198 Management of Conical Boile7's, 



45 gallons. The whole contents of the apparatus are 65 gallons, 

 and the work performed by the boiler is about the amount 

 which, on an average, I allot to that size. 



It is unnecessary to detail, in a practical publication like the 

 present, the exact mode of conducting the experiments. It is 

 sufficient to say, that every precaution was taken to secure 

 accuracy. The whole of the fuel being weighed and noted when 

 put on, the temperature was ascertained by two self-register- 

 ing thermometers, one in the stove, and one in the open air. 

 The observations were made at 8 p.m. and 8 a.m., or nearly so ; 

 and the mean temperatures given are the mean of these tem- 

 peratures thus noted, and of the minimum which occurred by 

 night in the interval. The mean temperatures by night thus 

 obtained, during the continuance of the experiments, are as 

 follow : — 



Outer air 



- mean 33-85° 



- minimum 21-5°, 



Stove - 



- ditto 65-62'' 



- ditto - 61°. 



The mean temperature artificially maintained was therefore 



31-77°. 



The fuel employed in the first experiment was oven-burnt 

 coke of the best quality, and cinders. The cinders were the 

 ordinary refuse of the house, freed from dust and ashes by 

 being sifted through a cross-wired iron sieve, having 16 aper- 

 tures to the square inch, being one of the finest generally em- 

 ployed in gardens : every thing, therefore, larger than a pea, 

 was retained as fuel ; and such cinders are best adapted to these 

 boilers. 



The fire was lighted on February 10th, and continued burn- 

 ing for the fourteen days during which the experiment was 

 continued. The attention it received and the ordinary manage- 

 ment were as follows : — Between 6 and 7 a. m. the fire was 

 cleared, and clinkered if necessary, the fire-brick plug being 

 taken out for this purpose ; 4^ lb. or one gallon of coke was 

 then thrown in, and about an hour afterwards 9 J lb. or one 

 peck of cinders. This fuel lasted till 5 p. m., when the fire was 

 again cleared and made up for the night, by the addition of 

 9^ lb. or a peck of cinders. Slight variations took place in the 

 management, as occasion dictated. The first peck of cinders 

 was sometimes put on at twice, when the fire happened not to 

 be sufficiently burnt down to receive them at once : on some 

 days rather more was burnt, as will be seen by the average. 

 The excess of coke, however, there indicated, arose principally 

 from a large quantity having been employed at first, till expe- 

 rience showed that it not only burnt to waste itself, but uselessly 

 destroyed the other fuel. In a short time, the quantity above 

 stated was found to be uniformly sufficient ; and I have since 



