Considerable additions have been made to the number of elements during late years. The reason of this is that the spectroscope has indicated the presence of metals previously unknown, and by the use of proper means they have been obtained in a separate condition. These substances are termed elements because we do not at present know how to split them up in such a manner as to prove that they are compounds. But it is not improbable that they are compounds, just as we now know that potash and soda are compounds; although before Sir Humphry Davy split them up into oxygen and a metal they were supposed to be elements. Indeed, recently much evidence has been brought to show that the substances which we call elements are really compounds.

It is from an examination of the spectroscopic character of the elements at different degrees of temperature that Lockyer has been able to obtain sufficient data to justify the definite formulation of the hypothesis that all the elements we know are really compounds, or, to speak perhaps more precisely, are really different forms of aggregation of one kind of matter.1 According to this hypothesis the matter of which the universe is composed was at one time equally distributed through space, and uniform in kind. The atoms then coalesced in various groups of two, three, or more; and these, again grouping themselves together still further, formed aggregates of more and more complex composition. These aggregates are, it is supposed, the elements with which we are acquainted. Most of those complex molecules are perfectly stable at ordinary temperatures; and so their composition remains constant under the conditions usual at the surface of this earth.

But when they are subjected to increased temperatures in the laboratory, rising from that of the Bunsen lamp to the electric arc, and then to the electric spark or to still higher temperatures in the sun, their spectroscopic appearances give evidence of decomposition into simpler molecules. When the elements are subjected to cold and pressure the molecules which compose them come closer together, and we get them forming a solid substance. Heat tends, by communicating vibrations to them, to shake the molecules further apart, and to produce a liquid condition. Still greater heat shakes the molecules further apart still, and produces a gaseous condition.

In all those conditions the molecules of the element become more complex by reduction of temperature or increase of pressure, and simpler by increase in temperature or reduction in pressure.2 Exceedingly great heat or electricity appears to shake apart still further the constituents of the element, so as to resolve it into simpler combinations of the elementary substance of which, according to the hypothesis, it is composed.

This shaking apart of the component elements is known to exist lncom1 Lockyer, Phil. Trans. 1874, p. 492 et seq.

2 According to another hypothesis, bodies are supposed to have molecules of one degree of complexity, and the difference between solid, liquid, and gaseous bodies is pounds, and to it the name of dissociation has been given. Thus when chalk or limestone is exposed to the action of heat it becomes dissociated into carbonic acid and lime, CaCO3 = CaO + CO2. This process is readily reversible by reversing the conditions. Thus the lime and carbonic acid which are dissociated by heat readily recombine in the cold CaO + CO2 = CaCO3.

When matter is solid the molecules of which it is composed are supposed to be large and close together. When in the state of vapour or gras, these molecules are smaller and much further apart.

Solid, liquid, or densely gaseous matter, when its molecules are agitated by heat, gives a continuous spectrum. Gaseous and vaporous matters, when their molecules are agitated at lower pressures or higher temperatures by heat or electricity, give a discontinuous spectrum consisting of bands or lines.

Between those extremes we have, as a rule, three other intermediate kinds of spectra : first, a continuous spectrum in the red; next, a continuous spectrum in the blue; next, a fluted spectrum, and after that the line spectrum already mentioned.

In all those kinds of spectrum, however, we are supposing that the elementary molecules are still intact; they are only more or less separated.

Compound bodies, like simple bodies, give definite spectra. The spectrum of a simple metal consists of lines which increase in number and thickness as the pressure of the vapour or its quantity in a given space is increased. The spectrum of a compound body consists chiefly of channelled spaces and bands which increase in the same manner. The greater the number of molecules in a cubic inch or cubic millimetre, and the more violently they are agitated, the more complex is the spectrum until it becomes continuous.

The smaller the number of molecules in a given space, the more simple is the spectrum, which then consists of a few lines only.

When a compound is exposed to heat, so as to dissociate it into its component parts the spectroscopic bands characteristic of the compound become thinner, and the lines of the metal increase in number, as shown in the accompanying diagram where the bands exhibited by calcium chloride in the flame of a Bunsen's burner, disappear, and are replaced by lines only, when an electric spark is used. When an element is treated with more and more heat and electricity it likewise gives exactly the same kind of evidence of dissociation - bands disappearing, and lines becoming thinner. Besides this, new lines make their appearance with every large increase of temperature.

Fig. 1.   Spectrum of calcium chloride

Fig. 1. - Spectrum of calcium chloride. (1) In the flame of a Bunsen's burner, showing the channelled spaces and bands of a compound. (2) In an electric spark, showing the lines of the element calcium. (After Roscoe.)