Next take the case of a metal, platinum or potassium, constituted, according to the atomic theory, in the same manner. The metal is a conductor; but how can this be, except space be a conductor, for it is the only continuous part of the metal, and the atoms not only do not touch, (by the theory,) but, as we shall see presently, must be assumed to be a considerable way apart. Space, therefore, must be a conductor, or else the metals could not conduct, but would be in the situation of the black sealing-wax referred to a little while ago.

But if space be a conductor, how then can shellac, sulphur, &c. insulate? for space permeates them in every direction. Or, if space be an insulator, how can a metal or other similar body conduct?

It would seem, therefore, that in accepting the ordinary atomic theory, space may be proved to be a non-conductor in non-conducting bodies, and a conductor in conducting bodies; but the reasoning ends in this, a subversion of that theory altogether, for if space be an insulator, it cannot exist in conducting bodies, and if it be a conductor, it cannot exist in insulating bodies. Any ground of reasoning which tends to such conclusions must in itself be false.

In connection with such conclusions, we may consider shortly what are the probabilities that present themselves to the mind, if the extension of the atomic theory which chemists have imagined be applied in conjunction with the conducting powers of metals. If the specific gravity of the metals be divided by the atomic numbers, it gives us the number of atoms, upon the hypothesis, in equal bulks of the metals. In the following table the first column of figures expresses nearly the numbers of atoms in, and the second column of figures the conducting power of, equal volumes of the metals named:

Atoms.Conducting power.
1·00gold6·00
1·00silver4·66
1·12lead0·52
1·30tin1·00
2·20platinum1·04
2·27zinc1·80
2·87copper6·33
2·90iron1·00

So here iron, which contains the greatest number of atoms in a given bulk, is the worst conductor excepting one. Gold, which contains the fewest, is nearly the best conductor; not that these conditions are in inverse proportions, for copper, which contains nearly as many atoms as iron, conducts better still than gold, and with above six times the power of iron. Lead, which contains more atoms than gold, has only about one-twelfth of its conducting power; lead, which is much heavier than tin and much lighter than platina, has only half the conducting power of either of these metals. And all this happens among substances which we are bound to consider at present as elementary or simple. Whichever way we consider the particles of matter and the space between them, and examine the assumed constitution of matter by this table, the results are full of perplexity.

Now let us take the case of potassium, a compact metallic substance with excellent conducting powers—its oxide or hydrate a non-conductor; it will supply us with some facts having very important bearings on the assumed atomic construction of matter.

When potassium is oxidized, an atom of it combines with an atom of oxygen to form an atom of potassa, and an atom of potassa combines with an atom of water, consisting of two atoms of oxygen and hydrogen, to form an atom of hydrate of potassa, so that an atom of hydrate of potassa contains four elementary atoms. The specific gravity of potassium is 0·865, and its atomic weight 40·; the specific gravity of cast hydrate of potassa, in such a state of purity as I could obtain it, I found to be nearly 2; its atomic weight, 57. From these, which may be taken as facts, the following strange conclusions flow: A piece of potassium contains less potassium than an equal piece of the potash formed by it and oxygen. We may cast into potassium oxygen, atom for atom, and then again both oxygen and hydrogen in a twofold number of atoms, and with all these additions the matter shall become less and less, until it is not two-thirds of its original volume. If a given bulk of potassium contains 45 atoms, the same bulk of hydrate of potassa contains 70 atoms nearly of the metal potassium, and, besides that, 210 atoms more of oxygen and hydrogen. In dealing with assumptions, I must assume a little more for the sake of making any kind of statement; let me therefore assume that in the hydrate of potassa the atoms are all of one size and nearly touching each other, and that in a cubic inch of that substance there are 2800 elementary atoms of potassium, oxygen, and hydrogen; take 2100 atoms of oxygen and hydrogen, and the 700 atoms of potassium remaining will swell into more than a cubic inch and a half; and if we diminish the number until only those containable in a cubic inch remain, we shall have 430, or thereabout. So a space which can contain 2800 atoms, and among them 700 of potassium itself, is found to be entirely filled by 430 atoms of potassium, as they exist in the ordinary state of that metal. Surely, then, under the suppositions of the atomic theory, the atoms of potassium must be very far apart in the metal, i. e. there must be much more of space than of matter in that body; yet it is an excellent conductor; and so space must be a conductor, but then what becomes of shellac, sulphur, and all the insulators? for space must also, by the theory, exist in them.

Again, the volume which will contain 430 atoms of potassium, and nothing else while in the state of metal, will, when that potassium is converted into nitre, contain very nearly the same number of atoms of potassium, i. e. 416, and also then seven times as many, or 2912 atoms, of nitrogen and oxygen beside. In carbonate of potassa, the space which will contain only the 430 atoms of potassium as metal, being entirely filled by it, will, after the conversion, contain 256 atoms more of potassium, making 686 atoms of that metal, and in addition 2744 atoms of oxygen and carbon.

These and similar considerations might be extended through compounds of sodium and other bodies, with results equally striking, and indeed more so, when the relations of one substance, as oxygen and sulphur, with different bodies are brought into comparison.