which may influence Geological Phcrnomcna . 297 



tion are detailed in the article of Laplace's work cited. But before entering 

 on a work of so much labour, it is quite necessary to inquire what prospect of 

 advantage there is to induce any one to undertake it. 



Now it certainly at first sight seems clear, that a variation of 3 per cent, 

 only, in the mean annual amount of solar radiation, and that arising from an 

 extreme supposition, does not hold out such a prospect. Yet it might be 

 argued, that the effect of the sun's heat is to maintain the temperature of the 

 earth's surface at its actual mean height, not above the zero of Fahrenheit's 

 or any other thermometer, but above the temperature of the celestial spaces 

 out of the reach of the sun's influence ; and what that temperature is, may be 

 a matter of much discussion. M. Fourrier has considered it as demonstrated 

 that it is not greatly inferior to that of the polar regions of our own globe; 

 but the grounds of this decision appear to me open to considerable objection*. 

 If those regions be really void of matter, their temperature can only arise, ac- 

 cording to M. Fourrier's own view of the subject, from the radiation of the 

 stars. It ought therefore to be as much inferior to that due to solar radiation, 

 as the light of a starlight night is to that of the brightest noon day ; in other 

 words, it should be very nearly a total privation of heatf , — almost the absolute 

 zero, respecting which so much difference of opinion exists ; some placing it 

 at 1000, some at 5000 degrees of Fahrenheit below the freezing point, and 

 some still lower, in which case a single unit per cent, in the mean annual 

 amount of radiation would suffice to produce a change of climate fully com- 

 mensurate to the demands of geologists. 



Without attempting, however, to enter further into the perplexing difficulties 

 in which this point is involved, which are far greater than appear on a cursory 

 view, let us next consider, not the mean but the extreme effects which a 

 variation in the excentricity of the earth's orbit may be expected to produce 

 in the summer and winter climates in particular regions of its surface, and 

 under the influence of circumstances favouring a difference of effect. And 

 here, if I mistake not, it will appear, that an amount of variation, which 

 we need not hesitate to admit (at least provisionally) as a possible one, may 

 be productive of considerable diversity of climate, and may operate during 



* Mem. de I' Acad. Royale des Sciences, 1827, torn. vii. p. 698. 



t The proportion of the light of the sun to that of the moon has been estimated by Bouguer 

 as 300,000 to 1. If we regard the illumination of full moonlight as only 100 times greater than 

 that of a bright starlight night, which is a very moderate supposition, we shall have a ratio of 

 30,000,000 to 1 for the illuminating power of the sun compared to that of all the stars in our 

 hemisphere, and consequently 15,000,000 to 1 for the ratio of the heating effect of the sun to 

 that of all the stars in both hemispheres. 



VOL. III. SECOND SERIES. 2 Q 



