94 Of the Management of Fire 



the width of these flues below is only 1 3 inches. The 

 walls of these flues are shown by dotted lines in 

 Fig. 17. 



The walls which separate the flues do not run quite 

 from one side of the boiler to the other; an opening 

 being left at one end of each of them, equal to the width 

 of one of the narrow flues, for the passage of the flame 

 from one flue into another, without its going from under 

 the boiler. 



The fire being made (on a circular grate) in the mid- 

 dle flue (see Fig. 1 8), the flame passes on in this flue to 

 its farther end ; and then, dividing to the right and left, 

 comes forward in the two adjoining side-flues. Having 

 arrived at the wall which supports the front of the boiler, 

 it turns again to the right and left, and, entering the two 

 outside flues, returns in them to the back of the boiler. 

 Here it went out (before the fire-place was altered) at 

 two openings left for that purpose in the wall which 

 supports the back part of the boiler, and the two cur- 

 rents of flame uniting entered a canal 7 inches wide and 

 1 6 inches high, which goes all round the outside of the 

 boiler. (See Fig. 20.) Having made the circuit of the 

 boiler, it went off by a canal (furnished with a damper) 

 into the chimney. 



From this description of the fire-place, it appears that 

 the flame and smoke generated in the combustion of 

 the fuel, in passing through those different flues, made 

 a circuit of above 70 feet in. .contact with the surface 

 of the boiler, before they were permitted to escape into 

 the chimney. This, I thought, must be sufficient to 

 give these hot fluids an opportunity of communicating 

 to the boiler all the heat they could part with, notwith- 

 standing the difficulties which attend their getting rid 



