EXPERIMENTAL STUDY OF HEAT LEAKAGE. 



739 



The high side thermometer bulb may be placed at some convenient 

 point in the path of the fluid ahead of the cup, preferably as close to 

 the cup as possible. The whole is immersed in a bath, the tempera- 

 tiu-e of which is that of the fluid at the high side thermometer. If the 

 Joule-Thomson eftect in the fluid is positive, the conductive flow of 

 leaking heat (aside from conduction down the thin, long and poorly- 

 conducting walls of the cup — a relatively small item) is in the direc- 

 tion of the flow of the fluid, antl ultimately consists in the transfer 

 of heat from a thin layer of the fluid immediately adjacent to the 

 plug wall on the high pressure side to a similar layer on the low pres- 

 sure side. This transfer of heat obviously has no effect on the tem- 

 perature drop between thermometers after the steady state has been 

 established, provided the thin layer of fluid on the outside of the plug 

 is prevented from receiving heat through the metallic walls of the 

 plug case from the bath. Aside from radiation, which is doubtless 



Figure 2. Diagram of radial flow plug. 



slight, heat may be gained, from the bath, by the fluid at any instant 

 adjacent to the high side of the plug, by conduction, provided the 

 temperature depression on the high side due to conduction through 

 the plug gives rise to an appreciable temperature gradient at the walls 

 of the plug case; that is, provided the " thin layer" of fluid mentioned 

 above is sufficiently thick. If this is the case, any mass of fluid will, 

 of course, also acquire heat before reaching the high side of the plug, 

 while it is moving through regions in which the temperature is not 

 that of the bath. Theoretically, at least in the ideal case of an infi- 

 nitely thin plug of material having a finite thermal conductivity, the 

 thickness of the "thin layer" is zero*, in the steady state, and while 

 it is probably small, at least in the ^^cinity of the closed end in such 



* The analysis on which this statement is based ignores changes in volume 

 and specific heat due to the throttling, and also changes in temperature caused 

 by the increased average velocity of the fluid while passing through the plug. 



