﻿416 Rev. J. J. Blake on 



caps ; i.e., would prove that his hypothesis is wrong, when he 

 assumes that gravitation is constant in spite of the different 

 densities of the oceans and continents. 



A second paper hy the same author *, in reply to another 

 criticism of his conclusions as to the existence of a level of 

 no strain at a small depth below the earth's surface, calls for 

 a few observations. 



It is certainly a gain to have elicited that, by Mr. Fisher 

 at least, " no claim is made that such a change in condition 

 exists at that depth in the actual earth." I thought that a 

 level of no strain had been " discovered," and that its actual 

 position in the earth was " proved " to be at so slight a depth 

 as to show that the surface-features had been produced by 

 some other cause than contraction, notwithstanding its 

 presence. 



Somehow or other, however, my two objections have got 

 mixed. The a priori objection is that in any cooling globe 

 such a level in such a position is extremely unlikely ; and the 

 a posteriori objection is that " in that case the surface-features 

 would not have resembled .... those that we see," which is 

 thus admitted by Mr. Fisher. That this objection does not 

 simply involve a petitio principii may be seen from the fact 

 that the compression at the poles is neglected in the investi- 

 gation, and this would make a difference two or three times 

 as great as the supposed depth of the level of no strain. 



But, as Mr. Fisher says, the primary objection is to the 

 method, and I am much obliged to him, and I think others will 

 be also, for calling attention to the investigation of the ques- 

 tion by Prof. R. S. Woodward, which is an admirable piece 

 of work, though I do not see that it helps Mr. Fisher much. 

 Prof. Woodward distinguishes between " free cooling,'" which 

 is considered by Lord Kelvin and Mr. Fisher, and "con- 

 ditioned cooling," which was dealt with by Poisson, Fourier, 

 and Riemann, and which is the actual state of affairs with 

 regard to the earth ; and he gives this curious reason for 

 assuming that the former may be adopted instead of the 

 latter, that " defects in the best data at present attainable, 

 and probable inaccuracies in our fundamental assumption, 

 render the results uncertain even in the first place of decimals." 

 In other words, when applied to the actual earth, both suppo- 

 sitions must give such loose approximations that it matters 

 little which is taken. How, then, can we hope to get results 

 of any value when we deal with such small quantities as a 

 mile or two in a spheroidal earth ? 



But, more than this, it seems to me that as soon as 

 * Phil. Mag. July 1894. 



