EXPERIMENTAL STUDY OF HEAT LEAKAGE. 737 



taken from the bushing p and from a similar bushing on the high 

 side. 



All axial flow plugs were of this general form and were contained 

 in the plug case shown. The plugs used differed in the size of the 

 holes drilled in S, in the dimensions and position of the cross channel 

 in S', in the character of the lagging in the vertical tubes T and T' 

 (which was of talc in some plugs and of 'poplox' in others) and in 

 the dimensions of the chamber between S and »S". In certain plugs 

 this chamber, somewhat enlarged for the purpose, was filled with 

 copper baffle-plates, the object being to secure a more intimate mixing 

 of the steam in its passage through the plug. In other plugs, these 

 baffle plates were replaced by a number of layers of fine copper gauze. 

 In certain plugs, additional baffles consisting of brass plates with inter- 

 vening layers of copper gauze were located in the vertical tubes T 

 and T' just above the holes in the block S. Various combinations 

 of these baffling devices were tried, chiefly to see whether any notice- 

 able effect in the observed values of ixCp with varying intimacy of 

 steam-mixing could be detected. The results of these experiments 

 were in general negative, no consistent variations in the results on 

 fxCp being traceable to changes in the baffling. 



In the plugs in which ' poplox ' was used for lagging in the vertical 

 tubes, the poplox, which is a very light cellular material, was con- 

 tained between two thin concentric brass tubes (corresponding to the 

 inner and outer surfaces of the annular cylinder of talc shown in the 

 tube T of Fig. 1). The inner of these tubes was used to afford an 

 electrical connection between the heating coil and the insulating plug 

 /, the wire lead shown in the figure being dispensed with. This was 

 done because it was suspected that some part of certain rather large 

 discrepancies in the measured values of /jiCp might be due to an acci- 

 dental proximity of the lead wire, heated by the passage of the electric 

 current, to the bulb of the thermometer on one side or the other of 

 the plug. 



The plug-case was swung in a cradle which also supported the 

 secondary superheater, consisting in this case of a helical coil of about 

 40 feet of | inch copper pipe (actual inside diameter = 0.49 inch) 

 of about the height of the plug case and arranged so that it enclosed 

 the plug case. This coil of pipe was connected at one end to the high 

 side of the plug, at the other to a primary superheater (see Figs. 5 & 6) 

 from which the supply of steam was drawn. The cradle, with the 

 plug case, coil of pipe and two motor-dri\'en stirrers, was suspended by 

 means of a block-and-falls in the oil bath which has already been 

 mentioned. 



