846 Transactions of the American Institute. 



should first select measures to ascertain, in a given instance, 

 whether . the steam is economically generated. As has been said, 

 the heat-prodncing power or evaporative efficienc}^ of a boiler is 

 measured by the number of pounds of water evaporated per pound 

 of cojil from a given temperature, say 212 Fahrenheit. We have, 

 therefore, to weigh the water evaporated, and the coal producing 

 the evaporation, a very simple thing, apparently, but one about 

 which there is nmcli misapprehension, resulting in statements 

 grossly erroneous and ridiculous. The water may be measured in a 

 tank or barrel, the contents of which has been ascertained, by care- 

 ful measurement, or by weighing water into it of a given tempera- 

 ture. When experimenting, the water in the tank should be 

 pumped out dry, if possible, or at least to a given mark ; the pump 

 then stopped, the tank reiilled to the proper height (the easiest way 

 is to overflow it), when the supply, can be shut off and the operation 

 repeated. The supply pipe should be arranged so that the water 

 can be seen entering the tank, and leakage detected while the pump 

 is working. The better way is to have a hose, to throw in and out 

 of the measuring tank. Before making an experiment, it should be 

 ascertained if the boiler foams, or raises water; if so, it must be 

 remedied before proceeding further. All leaks about the tank, 

 pump and boiler should be stopped ; and all extra pipes leading 

 water in or out of the boiler be disconnected, or frequently exam- 

 ined. The steam generated may be worked off in the engine, blown 

 off through the safety valves, or otherwise disposed of, so long as 

 no water is lifted with it. The latter is less liable to happen when 

 the evaporation takes place under considerable pressure. The 

 greatest care is necessary in commencing and ending experiments. 

 There are several methods of doing this. The first is, to measure 

 the temperature and height of the water in the boiler, and immedi- 

 ately upon starting the fire, to keep an account of the fuel con- 

 sumed until the close of the experiment ; then to weigli the coal and 

 ashes hauled out of the furnace. This involves a calculation to 

 ascertain the heating effect of the fuel used in generating steam. 

 It is of little value, for the purpose of comparison, foi^the shell of 

 the boiler and its surroundings (often a heavy mass of brick work) 

 has also to be heated, and of this no estimate can be formed. Ano- 

 ther plan often adopted is to get up steam with wood, and allow it 

 to burn low, leaving only sulhcient fire to start the coal. The 

 experiment is started when the first coal is put in the furnace, and 



