EXPERIMENTAL STUDY OF HEAT LEAKAGE. 



77 



It was deemed best to omit it because no reliable data for calculating 

 it are available. The question of the pressure coefficient of n is 

 briefly considered in a later section of this paper (Section IV, 2, h). 

 The pressure given in the title to Fig. 12 (4.02 kgm./cm.^) is the 

 average mean pressure for the nine types of plug. The average mean 

 pressure for the set-up S is also separately given. 



0.02 0.04 0.06 0.08 0.10 



^eoprocat offlo^/y hrs/kgm 



Figure 13. Curves of Fig. 12 shown with a common origin for comparison 

 purposes. 



6. Axial-flow Plugs Al, A2 and A3. 

 It will be noticed immediately (Fig. 14) that the plots of - vs. - 



y- J 



for plug A2 shows very well marked curvature. The plots for plugs 

 Al and A3 are not shown, but are similarly curved. The plots for 

 the other plugs (excepting possibly Ul) do not present certain evidence 

 of curvature which can be detected by the eye; but, with one excep- 

 tion (curve Ul), in all cases in which the constants of the plot have 

 been determined by the method of least squares, the residuals indi- 

 cate that the ix' line represents the observed values of fx' with con- 

 siderably more accuracy than the — line represents the observed 



values of ~- . One concludes, therefore, that the heat-leak in unit 



time is not even approximately proportional to the temperature drop 

 for the axial flow plugs, in which it is large, and that in general it 



