The substances are therefore all derivable from common roots, and in this respect, there is no distinction between ammonia and the two alkaline bodies connected with it; and further, there is no other hypothesis, except that of derivation from a common root, that will explain the identity of the essential characters of ammonia and the other alkalies.

Ammonia being formed by the combination of hydrogen and nitrogen, it may be expected that the elements of nitrogen will likewise exist in the alkalies. Accordingly, these alkalies comprehend the secondary root CON, and not only the alkalies, but all the alkaline earths, and all the metallic oxides. Finding, indeed, that nitrogen, or the elements of nitrogen, entered into the composition of ammonia, we might, à priori, have inferred that it entered into the composition of the alkalies; and if into the composition of the alkalies, into the composition of all the bodies with which the alkalies are related, namely, the alkaline earths and metallic oxides: and the fact, that hydrogen and nitrogen can and do form a body similar to that which a metal and oxygen form, demonstrates that the same particles of matter are common to ammonia and the alkalies.

AMMONIUM.

When ammonia in solution in water is placed in contact with mercury at the negative end of a voltaic battery, and the circuit is completed, the mercury increases to about 5 times its former volume. The same effect is produced by using salammoniac slightly moistened. A cavity being made in the salammoniac, it is placed on a plate of platinum attached to the positive end of a battery, and there is placed in the cavity about 50 grains of mercury, which is brought in contact with the negative end of the same battery by a platinum wire. A strong action takes place, with effervescence and the evolution of much heat, and the mercury, in a few minutes, increases in the remarkable manner referred to. The resulting compound is a solid substance, of specific gravity below 3, having the entire appearance of an amalgam of zinc. At the temperature of 70° or 80° this substance is very soft; but at 32° it becomes a firm crystalline mass. When acted upon by water, the mercury is revived, and ammonia remains in solution. When put in a glass tube, or when confined under naphtha or oils, the mercury separates, ammonia is formed, and a quantity of hydrogen is evolved. From these, and precise experiments regarding the substances evolved, it has been inferred that ammonia, united to an additional equivalent of hydrogen, combines with the mercury, and forms this substance, which, in its combination with mercury, exhibits the characters of a metal.

But how can we suppose a metal to be formed by a combination of hydrogen with nitrogen? We cannot even imagine such a result, if we hold a metal to be a simple

body; but the seeming anomaly disappears, when we regard, as we must do, a metal as a compound body, derived from the same roots as ammonium.

And, if one metal be formed by a combination of hydrogen with nitrogen, or the elements of nitrogen, we are bound to assume that all metals are so formed. For to hold that a single metal was formed by such a combination, and that the others were not, would be to violate every analogy founded on the physical and chemical characters of the bodies. If H+ N represents ammonium, we must find a common expression for all the bodies of the same class. But we can only find such an expression by referring them all to a common root; and, therefore, if ammonium be a metal, all the metals are compound.

Besides the compounds, ammonium, H1 N, and ammonia, H3 N, chemists have been led to the conclusion, that there exists another compound of hydrogen and nitrogen, H2 N, to which has been applied the term amidogen. This substance, indeed, has not been insulated; but, from certain actions, its existence or formation has been inferred, and the investigations of Dumas, and more recently those of Rose and Kane, have seemed to give a considerable degree of probability to the theory. But the formation of amidogen may be maintained on other grounds than those that have been taken, namely the law of continuity, as applied to chemical combinations, to which I shall immediately refer. Every substance being assumed to be derived from a root in the ascending order, the more immediate root of the series of bodies, of which ammonium is a sequence, is HN, or, as it may be called, monohydruret of nitrogen, and the series of the bodies will be represented thus:

The root H N corresponds with phosphorus, on the supposition that the combining weight of that body is 15.12, and that phosphorus, accordingly, is resolvable into HCO. It cannot be maintained that this actually is so, because we are in some doubt as to the true combining weight of phosphorus. But probabilities exist in favour of the opinion, founded on the analogy between phosphorus and the ammoniacal group of bodies. The resemblance between phosphoretted hydrogen and ammonia, both the result of the putrefactive fermentation of animal matter, was before adverted to; and phosphoretted hydrogen, upon the hypothesisH N+H3 — HN, which is the precise composition assigned to ammonium. Further, amidogen corresponds in the number and ratio of its elements with sulphur; and sulphur, we have seen, is directly derived from nitrogen.

Thus, in every view that we can take of the chemical constitution of bodies, we are compelled to refer them to a common origin. No other hypothesis will save us from contradictory conclusions, or accord with the results which experiment has established.

III.-LAW OF CONTINUOUS COMBINATIONS

The principle of chemical combinations which I have thus far endeavoured to establish is, that from certain roots or elements, all the other bodies in the descending order, simple or compound, may be derived. We have seen, that two suppositions may, with almost equal reason, be made; first, that all bodies may be resolved into hydrogen, carbon, and oxygen; and secondly, that all bodies may be resolved into hydrogen and carbon. Either supposition may be adopted, and this consistently with observed results; but the latter supposition is in accordance with the principle, that we shall not assign more causes for an effect than are necessary to explain it. If we shall suppose that all bodies are resolvable into hydrogen and carbon, as common roots, we can equally explain the resulting effects, as if we shall suppose that the roots are three, hydrogen, carbon, and

oxygen.

If carbon were a simple multiple of hydrogen, then we could explain its derivation from the molecules of hydrogen, thus-A molecule of hydrogen h = 1, combines with a molecule of hydrogen h, and forms the molecule h + h = 2. The new molecule h+h combines with the molecule h, and forms the molecule h+h + h=3; and this again with the molecule h; and so on in an arithmetical series, until we reach the molecule h +5 h=6=C.

But carbon, it appears from experiment, is not a simple