[22] The question of the molecular weight of calomel--that is, whether the mercury in the salts of the suboxide is monatomic or diatomic--long occupied the minds of chemists, although it is not of very great importance. It is only recently (1894) that this question can be considered as answered, thanks to the researches of V. Meyer and Harris, in favour of diatomicity--that is, that calomel is analogous to peroxide of hydrogen and contains Hg_{2}Cl_{2} (like O_{2}H_{2}) in its molecule if corrosive sublimate contains HgCl_{2} (like water OH_{2}). As a matter of fact, direct experiment gives the vapour density of calomel as about 118--that is, indicates that its molecule contains HgCl, whilst the molecule of the sublimate, judging also by the vapour density (nearly 136), contains HgCl_{2}; it might therefore be concluded that the mercury in the suboxide is not only monovalent (corresponding to H) but also monatomic, whilst in the oxide it is divalent and diatomic. Instances of a variable atomicity, as shown by the vapour density, are known in N_{2}O, NO, and NH_{3}, CO and CO_{2}, PCl_{3} and PCl_{5}, and it might therefore be supposed that the present was a similar instance. But there are also instances of a variable equivalency which do not correspond to a variation of atomicity--for example, OH_{2} (water) and OH (peroxide of hydrogen), CH_{4} (methane), C_{2}H_{5} (ethyl), and CH_{2} (ethylene), &c. Here, according to the law of substitution, the residues of OH_{2} and CH_{4} combine together and give molecules; OHOH = O_{2}H_{2} (peroxide of hydrogen) and CH_{3}CH_{3} = C_{2}H_{6} (ethane), &c. The same may be assumed also to be the relation of calomel to sublimate; the residue HgCl, which is combined with Cl in sublimate, corresponds to HgCl_{2}, and in calomel it may be supposed that this residue is combined with itself, forming the molecule Hg_{2}Cl_{2}. On this view of the composition of the molecule of calomel it would follow that in the state of vapour it breaks up into two molecules, HgCl_{2} and Hg, when the vapour density would be about 118 (because that of sublimate is about 136 and that of mercury about 100), and that in cooling this mixture (like a mixture of HCl and NH_{3}) again gives Hg_{2}Cl_{2}. It was therefore necessary to prove that calomel is decomposed in the state of vapour. This was not effected for a long time, although Odling, as far back as the thirties, showed that gold becomes amalgamated (_i.e._ absorbs metallic mercury) in the vapour of calomel, but not in the vapour of sublimate. Recently, however, V. Meyer and Harris (1894) have shown that a greater amount of the vapour of mercury than of