DISCOVERY 



137 



primaries than the furthest of the others that there 

 seems to be a fundamental difference between them. 

 Further, as they were not within the compressed gas 

 aroimd the primaries, their eccentricities would not 

 have altered much ; and actually they have very large 

 eccentricities. 



The anomalous satellites of Uranus and Neptune are 

 in a different position again, for they are comparatively 

 close to their primaries. There is no quite satisfactory 

 explanation of their remarkable motions. Perhaps for 

 some reason, as yet unknown, Uranus and Neptune 

 acquired abnormal rotations from the start (and the 

 retrograde sateUites of Jupiter and Saturn suggest that 

 there was something peculiar about the local motions 

 at the outer end of the original filament), and either by 

 tidal action, or by forcing the medium near them to 

 rotate with them, caused the orbits of their satellites 

 to turn round tiO the satellites revolved nearly in the 

 equatorial planes of the primaries. 



The moon may have been a primitive nucleus like 

 the retrograde sateUites of Jupiter and Saturn, but it 

 is not very likely, for several reasons. In the first 

 place, it is certain that the moon has been receding 

 from the earth for ages, and when the oldest known 

 sedimentary rocks were laid down it must have been 

 very much nearer than it is now. This follows from 

 the facts known about tides in shallow seas, and is 

 confirmed by two results of observation. If from the 

 data we have about the moon we calculate its position 

 at some past time, some thousands of years ago, it is 

 found that its observed position then was somewhat 

 further west, and the difference is nearly such as would 

 be expected from the theory of the tides. Again, 

 certain facts about the shape of the moon indicate 

 that it did not become solid when at its present dis- 

 tance, but at about a third of it. It is unhkely that 

 two separate nuclei were formed at such a short dis- 

 tance apart, so that the moon was probably initially 

 part of the earth. 



If the primitive earth-nucleus was broken up by 

 tidal action in passing near the sun, the change would 

 not have taken place in the same way as was suggested 

 for the great planets. The earth was much more dense 

 than these, and was probably largely liquid from the 

 Start. Now, a liquid or solid body broken up by tidal 

 action would not eject a long filament and form a series 

 of sateUites : it would break into two pieces, one large 

 and one smaU, very much like the earth and the moon. 

 I do not, however, think that the moon was formed in 

 this way. If a body was once broken up like this, 

 there is no reason why it should not happen again and 

 again, forming in succession four, eight, sixteen, and 

 more pieces, until the bits were so smaU that they 

 cooled and sohdified, when their cohesion might after- 

 wards hold them together. The solar system actually 



seems to contain two examples of this. The rings of 

 Saturn are composed of many tiny bodies, and are 

 generaUy believed to be the result of the destruction 

 of a sateUite in this way. .\lso the asteroids may have 

 been formed from a primitive small planet, broken up 

 by approach to the sun, or more probably to Jupiter, 

 in view of the large eccentricities of some of their orbits, 

 which could not have survived if they had been very 

 old. 



It seems more likely that the moon was formed after 

 the earth's orbit had become much more nearly circular. 

 Consider for a moment a man in a swing. If he is 

 puUed back for a moment and released, he performs 

 a few oscUlations and graduaUy comes to rest. But 

 if we give him a push every time he is nearest to us, 

 the osciUations get bigger every time. This increase 

 in the extent of a vibration when the force from outside 

 has a period equal to the natural period is known as 

 " resonance." Now, if the liquid earth was distorted 

 so that its equator became an eUipse, it would osciUate 

 backwards and forvvards about the circular form tiU it 

 came to rest, and the period of this motion would be 

 about two hours. But we can show that, if ever the 

 earth and moon formed a single body, this must have 

 rotated in about four hours, and hence the solar tide at 

 any place on it would give the requisite push or pull 

 every two hours. The osciUation would therefore 

 grow, and it is hkely that it became so huge that the 

 mass broke into two pieces. 



The sateUites of Mars offer a difficulty on account of 

 the smallness of their primary. For Mars must have 

 been almost wholly liquid at the start ; and if it was 

 broken up by tidal action, it should have produced 

 sateUites comparable in mass with itself, whereas they 

 are actually only a few miles in diameter. They may 

 be asteroids that have been attracted out of their 

 original paths and caused to revolve round Mars. 



The only other bodies unaccounted for by the 

 theory are comets and meteors. These are almost 

 certainly of more recent origin, and took no part in 

 the birth of the system. 



LITER.\TURE 

 J. H. Jeans, Problems of Cosmogony and Stellar Dynamics. 



(Cambridge University Press, 1919.) 

 Sir G. H. Darwin, The Tides and Kindred Phenomena in the 



Solar System. (John Murray, 191 1.) 

 H. Jeffreys, " The Early History of the Solar System," and 



" The Resonance Theory of the Origin of the Moon," in 



Monthly Nctices of the Royal Astronomical Society, 1917-18. 



Geography of Commerce and Industry. By R. S. 



Bridge, M.A. (Hodder & Stoughton, 4s. 6d. 



net.) 

 Volinneiric Analysis. 2nd Edition. By J. B. Coppock, 



B.Sc, F.I.C. (Pitman, ^s- 6d. net.) 



