ture produced by percuffion. Iron is easily treated to rednefs by hammering; yet it only fuffers a condenfation of Air, condenfed to, fcarcely raifes the most delicate thermometer a few degrees. Mr Dalton fhews indeed, by calculation, that the real increase of temperature is 50°; but this would be very trif ling in proportion to the caloric given out by hammering iron, if we were to eftimate what is wafted during the operation. The brittlenefs of iron hammered to rednefs, was afcribed by Dr Black to the deficiency of the caloric thus expreffed from it; and upon this hint, Dr Thomfon, filled with the philofophie fpirit of generalization, at once perceives, that brittlenefs feems in moft cafes owing to the abfence of the ufual quantity of caloric; and refers, for the illuftration of his opinion, to the phenomena of unanealed glafs. But he has not only failed altogether in proving that glafs, cooled quickly, contains at the fame temperature lefs caloric than glafs cooled flowly, but, in a fubfequent part of his work, he has, with more truth, afcribed the brittleness in the former cafe to its unequal contraction.

It is undoubtedly true, that water conftitutes a part of almost all mixtures in which a change of temperature takes place; but our author certainly goes much too far in ftating it to be effential; for, besides the mixture of gafes which he mentions as the only apparent exception to this rule, there is an extrication of heat and light when fulphur acts upon the metals, and probably in many other inftances. Water, indeed, is no more effential to the production of heat from mixture, than it is to chemical action in general.

Having concluded the subject of caloric, our author fubjoins fome obfervations on fimple bodies in general, which are only remarkable for the ingenuity with which he moulds nature to fuit his particular opinions, by first limiting the fimple fubftances to thofe concerned in combustion, and then fagaciöufly obferving, that oxygen is capable of uniting with all fimple confinable bodies.

We are now come to our author's Compound Subftances; but, from the very great extent of the work before us, and the immenfe variety of fubjects which it embraces, it would far exceed our limits to notice the whole of these in the manner they deferve; and as their arrangement has been already pretty fully explained, we fhall confine ourselves to a few curfory obfervations. In juftice to our author, however, we muft not omit to mention, that they will principally regard his own opinions, when they appear to us erroneous or doubtful. Whatever we pafs over in filence, is at leaft good; often excellent. In the account of "the compofition of water, fome errors have crept into his calculation, which differs from the ftatement given under hydrogen. As an accu

I 4

rate

rate knowledge of the conftituents of water is of great importance in analysis, and as the calculations have not been revised fince the real conftituents of carbonic acid have been discovered, we have corrected them from the original data. The quantity of gafes em. ployed were, Hydrogen Oxygen

Atmospheric air

25980.563

13475.198 French cubic inches,
15-

The carbonic acid was produced from a fmall quantity of carbon diffolved in the hydrogen. It weighed 26.9 grains, and contained 22.09 oxygen, and 4.81 carbon. From the hydrogen employed, the 16 inches in the refiduum must be deducted; and the remainder, 25964.563, multiplied by its weight per inch 0.040452, gives 1050.32 grains; from which, the 4.81 of carbon, being deduct ed, leaves 1045.51 grains as the real weight of hydrogen confumed. But the oxygen contained 404.256 cubic inches of azote mixed with it; which, with 465 of oxygen found in the refiduum, being deducted, and the 4 contained in the 15. atmospheric air being added, gives 12609.942 of oxygen. This, multiplied by its weight per inch 0.493986, gives 6229.33 grains; from which the 22.09 expended in the formation of carbonic acid, being deducted, leaves 6207.24 oxygen. There were therefore confumed,

which is but 7.75 French grains, or 6.36 Troy, more than the water obtained.

That important clafs of bodies, the Acids, are divided by our author into products, fupporters, and combustible acids; a divifion, which is of fome ufe in our general views of the subject, but, on the other hand, would become inconvenient if ftrictly followed in the detail. To this chapter fome obfervations on the acid principle are prefixed, in which our author endeavours to fhew that oxygen is not an effential constituent of acids. But we think

the

the matter ftill doubtful; for, befides the three undecompounded acids, the only others in which Dr Thomson has denied the prefence of oxygen, are, the Pruflic acid and fulphureted hydrogen. Now, the former certainly contains oxygen; for if Vauquelin's experiments were not of themselves fufficient to prove it, an oxide of carbon, charcoal, is admitted to be one of its constituents; and our acquaintance with the compofition of the latter is certainly not enough to allow us to affert that it contains no oxygen

The clafs of compound combustibles is exceedingly deficient It ought to have contained the greater part of animal and vege table fubftances; and our author's reafons for excluding them are moft unfatisfactory-They are too little known, and their utility as chemical inftruments is too inconfiderable!' A fyftem of chemistry ought to be complete in its arrangement, and totally independent of any effays on meteorology, mineralogy, or phyfiology. These form, it is true, beautiful applications of the fcience, and they cannot be understood without it; but they have no more pretenfions, than the chemical arts and other useful applications, to be forced in as effential parts of a fyftem of chemistry. To the chemift, each individual fubftance is the fame, from whatever kingdom of nature it may be derived, and to whatever purpose it may be applicable.

Our author gives an erroneous idea of the compofition of fixed oil, in afferting it to confist of carbon and hydrogen only. It is a compound oxide. Lavoifier's analysis, by burning oil with oxygen gas, gives the following refults.

In the Appendix we find our opinion confirmed by Dr Thomson himself, who, fpeaking of an oxide of fulphur he has difcovered con

taining

bon and hydrogen, exactly amount to the 15.79 grains of oil burnt. Therefore the conftituents of oil are,

The analyfis, given by our author, is that of Lavoifier; but Lavoifier was unacquainted with pure carbon, and gave that defignation to charcoal, which is an oxide of carbon. Therefore, in fpeaking of Lavoifier's carbon, Dr Thomfon fhould have always diftinguished it by the appellation of charcoal, and in all analyfes have remembered that it was an oxide, which he has feldom if ever done. His negligence in this respect is the more inexcufable, as, by ufing one term to exprefs two very different fubftances, he has often both misled himself, and rendered his ftatements ambiguous to others. Thefe obfervations apply ftill more strongly to his account of the compofition of wax and alcohol, because he has founded on the prefence of oxygen in these fubftances, as demonftrated by various experiments, to prove that the experiments of Lavoifier, from which that philofopher concluded that the former confifted of carbon and hydrogen, and the latter of carbon, hydrogen and water, are not to be depended on. Unexceptionable they are not; but, for the prefent state of the fcience, they are remarkably accurate; and until we have better data to go upon, we must confider them highly valuable. When the calculation from them is corrected, their compofition appears to be

Whether any of the hydrogen and oxygen exifted combined in the ftate of water, we have no means of ascertaining.

The Salts are the most important clafs of the fecondary compounds. The common diftribution of these into the two great families of the metalline, and earthy and alkaline falts, is properly retained; the genera of the latter being distinguished by the acid, and of the former by the metal they contain. The alkaline

- and

taining 6.2 per cent. oxygen, fays, I have fince found reafon to be lieve that it is this oxide, and not pure fulphur, which exifts in fulphureted hydrogen gas, and probably in all the hydro-fulphurets. '

and earthy falts are moreover divided by Dr Thomfon into the two orders of combustible and incombuftible; but it would have been more confiftent with other parts of his arrangement, to have formed a third order of the detonating or fupporting falts, which are at present claffed with the incombuftible. We may alfo, mention, that the ammoniacal falts are all combustible, and therefore, in fome inftances, do not properly belong to the fame order with the other fpecies of the genera.

The genera of the metalline falts are not eafily claffed in differ ent orders; but the fpecies of each genus form fevéral natural Our author has divided them into detonating, income buftible, combustible, metallic, and triple falts. This arrange ment is deficient with regard to unity; for it is formed upon two principles, which interfere with each other; the three firft divi fions being taken from the properties, and the two laft from the compofition of the falts. The two laft indeed appear to be altogether unneceffary; for all the metallic and triple falts are either detonating, incombustible, or combustible. They form, however, very natural fubdivifions of these groups. The falts are by far too numerous, for us to enter into any examination of our author's account of them. We may only mention, that he seems to have been rather hafty, notwithstanding Chenevix's excellent experiments, in annihilating the genus of oxymuriats; for it is certain, that many of them poffefs the property of bleaching, which, in all probability, depends on their containing oxymuria tic acid, fince neither the muriatic, nor hyper-oxymuriatic acid, deftroys vegetable colours. The hydro-fulphurets and foaps are the only other fecondary compounds noticed, although there are feveral other claffes of them.

Having finished his account of the fecondary compounds, Dr Thomfon proceeds, as ufual, to draw fome general inferences from the facts he has detailed; and, in the prefent inftances, he feems extremely unfortunate; for not one of the four he has stat ed is admiffible: 1. He has difcovered a fingular and remarkable correfpondence between fecondary compounds and fimple bodies; for neither of them poffefs that activity, that violent action upon other bodies, which distinguish primary compounds.' This is not fimply a mistake, it is a miftatement. Our author felects fuch fubitances, and places them in fuch circumftances, as fuit his purpose, although numerous facts exift in obvious and direct oppofition to his general conclufion: Does he confider combuftion as a proof of the inactivity of oxygen, and of the fimple com buftibles? or do the oxymuriats and metalline falts appear to him examples of the inertnefs of fecondary compounds? Nay, he himfelf has quoted potafs as the extreme example of the activity of primary compounds; but until he proves potafs to be a com

pound