394 PROF. H. L. CALLENDAR, PREFACE TO 



SUMMARY of Observations with Vacuum Calorimeters (continued). 



The value found for h increases, as one would naturally expect, when the distance 

 between the heating coil and thermometer is increased. Conversely h may be 

 diminished by reducing the distance, but it is necessary to allow sufficient space 

 for the mixing of the steam to a uniform temperature, and it was found 

 that D could not be reduced below 6 cm., even with the most effective spiral 

 mixing of the flow. Mixing the fluid by passage through wire gauze, which has 

 been commonly employed by other observers, was found to be very ineffective as 

 compared with the turbulent flow induced by the spiral. The maximum value of Ji 

 observed at a distance D = 11 cm. was only 3'5 per cent, of the heat supply for the 

 largest flow. The average was about 2 per cent., but none of the jackets maintained 

 a good vacuum when heated in steam, and they all cracked after a few days use. The 

 term &/Q 2 was generally less than 1 in 1000 for the large flow, but increased to nearly 

 1 per cent, when the distance was reduced to 6 cm. 



Since the measurement of the electric energy is made in watts, it is convenient to 

 calculate X and S in joules or watt-seconds per gr. degree. The values of X are 

 always a little greater than 2 for steam, and since the formulas involve only differences 

 between corresponding values of S and X, it saves a good deal of trouble in calculation 

 to work only the differences from the round value 2000 in each case. It is important 

 for the same reason to keep the differences between the flows as large as possible. 

 The observed values of X : were generally consistent to 1 in 2000 or better, but an 

 error of 1 in 2000 in the value of X : or X 2 would make an error of 3 or 2 in 2000 in 

 the result calculated from 3 flows. The order of agreement of the results, which 

 range_ from 2'052 to 2'070 is, therefore, as good as can be expected, considering the 

 range of variation of the conditions of the experiments. 



The results of this particular series of observations were reduced in the firsi instance 

 by employing the theoretical formula (7) for the variation of the specific heat over 



