BEHAVIOR OF HARD STEEL ANNEALED IN MOLTEN LEAD (330°).

From an inspection of the foregoing families of curves it appeared probable, in view of this relatively high temperature, that annealing effects of great magnitude would occur during the first minutes of exposure. Conformably herewith, the rods were subjected to 330° during consecutive intervals of 1", 30m, and 1h, and examined at the end of each of these times. The rods selected were:

No. 37; diameter (2p)=0.0820 cm.

No. 38; diameter (2p)=0.0616 cm.

No. 39; diameter (2p)=0.0483 cm.

The following results were obtained:

TABLE 25.-Hard steel annealed at 330°.
[Rod No. 37. 2 p=0.0820.]

Digest. The results thus far given adequately exhibit the general physical character of the process of tempering. For the sake of clearness, and with a view of partially eliminating such discrepancies as are due to incidental errors, the three individual values of thermo-electric power and specific resistance for each of the temperatures of annealing will be combined and their mean chosen for discussion. We thus arrive at the following relations:

Deduction. If these functionalities are constructed graphically, as has been done in Fig. 10 (time as abscissa, thermo electric constant as ordi

FIG. 10.- Hard wires annealed continuously at 0°, 66", 100°, 1850, 330°, and 1,000°, respectively. nate), we obtain a family of typical curves, the general character of which is distinctly pronounced and may be thus expressed:

The degree of hardness retained by a glass-hard rod, after having been subjected to the operation of annealing, is dependent both on the temperature to which it has been exposed and on the interval of time. during which this exposure has taken place, in such a way that the effect of time, though of predominating importance in the case of small values of temperature, is more and more negligible in proportion as

these values increase. The operation is always most effective in its earlier stages, and this efficiency decreases very slowly where the temperatures are low-very rapidly, indeed almost suddenly, where they are high. If the action of any temperature be indefinitely prolonged, the rod under its influence ultimately reaches an inferior and limiting degree of hardness characteristic both of the temperature chosen and the type of steel under experiment. In other words: the annealing effect of any temperature increases gradually at a rate diminishing continuously through infinite time, very slowly in case of low temperature (<1000), with extreme rapidity in case of high temperature (>200°), so that the highest of the inferior states of hardness possible at any given temperature is approached asymptotically. The effect of the intensity of glass hardness (strain) of the originally hard rod is not readily discernible.

The considerations set forth readily suggest the terminology: "maximum of permanent hardness for the temperature t"-an expression which will be used throughout the sequel to designate the highest of the inferior states of hardness, persistent at the said temperature, t.

The relatively large effect produced by low temperatures (<100°), when a glass-hard rod is subjected to their influence for a long period of time, is deserving of special mention. It follows that a perceptible annealing effect must be attainable with temperatures much lower than the lowest above employed; indeed that the only inferior limit in this respect is probably the temperature of the water in which the rod was originally chilled,-while even the chilled rod, though kept at constant temperature, cannot be regarded as existing in a state of thorough molecular equilibrium until the lapse of a considerable interval of time after the hardening has taken place. Herefrom it appears that in the thermo-electric experiments the danger of destroying the uniformity of temper of a glass-hard rod by overheating the hot junction is greatly to be apprehended. Quick work and low temperatures are, therefore, to be preferred at an unavoidable sacrifice of accuracy. Nor is soldering of the ends permissible, except with the most extreme and intelligent caution. We desire to advert, in conclusion, to the important bearing of this unlooked-for sensitiveness of hard steel, even to low temperatures, on all other physical properties depending on the temper of this material-permanent magnetism for instance. It is obvious that the nature of a purely magnetic phenomenon can only be satisfactorily investigated when the material carrying the magnetic quality, throughout the course of the experiments undergoes no permanent change, otherwise we virtually commence our investigation with one rod and finish it with another, obtaining data which are not immediately, if at all, comparable.

(648)

ON THE EFFECT OF HIGHER AND OF LOWER TEMPERATURES ON THE TEMPER OF STEEL ORIGINALLY ANNEALED AT A GIVEN INTERMEDIATE TEMPERATURE.

Effect of lower temperatures.-The marked tendency of glass-hard rods to suffer diminution of hardness, even when exposed to temperatures but slightly above mean atmospheric temperature, naturally led to another important inquiry almost the converse of the preceding. We refer to the behavior of a rod annealed at a given temperature to the prolonged effect of lower temperatures subsequently applied.

For the purpose of obtaining experimental data a steel wire (No. 26, 2p=0.085), which had on a former occasion been annealed in an oil bath at 250°, was exposed to steam at 1000 for one hour. The results of the thermo-electric measurements made before and after this final exposure were these:

30

The temperatures t in both series are approximately identical. The electromotive force e may therefore be regarded as dependent on the difference of temperature T-t only, and with this presumption graphically represented. If, with the object of interpolating graphically, the two curves be carefully constructed, and the values of e for the temperatures of the one set of observations be derived from the curve for the other, then a comparison of corresponding values of e shows clearly that no annealing due to 1000 is appreciable. In the table (29) the interpolated electromotive forces are distinguished by parentheses.

30 For e=a(T-1)+b (T2—12)=(T—t) (a+b [T—t]+2 bt)=ƒ (T−t) if t is constant.

(649)