■ 



Protective or Circuit Breaking Device. 273 



lower end of the stem with cement. This gave a firm suspen- 

 sion and prevented short-circuiting. The top of the tube was 

 closed with cement, care being taken not to make it air-tight. 

 The platinum wires were soldered to heavy brass bars just out- 

 side the tube. 



A furnace was built especially for the purpose of heating the 

 chromate. The construction is very similar to that described 

 in the Bulletin of the Bureau of Standards (vol. vi, No. 2, 

 p. 160, 1909). In building the furnace "Nichrome" ribbon 

 was used instead of platinum. The method of measuring the 

 resistance of the chromate was that of comparison." 



In the figure, G, is a Leeds & Northrup type K galvanometer 

 which was used with a telescope and scale at a meter's distance. 

 The sensibility of this instrument is 446 megohms with a resist- 

 ance of 120 ohms. An Ayrton shunt was used in connection 

 with it, as seen in the figure. S, is the variable standard with 

 which the resistance of the chromate was compared. 



The constants of the thermometer were determined and are 

 recorded with Table I. The ordinary formula for calculating 

 the temperature from the change of resistance and tempera- 

 ture coefficient was used. 



It was noted that when the chromate was heated to a high 

 temperature, it had a higher resistance after cooling than it 

 had before heating it. On a second heating to a temperature 

 considerably lower than the first, it showed little or no vari- 

 ation after cooling. At room temperature it was very stable. 



The curve given in fig. 2 was obtained from a piece that had 

 been heated twice for several hours with the range of temper- 

 ature used for the curve. Through this range the chromate 

 has a very large negative temperature coefficient. The points 

 on the curve are somewhat irregular. It was impossible to 

 determine the causes of all the variations which were observed 

 for the whole range of temperature. 



When the switch was on A and B with L open and the chro- 

 mate at a temperature of about 350° centigrade, the galvanom- 

 eter mirror swung to the extreme right of its zero position ; at 

 about 400° to the extreme left ; at 475° to the extreme right 

 again, and so on until at 800° there had taken place six distinct 

 reversals. Some of these movements, especially at high tem- 

 peratures, were quite violent and off the scale. On cooling, 

 the reversals were repeated in reverse order at about the same 

 temperatures. 



The curve given in figure 2 was obtained by impressing an 

 extra E.M.F. on the shunt marked sh in figure I. With this 

 E.M.F. balanced with the unknown E.M.F. in the circuit, a zero 

 position of the galvanometer was obtained. L was now closed 

 and a reading taken. 



