: ' ATMOSPHERIC PRESSURE BETWEEN 104 C. AND 115 C. 421 



value has been used in reducing the mean specific heats to the value corresponding to 

 zero rise of temperature. 



Throttling Effects in the Calorimeter Flwv-Tube. 



During its passage between the heating coil and the thermometer the steam 

 expands adiathermally and its. temperature is lowered. This effect depends on the 

 actual temperature of the steam and, in consequence, the cooling will be different 

 on the hot and the cold flows ; it is greater on the cold flow. However, it is only the 

 change in the amount of cooling which affects the results systematically ; the correc- 

 tion is such that it raises the observed value of the specific heat. The pressure 

 drop between the heater and the thermometer was estimated from the indications 

 of the manometer, Q, and the variation of the cooling effect was calculated from 

 equation (6), p. 387. With the straight flow-tube calorimeters the difference in the 

 cooling effects is about 0'0035 C. for a rise of temperature of 10 C. in the case of the 

 maximum flow, and for the medium and minimum flows the differences are still 

 smaller. Hence the throttling effects in the straight flow-tube calorimeters are quite 

 negligible. An additional proof that this is the case is afforded by the results of the 

 experiments made with these calorimeters, with different distances, I), separating the 

 thermometer 1 from the heating coil. The mean values of the specific heat show no 

 systematic variation with D greater than O'l per cent, and this is of the order of 

 accuracy of the experimental measurements. With the spiral flow-tube calorimeter 

 the cooling effects were larger, for they depend on the throttling occurring in the 

 spiral, and the pressure differences between the heating coil and the thermometer were 

 greater than in the case of the experiments made with the straight flow-tubes. An 

 approximate calculation showed that the systematic error introduced, due to the 

 neglect of these throttling effects, could not amount to more than O'l per cent, on the 

 specific heat. Only one set of observations was made with the spiral flow-tube 

 calorimeter and these experiments are, in comparison with the others, unimportant, 

 and as the throttling correction is small and rather uncertain it has not been made. 



(I designed the spiral flow-tube to be of such dimensions that the throttling effects 

 would be negligible, but in the piece of apparatus supplied by the makers the spiral 

 tube was constricted along its entire length.) 



Specimen Tables of Observations. 



A complete set of the observations made on May 1st, 1914, with the silica calori- 

 meter, is included in order to show the steadiness of the experimental conditions. 

 The vacuum jacket was exhausted by means of charcoal in liquid air, and throughout 

 the three experiments no discharge could be produced in the Geissler tube connected 

 with the jacket. The order in which the observations were taken has been described 

 on p. 414. The values of S c , h, and k calculated from these observations,- after the 

 necessary corrections have been made, are 2'0309, 0'00470 and 0'00005 respectively. 



