303-306] The Tidal Theory 285 



been mainly gaseous from their birth, we might naturally expect them to be 

 surrounded by systems of satellites moving approximately in the planes of 

 the orbits of their primaries. The genesis of the earth-moon system remains 

 indefinite, although not inexplicable, but the satellites of Mars present 

 something of a puzzle on the tidal theory, as indeed on any other theory of 

 cosmogony. 



305. This vague sketch of the tidal theory will, it is hoped, be read as 

 an indication of the possibilities open to the tidal theory, rather than as an 

 attempt to advocate the theory or to present it in a final form. The theory 

 is beset with difficulties and in some respects appears to be definitely un- 

 satisfactory. To the author it appears more acceptable than the rotational 

 theory, or any other theory so far offered of the genesis of the solar system ; 

 but an enormous amount of mathematical research appears to be needed 

 before the theory can either be advocated with confidence or finally abandoned. 



THE TIME-SCALE 



306. Throughout our discussion of the various processes of astronomical 

 evolution, the question of the time occupied by these processes has, inten- 

 tionally, been kept in the background. 



There are three main sources of information as to the time-scale on which 

 the duration of these processes must be measured: 



I. Lord Kelvin, in a well-known calculation, shewed that the gravita- 

 tional energy produced by the sun falling to its present configuration from a 

 state of infinite rarity would only suffice to maintain the sun's present rate of 

 radiation for about 20 million years. This estimate has been challenged on 

 the grounds that the sun may have other sources of energy, radioactive energy 

 in particular. 



II. Geological investigations have given various estimates of the time 

 which has elapsed since our earth assumed its present solid form. These are 

 not altogether consistent with one another, and a number of early estimates 

 must be definitely rejected in view of our recently gained knowledge of 

 radioactive processes, but after excluding doubtful estimates there remains a 

 number of apparently fairly reliable estimates which seem to converge to a 

 period of the order of 250 million years for the age of our earth. 



III. Some material for estimating the time-scale is provided by the 

 present motions of the stars. The time of transition of a star across our 

 galactic universe and back is of the order of 320,000,000 years, and the period 

 of motion round the milky way, or of rotation of the milky way, is of the 

 same order of magnitude. The stars shew such uniformity of velocity and 

 arrangement in different parts of space that we may reasonably suppose that 



