and of Heated Air fur Electricity. 183 



have been before described clearly prove; but I v;as anxious 

 to establish the same flict from an examination of the state of 

 tension of the poles. To ascertain the tension of the poles I 

 employed the gold-leaf electroscope of Bohnenberger, which 

 presents peculiar advantages in experiments upon voltaic elec- 

 tricity. The pile usually employed consisted of 100 pairs of 

 plates mounted with pump-water; they were arranged in two 

 columns and cai'efully insulated. 



Erman observed that if one pole of a voltaic pile be intro- 

 duced into an alcohol flame, the other being insulated, and 

 if the flame be touched by a wire in connexion with the ground, 

 the tension of the pole which terminates in the flame will 

 cease, while that of the insulated pole will increase. Here 

 the flame conducts the electricity of the pole with which it is 

 in contact. But when the insulation of the other pole was 

 removed, I found that then the deviation of the leaf of the 

 electroscope attached to the pole which had been introduced 

 into the flame did not apparently diminish when the flame 

 was connected with the ground. This does not arise from 

 the flame insulating under these circumstances the electricity 

 of the pole inserted into it, but from its conducting power 

 being so feeble that it is incapable of removing the tension of 

 that pole as rapidly as it is acquired. If the wire by which 

 the flame is connected with the ground be insulated, and its 

 free extremity placed on the cap of an electroscope, the de- 

 viation of the gold-leaf of the instrument will indicate the pre- 

 sence of the same kind of free electricity as that of the pole in 

 the flame. On the same principle a feeble current may ac- 

 tually be conducted by the flame when the opposite poles of a 

 battery are inserted into it without any sensible diminution of 

 the tension of its poles. By obstructing the passage of the 

 positive electricity, so as to reverse the second part of Ernian's 

 experiment, the accuracy of this explanation was established. 



The surface of contact of the positive pole with the flame 

 was rendered as small as possible by employing a very fine 

 wire of platina, while on the negative side a coil of platina, 

 exposing a far greater extent of surface, was suspended in the 

 flame; but on introducing a wire in communication with the 

 ground into the flame, the tension of the positive pole ceased, 

 while that of the negative pole inci"eased. Fluid conductors 

 were then interposed on the positive side, but the result was 

 the same. The following arrangement was next successfully 

 adopted. 



A platina wire was hermetically sealed into one end of a 

 glass tube about ^'^-th of an inch in diameter and 8 inches long, 



