THE question of the maintenance of solar energy is one that has been looked upon with deep interest by astronomers and physicists from the time of La Place downward. The amount of heat radiated from the sun has been approximately computed by the aid of the pyrheliomete>r of Pouillet and by the actinoineters of Herschel and others at 18,000,000 of heat units from every square foot of its surface per hour, or, put properly, as equal to the heat that would be produced by the perfect combustion every thirty- six hours of a mass of coal of specific gravity =1'5 as great as that of our earth. If the sun were surrounded by a solid sphere of a radius equal to the mean distance of the sun from the earth (95,000,000 of miles), the whole of this prodigious amount of heat would be intercepted; but considering that the earth's apparent diameter, as seen from the sun, is only seventeen seconds, the earth can intercept only the 2,^50-milJionth part. Assuming that the other planetary bodies swell the amount of intercepted heat hy ten times this amount, there remains the important fact that fftHSHo of the solar energy is radiated into space. and apparently lost to the solar system, and only -mf utilized. Notwithstanding this enormous loss of heat, solar temperature has not diminished sensibly for centuries, if we neglect the periodic changes apparently connected with the appearance of sun-spots that have been observed by Lockyer and others, and the question forces itself upon us how this great loss can be sustained without producing an observable diminution of solar temperature even within a human lifetime. Among the ingenious hypothesis intending to account for a continuance of solar heat is that of shrinkage, or gradual reduction of the sun's volume suggested by Helmholtz. It may, however, be urged against this theory that the heat so produced would be liberated throughout its mass, and would have to be brought to the surface hy conduction, aided perhaps by convection; but we know of no material of sufficient conductivity to transmit anything approaching the amount of heat lost by radiation. Chemical action between the constituent parts of the sun has also been suggested; but here again we are met by the difficulty that the production of such combination would ere this have accumulated on the surface, and would have formed a barrier against further action. These difficulties have led Sir Wm. Thomson, following up Mayer's speculation, to the suggestion that the cause of the maintenance of solar temperature might be found in the circumstance of meteorolites falling upon the sun from great distances in space, or witb an acquired velocity due to such fall , and be shows that each pound of matter so imported would represent a large number of heat units depending upon the original distance. Yet the aggregate of material that would thus have to be incorporated with the sun would tend to disturb the planetary equilibrium, and must ere this have shortened our year to an extent exceeding that resulting from astronomical records and observation. In fact, Sir William Thomson soon abandoned the meteoric hypothesis for that of simple transfer of heat from the interim' of a liquid sun to the surface by means of convection currents, which latter hypothesis appears at the present time to be supported by Prof. Stokes and other leading physicists of the day. But if either of these hypotheses could be proved, we should only bave the satisfaction of knowing that the solar waste of energy by dissipation into space was not dependent entirely upon loss of its sensible heat, but that its existence as a luminary would be prolonged by calling into requisition a limited , though may be large, store of energy II the form of separated matter. The true solut ion of the problem will be furnished by a theory, according to which the radiant energy which is now supposed to be dissipated into space and irrecoverably lost to our solar system. could be arrested and brought back in another form to the sun itself, there to continue the work of solar radiation. Some years ago it occurred to me tbat such a solution of the solar problem might not lie beyond the bounds of possibility, and although I cannot claim intimate acquaintance with the intricacies of solar physics, I have watched its progress, and have engaged also in some physical expcri ments bearing upon the question, all of which have served to strengtben my confidence. and ripened in me the determination to submit my views, not without some, misgiving to the touchstone of scientific criticism. For the purposes of my theory, stellar space is supposed to be filled with highly rarefied gaseous bodies, including hydrogen , oxygen. nitrogen, carbon, and their compounds, besides solid materials in the form of dust. Thisbeing the case, each planetary hody would attract to itself an atmosphere depending for its density upon its relative attractive importance, and it would not seem unreasonable to suppose that the heavier and less diffusible gases would form the staple of these atmospheres; that, in fact, they would consist mostly of nitrogen, oxygen, and carbonic anhydride, while hydrogen and its compounds would predominate in space. But the planetary system, as a whole, would exercise an attractive infiuence upon the gaseous matter diffused through * Paper read at the Royal Society, March 2, D.C. L.. LL.D., F.B S., Mem. !nst. C. E. by C. William Siemens,space, and would therefore be surrounded by aninterplanet- ary atmosphere, holding an intermediate position between the planetary atmospheres and the extremely rarefied stellar space. In support of this view it may be urged 1hat, in following out the molecular theory of gases as laid down by Clerk Maxwell, Clausius, and Thomson, it would be difficult to assign a limit to a gaseous atmosphere in space, and, further, that some writers, among whom I will here mention only Grove, Humboldt, Zoeilner, and Mattieu Williams, have boldly asserted the existence of a space filled wit h matter, and that Newton himself, as Dr. Sterry Hunt tells us in an interesting paper which has only just reached me, has expressed views in favor of such an assumption. Further than this, we have the facts that meteorolites whose flight through stellar, or at all events through interplanetary spacs, is suddeuly arrested by being brought into eollision with our earth, are known to contain as much as six times their own volume of gases taken at atmospheric pressure; and Dr. Flight has only very recently communicated to the Royal Society the analysis of the occluded gases of one of these meteorolites taken immediately after the descent to be as follows: ' CO. 0'12 CO 31 88 H 45'79 CH, 4 55 N 17 66 100 00 It appears surprising that there was no aqueous vapor, considering there was much hydrogen and oxygen in combination with carbon, but perhaps the vapor escaped observation, or was expelled to a greater extent than the other gases by external heat, when the meteorolite passed through our atmosphere. Opinions concur that the gases found occluded in meteorolites cannot be supposed to have entered into their composition during the very short period of traversing our - atmosphere, but if any doubt should exist on this head, it ought to be set at rest by the fact that the gas principally occluded is hydrogen, which is not eontained in our atmosphere in any appreciable quantity. Further proof of the fact that stellar space is filled with gaseous matter is furnished by spectrum analysis, and it appears from recent investigation by Dr. Huggins and others, that the nucleus of a comet contains very much the same gases found occluded in meteorolites, including “carbon, hydrogen, nitrogen, and probably oxygen,” while according to the views set forth by Dewar and Liveing, it also contains nitrogenous eompounds such as cyanogen. Adversely to the assumption that interplanetary space is filled with gases, it is urged that the presence of ordinary matter would cause sensible retardation of planetary motion, such as must have made itself felt before this; but assuming that the matter filling space is an almost perfect fluid, not limited by border surfaces, it can be shown, on purely mechanical grounds, that the retardation by friction through such an attenuated medium would be very slight indeed, even at planetary velocities. But it may be contended that, if the views here advocated regarding the distribution of gases were true, the sun should draw to itself the bulk of the least diffusible, and therefore the heaviest gases, such as earbonie anhydride, carbonic oxide, oxygen, and nitrogen, whereas spectrum analysis has proved on the contrary a prevalence of hydrogen. In explanation of this seeming anomaly it ean be shown, in the first place, that the temperature of the sun is so high, that such compound gases as carbonic anhydride and carbonic oxide could not exist within it, their point of dissociation being very much below the solar temperature; it haa been contended, indeed. by Mr. Lockyer, that none of the metalloids have any existence at these tempera! ures, although as regards oxygen, Dr. Draper asserts .its existence in the solar photosphere; there must be regions, however, outside that thermal limit, where their existence would not be jeopardized by heat, and here great accumulation of these comparatively heavy gases that constitute our atmosphere would probably take place, were it not for certain counterbalancing action. I here approach a point of considerable importance in my argument, upon the proof of which my further conclusions must depend. The sun completes one revolution on its axis in twenty- five days, and its diameter being taken at 882,000 miLs, it follows that the tangential velocity amounts to 12.3 miles per second. or to 4'41 times the tangential velocity of our earth. This high rotative velocity of the sun must cause an equatorial rise of the solar atmosphere, to which Mairan, in 17.i1, attributed the appearanee of zodiacal light. La Place rejected this explanation on the ground that the zodiaeal light extended to a distance from the sun exceeding our own distance, whereas the equatorial rise of t he solar atmosphere due to its rotation could not exceed 9-20ths of the distance of Mercury. But it must be remembered that La Place based his calculation upon the hypothesis of an empty stellar space (filled only with an imaginary ether), and that the result of solar rotation would be widely different, if it was supposed to take place within a medium of unbounded extension. In this case pressures would be balanced all round, and the sun would act mechanically upon the floating matter surrounding it in the manner of a fan, drawing it toward itself upon the solar surfaces, and projecting it outward in a continuous disk-like steam. By this fan action, hydrogen, hydrocarbons, and oxygen are supposed to be drawn in enormous quantities toward the polar surfaces of the sun; during their gradual approach, they will pass from their condition of extreme attenuation and extreme cold to that of compression, accompanied with rise of temperature, until, on approaching the photosphere, they burst into flame, giving rise to a great development of heat, and a temperature commensurate with their point of dissociation at the solar density. The result of their combustion will be aqueous vapor, and carbonic anhydride or oxide, according to the sufficiency or insufficiency of oxygen present to complete the combustion; and these products of eom- hustion. in yielding to the influence of centrifugal foree, will flow toward the solar equator and be thence projected into space. The next question for consideration is: What would become of these products of combustion when thus rendered back into space? Apparently they would gradually change the condition of stellar material, rendering it more and more neutral; but I venture to suggest the possibility, nay, the probability, that solar radiation would, under these cil cum- stances, step in to bring baek the combined materials to a condition of separation by a process of dissociation carried into effect at tbe expense of that solar energy whieh is now supposed to be lost to our planetary system. According to the law of dissociation, as developed by Bun3en and Sainte-Claire Deville, the point of dissociation of different compounds depends upon the temperature on the one hand, and upon the pressure on the other. Aceording to Sainte-Claire Deville, the dissociatiou tension of aqueous vapor of atmospheric pressure and at 2,800° C. is 05, or only half of the vapor car' exist as such, its remaining half being found as a mechanical mixture of hydrogen and oxygen, but that with the pressure, the temperature of dissociation rises and falls, as the temperature of saturated steam rises and falls with its pressure. It is therefore conceivable that the temperature of the solar photosphere may be raised by combustion to a temperature exceeding 2.800° C., whereas dissociation may be effected in space at a lower temperature. But these investigations had reference only to heats measured by means of pyrometers, but do not extend to the effects of radiant heat. Dr. Tyndall has shown by his exhaustive researches that vapor of water and other gaseous compounds intercept radiant heat in a most remarkable degree, and there is other evidence to show that radiant energy from a source of high intensity possesses a dissocwting power far surpassing the measurable temperature to which the compound substance under its influence is raised. Thus carbonic anhydride and water are dissociated in the leaf cells of plants, under the influence of the direct .solar ray at ordinary summer temperature, and experiments, in which Ihave been engaged for nearly three years, * go to prove that this dissociating action is obtained also under the radiant, influence of the electric arc, although it is scarcely perceptible if the source of radiant energy is such as can be produced by the combustion' of oil or gas. The point of dissociation of aqueous vapor and carbonic anhydride admits, however, of being determined by direct experiment. It engaged my attention some years ago, but I have hesitated to publish the qualitative results I then obtained, in the hope of attaining to quantitative proofs. These experiments eonsisted in the employment of glass tubes, furnished with platinum electrodes, and filled with aqueous vapor, or with earbonie anhydride in the usual manner, the latter being furnished with caustic soda to regulate the vapor pressure by heating. Upon immersing one end of the tube charged with aqueous vapor ill a refrigerating mixture of ice and chloride of calcium, its temperature at the end was reduced to 22° C.. corresponding to a vapor pressure, according to Regnault, of 1-1800 of an atmosphere. When so eooled no slow electric discharge took plaee on connecting the two eleetrodes with a small induction coil. I then exposed the end of the tube projecting out of the freezing mixture, backed by white paper, to solar radiation (on a dear summer's day) for several hours, when upon again connecting up to the inductorium, a discharge, apparently that of a hydrogen vacuum,was obtained. This experiment being repeated furnished unmistakable evidence, I thought, that aqueous vapor had been dissociated by exposure to solar radiation. The CO2 tubes gave, however, less reliable results. Not satisfied with these qualitative results, I made arrangements to collect the permanent gases so produced by means of a Sprengel pump, but was prevented by lack of time from pursuing the inquiry, which I purpose, however, to resume shortly, being of opinion that, independently of my present speculation, the experiments may prove useful in extending our knowledge regarding the laws of dissoeiation. Assuming, for my present purpose, that- dissociation of aqueous vapors was really effected in the experiment just described, and assuming, further, that stellar space is filled with aqueous and other vapor of a density not exceeding the 1-2000th part of our atmosphere, it seems reasonable to suppose that its dissociation would be effected by solar radiation, and that solar energy would thus be utilized. The presence of carbonic anhydride and carbonic oxide would only serve to facilitate the decomposition of the aqueous vapor by furnishing substances to eomhine with nascent oxygen and hydrogen. By means of the fan-like action resulting from the rotation of the sun, the vapor dissociated in spaee to-day would be drawn toward the polar surfaces of the sun to-morrow, be heated by increase in density, and would burst into flame at a point where both their density and temperature had reached the necessary elevation to induce combustion, each eomplete cycle taking, however, years to be accomplished. The resulting aqueous vapor, carbonic anhydride and earhonic oxide, would be drawn to • See Proceedings Roy. Soc.,..o1. xxx. . 1 March, 1880, and a paper read before Section A of the British Association 1 Sept., 1881, and ordered to be printed in the Report, ward the equatorial regions, and be then again projected into space by centrifugal force. Space would, according to these views, be filled with gaseous compounds in process of decomposition by solar radiant energy, and the existence of these gases would furnish an explanation of the solarabsorption spectrum, in which the lines of some of the substances may be entirely neutralized and lost to observation. As regards the heavy metallic vapors revealed in the sun by the spectroscope, it is assumed that these form a lower and denser solar atmosphere, not participating in the fan-like action which is supposed to effect the light outer atmosphere only, ' in which hydrogen is the principal factor. Such a dense metallic atmosphere could not participate in the fan action affecting the lighter photosphere, because this is only feasible on the supposition that the density of the in-flowing current is, at equal distances from the gravitating eenter, equal or nearly equal to the outfiowing current. It is true that the products of combustion of hydrogen and carhonic oxide are dense r than their constituents, but this difference may be balanced by their superior temperature on leaving the sun, whereas the metallic vapors would be unbalanced, and would therefore obey the laws of gravitation, recalling them to the sun. On the surface of contact between the two solar atmospheres, intermixture induced by friction must take plaee, however, giving rise perhaps to those vortices and explosive effects which are revealed to us by the telescope, and have been commented on by Sir John Herschel and other astronomers. Some of the denser vapors would probably get intermixed and carried away mechanieally by the lighter gases, and give rise to that eosmic dust which is observed to fall upon our earth in not inappreciable quantities; excessive intermixture would be prevented by the intermediary neutral atmosphere, the penumbra. As the whole solar system moves through space at a pace estimated at 150,000,000 of miles annually (being about one- fourth of the velocity of the earth in its orbit), it appears possible that the condition of the gaseous fuel supplying the sun may vary according to its state of previous decomposition, in which other heavenly bodies may have taken part. May it not be owing to such differences in the quality of the fuel supplied that the observed variations of tbe solar heat may depend ? and may it not be in consequence of such changes in the thermal eondition of the photosphere that sun-spots are formed? The views here advocated could not be thought acceptable unless they furnished at any rate a consistent explanation of the still somewhat , mysterious phenomena of the zodiacal light and of comets. Regarding the former, we should be able to return to Mairan's views, the objection by La Place being met by a continuous outward flow from the solar equator. Luminosity would be attributable to particles of dust emitting light reflected from the sun, or by phosphorescence. But there is another cause for luminosity of these partieles, which may deserve a passing.con- sideration. Eaeh partiele would be electrified by gaseous frietion in its acceleration, and its electric ten sion would be vastly increased in its forcible removal, in the same way as the fine dust of the desert has heen observed by Werner Siemens to be in a state of high electrification on the apex of the Cheops Pyramid. Would not the zodiacal light also find explanation by slow eleetric discharge backward from the dust toward the sun? and would the same cause not account for a great difference of potential between the sun and earth, whieh latter may be supposed to be washed by the solar radial current? May not the presence of the current also furnish us with an explanation of the fact that hydrogen, while abounding apparently in space, is praeti- cally absent in our atmosphere, where aqueous vapor, which may be partly derived from the sun, takes its place? An aetion analogous to this, though on a much smaller scale, may be set up also by terrestrial rotation giving rise to an electrical discharge from the outgoing equatorial stream to the polar regions, where the atmosphere to be pierced by the return flood is of least resistance. It is also important to show how the phenomena of comets could be harmonized with the views here advocated, and I venture to hope that these occasional visitors will serve to furnish us with positive evidence in my favor. Astronomical physicists tell us that the nucleus of a comet consists of an aggregation of stones similar to meteoric stones. Adopting this view, and assuming that the stones have absorbed in stellar space gases to the amount of six times their volume, taken at atmospheric pressure, what, it may be asked, will be the effect of such a mass of stone advancing toward the suii at a velocity reaching in perihelion the prodigious rate of 366 miles per second (as observed in the comet of 1845), being twenty-three times our orbital rate of mot ion? It appears evident that the entry of such a divided mass into a comparatively dense atmosphere must be accompanied by a rise of temperature by frictional resistance, aided by attractive condensation. At a certain point the increase of temperature must cause ignition, and the tieat thus produced must drive out the occluded gases, which in an atmosphere 3,000 times less dense than that of our earth would produce 6x3,0t)O=18,0OO times the volume of the stones themselves. These gases would issue forth in all directions, but would remain unobserved except in that of motion, in which they would meet the interplanetary atmosphere with the compound velocity and form a zone of intense combustion, such as Dr. Huggins has lately observed to surround the one side of nucleus, evidently the side of forward motion. The nucleus would thus emit original light, whereas the tail may be supposed to consist of stellar dust rendered luminous by reflex action produced by the light of the sun and comet combined, lis foreshadowed already by Tyndall, Tate, and others, starting each from different assumptions. These are in brief the outlines of my reflections regarding this most fascinating question which I venture to put before tbe Royal Society. Although I cannot pretend to an intimate acquaintance with the more intricate phenomena of solar physics, I have long had a conviction, derived princ:- Cally from familiarity with some of the terrestrial effects of eat. that the prodigious and seemingly wanton dissipation of solar heat is unnecessary to satisfy accepted principles regarding the conservation of energy, but that it may be arrested and returned over and over again to the sun, in a manner somewhat analogous to the action of the heat recuperator in the regenerative gas furnace. The fundamental conditions are: 1. That aqueous vapor and carbon compounds are present- in stellar or interplanetary space. 2. That these gaseous compounds are capable of being dissociated by radiant solar energy while in a state of extreme attenuation. 3. Th-it these dissociated vapors are capable of being compressed into the solar photosphere by a process of interchange with an equal amount of rpassociated vapors, this interchange being effected by the centrifugal action of the sun itself. If these conditions could be substantiated, we should gain the satisfaction that our solar system would no longer impress us with the idea of prodigious waste through dissipation of energy into space, but rather with that of well- ordered self-sustaining action, capable of perpetuating^solar radiation to the remotest future.