HOW THE TESTS WERE CONDUCTED. 15 



filled, accurate count having been made of the whole number. 

 Subsequently the quantity of water contained in the two bar- 

 rels is ascertained by weighing, and the rating of the meter is 

 quickly determined by calculation. 



When an engine is fitted with a surface condenser, the meas- 

 urement of the feed-water can be somewhat simplified by col- 

 lecting the water discharged by the air-pump. In this case the 

 same kind of apparatus can be used for weighing ; but the two 

 tanks are reversed, the water being discharged first into the 

 reservoir, and subsequently drawn into the weighing-tank, which 

 is placed below it, and, after being measured, thrown away. 



An approximate determination of the feed- water consump- 

 tion can be made by water-glass measurement, assuming that 

 the type of boiler is such that this method is applicable. The 

 feed-water is shut off from the boilers for a half hour's time, or 

 such period as is permissible, and the rate observed at which 

 the water disappears in the gauge glasses. Subsequently the 

 volume consumed in the observed time is computed from the 

 known dimensions of the space occupied, and from this 

 the weight of the water which has been evaporated. If the 

 water line is effected to any extent by the condition of the fires, 

 it is necessary in making these measurements to observe great 

 care that the conditions of the fires are the same at the end of 

 such a trial as at the beginning. In some boilers the increased 

 activity of the fire causes the water line to rise, while the dead- 

 ening of the fire has the opposite effect. With the damper wide 

 open, and the fire barred up and in an open or free condition, 

 there is great activity of the fire ; while with the damper closed 

 and new coal applied, there is, for the time being, a very marked 

 reduction in the intensity of the heat. The position of the 

 damper and the thickness and general characteristics of the fire 

 should be the same at one time as at the other. It is best to 

 observe this precaution in all cases, even when there is no sen- 

 sible effect produced by these changes in the fire. It is also 

 necessary that the gauge glass and the connections leading from 

 the water column to the boiler should be clear; a condition 

 which can be secured hy blowing them out a short time (say 



