III .
" On Waves in Liquids .
" By W. J. MACQUORN RANKINE , C.E. , LL. D. , F.R.S. Received April 16 , 1868 .
( Abstract .
) ( 1 ) Object of this Paper.-It has long been known that in an uniform canal filled with liquid , the speed of advance of a wave in which the horizontal component of the disturbance is uniform from surface to bottom is equal to the velocity acquired by a heavy body in falling through half the depth of the canal .
But , so far as I know , it has not hitherto been pointed out that a similar law exists for waves transmitting a disturbance of any possible kind in a liquid of limited or unlimited depth , provided only that the upper surface of the liquid is a surface of uniform pressure .
The object of this paper is to demonstrate that law , and to show some of its applications .
( 2 ) Velocity of Advance defined.-Throughout this investigation the velocity of advance of a wave will be defined to be the mean between the velocities with which the shape of the wave advances relatively to a surfaceparticle at the crest , and to a surface-particle in the trough respectively .
In ordinary rolling waves the velocities of particles in those two positions are equal and contrary , so that the speed of advance as above defined is equal to the speed of advance of the wave relatively to the earth .
A wave of translation in which the velocities of particles at the crest and hollow are not equal and contrary , may be regarded as produced by compounding the motion of a rolling wave with that of a current whose velocity is half the difference of the velocities of those particles .
( 3 ) Relation between hleight of wave and horizontal disturbance at the surface.-The followingrelation between the height of a wave and the horizontal disturbance of the surface-particles has already been proved and made use of by various authors ; and it is demonstrated here for convenience only .
Let uand --u1 be the velocities of a surface-particle at the crest and trough of a wave respectively .
Let a be the velocity of advance of the wave as defined in article 2 .
Conceive a horizontal current with the uniform velocity -a to be combined with the actual wave-motion ; the resultant motion is that of an undulating current , presenting stationary waves in its course ; and the forces which act on the particles are not altered .
The resultant velocity of a particle at the crest becomes -a+ u_ ; and the resultant velocity of a particle in the trough becomes ---u1 .
Let the height from trough to crest be denoted by Az ; then , since the upper surface of the liquid is supposed to be a surface of uniform pressure , the principle of the conservation of energy gives the following equation : V2 { ( p + U-nr ( a-u)r}i z on ... .i .
( 1 2 ) ( 4 ) Virtual Depth of Uniform Horizontal Disturbance.-By the phrase " virtual depth of uniform horizontal disturbance , " or , for brevity 's sake , virtual depth , I propose to denote the depth in the liquid to which an uniform horizontal disturbance would have to extend , in order to make the amount of horizontal distZrbance equal to the actual amount .
That is to say , conceive that a pair of vertical planes normal to the direction of advance , and each of the breadth unity , coincide at a given instant , one with the trough-line or farrow , and the other with the crest-line or ridge , which bound one of the slopes of a wave .
We will suppose this to be the front slope , merely to fix the ideas ; for similar reasoning applied to the back slope leads to the same results .
At a given depth Z below the surface , let -u't be the horizontal velocity with which particles are in the act of passing backwards through the plane of the trough , and + ut the velocity with which particles are passing forwards through the plane at the crest ; then the rate by volume at which liquid is passing into the space between those two planes is S u'dz+S u"dd ; the integrations extending from the surface to the bottom .
Let k denote the virtual depth ; then k=ud .
+_5u _d.e ( 2 ) 2u1 ( 5 ) Relation between Virtual Depth and Speed of Advance .
In an indefinitely short interval of time dt , the volume of liquid which passes into the space between the two vertical planes mentioned in article 4 , is 2kcu1dt ; and in order to make room for that volume of liquid , the front slope of the wave must sweep in the same interval of time through an equal volume .
But the volume swept through by the front of the wave is adtz ; so that , cancelling the common factor dt , we have the following equation : aAz= 2kcu ; but , according to equation ( 1 ) , As 1 ; which value being substituted in the above equation , gives 2aWul , 2k~T and therefore = ka , and a =V ; *. .
( 3 ) so that the velocity of advance of a wave ( defined as in article 2 ) is equal to that acquired by a body in falling through half the virtual depth ; and this is true for all possible waves in which the upper surface is a surface of uniform pressure .
( In article 6 of the paper , the speed of advance of a wave of translation is expressed by combining the speed of a rolling wave , / gk , with that of a supposed current , as stated in article 2 .
1868 .
] In articles 7 , 8 , and 9 the law which connects the speed of advance of a wave with the virtual depth is compared with the already known laws of the transmission of rolling waves in water of limited or unlimited depth .
The principal results may be summed up as follows .
Let T be the periodic time of a wave , in seconds ; h=-T2 the equivalent pendulum , that is , the height of the pendulum whose period is the same ; c= len ththe rolling 2r radius , being the radius of a circle whose circumference is equal to a wavelength ; u1 the greatest horizontal velocity , and w , the greatest vertical velocity of a surface-particle ; a the velocity of advance ; then V/ uu ^ ' ul 2and k= .
C=.2 .
) ut1 ut ( 10 ) Oblique Advance of Forced Waves.-Let s be the velocity with which a floating solid body is driven horizontally ; the wave which that solid body pushes or drags along with it is forced to advance at the velocity s also ; while the virtual depth of disturbance , k , bears some relation to the depth of immersion and figure of the solid body .
If the speed of advance corresponding to that depth , a= V-gk , is less than s , a pair of wave-ridges diverge obliquely from the path of the floating body towards opposite sides ; and the sine of the angle which each of those ridges makes with that path is a. Such is the mode of formation of the obliquely spreading waves which travel along with ships .
When the velocity of the floating body is less than the speed of advance corresponding to the depth to which it disturbs the liquid in its immediate neighbourhood , it is probable that the virtual depth of disturbance of parts of the liquid beyond the immediate action of the floating body adjusts itself to the velocity , and assumes the value- .
g 1 .
Possibility of Obliquely Advancing Tidal Waves.-It is possible that instead of a depth less than the virtual depth corresponding to the speed of advance of a tidal wave , the ridge of that wave may place itself in a position oblique to the parallels of latitude , according to the principle stated in article 10 .
It still remains to be ascertained , by the study of tidal observations , whether such phenomena take place in the tides of the ocean .
12 .
Terminal Velocity of TVaves.-It is known that in deep water all waves left free from the action of disturbing forces tend ultimately to assume the condition of free rolling waves whose velocity of advance depends on See Watts , Rankine , Napier , and Barnes , ' On Ship-building , ' Division I. Article 156 , p. 79 .
their periodic time , and is expressed by the equation a= 2- .
This , then , may be called the terminal velocity of a wave of a given period .
It follows that if a wave is raised through the disturbance produced by a solid body , that wave will at first travel with a speed depending on the virtual depth of the original disturbance ; but as it advances to a greater and greater distance from the disturbing body , the velocity of advance will gradually approximate to the terminal velocity corresponding to the periodic time , and the virtual depth will continually adjust itself to the changing velocity , and approximate gradually to the equivalent pendulum corresponding to the periodic time .
Such is the cause of the forward curvature of the ridges of the obliquely diverging waves which follow a ship* .

