ON AN EXPERIMENT BEARING UPON THE QUESTION OF THE DIRECTION AND A COIL of wire of 390 turns was driven at a very high rate of speed, the axis of the coil being the axis of rotation. When the coil reached 380 revolutions per second, the linear velocity of the wire in the direction of its own length amounted to 8,000 centimeters per second. By means of two brush contacts at the axis a current was sent through the coil while the latter was in motion. The magnetic moment of the coil was determined by means of a very sensitive astatic pair of magnets carrying a mirror. Readings were taken with the coil at rest and in revolution, the motion of the coil and the direction of the current being repeatedly reversed. If the electric current result from the flow of a fluid through the wire, in other words, if it may be considered as possessing direction and finite velocity, a motion of the conductor, with or against the current, should produce an appreciable influence upon the deflection of a magnet needle, even though the velocity of the current were very large as compared with that of the conductor. In order to render the detection of this presumably very small effect less difficult, the direct influence of the coil was elimnated by differential winding. Under these circumstances, when the coil was carrying as large a current as it could be made to do without injurious heating, the rotation of the coil was found to be without appreciable effect upon the magnetic moment of the same. The best results were obtained by sending 4.6 amperes of alternating current of 40,000 alternations per minute through the coil. At a velocity of the wire equal to 8,000 centimeters per second the rotation of the coil produced no effect upon the needle amounting to 0.2 millimeters deflection. The figure of merit of the coil and needle was determined by substituting a coil of continuous winding, its position with respect to the needle being the same as that of the rotating coil, and determining the current necessary to produce 1 centimeter deflection. The sensitiveness of the apparatus was found to be such that a current having direction and a velocity of 1,000,000,000 meters per second would have shown a change in its action upon the needle (when the motion of the coil was 380 revolutions, 8,000 centimeters per second) amounting to 0.1 centimeter deflection, an effect which could not have escaped observation. It follows from the above negative results that if the electric current consists in the flow of a medium or fluid through the conductor, the velocity of the same must be greater than the exceedingly high rate just mentioned. Foepl, who in some recent experiments (Wiedemann's Annalen, 1886), used an apparatus in most essential particulars similar to our own, but one by means of which only relatively very low velocities could have been detected, has reached the same negative conclusion. ON THE EFFECT OF THE ADDED TERM OF THE EQUATION OF THE QUADRANT ELECTROMETER ON ITS DEFLECTION CURVES. By Prof. T. C. MENDENHALL, President of Rose Polytechnic Institute, Terre Haute. Indiana. [ABSTRACT.] To the well known expression for the moment of the couple acting on the needle of the quadrant electrometer, M. Gouy has added a term depending on the difference of potential between the two quadrants and proportional to the square of this difference. The effect of this term on the deflection curves of the electrometer, as mounted according to Mascart, Thomson or Joubert, is investigated and found to be essentially as follows:-The Mascart equation still remains that of a straight line; the Thomson method reduces the equation to one of the third degree, no longer representing a parabola, but a curve with two branches, each having a point of inflection and an asymptote common to both. It is also shown that the first part of this may be assumed to be a straight line without sensible error, provided the potential of the needle is very large compared with that of the quadrants; thus, if the needle is charged to 5,000 volts, potentials below 100 volts may be compared by this method with less than one per cent error. . . . By the Joubert method of mounting, in which the needle and one quadrant are connected and the other is put to earth, the equation, which without the added term represents a common parabola with the origin at its vertex, is that of a curve somewhat similar to a parabola but with a point of inflection in each branch and an asymptote common to both branches. INCANDESCENT LAMPS CONSIDERED AS MACHINES FOR TRANSFORMING ELECTRICAL ENERGY INTO THE ENERGY OF LIGHT. [ABSTRACT.] By ERNEST THE author determined the ratio of the energy of the visible rays to the total energy of the lamp. This was done by cutting off the dark heat by a cell of about one decimeter thick, containing a solution of alum, and allowing the rays transmitted to fall on a delicate thermopile. The deflection of the galvanometer in circuit with the pile was then proportional to the energy of the light from the lamp. The alum cell was then removed, and the deflection corresponding to total radiation observed. The ratio of the two deflections gave the efficiency of the lamp at that candle power, when considered as a machine for producing luminous rays. This ratio was determined in one case by placing the lamp to be tested in a large glass calorimeter and measuring the heat given to the water and the total electrical energy supplied. In both cases corrections were applied, due to the fact that water or alum does not absorb quite all of the dark heat. In the case of alum this transmission amounts, in extreme cases, to 0.5 per cent. With water it is about 1.5 per cent. A correction was also made for the absorption of light by the water. The results show that the efficiency of a lamp as a machine does not correspond at all with its commercial efficiency. One lamp may have a larger "luminous efficiency" than another, and yet require a much larger expenditure of energy to bring it to the same candle power. For example, two Edison lamps were tested: one being quite new, the other old and commercially inefficient. The first gave 16 C. P. at 85 Watts, with a "luminous efficiency" of 4.0 per cent; the second gave 16 C. P. at 100 Watts, with an efficiency of 7.4 per cent. The values of the "luminous efficiency" of five different lamps at different candle powers are given below in per cents: The energy per candle power was also calculated from these results. This energy, or the mechanical equivalent of one candle power, decreases as the candle power rises. The following table gives the values of this mechanical equivalent in Watts: FLOATING DYNAMOMETER. By Prof. J. BURKITT WEBB, Stevens Institute, Hoboken, N. J. [ABSTRACT.] AT last year's meeting a new form of dynamometer was described with the aid of an illustrative model, which was more however in the form of a laboratory experiment than in that of practical application. During the past year two practical dynamometers have been constructed and have worked satisfactorily in all respects, so that there would seem to be no better method of testing dynamos and motors, especially those of great weight or unsymmetrical form. Especial forms have been devised for other purposes among which is a floating transmission dynamometer capable of giving exact results for scientific purposes. The photographs show an eight-hundred pound dynamometer mounted with a "Westinghouse Exciter" of nearly that weight. The "Ark," or "Floating Dynamometer," is applicable to all cases where a "Bracket Cradle " might be used and possesses some advantages which the latter has not. "OVERHAULING" IN A MECHANICAL POWER. By Prof. J. BURKITT WEBB, Stevens Institute, Hoboken, N. J. [ABSTRACT.] PROFESSOR BALL, in his book on "Experimental Mechanics,” has attempted to state a general law for such "overhauling." His statement is that whenever rather more than half of applied energy is lost in friction the apparatus will not overhaul. It has been shown by others that this statement is not always correct', though the entire fallacy of it may not have been perceived, the law, as we have worked it out, not having such a simple expression. We will simply show here how a mechanical power may be constructed which will not overhaul and yet lose much less than half the applied energy in friction. For example let the "power" be a simple lever with the power pulling upward and the weight consequently applied between the power and the fulcrum. Let the weight be applied at one inch and the power at two inches from the fulcrum and let the latter be a journal having a radius of five inches turning in a suitable box or bearing with a friction of one-fifth of the weight; i. e., with a coëfficient of friction equal to the tangent of the angle whose sine is one-fifth. Such a mechanical power will not overhaul and will waste in friction but one-quarter of the energy applied. This may suggest to any one having the necessary mechanical insight how to construct other non-overhauling mechanisms with other fractions of energy lost in friction. IMPACT IN THE INJECTOR. By Prof. J. BURKITT WEBB, Stevens Institute, Hoboken, N. J. [ABSTRACT.] SOME Comparisons have been made between pumps and injectors as to the mechanical work performed by them respectively in feeding a boiler; it is thus brought out very forcibly that while the injector wastes no energy it does very little work, most of the steam being used in warming the water. It will be interesting to notice the reason for this because it depends upon the simple principles of mechanics, which though embodied in the usual equations for the injector may not, in that form, be sufficiently appreciated. 1 Dr. Coleman Sellers, of Stevens Institute, Hoboken. N. J., showed this in a lecture to the upper classes so far as applied to an apparatus consisting of a number of parts. There are three places in the injector to which we will call attention in examining the action of a particle of steam; (a) in the boiler, (b) in the vacuum chamber, (c) in the mixed column of water and steam. At a the steam is practically in a state of rest, but possesses energy capable of doing work. In passing from a to b work is done in giving to the steam a high velocity, so that at b it possesses a large portion of kinetic energy by virtue of which it strikes the water a blow and sets it in motion. Suppose now that the steam, having a velocity v, strikes, say, fifteen times its mass of water and sets it in motion, the combined mass will have a velocity of only one-sixteenth of v, because the momentum before impact, v X 1, must be equal to that after impact, v 16 × (1+ 15). The kinetic energy possessed by the steam × 2 but afterward there will be no more then so that at impact fifteen-sixteenths of the kinetic energy of the steam disappears, being retransformed into heat by the concussion, it will thus be seen that the injector does primarily transform a large amount of heat into work, but that it uses this work in so uneconomical a way, in forcing the water into the boiler by impact, that most of it is wasted as mechanical work, though saved as heat. before impact is equal to 2 v2 162 NOVEL FORM OF ELECTRO-MAGNETIC TELEPHONE. By Prof. R. B. FULTON, University, Mississippi. [ABSTRACT.] THE author has devised a form of electro-magnetic telephone, which may be used as a receiver or transmitter, in which the lines of magnetic force, the direction of the induced currents, and the direction of the movement of the vibrating parts are each at right angles to the other two. This end is attained by dispensing with the ordinary spool electro-maguet and circular diaphragm. A tube of thin sheet iron, open at one or both ends, having a polygonal cross section, is magnetized, and has ordinary fine insulated telpehone wire coiled around the outside, the coils being each in a plane perpendicular to the length of the tube. One end of the tube is made a north, and the other a south pole. The coils are placed over those parts of the walls of the tube which are thrown into vibration when the tube is spoken into. The best results seem to be obtained with the above instrument as a transmitter, and the ordinary Bell receiver. A NEW PRESSURE INDICATOR OR RECORDER (EXPERIMENTAL MODEL). By Prof. W. H. BRISTOL, Stevens Institute, Hoboken, N. J. [ABSTRACT.] In this instrument the novelty consists in employing a special form of the well-known Bourdon spring tube in combination with a principle involved in the construction of a pressure indicator which was invented and |