VALUE OF THE BRITISH ASSOCIATION UNIT OF RESISTANCE. 203 
The contrivance designed by Lord Rayleigh by means of which our 24-unit 
standard was compared with the B.A. units needs a special reference : it has been 
described by Lord Rayleigh in his second paper on the value of the B.A. unit, an 
abstract of which was read before the Royal Society on March 9, 1882, while the 
paper is published in the Phil. Trans., Part I., 1882. 
Five coils each of approximately 5 ohms resistance were wound and enclosed in a 
box, from which the two electrodes, copper rods with amalgamated ends, of each coil 
protrude. By means of two series of mercury cups this system could be put either in 
series or in multiple arc. Then, if each coil of the series is so nearly equal to 5 units 
that we may neglect the square of the difference, it is easy to show that the resistance 
of the system in series is exactly 25 times that which it has when in multiple arc. 
The coil of 24 units and a single unit were arranged in series so that they could 
readily be put into connexion with Fleming’s bridge. The set of five 5-unit coils in 
multiple arc was compared with a single unit. The connexions were rapidly altered, 
and the five 5 units in series were compared with the 24 + 1 ; then again adjusting the 
connexions, another comparison between the 25 units in multiple arc and the single 
was made. 
' In this manner a value of the 24 ohms was obtained in terms of the single ohm, the 
result of the comparison being as already stated. 
Thus, whether we use as the values of R 0 and S u those found from comparison with 
the box or those determined by comparison with the coils, we have as values of the 
B.A. unit determined from this second part 
Series A . "98673 .... mean of 4 
„ B . . . . "98598 .... „ 2 
„ C . . . . "98716 .... „ 3 
While the mean result is 
*986706 ohm. 
The result obtained in June, 1881, Part I., as the mean of three complete sets, was 
"986350 ohm. 
Our discussion has shown us that the possible errors of this determination are 
considerably greater than that obtained in Part II. We will, therefore, give to each 
experiment in Part I. only half the weight of an experiment in Part II., and obtain 
thus as our final value for the ohm 
"9S665 ohm. 
The value obtained by Lord Rayleigh in his second experiments with the rotating 
coil (Phil. Trans., Part I., 1882) is 
"98651* ohm. 
* Since this paper was read Lord Rayleigh has obtained by two modifications of Lorenz’ method 
the two values ‘9867 and ‘9868. 
