1. The cohesion of B effects its complete separation; e. g. When ammonia is added to an aqueous solution of sulphate of alumina, the sulphuric acid at first divides itself between the two bases in the ratio of their chemical masses; but since the alumina is thus deprived of a portion of its sulphuric acid, and the remainder is not sufficient to hold all the alumina in solution, a portion of it is precipitated and thus removed from the sphere of chemical action: now since by this precipitation, the quantity of alumina contained in the solution, and therefore also its chemical mass, is diminished, the ammonia is enabled to rob it of another portion of sulphuric acid, thereby precipitating more alumina, diminishing the chemical mass of that which remains dissolved, again removing sulphuric acid, and so on,-till at length it appropriates all the acid and throws down the whole of the alumina. These successive decompositions follow each other so quickly, that the whole action seems to take place in a

moment.

2. The elasticity of B effects its complete separation; e. g. When hydrochloric acid is added to a solution of carbonate of potash in water, the potash at first divides itself between the two acids: the compound thus formed of part of the potash with the whole of the carbonic acid allows however a part of the carbonic acid, now less intimately combined, to escape as gas and thus to remove itself from the sphere of action; the chemical mass of the carbonic acid in the solution being thus diminished, the hydrochloric acid takes from it a fresh portion of potash, and sets free another portion of carbonic acid-and thus the action is repeated, as in the former case, till the whole of the potash has combined with the hydrochloric acid, and the whole of the carbonic acid has escaped.

3. The cohesion of A C effects the complete separation of B; e. g. If baryta dissolved in water be brought in contact with a mixture of sulphuric and nitric acids, in such proportion that for every molecule of baryta present there shall be 1 molecule of sulphuric and 3 of nitric acid, the baryta will at the commencement be divided between the two acids in the same proportion as the potash in the example above given. But potash forms with both the acids soluble salts, which therefore remain mixed; whereas the compound of baryta with a certain quantity of sulphuric acid is insoluble, falls down, and is removed from the sphere of action. The solution now contains, besides the combination of baryta with nitric acid, the excess of sulphuric acid which the precipitated sulphate of baryta was unable to take up. This free sulphuric acid takes from the nitric acid, in proportion to its chemical mass, a new quantity of baryta, which however is precipitated in combination with the proportional quantity of sulphuric acid; this sets free another portion of sulphuric acid, which again takes baryta from the nitric acid; and this repeated abstraction and precipitation goes on till all the baryta is thrown down in the form of sulphate and all the nitric acid is set free. In accordance with this explanation, Berthollet supposes that on bringing together two salts, whose acids as well as bases are different, four salts are always produced: thus nitrate of potash and sulphate of soda in solution produce a mixture which still contains a portion of these salts in the undecomposed state, together with sulphate of potash and nitrate of soda. In general therefore four salts are obtained; but if one of the two new salts is insoluble and separates itself from the sphere of action, the undecomposed salts yet remaining in solution produce, in consequence of the equal division of their elements, a new quantity of the insoluble salt; but as this always falls down, the decomposition goes on till the two original

salts are completely decomposed;-e. g., nitrate of baryta and sulphate of soda.

4. The elasticity of A C effects the complete separation of B; e. g. When peroxide of iron is heated to redness with charcoal, the oxygen ought to divide itself between the iron and the carbon in proportion to their chemical masses. But since the oxygen which combines with the carbon forms carbonic oxide, which escapes as gas and so becomes removed from the sphere of action, the remaining carbon continues to withdraw oxygen from the iron, till the latter is completely reduced to the metallic

state.

Review of Berthollet's Theory.-1. This theory does not establish the identity between affinity and universal attraction. Berthollet himsel supposes that different bodies have very different degrees of affinity for one another, without specifying to what extent the individual qualities of the molecules may exert a peculiar influence on their mutual gravitation and thus modify the laws of universal attraction.

2. Unacquainted with our present system of stoïchiometry, Berthollet supposed that two bodies can combine in any proportions whatever, and endeavoured to explain the fact that combination generally takes place in a few definite proportions only, by assuming that precisely when these proportions hold good, the compound possesses the greatest density, cohesion, or elasticity. But why does chlorine gas combine with hydrogen gas in one proportion only, and then without any condensation or expansion produce hydrochloric acid gas?

3. It is true that the quantity of a substance exerts some influence on its manifestations of affinity (p. 125); but unless adhesion also comes into play, this influence is not exerted by any quantity beyond that which is still capable of entering into combination. For example, since one atom of oxygen cannot combine with more than one atom of carbon, 100 atoms of carbon will have no more effect on the combination of any substance with one atom of oxygen than a single atom of carbon would; if this one atom cannot abstract the oxygen, neither will 100 atoms do it.

4. Berthollet's theory-that a body A divides itself between the bodies B and C in the proportion of their chemical masses-has an appearance of truth in those cases only in which the substances which act upon each other are contained in a liquid in which both they and their possible compounds are soluble; because in such cases it cannot for the most part be directly shown what compounds are contained in the liquid, whether AC and B according to the ordinary view, or A B and A C according to Berthollet's. But in some cases even of this kind, the incorrectness of Berthollet's theory may be distinctly shown. Boracic acid colours litmus wine-red, sulphuric acid turns it bright red. Now if sulphuric acid be gradually added to a warm solution of borate of soda in water which has been coloured blue with litmus, the liquid at first remains blue, because a combination of soda with excess of boracic acid is produced; on the addition of more sulphuric acid, boracic acid is set free, and colours the liquid wine-red; and not till all the soda has entered into combination with the sulphuric acid does a further addition of that acid give the liquid a bright red colour; but if sulphuric acid were present at the commencement of the action, either in the free state or combined with sulphate of soda in the form of an acid salt, the bright red colour would appear at once. (Gay-Lussac, Ann. Chim. Phys. 49, 323; also Pogg. 25, 619.) From the same cause, a solution of sulphate of potash or soda to which boracic acid has been added colours litmus only wine-red; but the

addition of of sulphuric acid immediately produces the bright red tint. (Dubail. J. Pharm. 18, 425). Hence boracic acid does not take soda from sulphuric acid or set that acid free. Hydrosulphuric acid and carbonic acid exhibit similar relations towards sulphuric acid. (Dumas.)Tincture of litmus is instantly bleached by chlorine water, but not till after several days by aqueous solution of iodine: now, a solution of chloride of sodium mixed with iodine should, according to Berthollet, produce a mixture containing chloride of sodium with excess of chlorine, and iodide of sodium with excess of iodine. But the orange-yellow mixture colours litmus green (from the yellow of the solution and the blue of the tincture): and a very small quantity of chlorine water immediately changes this green colour into the orange-yellow of the solution of iodine: this shows that no chlorine had been set free by the iodine. (Dubail; Gm.)-Phosphate of peroxide of iron is soluble in hydrochloric acid, but not in acetic acid. From its solution in hydrochloric acid it is completely precipitated by acetate of potash. Now if the potash had been divided between the hydrochloric and acetic acids, part of the hydrochloric acid would have remained free, and would have held some of the phosphate of iron in solution. (Gay-Lussac a. a. 0; also Ann. Chim. Phys. 70, 416.-Compare also Persoz. Chim. molecul. 346.)—The experiments of Soubeiran and O. Henry (J. Pharm. 11, 430, also Mag. Pharm. 15, 44, also N. Tr. 12, 1, 266) do not prove much in favour of Berthollet's views.

Other objections to Berthollet's theory of distribution may be deduced from the following facts. Oxalate of lead digested with water and as much sulphuric acid as is necessary to saturate the oxide of lead, is completely resolved into sulphate of lead and free oxalic acid. (Pfaff. Ann. Chim. 77, 266): Berthollet's remarks on this experiment (Ann. Chim. 77, 288,) are not satisfactory.-Hyperiodate of lead digested with water and a quantity of sulphuric acid somewhat less than that required to take up all the oxide of lead, yields a solution of hyperiodic acid free from sulphuric acid and from hyperiodate of lead. (Benckiser. Ann. Pharm. 17, 257.)-Chloride of silver mixed with water is easily converted by iron into metallic silver and chloride of iron, the latter remaining in solution. According to Berthollet the contrary result should be produced, since iron. is more coherent than silver, and chloride of silver is insoluble in water, while chloride of iron is soluble. These last experiments likewise show that insoluble substances, such as oxalate of lead, chloride of silver, &c., are by no means removed from the sphere of chemical action.-Similarly Gay-Lussac has shown (Ann. Chim. 89, 21) that a metallic oxide insoluble in water may completely precipitate another from its solution in acids (e. g., oxide of zinc may precipitate oxide of silver), provided it be added in sufficient quantity to saturate the acid.-3 atoms of iron fused with one atom of tersulphuret of antimony completely separate the antimony from the sulphur, though no solid or gaseous compound is formed, the melted sulphuret of iron lying in a stratum above the melted antimony.

It has also been shown (page 130) that hydrochloric acid decomposes carbonate of lime, and forms with the lime a perfectly neutral solution, even under a pressure sufficient to liquefy carbonic acid. Now since the hydrochlorate of lime is soluble and the carbonate insoluble, the contrary effect ought to be produced, according to Berthollet, as soon as the escape of carbonic acid is prevented. In a similar manner, hydrochloric acid decomposes sulphite of lime, although that salt is nearly insoluble, and sulphurous acid has less elasticity than hydrochloric acid, inasmuch as it is liquefied by smaller pressure.

To Berthollet must be conceded the great merit of having closely scrutinized the theory of affinity, examined it in a new light, and directed attention to the influence exerted by cohesion and elasticity on the manifestations of affinity. But he laid too little stress on the magnitude of affinity, and too much on the quantity in which substances act, and on the influence of cohesion and elasticity. He erroneously supposed that a body which separates in the solid state is removed from the sphere of action, that bodies are capable of combining in all proportions, and that a substance divides itself between two others in the proportion of their chemical masses. Second Hypothesis. Chemical combinations are produced by a peculiar power, called Affinity, different from universal attraction.

So long as it is assumed that universal attraction, as manifested in gravitation, acts only in proportion to the mass, and that the peculiar nature of a substance has no influence on its amount,-it is difficult to refer the manifestations of cohesion and adhesion, and impossible to attribute those of affinity, to its action. In chemical phenomena, the quality of a substance above all things determines the existence and strength of the attraction, and its influence cannot be replaced by that of quantity. Moreover, a high degree of affinity must be ascribed to the imponderable bodies, which are not subject to the laws of gravitation. So long therefore as it shall remain undemonstrated that gravitation is influenced by quality of matter, and that the hitherto so-called Imponderables possess weight, or else that the phenomena hitherto attributed to the affinities of these bodies are really due to other causes-so long will it be most advisable (as indeed most chemists at least tacitly do) to regard affinity as a peculiar power distinct from all others.

Third Hypothesis. The union of heterogeneous atoms is the result of Electrical Attraction. (Electrochemical Theories).

In some of these theories a common fundamental power is assumed which shows itself, sometimes as electrical, sometimes as chemical force; in others the combinations of ponderable substances, uninfluenced by any affinity of their own, are supposed to arise merely from the mutual attraction of the two electricities attached to their atoms, which attraction is itself regarded as a kind of affinity.

To the list of electro-chemical theories belong those of Winterl (N. Gehl. 6, 1 and 201);-of Sir H. Davy (N. Gehl. 5, 41, also Elem. of the Chemical Part of Nat. Phil.);-of Dumas (Phil. of Chem. p. 369), further developed and contested by Grotthuss (Phys. Chem. Investigations, 1, 44); of Ampère (Pogg. 2, 185);—of Becquerel (Ann. Chim. Phys. 24, 192);-of Ferré (Ann. Chim. Phys. 28, 417);-of Schweigger (Schw. 5, 49; 6, 250; 7, 302 and 515; 8, 307; 11, 54, 330 and 435; 14, 510; 25, 158; 39, 214; 40, 9; 44, 79; 52, 67);-and of Fechner (Schw. 52, 27).

The electro-chemical theory of Berzelius demands, as the fullest and most consecutive, a more detailed explanation.-Compounds usually called chemical are divided into two classes. The less intimate whose formation is attended with lowering of temperature-e. g. solutions of salts in water-must be regarded (since all solid bodies are not soluble in water) as resulting from a specific attraction (comp. page 34 1, 2); the atoms of the solid body diffuse themselves through the liquid, till each atom is surrounded by an equal number of atoms of the liquid.-The more intimate compounds are the really chemical or electro-chemical combinations. These result, not from any mutual affinity between their ponderable elements, but from that of the electricities attached to their atoms. The atom of each substance has two poles, on which the two opposite electricities

are accumulated in different proportions, according to the nature of the bodies. The atom of many bodies, oxygen for instance, has a large quantity of negative electricity attached to one of its poles, and but a very small quantity of positive electricity at the other; that of other bodies, potassium for example, has a large quantity of positive electricity at one pole and very little negative electricity at the other. Thus the elementary substances are divided into electro-negative and electro-positive. To each element however there belongs a particular proportion between the quantities of the two electricities. Oxygen has, of all the electronegative elements, the greatest quantity of negative electricity at one of its poles and the smallest quantity of positive electricity at the otherthen follows sulphur, then nitrogen, &c., and lastly hydrogen, in which the quantities of the two electricities are nearly equal. Of all electropositive substances, potassium has the largest quantity of positive and the smallest of negative electricity; and this inequality continually diminishes in other bodies, till we come to gold, in which the positive electricity predominates but little over the negative so that this element occupies the next place to hydrogen. According to this, the elements succeed one another in the electro-chemical series of Berzelius as follows, beginning with the electro-negative.

Electro-negative: O, S, N, F, Cl, Br, I, Se, P, As, Cr, V, Mo, W, B, C, Sb, Te, Ta, Ti, Si, H.

Electro-positive: Au, Os, Ir, Pt, Rh, Pd, Hg, Ag, Cu, U, Bi, Sn, Pb, Cd, Co, Ni, Fe, Zn, Mn, Ce, Th, Zr, Al, Y, G, Mg, Ca, Sr, Ba, L, Na, K.

In the combination of an electro-negative with an electro-positive body, the predominant negative electricity of the former unites with the predominant positive electricity of the latter. Before, however, combination takes place, the former substance exhibits negative, and the latter positive electricity in the free state; and the tension of the two electricities continually increases as the bodies approach the temperature at which combination takes place. Hence we have an explanation of electricity by contact. At the instant of combination, the negative poles of the atoms of the first body turn themselves towards the positive poles of those of the second; and since it is only in the fluid state that the atoms possess the mobility necessary for this arrangement, it follows that solid bodies have, generally speaking, no chemical action on one another. The two electricities of these poles now combine and produce heat or fire, whereupon they disappear. In every chemical combination, therefore, a neutralization of the opposite electricities takes place, by which heat or fire is produced in the same manner as in the discharge of the electrical pile or of lightning, excepting that these last-mentioned phenomena are not accompanied by any chemical combination, at least of ponderable bodies. Every chemical combination is therefore an electrical phenomenon depending on the electrical polarity of the atoms.

Since the electrical series does not accord with the order of affinitysince for example, the highly electro-negative substance oxygen has, according to experiment, less tendency to combine with the electro-positive body gold than with sulphur which stands next to oxygen in the electrical series-Berzelius supposes that, although in the atom of gold the positive electricity of the one pole is of greater amount than the negative electrity of the other, nevertheless the absolute quantity of positive electricity existing at one pole of the atom of gold is less than that which is present at one pole of the atom of sulphur,-the latter containing however a much greater quantity of negative electricity at its other pole than the