( 1«<^ ) 



these as ordiiiates taking the corresponding line belonging to lo<j t 

 as abscissa axis. 



In one stri]) of the diagram going from the critical state towards 



L{pr) 



higher temperatures and larger vobimes the graphs ot . 100 



pv 



and — 100 are the same. On the larger \olnmes side of this strip 



V 



the first method was chosen and on the smaller volumes side the 

 second. It appears that in this way a com|)rehensive repi-esentation 

 of the ditferences between the isotherms may be obtained for the 

 whole region. ') ^). 



Ojüv boundary curves and diameters^) obtained from experimental 

 data are shown, and those whicii would bo ol>taincd froju VII. 1 

 are omitted. The reason for this is that the determination of these 

 curves from VII. '1 necessitating a very prolonged calcidation, has 

 not yet been completed. 



eillior by di-awins tangents or by linear interpolation between t!ie observations or 

 l)y calculation from Vll. i ; of the three methods the last is to be |)referre(). It is 

 now assmned llial llie VII. 1. isotherms are practically parallel to the experimental 

 isotherms — indeed, the last melliod is based upon Ibis bypotbesis - ;m(l we 

 may, therefore, write 



d {lofi v) Zl {l(Ki r) 

 d {pv) A (jn-) 



(see note 5 on page 159). This assumption is correct to sullicient approximation 

 and is legitimate at all events where it is only a (lueslion of giving expression to 

 systematic cbanges in the differences, so that rigorously carc^ has to be taken only 

 tlial lliey are always subjected to llie same perfectly detinilc treatment. Tli.' above 

 now becomes 



dilogv) ^ ^ f Ac 



d{pv) ' V '■ 



from which — ^ is easily obtained. 



V 



1) Since observations upon argon in the region of great densities are not yet 

 available we shall no further discuss the exact shape of this strip. 



~) In our cboice and development of tbis manner of presenting the results we 

 bave gralefully availed ourselves of the experience gained by Mrs. van Rheedt- 

 1 Portland, née Sillevis in earlier calculations and constructions. 



'■'') For abscissae of points on the diameters we have taken values of tbe quantity 



loQ 1l in whicb volumes are expressed in the tbeoretical normal 



f I 1 ^ 



volume as unit. 



