upon in this chapter, may still, I apprehend, be considered as new,-so far, at least, as regards the determination of the actual and comparative influence of the various degrees of hardness or qualities of temper, by which the more important properties of magnets may be influenced. Much, indeed, of this portion of our subject, has been partially involved in the investigations in Part I.—“ on the ratio of the power of magnetised steel bars, as modified by difference of tempering;" and many of the general results might, by analogy, have been anticipated. But still, the investigations on the actual effects of tempering and hardness, on combinations of steel plates and bars, will not, I hope, be found unimportant, especially as affording additional information of much practical consideration in the construction and improvement of magnetical instruments. A summary of, or selection from, the investigations made on this subject-with bars or plates in several distinct conditions, as to temper —will afford all the information, as yet obtained, which has appeared to me to be of sufficient importance to be recorded. 1. Of the relative powers of combinations of magnetised steel plates, as affected by changes in the temper of the same series of plates. Captain Kater, in his investigations for his Bakerian lecture, published in the Philosophical Transactions, arrived at the conclusion, that increased energy was given to the small magnets designed for compass needles, by reducing the hardness in the middle of the needles after their having been tempered at a blue throughout. As, however, the investigations of this scientific experimentalist had been chiefly made with small masses of steel, and, generally, with thin plates, it by no means followed that this was a general law, -as we shall hereafter find it did not in the case of other results obtained with like masses. On the contrary, it will be abundantly proved, that the principles which affect the proportionate powers of small and thin magnets, of unequal degrees of hardness or difference of quality, will not apply to a comparison of the powers of large masses, under the like conditions. In order to determine what might be the effect, of reducing the hardness of a combination of tempered steel plates, on their magnetic capabilities, I made trial of different sets, some being two feet in length, some 8.7 inches, and others six inches; the plates of each set being of corresponding size, quality, and temper. The first set, subjected to investigation, consisted of plates of ordinary steel, cut out of a parcel obtained at the shop of an ironmonger and tool-maker in London, 8.7 inches in length, 65 inch in breadth, about 035 inch in thickness, and weighing, on an average, 370 grains. These were 24 in number, consisting of a selection, out of the original series, of the weaker plates, or those of an inferior degree of retentiveness. Their degree of hardness was, I considered, that of a low spring temper, pretty regular throughout. The whole series, being strongly magnetised, was first laid together in one fasciculus, and their order of tenacity, for magnetism, individually determined. Each plate, having been previously marked with a separate number, had its several capabilities registered. Two sets of experiments were then undertaken with this series, in their original condition as to temper, arranged in their order of retentiveness or strength, for the determination of the progressive accumulation of power; first, commencing with the strongest plates, and next, after their being remagnetised, commencing with the weakest, or in the reverse order of the series. These experiments being completed, the whole of the plates were then reduced in the middle to the extent of about six inches, by laying them in succession across an iron bar, heated to redness, till the surface above became of a deep purple colour, and then letting them cool slowly, leaving nearly one and a half inches of each end in the original state. In this condition they were re-magnetised, and their powers, in progressive combination, again determined. Twelve of the series, from the weakest side of the entire fasciculus of plates, were then heated throughout their length in the same manner as before, and their powers also examined. The whole of the results are given in the following table. The first division exhibits the actual and comparative powers of the whole series, commencing with the strongest plates, in two different conditions,-that of their original temper, and that of their supposed reduced state, after being heated in the middle. The second division shews the powers of the series in a reverse order, commencing with the weakest plates, in their original state as to temper—with the comparative powers of a part of the same series when reduced throughout their length, by heating as before. TABLE exhibiting the Comparative Powers, in combination, of the Series of smaller Plates, 8-7 inches in length, in three several states of temper. 1234567 9.42 171 10.56 193 24 8.52 156 15.55 285 17.54 18.21 332 20.46 19.22 351 21-55 402 5 21.7 366 23.34 436 20 8.58 158 6 21.38 397 23.42 7 22-25 413 24.31 439 19 323 23 14.22 256 14.29 258 309 325 18.33 336 19.26 353 19.54 362 18 21 27.42 525 30.23 586 4 27.38 524 The results of these several effects of heating, and (as it is assumed) reducing in temper, were not such as had been expected. For, from all foregoing experiments, I had been led to conclude, that, whatever might have been the effect of the reducing of the temper in single thin plates, or small masses, the effect in combination would have been to diminish the energy of the plates, whereas a considerable gain of |