ELE("rRICAL ACTIVITY OF <JRE BODIES. 329 



partiiil circuit, moreover, remaining' fixed — is tested once f'oi- all before com- 

 mencing the experiments. 



It is often desirable, before inserting- the Daniel!, t(» determiiu; wlu'tlier 

 the circuit is in order and without a break. This may be easily accom- 

 plished by touching- with the linger a copper part of it, so that a secondary 



circuit, T. C, wire, galvanometer, wire, body, earth, T. C, or F. C, wire 



body, earth, P. 6'., is produced, respectivel}'. The electromotive force acting 

 in this case is that of zinc-copper, but in consequence of the very large 

 i-esistance of the finger contact the current, though distinctly i^erceptiljle, 

 is too weak to produce any appreciable polarization. 



In spite of all these safeguards, however, a close inspection of the 

 recorded values still revealed discrepancies which had not been avoided. 

 Accordingly the method of procedure was further improved by the follow- 

 ing additions: To eliminate as much as possible the effect due to the termin.il 

 bags, a variation was introduced by which the results from different bags 

 could be compared. Four of these. A, B, C, and D, were generalh' employed, 

 which, when combined, two and two, in the maimer shown in the diagram, 

 gave three separate and distinct values for the lode electromotive force e. 

 The electromotive force between any two bags, A and B, is represented b\- 

 AB, between A and C by A\C, etc. 



Original position. First exchange. Second exchanfje. 



After the second exchange, the bags again have their original posi- 

 tion with reference to the holes. The corresponding measurements, there- 

 fore, check one another, while from their mean any linear variation of their 

 own electromotive force is eliminated. Each series gives a value for e. 

 With this method of triple measurement the series was completed by deter- 

 mining all the electromotive forces between P. (J. and each of the T. C.'s, 

 starting with the one nearest P. C. and ending with the most remote. After 

 tills the whole set was again repeated, starting, however, with the extreme 



