200 Mr. J. E. Myers on a new Volumenometer. 



S.G. = -172!. Volume = 10*17 9 cub. centim. 

 •172 2 . 10-17 8 . 



•172j. 10-17 9 . Mean error = *002 cub. 



•172 3 . 10*17!. centim. 



IV. Determination of Specific Gravity of very finely 

 divided Cork Dust. 



A value greater than the one above given for cork was of 

 course expected. The cork dust was contained in a small 

 cylindrical glass vessel of about 3 cub. centim. capacity and 

 almost rilled it. The experiments prove, however, that the 

 volume actually occupied by the cork was only about 0*569 

 cub. centim. 



S.G. = '95 6 . Volume = *56 9 cub. centim. 

 •95 4 . -57J. 



•94 9 . -57 4 . Mean error ='002 cub. 



*96 2 . *56 6 . centim. 



•95 9 . -56 8 . 



•95 4 . -57!. 



A slight difficulty occurs when one is letting out mercury 

 from the vessel by opening the tap, in consequence of the 

 spasmodic change in level of the mercury in the capillary 

 tube. A more accurate method of procedure consists in 

 having the capillary tube graduated, and observing as 

 follows : — Read the position of the mercury column when it 

 is just above or below the standard position, then open the 

 tap to permit a small drop of mercury to escape, which is 

 separately collected and weighed, so that, after reading the 

 position of the mercury column, a simple linear interpolation 

 suffices to calculate the exact quantity of mercury which 

 would have escaped if the equilibrium had been exact. 



The following are successive determinations of the volume 

 of an iron cylinder, employing this mode of procedure : — 

 Volume = 6'50 7 cub. centim. 

 6*50 8 

 6-50 9 . 

 6'50 8 . 



6'51 . Mean error ='0009 cub. centim. 



6-50 9 . 

 6-50 7 . 



6*50 9 . Volume of cylinder deduced from 



6'50 8 . measurement = 6*50 9 cub. 



6*51 . centim. 



Mean=6»50 85 . 

 The instrument would probably be very suitable for the 

 accurate determination of the specific gravity of samples of 

 gunpowder. 



