216 



The effect of the speed is evident by comparing, for ex- 

 ample, Nos. 4, 8, and 14, where the load is the same number. 

 However, when it is equal, the effect per horse power must be 

 the same. By equating the values oi E, under this condition, 

 the relation of g and £ may be determined. Thus, Nos. 16 

 and 8 gave 



1-319 (</ + 5£)=l-736 (^ + 3£); 

 from which 



e=gx 0-3240. 



The mean of eight such* gives it = ^ x 0'3349 with no very 

 great discordance ; and substituting this value in E, the num- 

 bers given in the last column of the table are the result. 



Taking means of those that are adjacent, and arranging 

 them according to the revolutions of the turbine, per minute, 

 they give : 



Mean of 







Eevol. 









Effect. 



Nos. 3, 13 ... . 38-9 .... 3624 



4,12,3 







510 









4-035 



11, 19, 10 







62-4 









4030 



2, 18 







70-4 









3-824 



9, 17 







76-9 









3-709 



8, 16 







830 









3-504 



7, 15 







90-5 









3077 



6, 14 







96-6 









2-654 



It seems from this that the maximum effect is produced 

 when it makes about fifty-four revolutions per minute; and 

 that the effects at the speeds which correspond to those of the 

 brake experiments are nearly in the same ratio. 



Some practical results may be deduced. 



1. The close approximation of the values of - shows that 



* When the speed is not exactly the same, interpolation is used. 



