EFFECT OF VARIOUS CONDITIONS UPON ECONOMY. 61 



it to a nearly closed position, so as to secure regular variations 

 of pressure over a uniform range. These tests show the 

 tendency to loss produced by an unsteady draught. They do 

 not show, however, what may happen in the ordinary work of 

 operating boilers. In the case given, the firing was good for 

 both methods of regulation. In ordinary work variations 

 occur not only in firing but in other respects, and the economy 

 of automatic regulation may be greater or even less than that 

 obtained in the special case cited. Owing to the important 

 part which the personal element of the fireman plays there 

 must be a large difference in the result produced in different 

 cases, and the tests on some large plants, referred to in the 

 memoranda relating to Boiler No. 10, one of which gave a 

 result favoring automatic regulation, and the other favoring 

 hand regulation, show this to be true. 



The use of an artificial draught, made by a fan supplying 

 warm air to the ash pit of Boiler No. 8, produced a result 

 which was not materially different from that obtained with a 

 natural supply of unheated air. The fuel was a mixture of 

 anthracite screenings and Cumberland coal. There was the 

 slight gain of 2 per cent, figured on the w r eight of coal used, 

 and less than 1 per cent, figured on combustible. 



The addition of heating surface to Boiler No. 2, by introduc- 

 ing pipes beneath the shell through which the feed- water 

 passes before it enters the boiler, seemed to increase the 

 evaporation per pound of coal 3.5 per cent. This gain does 

 not count for much, because at the time when the pipes were 

 introduced other changes in the line of improvement were 

 made, and this percentage simply represents the collective 

 result. It is noted in these tests that the temperature of the 

 escaping gases is low, being in the first test only 355 degrees. 

 Little benefit can be expected from an increase of furnace sur- 

 face in a case like this, where there is an indication, in the 

 small amount of waste heat, that the heating surface already 

 provided absorbs about all the heat that can be given up to the 

 boiler. 



The tests on Boiler No. 43 furnish an instance where the 



