CARBON DIOXIDE AT DIFFERENT TEMPERATURES. 



205 



readings in the two calibrations differed by an amount corresponding to as much as 

 0'001 C. in the main experiments, and indeed the coils themselves were remarkably 

 accurate, the average error amounting to less than O'Ol cm. of bridge wire. A move- 

 ment of 10 cm. on the bridge wire corresponded to an alteration of O'l ohm in one of 

 the arms of the bridge, or to an alteration of 2 C. in the temperature of one of the 

 thermometers. The galvanometer used with the bridge was of the astatic type, 

 with a quartz fibre suspension, and the sensitiveness in the main experiments was 

 such that a movement of O'l mm. on the bridge wire caused an alteration of 10 

 divisions in the deflection on reversing the current. A reversing key was used 

 in the battery circuit throughout, so that 

 thermo-electric effects could be eliminated by 

 balancing until there was no alteration of 

 the galvanometer deflection on reversing the 

 current. 



The accompanying diagram, fig. 5, shows the 

 arrangement of the thermometer leads, the 

 leads which are drawn side by side being equal 

 in resistance. H and K are the two platinlim 

 thermometers, the smaller loops representing 

 the compensating portions. It will be seen that the arrangement of the leads is 

 such that the thermometers can be joined up to the bridge either separately or 

 differentially. 



The usual apparatus was employed for the determination of the fixed points. 

 Since it was on the thermometer H that the rise iu temperature in the main 

 experiments was measured, it was essential that this thermometer should be guarded 

 against all possible sources of irregularity. The eight leads associated with this 

 thermometer and the heating coil caused some difficulty at first in the accurate 

 determination of the fundamental interval, since they tended to conduct heat away 

 from the space surrounding the thermometer coil, but special precautions were taken 

 which successfully eliminated this effect. It would have simplified matters if the 

 heating coil had been wound on the prolongation of the axis of thermometer K, as H 

 would then have had only four leads associated with it. Full details of the determina- 

 tions of the fundamental intervals have been deposited in the archives of the Royal 

 Society, but the following is a summary of the results : 



THERMOMETER H. 



Experiment 1. Fundamental Intervals 

 2. 

 3. 



Mean ,, ,, 



Resistance at C. . 



5-1910 ohms. 

 5-1920 

 5-1917 

 5-1916 

 13-3761 



