THE PROPAGATION OF ELECTRICITY. 171 



metre, in accordance with the plan of Dr. Wollaston, while their upper extremi- 

 ties communicate respectively with two ends of" the wire of a galvanometer, 

 whose coils are well isolated by means of resin. The supports which bear the 

 platiua wires are movable along a division in such way that the two extremi- 

 ties of the wires immersed in the water may be made to approach one another as 

 closely as possible, or be separated very nearly the whole length of the stratum 

 of water. By means of a micrometric screw, the relative distance of the two 

 points of platina may be so varied as to be appreciable to nearly the tenth of 

 a millimetre. These two points draw off an almost insensible proportion of the 

 electric current which traverses the trough filled with water — a proportion, how- 

 ever, which suffices to act in a distinct manner on the needle of the galvanometer. 

 The proportion drawn off depends for a current of constant intensity on the dis- 

 tance of the two points, so that, if the intensity be variable, it is the variable 

 distance to which it is necessary that the two points shall be brought, in reference 

 to one another, in order for the indication of the galvanometer to remain con- 

 stant, which measures the proportion drawn off in each case, and thus, by a ratio 

 easily determined, the absolute intensity of the current. 



Finally, a good pneumatic pump, to which a second complementary one may 

 be joined, enables us to bring the gas to an advanced degree of rarefaction. As 

 to the elastic force of the gas, that is measured by a manometer of mercury 

 very carefully constructed, with which, by means of a cathetometer, a difference 

 of pressure of even the fiftieth of a millimetre may be appreciated. 



$ I.— GENERAL PHENOMENA PRESENTED BY THE TRANSMISSION OF 

 ELECTRICITY IN RAREFIED GASES. 



The Ruhmkorff apparatus, of which I have availed myself, gives in the in- 

 ducted wire two successive and alternately contrary discharges. Hence, if 

 these discharges encounter in the circuit which they traverse only good con- 

 ductors, such as metallic wires, and even distilled water, no deviation is re- 

 marked in the galvanometer, because the discharges being alternately in a 

 contrary direction, and in rapid succession, their opposed double acti(m is 

 neutralized. But if the circuit comprises an elastic fluid very much rarefied, 

 the resistance which it opposes to the passage of the two successive discharges 

 causes one of them to predominate, so that the phenomena take place as though 

 there were but a series of discharges all in the same direction. The explana- 

 tion of this difference is, that the two discharges, or inducted currents, though 

 eqxial in quantity, have not the same tension, the direct, which have a less 

 duration, having a stronger tension. It thence results that when the circuit is 

 interrupted by a body which is a bad conductor, such as an elastic fluid more 

 or less rarefied, the direct currents can alone be transmitted, so that the direc- 

 tion of the inducted current which traverses the elastic fluid is the same with 

 that of the inductive current, and the latter changing, the other changes at the 

 same time. 



The pressure at which a discharge of a given intensity begins to pass through 

 a gas varies with the nature of that gas, with its degree of rarefaction, and with 

 the dimensions and form of the vessel which contains it. Moreover, the dis- 

 charge does not pass immediately upon the electrodes being put in communica- 

 tion with the poles of the Ruhmkorff apparatus. For that a certain time is 

 necessary — a time so much longer in proportion as the resistance is greater, 

 whether arising from the nature or density of the elastic fluid, or from the effect 

 of the form and dimensions of the vessel. Thus, in a long tube, from 2 to 

 5 centimetres in diameter and from 30 to 50 centimetres in length, it requires 

 several minutes before the discharge can be transmitted, however rarefied the 

 gas. But the first discharge having once passed, the succeeding ones pass 

 with facility, and follow one another so rapidly as to produce on the galva- 



