]\gt ; The tional Sbility of the Earth .
By A. E. H. LoYE , , Sedleian Professor of Natural Philosophy in the University of Oxford .
( Received February 16 , \mdash ; Read March 14 , 1907 .
) ( Abstract .
) The problem of determining the conditions , as regards elasticity , in which a body of planetary dimensions can be gravitationally unstable , was attacked by J. H. Jeans .
* To turn the difficulty arising from the fact that such a body is necessarily in a state of " " initial stress by which gravitation is balanced throughout its interior , he introduced an artificial external field of force to balance gravitation in the undisturbed body , and he thus dispensed with initial stress .
A more direct method of treating the problem has been advocated by Lord Rayleigh .
In this method the stress at any point of the body , when disturbed from its equilibrium state , is taken to be compounded of two stress-systems ; an initial stress of the nature of hydrostatic pressure , balancing gravitation in the undisturbed body , and an additional stress connected with the strain ( reckoned from the initial state ) by the same formula as connect stress with strain in a homogeneous isotropic elastic solid body which is slightly strained from an initial state of zero stress .
In the first part of the present paper the equations of vibratory motion of the planetary body , assumed to be initially spherical , are formed in accordance with this method , and they are solved by means of spherical harmonics .
The frequency equation is obtained , and the conditions that it may be satisfied by a zero value of the frequency are found .
When these conditions are fulfilled the body is in a state of gravitational instability .
Let denote the radius of the sphere , the density of the body in the initial state , its rigidity , its modulus of compression , the constant of ation .
When the body is disturbed let denote the displacement of any point , and the dilatation , so that ; also let denote the additional gravitation potential due to defect , or excess , of density , and to displacement of mass across the initial bounding surface .
Then satisfies the equation *London , ' Phil. Trans Ser. , vol. 201 , 1903 , p. 157 .
London , ' Roy .
Soc. Proc Ser. , vol. 77 , 1906 , p. 486 .
The Gravitational bility of the and , if the origin of co-ordinates is at the centre of the undisturbed sphere , the equations of vibratory motion are three equations of the } ) The quantities must satisfy these equations , and they must also satisfy certain boundary conditions .
One of these conditions is that the dilced bounding surface is the surface separating the region of space which contains the matter of the body from the external region .
This condition is expressed as an equation which holds at the bounding surface , by the potential as due to a volume distribution of density within this surface and a superficial distribution upon it .
The remaining boundary conditions express the vanishing of the surface tractions at the dispIaced bounding surface .
These conditions also are replaced by equations which hold at the initial bounding surface , and these latter equations can be interpreted in the statement that the traction on the initial surface is a pressure , equal to the weight ( per umit of area ) of the material heaped up to form the radial inequality .
When the body is vibrating in a normal mode , the dilatation at any point is proportional to the product of a certain function of the distance .
of the point from the centre and a spherical surface harmonic of degree The displacement at any point is expressible in terms of this harmonic and its diff'ereutial coefficients , by the aid of certain metions of which are determined , and the radial component of displacement at any point is proportional to the product of a function of !
and the same harmonic .
The harmonic inequality of the initial bounding surface follows the same law as the distribution of excess density at points of any spherical surface concentric with the boundary .
The condition that the frequency equation may be satisfied by a zero value of the frequency is an equation to determine the number , which is denoted by .
The equation contains the number , which is denoted by ; it also contains the integer , which is the of the spherical harmonic specifying the mode of vibration .
The homogeneous spherical planet is gravitationally unstable as regards displacements of the type specified by harmonics of degree if the equation in question has a real positive root .
It is proved that the equation has one , and only o1le , real positive root in each of the cases and ; it is proved further that it has no real root in either of the cases or .
The critical value of in the case ( radial displacements ) lies between 3 and 4 if and between 4 and if is very small .
The critical value in the case Prof. A. E. H. Love .
[ Feb. 16 , lies between and if , and between and 8 if is very small .
The following table shows the corresponding values of the modulus of compression in dynes per square centimetre , in a body of the same size and mass as the earth:\mdash ; very s The modulus of compression of ordinary steel is about , that of glass , that of mercury The elastic constants of the earth , in its present state , can be estimated from the observed velocities of propagation of the three types of waves which are transmitted when a great earthquake takes place .
There are two sets of " " preliminary tremors\ldquo ; propagated directly through the earth with nearly constant velocities of about 10 kilometres per second and 5 kilometres per second , and a main shock\ldquo ; propagated over the surface with a velocity of about 3 kilometres per second .
The two sets of tremors have been identified with waves of dilatation and distortion , and the " " main shock\ldquo ; with superficial waves of the type first investigated by Lord In the present paper reason is given for thinking that the manner of propagation is not much affected by gravitation and initial stress , and thus the observed values of the velocities of propagation of earthquake tremors and shocks would yield ( i ) for the seismic effective modulus of compression of the earth as a whole the value dynes per square centimetre ; ( ii ) for the seismic effective idity of the earth as a whole the value dynes per square centimetre ; ( iii ) for the seismic effective rigidity of surface rocks a value approximately equal to 6 dynes per square centimetre ; ( iv ) for the seismic effective value of the number : .
The corresponding value of is .
As this is well below any critical value , it is concluded that the present state of the earth is far removed from one of gravitational instability .
In the second part of the paper an attempt is made to trace some of the consequences of the hypothesis that the elasticity of the earth was once * London , 'Proc .
Math. Soc vol. 17 , 1885 , p. 4 , or ' Scientific Papers , ' vol. 2 , p. 44 ] .
1907 .
] The Gravitational Stability of the Earth .
too small to secure stability in a homogeneous spherical configuration .
J. H. Jeans , in the paper already cited , concluded that , if the earth at one time had so small a modulus of compression that a spherically symmetrical state of aggregation would have been unstable , it would have tended to take up a state in which there would have been excess of density in one hemisphere and defect in the antipodal hemisphere .
To a first approximation the surface would have been spherical , but the centre of gravity would not have coincided with the centre of figure .
Jeans found that , when a second approximation is made , the figure would have been slightly\ldquo ; pear-shaped , \ldquo ; a blunt end , a sharper end and a waist .
In the same year , 1903 , W. J. Sollas*published a paper in which , from an examination of the distribution of land and watel , he also concluded that the figure of the earth resembles \ldquo ; pear The two pears have , however , little in common beyond the name ; and Sollas ' description of the figure points unmistakably to a mathematical description in terms of spherical harmonics of the third degree , while Jeans ' theory is based on the occurrenoe of harmonics of the first and second degrees .
If a homogeneous spherical planet were unstable at all , it would be unstable as regards radial displaoeulents .
It would , therefore , tend to pass into a state in which the density near the surface is less than the mean density .
This is the actual condition of the earth , and it is at once accounted for by the hypothesis that the modulus of compression was once too small for a homogeneous state to be stable .
The result , obtained in the first part of the paper , that the critical value of in the case is not very high in comparison with that which was found in the case , points , to the conclusion that , if the mean modulus of compression of a planet , with a spherically symmetlical distribution of density , were sufficiently small , the planet , though stable for radial disturbances , might be unstable for disturbances specified by harmonics of the first degree .
If the planet were at rest it would tend to take up such a form and state of ation as that described by Jeans .
There would be excess of density in one hemisphere , and defect in the antipodal hemisphere , the excess or defect at any poiut being proportional , at any distance from the centre , to a spherical surface harmonic of the first degree , that is to say , to the distance of the point from the bounding plane of the hemispheres .
The centre of gravity would be displaced relatively to the centre of figure towards that side where the density is in excess .
The equipotential surfaces which lie outside the boundary of the figure and sufficiently near to it on the side where the density is in excess would cut it on the side where the density is in defect .
If this were the case with the earth , the lithosphere ( or nucleus ) would cut the geoid ( the equipotential Quart .
J. vol. 69 , 1903 , .
180 .
Prof A. E. H. Love .
[ Feb. 16 , surface with which surface of the ocean coincides ) so as to protrude from it towards one side .
The land of the globe would consist of a single circular island which might occupy any fraction of the area of the surface .
If a second approximation were made in the manner indicated by Jeans , a comparatively small island would be introduced at the antipodes of the central portion of the circular island indicated by the first approximation .
If the earth were at rest , and the distribution of land and water were due to a past state of gravitational instability , the map would have the form just described .
It is necessary to consider the modification of these results on account of the of the earth and the attraction of the moon .
attraction was , at one time , a nearly constant force , for the period of revolution of the moon was once nearly equal to the period of rotation of the earth .
The figure of the earth would then have been more nearly an ellipsoid with three unequal axes than an oblate spheroid ; and the inequalities of the attraction at different points of a spherical surface , having its centre at the centre of gravity , would have been expressible by means of a disturbing potential the terms of which would have been spherical surface harmonics of the second degree .
In the present paper an examination is given of the effect of rotation upon a planet of which the modulus of compression is so small that , in the absence of rotation , a spherically symmetrical state of ation would be unstable .
It appears that the type of disturbance would not be specified exclusively by harmonics of the first degree , but certain harmonics of the third degree would be introduced .
When forces of the type of the attraction of the moon , involving a disturb potential expressed by means of harmonics of the second degree , are taken into account , it is found that all the harmonics of the third degree would be introduced .
Both the lithosphere and the geoid would be nearly spherical , more nearly oblate spheroids , still more nearly ellipsoids each with three unequal axes ; and finally there would be a wrinkle on the ellipsoidal surface of the lithosphere expressible by meaus of harmonics of the third .
There would also be excess of density on one side of a diametral plane and defect on the other side .
The surface of the lithosphere , as determined by difference of level above or below the geoid , would be expressed by a sum of harmonics of the first , second , and third degrees .
If ever the earth as a whole had so small a modulus of compression that a spherically symmetrical state of aggregation would have been unstable , the lithosphere would have consolidated in such a shape , and with such a situation relative to the geoid , as has been described ; and we should expect to find traces of this shape and situation in the general features of the distribution of land and water , and still more in those features of the continental blocks and oceanic regions which are prominent at considerable depths below sea-level .
1907 .
] The bility of the In the paper a spherical harmonic analysis , as far as harmonics of the third degree , is made of a distribution in which the value is assigned to those points which are above sea-level , the value to those points which are below mean-sphere-level , and the value to intermediate points .
The coefficients of the various harmonics are determined by a method of approximate integration , and the following spherical surface harmonic is thus found:\mdash ; sin sin ; the quantities denoted by are co-latitude measured from the North Pole and longitude east of Greenwich .
A number of contour lines of the harmonic are drawn ; and the results seem to show a sufficient agreement with the raphical facts to lend considerable support to the hypothesis which led to the conclusion that harmonics of the first , second , and third degrees should be predominant .
From the fact that there are dynamical reasons why the lithosphere should have consolidated in a certain type of shape , and that its actual shape is approximately of this type , it appears probable that the shape has not changed greatly , but that many of the changes which have taken place have been of the nature of adjustments of the figure to changing dynamical conditions .
This conclusion carries with it important consequences as to the eneral trend of events in history .

