168 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 11, NO. 7 



variations produced no measurable eflfect in the temperature of the steam 

 space, thus indicating the absence of both superheating and excess pressure 

 in the steam space. 



Extended experiments were made to determine the reasons for the fortui- 

 tous errors of about 0.005° usually observed in precise steam point determina- 

 tions. Apparently such errors are due to irregular fluctuations in atmos- 

 pheric pressure and to errors in measurement of barometric pressure, indicat- 

 ing that improvements in this respect will require the use of a closed system 

 and better temperature control of the barometer. 



The instrument described was exhibited and operated. 



Discussion: The paper was discussed by Messrs. White, Tuckerman, 

 SosMAN, Heyl, and Beall. 



T. S. Sligh, Jr.: Thermostatics. 



In this paper it is shown that the degree of temperature regulation attain- 

 able in a thermostated bath depends upon the thermal and space relations 

 of the various elements of the bath to an equal if not greater extent than 

 upon the characteristics of the thermoregulator alone. 



The three desirable characteristics of a thermoregulated bath are: steadi- 

 ness of controlled temperature, constancy of controlled temperature, and 

 range of regulation. 



Steadiness is measured in terms of the periodic variations of the instantane- 

 ous temperatures of the working-space as the regulator operates. It is de- 

 sirable that the amplitude of this variation be small and that the period be 

 short. 



Constancy is measured in terms of the variations of the time-average tem- 

 perature of the working-space as measured over a complete cycle of operation 

 of the regulator with variations of average rate of energy gain or loss from 

 the bath. 



Range may be measured in terms of the average energy input controlled 

 by the regulator or its equivalent, the number of degrees of change in sur- 

 roundings temperature for which the regulator is able to compensate. 



It is shown by reference to the characteristic equations for thermostatic 

 regulators as derived in the author's paper ^ that some gain is to be secured 

 by increasing the sensitivity of the regulator but, in view of present practice, 

 it seems that increased constancy, steadiness, and range can best be secured 

 by reducing the over-all lag of the system. This lag is made up of the lag of 

 the heater, the lag, or lapse, due to the time required for the water to circulate 

 from the heater to the regulator bulb, and the lag of the regulator bulb. The 

 lag of the heater may be made small by the use of a heating coil of small heat 

 capacity and large dissipating surface. The stirring lapse may be made 

 small by securing vigorous circulation and by placing the regulator bulb as 

 close to the heater as is possible without the inclusion of a temperature 

 gradient region between the regulator bulb and the working space. The 

 regulator lag may be made small by the use of bulbs of large surface volume 

 ratio and by securing a vigorous circulation of water past the bulb, thus 

 reducing the thermal resistance of the bulb surface. 



Great steadiness together with large range may be secured by the use of 

 the oscillating contact regulator which is described in the paper referred to 

 above. 



' Some characteristics of the Couy thermoregulator. Journ. Am. Chem. Soc. Jan. 1920. 



