TRANSACTIONS OF SECTION G. 571 
cold junction of the thermo-couples. These couples give an electromotive 
force of 40 micro-volts per degree C., giving on a pointer galvanometer a 
deflection of 60 mm. for a rise of 20° C. It is thus possible to estimate the 
temperature rise of the bomb to at least 1° C. or to 33, of the total rise in 
temperature. 
‘he loss of heat from the bomb is due to three causes: (1) conduction along 
the constantan discs; (2) convection currents in the air-space surrounding the 
bomb; and (3) radiation from the walls of the bomb. 
The effect of the loss by conduction is only to lower the maximum tempera- 
ture obtainable as compared with what it would have been if the supporting 
discs were non-conductors. The losses due to convection and radiation are 
negligible as compared with those due to conduction, 
The coal under examination is placed on a small tray in the bomb, the latter 
being filled with oxygen at about 200 lb. pressure. The coal is fired electrically, 
the maximum temperature rise on the galvanometer taking place about a minute 
and a half after ignition. 
Experiments show that the gas pressure can be varied from 150 lb. to 250 lb., 
and the weight of coal burnt from 0:2 to 0°7 gramme, without any difference in 
the results obtained. The instrument is standardised by burning either sugar, 
carbon, or standardised coal briquets, the calorific values of which are known. 
The instrument is easy to operate, and gives consistent results. 
5. The Hysteresis Loss in Iron due to a combined Pulsating and Rotat- 
ing Magnetic Field.1 By Tuomas F. Watt, D.Sc., M.Eng., 
A.M.Inst.C.h. 
There is a considerable amount of experimental data available with regard 
to the hysteresis loss in iron due to a pulsating magnetic field, and due to a 
rotating magnetic field, each acting separately, but, as far as the writer is aware, 
no experiments have been made to determine the hysteresis loss due to the simul- 
taneous action of a pulsating magnetic field and a rotating magnetic field. 
Cases occur in practice in which the field pulsates and the iron at the same 
time rotates. An example of this is the case of a plain series single-phase com- 
mutator motor, in which type of machine the iron of the cores is subjected to 
the pulsating field, due to an alternating current in the windings, while at the 
same time the rotor core is rotating. ‘lhe question then arises whether it is 
permissible to ascertain the hysteresis loss due to each source of loss as though 
it were acting alone, and then add the two quantities in order to obtain the total 
hysteresis loss. Otherwise stated : Do the two hysteresis effects superpose without 
any mutual reaction? The experiments described in the paper were intended to 
decide this question. 
The‘ running down’ method was used to separate out the hysteresis losses, 
and the points on the ‘running down’ curves were taken by means of a strobo- 
scopic disc. The machine used for the experiments had ball bearings and was 
therefore particularly adapted for this method of testing. 
The first part of the investigation consisted in determining the value of the 
rotational hysteresis torque for a series of values of direct current in the stator 
winding, and a curve was obtained showing the connection between these two 
quantities. If, now, an alternating current of known wave form be supplied to 
the stator winding, the mean value of the retarding rotational hysteresis torque 
may be deduced from the above curve because, for each ordinate of the alter- 
nating wave of current the corresponding rotational hysteresis torque may be 
read off from the above curve, and thus a wave of hysteresis torque may be con- 
structed. The mean ordinate of this wave will give the mean rotational hysteresis 
torque for the given wave form of current, and is the quantity which is effective 
in producing the retardation. This assumes that the pulsating hysteresis torque 
in the rotor has ‘no influence on the rotational hysteresis torque. Of course the 
pulsating hysteresis itself cannot produce a retarding torque on the rotor. A 
curve was shown deduced in the way described and connecting the rotational 
miurtarenia torque, and the alternating-current amperes, supplied to the stator 
winding. 
1 See The Electrician, September 20, 1912. 
