124 BRIDGMAN, 



percentage decrease of resistance was 6.68% at 6000 kg., and 12.07% 

 at 12000 kg. At 96°, up to 3000 kg., the coefficient is about 4% 

 greater. The resistance decreases Hnearly with temperature between 

 0° and 100°, the total decrease for 100° being 2.56% of the resistance 

 at 0°. The readings on this carbon were entirely regular, showed little 

 difference between ascending and descending values, and an almost 

 perfect recovery of the zero. 



The first specimen of Acheson graphite was cut from a rod of 1.5 

 inches diameter to about the same dimensions as the gas carbon. 

 Measurements were made by the same method. Three complete 

 series of readings were made, at 0°, 51°, and 97.4°. The readings were 

 not so regular as with the gas carbon, showing large hysteresis effects, 

 rising at the maximum to 12% of the total pressure effect, and there 

 were also parasitic e.m.f.'s so large as to necessitate a special arrange- 

 ment of the constants of the circuit. The sign of the eft'ect is positive, 

 the reverse of what it is for gas carbon, and there is very considerable 

 departure from linearity with pressure, the coefficient becoming 

 numerically less at the higher pressures. The total fractional increase 

 of resistance under 12000 kg. decreases with rising temperature, being 

 4.75%o at 0°, 4.23% at 50°, and 4.23%o also at 100°. The large 

 departure from linearity may be judged from the fact that at 6000 kg. 

 at 0° the increase of resistance is 3.44%, which is 72% of the increase 

 under 12000 kg. The resistance decreases with rising temperature 

 at atmospheric pressure, and the change is not linear. At 0° the 

 resistance on an arbitrary scale is 1.0000, at 50° 0.9135, and at 100°^ 

 0.8687. 



The second specimen of graphite, supposed to be especially pure, 

 was subjected to a special preliminary seasoning in order to eliminate 

 as far as possible the pores. It was sealed into a lead tube and 

 subjected to a fluid pressure of 12000 kg. on the outside of the tube. 

 The diameter of the specimen was reduced by about 2%, but there 

 were a great many pores still visible to the naked eye. It w^as further 

 seasoned by heating to 125° in vacuum; this treatment should have 

 removed all moisture, of which no traces, however, were evident. 

 This specimen was cut to the same dimensions, and measurements 

 made in the same way as on the two preceding samples. Only one 

 run was made, at 50°. This was terminated b}^ an explosion at 12000 

 kg. In view of the unreproducible character of the results it did 

 not seem worth while to repeat the effort to obtain a complete set of 

 readings. As with the other sample of x\cheson graphite, the effect 

 of pressure is to increase the resistance, but the change was much less 



