ADMISSION OF AIR OVER FUEL. 55 



bustible, and this is the highest given in the paper for horizon- 

 tal tubular boilers. 



A fourth method consists in first passing the air back and 

 forth through passages running lengthwise of the walls of the 

 setting and then discharging it through perforations, part of 

 which are located on the top of the bridge wall and part on the 

 two sides of the furnace. The object sought is to supply the 

 air in a somewhat heated condition. Several tests are given 

 which show the effect of this method. Tests No. 11 and No. 12 

 were made for this purpose. A mixture of two parts pea and 

 dust and one part Nova Scotia culm was used, and in one test 

 airw r as supplied through the registers in the fire-doors, and in 

 the other through the passages in the walls. The use of the air 

 passages secured an increased evaporation of 2 per cent, per 

 pound of coal, while the result, when figured on combustible, 

 was the same in both cases. Vertical boiler No. 54 is provided 

 with air passages running up and down in the circular wall of 

 the setting, and discharging at numerous points over the fire 

 through perforations. Test No. 109, made on this boiler with 

 air admitted, gave 4.3 per cent, less evaporation per pound of 

 coal and 2.3 per cent, less per pound of combustible, than test 

 No. 110, made with air excluded. The fuel was a mixture of 

 two parts anthracite screenings and one part Cumberland coal. 

 Boiler No. 20, in which air was admitted in accordance with 

 this method, compared with an exactly similar boiler (No. 21) 

 in which air was not admitted, fired with a mixture of three 

 parts pea and dust and one part Cumberland coal, gave 4.5 

 per cent, less evaporation per pound of coal and 4.7 per cent, 

 less per pound of combustible. Boiler No. 27, with air 

 admitted and fired with Nova Scotia coal, gave 1 per cent, 

 higher evaporation per pound of coal and 1.5 per cent, higher 

 evaporation per pound of combustible than Boiler No. 26, 

 which was precisely similar, except that no provisions were 

 made for admitting air through the walls. In both of these 

 cases the registers in the fire-doors were open. They pre- 

 sented an open area of 22 square inches, or one-fourth of one 

 square inch for one square foot of grate surface. The openings 



