1875.) Physics. 129 
E is alternately positive and negative. The ends of the wire of the coil are 
connected with a commutator on its axis (as in an ordinary Delezenne’s circle), 
which causes the current in all external portions of the circuit to flow in only 
one direction. This being arranged, all that would be wanted, in order to 
know the resistance of the wire of the coil, would be to measure the strength 
of this current, and calculate the electromotive force by observing the speed 
of rotation. But it is plain that the resistance so calculated would include 
the resistance of the commutator, which is variable and wholly irregular, so 
some modification is necessary which shall entirely eliminate this source of 
error. Prof. Foster manages this by applying Poggendorff’s method of 
measuring E.M.F.—that is, by stopping all current in the coil with a branch 
of an auxiliary stronger current. A constant battery (shown in the diagram) 
has its circuit completed through a Siemens unit, a standard coil, or other 
conveniently-mounted resistance, A B, which it is desirable to know in 
absolute measure. A certain difference of potential exists between the ends 
of the wire A B, and this can readily be altered by changing the resistance of 
other parts of the circuit till it is precisely equal to the eleGromotive force 
generated in the coil when revolving at some constant rate. There is a 
delicate Thomson’s galvanometer, g,in the coil circuit, which shows when the 
two differences of potential are equal by the absence of all current through 
it; and when this point has been attained, the battery circuit is totally unaf- 
fected by its connection with the coil. The strength of the current in AB is 
now to be very carefully measured in absolute measure by an accurate abso- 
lute tangent-galvanometer, G; this current is produced by an electromotive 
force E (which we have arranged to be equal to the difference of potential 
between the ends A and B) acting against the resistance R of the wire AB; 
but' the electromotive force of the coil is also E, and therefore it could produce 
exactly the same strength of current against the same resistance R. Thus 
the strength of the current which the calculated eleCtromotive force E of the 
coil could produce against the required resistance R is measured when there 
is no current at all circulating in the coil. The absolute resistance of the wire 
AB is now known, for— 
2aH 
eee ee Eds ray Li Bitty 
C kHtand kéttand’ 
k is the constant depending on the particular tangent-galvanometer, and has 
of course been previously determined. The horizontal intensity of the earth 
is eliminated from the result, as it appears both in E and C; and this is im- 
portant, since it is a thing very difficult to determine accurately, and is liable 
to local disturbances. The advantages of the whole method which are pecu- 
liar to itself are—first, that the wire whose resistance is to be measured may 
be of the most convenient form and material for accuracy and constancy, and 
any wire whatever may be measured just as easily as any other; second, that 
the strength of the current is not measured by some arrangement inside the 
revolving coil, but by an independent accurate tangent-galvanometer, whose 
sensitiveness may be so adjusted that the needle may be deflefted about 45°; 
and, further, that by applying Bosscha’s modification of Poggendorff’s method 
(that is, after measuring as before, add a certain resistance to the wire A B, 
and then compensate it by a known addition to the external circuit, and so on) 
the E.M.F. of any cell may be compared dire&tly with that of the coil, and so 
found in absolute measure. In the discussion which followed, Prof. Guthrie 
and the President (Prof. Jellett) both doubted whether the irregularities in the 
resistance of the commutator were wholly got rid of by the method. Prof. 
Foster explained that, as when the measurement is to be made there is no 
current in the coil, a change of resistance can have no effect. If the circuit 
were properly balanced, the needle of the galvanoscope would remain at rest, 
whether the resistance of the commutator were 0 or «; the only thing is that 
if the resistance of the commutator were infinite, the coil would be disconne@ed, 
and there would be no experiment; so the effect of an increase of resistance 
at the commutator is to diminish somewhat the sensitiveness of the method, 
VOL. V. (N.S.) R 
