ON CONSUMPTION OP FUEL AND PREVENTION OP SMOKE. 107 



Viewing the subject in this light, it will not be uninteresting if we attempt 

 to exhibit some of the important and exceedingly curious changes which take 

 place in the ordinary process of heating a steam-engine boiler. 



For these experiments we are indebted to Mr. Henry Houldsworth of 

 Manchester ; and, having been present at several of the experiments, 1 can 

 vouch for the accuracy with which they were conducted, and for the very 

 satisfactory and important results deduced therefrom. 



In giving an account of Mr. Houldsworth's experiments, it will be neces- 

 sary to describe the instrument by which they were made, and also to show 

 the methods adopted for indicating the temperature, and the changes which 

 take place in the surrounding flues. 



The apparatus consists of a simple pyrometer, with a small bar of copper 



t or iron (a in the following sketch) fixed at the extreme end of the boiler, 



and projecting through the brick-work in front, where it is jointed to the arm 



of an index lever b, to which it gives motion when it expands or contracts 



by the heat of the flue. 



The instrument being thus prepared, and the bar supported by iron pegs 

 driven into the side walls of the flue, the lever (which is kept tight upon the 

 bar at the point e by means of a small weight over the pulley at d) is at- 

 tached and motion ensues. The long arm of the lever at d gives motion to 

 the sliding rod and pencil^, and by thus pressing on the periphery of a slowly 

 revolving cylinder, a line is inscribed corresponding with the measurements 

 of the long arm of the lever, and indicating the variable degrees of tempera- 

 ture by the expansion and contraction of the bar. Upon the cylinder is fixed 

 a sheet of paper, on which a daily record of the temperature becomes in- 

 scribed, and on which are exhibited the change as well as the intensity of 

 heat in the flues at every moment of time. In using this instrument it has 

 been usual to fix it at the medium temperature of 1000°, which it will be 

 observed is an assumed degree of the intensity of heat, but a sufficiently near 

 approximation to the actual temperature for the purpose of ascertaining the 

 variations which take place in all the different stages of combustion conse- 



