subject, either as to the nature of the steel to be used for the best results, or the methods of treatment, hardening, tempering, or the like to be employed. This led to the carrying on through several years of thousands of experiments, and the development of methods of rapid measurement of the flux between the poles of the magnets, as well as methods of heating, forming, hardening, tempering, seasoning and testing for permanence. Very many analyses of different steels were also made, with the object of discovering the effect of the differing proportions of carbon, tungsten, silicon, etc. I notice that Dr. Guthe has not included a ring of tungsten steel now almost universally used for magnets, among those subjected to test. In our work we soon found that no steels without tungsten gave such relatively high induction together with other desirable qualities, as steel containing two to three per cent. of tungsten. Amounts up to four or five per cent. of tungsten gave no advantage, while with less than about two per cent. the induction fell off and the metal approached ordinary steels in magnetic qualities. It was found that the treatment recommended by Barus and Strouhal for seasoning, and referred to by Dr. Guthe, was not sufficient to ensure permanency even where the form of the magnet was, as in our meter magnets, quite favorable to retention of magnetism. The effect of heating to 100° C. for several hours, caused a loss or diminution of flux which varied with the composition and prior treatment of the steel. Extending the time of the heating to twenty-four hours or more, gave scarcely more effect than an hour or two. The object was, of course, to reduce the flux to that amount which the steel would retain permanently. In the course of this work it was determined that repetitions of the heating to 100° C. after cooling would cause a further loss of magnetism, but at each repetition the percentage became less so that there was at last no discoverable change. A magnet so treated is practically permanent at ordinary temperatures. Whether the effect obtained is due to the removal of that portion of the flux which may be considered as unstable or to a magnetic hardening was unknown, but it would appear from Dr. Guthe's paper that both effects may have been produced, assuming that tungsten steel behaves in the same ways as that of the rings tested by him. We developed a process of seasoning magnets which consisted in subjecting them after hardening and magnetizing to successive immersions in boiling water and in cold water at intervals of a few minutes; a treatment carried on for a sufficient length of time to ensure stability of the flux density in the air-gap between the poles. Certain facts alluded to in Dr. Guthe's paper have been known to us for a long time, and were developed in our practice. At one time we arranged to temper our magnets after hardening, by heating them to temperatures just short of those giving the blue oxide coat. This was found to be a matter of difficulty in practice on the large scale, though it undoubtedly increased the induction. It was preferred to obtain a similar result by carefully excluding steels with too high a percentage of carbon, and by hardening the magnets by care in selecting the proper hardening temperature. Thus in fact the same principle was developed in our work as is alluded to by Dr. Guthe, namely, that too high temperatures of hardening give weaker magnets or those in which the steel is magnetically as well as mechanically harder and of too low magnetic susceptibility. It was in fact found that the desired flux could not be retained by such overheated steel when magnetized. I think that the fact that with ordinary steels the strongest magnets are produced when the hardened steel has been heated to a blue, is well known, and has been practiced for a very long time by magnet makers. The magnets of the old magneto machines were nearly always blued. It has been my practice also for fifteen to twenty years to adopt a similar process in dealing with ordinary steel. After hardening, the brightened steel was carefully heated by a blow-pipe until the middle or neutral region of the magnet (bar or horseshoe) was blued, while the poles were kept of straw color, the one color fading into the other along the bar. A strong magnet is thus made and one which is relatively permanent. Here The point made by Dr. Guthe that there is a proper temperature for annealing iron or mild steel in order to secure the greatest freedom from hysteretic loss has also been found to be the case in our practice with iron and steel. The range of temperature for the best effect is found not to be a wide one. again there is a widely known mechanical fact which bears upon the point under consideration. It is this, that some steels heated to that temperature which just fails to harden them, are on quenching made very soft. The range of temperature is narrow, but evidently has the bearing upon the recalescence temperature pointed out by Dr. Guthe. This may be observed by heating to redness one end of a bar and quenching it, when it will be found that just back of the hardened end is a very soft portion; noticeably so. It would appear probable that whatever softens the metal mechanically may have a similar effect magnetically. It would take more time than is at my disposal to touch upon the many other interesting points brought out by Dr. Guthe's paper. I have alluded above to some of the more practical points useful to the constructor. [COMMUNICATED after adjournment, by J. STANFORD BROWN.] This paper is very interesting. The chemical composition of steel for magnets, however, is perhaps of more practical importance than might be inferred from the paper, or even from the remarks added by Prof. Thomson. The use to which magnets are put largely, or you may say entirely, determines the required chemical composition. Steel for permanent magnets will be entirely different from that for electro-magnets. Magnets made to sustain a load are satisfactorily made from a steel which would fail utterly for the controlling magnets of our beautiful Thomson wattmeters, and so on. The following specification for magnet steel was a few days since submitted to the writer : C.75; W 4.5 to 5.5; P .03; S .02; Si .10 or less; Mn .12 or less. Such steel is somewhat difficult to make and is consequently expensive, retailing at 40 cents or more per pound, according to quantity. The specification is peculiarly low in C for the amount of W. With so high W the carbon could more usually run from .95 to 1.05. A highly satisfactory magnet steel for certain uses runs: C.65; Mn .35; W 2.95. On the other hand, for another kind of magnets, C.65; Mn .60; W nil; Si .25, and P .065 is supplied. Here the high Si is noticeable. Sometimes crucible steel is used, and other times a cheap open hearth steel seems to answer. The electrical companies have an enormous amount of data and know thoroughly what they want, but unfortunately they still seem to regard it of too high commercial value to be contributed to the advancement of science and their competitors. REPORT OF COMMITTEE ON UNITS AND STANDARDS. To the President and Council of the American Institute of Electrical Engineers. Gentlemen :— Your Committee on Units and Standards having carefully considered the communications which you have referred to us, recommend : 1st. That Hefner-Alteneck Amyl-Acetate Lamps furnished with test certificates from the Physikalisch-Technische Reichsanstalt at Charlottenburg, Berlin, should be temporarily adopted as concrete standards of luminous intensity, or candle power. 2nd. That in measuring the mean horizontal luminous intensity or candle power of an incandescent lamp, a Lummer-Brodhun photometer screen be adopted, and that the incandescent lamp be steadily rotated about a vertical axis through its axis of figure at a uniform speed of approximately two revolutions per second. Your Committee believe that the adoption of these recommendations would lead in practice to a much greater degree of uniformity in results of measurements of the candle power of incandescent lamps, by different and remote observers than is now usually attainable. Although incandescent lamps are at present rated by their horizontal candle power, yet, since the only true criterion of the total quantity of light emitted by a lamp is its mean spherical candle-power, we recommend that the rating of lamps should be based upon their mean spherical candle-power so far as is commercially practicable. January 19th, 1897. Yours respectfully, A. E. KENNELLY, Chairman. W. E. GEYER, AMERICAN INSTITUTE OF ELECTRICAL ENGINEERS. NEW YORK, April 28th, 1897. The 115th meeting of the INSTITUTE was held April 21st, at 12 West 31st Street, and was called to order by Vice-President Steinmetz at 8.15 P. M. The following associate members were elected by the Executive Committee April 28th. Name. ABBOTT, HENRY Address. President, Calculagraph Co., 2 BROWNE, SIDNEY HAND Consulting Electrical Engineer, CARTER, HENRY W. GREENWOOD, FRED. A. HOAG, GEO. M. Attorney and Expert in Patent City Electrician, City of Cleveland, KAMMERER, JACOB A. General Agent, The Royal Elec- LOVEJOY, D. R. MATHER, EUGENE HOLMES Superintendent and Electrical RALSTON, LOUIS C. Engineer, New Haven Street Railway Co., Exchange Building, New Haven, Conn. Graduate Student, Cornell University, 1170 Madison St., Oakland, Cal. Endorsed by Louis Duncan. F. S. Hunting. M. C. Canfield. Theo. Stebbins. Fredk. Bedell. Edw. L. Nichols. |