In this paper the term galactic system will mean not only the milky way, but also the caps of nebulas at its poles. The term milky way will be used to imply the galaxy itself. The term cosmic system will be applied to aggregations of dimensions comparable to the Magellanic clouds and to our own galactic system. (In my earlier papers these masses were called universes; but Lord Kelvin and Lord Rayleigh independently pointed out that this name might mislead, I have consequently substituted cosmic systems.) The term cosmic system of the first order applies to aggregations without definite structure, in which no general collision has occurred. In systems of the second order, a single general collision has taken place, and the symmetry is perfect. All other cosmic systems are of the third order. In these more than one general collision has occurred, and there is too much symmetry to be of the first order, and too little symmetry to be of the second. The visible universe (the galactic system of which our solar system is a part) is consequently a system of the third order.
The group of recently-discovered monatomic elements that have no combining power, namely, helium, neon, argon, crypton, and xenon, I call cosmic pioneers. They are practically page 217 always independent atoms, and probably play an important part in laying the foundation of an incipient cosmic system, helium being the most important. Possibly these elements have no other function than this, as deductions from their properties suggest that they must largely pass out of cosmic systems before the system matures. Hydrogen plays the same initial part, but it is more than a cosmic pioneer; it has important functions in cosmic systems of all orders.
This paper is chiefly devoted to grazing and whirling collisions of celestial bodies.
Grazing collisions of stars were discussed before the Royal Society by Dr. Johnstone Stoney more than thirty years ago. The formation of double stars and new stars by such an occurrence was suggested by him, as well as the probability of the existence of dead suns in countless hosts.
The especial point I wish to forward is that a grazing impact will generally result in the formation of a new body, whilst the two struck stars proceed on their journey; as it were, flint and steel have struck and have cut off a part from each other, that results in an intensely heated spark.
Because the non-colliding parts are but little affected by the collision I call such a phenomenon a "partial impact."
With stars of the same order of dimensions that our sun has, the velocity developed by mutual attraction will be hundreds of miles a second. When by impact this motion is converted into heat in the coalesced parts, the temperature will be practically the same, whatever the amount struck off; if the graze be small, the attractive power of the new body will also be small, and it is evident that the velocity of the molecules may be great enough for every molecule to have more than the critical velocity; each molecule as it reaches the surface will consequently leave the body, never to return.
The temperature may easily be from ten to a hundred million centigrade. It will of course vary with the chemical composition.
Thus is produced in less than an hour a fiery mass expanding about a million miles an hour, and this increase in size will cause for a time an increase in brilliancy. Presently, however, the radial direction of the molecules will tend to cause their motion to become parallel, and will lessen the number of molecular encounters, and consequently the amount of radiation, so that after a time our bright star becomes a planetary nebula. In a year or so each molecule is wandering alone; the brilliant body and the nebula are gone. The body has lost its light, not by cooling, but by being too hot to hold together.page 218
The two stars that struck have been heated where they were sheared, and they are separating at a speed of hundreds of miles a second. Hence the spectrum of our nova is made up of a continuous spectrum, with broad, bright, indistinct bands produced by the expanding gas, and on this band are superimposed two other lines, bright or dark, dependent on the position from which we view the lake of fire produced by the impact.
Obviously the tangential retardation will cause rotation, and the cut stars may alternately show their light and dark faces. Thus two variable stars are produced at once; generally this variability will tend to die out more quickly in one than in the other, yet there are many such pairs still existing.
It is certain that such pairing is not the result of chance. Whatever the explanation offered to account for variable stars must account also for the existence of pairs. (The accompanying diagram represents such a series of phenomena. With bodies of solar density the time taken to produce the changes shown in the series is less than two hours. The mass of the bodies makes no difference in the time, as with bodies of equal density the velocity acquired by gravitation is proportional to the diameter.)
The middle body attracts and retards the escaping stars, and may wed them into a pair.
Then, were no other agency to come into play, the pair would return to impact again, but long before they attain aphelion distance the central mass (consisting as it does of gas above the critical velocity) will have fled into space.
Hence the only force that attracts the stars back again is their own mass, and consequently, instead of colliding, the stars move in the ordinary double-star orbit. Double stars, when first connected, would be variable, and would be associated with nebulæ; this is actually the case, and any satisfactory account of double stars must explain these facts.
If the two stars had had a considerable proper motion they would not have been orbitally connected, and they would constantly increase their distance from each other.
This is doubtless the condition of the unassociated variable stars that are in pairs, and it is possible that their increase in distance could be observed.
Supposing too much had been cut off and coalesced, and the attraction were consequently too great for the heat to give every molecule more than its critical velocity; on attaining equality of temperature the light atoms would rob the others of their energy and escape.page 219
Diagram showing an Impact of two dead Suns, forming a temporary and two variable Stars.page 220
Hydrogen at the same temperature has sixteen times the tendency to escape that oxygen has, and two-hundred and eight times that of lead. This tendency of the chemical elements to sort themselves I call "selective molecular escape."
Hence at every cosmic impact of dense bodies some) light molecules leave with such extreme velocity as to escape not merely the mass, but the galactic system altogether. These molecules wander in space, perchance to other cosmic systems.
Another agency is at work giving motion to free molecules. Radiant energy is caught by cosmic dust of all dimensions. Sir W. Crookes's experiments on "Radiant Matter" suggest that free molecules do not take up or give out radiation. (Dr. Johnstone Stoney has lately suggested to me that this point is unimportant, as even should the molecules absorb radiation, this energy will increase the velocity of the succeeding rebound.) But when slowly moving light molecules touch this heated dust, it will bound off in the same way that molecules fly with increased velocity from radiometer-vanes. Thus radiant energy is converted into heat, and this into potential energy.
There are other agencies by which light atoms are liberated from cosmic systems to wander indiscriminately. Such atoms do work against the attraction of systems, and where potential is highest they move slowest.
Where they thus linger they tend to accumulate. The potential of this part of space lessens, and the work required to reach these positions not being so great as at first, oxygen and other heavier molecules get there, increasing the density; and oxygen also tends to produce non-volatile compound molecules.
These would coalesce; but helium and the other cosmic pioneers do not combine, they remain permanently gaseous. Thus a primary cosmic system is incipient. Dense bodies sent out of cosmic systems by the interaction of three bodies would generally pass through old cosmic systems where matter is in dense masses, but evidently not through such vast gaseous aggregations as the incipient cosmic systems. The bodies would be retarded by the friction produced, and perchance volatilized, forming nucleii in the general mass; their mutual attraction would cause denser aggregations to occur, and a cosmic system of the first order would be produced. Two such systems colliding produce a system of the second order. This, colliding with any other cosmic system, produces a system of the third order. Our own galactic system is very probably a tertiary system.page 221
The kinematic condition of the impact would exactly produce such a system. It is now known to be a rough double spiral of stars, with sprays and streams of stars and two caps of nebulae. It is not difficult to picture the kinematic conditions necessary to form such a rough ring, or double spiral of stars with polar caps of nebulous matter.
Let us assume a complete whirling coalescence of two cosmic systems in which the part coming into collision is considerable. This heated part is in the centre of the system. Here all the material is volatilized, and the pressure produced can find no relief save axially; hence the system is, as it were, a short cannon open at each end, and the discharged gas spreads itself over the poles of the system.
This discharge, that is commenced by pressure, is finished by molecular escape. Globular nebulæ form in this gaseous matter by the attraction produced by wandering bodies plunging into the gas. The globular nebulæ so produced attract one another and become double nebulæ; they are then wrought into spindles, spirals, dumbbells, or rings by the kinematic peculiarities of the varying depths of impact.
It is significant that temporary stars, planetary nebulæ, and all the bodies likely to be produced by the impact of stars are in the milky way; and all the forms of nebulae deduced as resulting from the impact of nebulae are where we should expect them to be, namely, at the poles of the milky way.
If this generalization represents the mode of nature's action, then there is a possibility that the entire cosmos is immortal, and the present order but a phase of an eternal rhythm.
|(1)||Diffusion of heat by radiation.|
|(2)||This radiation, falling on the dust of space, heats it.|
|(3)||The heat of this cosmic dust is taken away by slowly moving light molecules having their velocity increased.|
|(4)||Free molecules are also sent out of systems by partial impacts, by selective molecular escape, and other agencies.|
|(5)||Free molecules will remain longest in the position of maximum potential where their motion is least, and will consequently tend to aggregate in the empty parts of space.|
By the interaction of three bodies the velocity acquired by one sometimes takes it out of the cosmic system.page break page 223
|(7)||Hydrogen and the cosmic pioneers then become a trap for wandering bodies that tend to be stopped and converted into dense nebula?.|
|(8)||These dense nebulas tend to attract surrounding gas; they cool and shrink, some ultimately forming solid bodies.|
|(9)||These bodies, by mutual attraction, give density to the new cosmic system.|
|(10)||Such systems are of the first order.|
|(11)||The impact of systems of the first order produces systems of the second order.|
|(12)||Any other impacts produce systems of the third order, of which our galactic system is a type.|
|(13)||The coalescence of two cosmic systems does not necessarily, as a final result, produce a system of a larger mass than the two original systems from which it was formed, as many agencies are tending to send matter out of the coalesced mass.|
|(14)||It is thus seen that dissipation of energy is but a part of a complex cyclical process; and there is consequently the possibility of an immortal cosmos in which we have neither evidence of a beginning nor promise of an end, the present being but a phase of an eternal rhythm.|
The accompanying diagrammatic scheme illustrates these agencies. It must be noted that bodies and systems are printed in italic capitals; and where several such are one above another it implies sequence of phenomena.
* Communicated by Prof. A. W. Rücker, Sec. R.S.