734 TRUEBLOOD. 



I. INTRODUCTION. 



Some time ago, the writer undertook the experimental determina- 

 tion of the Joule-Thomson effect and of the product of this by the 

 specific heat at constant pressure, in- superheated steam, as functions 

 of pressure and temperature. The theoretical and practical impor- 

 tance of such measurements as these has been fully set forth in a recent 

 paper by Da\is.^ In the experimental work, it became evident 

 with the first results obtained that the principal problem was the 

 reduction of heat leak to such a point that its effect might be elimi- 

 nated without an excessive number of observations under varied 

 conditions of pressure drop and flow at a single point of the (p, T) 

 plane. The experimental study of this problem developed chiefly 

 in the direction of making measurements of n, the Joule-Thomson 

 coefficient, with plugs of different types, under a considerable variety 

 of circumstances in respect of the arrangement of the containing case, 

 lagging, etc., the determination of the product nC^ being temporarily 

 abandoned. These experiments have involved altogether nearly one 

 hundred separate measurements of ju with three distinct types of 

 radial flow plug case, one of which was especially designed for the 

 purpose of studying the effect of variations in lagging, and with one 

 type of case for what may be called, for the sake of convenient 

 distinction, an axial flow plug. Of these measurements of n, all but 

 fourteen were made at practically the same pressure and temperature. 



The chief intent of the present paper is the presentation and dis- 

 cussion of the results of these measurements of ju at a single point of 

 the plane, as a study of the question of heat leak in throttling calori- 

 metry, and particularly as a study of the behavior of the radial flow 

 plug as applied to throttling calorimetry; for while a considerable 

 proportion of the runs were made with axial flow plugs, the general 

 type of plug and containing case used in all of these was of an early 

 design, in which the heat leak was unnecessarily large, so that the 

 interest these axial flow runs have for the purposes of this paper lies 

 chiefly in the fact that they illustrate rather strikingly the heat-leak 

 possibilities in throttling apparatus not designed primarily to avoid 

 heat-leak, rather than in any basis they afford for a fair comparison 

 of the relative merits of correctly designed axial flow and radial flow 

 plugs. 



1 Davis, Phys. Rev. (2) 5, 359 (1915). 



