38 PROCEEDINGS OF THE AMERICAN ACADEMY 



25% of all the steam used by the engine, is not recovered in this 

 way, and only reappears during the back stroke, when it is worse 

 than useless, as it increases the back pressure which the returning 

 piston must overcome. Engineers have been led to this theory by 

 a study of indicator diagrams, that is, diagrams made automatically 

 while the engine is in operation, and recording the steam pressure 

 within the cylinder at every point of the forward and backward 

 stroke (see Figure 5). Such diagrams commonly show the weight 

 of steam in either end of a cylinder to increase in the latter part of 

 each stroke, as the piston moves forward after the admission valve 

 is closed, but the greatest weight of steam shown by the indicator 

 diagram at any part of the stroke is less than the weight of steam 

 and ivater passing through the cylinder at each stroke. Accord- 

 ingly, it is argued that a considerable part of that which comes 

 from the boiler as steam goes through the cylinder as water, at 

 least during that part of the stroke when it would, as steam, be of 

 service. 



But there have not been wanting some to maintain that the ob- 

 served, or apparent, effect, as shown by the indicator diagram, is 

 too large for the alleged cause. In January, 1889, Mr. Dickerson 

 of New York, a distinguished patent lawj'er, now dead, speaking 

 before the Electric Club of New York, advanced the proposition 

 that the peculiar character of the indicator diagram, which is sup- 

 posed to show cylinder condensation and subsequent re-evaporation, 

 is really due to leakage of steam past the engine valves, past the 

 exhaust valve in the early part of the stroke, past the admission 

 valve during the latter part of the stroke, the whole effect being to 

 make the expansion cwve, so called, of the diagram too steep at 

 the beginning, and too nearly horizontal at the end. He made the 

 statement that the steamers in the waters about New York City will 

 make four or five miles an hour with the valve between the boiler 

 and the cylinder closed. I do not know how accurate that state- 

 ment was, but I have never seen it contradicted. Mr. Dickerson's 

 argument excited my interest, and I cast about for some method of 

 studying the problem of so called cylinder condensation, not, as 

 engineers had done and are still doing, bj r means of the indicator 

 diagram, but through the cylinder wall by means more familiar to 

 the physicist than to the mechanical engineer. I wished to deter- 

 mine by actual trial how great are the fluctuations of temperature 

 occurring during a complete forward and back stroke of an engine 

 at- a given depth of metal from the surface touched by the steam. 



