82 PRESIDENTS ADDRESS—-SECTION C. 
cooling will sink slowly down below the surface. The shell of 
greatest cooling will be the shell of greatest circumferential 
contraction, and the amount of circumferential contraction will 
diminish inwards and vanish at the level where no cooling is 
taking place. The circumferential contraction will also diminish 
outwards towards the surface, where also there will be no 
circumferential contraction, owing to the mean surface tempera- 
ture being kept constant by radiation from the sun. But radial 
contraction would also be going on, and this would vary along 
any one radius, being nothing at the level of no cooling and 
greatest at the surface, where it is the sum of the whole radial 
contraction. Now, as the radius is less than the circumference, 
the mean rate of radial contraction is less than the mean rate of 
circumferential contraction ; consequently, the shell of greatest 
cooling, where circumferential contraction is at its maximum, 
must be in a state of tension, while the surface, where radial 
contraction is at its maximum, must be in a state of compression. 
The crust will be formed by an outer shell of compression resting 
on an inner shell of tension, the one passing gradually into the 
other. The level where the one passes into the other will be 
that Jevel where the radial and circumferential contractions are 
equal. This is the level of no strain, outside of which the 
compression gradually increases until it reaches its maximum at 
the surface. 
The subject has been taken up by Mr. C. Davison, Professor G. 
Darwin, and the Rev. O. Fisher. On the supposition that the 
rate of cooling varies as the square root of the time that has 
elapsed since the consolidation of the globe, Mr. Davison 
calculates that if the crust solidified at a temperature of about 
7000° F., about 174,240 millions of years ago, the cooling would 
have penetrated 400 miles into the interior, the level of greatest 
cooling would be 72 miles, and the level of no strain five miles 
below the surface. The supposed data are, however, far too 
great, and Professor Darwin, taking the more reasonable assump- 
tion that solidification took place 100 millions of years ago, 
calculates the level of no strain at two miles deep. He also finds 
that, in ten millions of years, 28} miles of rock on a great circle 
would be crushed up—that is, rather more than one mile ina 
thousand, Also, that 228,000 square miles would be piled up 
on the top of the subjacent rocks—that is, a cone with a base 
of 228,000 square miles and a height of two or, possibly, three 
miles would be crushed. This would make, in ten millions of 
years, a mountain chain about a half or a third the size of the 
Himalaya, Evidently the results are much too small to account 
for mountain building during the Cainozoic era. 
Mr. Fisher, on the supposition that the present internal tem- 
perature gradient is 1° F. in 51 feet, finds that, if the temperature 
of soliditication was 7000° F., the level of no strain would be 
