PREMONITIONS OF ISOSTASY 177 



such reasoning more than that conditions are quite compatible with the 

 hypothesis of increase of homogeneity with deptli. 



Thus it follows from Stokes's theorem that the observed deflections of 

 tlie plumb-line may be due entirely to the heterogeneity of the earth's 

 outer sbell; but it does not follow tbat only the outer layer is heteroge- 

 neous. If the center of the batholith I have imagined were at 8I/2 miles 

 beneath the surface instead of at 414 miles, its mass remaining un- 

 changed, it would cause a deflection of only a fourth of 23" or, say, 6". 

 Consequently deflections alone give no direct information as to the dis- 

 tribution of density in depth. 



If indeed it could be assumed that the primitive earth was fluid 

 throughout, or to a great depth from its temporary surface, such irregu- 

 larities as are observed at the present surface could scarcely be supposed 

 to exist far below it. To be sure, there is seemingly no limit to thinkable 

 viscosity; an earth can be imagined so viscous that density would not 

 efi'ectively control configuration. But there is enough evidence of gravi- 

 tative differentiation of rocks to show that many magmas yield at a sensi- 

 ble rate to the stresses produced, even by very small phenocrystic crystals, 

 and therefore with much greater velocity to bodies of batholithic dimen- 

 sions. The respective velocities for highly viscous magmas are doubtless 

 proportional to the cross-sections. 



Laplace, Pratt, and Kelvin were all convinced that the earth had been 

 fluid originally, while Stokes considered the evidence very strong, though 

 not conclusive. Recent investigations in isostasy strengthen this evi- 

 dence and seem to confirm Laplace's view as to the superficiality of the 

 heterogeneous shell. ^* 



So far as I can see, the investigations which have been passed in review 

 lead to no definitive conclusion as to tbe condition of tlie interior of the 

 earth, although they leave no question that the material of the globe is 

 arranged nearly as if it had been fluid. On the other hand, they estab- 

 lish a presumption that irregularities in density, such as are encountered 



"Thomson and Talt (Nat. Phil., 1883, section 821) summarize tlie evidence then avail- 

 able as follows : 



"There is, as we shall see in later volumes, a great variety of convincing evidence In 

 support of the common geological hypothesis that the upper crust was at one time all 

 melted by heat. This would account for the general agreement of the boundary of the 

 solid with that of fluid equilibrium, though largely disturbed by uitheaval and shrinkiugs 

 In the process of solidiru'atioii, which has probably been going on for a few miUiou years, 

 but is not quite complete (witness lava llowing from still active volcanoes). Tbe oblate- 

 ness of (lie deeper layers of e(iual density vvliicli we now infer from the ligure of sealevel, 

 the observed density of the upper crust, and Cavendish's weighing of the earth as a 

 whole, renders it highly probable that the earth has been at one time melted not merely 

 all round its surface, l)ul either throughout or to a great deptli all round." 



It is probable (hat the liquid portion of the earth was approximately In couvectlve 

 equilibrium, and that consolidation began at the centei*. 



