Of the circumstances which occasion a change in the properties of bodies, some are the addition of what are properly called _substances_, or things that are the objects of our senses, being _visible_, _tangible_, or having _weight_, &c. Thus the addition of an acid changes an alkali into a neutral salt. But other changes are occasioned either by a change of texture in the substance itself, or the addition of something that is not the object of any of our senses. Thus, a piece of steel becomes a magnet by the touch of another magnet, and a drop of glass acquires the property of flying asunder by a small fracture, in consequence of falling when red hot into cold water. Such also, in the opinion of some, is the difference between hot and cold substances.
Till the nature of the cause be ascertained, it is convenient to make use of the term _principle_, as including both of the above-mentioned causes of the change of properties in bodies. Thus, whatever be the real cause of gravity, or of inflammability, we may speak of the _principle of gravity_, or of _inflammability_; whether, with Newton, we suppose gravity to be occasioned by a fluid pervading the whole universe, which he termed _æther_, and whether inflammability be caused by the presence of a real substance called phlogiston, or not. In this manner we use the letters _x_ and _y_ to denote unknown quantities in algebra.
When changes are made in substances by the addition of other substances, they make what is called a _chemical union_; and in this case the properties of the compound cannot with any certainty be deduced from those of the component parts, but must be ascertained by fresh experiments. Thus, from the specific gravities, or the degrees of fusibility, of two metals, those of the compound cannot be predicted. Neither water nor acid of vitriol will separately dissolve iron, so as to produce inflammable air, but both together will do it. However, the properties of similar compounds are similar to one another. Thus, all metals dissolved in acids are precipitated by mild alkalis. This chemical union of two substances we ascribe to a certain _elective attraction_, or _affinity_ that subsists between them, in consequence of which they unite with one another whenever a proper opportunity offers, in preference to those substances to which they were before united. Thus the vitriolic acid, having a stronger affinity with the vegetable alkali which is the basis of nitre, will unite with that alkali, and with it form another compound, called _vitriolated tartar_, while the acid of nitre, being detached from its base, is collected separately.