ELECTRICAL CURRENTS OF THE EARTH. 213 



fixed, wliicli bore a small arm of irou with a dimiuutive pulley. By 

 meaus of a silk thread and of the pulley, 1 elevated a copper wire cov 

 ered with gutta-percha, which, at the upper extremity, was terminated 

 by a large uncovered i)ortiou coiled spirally, into which I introduced a^ 

 sort of cornet, formed of divers layers of touch- wood, and kindled it at 

 the moment of the experiment. The lower extremity of this copper wire 

 was united to the ball of an electroscope attached to a dry galvanic 

 pile. 



EXPERDIENTS WITH A MIXED CIRCUIT OF A LENGTH NOT GREATER 

 THAN A KIL03IETRE* IN A HORIZONTAL STRATUM. 



Experiments under these conditions have been often tried, by placing 

 the electrodes of zinc sometimes in contact ; sometimes at distances 

 varying from 10 to 20, 50, and 100 metres. It is, in fact, by these ex- 

 periments which I am accustomed to make i>reviously to undertaking 

 those with much longer circuits, that I satisfy myself that there is no 

 heterogeneity between the electrodes of zinc formed in the manner above 

 described. I have been thus enabled many times to ascertain that if 

 on such an occasion there was a slight deviation it depended on the 

 vases of terra-cotta, and that there was no regularity in the currents 

 thus obtained on transporting the vases to ditfei'ent distances within 

 the above limits. In fact, the current is sometimes found to increase on 

 a Avider separation of the vases than to diminish or even disappear, and 

 sometimes to an inverse order on the removal of the vases to still greater 

 distances. In these cases I have always succeeded in recognizing' that 

 there was a difference in the physical qualities of the ground. Thus a 

 current arises if one electrode be placed in a soil charged with loam, 

 and the other in an argillaceous soil ; and in operating on the sands 

 adjacent to the sea, a current supervenes if the electrodes be stationed 

 at different distances from the beach. But the effects of these differ- 

 ences of soil are only manifested when in contact with the vases con- 

 taining the electrodes. Ilence, if it be found that a current exists be- 

 tween two points at a distance of 15 or 20 metres (19 or 60 feet) from 

 one another, we may b^ sure of causing it to cease by excavating at 

 those points two holes, which need not have a diameter of more than one 

 metre, (3 feet,) and filling them with the same earth, into which the 

 vases with the electrodes are then to be introduced. It is advisable, 

 therefore, to pursue this course Avhen oj^erating with mixed circuits at 

 great distances, provided it be not previously ascertained, as I have al- 

 ways attempted to do when practicable, that the deviation remains in- 

 variable on reversing the position of the electrodes and their vases. In 

 order to remove all doubt as to whether the earth in which the electrodes 

 were sunk might not influence the results found when the circuits were 

 very long, I have been accustomed to make, at each of the extreme sta- 

 tions, four or five holes at a distance of 10 or 20 metres (33 or 06 feet) 

 one from another, and to proceed forthwith to the i^roof of homogeneity 

 bj- changing the position of the vases. 



I pass now to a description of the experiments made on such mixed 

 circuits as are much longer than those just described, on circuits, namely, 

 having the full length of a kilometre, (3,281 feet.) With a view to ex- 

 X^eriments of this kind, I selected a large, horizontal meadow, adjacent 

 to the Arno, and forming part of the Cascine of Florence. Tlie earth of 

 this meadow, at least to the depth of the 25 or 30 centimetres (10 or 12 

 inches) requisite for imbedding the vases of the electrodes, x)ossessed 



* Kilometre = 1093.6389 yards. 



