Mr. J. E. Myers on a new Volumenometer, 199 



invariably be the case small holes have been drilled through 

 each of the screw-caps E and H, which holes are perfectly 

 closed by means of screws (not shown in the diagram) before 

 the application of pressure in each experiment. The internal 

 volumes of A and B are each approximately 16 cub. centim., 

 and the capacities of C and D as above stated 20 cub. centim. 



It is difficult to compare the instrument with those due to 

 Paalzow and Riidorff under similar conditions. In the present 

 instrument air under about two atmospheres' pressure is em- 

 ployed, while in previous instruments rarefaction has always 

 been resorted to. This alone suffices to produce increased 

 accuracy. Paalzow shows, in the paper above referred to, 

 that with his instrument a difference of pressure of 0*1 millim. 

 involves a change of volume =0*016 cub. centim. In Professor 

 Stroud's instrument the same pressure-difference requires a 

 change of volume =0*0023 cub. centim., — that is, the accuracy 

 is 7 times as great. Riidorff does not give the dimensions of 

 his instrument, so that comparison cannot be made in this 

 manner. From a series of results which he gives, however, 

 the calculated mean error =0*008 cub. centim., and this is 

 much larger than the error deduced from any of the series of 

 results given below. 



The following are specimens of some of the earlier results 

 which have been obtained. At Prof. Stroud's suggestion the 

 first uncertain figure (or estimation figure) in the result is 

 indicated in small type : — 



I. Determination of the Volume of an Iron Cylinder. 



v = 3*97 6 . 

 3-97 x . 



3*97 8 . Volume deduced from measurement =3*97 7 . 

 3*96 8 . 



II. Determination of Specific Gravity of CuS0 4 5H 2 0. 

 Weight of CuS0 4 employed was about 6 grammes. 



The following are the results of successive determina- 

 tions : — 



S.G. = 2'28 . Volume = 2-64 cub. centim. 



2*28 8 . 2-63 . 



2-28 4 . 2-63 5 . 



2*28 5 . 2-63 3 . Mean error =-003 cub. 



2-28 6 . 2*63 2 . centim. 



2*28 4 . 2-63 5 . 



2-28 . 2-64 . 



III. Determination of Specific Gravity of a piece of Cork. 

 The volume of cork employed was about 10 cub. centim. 



