414 ME. J. H. BEINKWORTH ON THE SPECIFIC HEAT OF STEAM AT 



lowered sufficiently to allow a full stream of gas to pass to the burner ; this tap was 

 then closed. As soon as the steam was generated it passed into the outer annular 

 jacket surrounding the calorimeter, and a small current of steam escaped through the 

 drain-tube, d 2 . The pressure gradually rose until the difference between the levels on 

 the gauge attached to M was the same as that between the levels of the mercury in 

 the regulator, when the tap, H, was again opened and the regulator put into action. 

 These conditions were maintained for about half-an-hour, in order that the copper 

 discs in the side-heating tube and the calorimeter itself might be warmed up before 

 passing steam into the inner jacket. At the end of this period the tap, X, fig. 4, was 

 slowly opened, thus allowing steam to enter the separator and to pass through the 

 throttle into the calorimetric apparatus. It was possible to obtain an approximate 

 reading of the temperature of the boiling-point of steam under the given barometric 

 conditions whilst this gradual heating up was taking place. For during this initial 

 warming of the calorimeter by the inflowing current of steam a small amount of 

 condensation necessarily took place in the flow-tube, this moisture being evaporated 

 at a constant temperature after the passage of the steam for a few minutes. 



At the end of about 40 minutes the thermometer reading was quite steady, and 

 after making allowances for changes in the barometric height, agreed to within 

 0'03 cm. of bridge-wire with the cold temperature reading taken after the heating. 

 In some of the earlier, and in all the final, experiments, these " initial cold tempe- 

 ratures " were recorded until the indicated temperature was steady for about ten 

 minutes, but since it was found that the cold temperatures obtained after the 

 heating (i.e., three-quarters to one hour later) agreed so well with the initial values, 

 in many experiments these initial readings were not recorded but only noted, so that 

 it was known when the calorimeter was heated up, and the conditions practically 

 constant. The electric current was then started through the heating coil, adjusted 

 to give the desired rise of temperature, and approximate balances noted on the 

 potentiometer. After another interval of from 20 to 40 minutes, depending on the 

 flow, the balance-point on the bridge again became steady to about 0'02 cm. or 

 0'004 C., and the main calorimetric observations were commenced. At a definite 

 instant a weighed half-litre flask was put under the end of the condenser, left there 

 for exactly ten minutes, in the case of the maximum flow, and then taken away. 

 Ten seconds afterwards it was replaced by another flask, and a second observation 

 of the flow during another ten-minute interval was taken. The duration of the 

 experiments with the medium and minimum flows was 15 minutes. In most cases 

 the amounts of steam condensed in two consecutive experiments, under as nearly as 

 possible identical conditions, agree to about 1 in 1000, but when the difference in 

 the observed flows was greater than this, due to some real change in the current of 

 steam, the variation had often been expected, since a change in the temperature 

 readings, not due to a change of barometric pressure nor to a change in the rate of 

 supply of electrical energy, had pointed to such an effect. The readings of the 



