employed for determining the Ohm. 21 o 



A difficulty presents itself here, siuce induced currents are 

 produced in A on opening and closing the current which 

 flows through the combination of wires. We must therefore 

 either employ a constant current of such feeble intensity that 

 the heating of the circuit shall be imperceptible, or, according 

 to the method of Kohlrausch*, who has also worked out the 

 necessary calculations, we must employ an induction-appa- 

 ratus which sends alternating currents through the circuit. 

 A special source of error is found in making contact with the 

 mercury column contained in an accurately-calibrated care- 

 fully cleaned glass tube. The ends of the tube terminate in 

 glass vessels of diameter relatively large. The apparatus is 

 best filled with mercury by pouring mercury into one of the 

 vessels and inclining the apparatus towards the same side, 

 then, after closing the openings, repeatedly exhausting the 

 apparatus as completely as possible, gently warming the tube, 

 then allowing pure air which has passed over phosphoric 

 anhydride and wadding to enter, and finally, after exhau sting- 

 again, allowing the mercury to flow into the tube by placing 

 it in a horizontal position. If the electrodes are immersed in 

 the vessels containing mercury, then the resistance of the mer- 

 cury in the tube is increased by that of the mercury contained 

 in the vessels up to the ends of the electrodes. This resistance 

 may be calculated on the assumption that the vessels are infi- 

 nitely large. It might perhaps be advantageous to determine 

 it directly by compensating the resistance of a tube filled with 

 mercury with two contact-vessels by a nearly equal resistance 

 on Wheatstone's bridge, and then to cut the tube at one or 

 more points and to introduce wider mercury vessels of suitable 

 form, and then again determine the resistance. We might 

 also introduce two tubes, of lengths m and n and of equal sec- 

 tion, between the same contact-vessels, and compare their 

 resistances, and from the data so obtained calculate the resist- 

 ance of the vessels. 



In order to avoid this separate comparison of the resistance 

 of the coil determined in absolute measure with a mercury 

 column, Carey Foster f and Lippmann J have almost simul- 

 taneously suggested (L.) or carried out (C. F.) the ingenious 

 plan of employing "Weber's fourth method, as adopted by the 

 British Association, and PoggendorfFs compensation-method 



* Poggendorff ; s Annahn, 1871, vol. cxlii. p. 418. 



t Carey Foster. ' Electrician.' 1881, vol. yii. p. 266 ; Bciblatter, to!. \i. 

 p. 133. 



X Lippniann. Comptes Bendus, 1881, vol. xciii. p, 713 ; Beibllitter, 

 vol. vi. p. 43. 



Phil. Mag. S. 5. Vol. 14. No. 88. Oct. 1882. T 



