408 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 11, NO. 17 



temperature constant within =t 0.1°C.for hours without any atten- 

 tion. At higher temperatures the furnace cools slowly, 2 or 3° per 

 hour at 1450°. This cooling, which is perfectly regular, is undoubt- 

 edly due to slow changes in the furnace for the resistance of the heat- 

 ing coil is held constant by the regulator. The cooling can be com- 

 pensated for by an occasional manual adjustment of the slide wire 

 contact on the bridge. 



1070 



1060 



1600 jD. 



13 



s_ 



0) 



a 

 £ 



0) 



1050 



i ii » » » >• » 



Time 



-^0 



Fig. 3. 



Curves to illustrate the performance of the regulator in the steady heating of the 

 furnace to determine a melting point. 



In a particular experiment, a platinum wound furnace of the type 

 described was being maintained by the regulator at 1269°. Its re- 

 sistance at this temperature was 9.9 ohms. The galvanometer 

 contacts were set to close at ± 5 millivolts, corresponding to a change 

 of about ± 1° in the temperature of the heating wire. With the 

 rheostat adjusted so that the times of heating and cooling were ap- 

 proximately equal, and the current 10.0 and 7 . 3 amperes respectively, 



