210 ELECTRICAL CURRENTS OF THE EARTH. 



earth. Nor in most cases are we iuformed whether or not the nniuhers 

 reported were obtained by experiments made, as is (piite probal)le, at 

 the time when the lines Avere in service for telegraphic coriespondeuce. 



It wonld bo useless, we think, to eidarge critically upon the experi- 

 ments to v.inch we allude or the resnlts obtained. To such criticism 

 a distinguished Swiss physicist, M. Dufour, of Lausanne, has lately 

 devoted himself, and we content ourselves with citing, in his own words, 

 the conclusion to which he has arrived: "It is qnite evident that if 

 researches are to be undertaken respecting the electric currents of the 

 earth, offering any solid guarantee of exactness, it is necessary to employ 

 special lines and such as are absolutely independent of the telegraphic 

 reticulation." 



Anuing our i)redecessors in these researches, Lamont alone seems to 

 have bestowed some previous attention on the method of experimenting 

 and on the causes of error incident to the methods followed. Hence 

 the eminent astronomer of Monaco confesses that he had not yet found 

 in these experiments a iw'int of dcpartiire sufficiently secure^ i\u(\ closes his 

 memoir v.ith the admission that what he had publislied up to that time 

 ought to he regarded only as a few general and preliminary indications. In 

 a word, 1 do not think it an exaggeration to affirm that it would be 

 impossible, from all the researches which I have cited, to draw the 

 demonshation of the existence of the phenomenon of an electric current 

 which circulates in a metallic line extended along the earth, and insu- 

 lated from it, having its extremities sunk in the ground, inde|)eudently 

 of the heterogeneity of the electrodes and of the various causes of error 

 -introduced into those experiments; meaning by that phrase causes 

 already known, and which have nothing to do with a proper electric 

 stratum of the earth. 



METHOD OF EXPERIMENTING. 



The description of this method should embrace the metallic part of 

 ithe mixed circuit, the comuumications between the extremities of the 

 metallic line and the ground, and the instruments used to detect land 

 measun^ the current. 



Metallic line. — I will state, in the first place, that none of the experi- 

 ments described in this memoir have been executed upon a wire per- 

 taining to a telegraphic line composed of several other wires. When- 

 ever I have used a telegraphic line, it has consisted of a single wire; 

 ■ and tlie experiments were made eitlier during montlis when the tele- 

 . graphic service was not conducted by that wire, or at hours when that 

 service was li:nown with certainty to be suspended. Before commencing 

 the experiment the line was examined throughout its course, and pro- 

 tected by the removal of any possible contact with the boughs of trees 

 or the v.'alls of houses, and by the renewal of the solderings of the 

 junctions. The line w^as ibrmed of the usual iron wire of telegraphic 

 connections, 3 or 4 millimetres (J of an inch) in diiimeter. Its extremi- 

 ties were united to the instruments, and the electrodes sunk in the 

 ground by means of a piece of copper wire covered with gutta-percha, 

 freshly soldered outside of the telegraphic offices, and all communica- 

 tion was interrupted between the line and the usual wires entering 

 the offices. The insulation of the line was always tested before com- 

 mencing the experiment, and was always such as not to impart 

 during dry days a sensible deviation to a galvanometer of 2,000 coils. 



In many of the experiments which we shall report, a copi)er wire, 2 

 millimetres (^V of an inch) in diameter and covered with gutta-percha, 



