134 Dr. Schroeder van der Kolk on Gases, 



At 4° and 0*5 metre pressure, 



H 422-257 



N ... 30-1220x14=421-708 

 CO 2 . . . 191761x22=421-874 



At 100° and 0*76 metre pressure, 



H 422-337 



N ... 30-1386x14=421-940 

 CO 2 . . . 19-1806x22=421-972 



At 100° and smaller pressure the calculation could only be made 

 for carbonic acid ; as in the case of nitrogen, the value of k at 

 100° is only known for 0*76 metre pressure. 

 At 0-5 metre we get 



N 422-257 



CO 2 . . . 19-1918x22 = 422-221 



I believe that from this the conclusion may with certainty be 

 drawn, that Gay-Lussac's law of volumes is quite valid ; and that, 

 therefore, from the atomic weights the ratio of the densities for 

 the limiting condition of the gases may be accurately calculated. 

 It is true that this in general has not been disputed ; but when 

 a physical principle is to be considered quite accurate, and 

 therefore applicable as the starting-point of a theory, a new con- 

 firmation is never superfluous. 



It must here be expressly remarked, that the formulae ob- 

 tained give the values of k accurately only within the data taken 

 for the calculation. If the above values are calculated for a very 

 small pressure, for instance - 2 metre, or for 200°, greater differ- 

 ences are obtained, and sometimes even in opposite directions. For 

 air, the values of k at 200° and different pressures were found 

 almost exactly equal to each other ; the formulae of the preceding 

 paragraph give then far greater differences than at 100°. Yet 

 this does not disagree either with the theory or with the formulae 

 obtained. These do not in general give the real law of the 

 changes in k; the true formula is undoubtedly far more com- 

 plicated. Yet, within the limits of pressure and temperature 

 for which they are calculated, the empirical formulae are suffici- 

 ently accurate; that is, the deviations which occur are not 

 greater than the errors of observation. Within these limits 

 they may safely be used in testing a theory. 



If it were desired to calculate the formula with greater accuracy 

 and in greater generality for these gases', they might be sub- 

 jected to the condition of giving the theoretical density for pres- 

 sure = 0, and very high temperature. Yet this calculation would 

 be almost useless, and could not give much greater accuracy ; 



