134 ROYAL SOCIETY OF CANADA 



vaeuuiu and lbs. of feed water discharged were all plotted out as ordi- 

 nates on a time base. The lbs. of feed were also plotted out as ordinates 

 on a base of revolutions. By this method it would have been possible to 

 detect any inaccuracy of observations, especially with regard to the 

 important one of the amount of feed water used. In all the trials how- 

 over it has been found that the points representing the lbs. of feed are 

 remarkably near to being in a straight line, which affords satisfactory 

 evidence, that if conducted with care, short trials may be made to yield 

 at least as accurate results as long ones. 



The cards have all been measured by the ordinate method and the 

 means of the results obtained were used in drawing the combined 

 indicator cards shown in the plates. These cards however only represent 

 the means from the diagrams taken from one end of the cylinder, as 

 it was not thought advisable to obtain a mean of the cards from both 

 ends of the cylinder and expand it. It has been assumed that the 

 cylinder feed is divided equally between each end of the cylinder in 

 making the dryness calculations. Although the various ligures of dry- 

 ness given may not be the correct absolute ones, yet it is thought that 

 they truly represent the relative conditions of the steam throughout the 

 various portions of the stroke. The diagrams were drav/n because it was 

 thought they show much more readily, than is done by simple numbers, 

 the effects that the varying cylinder ratios, etc., have "upon the initial 

 condensation of the steam and its changes in dryness throughout the 

 different stages of its expansion. 



As there are so many diffei'ent wa3^s of showing combined cards, it 

 may be advantageous to describe the manner in which these particular 

 ones were drawn. 



It was first assumed that the steam at the beginning of compression 

 was dry and saturated. The pressure at this stage could be obtained 

 from the indicator diagram. The volume of the cushion steam was then 

 the volume of the cylinder that the piston yet required to sweep out be- 

 fore reaching the end of its stroke plus the clearance volume By 

 consulting steam tables its weight could be found, and also the volume it 

 would occupy at the admission pressure of the cylinder. This was done 

 for all the cylinders used in any particular trial. The line A B (Fig. 1, 

 Plate 1) was then drawn. To the left of it were set off these calculated 

 volnmes and from the jxjints thus ol)tained vertical lines were drawn. 

 For the L.P. cylinder the line C D was drawn and similar ones for the 

 other cylinders. This represented the amcmnt of steam that remained 

 permanently in each cylinder. To the I'ight of C I), etc., were measured 

 the respective clearance volumes of tlie different (ylinders and from these 

 points again were drawn vertical lines ; to the right of these last obtained 

 lines were measured the cylinder volumes. It will be noticed that in 

 some cases the card commences to the left and in others to the right of 



