626 Dr. W. M. Thornton on the 
agreement within 2 per cent. The method of testing was 
now as originally designed. The ring windings were con- 
nected to the generator ee or reversing and spring- 
break change-over switches *. The operations were then :— 
1. Make the circuit and Mine the current to reach a steady 
state. 
2. Short-circuit the windings and allow the current to fall 
to zero. 
3. Reverse the coil-connexions and repeat the process. 
This was done several times for each current, the quanto- 
meter readings being eventually reduced as in ordinary 
ballistic testing. 
The instrument is so easily w orked, and its readings are so 
constant when there is a strong field in its air-gap, that if it 
were better known it would no doubt be used for magnetic 
testing on a large scale in preference to any other. For 
measuring large changes of magnetization, these methods, 
the only two which are not ballistic, represent extremes. In 
the first with a recording voltmeter the best results are 
obtained when the current is varied as slowly and uniformly 
as possible; in the second with the quantometer, when the 
change is as rapid as the inductance of the winding and the 
tailing-off effect of the core currents will permit. The latter 
do oe affect the accuracy of either method. 
. The rings exhibit in a very marked manner the effects 
of ee currents and of change of permeability in the core 
on the growth of the magnetizing current. Ourves I,, L, 
fio. 3, were taken, using an oscillograph as a dead-beat 
ammeter, and recording the deflexions photographically, as 
previously described, but with the drum making one revolu- 
tion asecond. They are remarkable on account of the pause in 
the rate of growth observed in each, and for the difference 
between the two curves. I, corresponds toa molecular change 
from +B,, the residual magnetism, to + B,,, the maximum 
induction-density; I, from + B, to —B,,. The magnetic move- 
ment is therefore much greater in the second case, and the 
current is accordingly more retarded. 
In order to study the reaction of the core currents on the 
magnetizing circuit, it is necessary to compare the curves of 
fig. 3 with those obtained when the core is laminated. For 
this purpose a 20-kilowatt transformer was used, having 
about the same weight of ironastherings. The high- -tension 
winding was excited from storage-cells, and on closing the 
switch the current rose as shown in fig.4(p.628). Here the 
* Thid. vol. lii: p. 479, “On a master switch for Central Stations.” 
