310 ON THE ELECTRICAL CURRENTS OF THE EARTH. 



tlie ground in which the tenninal cavities were dug, and the deviation was not 

 found to vary in a lapse of many days, provided there was neither tempest nor 

 rain. After rain, the deviation was constantly seen to increase. I have satis- 

 fied myself, by measuring a constant current transmitted in this mixed circuit, 

 that the augmentation which followed rain was only the effect of the better con- 

 ductibility of the earth depending on a state of greater humidity in the teiTcstrial 

 stratum immediately in contact with the extremities of the line. And, in fact, 

 it could be obtained V>y pouring around the cavity in which the electrodes were 

 plunged, within a radius of two to three metres, a few buckets of water. 



I have tried the injmersion of the electrodes in the water of a well, which was 

 effected by a very simple contrivance. For this purpose, I take a thick square 

 piece of cork and fix, in a hole made in this cork, porous vessels filled with sul- 

 phate of zinc. The cork suspended by a cord floats on the water of the well 

 into which the porous vessels descend ; by means of a copper wire covered with 

 gutta-percha and bound to the cord, the electrode of zinc was introduced into the 

 porous vessel and comuuinicated with the line. I was thus able to establish the 

 mixed line, employing the well water as extremities of the teiTestrial stratum, in 

 which the electrodes were sunk. With this aiTangement, also, I have realized an 

 ascending cuiTcnt in the wire, and the deviation was only a few degrees greater than 

 that of the cmTent obtained by using the artificial cavities or pits which I have de- 

 scribed above. By using the wells we have this advantage : that the conditions 

 of conductibility of the teiTestrial strata into which the electrodes ai'e introduced 

 remain invariaV)le. It is necessary to ascertain in advance that the waters of the 

 two wells, when those which we employ are in two cavities formed in the earth 

 at a short distance from one another, do not yield an electric cun-ent. I have 

 varied as far as possible the excavations situated at different levels, but in all 

 cases have found the current in the metallic line to be an ascending one. I was 

 even enabled to divide the line at the hill of Turin, a length of nearly 600 metres, 

 about midway where there existed a well, and this remarkable and constant 

 result was realized : that, notwithstanding the greater resistance of the entire 

 line, the cuiTcnt, which continued to be ascendant in the two halves, had still a 

 less intensity in the two lines taken separately than in the entire line. 



I have had an opportunity of observing in these lines the eftects of two or 

 three storms during the month of July. I will first remark, that I have satisfied 

 myself that in leaving one of the extremities of the line in conmiunication with 

 the electrode and the earth, and the other in the air, I had never any trace of a 

 current, even when using a galvanometer of 24,000 coils. I have often made 

 the experiment of putting an isolated metallic vessel, at the end of a wooden 

 staff from seven to eight metres high, in communication with the extremity of 

 the line which was in the air; placing in the vessel sometimes lighted coals, 

 sometimes touch-wood, sometimes shavings saturated with burning alcohol, in 

 order to obtain a large flame and a current of heated air. In all these experi- 

 ments, whichever might be the extremity of the line immersed in the water or 

 raised in the air, I have never obtained a sign of the current in the most delicate 

 galvanometer, provided care were obsei"ved to isolate the line effectually and no 

 account were taken of the indications of the galvanometer at the moment when 

 it was necessary to touch the line with the hands. 



Neither, during storms, have I observed, with the line, which was only 600 

 metres in length, any deviation in the needle at the moment when lightning 

 flashed between clouds, provided the two extremities of the line are not in com- 

 munication with the ground. When this communication is established and a 

 deviation of the needle has resulted from the terrestrial current, a sudden move- 

 ment is seen to take place at each flash, such as would be occasioned by the dis- 

 charge of the torpedo fish. I observed at the same time the galvanometer and 

 an electroscope of dry batteries (a piles sechesj communicating with an iron wire 

 from seven to eight metres long, well isolated and raised in the aii', and having 



