tisfactory that they were neceffary to exprefs the ideas he meant to convey; namely, that we are able to confine the first fet in veffels, but that the fecond cannot be confined in any vessel.' But when we come to find the ufe he makes of thefe terms, and efpecially when he adds, that all the terms that have been hitherto employed to characterife thefe two fets of bodies, convey fome hypothefis or other, which, in a work of this kind, it is neceffary as much as poffible to avoid,' we find ourselves compelled to object, firit, that the application of his terms is inconfiftent with fact; and, fecondly, that other terms have been already employed which conveyed no hypothefis. It appears to us, that whatever can be excluded, can also be confined; and whatever can be impeded in its progrefs, is not abfolutely unconfinable. The bodies which he calls unconfinable are, light, caloric, electricity, and magnetifm. The two last are not treated of in this work. With regard to caloric, the difficulty with which it permeates certain bodies is well known. Clothing is ufed to confine the warmth of the body; our furnaces are conftructed of bad conductors, to prevent the diffipation of the heat; and caloric may be abfolutely confined in a veffel of ice, as long as the veffel itself will laft. Light is ftill more confinable. Every room is furnished with fhutters to exclude it, and the dark-lanthorn was contrived to confine it. The circumftance of these bodies poffeffing no determinable gravity, or being imponderable, which has been already employed to characterise them, is the fimple enunciation of a fact, and not liable to fimilar objections. Dr Thomfon's confinable bodies are fubdivided into oxygen, fimple combustibles, and fimple incombuftibles.

The compounds are divided into primary and secondary. The former confift of two or more fimple bodies united together, and the latter of two or more compounds. In this arrangement, there is no place allotted for the combinations of compounds with fimple bodies. But befides this omiffion, the divifion of compounds into primary and fecondary in the present state of the fcience, must be entirely arbitrary or hypothetical. It is not only exposed to the general objection arifing from our total ignorance of what bodies are really fimple; but many of our author's primary compounds are, in fact, only known to us in a ftate of farther compofition. Almost all the acids, as objects of our fenfes, are compounds of acid and water; and many of them, independently of this, are fecondary compounds in the ftrict fenfe of the word. But, waving as frivolous this objection, which applies to all fubftances compofed of two ingredients which combine in more proportions than one, we are completely ignorant of the nature of the combinations formed by the

union of three or more fimple fubftances. We do not know whether the immenfe variety of thefe are primary compounds, refulting from various proportions of A, B, C, or whether A firft unites with B, and then forms a fecondary compound with C. Indeed, according to the opinions which Dr Thomfon afterwards advance's when treating of affinity, there can be no fecondary compounds.

In his fubdivifions of the primary compounds, Dr Thomson entirely renounces his general principles, and arranges them under the five heads of alkalies, earths, oxides, acids, and compound combustibles-natural claffes which have been adopted by all preceding fyftematic writers. In thus deviating from his own peculiar principles, Dr Thomfon is certainly inconfiftent; nor are we fatisfied with his reafons for being fo. Thefe are ftated in a preliminary note, in which he gives us a fketch of the arrangement he might have followed; and adds, but in the prefent imperfect ftate of the fcience, the advantages attending this arrangement would not compenfate for the violence of the changes which it would introduce. It would oblige us to claflify substances together, which have always been confidered as diftinct; and to feparate many bodies which have hitherto been always grouped together. Befides, we would be forced to ómit a number of fubftances which are ftill undecompounded, and which are not the lefs important, because they cannot with propriety be introduced among the fimple fubftances.' The firft argument, we confefs, we did not expect from our author, who on fo many occafions calls in queftion opinions the most univerfally received, and advances others fo contradictory to all former obfervation, that he is in greater danger of being accufed of temerity, and affectation of fingularity, than of being blamed for blindly following eftablifhed authorities, or regarding the prejudices of others. The fecond argument militates directly against the whole fyftem. For what are we to think of an arrangement, in which a number of important bodies cannot with propriety be claffed among the fimple fubftances, and yet are pofitively excluded from any place among the compounds? We are the more aftonifhed at this reafon being given by our author for not following up the principles of his arrangement, as the objection might have been easily removed, and indeed, in strict conformity to the philofophy of arrangement, ought to have been removed, by adopting the fame characters for diftinguifl ing the orders of the compounds with thofe employed for the fimple fubftances. Compound fupporters, combuftibles, and incombustibles, would have comprehended every compound body. Under the head of fecondary compounds, are included fome commonly received and natural families; but the claffification is exceedingly defective.

The

The third book of the first part treats of Affinity.

The fecond part of this great work contains the chemical examination of nature; a fubject certainly most intimately connected with the science of chemistry, but fo far diftinct from it, as not to be a neceffary part of an elementary treatife, which ought to contain a complete view of the principles of the fcience, totally independent of its application to any purpose whatever.

We have been thus particular in our account of the arrangement of this work, becaufe fo much fuperiority has been afcribed to it by its author, and, in our opinion, without fufficient reafon. It is every where inconfiftent with its own principles; it is incomplete; it fometimes clailes together bodies which have. little analogy; but more frequently divides and fubdivides the account of a connected fubject into minute portions, which are fcattered through very diftant parts of the work. If it had been rendered totally independent of hypothefis, and completed on the fame general principles, it might have afforded a very good tabular view of the fcience; but, as the outline of a detailed fyftem, or as the text-book of a lecturer, it feems to be by no means preferable to thofe in common ufe.

In

It now remains to examine the execution of the work. doing this, it will only be poffible to notice fuch parts as appear particularly interefting, more efpecially thofe in which an author, of fo great reputation, has committed errors, which might miflead readers lefs difpofed to question his infallibility than we are. The definition of Chemistry is as unintelligible as abstract definitions of science ufually are. It is faid to be that science which treats of thofe events or changes which are not accompa➡ nied by fenfible motions.' Dr Thomfon may probably confider it as an initance of vulgar prejudice; but we must confefs that we have always confidered the burfting of a bomb-fhell and the elevation of the pifton of a fteam-engine as fenfible motions. The definition is alfo particularly defective, in taking no notice of the most indifpenfable condition of chemical action, namely, the reciprocal action of at least two kinds of matter, and the change of properties occafioned by it.

After mentioning concisely the different epochs in the hiftory of chemistry, Dr Thomfon enters upon his account of the fimple bodies. The general manner in which he treats each of thefe, is, firft to tell how it is procured, then to mention its phyfical properties, and, laftly, to detail its mode of action upon thofe other bodies which have been already defcribed. The account of oxygen is neceffarily very brief, as not one of its chemical properties can be explained, without. fuppofing fome previous chemical knowledge. This inconvenience is not peculiar to Dr Thomfon's arrangement, and indeed cannot cafily be a

voided.

voided. But it would furely have been better to have commenced with a clafs of tangible bodies, than to have plunged his readers, in the very first page, into all the intricacies of the defcription of a pneumatic apparatus, and to have perplexed them with different kinds of gafes, before they can be fuppofed to have formed any conception of air being a body poffeffing chemical properties, or indeed to have any conception of chemical properties at all. This fection concludes with an explanation of affinity; which fhews, in the first place, that Dr Thomfon cannot proceed a fingle ftep, without explaining the general doctrines of chemistry; and, fecondly, how few facts are fufficient to make them intelligible even to beginners.

The fecond chapter treats of the fimple combuftibles; and, firft, of fulphur; which gives Dr Thomson an opportunity of briefly explaining the theories of combuftion propofed by Stahl and Lavoifier. The other fections treat of phosphorus, carbon, and hydrogen. Carbon is here fynonymous with diamond; of courfe, Lavoifier's carbon becomes carbonous oxide, or charcoal. But, throughout the whole work, Dr Thomson has neglected to attend to this diftinction; and confequently, in the very next chapter, we are prefented with a pretty full account of the carburetted hydrogen gafes, chiefly taken from the experiments of Mr Cruickshank; although we think that neither their ingenious contriver nor Dr Thomson has drawn from them the conclufions they warranted. For example, in the third experiment, 16 grains of carburetted hydrogen were detonated with 40 of oxygen; the products were 36 grains of carbonic acid gas, and 20 of water. Now, 36 grains of carbonic acid gas are compofed of 29.56 oxygen, and 6.44 carbon; and 20 of water of 17.12 oxygen, and 2.88 hydrogen. From the total quantity of oxygen in the product, 46.68, deduct the 40 added, and we have 6.68 oxygen, 2.88 hydrogen, and 6.44 carbon, as the elementary constituents of the 16 grains of carburetted hydrogen.

The following table exhibits a view of the conftituents of all the fpecies calculated in this way, and as given by Dr Thomson. Oxygen. Carbon. Hydrogen. Water. 35.54 + 46.40 + 18.06

Carb. hyd, from ftag-}

nant water,

Thefe differences arife from two caufes. Mr Cruickshank did not calculate the constituents of the carbonic acid produced according to its analysis by Morveau; and he fuppofed that the whole water produced existed in the gas in a state of folution. But it is highly improbable that any gas is capable of diffolving its own weight of water; and as the carbonic acid gas produced must be faturated with vapour, it is, on the contrary, probable, that the whole water precipitated was formed at the moment of decomposition. We therefore think ourselves warranted to conclude, that thefe gafes are not carbureted hydrogen, holding their own weight of water in folution, but that they are hydro-carbonous oxides, and therefore should be referred to a different place in Dr Thomson's arrangement from that which is now affigned them.

The chapter on Combustibles is concluded, as indeed all the chapters are, with fome general reflections. The next chapter treats of the Simple Incombuftibles, in which the affociation, of azote and muriatic acid as analogous bodies, is the only thing re markable. An account of the metals concludes the ample confinable bodies.

The fecond divifion of the fimple bodies comprehends those which Dr Thomfon has denominated unconfinable. In this work, he only treats of light and caloric; but we are happy to learn, that an author fo capable of doing juftice to the fubject propofes to confider electricity in a feparate work. The chapter on Light is exceedingly well written. That on Caloric, is perhaps the moft remarkable in the whole work, from the very curious fpe culations which it contains.

Caloric paffes through fome bodies with the velocity of light, and through others extremely flowly. Its motion in the first cafe in which it is faid to be tranfmitted, is explained by fuppofing the particles of caloric to repel each other; but the explanation of its fecond kind of motion is not so easy. It has hitherto been confidered as the operation of a pofitive force, which has been called the conducing power; and bodies were hamed good or bad conductors, in proportion as caloric moved through them with greater or lefs facility. Dr-Thomfon has, however, advanced an hypothefis directly the reverse of this; for, obferving that caloric is tranfmitted through fome bodies with immenfe velocity, he has feen the neceflity of accounting for the retardation of its progrefs when it enters conductors. This he afcribes to an attraction or affinity exifting between the particles of caloric and thofe of the conductor, by which the caloric is entangled and detained; fo that bodies which were ailed bad conductors, are in fact good retarders; and the con

VOL. IV. NO. 7.

I

ducting