850 REPORT— 1892. 



densed steam is re-evaporated partially during expansion, but mainly during- 

 exhaust, and serves as a mere carrier of heat from boiler to condenser, in conditions 

 not permitting its utilisation in producing work. 



It became clear from Ilirn's experiments, if not from the earlier experiments of 

 Isherwood, that for each engine there is a particular ratio of expansion for which 

 the steam expenditure per horse-power is least. Professor Dery has since deduced 

 from them that the practical condition of securing the greatest efficiency is that 

 the steam at release sliould be nearly dry. In producing that dryness the jacket 

 Las an important influence. In spite of much controversy amongst practical 

 engineers about the use of the jacket, it does not appear that any trustworthy 

 experiment has yet been adduced in which there was an actual loss of efficiency 

 due to the jacket. In the older type of comparatively slow engines it is a rule 

 that the greater the jacket condensation, the greater the economy of steam, even 

 ■when the jacket condensation approaches 20 per cent, of all the steam used. It 

 appears, however, that as tl;e speed of the engine increases, the influence of the 

 jacket diminishes, so that for any engine there is a limit of speed at which the 

 value of the jacket becomes insignificant. 



Among steam-engine experiments directed specifically to determine the action 

 of the cylinder walls, those of the late Mr. Willans should be specially mentioned. 

 Mr. Willans' death is to be deplored as a serious loss to the engineering profession. 

 His steam-engine experiments, some of them not yet published, are models of what 

 careful experiments should be. They are graduated experiments designed to indi- 

 cate the effect of changes in each of the practically variable conditions of working. 

 They showed a much greater variation of steam consumption (from 46 to 18 lb. 

 per indicated horse-power hour) in different conditions of working than, I think, most 

 practical engineers suspected, and this has been made more significant in later experi- 

 ments, on engines worldng with less than full load. The first series showed that in 

 full-load trials the compound was superior to the simple engine in practically all 

 the conditions tried, but that the triple was superior to the compound only when 

 certain limits of pressure and speed were passed. 



As early as 1878 Prof. CotteriU had shown that the action of a cylinder wall 

 ■was essentially equivalent to that of a very thin metallic plate, following the tem- 

 perature of the steam, the exceedingly rapid dissipation of heat from the surface 

 during exliaust especially being due to the evaporation of a film of water initially 

 condensed on its surface. In permanent refjime the heat received in admission 

 must be equal to that lost after cut off. In certain conditions it appeared that a 

 tendency would arise to accumulate water on the cylinder surfaces, with the effect 

 of increasing in certain cases the energy of heat dissipation. Recently Prof. Cot- 

 teriU has been able to carry much further the analysis of the complex action of 

 condensation and re-evaporation in the cylinder, and to discriminate in some degree 

 Ijetween the action of the metal and the more ambiguous action of the water film. 

 I3y discarding the less important actions, Prof. CotteriU has found it possible 

 to state a semi-empirical formula for cylinder condensation in certain restricted 

 cases, which very closely agrees with experiments on a wide variety of engines. It 

 is to be hoped that, with the data now accumulating, a considerable practical 

 sdvance may be made in the clearing up of this complex subject. There are, no 

 doubt, some people who are in the habit of depreciating quantitative investigations 

 of this kind. They are as wise as if they recommended a manufacturer to carry 

 on his business without attending to his account books. Further, the attempt to 

 obtain any clear guidance from experiments on steam-engines has proved a hopeless 

 failure without help from the most careful scientific analysis. There is not a 

 fundamental practical question about the thermal action of the steam-engine, 

 neither the action of jackets or of expansion or of multiple cylinders, as to which 

 contradictory results liave not been arrived at, by persons attempting to deduce 

 results from the mass of engine tests without any clear scientific knowledge of the 

 conditions which have afi'ected particular results. In complex questions funda- 

 mental principles are essential in disentangling the results. Intt^preted by what 

 i« alrei'lv known of thermodynamic actions, there are very few trustworthy engine 

 tests which do not fall into a perfectly intelligible order. There i.- only one known • 



