]\gt ; Lord Rayleigh .
( 4 ) A study of the necessary interval between current reversals in the process of netisation has been made , and it ; is shown that the delay in reversal of magnetic phenomena in considerable masses of iron , due to eddy currents , is extremely small when the magnetic inductions are less than C.G.S. units .
the Dynamics ] Revolving By LORD AYLEIGH , O.M. , F.R.S. ( Received December 8 , 1916 .
) So much of depends ultimately upon the dynamics of revolving fluid that it is desirable to formulate as clearly as possible such simple conclusions as are within our reach , in the hope that they may assist our meant when an exact analysis seems impracticable .
An important contribution to this subject is that recently published by Dr. Aitken .
* It formed the starting point of part of the ation which follows , but I perhaps to add that I do not share Dr. Aitken 's views in all respects .
His paper should be studied by all interested in these questions .
As regards the present contribution to the theory it may be well to premise that the limitation to symmetry round an axis is imposed throughout .
The motion of an inviscid fluid is governed by equations of which the first expressed by rectangular co-ordinates may be written , ( 1 ) where ) , ( 2 ) and is the potential of extraneous forces .
In ( 2 ) the density is either a constant , as for an incompressible fluid , or at any rate a known function of the .
Referred to cylindrical co-ordinates , ; , with velocities eckoned espectively in the directions of increasing , these equations become ( 3 ) ' ' The Dynamics of Cyclones and Anticyclones.\mdash ; Part 3 'Roy .
Soc. Edin .
Proc. , vol. 36 , p. 174 ( 1916 ) .
Compare Basset 's ' Hydrodynamics , ' S19 .
the of Revolving Fluid .
( 5 ) For the present purpose we assume symmetry with respect to the axis of , so that , and ( assumed to be singhed ) are independent of So simplified , the equations become , ( 6 ) , ( 7 ) , ( 8 ) of which the second may be written , ( 9 ) signifying that may be considered to lnove with the fluid , in accol.dance with Kelvin 's general theorem respecting " " circulation.\ldquo ; If , be the initial values of , for any particle of the fluid , the value of at any future time when the particle is at a distance from the axis is given by Respecting the motion expressed by , we see that it is the same as might take place with , that is when the whole motion is in planes passing through the axis , provided that we introduce a force along equal to .
We have here the familiar idea of " " centrifugal force and the conclusion might have been arrived at immediately , at any rate in the case where there is no motion .
It will be well to consider this case more in detail .
The third equation ( 8 ) shows that is then independent of , that is a function of ( and t ) only .
It follows from the first equation ( 6 ) that also is a function of only , and .
Accordingly by .
( 10 ) If , the potential of impressed forces , is independent of , so also will be and , but not otherwise .
For example , if gravity ( g ) act parallel to measured downwards ) , , ( 11 ) gravity and centrifuigal force contributing independently .
In ( 11 ) will be Lord Rayleigh .
constant if the fluid is incompressible .
For ases following Boyle 's law , .
( 12 ) At a constant level the pressure diminishes as we pass inwards .
But the corresponding rarefaction experienced by a compressible fluid does not cause such fluid to ascend .
The heavier part outside is prevented from coming in below to take its place by the force.* The condition for equilibrium , taken by itself , still leaves an arbitrary function of , but it does not follow that the equilibrium is stable .
In like manner an incompressible liquid of variable density is equilibrium under ravity when arl.anged in horizontal strata of constant density , but stability l.equires that the density of the strata everywhere increase as we pass downwards .
This analogy is , indeed , very helpful for our present purpose .
As the fluid moves ( and finite ) in accordance with equations ( 6 ) , ( 7 ) , ( 8 ) , ( vr ) remains constant ( k ) for a ring consisting always of the same matter , and , so that the centrifugal force actilg upon a portion of the fluid is inversely , and thus a known function of position .
The only dilference between this case and that of an incompressible fluid of variable density , under extraneous forces derived from a potential , is that here the inertia concerned in the motion is liform , whereas in a variably dense fluid moving under , or similar forces , the inertia and the weight are proportional .
As ards the question of the difference is immaterial , and we may conclude that the equilibrium of fluid revolving one way round in cylindrical layers and included between coaxial cylindrical walls is stable only under the condition that the circulation ( k ) always increases with .
In any portion where is constant , so that the motion is there " " irrotational the equilibrium is neutral .
An important particular case is that of fluid moving between an inner with angular velocity and an outer fixed cylinder .
In the absence of viscosity lvhe rotation of the cylinder is without eff'ect .
But if the fluid were viscous , equilibrium would require , that the circulation diminishes outwards .
Accordingly a fluid without viscosity cannot stably move in this manner .
On the other hand , if it be the outer cylinder that rotates while the inner is at rest , , and the motion of an inviscid fluid according to this law would be stable .
When the fluid is viscous the loss of circulation near the bottom of the containing veHHel modifies this conclusion , as explained by James Thomson .
's ' Hydrodynamics , ' S333 .
On the Dynamics of Revolvin .
Ftuids .
l5l We may also found our argument upon a direct consideration of the kinetic energy ( T ) of the motion .
For is proportional to , or .
Suppose now that two rings of fluid , one with and and the other with and , where , and of equal areas or are interchanged .
The corresponding increment in is represented by , and is positive if so that a circulation always increasing ontwards makes a minimum and thus ensures stability .
The conclusion above arrived at may appear to conflict with that of Kelvin , finds as the condition of minimum that the vorticity , proportional , must increase outwards .
Suppose , for instance , that , increasing outwards , while decreases .
But it would seem that the valiations contemplated difl'er .
As an example , Kelvin ives for maximum energy from to from to ; and for minimum energy from to , from to In the first case , and in the second case ; so that the moment of momentum differs in the two cases .
In fact Kelvin supposes operations upon the boundary which alter the moment of momentum .
On the other hand , he maintains the strictly two-dimensional character of the ible variations .
In the problem that I have con- sidered , mmetry round the axis is maintained and there can be no alteration in the moment of momentum , since the cylindrical walls are fixed .
But the variations by which the passage from one two-dimensional condition to another may be effected are not themselves two-dimensional .
The above reasoning suffices to fix the criterion for stable equilibrium ; but , of course , there can be no actual transition from a configuration of unstable equilibrium to that of permanent stable equilibrium without dissipative forces , any more than there could be in the case of a heterogeneous liquid under gravity .
The difference is that in the latter case dissipative 'Natu re .
' vol. 23 , October , 1880 ; 'Collected ' vol. 4 , p. 175 .
Lord Rayleigh .
forces exist in any real fluid , so that the iluid llltimately settles down into stable equilibrium , it may be after many oscillations .
In the present roblell Odinary viscosity does not meet the requirements , as it would interfere with the of the circulation of given rings of fluid on our easoning depends .
But for purely theoretical purposes there is no inconsistency in supposing the motion esisted while the the -motion is unresisted .
The next RupPoSitio to in order of simplicity is that is a function of and only , and that , or at most a finite constant .
It follows that is independent of , while ( 6 ) becomes , ( 13 ) deterlllining the pressnre .
In the case of an incompressible fluid as a function of is determined by the equation of continuity , where is a function of only ; and when and the initial circumstances are known , follows .
As the motion is now two-dimensional , it may conveniently be expressed by means of the vorticity , which moves with the , and the stream-function , connected with by the equation .
( 14 ) The solution , appropriate to our purpose , is , ( 15 ) where A and are arbitr'ary cousta1lts integration .
Accordingly ' .
( 16 ) In eneral , A and are functions of the time , and is a function of the time ab wcll as of A simple partlcular case is when is initially , and therefore permanently , uniform the fluid . .
( 17 ) Let further that initially the motion is one of pure rotation , as of a solid body , so that initially , and that the1l the outer wall closes in .
If outer radius be initially and at time equal to , then at time , ( 18 ) that ( 17 ) is still applicable under appropriate boundary conditions the fluid is viscous .
On the Dynamics of Revolving Fluids .
since remains unchanged for a given ring of the fluid ; and correspondingly , Thus , in addition to the motion as of a solid body , tho fluid acquires that of a simple vortex of intensity incyeasing as diminishes .
If at any stage the motion ceases , ( 6 ) ives and thus const .
Since , as a function of ] increases as diminishes , same is true for the difference of pressures at two given values of , say and where .
Hence , if the pressure supposed constant at , it must continually increase at If the flnid be supposed to be contained between two coaxial cylindrical walls , both walls must move inwards ether , and the process comes to an end when the inner wall reaches the axis .
But we are not obliged to an inner wall , or , indeed , any wall .
The fluid passing inwards at may be supposed to be .
And it remains the that , if it there pass at a constant pressure , the pressure at must continually increase .
If this pressure has a limit , the inwards flow must cease .
It would be of interest to calculate some case in which the motion is less simple , for instance , when fluid is removed at a point instead of uniformly along an axis , or inner cylindrical boundary .
But this seems hardly practicable .
The condition by which is determined requires the expression of the motion of individual particles , as in the so-called Lagrangian method , and this usually presents great difficulties .
We may however , formulate certain conclusions of a general character .
When the motion is slow elatively to the motion , a kind of equilibrium theory\ldquo ; approximately meets the case , much as when the slow motion under gravity of a variably dense liquid retains as far as possible the horizontal stratification .
Thus oil water is drawn off by a syphon without much disturbing the water underneath .
When the density varies continuously the situation is more delicate , but the tendency is for the syphon to draw from the horizonta .
stratum at which it opens .
Or if the liquid escapes slowly through an aperture in the bottom of the vesse.l , only the lower strata are disturbed .
In like manner when fluid is drawn off in the neighbourhood of a point situated on the axis of rotation , there is a tendency for the surfaces of constant circulation to remain On the of Revolving Fluids .
cylindrical and the tendency is the more decided the greater the rapidity of lotation . .
liquid is drawn always from along the axis and not symmetrically in all directions , as when there is no rotation .
The above is , in substance , the reasoning of Mr. Aitken , who has also described a simple nent in illustration .
P.S.\mdash ; lt may have been observed that according to what has been said above the stability of fluid motion in cylindrical strata requires only that the of the ation increase outwards .
If the circulation be in both directions , this disposition involves discontinuities , and the stability exists only mder the condition that symmetry with respect to the axis is rigorously maintained .
If this limitation be dispensed with , the motion is certainly unstable , and thus the stability of motion in cylindrical layers really requires that the circulation be one-signed .
the general question of the two( motion of luids between hxed co-axial cylindrical walls reference may be made to a former paper .
* The motion in o1l strata is stable provided that the " " rotation either continually increase or continually decrease in from the axis demonstration is on the same lines there set out for plane strata . .
Math. So vol. 11 , ) ( 1880 ) ; 'Scientilic ) ' vol. 1 , p. 487 .

