and decompofition is certainly no test of the ftrength of affinity. The affinities of the acids follow the order which has long been recognized in the metallic falts. It is true, they are apparently different in the falts from which these tables have been calculated: but that, according to Dr Thomfon, is of no confequence, as the muriats are all more foluble than fulphats. With regard to the carbonic acid, its affinities as calculated from thefe tables are inconfiftent with fact; but they must not be taken into confideration, because the compofition of the carbonats is very imperfectly determined. This kind of reafoning, however, we cannot admit. The compofition of the carbonats was afcertained by Kirwan as well as that of the other falts, and is equally entitled to our confidence; and although, in confequence of the action of mass, elafticity and cohefion, the order of affinity may be different from that of decompofition, it furnishes no argument to prove that Berthollet's hypothefis is more probable than the directly oppofite one of Kirwan, or than any other which may be imagined by any fucceeding philofopher. It is, on the contrary, in favour of Kirwan's hypothesis that it in general coincides with the order of decompotition; for the action of mafs, cohesion and elafticity, may enable him to explain the few apparent exceptions. But Berthollet, although he were to fucceed in the more arduous task of proving that the order of decompofition is in almost every inftance wrong, has not advanced one ftep in establishing the probability of that which he has adopted. Now, befides the affinities of carbonic acid, there are others, derived from his hypothefis, which cannot be accounted for. For example, the affinity of lime to fulphuric acid is ftated to be stronger than that of potafs or foda, and its affinity to nitrous and to muriatic acid weaker than that of magnelia: the affinity of muriatic acid, again, to foda, is stated to be nearly twice as ftrong as that of fulphuric acid: which are all contrary to the order of decompofition, and oppofed alfo, in thefe inftances, by the action of cohefion. Berthollet's hypothefis, therefore, appears to us inconfiftent with fact. Another way of examining the validity of any hypothefis of this nature, is to carry them as far as they will go, and fee to what conclufions they will lead. Now, if the principles of either Kirwan or Berthollet were true, the affinities of bafes for all acids, and of acids for all bafes, fhould follow the fame ratio; which is alfo contrary to fact. Thefe fpeculations, therefore, do not feem to have increased our knowledge of the comparative affinities, of bodies; and we must ftill refort to the humble and tedious method of experiment to af certain them. The next fubject treated of, is Compound Affinity, concerning which we find nothing very remarkable. It does not appear to us by by any means certain, that faline folutions, which may be mixed without precipitation, combine; for example, that when folutions of fulphat of potafs and muriat of foda are mixed, thefe compound falts do not remain entire, but that a folution is formed, containing fulphuric acid, muriatic acid, potafs and lime, uniformly combined; for, upon the fame principle, there fhould be no fecondary compounds, and the phenomena of chemistry fhould be different from what they really are. The effect of the infolubility of falts, as explained by Berthoilet, is true to a certain extent; but it is not without exceptions. In the tables of affinity for nitric and muriatic acid, calculated on his own principles, frontian is placed below foda and potafs, although the falts of itrontian are the moft foluble. The laft chapter is on Repulfion; and it might have been entirely omitted, without any injury to the book as a fyftem of chemistry. To moft of his readers, it will be totally unintelligible, and by many it will be efteemed as a wonderful effort of learning and ingenuity. Notwithstanding the great length of thefe obfervations, a volume and an half fill remain to be noticed, containing the fecond part of the work, entitled the Chemical Examination of Nature. It will not, however, detain us long; as we confider by far the greatest part of what is here collected under this title, as mifplaced in a fyftem of chemistry; and the remainder is merely the application of the knowledge contained in the former part, to the examination of nature. The means of analyzing the atmosphere, mineral waters, minerals, and animal and vegetable fubftances into their immediate principles, and the investigation of whatever chemical changes they undergo, belong properly to chemistry, and would have formed a very natural fequel to a general system of the fcience; while the greater part of the meteorology, mineralogy and phyfiology belong to other departments. The account of the atmofphere is in general well executed; but Dr Thomson has committed an error in his calculation of the proportion of weight of its constituents. From his own data, intead of 74 azotic gas and 26 oxygen, the refults are 75.12 and 24.88; but he has fuppofed the relative specific gravity of oxygen gas to that of azotic gas to be as 135:115, whereas they are as 1356:1189: the real refults are 75.67 azotic gas, and 24.33 oxygen. Our author differs from Mr Dalton in believing atmospheric air to be a chemical compound. Only one of his arguments, however, appears to us to be relevant, viz. that derived from the experiments by which Humboldt and Morozzo endeavoured to establish a difference of properties between atmospherical air and an artificial mixture of its conftituents, though the refult was owing to an excefs of oxygen in their mixture. In fpeaking of the comparative merits of the muriatic and nitric acid fumes in deftroying contagion contagion, Dr Thomson certainly does not speak from experience, when he prefers the former, not only on account of their fuperior efficacy, but also because the latter are attended with inconvenience, from being almost always contaminated with nitrous gas. To what inconvenience he alludes, we know not; but it is certain that the nitric acid fumes, diffufed according to Dr C. Smyth's directions, do not render the removal of the patients during the fumigation at all neceflary, which the muriatic acid gafes do Mineralogy, we are told, is that branch of chemistry which treats of Minerals;' and in conformity with this opinion, Dr Thom fon has filled almost a volume of his work with this subject. But Mineralogy is certainly a branch of Natural history, which is as intimately connected with the phyfical as with the chemical properties of its objects. If Dr Thomson believed himself qualified to write a better fyftem of mineralogy than any of those we pof fefs, it would have been highly acceptable as a feparate publica tion; but we think that, by introducing it in this work, he has unneceffarily increafed its expence. In compiling it, our author is principally indebted to Haüy and Brochant. In the arrange ment, indeed, he seems to think he poffeffes confiderable merit, though we cannot perceive upon what grounds. The principle is taken from Bergman; and in its application, Dr Thomson devi ates from it almoft as frequently as he adheres to it. In other fyftems, minerals have been claffed in genera, according to the nature of the earth from which they derive their characteristic properties; and from this characteristic earth the genera have rea ceived their names. Dr Thomson claffes them in genera accordi ing to the proportions of their conftituents, and gives them fym bolic names, formed by arranging the first letter of every fubitance which enters in any confiderable quantity into their compofition, in the order of their proportions. Now, it appears to us, that every argument which Dr Thomfon adduces against the common arrangement, applies as forcibly against this, and that it is attended with other infurmountable inconveniences. Before any fpecimen can be arranged, it must not only be analyzed, but its analyfis must be perfect; and even if analysis were as eafy as it is difficult, it would often oblige us to place different fpecimens of the fame mineral in different parts of the fyftem. To prove the truth of this opinion, we need only examine a few of Dr Thomson's genera. The first is entitled A, which, according to his principles, are minerals confifting entirely of alumina. It contains two fpe cies; the first, Diafpore, contains alfo 7 water and 3 oxide of iron, and should therefore be defignated by the fymbol A W, if not A WI; the fecond,' Native Alumina, contains only 45 Aluhina, 27 water, and 24 falphat of lime-its fymbol is therefore K2 AWL, AWL, or, as Dr Thomson overlooks falts, A W. The fecond genus is AS. The firft fpecies, Corundum, contains the oriental ruby and fapphire, correctly placed here according to M. Chenevix's analyfis; but, according to Mr Klaproth's, fapphire belongs to A, of the imperfect corundums; that from China, as well as Emery, belong to AIS, as the quantity of iron exceeds that of filica. The fecond fpecies, Chryfoberyl, contains 6. of lime, and therefore belongs to A SL. The third, the Topaz, is right, as well as the Fibrolite, alfo numbered the third by mistake, and the fourth Sommite. The third genus is A M. The first fpeeies, Spinel Ruby, belongs, by Vauquelin's analyfis, to AMC, and by Klaproth's to A SM; and the fecond, the Ceylanite, to AIM. From these three first genera our readers will be able to judge of the others. In this edition, a chapter is added on compound minerals, tranflated from Brochant; and the last chapter treats of the analyfis of minerals. .. The fourth book treats of Vegetables; and the fifth, which concludes the work, of Animals. No part of the work has undergone fo many alterations in this edition as the chapter which treats of the ingredients of vegetables. The author's ideas on the importance of this fubject feem to have undergone a very great change, and to this change of opinion his readers are indebted for much very valuable information; for, instead of 60 pages, it now occupies 160; although there is very little of it, except what is derived from his own experiments, which was not known to pharmaceutifs when the former edition was published. But vegetable chemistry has become fashionable, and Dr Thomson has applied to it with very great fuccefs, in his experiments on gum, farcocol, and the bitter principle. An appendix is added, containing thofe difcoveries of importance which were made during the printing of the work; and we are forry that we must conclude our analysis, by lamenting that the index is not more copious. Dr Thomson has, in general, adopted M. Chenevix's nomenclature; but we have occafionally obferved deviations inconfiftent with it, as tannat and other ats for combinations of fubftances which are not acid. Thefe, however, we believe to be acci dental. Dr. Thomson's method of diftinguishing the degrees of oxidation in the metallic oxides, by prefixing the first syllable of the Greek ordinal numbers to the word oxide, as prot-oxide, deutoxide, &c., and the maximum of oxidation by per-oxide,, we think is an improvement. On the other hand, we truft that our author's example will induce no one to follow him in diftinguishing thofe metalline falts which contain the metal in the state of per per-oxide, by prefixing the particle oxy to the name of the acid, as that form of expreffion has already another much more natural meaning. Capacity for caloric,' is alfo ufed by Dr Thomfon to exprefs the quantity of caloric in equal bulks of bodies, although it has hitherto always had a reference to equal weights. Our author feems alfo to have a very great diflike to fuperfluous letters, not only in the names of fubftances, but also in those of the German chemifts; but Hermitad, Humbolt, Weftrum, &c. will appear to a German eye as awkwardly exotic as Tomson would do to our author's. The references to authorities with which this work abounds, are extremely valuable; and, in general, Dr Thomson gives a due degree of credit to the difcoverers of particular facts; and if, in fome inftances, through ignorance or inadvertence, the real difcoverer is not mentioned, in others his praise ahnoft amounts to flattery. For example, his gratitude to that excellent chemift Mr Hatchett, for having communicated to him his unpublished experiments on refins, has led him to exaggerate their importance to a degree that we conceive must be difpleafing to that gentleman's modefty, especially as moft of the facts, which Dr Thomson feizes every poflible opportunity of announcing as Mr Hatchett's difcoveries, were previoufly known. His general ftatement is in the following words: Hitherto it has been affirmed by all chemists, ancient and modern, that the alkalies do not exert any action on refins. Fourcroy, for instance, in his last work, affirms this in the most pofitive manner; but the experiments of Mr Hatchett have demonftrated this opinion to be completely erroneous.' And after stating the experiments, he proceeds, Nothing can afford a more ftriking proof, than this, of the neceffity of repeating the experiments of our predeceffors before we put implicit confidence in their affertions. The wellknown fact, that the foap-makers in this country conftantly mix rofin with their foap; that it owes its yellow colour, its odour, and its eafy folubility in water to this addition, (?) ought to have led chemiits to have fufpected the folubility of refins in the alkalies. No fuch confequence, however, was drawn from this notorious fact. In oppofition to all this, we thall quote only one odern chemift, Gren, who exprefsly fays that the refins alfo form, with the cauftic alkalies, foapy combinations. Again, It has been fuppofed alfo,' fays Dr Thomfon, that the acids are incapable of acting upon the refins; Fourcroy is equally positive with regard to this; and Gren fpeaks of it in fuch a manner that every reader muft conclude that he had tried the effect of nitric acid upon refins. Yet Mr Hatchett has afcertained this spinion likewife to be erroneous, at leaft as far as nitric acid is concerned.' K 3 |