OCTOBEK 30, 1896.] 



SCIENCE. 



633 



this is improbable because of the slow rota- 

 tion of the sun, but Perry has given reasons 

 for an opposite conclusion. 



The collapse of the first argument of tidal 

 retardation, and of the second of the cool- 

 ing of the earth, warn us to beware of a con- 

 clusion founded on the assumption that the 

 sun's energy depends, and has ever de- 

 pended, on a single source of which we 

 know the beginning and the end. It may 

 be safely maintained that such a conclusion 

 has not that degree of certainty which justi- 

 fies the followers of one science in assuming 

 that the conclusion of other sciences must 

 be wrong, and in disregarding the evidence 

 brought forward by workers in other lines 

 of research. 



We must freely admit that this third 

 argument has not yet fully shared the fate 

 of the two other lines of reasoning. Indeed, . 

 Prof. George Darwin, although attacking 

 these latter, agrees with Lord Kelvin in re- 

 garding 500 million years as the maximum 

 life of the sun.* 



We may observe, too, that 500 million 

 years is by no means to be despised ; a great 

 deal may happen in such a period of time. 

 Although I should be very sorry to say that 

 it is sufiicient, it is a very difierent offer 

 from Prof. Tait's 10 million. 



In drawing up this account of the physi- 

 cal arguments, I owe almost everything to 

 Prof. Perry for his articles in Nature (Jan- 

 uary 3 and April 18, 1895), and his kind- 

 ness in explaining any difficulties that arose. 

 I have thought it right to enter into these 

 arguments in some detail and to consume 

 a considerable proportion of our time in 

 their discussion. This was imperatively 

 necessary, because they claimed to stand as 

 barriers across our path, and, so long as 

 they were admitted to be impassable, any 

 farther progress was out of the question. 

 What I hope has been an unbiassed examin- 

 ation has shown that, as barriers, they are 



* British Association Reports 1886, pp. 514-518. 



more imposing than effective ; and we are 

 free to proceed and to look for the conclu- 

 sions warranted by our own evidence. In 

 this matter we are at one with the geolo- 

 gists ; for, as has been already pointed out, 

 we rely on them for an estimate of the time 

 occupied by the deposition of the stratified 

 rocks, while they rely on us for a conclusion 

 as to how far this period is sufiicient for the 

 whole of organic evolution. 



First, then, we must briefly consider the 

 geological argument, and I cannot do better 

 than take the case as put by Sir Archibald 

 Geikie in his Presidential Address to this 

 Association in Edinburgh in 1892. 



Arguing from the amount of material re- 

 moved from the land by denuding agencies, 

 and carried down to the sea by rivers, he 

 showed that the time required to reduce the 

 height of the land by one foot varies, ac- 

 cording to the activity of the agencies at 

 work, from 730 years to 6,800 years. But 

 this also supplies a measure of the rate of 

 deposition of rock ; for the same material is 

 laid down elsewhere, and would, of course, 

 add the same height of one foot to some 

 other area equal in size to that from which 

 it was removed. 



The next datum to be obtained is the 

 total thickness of the stratified rocks 

 from the Cambrian system to the present 

 day. '' On a reasonable computation these 

 stratified masses, where most fully devel- 

 oped, attain a united thickness of not less 

 than 100,000 feet. If they were all laid 

 down at the most rapid recorded rate of de- 

 nudation they would require a period of 

 73 millions of years for their completion. 

 If they were laid down at the slowest rate 

 they would demand a period of not less 

 than 680 millions." 



The argument that geological agencies 

 acted much more vigorously in past times 

 he entirely refuted by pointing to the char- 

 acter of the deposits of which the stratified 

 series is composed. " We can see no proof 



