94 DENSITIES OF OXYGEN AND HYDROGEN 



c'oiinmted liy Sainvu to tliose wliidi winilil ulitaiii in tlio iiiiiiiiKliHc(l ('(|iiati<>ii, we 



A := 0.0020 



a = 0.0002 



But, as Sari'au it-inarks, iIk' [Houf tliat a lias a sensible inagriitiule is iiisurtioieiit. 



From Amagat's later series of observations,* no value of A can be computed 

 w Inch will satisfy the determinations at all piessui-es, as Professor William Ilarkness 

 lias assured me. That Nau der Waals' e(]uatiou needs modification is attested by 

 the more oi- less successful attempts of Clausius, Sarrau, and Amagat. But it will 

 answer our purpose, provided that we ooin[)ute its constants fi'oin observations at 

 pressures not far fiom those to which we desire to ai>i)ly it. If we compute A and 

 (I fiom various combinations of observations at thiee ditterent pressures, we get the 

 following values : 



VALUKS OK A AM) (I I- UO.M AMAGAt's KXl'KKIMKNTS. 



Pressures used in lompulation Values of A 



in Atmospheres. 



0.002284 

 0.002174 

 0.002127 

 0.002022 

 0.002006 



Values of a. 



— 0.000035 



-j- 0.000016 



0.000167 



( hily the i)Ositive, or the iiunit'rically .^nialliT negative, root is given. It is plain 

 that constant values of A and tr will not .satisfy the observations; we can scarcely 

 do better than to assunu', w illi Sarrau, 



.•\ = 0.0020 

 a = 0.0000 



Substituting these values in the expression for the [ncssurc of tlic nii.xturc of 

 two volumes of hydrogen .nut one volume of o.xygen when occupying three volumes 



we have 



P -|- p = 1.000389 



We may therefore coriect for the deviation of the mi.xed gjises from the 

 density computed l>y Boyle's law, by ninltiplying the observed den.sity by the 

 factor l.DOO;?S;); we thus lind 



0+ 2H — 3.001 167M 

 " ~ ' 1.000389M - H. 



* Aiitui/(s (if Chimie et dc Physique (5), 29, 68 



