8 Sir William Thomson [Jan. 21, 



density of the sun, being about 1 • 4 times the density of water, or 

 about a quarter of the earth's mean density. In reality the density 

 in the upper parts of the sun's mass must be something less than 

 this, and something considerably more than this in the central parts, 

 because of the pressure in the interior increasing to something 

 enormously great at the centre. If we knew the distribution of 

 interior density we could easily modify our calculations accord- 

 ingly ; but it does not seem probable that the correction could, 

 with any probable assumption as to the greatness of the density 

 throughout a considerable proportion of the sun's interior, add more 

 than a few million years to the past of solar heat, and what could be 

 added to the past must be taken from the future. 



In our calculations we have taken Pouillet's number for the total 

 activity of solar radiation, which practically agrees with Herschel's. 

 Forbes * showed the necessity for correcting tlie mode of alloAving for 

 atmospheric absorption used by his two predecessors in estimating 

 the total amount of solar radiation, and he was thus led to a number 

 1 • 6 times theirs. Forty years later Langley,t in an excellently 

 worked out consideration of the whole question of absorption by our 

 atmosphere, of radiant heat of all wave-lengths, accepts and confirms 

 Forbes's reasoning, and by fresh ol)servations in very favourable 

 circumstances on Mount Whitney, 15,000 feet above the sea-level, 

 finds a number a little greater still than Forbes (1*7, instead of 

 Forbes' 1 • 6, times Pouillet's number). Thus Laugley's measurement 

 of solar radiation corresponds to 133,000 horse-power per square 

 metre, instead of the 78,000 horse-power which we have taken, and 

 diminishes each of our times in the ratio of 1 to 1 • 7. Thus, instead of 

 Helmholtz's twenty million years, which was founded on Pouillet's 

 estimate, we have only twelve millions, and similarly with all our 

 other time reckonings based on Pouillet's results. In the circum- 

 stances, and taking fully into account all possibilities of greater 

 density in the sun's interior, and of greater or less activity of 

 radiation in past ages, it would, I think, be exceedingly rash to 

 assume as probable anything more than twenty million years of the 

 sun's light in the past history of the earth, or to reckon on more 

 than five or six million years of sunlight for time to come. 



We have seen that the sun draws on no external source for the 

 heat he radiates out from year to year, and that the whole energy of 

 this heat is due to the mutual attraction between his parts acting in 

 conformity with the Newtonian law of gravitation. We have seen 

 how an ideal mechanism, easily imagined and understood, though 

 infinitely far from possibility of realisation, could direct the work 

 done by mutual gravitation between all the parts of the shrinking 

 mass, to actually generate its heat-equivalent in an ocean of white- 

 hot liquid covering the sun's surface, and so keep it white-hot while 



* 'Edin. New Phil. Journal,' vol. xxxvi. 1844. 



t 'American Journal of Science,' vol. xxvi. March, 1883. 



