Notices respecting New Books. 51 



Now the numbers which express the relation -— - increase regu- 



(1) 



larly with the pressure. Thus, for instance, the difference between 

 the elastic force observed and that deduced from Mariotte's law rises 

 to 1 14 millims. under the pressure of 9336*41 millims., and of 

 18548 98 millims. 



Nitrogen gas in like manner becomes more compressed than it 



(r) ■' 



should according to Mariotte's law ; and the relation ' augments 



(^) . 



regularly with the pressure ; it increases, however, less rapidly than 

 with atmospheric air. It is therefore very probable that oxygen gas 

 would depart still further from Mariotte's law than common air and 

 nitrogen. 



Carbonic acid gas departs very notably from Mariotte's law, even 

 from the point of a single atmosphere of pressure ; the relation 



y 



which is, for — i equal to 2, from 1'0076 when the initial 



pressure is a single atmosphere, becomes 1*0999 when the initial pres- 

 sure is 12*66 atmospheres. There are even still greater aberrations : 

 thus, when the volume 1 of carbonic acid gas, having an elastic force 



of 6820 millims., is reduced to the volume x-77:, its elastic force be- 



(r) 



comes 20284 millims., and the relation .p^ acquires a value of 



M772. ^ (p;) 



Mariotte's law therefore cannot be regarded even as an approxi- 

 mative law for carbonic acid gas, when this gas is observed under 

 rather high pressures. 



Hydrogen does not follow Mariotte's law better than the other 

 gases, but what is very remarkable, it departs from it in the con- 

 trary direction. Its compressibility, far from increasing, as in the 

 rest, diminishes in proportion as the pressure augments. The elastic 

 force of this gas is therefore analogous to that of a metallic spring, 

 which offers a greater resistance in proportion as it is subjected to a 



greater compression. The relation -,j^\- is for hydrogen always 



(pi) 



sensibly less than 1, but greater than 0*99000 within the limits of 

 M. Regnault's experiments. 



E2 



