( 556 ) 



tlie noi-nial voliinic ot' lliis iiiaiióiiictoi'. Tlie injiMomcter was i-cnio\p<l 

 tVom lli(' iiressiirc cn liii<l<'r and llic iioniial \ (tlimuM'cdrk'i'iiiiiiod, and 

 it appeared thai tVom -22.114 rv. (Mav 1» lÜOl) il had falh^i to 

 22.05(5 {'v. (Aug'. 23 '02). Aftoi- this Iho inaiKtiiietor has !)oeii cali- 

 hrated foi- the second time (calihration />)') and üINmI with hych'o.ücii; 

 ot' each ot' tlie isothennals (h'lerniined scune points haxc heen tested 

 bv ni(Nins of the hydroucn manonuMer. ('olninn (' of lahh' X\ uives 

 the mean rehitioii of the |)i'essiii'e nieasnr(Ml w ith the hy(h-oi>eii mano- 

 meter and tlial measnivMl with the air manometei-, where for the 

 normal volnme of the latter we have taken the mean between 

 those before an<l after the exi)erimejits. C'olnmn A uives the date, 

 B the temperatnre relating- to the isothermal. 



tablp: XV. 



Fi'oni colnmn C we cannot derive a regnlar conrse in this short 

 peiioil, so that for tlu>sc isothermals 1 ha\e ninl1ii)lied all the jires- 

 snres measnred with the aii- manometer by the coefticient J. 0018. 

 On .Inne 5, H and Ang. 21 points of the boi'dcr curxc have 

 been dctei-niiiuMl. These conld only be bronuht to liarnioni/,e with 

 those determined later with the hydrogen manometei' by niiilli|»lying 

 the pre.ssni'es b\ coefficients which are combined in the following 

 table together ^vitll that afore-mentioned. Hence this shows the course 

 of the variation during that periofl. 



If we compare these tignres with the values for the nornjal volnme 

 before and after the measurements, it ap|>ears that almost the entire 

 variation has taken place during this last period. From .May '01 to 



1) Comp. this Comm. t. p. 034. 



