]\gt ; [ 85 ] 11 .
On Disturbance of Level Arising om the Load of Neighbouring Oceanic K. TERAZAWA , Ri-Gakushi of the University of Tokyo .
Communicated by Sir JOSEPH LARMOR , F.R. Received April 17 , \mdash ; Read June 1 , 1916 .
I. Introductory .
IN his important observations on the lunar deflection of gravity , Dr. O. HECKER has pointed out that the force acting on the pendulum at Potsdam is a larger fraction of the moon 's force when it acts towards the east or west than when it acts towards the north or south .
Various explanations of this anomaly have been proposed ; among them one , suggested by Prof. A. .
H. LOVE , that a possible cause may perhaps be found in the attraction of the tide wave in the North Atlantic .
Recently Prof. A. A. MIOHELSON has found a similar result in his arduous task , cessfully achieved , of obtaining a continuous record of the lunar perturbation of a very long water-level at Chicago .
Prof. Sir J. kindly suggested to me a query , whether the excess-pressure of the tide in the North Atlantic would affect much the measurement of water-level at Chicago , owing to the elastic depression of the earth 's surface that it would produce .
In the present paper the calculation is undertaken in order to ascertain to what extent the consideration of tilting of the ground is important for the of this geodynamical discrepancy , on the assumption that the earth consists of elastic matter of uniform density and of invariable elastic constants , and that the North Atlantic may be represented roughly as a circular basin and that the tide in it is uniform or else elliptic .
The curvature of the earth is neglected , which is , of course , admissible for a first estimatio1l .
The solution of the equation of equilibrium concerning an elastic body with plane boundary , otherwise extending to infinity , subject to certain boundary conditions , * 'Some Problems of Geodyuamics , ' Cambridge , p. 88 .
'Th .
Astrophysical Journal , ' vol. 39 ( 1914 ) , p. 105 .
VOL. CCXVII .
July 12 , 1916 .
; MR. K. TERAZAWA ON PERIODIC DISTURBANCE OF LEVEL has been obtained by LAME and CLAPEYRON* and by J. ; the former by making use of FOURIER'S theorem , and the latter by introducing several kinds of potential function .
, The solution for the where the boundary condition is a normal pressure distributed symmetrically round a point on the surface and the body is free from bodily force is given by Prof. H. LAMB in a form of definite I have tried to solve the problem , independently of LAMB , and obtained the solution answering to any distribution of a normal pressure , from which the solution of LAMB can be derived as a special case .
S For the present purpose let us confine our attention to the simplest case in which the boundary is.subjected to a symmetrically distributed normal pressure .
Take the centre of the area , in undisturbed state , on which the given normal pressure is distributed , as the origin of the co-ordinates , the inward normal to the undisturbed surface as the -axis , and denote the distance of any point from the -axis by Then corresponding to the normal pressure , ( 1 ) given at the surface , we shall have the following expressions the displacement components:\mdash ; ( ) ( ) , ( 2 ) , ( 3 ) in which stands for ( ) , ( 4 ) being LAME'S elastic constants , , BESSEL'S functions of zeroth and first order , and the radius of loaded area .
Now , suppose the normal pressure ( 1 ) to arise entirely from the tidal loading in the North Atlantic basin , then the tidal loading would come into piay to affect the water-level measurement at Chicago in two accounts : the one is the direct attraction of the material l.oad , the other the deformation of the ground by the pressure *'CRELLE'S Journal , ' vol. 7 , p. 400 'Application des .
Potontiels Paris ( 1885 ) .
'Loud .
Math. Soc. Proc vol. 34 , p. 276 ( 1902 ) .
S The solution , with numerous examples , will be published FROM THE LOAD OF NEIGHBOURING OCEANIC TIDES .
by that load .
Dr. C. CHREE , afterwards more completely Prof H. find a formula , by using the formula obtained by BOUSSINESQ , to calculate the of the direction of gravity due to the attraction of a material load on the surface of the earth .
The same result can be attained , of course , from our solution .
The expression for the vertical displacement at a point on the surface oan be transformed into by making use.of NEUMANN'S addition theorem for BESSEL'S functions , where stands for .
On the other hand , if we denote the attraction constant by , and gravity , prior to application of the load , by , then the gravitation-potential at a point on the unloaded surface due to the loading can be expressed by provided the height of the loading material is negligibly small upared with the distanoe of the point under consideration from any point in the loaded area .
omparing the above two expressions , we have Thus the , direction of gravity becomes in consequence of the attraction of the loading material inclined to the vertical at the angle which will be determined by , ( 5 ) while its tilting effect is expressed by . . . . . . .
( 6 ) The total effect of the loading will thus be , . . . . .
( 7 ) in a close approximation .
*'Phil .
NIag V. , 43 , p. 177 ( 1897 ) .
, ' Sug .
Buts .
Kizi , ' , p. 208 ( 1912 ) .
III .
Loading .
Let be the radius of the loaded circle , the height of the material loading , which is supposed to be uniform , and its density , then we have to put for ( 8 ) On this supposition , we get ( ka ) , therefore ( ) , ( 9 ) .
( 10 ) The other components of the displacement and those of stress can be expressed similar forms .
But it is unnecessary to write them down here as they are out of our present purpose .
The integrals required here cannot be evaluated in a very simple way .
Some of them are closely connected to the magnetic potential due to a circular current , or to the velocity-potential and stream function of a circular vortex , and have been discussed by various authors .
In his paper on the inductance of circular Prof. NAGAOKA has devised a comparatively simple method which may be applied to evaluate all the integrals needed for the calculation of the displacement and stress in the present problems .
Let us follow his method and describe it here briefly .
Put , then by NEUMANN'S addition theorem for BESSEL'S function we have 'Phil .
Mag , 6 ( 190$ ) , p. 19 .
FROM THE LOAD OF HBOURING OCEANIC TIDES .
use of these formulae we obtain ( ka ) ( ka ) To find integrals , put , ( 11 ) ( 12 ) ' so that and change the integration variable from to by then Put again , then , since is real and lies between , and , and correspond to , and respectively , if we take to be positive* ; where and denote the real and imaginary half-period .
Thus . . . . . . . . . .
( 13 ) the evaluation of the other integrals , write .
( 14 ) to be negative , then and correspond to and will , as a matter of course , be obtained after integration .
then we shall have Now we have . . . . .
( 15 ) Hence , ( 16 ) .
( 17 ) The term enters because of the many-valued property of a logarithm .
actual value of and will be determined by the following consideration .
From the definition of and , it follows immediately that accordingly The last inequality shows that the value of must be one of the following:\mdash ; , ( 18 ) where and denote any integers , positive or negative , or zero , and number less than unity .
To determine the value of in the formulae 16 ) ARISING FROM THE LOAD OF NEIGHBOURING OCEANIC TIDES .
the value of given in of ( 18 ) , observe that the integral on the left-hand side of ( 15 ) and the function change their values continuously as varies from to 1 , while remains unchanged in this variation .
In the limit as , the value of the integral is nil and and therefore we have Similarly for the value of given in ( ii .
) of ( 18 ) , proceeding to the limit , we find .
value of will be obtained from In the present case we have It may easily be shown that the value of is a positive imaginary quantity for the value of given in of ( 18 ) and a negative imaginary quantity for ( ii .
) .
Therefore we have to take ( 19 ) for ; and , ( 20 ) for ; where Lastly , one more integral which can be evaluated without the knowledge of elliptic functions is for . . . . .
( 21 ) Substituting these values of integrals in ( 9 ) , we have in which the simplest value of representative .
At the surface it reduces simply to ( 22 ) in of ( 18 ) is taken as the , ( 23 ) where .
( 24 ) Similarly for the value of , we have .
( 25 ) For practical purposes of calculation , it will be very convenient to transform the expressions into JACOBI'S series ; being defined by After WEIERSTRASS , if we put then can be computated from .
ARISING FROM THE LOAD OF NEIGHBOURING OCEANIC TIDES .
At a distance from the edge of the loaded circle , is a fairly small quantity and consequently the terms after second or third may be dispensed with .
The series of the functions needed here are as follows:\mdash ; , ; At the point near the edge of the loaded circle , the above expansions cease to b.e applicable .
For this case , our object will be accomplished by using the quantity instead of , which is defined by is calculated from , in which For example , to calculate near the edge , we .
proceed like this : By the aid the relation the function may be transformed into Making use of the transformation formulae of Theta-functions it will be easily shown that VOL. CCXVII .
and , consequently we have The series for the functions required here are In the accompanying diagram the courses of and its functions of the distance of the point of observation from the circle , with the proper scale .
The lower curve thus represents the vertical displacement and the upper one the corresponding slope outside of the If we liken the North Atlantic to a circular basin of a large radius and determine the relative position of Potsdam or Chicago referred to the centre of it , the attraction effect and the tilting effect of the periodic filling and emptying of tide , which might ARISING FROM THE LOAD OF NEIGHBOURING OCEANIC TIDES .
the extra east-west force , in observations of lunar disturbance of may be computed .
If we suppose the place of the observation to be not very near to the circular basin , the effect , as we see from above diagram , is of course small , but it increases rapidly as the edge is approached .
For example , suppose the radius of the North Atlantic basin to be 2000 km .
, the position of Chicago to be 3000 km .
from the centre , and the level of the water in this area to be raised one meter , then and Further assume that the density of sea water is 1 , and in c.g. .
units , then we shall have Accordingly the total effect amounts to It will be noticed that the effect oftilting is about three times as great as that of the attraction , so far as the material constants are assumed as above .
According to Lord KELVIN , * who initiated these investigations , the direct lunar effect on the deviation of a.plumbline is a maximum when the moon is at the altitude 45 degrees and amounts to nearly .
The total effect of a tide of amplitude one metre ( which is possibly two or three times the actual amount ) found here is not small enough to be neglected compared with the direct effect of the moon , the former is nearly half the latter .
As the tilting effect and the attraction effect of the tide wave are directly proportional to the height of the tide , the total effect oscillates in time in accordance with the law which the tide obeys .
There is , in general , a ifference in phase between the lunar effect and the tidal effect , which is worthy of closer investigation .
But we must bear in mind that the calculation adopted here is nothing but a rough estimation of order of magnitude , since the North Atlantic is far from circular , the tidal loading in it is never uniform .
* Natural Philosophy , ' Part II .
, p. 383 .
Fortunately , the phase ence of both effects in MICHELSON'S experiment may be neglected in a rough estimation , owing to the relative position of Chicago and the centre of the North Atlantic .
Nevertheless the above analysis shows that the tidal effect on the measurement , even at a point as far from the coast as Chicago , plays an and cannot be regarded as a small correction .
IV .
Elliptic Loading .
Next , let us suppose that the tide in the North Atlantic is not uniform , surface is given by the equation viz. , the excess pressure on the bottom due to the tide diminishes on approaching coast so as to amount to for ( 26 ) In this case the function becomes ( 27 ) Therefore we have ( ) ( ) , ( 28 ) and ( ) .
( 29 ) The integrals contained in the above can be obtained by expanding the trigonometrical functions into power series of and making use of the formula ( ) .
Thus ( 30 ) ARISING FROM THE LOAD OF NEIGHBOURING OCEANIC TIDES .
where These series converge for , and are applicable in this region .
At the surface we have to put and .
Since we have .
( 31 ) Consequently . . . .
( 32 ) for To find the expression for the vertical displacement within the loaded circle we proceed as follows:\mdash ; Making use of the power series of BESSEL'S function , we have ( ) .
( 33 ) where stands for The evaluation of the function can be undertaken by the aid of the formulae A little calculation will give us and , in general , Thus the integral on the left-hand side of ( 33 ) can be expanded in an ascending power series of which probably converges for limited values of if the value of is fixed .
This series and that found in ( 30 ) have a common region in .
which they are both convergent and therefore they must be congruent to each other in that region .
On the proof of this proposition we shall not enter , but we shall find the region of convergency of the latter series at the boundary .
Expand into a power series of , supposing to be sufficientJy small , then the first term of it will be !
If we retain only the terms which contain the first power of in the series of ( 33 ) , its general term will then be The series which has this expression as its general term converges obviously for the values of smaller than .
Thus the expansion ( 33 ) applies for when is an infinitesimal .
Since , for , we have ( ) , ( 34 ) for Hence ( 35 ) for Quite similar arguments may be employed to find the expression for the tiltin We shall have ( 36 ) ARISING FROM THE LOAD NEIGHBOURING OCEANIC TIDES .
Consequently , ( 37 ) for The general march of the function and are shown in the attached ( fig. 2 ) in a suitable .
The lower curve represents the vertical displacement and the upper one the slope of the ground .
The course of the curves is very similar to that of fig. 1 , as we would expect , except at the points very near to the edge of the loaded area , where , as there is no discontinuity in the load in this case , the slope remains finite .
Le .
us take an example , with the same assumption regarding the various constants and the position of the point of observation as in the former example , and suppose that the total amount of the load is the same as before , i.e. , the mean height of the tide is one metre .
Then we shall and the same a6 the result in the former example .
If we suppose the place of the observation to be nearly at the edge of the loaded area , then and amounting to more than the

