450 Messrs. Guthrie and Boys on 



friction appears in the damping of the swing of the compass- 

 needle by neighbouring metallic masses, and by the evolution 

 of heat in the metallic disk revolving between magnetic poles. 



It occurred to one of us that such pursuit as occurs in 

 Arago's experiment might be made use of with advantage to 

 measure the rate at which machinery is moving. For we 

 have merely either (1) to connect a revolvable magnet with 

 the machinery, and measure the angle through which it turns 

 a copper plate in its near neighbourhood, which is restrained 

 by increasing torsion, spring, or weight moment, or (2) to 

 connect a copper plate with the machinery so as to revolve in 

 the neighbouroood of a magnet restrained by similar means 

 or by the earth's directive magnetism. The latter plan indeed 

 might seem to have the obvious advantage that, as in the tan- 

 gent-galvanometer, the amount of the effect is independent of 

 the strength of the magnet. But this advantage is probably 

 more than counterbalanced by the limitation of the angle of 

 deflexion to less than a right angle, whether the magnet swing 

 in a horizontal or a vertical plane. 



We have in the following experiments adopted the plan of 

 making a permanent steel or electromagnet revolve in the 

 neighbourhood of the conductor, and governing the motion of 

 the latter by the torsion either of a hair spring of a watch or 

 fine platinum wire. 



The accuracy of measurement which can thus be reached, 

 and the greatness of the effects, led us to the hope that the 

 electrical conductivity of liquids might be detected and 

 measured by the same means — a result much to be desired, 

 because, while the effect is due to the passage of currents in 

 the liquid, such passage is wholly unaccompanied by electro- 

 lysis and its attendant incubus, polarization. 



Before undertaking this latter investigation we have re- 

 examined experimentally the results previously obtained with 

 metallic conductors by others; and as our method differs 

 somewhat from those previously employed, and our results are 

 more extended and in some cases at variance with those of 

 former experimenters, we venture to lay them before the 

 Society*. 



Our oivn Experiments. 



Our experiments w T ere mapped out as follows : — 



Other things being the same, — 

 I. Vary rate of rotation. 

 II. Vary distance between the elements. 



* At the recommendation of the Publication Committee of the Physical 

 Society, we omit an historical summary of the previous investigations of 

 others. 



