THERMO-ELECTRIC QUALITY UNDER PRESSURE. 335 



was drawn from 0.06 to 0.03 inches diameter, and annealed to redness 

 after the final drawing. The commercial wire must have been of 

 unusual purit}-, because the average temperature coefficient of re- 

 sistance at atmospheric pressure between 0° and 100° had the high 

 value 0.00596, and its resistance shows the same striking departure 

 from linearity with temperature as does the pure iron. 



At atmospheric pressure the thermal electric behavior of these 

 various specimens against lead is given by the following formulas. 

 Ingot iron, annealed, 



E = (16.18 /-0.0089^2_0.000086i3) X 10"^ ^.^i^s, 



p = (16.18 -0.0178 < -0.000258^2) (^ _|_ 273) x lO'^ volts, 



a = (-0.0178-0.000516 (t + 273) X lO'^ volts/°C. 



Ingot iron, hard drawn, 



E = (15.92 ^-0.0106 ^—0.000056 f^*) X 10"^ volts, 



P = (15.92 -0.0212/ -0.000168/2) (j. ^ 273) X lO'^ volts, 



a = (-0.0212-0.000336 (t + 273) X IQ-^ volts/°C. 



Commercial iron, annealed, 



E = (16.56 f-0.0033f— 0.000122/3) ^ lO^^ volts, 



P = (16.56 -0.0066/ -0.000366/2) (/ + 273) X lO'^ volts, 



a = (-0.0066-0.000732/) (/ + 273) X lO'^ volts/°C. 



The behavior of these different specimens is therefore not unlike. 



Under pressure, each specimen was subjected to the same treat- 

 ment. They were seasoned twice by applications of 12000 kg. at 

 room temperature, and runs were then made as usual at 25°, 50°, 75°, 

 and 95°. The behavior of all was qualitatively the same, with numer- 

 ical differences. A longer time than normal was always required to 

 reach steady readings, and there were rather large irregularities, 

 suggesting that all three samples were in a state of incomplete internal 

 equilibrium. This is what would be expected from the behavior of 

 iron in other particulars. 



Of course the greatest interest attaches to the annealed ingot iron, 

 this approaching most closely to pure iron in a state of complete ease. 

 Even this shows much hysteresis, which increases greatly at the higher 

 temperatures. At 25°, the e.m.f. is negative throughout, and is small. 

 For the first 4000 kg. the curve is exceedingly flat, and hugs the axis 

 closely. The curvature is downward, the negative effect increasing 

 numerically by larger fractions of itself for equal pressure increments 

 at the higher pressures. At 25° the hysteresis is one eighth of the 



