130 Dr. Schroeder van der Kolk on Gases. 



Temp. Pressure. Hydrogen. Air. 



t= 4, *i = §3J=0-858, k, = 0-999945, £, = 1-000177, 



/ = 100, £ 2 =fJ§ = l-155, * ;/ =1000069, *a= 1-000601. 



t= 50, h 3 = l * y// = 1-000000, 



273°, as being sufficiently accurate, is here taken instead of 

 273°-15. 



We have now the proportion 



323*8-277 x 1-000177 : 373 x 1-000601-277 x 1-000177 



= 323-277 x 0-999945 : 373 x 1-000069-277 x 0-999945, 

 from which is obtained 



* 3 = 1 -000403 K. 



The above values of k ought properly to be multiplied by K, which 

 is constant for each gas. But as we have here only to do with 

 changes of k, I have omitted to do so ; K, in fact, entirely dis- 

 appears from the proportion. 



For calculating the empirical formula I calculated the values 

 for —87°. For a pressure of 2 metres we have, then, 



T Absolute 



lemp. tem p Pressure. Hydrogen. Air. 



-87°, 186, 2 k l =1-000429, 



4, 277, f^ = 2-980, *„ =1-000928, * 2 =0'997627, 

 100, 373, ff§=3-990, *, w = 1-001417, * 8 = 1-000286, 



and the proportion 



373*3 -277* 2 :277* 2 -186*j 

 = 373*„ / -277* y/ :277* // -186*,. 



from which k Y can be calculated. 



Although it follows from Regnault's observations that, in the 

 case of hydrogen, * does not change with the temperature be- 

 tween 0° and 100°, this is of course only considered to be proved 

 within the limits of the accuracy of these experiments. It is 

 not very probable that *is absolutely independent of t; for at no 

 temperature would hydrogen then approximate to the ideal gas, 

 as do other gases : the question is, whether an influence of this 

 kind is still imperceptible between such wide limits as —87° and 

 100°. If, however, * ; is too great, k lu must, owing to continuity, 

 be too small at 100°, since k n must be considered correct ; and, 

 inversely, the last two members of the above proportion will 

 then be either both too large or too small; and as 4° is almost 

 halfway between —87° and 100°, and the changes are in any 

 case very small, the ratio will not be perceptibly ahered. 



The following results are thus obtained : — 



