CARBON DIOXIDE AT DIFFERENT TEMPERATURES. 223 



were taken as before, as were also the various temperature readings. The taps T! 

 and T 2 (fig. 2) were then turned so as to cause the gauge to measure the excess 

 pressure at the high pressure ends of the tubes over that of the atmosphere. While 

 the oil was settling down in the gauge, the potentiometer reading across the cadmium 

 cells was again taken, and the barometer was read. The gauge reading having been 

 taken, the electric current was stopped, and the gas was allowed to flow through the 

 calorimeter to cool it down. At the end of about half-an-hour the balance on the 

 bridge became steady, and usually agreed with that obtained at the beginning of the 

 experiment to less than O'l mm. The balance point having been obtained, the bridge 

 connections were again altered, and the actual temperature of the gas was measured 

 with the thermometer K. The current of gas was then stopped, the throttles for 

 altering the flow were changed, and tested for leaks, and another run was taken in a 

 similar manner, the electric current being adjusted so that the rise in temperature 

 was almost the same as before. 



The experiments at steam temperature were performed in a similar manner, except 

 that the current of water round the calorimeter jacket was replaced by a current of 

 steam, which was allowed to flow for about an hour before the experiment was started. 

 In some of the experiments at steam temperature the steam was supplied from two 

 cans of water arranged in parallel ; when one can showed signs of giving out the 

 other was started. It was found that under these conditions the temperature 

 readings did not remain absolutely constant, for the pressure of the steam varied 

 with the amount of water in the cans. I consequently fastened a horizontal tube 

 into the side of a can, bent the end up vertically, and fixed it under a tap which was 

 allowed to drip upon it at a rate slightly greater than the rate of ebullition of the 

 water in the can. Thus the level of the water in the can was kept constant, the 

 excess of water running over the edge of the tube. The vertical portion of the tube 

 was of fairly small bore, so that if the rate of ebullition altered momentarily, as it 

 sometimes did, the sudden fall of water in the tube due to the fall in pressure did not 

 cause any appreciable quantity of cold water to suddenly rush into the can. 



(19) The Heating Effect of the Current Leads. A main run having been taken 

 in the manner above described, and the balance point on the bridge having been 

 redetermined with no electric current flowing, the external connections to the current 

 and potential leads were changed so that the current could flow through the leads Pp 

 (fig. 4). The electric current was adjusted to practically the same value as in the 

 main run, a,nd the rise in temperature of the gas was measured. The connections 

 were then altered so that the current flowed through the leads Qq, and the tempe- 

 rature was again measured. A large number of experiments were performed, details 

 of which are preserved in the archives. The following table gives the mean values of 

 (Bt/t) 100 for the large and small flows at steam and air temperatures; 8t represents 

 the rise in temperature due to the leads under consideration, corresponding to the 

 vise iu temperature t for the same current in the main experiment. The difference 



