The fifth, or laft part, will contain feveral detached articles, such as the defcription of feveral additional experiments, machines, &c. which cannot conveniently be inferted in the preceding divifions.

The axioms of philofophy, or the axioms which have been deduced from common and conftant experience, are so evident and so generally known, that it will be fufficient to mention a few of them only.

I. Nothing has no property; hence,

II. No fubftance, or nothing, can be produced from nothing.

- III. Matter cannot be annihilated, or reduced to nothing.

Some perfons may perhaps not readily admit the propriety of this axiom; feeing that a great many things appear to be utterly destroyed by the action of fire; also that water may be caused to difappear by means of evaporation, and fo forth. But it must be obferved, that in thofe cafes the fubftances are not annihilated; but they are only dispersed, or removed from one place to another, or they are divided into particles fo minute as to elude our senses. Thus when a piece of wood is placed upon the fire, the greateft part of it difappears, and a few afhes only remain, the weight and bulk of which does not amount to the hundredth part of that of the original piece of wood. Now in this cafe the piece of wood is divided into

its component fubftances, which the action of the fire drives different ways: the fluid part, for instance, becomes fteam, the light coaly part either adheres to the chimney or is dispersed through the air, &c. And if, after the combuftion, the scattered materials were collected together, (which may in great measure be done), the fum of their weights would equal the weight of the original piece of wood.

IV. Every effect has, or is produced by, a caufe, and is proportionate to it.

It may in general `be obferved with respect to those axioms, that we only mean to affert what has been conftantly fhewn, and confirmed by experience, and is not contradicted either by reason, or by any experiment. But we do not mean to affert that they are as evident as the axioms of geometry; nor do we in the leaft prefume to prescribe limits to the agency of the Almighty Creator of every thing, whofe power and whofe ends are too far removed from the reach of our understandings.

Having ftated the principal axioms of philofophy, it is in the next place neceffary to mention the rules of philofophizing, which have been formed after mature confideration, for the purpose of preventing errors as much as poffitle, and in order . to lead the ftudent of nature along the shorteft and fafeft way, to the attainment of true and ufeful knowledge.-Thofe rules are not more than four;

viz.

I. We are to admit no more caufes of natural things, than fuch as are both true and fufficient to explain the appearances.

II. Therefore to the fame natural effects we muft, as far as poffible, affign the fame causes.

III. Such qualities of bodies as are not capable of increase or decrease, and which are found to belong to all bodies within the reach of our experiments, are to be esteemed the univerfal qualities of all bodies whatsoever.

IV. In experimental philosophy we are to look upon propofitions collected by general induction from phenomena, as accurately or very nearly true, notwithstanding any contrary hypothefes that may be imagined, till fuch time as other phenomena occur, by which they either may be corrected, or may be fhewn to be liable to exceptions.

With respect to the degree of evidence which ought to be expected in natural philofophy, it is necessary to remark, that physical matters cannot in general be capable of fuch abfolute certainty as the branches of mathematics.-The propofitions of the latter fcience are clearly deduced from a fet of axioms fo very fimple and evident, as to convey perfect conviction to the mind; nor can any of them be denied without a manifest abfurdity. But in natural philofophy we can only say, that because fome particular effects have been conftantly produced under certain circumftances; therefore they will most likely continue to be produced as long

as the fame circumftances exift; and likewise that they do, in all probability, depend upon those circumftances. And this is what we mean by laws of nature; as will be more particularly defined in the next chapter.

We may, indeed, affume various physical principles, and by reasoning upon them, we may strictly demonstrate the deduction of certain confequences. But as the demonftration goes no farther than to prove that fuch confequences muft neceffarily follow the principles which have been affumed, the confequences themfelves can have no greater degree of certainty than the principles are poffeffed of; fo that they are true, or falfe, or probable, according as the principles upon which they depend are true, or falfe, or probable. It has been found, for instance, that a magnet, when left at liberty, does always direct itself to certain parts of the world; upon which property the mariner's compass has been conftructed; and it has been likewise observed, that this directive property of a natural or artificial magnet, is not obftructed by the interpofition or proximity of gold, or filver, or glass, or, in short, of any other fubftance, as far as has been tried, excepting iron and ferrugineous bodies. Now affuming this obfervation as a principle, it naturally follows, that, iron excepted, the box of the mari ner's compass may be made of any substance that may be moft agreeable to the workman, or that may best answer other purposes. Yet it must be

confeffed,

confeffed, that this propofition is by no means fo certain as a geometrical one; and ftrictly speaking it may only be faid to be highly probable; for though, all the bodies that have been tried with this view, iron excepted, have been found not to affect the directive property of the magnet or magnetic needle; yet we are not certain that a body, or fome combination of bodies, may not hereafter be discovered, which obftruct that property.

may

Notwithstanding this obfervation, I am far from meaning to encourage fcepticism; my only object being to fhew that juft and proper degree of conviction which ought to be annexed tophyfi cal knowledge; fo that the ftudent of this fcience may become neither a blind believer, nor a useless fceptic*.

Befides a strict adherence to the abovementioned rules, whoever wishes to make any proficiency in the study of nature, fhould make himself acquainted with the various branches of mathematics; at leaft with the elements of geometry, arithmetic, trigonometry, and the principal properties of the

conic

Scepticism or skepticism is the doctrine of the fceptics, an ancient fet of philofophers, whofe peculiar tenet was, that all things are uncertain and incomprehenfible; and that the mind is never to affent to any thing, but to remain in an abfolute ftate of hefitation and indifference. The wordfceptic is derived from the Greek one, which signifies confiderate, and inquifitive.