ON THE PHYSICAL STlirCTlIRE OF THE EARTH. 215 



CoDsequeiitly tbe processional motion of such a slicll would bo the same 

 as that of a homegeneous si>lioroid of the same ellii)ticity. If e = tj^-, 

 it appears that the value of precession for such a spheroid would bo 

 57", while its observed value is 50"-l.* Now, as it is impossible to 

 admit such a dilierence where the result of observatiou is so well estab- 

 lished, we must conclude that the solid shell of the Earth, composed of 

 nearly equi elliptic strata, can not extend to its center —in other words, 

 that the Earth can not be altogether n solid from its surface to its center. 

 On the other hand, the Jluid nucleus contained within the shell can not 

 be devoid of friction and viscidity, but must possess these proi)orties 

 in common with all fluids that have ever beeu observed on the Earth's 

 surface. These properties of the liquid may, as I have long since an- 

 nounced, cause the shell and liquid uu(;leus to rotate together as one 

 solid mass. The same conclusion was afterward put forward by M. De- 

 launey; and experiments made under his direction, and afterward, 

 at the instance of the JJoyal Irish Academy, by me, show that in rota- 

 ting glass vessels tilled with water the amount of friction and viscidity 

 is such as to render any difference of slow motion between the liquid 

 and its containing vessel insensible. With liquids so viscid that water 

 is in comparison limpid, such as pitch, honey, and especially volcanic 

 lava in a fused state, the results wonld be absolutely decisive. To 

 this class of li(puds the fluid matter of the Earth's interior, so far. as 

 It has come under observation, undoubtedly belongs; and hence the 

 •overwhelming certainty of our general conclusions as to the connection 

 between the Earth's structure and its rotation. 



(8) If the tendency of the solid crust is to become more elliiitical at 

 its inner surface as it increases in thickness, some interesting conse- 

 quences appear to follow. If the shell were unaccompanied by the 

 nucleus, or if no friction existed at their surfiices, the changes in the 

 relations of the principal moments of inertia of the shell might be sup- 

 posed to cause its rotation to become unvStable, so as to bring about 

 •conditions which might result in a change of the axis of rotation. It 

 is easy to show on the most favorable suppositions that this could not 

 'Occur. The increasing ellipticity of the inner surface of the shell would 

 be due to the increasing oblateness of the surface of the fluid nucleus, 

 and this would be at its maximum if the noucleus approached a state 

 •of homogeneity; but the fluid can not approach this state unless the 

 radius of the nucleus is so small that the variation in density due to 

 pressure ])ecoraos insensible, whence all its strata would i)ossoss the 

 same density. This condition with a certain thickness of the solid shell 



* A revision of tlie nnuierical daia from recent astronomical resnlts leads me to con- 

 clude that the precetsion for the solid spheroid would he a litile less, and about 55'/ 

 instead of 57". This I propose to jirove in a short i)apcr, entitled " Note on the an- 

 nual precession calculated on the hypothesis of the earth's solidity." This note 

 [appended to this article] leaves the general conclusions of the present paper un- 

 ^iltered. 



