BA r.D STEAM, ASU O7 iPOJUL 17 



in which A is the reciprocal of the mechanical equivalent of beat, T U toe 

 "m the absolute zero, and * is the volume of one unit of weight 



of tl from wuich the vapor ia formed. The differential co-efficient 



can be calculated by aid of the equation* on page 1 1 . 



temperature ia obtained bj ado 7 to the 



rees Centigrade, or 460.7 to the temper: legrees Fi 



volume* and denitk of saturated steam given in Tablea I, II. 

 ! I were calculated by this method. 



It is ,f interest to consider the degree of accuracy that may be expected 



method of calculating the density of saturated vapor. The value 



of r ropcnds on A a; f the first, Regnault gives three figures hi the data 



from whirl, t!i- i-inpirirnl equation is deduced, and the experimental work 



does not indicate a greater degree of accuracy. The fourth figure, if stated, 



unite. The value of 7 is com* 



urea, of which the last may be in error by two unit*. 



A, as !: i iiimr.i l-\ How-land. lias four figures, the last being uncertain to 



the extent of one or two units. The differential co-effi< ? ia deduced 



from ions for calculating ;>; and those equations are derived from 



English system give a pressure of 14.6967 



|M>umU nn the square inch; 1-ut the s^citic volume calculated by : 

 K-lua' .ilr Filiation C - 



The mean. - !rm either ty aln.ut one in seven luu 



i *luc to the f:i s represented by EquaUona 



/> aim c common tempcratun it i. not have a common 



tangent. - .- filiations are cmpnirnl n:,-l \.--i l..-jir:l. t! 



11 ( alcuhited specific volumes are affec- 

 y. The greatest probable error is in ilrt.-rmS: 

 which it may IK? about one in one thousand. The error intrtxlucvtl into thi.i 



.- tlu- values of A in common use, that is, 772 instead of 

 is abo 



Tata and Fairbairn's Experiments. In 1860 an attempt was made by 

 i ic the >|.e.-i:ic \-.. in MI.- <>r .steam by direct 



, taken f ron Tranmtin**. 



results of all their experiineuU, to. U the rnhm t 



calculated by their empirical formula, 



49&18 



