( 638 ) 



meter is required, in the same manner as would be necessary if we 

 employed the method of constant volume. For, ^^'e have already 

 mentioned in § J that this compression apparatus is suitable for this 

 method. The measurement is made liv shutting l\^o under known 

 pressure and allo^ving• the compressed gas to expand through the 

 capillaries ƒ//' and 7/ into the volumenometer and reading as before. 

 This second xolumelric measurement, with its necessarv corrections, 

 gives the determination of the normal volume after the measurements 

 at high pressure and com})ares with the second normal volume 

 determination of the original method. 



§ 20. A?! accurate volumenometer. The volumenfmieter mentioned 



above in § 19 was designed to give isothermal measurements of an 



1 



accuracv of ui) to 60 Ats pi-essure. Hence the same accuracv 



10000 * ^ 



was desired as with the standard piezometers of Comm. n" 50 June '99, 



while at the same time the determination of the deviations from 



Boyle's law at ordinary pressures was kept in view. The most 



analogous apparatus is that employed bv Leduc; that of Witkowski, 



who has used a form more closely analogous \\'\{\\ ours, does not 



appear to have been designed for high accuracy. 



The measuring vessel E^ (PI. II tig. 1 and more in detail tig. 2) 

 where the gas is shut off by mercury entering through Ei^^^ consists 

 principally of 5 bulbs A\ such of 250 cc. and a smaller bulb j^\ 

 of 25 cc. capacity. These are separated by short really cylindrical 

 portions Ei^ .... Eb^ , on each of ^vIlich there is a mark, near to 

 which the mercury meniscus is brought for the measurements. 



At the lower end of the measuring vessel is a contrivance after 

 the scheme of Comm. n". 27, for catching any dust or stray gas which 

 may perchance come from the rubber tul)e at Cl-^ PI. II fig. 2. At the 

 upper end the vessel terminates in a capillary .tube 7^690 which is 

 divided almost immediately into two Eb^i and Eb^.^. One of them is 

 terminated near the apparatus by a cock r.. The other ends in a 

 glass or steel capillary terminated also by a cock. On PI. II fig. 2 

 e. g. the volumenometer is connected to the mixing apparatus F by a 

 capillary tube soldered on to it after it has been mounted in Ea by 

 the cock r^. At ^693 (PI. II fig. 1) e.g. r, the steel capillary ƒ//' 

 proceeding from the three way cock mentioned in ^ 19 may be 

 connected. The small bulb Z^a is calil)rated l)y mercury at the same 

 time as the larger and serves to determine the small volumes above Eb^^. 



To keep the temperature of the gas constant and uniform the 

 measuring vessel is firmly fixed to the bottom of the copper case ^„0, 



