43ti 



1)K. CHAULKS II. I,KKS o\ TIIK HFFKl'T OF TEMPEfcATUfcE ON TIlK 



This method was in turn discarded, and it was eventually found that by revertini: 

 to the resistance method and using thin spirals of wire, instead of the straight wires 

 used previously, the difficulties due to the contraction of the medium were overcome. 



Description of Apparatus. 



The apparatus in its final form (fig. 1) consisted of two vertical narrow glass tubes, 

 AI, A 2 , 5 '8 centims. long, '17 ceutim. internal, "25 ceutim. external diameter, to the 

 upper ends of which were fused two wider vertical tubes, B,, Bj, 22 centims. long, 



'22 centim. internal, '30 centim. external 

 diameter. The tubes were fixed together so 

 that the axes of the narrower tubes were 

 parallel to each other, 79 centim. apart. 



Down each of the wider tubes passed two 

 bare No. 22 ('071 centim. diameter) copper 

 wires, separated from each other by a thin 

 glass tape extending along the centre of the 

 tube. To the lower ends of the two copper 

 wires in each tube were brazed the ends of 

 a piece of No. 29 ('033 centim. diameter) 

 platinoid wire, 1 1 '2 centims. long, bent at its 

 middle so that the two halves lay alongside 

 each other in the narrower tube, insulated 

 from each other by a narrower glass tape 

 continuous with that in the wider tube. 

 The space around the wires in the narrower 

 tubes was filled with sealing-wax, and the 

 lower ends of the tubes closed. In what 

 follows, these tubes will be called the heating 

 tubes. 



Round the centre of each narrow tube 

 was wound 14 '5 centims. of No. 40 ('0122 

 centim. diameter) platinum wire, P,, P 3 , obtained from Messrs. JOHNSON and MATTEY 

 as platinum thermometer wire. The coils of the spiral were at such a distance apart 

 that the total length of spiral along each tube was I "2 centims. 



The lower ends of both spirals were soldered to the end C of a No. 20 ('091 ceutim. 

 diameter) bare copper wire, which was surrounded by a thin glass tube of '22 ceutim. 

 outside diameter, fixed parallel to the heating tubes and 'G ceutim. away from them. 

 The lower end of this tube was bent towards the plane of the heating tubes, so that 

 the end of the copper wire to which the spirals were attached came into the plane of 

 those tubes. 



Fig. 1. 



