306 ON THE ELECTRICAL CURRENTS OF THE EARTH. 



uiKlergo a cliange of sign when the position of the plates was clianged or 

 their hetcrogeneonsness was moditied by causing the current of a batteiy to pass 

 in a given direction. These currents disappear, or are considerably weakened, 

 by employing })lates and liquids as homogeneous as possible. By employing 

 more sensitive galvanometers and quite homogeneous plates of copper, it 

 will readily be recognized that the slightest difference in the composition of 

 the water of the terminal wells suffices to excite currents. It need scarcely be 

 added that in operating upon telegraphic lines, it is necessary to take account 

 of the secondary polarities which the currents of the battery develop, some- 

 times in one direction, sometimes in the other. Telegraphic lines have also other 

 causes of error due to the variable contact of the wire with the posts. From the 

 moment when I proposed to study '^his subject, I felt convinced that, before 

 aught else, it was necessary to possess a method by which would be realized 

 the condition of having long conducting wires, perfectl}^ isolated, extended in 

 definite directions, the connnunications of which with the earth should be abso- 

 lutel}^ homogeneous, and which should form mixed circuits, all endowed with the 

 same conductibility. It was in the following manner that I attained these 

 objects: 



The wire Avhich I employed was of C(jpper, two millimetres in diameter, and 

 covered with gutta-percha; this wire was suspended by means of a sort of cleft, 

 wrought in the top of a rod or slender post of wood, such as is in use here for 

 military telegraphic lines. These wooden rods were planted at a distance of 25 

 or 30 metres from one another, in two lines exactly traced, one in the magnetic 

 meridian, the other perpendicular to the meridian. Each of these lines was six 

 kilometres in length, the place where they were established being the plain of 

 Saint Maurice, 22 kilometres from Turin, a plain set apart for military exercises. 

 The communications between the extremities of the wire and the earth were 

 effected in the following manner. At the extremity of each line I caused to 

 be dug a kind of pit of rectangular form, two metres in depth and length, and 

 one in l)readth; at the bottom of this pit was formed a cavity much smaller, and 

 which might be termed a sort of capsule, having a width and depth of 30 centi- 

 metres. A 1)ed of clay, such as is used in the fal»rication of pottery, was carefully 

 spread over the interior surface of this capsule, so as to prevent the water from 

 percolating too rapidly through its wall. The same water, being that from a river, 

 was employed for the four cavities, and the person appointed to superintend at 

 each extremity had a supply of this water at hand, in order to maintain it constantly 

 at the same level. Lastly, a porous cylinder, such as is used for the batteries of 

 Daniell, filled with a saturated and neutral solution of the sulphate of zinc, was 

 plunged in the water at the centre of the cavity, and the wire of the line was 

 united to a plate of zinc perfectly amalgamated, and which in turn descended 

 into the solution of the sulphate. The porous cylinders thus prepared and the 

 plates employed were tested in advance, and this testing was renewed from time 

 to time, so as to be sure that the plates were perfectly homogeneous. It rarely 

 happens that two plates once rendered properly homogeneous undergo alteration 

 for several days when they remain constantly immersed in the solution. Should, 

 however, a slight heterogeneity a])pear, it will suffice to wash and amalgamate 

 them anew, in order to render them again homogeneous. The two lines also 

 must be ascertained in advance to have the same conductibility. In a unifonu 

 plain, like that in which I operated, the pits being excavated in nearly the same 

 stratum, the differences of conductibility could not be great; but I succeeded 

 in rendering them equal by deepening by a few centimetres the cavities made at 

 the bottom of the pits of that line which was found to be most resistant. 



In this manner the conditions of the circuit which I deem essential for these 

 experiments were realized. It is proi)er to remark that, wishing to test in advance 

 two similar excavations, with cavities at the bottom as aljove described, and 

 formed at a distance of live to six meters from one another, I found no trace of a 



