30 REPORT—1 874, 
R by #'. As before, in the straight line ABD take BD to represent the added 
resistance R, and draw BC and DE so that the angles ABC and ADE are respec- 
tively equal to 6 and #'; and let BC and DE intersect in F. Draw FA to bisect the 
external angle EF'B; and let A be the point where the bisector of this angle cuts the 
original straight line ABD, Then the permanent resistance will be represented by 
AB, and the electromotive force by the radius of the circle drawn about the centre 
A so as to touch the straight lines BC and DE, the unit electromotive force being 
that which, in a circuit of unit resistance, would give a current capable of deflecting 
the sine-galvanometer employed through an angle of 90°. 
The experiments give 
e=pin Bao’ and c=sinp'=AB__ AE _ 
AB’ AD AB+BD’ 
AC and AE being both of them radii of the circle drawn about the centre A. 
Consequently 
yr AB 7+R AB+BD 
—= =, and = 
ee NC} e Ania’ 
whence 
r e R 
which proves the construction. 
It is evident that, with a perfectly constant battery, if a series of experiments are 
made by giving various values to the resistance R, all the lines drawn in the same 
way as BC and DE will pass through the same point in the case of a tangent- 
galvanometer, and will all be tangents to the same circle in the case of a sine- 
galvanometer. 
Suggestions for a Redetermination of the Absolute Electromagnetic Units of 
Resistance and of Hlectromotive Force, By Prof, G, CO, Fostur, F.2R.S. 
On Ohm’s Law. By Anruur Scuuster, Ph.D. 
Ohm’s law has often been subjected to an experimental verification. None of the 
methods employed, however, were sufficiently delicate to prove the law between very 
wide limits. The author thinks that the following method will allow us to judge 
with far greater certainty whether Ohm’s law is rigidly true. If we send rapidly 
alternating currents through a galvanometer they will not affect the position of the 
needle, as the two currents going in opposite directions will balance each other, if 
the circuit is inits normal state ee paper on unilateral conductivity, p.31). If we 
send in rapid succession two currents through the galvyanometer which have different 
intensities, they will have the same effect as one current, the intensity of which is the 
arithmetic mean between the two currents. Let the electromotive force of one of 
the currents be E+-2, of the other E—w; then, if the resistance is the same in both 
cases, the two currents will have the same effect as a single current produced by an 
electromotive force E. On the other hand, if this is not the case, the resistance for 
the electromotive foree E+z must be different from the resistance for the electro- 
motive force E—«. 
A weak constant current is sent through a galvanometer and a coil in which a 
magnet was rotated. The currents induced by the rotating magnet had no effect on 
the galvanometer-needle when the constant current did not pass. When the current 
passed, however, the rotating magnet always increased the deflection. In order to 
explain this result we are obliged to make one of the two following suppositions :— 
1. The resistance of the wire decreases as the current increases. 
2. The self-induction of a wire involves a term depending upon the strength of a 
current and approaching a limiting value as the current increases, 
