418 EEPOKT— 1897. 



Experiments on the Condensation of Steam. Bij H. L. Callendar, 

 M.A.,F.R.S., Professor of Physics, and J. T. NicoLSON, B.Sc, Pro- 

 fessor of Mechanical Engineering, McGill University, Montreal. 



[Ordered by the General Committee to be printed in extenso.'] 



Part I. — A New Apparatus for Studying the Mate of Condensation of 

 Steam on a Metal Surface at Different Temperatures and Pressures. 

 By Professor H. L. Callendar, a7id Professor J. T. Nicolson. 



As the result of some experiments by electrical methods on the measure- 

 ment of the temperature changes of the walls and steam in the cylinder of 

 a working steam-engine, which were made at the McDonald Engineering 

 Building of McGill University in the summer of 1895, the authors arrived 

 at the conclusion that the well-known phenomena of cylinder condensation 

 could be explained, and the amount of condensation in many cases 

 predicted from a knowledge of the indicator card, on the hypothesis that 

 the rate of condensation of steam, though very great, was not infinite, but 

 finite and measurable. An account of these experiments was communi- 

 cated to the Institution of Civil Engineers in September 1896, and will, it 

 is hoped, be published in the course of the ensuing session. In the mean- 

 time, the authors have endeavoured to measure the rate of condensation 

 of steam under different conditions by a new and entirely different method, 

 with a view to verify the results of their previous work, and also to estimate 

 the influence, if any, of the film of water adhering to the walls of the cylinder. 



In considering the condensation of steam on a metal surface, it is usually 

 assumed that the surface exposed to the steam is raised up to the 

 saturation temperature corresj)onding to the pressure of the steam, and 

 that the amount of condensation is limited by the resistance of the water- 

 films to the passage of heat from the steam to the metal and from the 

 metal to the water. If the steam contains air, there may also be a 

 considerable resistance due to the accumulation of a film of air on the 

 surface, but it is comparatively easy to exclude this possibility in 

 experimental work. 



In the steam-engine experiments above referred to, it was practically 

 certain that the water-film due to the cyclical condensation never 

 exceeded one-thousandth of an inch in thickness, and that the resistance 

 offered by it was unimportant. At the same time, it appeared clear that 

 the temperature of the surface of the metal at its highest was considerably 

 below the saturation temperature of the steam, a condition which could 

 only be explained by supposing the rate of condensation of steam on a 

 surface to be limited by some physical property of steam itself, apart from 

 the resistance of the condensed film of water. Interpreted in this manner, 

 the experiments led at once to the conclusion that the rate of condensation 

 at any moment was simply proportional to the difference of temperature 

 between the saturated st^am and the surface on which it was condensing. 



The limit thus found was shown to be capable of explaining many of 

 the phenomena of cylinder condensation in a rational manner, but the 

 method by which it was establisned was of an indirect and somewhat 

 intricate character, and appeared to require some simpler and more direct 

 confirmation. 



If the rate of condensation of steam were really infinite, it should be 



