448 



NATURE 



[August 8, 191 8 



parable with the mean density of a sphere of radius 

 r and mass equal to that of the sun. Planets hav- 

 ing been formed at many different distances from 

 the sun, it follows that the mass must have origin- 

 ally been widely distributed through the system. 

 The distribution of density and velocity being thus 

 known within sufficiently narrow limits, it can be 

 shown by the principle of the constancy of angular 

 momentum that if planets of the sizes of ours were 

 formed, the resulting central body could not pos- 

 sibly rotate so slowly as the actual sun. There is 

 no agency capable of reducing this rotation, and 

 it seems necessary to abandon completely those 

 hypotheses that require the solar system to have 

 been formed by the gradual condensation of a 

 nebula. 



We are led to inquire next whether planets could 

 come into being by a more rapid or catastrophic 

 process. Projection from the sun is not a possible 

 origin, for a body started in this way must neces- 

 sarily strike the sun again on its return and be 

 reabsorbed ; further, there is no reason why all 

 should revolve in the same direction. The tidal 

 theory appears to give a better account of the 

 present state of the system. According to this, a 

 star much more massive than the sun approached it 

 very closely, and raised on opposite sides of it two 

 projections, ju^t as the moon raises tides in the 

 earth ; but the scale of the disturbance was in this 

 case so enormous that the sun's gravitation was 

 unable to prevent a rupture from occurring. Thus 

 either one or two streams of matter were shot out 

 in a time comparable with a few months or years. 

 Being longitudinally unstable, they broke up into 

 a series of detached masses, perhaps before the 

 parts projected later had actually separated from 

 the sun. That such rupture could occur has been 

 proved by Jeans. The attraction of the disturbing 

 body produced the direct revolution (in the same 

 sense as the motion of the star relative to the sun) ; 

 some of the revolving matter returned into the sun 

 and gave it a direct rotation. The angular momen- 

 tum thus acquired was, of course, derived from the 

 transverse motion of the disturbing body relative 

 to the sun. 



The size of the deformed body has little influence 

 on its chance of being broken up. Thus the de- 

 tached masses might well have produced systems of 

 satellites affd developed direct rotations in the 

 majority of cases, though complete uniformity 

 could scarcely be expected on account of the num- 

 ber of complicating factors. The fission would 

 cease when the star had receded a sufficient dis- 

 tance ; thus the outer nuclei, being the first ejected, 

 would produce most satellites. It seems possible 

 also that some of these would be formed when the 

 nucleus and the sun were on opposite sides of 

 the star, and that the motion would therefore be 

 retrograde. All the bodies, having recently 

 formed part of the sun, would naturally be very 

 hot. 



The system after the passage of the star would 



therefore include a central sun surrounded by a 



number of heated planets, moving in direct 



orbits, and attended by satellites ; the most re- 



NO. 2545, VOL. lOl] 



mote planets would have most satellites. The 

 rotation of the sun would be direct ; the rotation 

 of each planet would be in the same sense as the 

 revolution of most of its satellites, and in most 

 cases this also would be direct, though a few ex- 

 ceptions might well occur, especially in the outer- 

 most sub-systems. In every point this agrees with 

 the existing solar system. The heated interior of 

 the earth, the building of mountains by compres- 

 sion, and the present heated state of the greater 

 planets are readily accounted for. The occurrence 

 of three retrograde satellites on the outskirts of 

 otherwise direct sub-systems presents a difficulty, 

 but not, I think, a serious one. 



In addition to the planets and satellites, however^ 

 there would be a considerable amount of gaseous 

 matter too light to be condensed into the nuclei, 

 and probably consisting mainly of hydrogen. This 

 would be pushed round by the planets as they 

 moved, but its resistance to oscillatory motions 

 would steadily reduce the eccentricities of their 

 orbits, which would initially be considerable. At 

 the same time its own viscosity and diffusion 

 would cause it partly to dissipate into outer space 

 and partly to be reabsorbed into the sun. The 

 zodiacal light is probably the last remnant of it. 

 The actual eccentricities of the planetary orbits 

 being now small, but definitely different from zero, 

 it seems that the time the medium took to degene- 

 rate and the time needed to produce a considerable 

 effect on the eccentricities must have been of the 

 same order of magnitude. These are capable of 

 being estimated in terms of the density of the 

 medium, the first being pro{5ortional and the 

 second inversely proportional to it. Thus the con- 

 dition that they are of the same order of magnitude 

 makes it possible to estimate very roughly both 

 the density and the time needed for the changes, 

 which is found to be of the order of 3x10^ years, 

 as nearly equal as could be expected to the age of 

 the earth indicated by its radio-active constituents 

 (about I -6x109 years). At the same time large 

 condensations would form around the larger 

 planets, and the resistances offered by these would 

 be so great, being proportional to the squares of 

 the masses, that their eccentricities would dimin- 

 ish rapidly. Thus it would be expected that, on 

 the whole, the larger planets would have smaller 

 eccentricities than the smaller ones ; this is the 

 case, Mercury and Mars having larger eccentri- 

 cities than Jupiter, Saturn, and Uranus, and the 

 earth than Neptune. Venus has a somewhat 

 smaller eccentricity than Neptune, but otherwise 

 the agreement is remarkable. 



It is certain that on the tidal theory the primi- 

 tive nuclei must have been very hot, and probably 

 fluid. Whether the larger ones were liquid or 

 gaseous is very doubtful, but the smaller ones, 

 including the asteroids and most of the satellites, 

 can be shown to possess too little gravitative 

 power to have been able to hold together in the 

 gaseous state, and must therefore have been 

 liquid or solid at the start. The satellites prob- 

 ably liquefied almost as soon as they were ex- 

 pelled from their primaries or from the sun^ on 



