280 Profs. Dewar and Fleming on the Electrical 



time of making each set of measurements and the proper cor- 

 rection applied. Another correction of importance was that for 

 resistance of the connecting leads and the copper terminals of 

 each coil. A loop of copper wire was prepared of the same 

 metal and of the same total length as the standard size 

 employed as terminals of each resistance-coil. This blank 

 was always measured before and after each set of experiments 

 on the coils, and so arranged that it was under the same 

 temperature conditions as the terminals of the coils. This 

 together with the resistance of the short thick bridge leads, is 

 always called " resistance of connexions " in the Tables, and is a 

 correction invariably applied in each case. The galvanometer 

 employed with the bridge was a Pitkin-Holden Galvanometer 

 of the suspended coil type, having a resistance of about 10 

 ohms and nearly perfectly dead-beat. One cell of a Helsen 

 dry battery was used, and the usual battery and galvano- 

 meter keys inserted in their respective circuits. 



The temperatures above zero Centigrade were obtained by 

 immersing the coils in a beaker of melted paraffin oil standing 

 in a large zinc bath full of water. This water was heated to 

 varying temperatures up to about 98° C. The paraffin was kept 

 w T ell stirred, either by a mechanical stirrer or, better, by passing- 

 si owly bubbles of air through it from an elastic bag full of 

 air. The temperature of the paraffin oil was taken by a 

 mercury-in-glass thermometer, the scale-readings of which 

 were corrected for us by Mr. Casella by comparison with 

 those of a standard mercury thermometer which had been 

 checked at Kew. Temperatures close to zero Centigrade 

 were chiefly obtained during the long frost in January 1893, 

 by placing the coils in paraffin oil in beakers which were 

 placed in the zinc bath filled with broken ice. Coils waiting 

 to take their turn to be measured were kept standing in other 

 beakers of paraffin oil also placed in the same large zinc bath. 

 Those who have conducted similar experiments know well that 

 the great difficulty in such resistance measurements is to 

 make sure that the coils have reached the temperature of the 

 surrounding fluid and to avoid changing their temperature 

 by the current employed to measure them. For this reason 

 very weak currents have to be used with the bridge. The 

 temperatures below zero Centigrade were obtained by the use 

 of a bath of carbonic acid snow, and ether, or paraffin oil, 

 which gave temperatures according to desire from 0° C. down 

 to about —78° C, and a series of measurements was thus 

 made at about —40° and —78°. By the employment of 

 liquid ethylene a temperature of —100° C. could be reached. 

 Liquid oxygen was then employed as a refrigerating agent to 



