QUARTZ FIBRES AND ITS TEMPERATURE COEFFICIENT. 425 



inside. K and N are two copper rods which serve as terminals for connecting the 

 platinum tube to the secondary of the transformer. They were bent, as shown in 

 the figure, and dip into two mercury cups. The other ends of the two water-jackets 

 are closed by indiarubher bungs, and the water circulates in the manner indicated by 

 the arrows in the figure. 



The part of the platinum tube to be heated is surrounded by a glass jacket of the 

 same diameter as the water jackets, two side tubes being provided for the wires of 

 the thermojunction. This jacket serves to keep off draughts. It is covered externally 

 with layers of sheet asbestos, which also assist in maintaining the part of the tube 

 surrounding the fibre at a uniform temperature. 



The platinum tube, with its accessories, is supported on a stand which rests on 

 levelling screws above a stout wooden lx>x, rigidly fixed to the wall of the laboratory. 

 The vibrator hangs inside this box, as shown in the diagram. The dimensions of the 

 box are 18 centims. square by 10 centims. deep, and there is a hole in the centre of 

 the top through which the end of the platinum tube protrudes. In order to make 

 an air-tight joint where the tube enters the box, and yet to place no restriction on 

 its expansion when heated, the hole in the box was covered with a piece of sheet 

 indiarubber, through the centre of which a small hole had taen made, and this fits 

 tightly round the platinum tube. The front and tattom of the lx>x can be removed, 

 and the suspended system can Ije lowered by means of the rod E so as to fix the 

 vibrator in position. There is a small window of optically worked glass in the front 

 of the box, and through this the mirror M is viewed. 



The only manner in which the platinum tube is held is by a clamp round the top 

 water-jacket ; it is thus free to expand in a downward direction, and no bending 

 occurs when it is heated. The glass jacket is held in position by a second clamp, 

 shown in the figure. 



The vibrator is set in oscillation by the usual device, already described, and its 

 temperature is indicated by the thermometer T, which enters through the side of tin- 

 wooden box and has its bulb about a millimetre distant from the cylinder. 



Before the experiments with the quartz fibres were begun, some experiments were 

 made to ascertain how nearly the temperature as indicated by the thermocouple 

 could be taken to be that of the inside of the tube ; and also how the temperature 

 inside the tube varied over the 10 centims. occupied by the fibre. For this purpose 

 a second junction of the same two metals, viz., pure platinum and rhodo-platinum, 

 was made and arranged on a mica frame, so that it could be placed at any desired 

 position inside the platinum tube. The galvanometer used was a sensitive one of 

 the d'Arsonval type. This was connected in circuit with a resistance of 1000 ohms 

 and a key by means of which t-itlu-r junction could lie included in the circuit. With 

 this arrangement it was found that the temperature inside was never very different 

 from that of the outside, the greatest difference obtained Ixjing 20 C. when the 

 platinum tube was at 1000 C. The temperature was also constant to within these 



VOL. cciv. A. 3 I 



