252 i On the Non-Periodic or Residual Motion of Water Moving in Stationary Waves .
By Mrs. Hertha Ayrton .
( Communicated by Professor J. H. Pointing , F.R.S. Received December 7 , 1907 , \#151 ; Read January 30 , 1908 .
) [ Plates 4\#151 ; 6 .
] It is well known that when water moves in stationary waves , the particles do not , like pendulums , simply swing to and fro , returning after each oscillation to the points from which they started ; but that each element takes up a new position after each oscillation , so that it traces out a path for itself , only returning after many oscillations to its starting point .
Part of this non-periodic or residual motion , as I shall call it , in stationary waves , has been traced out mathematically by Lord Rayleigh .
The object of the present paper is to show , experimentally , what it is , as completely as possible .
In his classic paper " On the Circulation of Air observed in Kundt 's Tubes , and on some Allied Acoustical Problems , " Lord Rayleigh examined , among other things , the influence of the bottom of a horizontal vessel on the motion of water moving in it in stationary waves , and he came to the following conclusion : calling places of maximum horizontal motion loops , and places of maximum vertical motion nodes , " near the bottom the fluid rises from the bottom over the nodes , and falls back again over the loops , the horizontal motion near the bottom being thus directed towards the nodes and from the loops .
" Quite close to the bottom , on the contrary , he found that the motion was in the opposite direction , from the nodes and towards the loops .
Fig 1 shows these two sets of vortices diagrammatically .
B A E D* F Mode Loop Node Loop Node Fig. 1.\#151 ; Diagram of Lord Rayleigh 's Residual Vortices in one complete stationary wave of water .
In the course of some experiments on the motion of heavy particles under stationary waves , finding that it was necessary to trace out the complete Ayrton Roy .
Soc. Proc. , A. vol. 80 , Plata P- Level of wAber aG rest Fig. 3 .
Instantaneous Photograph of Water permeated with Bronze Powder , oscillating in half a stationary / 3 .
Diagram of Residual Motion indicated by Bronze Powder in Water , oscillating in half a stationary we ABC , A'B'C ' , Lord Rayleigh 's upper vortices ; DEF , D'E'F ' , " end vortices .
" ( Shallow water .
) a P- Level erf wAber Ab resG Node Fig. 4 .
Instantaneous Photograph of Water permea/ ted with Bronze Powder , oscillating in half a stationary / j3 .
Diagram of Residual Motion indicated by Bronze Powder in Water , oscillating in half a stationary ABC , A'B'C ' , Lord Rayleigh 's upper vortices ; DEF , D'E'F ' , " end vortices .
" Depth of water nearly a wave-length .
Non-Periodic or Residual Motion of Water , etc. 253 residual motion of the water , I tried to render this visible , oscillating the water in the simplest possible way , i.e. , in time with its natural swing , so that it formed half a stationary wave .
The first attempts , which were made in a trough 3 feet long , were entirely unsuccessful : I found it impossible , with the help of any visualising agent I could think of , to disentangle , with the eye , the residual motion from the to-and-fro movement .
In a trough only 10 cm .
long , however , oscillated in the same way , the residual motion to within less than a millimetre of the bottom could easily be seen , even when pepper was the visualising agent ; but still better when this was replaced with bronze powder , rubbed down in a little gum with a stiff brush and then well washed .
As a certain amount of residual motion continues for some time after the oscillations have ceased , a good way to observe , it is to oscillate the trough till plenty of bronze powder has been set in motion , and then to bring it to rest .
When this was done with the little 10-cm .
trough , Lord Rayleigh 's upper vortices , ABCD , etc. ( fig. 1 ) were clearly discernible ( inner vortices , fig. 2 , Plate 5 ) though the motion of the thin layer quite close to the bottom could not be detected in this particular way .
What was plain , however , was that there was some residual motion , even quite near to the bottom , that could not be accounted for by either of Lord Rayleigh 's two sets of vortices ; for instead of his upper pair spreading horizontally along the whole length of the trough , as they should have done , they occupied a part of the length only\#151 ; at the middle\#151 ; and were flanked on either side by much more active vortices , that monopolised the whole of the ends of the trough from top to bottom .
Moreover , the horizontal space occupied by each of the two sets of vortices varied with the depth of the water .
When this was small compared with its length\#151 ; say 1 cm .
deep to 10 cm .
long\#151 ; Lord Rayleigh 's upper vortices ( ABC , A'B'C ' , ( 3 , fig. 3 ) , took up most of the length of the trough , and the end vortices ( DEF , D'E'F ' , / 3 , fig. 3 ) occupied very little room at the bottom , although they always met in the middle near the top of the water .
As the water was deepened , however , the end vortices became longer and the middle ones shorter ( fig. 4 , Plate 4 ) till , with water as deep as half a wave-length , the middle vortices were practically squeezed out of existence by the expanded " end vortices " ( fig. 5 ) , while small corner vortices with a languid movement appeared at F and F ' ( fig. 5 ) .
For these experiments with deep water I found aluminium powder better than bronze powder , on account of its lightness .
This visual test , therefore , brings out one very important point , namely , that the complete residual motion of the water , even quite close to the bottom , is dependent on the depth of the water compared with the length of 254 Mi 's .
H. Ayrton .
Non-Periodic or Residual [ Dec. 7 , the stationary wave , and cannot be determined without taking that relation into account .
Level of water cit rest Fig. 5.\#151 ; Diagram of Kesidual Motion indicated by Bronze Powder in half a stationary wave .
Depth of water greater than half a wave-length .
The foregoing experiments were , as I have mentioned , made with water moving in half a stationary wave , so that the loop was at the middle of the trough , and the two nodes were at the ends .
It now seemed important to find out whether the " end vortices " would come into play at any node , or whether they only appeared at nodes at the ends of the vessel .
Accordingly I oscillated the water\#151 ; still permeated with bronze powder\#151 ; in such a way as to obtain three loops and four nodes , i.e. , in one and a half stationary waves .
Immediately double " end vortices " came into play at each of the two interior nodes , as well as the single ones at the end nodes , so that the residual vortices appeared as shown diagrammatically in fig. 6 .
Thus , what I have called " end vortices " are not confined to the ends of the trough ; they come into play at each node of any series of stationary waves .
The almost vertical light line down the middle of the trough in a ( fig. 4 ) is a good example of what is seen in all experiments with stationary waves in water permeated with visualising powder .
At each vertical plane of Ayrton .
Roy .
Soc. Proc. , A. vol. 80 , Plate \#163 ; .
_ _ Level _of_ water _ \amp ; 6 _ resb__ Fig. 7 .
a. Instantaneous Photograph of Curves formed during first few Oscillations in Water moving m half a state wave , by water-colour stain placed along bottom of trough , ft. Diagram of Curves showing Residual Ma in Water moving in half a stationary wave during first few oscillations .
Fig. 2.\#151 ; Instantaneous Photograph of Water permeated with Bronze Powder , just after it has ceased oscillating in half a stationary wave .
i i M- Fig. 8\#151 ; Instantaneous Photograph of Curve formed , during first few Oscillations i Water moving in half a stationary wav by water-colour stain placed near middle o bottom of trough on one side , 1907 .
] Motion of Water Moving in Stationary Waves .
demarcation between two residual vortices such a line is visible .
These planes occur , as is clear from fig. 6 , at all the loops and at all the nodes , Level of water at rest Fig. 6.\#151 ; Diagram of Residual Yortices indicated by Bronze Powder in Water oscillating , in one and a half stationary waves .
except the end ones , of any system of stationary waves .
In a ( fig. 4 ) where* water oscillating in half a stationary wave is depicted , there is one singlevertical plane of demarcation only , viz. , at the loop , which is at the middle of the trough , hence the vertical line in the middle .
I have said that the visualising powder only enables one to follow the residual motion of the water to within a millimetre or so of the bottom of the trough .
That means that Lord Rayleigh 's thin vortices ( DAEF , etc. , fig. 1 ) cannot be detected by means of it .
I noticed , however , that when the water was quite clear and only a thin layer of powder , or a small drop of water stained with water-colour paint , rested at the bottom , this rose , during the first few oscillations , in the shape shown in fig. 7 , moving in the directions indicated by the arrows in the diagram ( / 3 , fig. 7 ) .
Here ABCD and A'B'C'D ' were clearly Lord Rayleigh 's upper vortices , and what went on close to the bottom could be inferred from the fact that a clear space of water devoid of powder quickly formed between AB , for instance , and the bottom , while the powder continued to be carried along AB and thence upwards to G and down again to D. The inference was that the powder at the bottom was being driven , by the residual current there , in the direction from E to A , so that it all gradually got carried away from the bottom back along AB , up BC , and so on .
If this were so , then I had ocular proof of a part of Lord Rayleigh 's thin vortices ( DAEF , etc. , fig. 1 ) between AB and the bottom ( / ?
, fig. 7 ) ; for the powder and , therefore , the residual current , went from the nodes E , E ' , towards the loop A along the bottom , and from the loop towards the nodes along AB , AB ' .
VOL. LXXX.----A .
s 256 Mrs. H. Ayrton .
Non-Periodic or Residual [ Dec. 7 , To make sure that this inference was correct , I used clear water without powder , and squirted on to the bottom of the trough , with a pen-filler , a bright green solution near the left end , and a bright yellow one near the middle on the same side , holding the filler quite close to the bottom , so that the water should not be stained higher up .
Having waited till the water had come to rest again , I then oscillated the trough as before , the idea being that the green colour would visibly travel along the bottom to the middle , while the yellow would move above it along AB and then up round CD \lt ; A fig- 7 ) .
This actually happened .
After a few oscillations the green stain was entirely in the middle of the trough at the bottom , while the yellow travelled towards the end of the trough and then rose , as shown in fig. 8 ( Plate 5 ) , leaving a clear space at the bottom between itself and the trough ; indeed , for an instant it was possible to see green travelling along the bottom towards the middle of the trough , and yellow along AB , above it , towards the end , the two colours passing one another , travelling in opposite directions ; after the green had reached the centre of the trough , it also moved back to the end , and the two colours became mixed .
But the first effect was very striking , and proved beyond doubt , experimentally , the existence of the residual currents along the bottom from node to loop and return currents a little higher up , from the loop towards a node , that Lord Rayleigh had predicted .
The question now arose , where does the downward flow in these thin vortices take place ?
There is , of course , a downward flow at the nodes , but that current extends from top to bottom of the water and belongs to the " end vortices .
" If Lord Rayleigh 's thin vortices were separate and distinct from the " end vortices , " there ought to be a downward flow at some such point as B ( 0 , fig. 7 ) .
But the fact that the water under B always became clear so quickly after the oscillations were started , and thenceforward remained so , seemed to show that there could be no residual downward flow there , otherwise some of the stain or powder would be seen to move downwards from B to the bottom .
This never happened , however ; on the contrary , stain squirted right , over B , even though it was , as usual , heavier than water , always travelled upwards along such a line as BC , but never downwards .
It seemed to me , therefore , that the " end vortices " and Lord Rayleigh 's thin vortices must really form one vortex of the peculiar shape shown in ABCDEFA , AB'C'D'E'F'A ( 0 , fig. 9 , Plate 6 ) .
To test this , I squirted a little green stain on to the bottom of the trough , quite at each end , and waited till it had fallen perfectly flat .
I then oscillated the trough so that the water moved in half a stationary wave , and saw the stain spread itself as shown in a ( fig. 9 ) .
It first crept along the bottom to Ayrton .
Roy .
Soc. Proc. , A. 80 , Plate 0- Level of water at rest Fig. 9 .
a. Instantaneous Photograph after many Oscillations of Water moving in half a stationary wave .
Besidual in shown by water-colour stain placed on bottom of trough when water was at rest .
/ 3 .
Diagram of \gt ; of principal Besidual Vortices in Water oscillating in half a stationary wave .
ABCDEFA , AB'C'D'F -outer vortices ; ABCG , AB'C'G ' , inner vortices .
Depth of water less than a quarter of a wavedength .
Fig. 10.\#151 ; Instantaneous Photograph of Water moving in half a stationary wave .
Besidual motion shown by water-colour stain placed on bottom of trough when water was at rest .
Depth of water more than a quarter of a wave-length .
Fig. 11.\#151 ; Instantaneous Photograph of Wate permeated with Bronze Powder , oscillatin in half a stationary wave , after about si oscillations , 1907 .
] Motion of Water Moving in Stationary Waves .
257 the middle , A ( / ?
, fig. 9 ) then rose up and travelled backwards in such a line as AB ( taking the left side only ) ( / ?
, fig. 9 ) , then turn back again along BC , branching off at C , where part of it rose up CD and turned back along DE and down towards F , while part went down towards A and turned back towards G. The only interpretation that it seems possible to give to such a series of motions as this is that Lord Bayleigh 's thin residual vortices and the " end vortices " form part and parcel of one another and together make a vortex having an outer layer shaped like ABCDEFA ( / 3 , fig. 9 ) .
Another layer of this odd vortex must have a shape like HKM , another like NOP , and so on .
It seemed , at first , impossible to believe in the existence of even a residual vortex of such a form , but I have applied many tests , and each one has only shown more forcibly that this , and no other , is the real form of the vortex .
I have , for instance , injected tiny drops of stain into the water at such points as N , B , H ( / 3 , fig. 9 ) , and oscillated the water in half a stationary wave before they had had time to fall , and whichever point they were placed at they always moved as if they were being carried along by such a vortex as ABCDEFA .
I have placed drops of stain successively at various points along the same horizontal line a short distance from the bottom , and watched , while oscillating in the same way , to see whether there was any sign of a downward current anywhere except at an end of the vessel\#151 ; but there was none .
There is , therefore , I consider , no room to doubt that , when the depth of water is less than half the wave-length , the residual motion does actually take the courses indicated in / 3 , fig. 9 .
As the pair of vortices ABCDEFA , AB'C'D'E'F'A ( / 3 , fig. 9 ) entirely embrace the pair ABCG , AB'C'G ' , I propose to call the first pair the outer residual vortices , and the second the inner ones .
In order to trace with more accuracy the exact way in which the height and length of the residual vortices altered as the water was deepened , and also to see what happened when the depth of the water was greater than half the length of the stationary wave , I used a trough 15 cm .
high , 1 cm .
wide , and 10 cm .
long , with water-colour stain placed at the bottom while the water was still .
Oscillating this so that the water formed the usual half stationary wave , I found that the inner vortices , which occupied practically the whole length of the trough when the water was very shallow " , e.g. , 5 mm. or so deep , diminished continually in length as the depth of the water increased , till they disappeared when this was half a wave-length .
Their height , on the other hand , increased at first as the water wras deepened , and attained a maximum when the depth of water was about one-eighth of a wave-length .
After this , the height of the inner vortices diminished again , s 2 258 Mrs. H. Ayrton .
Non-Periodic or Residual [ Dec. 79 and became practically zero when their length became zero , i.e. , when the water depth was equal to half a wave-length .
As for the outer vortices , they evidently consist of three parts ( / 3 , fig. 9 ) , REFQ occupying the whole depth of the water between a node and the nearest point to that node of the corresponding inner vortex ABCG ; BCDR extending over the inner vortex to the top of the water ; QAB lying between the inner vortex and the bottom .
Deepening the water increases the length of REFQ and the height of BCDR , but diminishes the length of both BCDR and QAB , till , when the depth of the water is half a wave-length , REFQ extends the whole way from loop to node and from top to bottom of the water , except for the small place occupied by the ineffective little corner vortices already noticed , F , F ' ( fig. 5 ) .
A comparison of a ( fig. 7 ) , a ( fig. 9 ) , and fig. 10 is very instructive in showing the way in which deepening the water , while retaining the same wave-length , alters the length and height of the inner residual vortices .
In the shallow water in fig. 7 these are nearly as long as the trough , and reach nearly to the top of the water .
In the deeper water in fig. 9 they do not extend so far along the trough nor to the top of the water , and yet they are about the same height as those in fig. 7 .
In fig. 10 , where the water is still deeper , the inner vortices do not occupy more than about half the length of the trough , and are only about a third of the height of those in fig. 9 .
It seemed just possible that , if the water were made considerably deeper than half a wave-length , such vortices as F and F ' ( fig. 5 ) might spread along to the middle of the trough\#151 ; to the loop\#151 ; and become important .
No such thing happens , however ; it is difficult , even when the oscillations are very violent , to get any motion at all in the water lower down than half a wave-length , and the small corner residual vortices never spread under the outer ones ; they rise a little higher , perhaps , as the water is deepened , but they remain always feeble and always in the corner , with their lower boundaries on a level with those of the outer vortices , and below this there is no apparent residual motion at all .
Practically , then , in water as deep as , or deeper than , half a wave-length , the residual vortices are reduced to a single pair , of which the upper portion moves from loop to node and the lower from node to loop , and this single pair always extends to the top of the water , but moves over a layer of still water at the bottom , when the depth of ihe water is much more than half a wave-length .
It is interesting to notice that , whatever the depth of the water may be , in consequence of the way in which the outer residual vortices enclose the inner ones , the direction of the residual motion on top is always from loop 1907 .
] Motion of Water Moving in Stationary Waves .
259 to node , that at the bottom from node to loop , and that at the nodes downwards ( save for the small corner vortices ) .
At the loops , on the contrary , although the residual motion of the upper part of the water is invariable , namely , vertically upwards , that of the lower part depends on the depth of the water , being downwards for any depth less than half a wave-length , and upwards\#151 ; moving in the same direction as the upper part\#151 ; when the depth is greater than this .
Mathematically speaking , I suppose , there is always an infinitesimally small pair of inner vortices , and , therefore , a small downward current between the bottoms of the two outer vortices , however deep the water may be , but practically in water as deep as , or deeper than , half a wave-length these do not exist .
Fig. 11 is a photograph of water , permeated with bronze powder , which had been oscillated in half a stationary wave some half dozen times only , it shows clearly where the residual motion first becomes perceptible , viz. , at the top of the water , at the nodes ; for there is no apparent disturbance anywhere else .
The friction of the water against the front and back of the trough causes the residual motion to be slowest close to the glass and greatest half way between the front and back .
This gives to the water-colour stain , that shows the residual motion in each quarter wave-length , a shape as of a sail bellying in a wind , which drives from the end of the trough towards the middle .
The effect can be seen to a certain extent in a ( fig. 9 ) , but is much more clearly visible in the actual experiment .
One final feature of residual motion , that is rather curious , must be mentioned .
It has been said that , after the oscillations have been stopped , the residual motion continues , and is even more apparent , owing to the cessation of the to-and-fro movement .
In a short time , however\#151 ; perhaps a minute after the oscillations have ceased\#151 ; a part of the water begins to move in the opposite direction to that which it had previously taken , and this tendency gradually spreads till the whole motion of the liquid is reversed , though it is more confused than it was before .
Where originally there was upward movement , there is now downward , and vice versed ; motion from right to left has become motion from left to right , and so on .
It is as if a set of water springs had been wound up , and now proceeded to unwind themselves .
To sum up the results of this investigation ; the residual motion of water moving in stationary waves takes the form of two principal sets of vortices n each half stationary wave\#151 ; the outer ones extending from loop to node at top and bottom , but embracing the inner ones in the middle\#151 ; the inner pair , entirely surrounded by the outer pair , having their largest part at a loop , Profs .
C. S. Myers and H. A. Wilson .
[ Jan. 2 , and tapering to points near the bottom , between this loop and the corresponding nodes .
Both the length and the height of the inner pair depend on the depth of the water , the length diminishing steadily as the depth of the water increases , and the height increasing till the depth of the water is about one-eighth the wave-length , and then diminishing till the water is as deep as half a wave-length , when they are practically reduced to nothing .
The outer pair of vortices , since they entirely surround the inner ones , naturally grow as these diminish , until at last , when the depth of the water is half a wave-length and the inner vortices cease to exist , the whole length of the half wave is occupied by two oblong residual vortices moving from loop to node above and from node to loop below , the vertical motion being upwards at the loop and downwards at the node .
On the Perception of the Direction of Sound .
By Professor C. S\gt ; Myeks , M.A. , M.D. , and Professor H. A. Wilson , D.Sc .
, F.R.S. , King 's College , London .
( Received January 2 , \#151 ; Read January 16 , 1908 .
) The following paper contains an account of a series of experiments on the perception of direction of sound which were undertaken with the object of investigating the nature of the influence of phase differences between the vibrations at the two ears .
Lord Rayleigh* has shown that such differences help to determine the apparent direction of the sound , the sound appearing to be on the side at which the phase is the more advanced .
Professor Moref arrived at a similar conclusion to Lord Rayleigh by experiments of a different character .
The following paper also contains a theory of the influence of phase differences which appears to offer a possible explanation of the observed effects .
Most of our experiments have been done with an apparatus similar in principle to Professor More 's , but permitting of a continuous variation of the difference of phase .
The apparatus consisted of a brass tube , AB ( fig. 1 ) , about 250 cm .
long and 2*5 cm .
in diameter , with a short T-piece soldered on to it at its middle point .
This tube could slide freely in two slightly larger brass tubes , CD and OF , which were supported horizontally a definite * 'Phil .
Mag. , ' February , 1907 .
t 'Phil .
Mag. , ' April , 1907 .

