PHYSICS, PROGRESS OF, IX 1896. 



first, that when the circuit absorbs energy from tho 

 external wave system till it has acquired ft. maxi- 

 mum of oscillation: second, tlie stationary stage, 

 when the circuit absorbs only as much energy as is 

 necessary to compensate for that which it gives off; 

 and. third, the stage when the oscillation is dying 

 out. after the cessation of the primary wave. Tin- 

 distribution of energy in each statre can be calcu- 

 lated. Zeemato (Amsterdam Royal Academy <*' 

 ences. Sept. '-26. !S!l5)has investigated the variation 

 of absorption of electric waves by electrolytes, with 

 the wave length and the concentration. He finds 

 that the coefficient of absorption changes as the 

 square root of the conductivity of the solution, and 

 does not change if the conductivity and the wave 

 length vary in the same ratio. In another paper 

 (ibid.. Oct. 26, 1895) he finds (1) that the energy di- 

 minishes in the logarithmic ratio : (2) that if the wave 

 length is 6'5 metres, the energy has decreased to one 

 third of its original value when the wave has passed 

 through 6'5 centimetres of a solution of common 

 salt, whose resistance is 3,200-1 10 that of mercury. 

 Drude (Wiedemann's "Annalen," May) finds that 

 short electric waves are more strongly damped in 

 alcohol and glycerin than in water or aqueous solu- 

 tions. Theoretically the damping should increase 

 with the conductivity, yet these badly conducting 

 bodies damp as effectually as a 5-per-cent. solution 

 of copper sulphate, which is 6,000 times more con- 

 ducting. Alcohol and glycerin also show anomalous 

 dispersion for rapid electric oscillations. 



Electric Wares : Refraction and Dispersion. 

 Cole (" Physical Review," July-August) finds that 

 for waves 260 centimetres long the index of refrac- 

 tion of alcohol is 5 - 24 and that of water is8'95. For 

 waves 5 centimetres long the indices are 3;2 and 

 fci-s.~) respectively, showing that the dispersion by al- 

 cohol between these limits is considerable. Lampa 

 (Vienna " Sitzungsberichte '') has determined the 

 refractive indices of several substances for electric 

 waves 8 millimetres in length as follow : Paraffin. 

 1-524; ebonite, 1'739: crown glass. 2'381 ; flint glass, 

 2-899; sulphur. 1-802: benzole, 1-767; glycerin, 

 1-843; oil of turpentine. 1'782; oil of vaseline, 

 1-626 ; oil of almonds. 1-734 : absolute alcohol. 2'568 ; 

 distilled water. 8'972. Cohn and Zeeman (Wiede- 

 mann's *' Annalen." January) find that when elec- 

 tric waves are propagated in water no dispersion is 

 observed for frequencies ranging from 27 to 97 mil- 

 lions, and that the index of refraction for such 

 vibration equals the square root of the dielectric 

 constant. 



Electric \Vai-es: Effect on Resistance. Aschki- 

 nass (\Viedemann's Annalen," March) finds that 

 tin-foil gratings have their resistance lowered by 

 electric rays, but the original resistance is restored 

 by shock or heat. Von Lang (Wiedemann's " An- 

 nalen." January) finds that when a rod of carbon 

 stands lightly on a carbon block there is a degree 

 of contact whose resistance is abruptly diminished 

 by electric waves in the neighborhood, even those 

 produced by an electrophorus. The resistance is 

 restored by a slight concussion. Zinc and alumin- 

 um give similar results. Gulik (Amsterdam Royal 

 Academy of Sciences. Jan. 25). in investigating the 

 cause of the variation of resistance in microphonic 

 contacts brought about by electric vibrations, has 

 ascertained that the ends of a current chain, when 

 brought very close together, attract each other if 

 the wires are exposed to Hertz waves. If the air 

 gap is longer than 4 microns, small sparks result. 



Electrographs. Robinet and Perret (" La Na- 

 ture." May 9) have made electrographs by placing 

 a sensitive plate beneath a negative, sheltered from 

 ordinary light, in a sort of condenser formed of air 

 films. An exposure of thirteen minutes gave an 

 excellent result. The plate is more rapidly acted 



upon when the frequency of the alternating cur- 

 rent that is UM'd is greater, and the impre^ion de- 

 pends also on the nature of the condenser : 

 (cnpper and lead giving 'he l-t results) and on 

 their distance. The experiments differ from tl 

 performed several years ago by Xiewenglowski and 

 Lamoureux in the shortness i.f exposure and in the 

 fact that phosphorescence plays no part in tl. 

 tion. 



Hall Effect. Bagard (Pan's Academy of Si-imces. 

 Jan. 13) finds, contrary to conclusions already 

 reached by Roiti, that Hall's effect occurs with 

 liquids. lie has proved this experimentally for 

 saline solutions, and finds that the results for zinc 

 sulphate are of the same order of magnitude as 

 those for metallic bismuth. 



Edison Efft-H. Fleming (London Physical So- 

 ciety, March 27) has experimented on the Edison 

 effect in an incandescent lamp that is. on the 

 passage of a current between a metal plate inside 

 the loop and the positive (not the negative) lead of 

 the lamp. He finds that after a certain time, de- 

 pending on the position of the plate, its potential 

 falls to that of the negative plate. This time is 

 longer if the loop is outside the filament. The 

 space between the plate and the negative lead has 

 a kind of unilateral conductivity, a low-voltage 

 battery being able to send a current from plate to 

 lead, but not in the opposite direction. If the volt- 

 age of the lamp current be raised, a current can be 

 passed from the plate to the negative lead, while a 

 galvanometer properly connected will show that a 

 current is passing from the positive lead to the 

 plate. In this condition the lamp is very sensitive 

 to the influence of a transverse magnetic field. 

 Prof. Fleming thinks that his experiments show 

 that the resistance of a vacuum tube to the passage 

 of a discharge would be greatly reduced if the 

 cathode were made incandescent. Salomons (Lon- 

 don Physical Society. Feb. 28) shows that if an 

 incandescent lamp through which an alternating 

 current passes is brought near a continuous-current 

 electro-magnet the filament vibrates, and may be 

 broken. The number and position of the nodes are 

 independent of the natural period of the filament, 

 but depend on the frequency of the current. 



Electrification of Vapors. Schwa Ibe (Wiede- 

 mann's "Annalen," July) finds that the vapors 

 rising from electrified liquids can not bear away 

 with them any of the electricity. This goes to dis'- 

 prove Exners theory of atmospheric electricity. 



Electric Measurements. Campbell (London Phys- 

 ical Society. May 22 < employs for the direct measure- 

 ment of the frequency of alternating or pulsating 

 currents a steel wire or a steel spring acted upon 

 by an electro-magnet through which the current 

 s. The tension of the wire or the length of 

 the spring is varied till maximum resonance is ob- 

 tained, which point is found by a small contact 

 piece. The instrument will measure frequencies of 

 40 to 150 double vibrations a second. Starke (Ber- 

 lin Physical Society. Feb. 28) has devised a simple 

 method of determining the electrical constants of 

 solid bodies, based on introducing into one arm of 

 a Wheatstone bridge a condenser between whose 

 plates fluid mixtures of dielectrics with varying 

 electrical constants may be placed. A mixture is 

 sought, such that the immersion in it of the given 

 solid does not alter the constants of the mixture. 

 Leake. Leventhorpe, and YVhitehead (London Royal 

 Society. l>ec. 5. 1*115). by using a method adopted 

 by Ayrton and Mather for accurate calibration of 

 electrostatic voltmeters, find that measurements of 

 steady potential differences of about 2,000 volts 

 made in the ordinary way with the Kelvin absolute 

 electrometer are, on the average, 1'5 per cent, too 

 large an error due to imperfect action of the 



