Manchester Memoirs, Vol. xliii. {\Z(^g), No. 4.. 5 



hard, elastic spheres, frictionless when in collision with 

 other molecules, or with the walls of the containing vessel. 

 It does not appear probable that molecules are really of 

 such a form, or possessed of such properties. But there 

 can be no doubt whatever, that, granting the premises, the 

 conclusions will follow. They are derived from common 

 mechanical conceptions. 



Until the year 1894, the only gas of which the ratio 

 of specific heats attained the value of i "66 was mercury 

 vapour. The measurement was made by Kundt and 

 Warburg in 1868. Now, there are other reasons in the 

 case of mercury which render it probable that its molecule 

 is of a simpler nature than that of other gases. We are 

 accustomed to compare the density of gases with that of 

 hydrogen, taken as unity. And as, according to the law 

 discovered by Avogadro in 181 1, equal volumes of gases, 

 at the same temperature and pressure, contain the same 

 number of molecules, the density of a gas, compared with 

 that of hydrogen, shows the relative weight of its molecule, 

 compared with that of a molecule of hydrogen. Now, 

 mercury -gas is too times as heavy as hydrogen, and hence 

 its molecule is 100 times as heavy as a molecule of 

 hydrogen. But the specific heat of liquid mercury is one- 

 two hundredth part of the supposed specific heat of solid 

 hydrogen ; and, as the specific heats of elements are known 

 to stand to each other in the inverse ratio of their atomic 

 weights, it follows that the atomic weight of mercury is 

 two hundred times that of hydrogen. The atom of mer- 

 cury is then 200 times as heavy as the atom of hydrogen, 

 while the molecule of mercury is only 100 times as heavy 

 as the molecule of hydrogen. From this it follows that a 

 molecule of mercury must contain only half as many 

 atoms as a molecule of hydrogen. The simplest statement 

 of this conclusion is that the molecule of hydrogen con- 



