380 ROMBERG. 



difficult, and need not be discussed. The value of Aq or (Ti— T\) — 

 (T'o— T2) is measured as (i?i— R'l) — iR'2— R2) on a Carey Foster 

 bridge. The changes in Ao thus measured are in terms of a tempera- 

 ture scale defined by R\ and R'o in their respective temperature 

 regions. They may be subsequently expressed in terms of a single one 

 of the thermometers used, and finally reduced to the gas scale by 

 comparing this thermometer with one that has been tested for its 

 " delta." These reductions involve a determination of the differences 

 between the temperature coefficients of the thermometers Ti and T^, 

 and between one of them and T3, the thermometer whose "delta" is 

 known. There is no need of accurate knowledge of the resistance of 

 any thermometer in any unit, nor of the relative resistance of any pair 

 of them, nor of their temperature coefficients. Nowhere in the work 

 is there any need of knowledge of the heat loss, of the heat capacity 

 of the apparatus, or of the work of the flowing water. No absolute 

 precision measurement is required in the entire calorimetric work, and 

 but a single such measurement, that of the Callendar "delta" of one 

 of the thermometers, in the entire work. 



The apparatus may best be described by following the course of the 

 water tlu-ough it. Water enters at A (see Fig. 3) from a regulating 

 valve from which it flows under a head of about eight inches, kept 

 constant by an overflow. It flows into a steam heater B, where it 

 boils. An arrangement not shown allows air to escape from the top 

 of this heater. The water is then cooled to about 80° by a cold water 

 circulation, and enters a coil of about 100 feet of 3/8" copper tubing 

 in the hot thermostat, w^here it is brought to the inflow temperature — 

 usually 78°. This temperature is kept constant by a regulator not 

 shown, which vaW be described later. From the thermostat the water 

 enters the thermometer tube Ti, which is lagged with wool, and is 

 placed in a 4" tube jacketed by the thermostat itself. It then passes 

 through a Dewar tube D, and a straight glass tube, into one of the 

 copper tubes of the thermal coupling, which consists of two tubes 

 coiled together in a double layer, in such a manner that the hot water 

 flows in alternate turns of each layer. It then passes up in tubes D'l 

 and r'l situated behind Di and Ti, and passes a second thermometer 

 inside the same jacketing tube. It then passes to the cooler, where a 

 tap-water circulation brings it to a temperature of about 12°. It then 

 traverses the cold thermostat coil, where it is brought to about 16°; 

 then the thermometer tube To, etc., and takes up in passing through the 

 second tube of the coupling a quantity of heat which may be con- 

 trolled by varying the amount of mercury in the bottom of the Dewar 



