346 



REPORT 1887. 



again during the time the smaller current is on, the average temperature in both 

 cases being the same. The upper part in the diagram illustrates this. 



However with the same frequency in changing currents the average tempera- 

 ture, while the larger current runs, may be made very different from the average 

 temperatiu'e while the smaller current is on, by immersing the wire in liquid, for 

 the general temperature of the wire can then be only slightly above the temperature 

 of its surroimdings, so that the rate of cooling can no longer be considered the same 

 throughout. The lower part of the diagram is intended to roughly represent this. 

 The temperature curve is concave to the lower side while the larger current runs, 

 but is convex as the wire falls in temperature during the period of the smaller cur- 

 rent. The ' temperature effect ' if thus introduced would have to be met by increas- 

 ing the speed of the contact breaker, or otherwise smaller currents should be used, 

 which of course would diminish the refinement in the determination of ' A ' corre- 

 spondingly. 



Now this is similar to the case of a liquid electrolyte, the smaller arm of the 

 biidge rapidly losing heat by convection from its necessary proximity to the larger 

 bodies of liquid at both ends. The actual rise in temperature in one experiment 

 was ascertained, through the increased resistance, to be much less than ten degrees, 

 while in some of the determinations made with metals the wire fused dui'ing the 

 experiment so high a temperatiu-e was reached. 



The greatest speed found necessary for the contact breaker in the determination 

 with metal conductors, when ' h ' was ascertained to be less than 10~'-, was about 

 100 per second. The fastest of the three forks we have successively employed is 

 about 160 per second, but a much faster fork has been prepared, though as yet it 

 has not been got to work satisfactorily. 



The diameter of the smaller arm has been reduced considerably since last year. 

 One hole of "0027 cm. was bored with a specially prepared needle in extremely 

 thin mica. So that not much more can be done in this direction. The density of 

 the greatest current which could be used through this without ' heating effect ' 

 was about ten amperes per square centimetre. 



On the Resistance of Hydrated Salts. By Dr. B. Wiedemann. 



With a view to the settlement of the question, whether or no the conductivity 

 of a salt depends on the quantity of water to which it is attached in solution, the 

 conductivity of solutions of copper chloride at different temperatures has been 

 determined. At the lower temperatures the solution is blue, and at the higher 



