Evaporation of different Liquids at their Boiling-points. 45 



uniform ; because we are only concerned with comparative 

 measurements, which are simply proportional to the lengths 

 measured. 



For the same reason the standard wire of known resistance 

 is really unnecessary, and was omitted in all the remaining 

 experiments. 



•& 



Si / s 2 



^ & — 4 <k 



z. 



S x S 2 are the platinum spirals of the lamps. 



m x m 4 are mercury-cups by which connexion is made with S x and S 2 . 



B 1} B 2 are the batteries supplying the current in the primary and secondary 



circuits respectively. 

 C, D are the terminals of the fourfold potentiometer wire. 

 A is a sliding contact, and has to be moved until the fall of potential on 



the bridge between C and A is equal to the fall of potential 



between m 4 and m 3 , or m 2 and m x . 

 G is a simple reflecting-galvanometer. 



It became evident at once — working in these improved 

 conditions — that the whole method had been made very much 

 more sensitive and accurate than before ; hence it was of the 

 utmost importance to repeat the fundamental comparison 

 between alcohol and water from which all the other results 

 were calculated. These experiments are entered in Table IY. 



It was at this point that we gave up the idea of taking 

 water for our standard, and looked round for some other 

 liquid to use in its place. It is quite clear, from the numbers 

 obtained, that there is something in the water itself that 

 interferes with the experiments. Probably there is some 

 electrolysis during the passage of the current along the un- 

 insulated wire ; it appears impossible by any ordinary means 

 to get water so pure that its conductivity may be disregarded. 



