59

represent the constitution of ethylene. But it may be objected that this is pure fiction, and that it would be simpler to admit that in ethylene carbon plays the part of a trivalent element, H,C-CH,, the two carbon atoms being united by a single exchange of atomicities. It is not a fiction, it is in accordance with facts, for we must not forget that all known hydrocarbons contain an equal number of hydrogen atoms. This would not be the case if carbon could play the part of a triatomic or trivalent element; if so, methyl, CH,, and ethyl, C2HË, should exist in a free state. We must, therefore, conclude, taking experiment as our authority, that in the combinations of carbon and hydrogen carbon is never bivalent, as it is in carbon monoxide; methylene, CH2, does not exist that it is never trivalent; methyl, CH,, and ethyl, C2H,, do not exist. It is therefore quadrivalent or tetratomic, and we are thus led to admit that carbon atoms have the faculty of exchanging with each other several units of saturation. But the combination thus constituted is in a state of unstable equilibrium, which is destroyed by the intervention of chlorine. The latter can fix itself upon molecules so formed, thus destroying the double link and constituting a perfectly saturated molecule. It is true that the affinity of carbon for carbon is strong; but when two atoms of this simple body have exchanged two units of saturation this affinity cannot stand, as far as the second bond is concerned, against that of chlorine, which tends to fix itself on to both the groups CH2. The following formulæ will explain this view of the case :

Ethylene and analogous radicals have, therefore, the power of directly fixing chlorine and other elements, because they contain atoms of carbon, the combining capacity of which is not exhausted; since it is twice. exerted between carbon atoms, it can still be manifested towards the atoms of chlorine. The latter severally fix themselves upon an atom of carbon in olefiant gas, although they refuse to unite directly with free carbon; the affinities of this body are, in fact, very different according as it is considered in the state of a simple compact and condensed body, C1, or in a state of combination with hydrogen and in a gaseous form.'

It is a circumstance worthy of remark that chlorine or bromine, which do not unite with free carbon, as does oxygen, can fix themselves directly upon the unsaturated hydrocarbons, which oxygen cannot do. Are we to conclude that the chlorine is attracted, not by the carbon, but by the entire ethylene group acting as a radical, as was formerly supposed? This would be going a step backwards. In my opinion it is unquestionably the unsaturated carbon which attracts or admits the chlorine: it attracts it because it occurs in gaseous combination with hydrogen; it admits it because there are two vacant places in the system. The saturated hydrocarbon C„H。 also attracts chlorine; but, as there is no vacant place in the system. it can only admit it by losing two atoms of hydrogen. The hydrogen atoms seem, therefore, to exercise an influence upon the property possessed by carbon of fixing chlorine-that is to say, of admitting this element into its sphere of action. Such an influence is exercised in other instances and by other elements. Ethylene, which fixes chlorine, is incapable of directly fixing oxygen, but dibromethylene, CH¿Bг2, can fix it, according to Demole, to form the compound CH„BгO-Br (bromacetyl bromide). In this case, as in the former, the affinities of carbon have been modified by the intervention of other elements-hydrogen or bromine.

II.

The foregoing considerations upon the hydrocarbon radicals apply to all compounds capable of directly fixing elements, which compounds, in virtue of this property, resemble radicals. These elements are attracted by one or other of the unsaturated atoms contained by the compound in question. Let us take some examples.

Carbon monoxide can directly fix oxygen or chlorine because the bivalent carbon which it contains is not saturated. In carbonyl chloride and carbon dioxide. the carbon has become quadrivalent. Like carbon monoxide, sulphurous acid gas can fix oxygen or chlorine, and it is the sulphur which attracts these elements. In sulphuryl chloride and in anhydrous sulphuric acid the sulphur has become sexvalent.1

Phosphorus trichloride, in fixing directly two atoms of chlorine, behaves, in some respects, like a radical, and it owes this property to the unsaturated phosphorus

Cl

Cl

29

1 Sulphurous acid gas being S1O, or = S1O2, sulphuryl chloride is но HOS.

O sulphuric acid /OCI The formula S" which might be attributed to sulphuryl \OCI' chloride, does not seem to us probable, since oxygen does not possess any tendency to unite with chlorine, and because the properties of sulphuryl chloride are not those of a body containing the hypochlorous residue OCl, which would decompose with explosion. We may add that selenium and tellurium are manifestly quadrivalent in the tetrachlorides and sexvalent in the anhydrides and selenic and telluric acids.

which it contains. The difference between the two views of the mode of action displayed by radicals is here shown in a most striking manner. It was formerly maintained that phosphorus pentachloride should be regarded as a combination of phosphorus trichloride with chlorine; the trichloride exists in it as a whole, as a radical endowed, as such, with a power of combination. We say now that the trichloride can take up chlorine because the phosphorus which it contains is not saturated; in the pentachloride phosphorus is united directly with five atoms of chlorine, and when phosphorus trichloride takes up two atoms of chlorine the latter are attracted by the unsaturated atom of phosphorus. When phosphene fixes hydriodic acid, or when ammonia unites directly with hydrochloric acid, they also act as radicals, and owe this property to the atom of phosphorus or of nitrogen which they contain, both of which show a tendency to become further saturated. In hydriodate of phosphene (phosphonium iodide), as in hydrochlorate of ammonia (ammonium chloride), they become quinquivalent.

In organo-metallic radicals properly so called we find properties of the same order, which we interpret in the same manner. And it must be confessed that these ideas upon the saturating capacity of elements, a capacity varying with the combinations in which they occur, are the natural consequence of the experiments undertaken twenty years ago upon the class of compounds in question. We refer to the classical discoveries of Frankland, Baeyer, Cahours, and the ingenious views which they introduced into science.

When Frankland compared with each other stannic iodide, stannethyl iodide, and stannic ethide, expressing the composition of these bodies by the formulæ

surely he showed by this notation that in these three bodies iodine and ethyl are combined in the same manner with tin, and that stannethyl, SnC,H,, only plays the part of radical because the tin which it contains tends to pass into the state in which it exists in stannic iodide. Stannethyl, SnEt, has just as much claim to be considered as a radical as stannous iodide, and in both cases it is the tin itself, and not the radical considered as a whole, which attracts the iodine.

And in his masterly statement of the theory of the saturation in the organo-metallic compounds of tin Cahours referred the power of attracting either chlorine, methyl, or ethyl, in order to attain a stable molecular equilibrium, to the tin itself, so that the general composition of all these saturated compounds might be expressed by the formula

SnX.2

These formula are on the old notation: C = 6, Sn = 59.

= 118,

2 Cahours wrote Sn,X. With the atomic weight of tin, Sn = this expression becomes SnX, and the saturated compounds of tin receive, in consequence, the following formulæ :—