﻿effect of Temper, in case of Steel. 279 



distinction might be more delicate than the estimation of pre- 

 cipitated carbon. To test this inference we selected rods ISTos. 

 1 to 12, quenched uniformly glass-hard by our method. These 

 were then broken in the middle and the first halves of each 

 rod left in the glass-hard state ; the other halves were annealed 

 in pairs at 20°, 100° 4 h , 185°, 360° 450°, 1000°, respectively. 

 In order to anneal these (long) rods uniformly, we used a spe- 

 cial device by which they were drawn vertically upward, 

 through a zone of constant temperature, by clockwork. If h 

 be the height of this zone and p the radius of the disk or drum 



revolving once per hour, then t=- is the time of annealing 



in hours. Again, if we make A=2jt, then the time of exposure 

 in hours is the reciprocal of the radius of the disc in centime- 

 ters. This method of annealing is exceedingly convenient for 

 long geometrical figures, inasmuch as it requires only a very 

 narrow cylindrical zone of constant temperature.* It is there- 

 fore applicable at all temperatures, and almost invaluable for 

 high temperature work (500° to 1000°) where sharp distinc- 

 tions of time are unnecessary and zones of definite constant 

 temperature are not easily produced. In this way we obtained 

 six pairs of hydro-electric couples, each of which consisted of 

 glass-hard steel and the same steel annealed at one of the tem- 

 peratures specified. Our first results were investigated with a 

 zero method. They showed clearly that annealed steel is hydro- 

 electrically positive with reference to hard steel, and that elec- 

 tromotive force increases with difference of temper. But in 

 view of the large polarization discrepancies incident to these 

 measurements, the electrometer is preferable to the zero instru- 

 ment. 



The following data were obtained with Mascart's apparatus 

 adjusted to indicate O'OOl volt, accurately. The electrodes of 

 the steel couples were immersed in a concentrated solution of 

 zinc sulphate (pure), contained in a U-tube, the two limbs of 

 the tube receiving the two steel wires. We kept them scrupu- 

 lously bright by repeated scouring with sand paper. In the 

 tables we give the electromotive forces e of the divers couples 

 of hard and tempered steel, as well as the probable mean error 

 of each. The means of the two values of e for each tempera- 

 ture of annealing are given in the second horizontal row and 

 fairly exhibit the hydro-electric effect of temper in question. 



The following data are the mean results of four series of 

 measurements of five observations per series. The rods were 

 scoured before beginning the first and the third of these series. 



* 27r x 7r(0"3)cm , or even a narrower cylindrical figure being sufficient. We 

 have in view a similar method of quenching by which long continuous steel wires 

 are drawn rapidly through a zone of high temperature into water. 



Am. Jour. Sol— Third Series, Vol. XXXII, No. 190.— October, 1886. 

 18 



