740 



TRUEBLOOD. 



plugs as have actually been used, the experiments to be discussed 

 later show that it is essential to shield the fluid from the effects of this 

 conductive action by interposing lagging between the plug and the 

 walls of the containing case (see Fig. 4). 



An important effect which is analogous in its origin to the effect due 

 to heat conduction through the body of the plug arises, if certain pre- 

 cautions are not taken, from the rather peculiar geometry of the radial 

 flow plug. At its open end, the plug is necessarily brought, for pur- 

 poses of mechanical support, into good thermal contact with the 

 metal walls of the case. As a result, fluid which has passed the low 

 side thermometer is provided with a thermally short path of communi- 

 cation with fluid which has not yet passed through the plug. This 

 results in a depression of the temperature of the latter, which, joining 

 the low side fluid at a temperature therefore lower than then exists 

 there, contributes to a still further depression of the temperature of 

 the fluid which has not yet passed. This cumulative action, similar 

 in principle to that taking place in certain types of liquid air machines, 

 is limited by the ability of the bath to supply heat to the low side 

 fluid. The reason that a similar action does not take place b^' con- 

 duction through the plug walls is that, at least near the closed end of 

 the plug, the direction of flow is parallel to the temperature gradient 

 in the plug for a considerable distance on both sides of the plug; 

 undoubtedly the effect does exist in some degree near the open end 

 of the plug, where the stream lines even a very short distance from the 

 plug walls are approximately axial. 



The effect of the action just described on the observed temperature 

 drop is indirect, since the fluid directly affected does not pass the low 

 side thermometer. But the temperature depression produced in the 

 neighborhood of the open end produces indirectly, by conduction 

 through the walls of the plug, the walls of the case, the inside lagging 

 and, doubtless, to some extent through the fluid itself, a depression 

 of the temperature of fluid which does pass the low-side thermometer; 

 that is, there is an outward leak of heat from fluid between ther- 

 mometers to fluid which has passed the low side thermometer, with 

 the result that the observed temperature drop is too high. 



This effect, of the existence of which experimental evidence is given 

 later (see IV, 2, d) will be called the 'regeneration effect' hereafter. 

 The remedy for it is to provide sufficient lagging between the metallic 

 support of the plug and the high side fluid. This lagging will be called 

 'internal end-lagging' or simply 'end-lagging' Avhenever it is neees- 

 sarv to refer to it in what follows. 



