There are some cases in which bodies appear to combine without any limitation, as spirit of wine and water, sulphuric acid and water; but these must be considered as conditions of mixture rather than of chemical combination. The composition of bodies is fixed and invariable, and a compound substance, so long as it retains its characteristic properties, must consist of the same elements united in the same proportions. Thus, sulphuric acid is invariably composed of 16 parts of sulphur and 24 parts of oxygen. Chalk, whether formed by nature or by the chemist, yields 43·71 parts of carbonic acid, and 56·29 parts of lime. The rust which forms upon the surface of iron by the action of the atmosphere, is as invariable in its composition as if it had been formed by the most delicate adjustment of weight by the most accurate manipulator, being 28 parts of iron and 12 parts of oxygen. This law is the basis of all chemical inquiry, all analytical investigations depending upon the knowledge it affords us, that we can only produce certain undeviating compounds as the results of our decompositions. We are not in a position to offer any explanation which will account for these constant quantities in combination. The forces of cohesion and elasticity have been advanced in explanation, on the strength of the fact that the solubility of a salt in water is regulated by cohesion, and that of a gas by its elasticity. Although it may appear that some cases of chemical combination are due to these powers,--as, for instance, when the union of oxalic acid or sulphuric acid with lime produces an insoluble salt,--we cannot thus explain the constant proportions in which the metals, sulphur, oxygen, and similar bodies, unite. It is quite certain there is a power or principle, which we have not yet reached, upon which are dependent all the phenomena which we now embrace under the term chemical affinity. Another law teaches us that when compound bodies combine in more than one proportion, every additional union represents a multiple of the combining proportion of the first. With the difficulty which arises from the sub-multiple compounds we cannot deal:--further research may render their laws less obscure. We have seen that 8 parts of oxygen unite with 1 of hydrogen and 14 of nitrogen. It also unites with 110 of silver, 96 of platinum, 40 of potassium, 36 of chlorine, and 200 parts of mercury, giving rise to-- Water 9 Nitrous oxide 22 Oxide of silver 118 Oxide of platinum 104 Potash 48 Oxide of chlorine 44 Oxide of mercury 208 In these proportions, or in multiples of them, and in no others, will these bodies unite with the acids or other compounds. It will, of course, be understood that any other numbers may be adopted, provided they stand in the same relation to each other.[208]