ON THE ABSOLUTE UNIT OF ELECTRICAL RESISTANCE 169 



grale ' as well as by the expansions in terms of the modulus after trans- 

 forming them hy the Landen substitution. 



The Circle. The circle whose constant we have called G" and which 

 was around the galvanometer whose constant was G, was a large wooden 

 one containing a single coil of No. 22 wire. 14 To prevent warping, it 

 was laid up out of small pieces of wood with the grain in the direction 

 of the circumference, and was carefully turned with a minute groove 

 near one edge in which the wire could just lie. It was about 5- cm. 

 broad, 1-8 thick and 82-7 cm. diameter. As the room had no fire in 

 it, the circle remained perfect throughout the experiment. The wire 

 was straightened by stretching and measured before placing on the 

 circle, which last was done with great care to prevent stretching; after 

 the experiment it was measured and found exact to T ' T mm. 



The circle was adjusted parallel and concentric with the coils of the 

 galvanometer, but at a distance of 1-1 cm. to one side, in order to allow 

 the glass tube with the suspending fibre to pass. The length of wire 

 was 259-58 cm. which gives a mean radius of 41-31344 cm. These data 

 give G" = -151925. Preliminary results were also obtained by use of 

 another circle. 



Chronometer. To obtain the time of vibration, a marine chronometer 

 giving mean solar time was used. The rate was only half a second 

 per day. 



Wheatstone bridge. To compare the resistance of the circuit with the 

 arbitrary German silver standard, a bridge on Jenkin's plan, made by 

 Elliott of London, was used. A Thomson galvanometer with a single 

 battery cell gave the means of accurately adjusting the resistance, one 

 division of the scale representing one part in fifty thousand. 



4 Thermometers. Accurate thermometers graduated to half degrees 

 were used for finding the temperature of the standard. 



The arbitrary standard. This was made of about seventy feet of 

 German silver wire, mounted in the same way as the British Association 

 Standard. Immediately after use, two copies, one in German silver and 

 the other in platinum-silver alloy, were made. It had a resistance of 

 about 35 ohms. The temperature was taken as 17 C. 



To obtain the accurate resistance of this standard in ohms, I had two 

 standards of 10 ohms and one of 1, 100, and 1,000 ohms. The 1-ohm, 

 and one of the 10-ohm standards, were made by Elliott of London, and 



u ln another part of my paper I have criticised the use of wooden circles for coil, 

 but it is unobjectionable in the case of a single wire, especially when the needle i& 

 suspended near its centre. 



