Trans. N. Y. Ac. Set. 112 j^'eb. i 



J) 



various conditions, and finally, to show the method devised by the 

 author for laying down curves of efficiency, and to show what conclu- 

 sions follow from their study. 



It was shown that in no case, in steam engines as to-day constructed, 

 can the expansion-line or the curve of mean pressures be such as would 

 be obtained in a non-conducting cylinder. Steam must always be more 

 or less condensed at the beginnmg and must always carry away heat by 

 its re-evaporization at the end of the stroke. The steam-jacket checks 

 the first operation, but accelerates the last, and, with wet steam, may 

 scarcely decrease the evil that it is designed to prevent. 



It was stated that the adiabatic curve may be closely represented by 

 a regular curve of the hyperbolic class, jz^j 2/1" = p v'\ the exponent n 

 varying with the proportions of steam and water in the mixture' at the 

 commencement of the expansion, which is assumed to take place in a 

 non-conducting cylinder. Zeuner finds the value of n to be nearly n 

 = 1.035 + o-i '1^. -^ being the proportion of steam present. 



Table I, appended to the paper, gave the values at the ratio of mean 

 pressure to initial pressure ^, for various mixtures from steam i.oo, 



water o, to steam 0.50, water 0.50, assuming the formula to be practically 

 accurate within that range. 



With these are given the adiabatics for superheated steam, ;/ == 

 I-333- 



Pm 



Table II gave the values of -~ for steam-expansion in a jacketed 

 metal cylinder, in which it is kept just dry and saturated by heat from 

 the jacketed sides and ends ; the values for wet air compressed in air- 

 compressors, in which ;/ is frequently found to be 1.2 ; and for peculiar 

 cases in actual steam engines in which leakage cr re-evaporation, or 

 both, raise the terminal pressures greatly, giving ;/ = 0.50, n = 0.75. 



It is, as yet, impossible to predict which of these curves will be found 

 in ordinary engines, and the engineer is compelled to rely entirely upon 

 the "indicator" for information of this character ; this instrument 

 gives him a more or less exact graphical representation of the changes of 

 pressures and volume throughout the stroke. The greatest possible 

 variety of curves are found to occur in such cases, but they approach 

 the adiabatic more nearly, as the steam is drier and, as the speed of 

 piston is increased, rarely departing far from the common hyperbola in 

 good engines. 



Pm 



The values of -7^ given in the tables are plotted on Plates I and II, 



and from these a better idea can be obtained of the method of varia- 

 tion of mean pressure and of work done with variation of the ratio of 



