﻿932 Mr. A. H. Davis on Natural 



Experimental Observations. 



Certain general characteristic effects were noticed during 

 the experiments. 



With slight heating of the wire the conditions ultimately 

 set up were usually quite steady, except perhaps for a slight 

 drift with the more viscous liquids due to the fact that the 

 prolonged heating necessary before the steady state was 

 reached gradually increased the temperature of the whole 

 mass of the fluid. However, with greater heating, unstead- 

 iness supervened, the galvanometer deflexion and also the 

 current changing spasmodically, as though the convection 

 current had become turbulent or reached some critical 

 condition. 



Various measurements were made of the drift mentioned, 

 which was found to be of the order expected from the 

 known energy dissipation in the volume of liquid used. 

 Observations were taken while a heating current was 

 passing, sometimes of the decrease of current necessary as 

 time elapsed to keep the temperature of the wire constant, 

 and sometimes, with constant current, of the ultimate rate 

 of temperature rise of the wire. Observations were also 

 taken of the resistance of the cold wire just before the 

 heating current was passed, and again at such a time after 

 switching off that all local heating of the liquid had been 

 dissipated. In the case of glycerine this dissipation occupied 

 some minutes, and a true drift of 0*001 ohm during an 

 experiment appeared as great as 01 ohm when taken 

 too soon after switching off. Results were satisfactorily 

 consistent, and in general were equivalent to a change in 

 6 of the order of one quarter of one per cent, per minute. 

 The less viscous fluids attained the steady state too quickly 

 for this drift to have appreciable effect. 



Another effect was noticed with the more viscous liquids. 

 For instance, with the b-mil wire immersed in olive oil 

 slight heating currents gave a tendency for the galvano- 

 meter to drift in a certain direction as though the whole 

 mass were being gradually heated up. With greater 

 heating, however, an opposite drift was observed. This 

 was found to be due to oscillations having a period of a 

 minute or so, which slowly died out and left the usual 

 steady state. Thus, while on switching on a very slight 

 heating current the wire merely rose gradually in temper- 

 ature to its final value (2° C. excess), reaching this in about 

 three minutes, it was found that for rather greater heating 

 (5° C.) the temperature of the wire first overshot its final 

 value by about 1/8 ° C, reaching a maximum in about 

 one minute, and then fell gradually to its steady state. With 



