00 Mr. H. H. Poole on Thermal Conductivity and Specific 



supplied to the furnace ; C is one of the mercury cups 

 leading the current in and out ; the wires D are connected 

 to direct current 50 volt mains, and the contacts shown on 

 either side of A with two small electromagnets whose arma- 

 tures are, as shown, attached to Morse telegraph keys. 

 These magnets contain a great number of turns of fine wire, 

 so that they are energised by a very small current. The 

 Morse keys are held in the position shown by light adjust- 

 able springs not shown in the diagram. L is a 16 C.P. 

 200-volt carbon lamp used as a high resistance. M is a 

 permanent magnet motor, which is connected to a small 

 windlass made out of an old clock from which is suspended 

 the bundle of iron wires forming the core of the choking coil 

 controlling the main alternating current. 



The operation is as follows : — When the power is exactly 

 equal to that for which the torsion head of the instrument is 

 set, A is, as shown, midway between the platinum contact- 

 pieces on either side ; but if the power increases or decreases, 

 A makes contact with one or other, thus eneroisino- the 

 corresponding relay magnet and depressing the handle end 

 of its Morse key, so that the local battery drives M one way 

 or the other, thus raising or lowering the core of the choking 

 coil until the power returns to the right value, when the key 

 is released, thus short-circuiting the motor which stops very 

 quickly. The use of a large voltage and very little current 

 across the contacts at A makes the apparatus much more 

 reliable, as contact is made more certainly and there is less 

 sticking. For the most sensitive control the platinum wires 

 constituting the contacts should be sharply pointed at the 

 ends. This makes the contacts more reliable, but tends to 

 cause sticking and consequent hunting ; this w r as completely 

 cured by suspending the buzzer B (an electric bell with the 

 gong removed) from a small wooden arm attached to the 

 wattmeter. When either key is depressed B commences to 

 work, and the very slight vibration set up prevents any 

 sticking of the platinum points. B was only added towards 

 the end of the work, but is a distinct improvement. As 

 the cup C forms part of both the direct and alternating 

 circuits, it was found desirable to earth it at E through 

 a fuse. 



This apparatus worked most satisfactorily. The torsion 

 head was set to the required reading and the apparatus left 

 to adjust the power. The control could be made very close, 

 as the gaps between A and the platinum contacts could be 

 reduced to a small fraction of a millimetre. 



As before, the energy developed in the central wire was 



