The genesis of the elements has, in truth, not yet been made the subject of coherent speculation. Current ideas regarding it imply a double course of change, by aggregation first, and subsequently by disintegration. And this should give us a twofold series of elements. On one side there should be fixed survivals from the advancing process, on the other, products of decomposition, continuously evolved, and even now accumulating. If the claim of helium to take rank among these last should be finally established, our conceptions of the nature and history of nebulæ might have to undergo a strange inversion; but the outcome of the researches in progress is still uncertain, and may be far off. It is, nevertheless, quite clear that the electronic theory of matter supplies no genuine explanation of the source of energy in the universe. What is given out when the atoms go to pieces must have been stored up when they were put together. Whence was it derived? This is the fundamental question which underlies every discussion concerning the maintenance of the life of suns. It is unanswered, and probably unanswerable. FOOTNOTES: [Footnote 100: It must at the same time be borne in mind that their total darkness is not proved. All that is certain is that their spectra are not bright enough to leave any impression on the exposed plates.] [Footnote 101: Thomson and Tait, _Natural Philosophy_, Appendix E, p. 494, edition 1890.] CHAPTER XIV OUR OWN SYSTEM Our sun is clearly middle-aged. It bears none of the marks associated with juvenility in stars, while its decrepitude is in the distant future. It is crossing, most likely, a level tract where recuperation so nearly balances expenditure that radiation can be maintained for an indefinite time at a high and fairly uniform standard. Stars of the solar type pursue the even tenor of their way with particularly few interruptions. They show little tendency to intrinsic variability. Their periodicity, when it exists, is due to the presence of a companion. Variables, in other words, belonging to the spectral family of our sun, are binary systems; and they are usually, if not always, non-eclipsing binaries, on the pattern of δ Cephei. Light changes can thus be impressed upon sunlike stars by external influence; they do not conspicuously arise through native instability.