive 

 scraping, the contacts being employed as in actual use, and 

 the whole being supported for this purpose at a convenient 

 height by a suitable spider-frame. 



When complete, two more coats of shellac were applied all 

 over the interior of the box to protect against any possible 

 inward leakage of mercury. 



It is not of great importance that the coils of a potentio- 

 meter of this kind should be of negligible temperature- 

 coefficient, so long as this is the same for all the different 

 sizes of ware employed ; but it is desirable that the series- 

 resistance, used to reduce the E.M.F. of the accumulator to 

 that of the standard cell, should be of the same material as the 

 potentiometer, so that the effect of temperature changes will 

 be compensated. Placing the standard cells and also the 

 accumulator in double-walled boxes packed with cork-clippings 

 materially adds to the convenience of working, obviating 

 the necessity for so frequently checking the compensating 

 current. 



Mercury reversing-keys in the circuit both of the battery 

 and ther mo junctions are a convenience, and where many 



