OUPblog > Language > Dictionaries & Lexicography > On The Definition of “Siphon”

On The Definition of “Siphon”

Recently much ado has been made about the definition of siphon in the Oxford English Dictionary (OED). Below the editors of the OED respond to the controversy.


“The full OED is currently undergoing its first major revision and update for one hundred years. At present just over one quarter of the text has been revised and published at www.oed.com.

The OED entry for siphon is at present unrevised: the existing definition for the main sense states that a siphon works ‘by means of atmospheric pressure, which forces the liquid up the shorter leg and over the bend in the pipe’. Since this definition was published, in 1911, this explanation has been queried, most recently by Dr Stephen Hughes. Dr Hughes states that it is not atmospheric pressure but the cohesion of the water that makes a siphon work: the water in the down arm in effect pulls the water in the up arm.

As a comprehensive historical dictionary, OED seeks to present and evaluate the widest evidence of usage. The way the action of the siphon is explained has varied, and the picture is more complex than reflected either in OED‘s definition or in that of Dr. Hughes. Indeed, there is continuing debate among scientists as to which view is correct. A common explanation is that atmospheric pressure acts on the liquid in the upper container to push it into the tube as gravity pulls downwards the liquid in the lower length of tubing (so that a vacuum would otherwise form in the space vacated). Recently the rise of the liquid in the ‘up’ tube has been attributed to the cohesiveness or tensile strength of the moving liquid rather than to atmospheric pressure: the falling liquid in the ‘down’ arm in effect drags or pulls that in the ‘up’ arm. We would expect to reflect this debate in the fully updated entry for siphon, due to be published later this year.”

15 Responses to “On The Definition of “Siphon””
  1. [...] This post was mentioned on Twitter by paigecasey and Christian Purdy, Lauren. Lauren said: RT @paigecasey: RT @PurdyOxford: OED editors respond to "siphon" definition discussion http://bit.ly/dounbN [...]

  2. Mind Builder says:

    It should be clear that both theories are partly correct. there are actually two different kinds of siphons that operate on two different principles. In every siphon I know of, gravity is ultimately the motivating force. And in some siphons, such as the supposedly demonstrated siphon in a vacuum, liquid cohesive forces must play a critical or exclusive role in raising the liquid. But the simple demonstration of a working siphon that starts out with a bubble at the top, preventing the liquid in each side from touching, and thus eliminating the possibility of liquid cohesion from contributing to bringing the liquid up the shorter tube, proves that in common siphons, air pressure plays a major role.

    It is important to note however that it is not the difference in atmospheric pressure between the upper and lower reservoirs that drives the siphon. In fact, if anything, that insignificant difference would tend to defeat the siphon.

    I don’t think your definition should reflect controversy in the general scientific community, regardless of the previous position of Dr Hughes. You should first contact Dr Hughes and see if he will revise his position after looking at the Wikipedia article, and then consult a few other physicists after they have reviewed the evidence. I expect you will find unanimous agreement among the experts on the important aspects.

    Overall, I would call the old OED definition correct, as far as it went, but leaving out the academic case of the vacuum siphon, and the important element of gravity.

  3. Chandra M.S. says:

    It is puzzling, mind-boggling to me to see such a struggle to “fit a theory” for a natural phenomenon – Siphon. Not withstanding various types of siphon and their endless variations and possibilities, siphon phenomenon rests on a very basic, natural Fundamental Principle of Motion – Newton’s First Law of Inertia – A body remains at REST forever, unless a Force is applied!

    Dr. Hughes claim that gravity alone is responsible for flow of water in a siphon is incorrect, at best misleading and inadequate! since gravity was never absent! Therefore, gravity can NOT initiate a specific motion at a specific time!

    Also, a Motion, such as water/liquid flow is accurately determined precisely by the Resultant Force having a magnitude and direction, i.e. a Vector.

    Having found its own level, water in a tank remains at REST in an equilibrium state – No motion, flow.

    A precise equation involving all the forces can be written, and it suffices here to know that pressures at either ends of the pipe factors into on either sides of the equation, thus the pressure difference becomes a factor.

    When a tube(inverted U) is inserted into water, nothing happens, rightly so, because the pressures on either ends are equal and hence they cancel out! But, that doesn’t mean it is FREE from the dependency on the pressure terms! In fact, it proves just the opposite! i.e., in other words, a pressure difference is NEEDED to begin the flow, disturbing the equilibrium state of water or liquid. This exactly what is done by ‘sucking’ air out of tube, creating a pressure difference. Thereafter, gravity, having overcome and countered the pressure, takes over to move the water/liquid. The flow will continue, of course, even after ‘sucking’ force is removed, yet it is conditional on certain condition, i.e., no bubble large enough to ‘break’ the flow. But, this doesn’t mean gravity alone is responsible for the motion! In fact, again as I’ve stated above, it is the Resultant force that is at work here, and it can be precisely measured and calculated, that determines the flow entirely.

    In fact, viscosity of the liquid also gets factored in the precise calculations of how much ‘sucking’ force is needed, the rate of flow, etc.

    Therefore, it is inaccurate and misleading to ‘generalize’ siphon principle by its specific case of equal pressures – canceled terms – case. In some practical situations, often to deal with pressure differences. The idea of ‘tensile strength’ of water or other liquids is ridiculous, as it is imprecise, and is already accounted quantitatively by the scientific concept of viscosity of liquids, by Reynolds number, etc.

    If we can not ‘define’ a principle by simple sound-bite, we shouldn’t attempt to do it so, as it would be inadequate, besides being misleading and inaccurate and unscientific.

    Reducing to simplistic words, by taking sides or inventing ‘theories’ to fit, etc. to describe a natural phenomenon such as siphon is causing this endless controversy, nothing scientific about it!

    If it takes a line or two or more to explain clearly, so be it!

    Also, note that air atmospheric pressure itself is due to gravity which is naturally present at all times! Its the interaction of different forces that governs a motion!


  4. Mind Builder says:

    Chandra – Gravity causes the flow in the siphon because although the gravity is there before the siphon is started, it is usually opposed by another force like your thumb pressed on the end of the tube. When the opposing force is removed, then the gravity acts to make the liquids move. Also see the Z-tube on Wikipedia for a clever way to test liquid tensile strength.

  5. Chandra M.S. says:

    z-tube – What’s that got to do with Siphon? NOTHING!

    You’re just cooking up completely something that it has nothing to do Siphon, what-so-ever!!!!

    Your concept of tensile strength is at best a voodoo, unscientific thing, ill conceived property, which is already accounted scientifically – viscosity! FYI, there is also, a well defined nascent property of liquids called as surface tension. In capillary flow it matters. You need to study States of Matter and how they occur and interact, Laws of Mechanics, Fluid mechanics!

    Siphon, is a specific principle, a natural phenomenon that occurs even in Nature. Weather phenomenon of Low pressure and High Pressure that occur in the upper atmosphere very scientifically! That’s the reason, why they can be computed precisely, and hence their effect can be forecast-ed.

    The LOW pressure front, rotating in the anti clock-wise, churns the ocean, generating an UPWARD draft which then ‘lifts’ (causes evaporation) water from the ocean to form a storm and contains the energy that is EXACTLY computable (if you know a lot of data)!! The upward draft thus created lifts water against the gravity! This is Siphon phenomenon – not some concocted wise experiment that is nothing to do with Siphon! Weather stations use supercomputers to solve this basic phenomenon with huge amount of data! It takes better than 50 hours computer time to solve for temperatures, pressure, winds, rain amounts, etc.

    Due to Law of Conservation of Energy, we also know exactly how that energy in the storm, rain clouds, etc would come down and affect the areas below. It is very tedious and painstaking computations, obviously due to the scale of it and amount of data collected.

    Therefore, siphon is a siphon, where there is a lift (force) of water/liquid is necessary above its own equilibrium level, and then when such lift happens to deliver the liquid directly in-line with gravity (facing ‘downwards’), the water/liquid would come down, at a rate of flow, again, specified and calculated exactly by knowing all the forces involved. This is Fundamental Mechanics, when further considering the liquid, gas(air) properties, becomes specific – Fluid Mechanics.

    Mechanics always are dynamic in nature, and hence their general behavior is represented by Partial-differential equations! Not just some voodoo thoughts or concepts.

    There is nothing magic about gravity that ‘it does’ in siphon!

  6. Mindbuilder says:

    Back in 1948 Nokes reported making siphons work in vacuum with boiled water, mercury, and dibutyl phthalate (a low vapor pressure liqud) The Z-tube demonstration of liquid tensile strength is very relevant here.

    What is it that powers a vacuum siphon if not gravity? Note that pendulum clocks are often powered by gravity in the form of weights hanging from cables wrapped around drums. In the pendulum clock, when you have wound the cables around the drums and you let go, the weights start powering the clock. When you fill the siphon tube with liquid and then let the liquid go, the weight of the liquid powers the siphon. It’s the release of gravitational potential energy from the water moving to a lower level.

    Note from my first post up above that I understand that common siphons use atmospheric pressure rather than liquid tensile strength. But both types are still powered by gravity.

    Check out the Wikipedia page on siphons and its associated discussion page.

  7. I tried to sign up to your RSS however it doesn’t appear to be working. I’ll come back later.

  8. Chandra M.S. says:

    Hi Mind…
    None of those you say, you list, are anything to do with the concept or the phenomenon called SIPHON!! The Law of Energy Conservation along with Laws of mechanics what makes EVERYTHING work, even including your irrelevant experiments! But, you want to think there is some mystery outside the scientific laws, that needs to be understood (as you claim you know it!), that mysteriously makes gravity alone responsible for something it is not! Gravity factors into EVERYTHING on this EARTH, or any other planet! That doesn’t mean, arbitrarily reduce (anything) to simply gravity, and some notion such as tensile strength of liquid. Why are you hung up on this thing, I can’t understand, when a well defined and quantified scientific concept – viscosity – does account it already?? and the Mechanics takes into account very accurately.

    But…for some reason, you’re using some specific cases, which are nothing to do with siphon, using them as some how connected to siphon?? Don’t try to CONFUSE the issue! with your counter questions about vacuum siphon and so forth…When you solve using the same mechanics laws, you would clearly know that gravity is only one factor among other forces exerting.

    The 3 States of Matter – is a very scientifically developed study by itself, and a complete analytical equations (PDE) have been written and solved, involving properties like viscosity, etc.!! Even, highly computational intensive numerical solutions have been devised to accurately predict and verified in 1940′s to 60′s! Your concocted notion a scientific-sounding phrase ‘liquid tensile strength’ has no significance besides viscosity-of-fluids, and you’re confusing others by such claims.

    Don’t bring some clever experiments, that are nothing to do with straight forward practical significant mechanics and physics of the Nature.

    Yes, you’re free to concoct what-ever you like, but that physics has to bend to your whims and fancies is your fantasy and voodoo fanaticism.

    So again, to reiterate, its a common phenomenon of basic motion in Fluid Mechanics caused and controlled by pressure difference, despite the gravity that is uniformly present every where and at all times! Geometry, Mass, etc., all factor into mechanics!

    An airplane, the Jet Plane, all modern planes, work on just ONE basic, simple principle – pressure difference caused between the upper and lower aerodynamic air (fluid) flow – of the contour of wings! In fact, its only a tiny delta difference, that’s good enough to give a LIFT to thousands of pounds of load weight of airplane countering the gravity! Here is a practical example of application of aerodynamic principle that is at work! In fact, ice build up on the front-edge of the wings becomes a danger to the stability of the flying aircraft, only because, the air pressure difference on the wing edge gets disturbed, could not be maintained! Imagine that!!

    Other well known and highly useful mechanical use of the fluid is in hydraulic control systems used as a backup, since fluid used to apply a controlled force and pressure is a SURE THING, as they it can be operated manually.

    Boeing, NASA, all the academia uses the same PDE’s, use the same computation methods, to solve and arrive at the same results each and every day! If what you say – gravity alone causes everything, ignore pressure difference etc. then why bother to do anything, or why should you have to sweat and toil to solve complex PDE’s or build wind-tunnels, etc.,!? not to mention spend millions of dollars of fuel cost, etc. No need to spend zillions to launch shuttle!

    I’m sorry, unfortunately, I can’t continue with your unscientific and irrelevant claims in order to change the definition of Siphon! I don’t see any point of it, to continue this discussion further, as you do not want to accept the well established and (practically) proven scientific, natural laws, and applied mechanics!

  9. Pedro Flecha says:

    Dr Hughes statement that it is not atmospheric pressure but the cohesion of the water that makes a siphon work seems adequate also to explain capillarity (adhesion). The combination of both phenomena have been used for milennia in the Andean Civilization.

  10. Stephen Hughes says:

    Since publication of my paper ‘A practical example of a siphon at work’ (http://eprints.qut.edu.au/31098/) there has been considerable debate about how siphons actually work. From reviewing numerous web postings and emails that people have sent me from all over the world, there appears be two competing views or models. One is that water flowing out of a siphon tube reduces the internal fluid pressure enabling air to push water into the siphon and up over the crown. In this model both gravity and atmospheric pressure are considered essential to the operation of a siphon. An alternative model is that water flowing out of a siphon pulls water into the tube and up over the crown due to cohesion between water molecules. In this model, only gravity is necessary to explain siphon flow.
    There appears to be a consensus that dictionary definitions should mention gravity as a motive force in a siphon. However, the question implicit in the introduction to this blog is, should the definition cite both gravity and atmospheric pressure or only gravity as the motive force? The main objection to the gravity only model is the assumption that water cannot take any tension, i.e. cannot be stretched without being pulled apart. In everyday life this does seem to be the case as, for example, it doesn’t take a huge amount of effort to drink a glass of water.
    However, in certain circumstances water can be very difficult to break apart. For example, if a flat-bottomed glass is pressed down onto a wet tablemat and the glass lifted, the mat often stays stuck to the glass. In this case, water molecules act like glue between the glass and tablemat. An even better demonstration of the strength of water bonds is to try and prise apart two flat blocks of glass stuck together by a thin layer of water.
    A closely related issue to how siphons work is surface tension. If a slowly dripping tap is observed closely it will be noticed that water drops hanging from the tap in the process of forming have a rounded shape. A slow motion film of falling water drops shows that they are spherical in shape (http://www.youtube.com/watch?v=Yi3LW5riHfc&feature=related ). This happens because water molecules at the surface of the drop are pulled inwards by water inside the drop. This pulling is effected by hydrogen bonds. The internal cohesion between water molecules can be thought of as internal surface tension or volume tension. The volume tension of water makes water molecules difficult to prise apart and so water has a tensile strength. Liquids other than water can also be siphoned, e.g. petrol and mercury and in these cases molecules are connected via van der waals forces.
    A phenomenon known as cavitation is also testament to the high tensile strength of water. Cavitation occurs when a stream of water moves at high speed creating enough force to rip water molecules apart. Cavitation often occurs next to propellers. A diesel mechanic friend of mine has given me two parts form a large diesel engine that show cavitation damage. One part is a propeller from a water pump and the other a water jacket from around a piston. Both parts have pits a few mm across in some cases.
    An interesting question is how is it possible for water to produce such damage? What happens is that water is ripped apart creating a vacuum that quickly fills with air dissolved in water creating a bubble, However, this bubble is no ordinary bubble as it is surrounded by water in a high state of tension. This tension causes the bubble to suddenly collapse heating the air to several thousand degrees (approaching the surface temperature of the Sun), which is capable of melting metal – hence the pits seen. An intriguing creature known as the pistol shrimp uses cavitation to shoot superheated bubbles of water (aquatic bullets) to kill its prey (http://www.youtube.com/watch?v=eKPrGxB1Kzc ).
    Although atmospheric pressure is not the motive force in a siphon, atmospheric pressure is one of the three ways that can be used to prime a siphon. For example, a siphon can be primed by sucking on the end of the siphon tube; either by mouth, (not recommended for health and safety reasons) or using a syringe. Another way of priming a siphon is to immerse the tube in a bucket of water to replace the air with water. (I’ve used this technique on numerous occasions in my laboratory). A third method is to use a pump to force water into a siphon tube and at the same time force air out.
    Experiments can be performed with a large bubble at the entrance of a siphon tube. In this case there is obviously no water between the water in the reservoir and water higher up in the siphon tube and therefore there can be no cohesive connection between the two bodies of water. However, a siphon can still be made to work. Does this prove that atmospheric pressure is essential to the operation of a siphon? No. It demonstrates that in some circumstances when a large bubble enters a siphon, water flowing beyond the bubble is able to expand the bubble reducing the pressure enough to enable atmospheric pressure to push water into the tube behind the bubble.
    The bubble-in-a-siphon situation is directly analogous to using a syringe to prime a siphon. This is fairly easy to demonstrate. Get hold of a large syringe, e.g. 50 ml. Attach a length of tube, say 3 – 6 mm in diameter and a few tens of cm in length. Push the plunger of the syringe in so that air is expelled from the barrel. Immerse the open end of the tube in water and pull on the syringe. Water will be drawn into the syringe and if the capacity of the syringe is great enough water will reach the barrel. Detach the tube from the syringe and drop it lower than the level of the water in the reservoir at the other end of the tube. Water should flow out of the tube and you’ve got yourself a siphon. This is not proof that atmospheric pressure is the driving force in a siphon but rather a special case of atmospheric pressure being used to prime a siphon. Once the siphon has been reprimed it operates as per normal through molecular cohesion.
    Since atmospheric pressure increases closer to the ground, atmospheric pressure actually slightly retards the flow of water out of a siphon,. However, the effect is very small – the retarding effect of the atmosphere in a siphon with a height difference of one metre (the difference between the water level in the upper reservoir and the siphon outlet) is equivalent to the height of the siphon being less by just one mm. The effect of the atmosphere would of course not be noticed when using a siphon to empty a fish tank, for example. However, an important point is that in normal use the atmosphere is slightly retarding and therefore does not impel the water even slightly.
    If the pressure of the air above the upper reservoir of a siphon is increased (e.g. by blowing into an airtight container enclosing the reservoir) flow out of the end of the siphon is increased. If air pressure above the upper reservoir is reduced flow decreases. It follows that there must be a pressure just below the pressure that increases flow and just above the pressure that decreases flow that neither increases or decreases flow. This pressure is very close to atmospheric pressure – in fact very slightly above to overcome the retarding influence of the atmosphere as discussed above. So, when atmospheric pressure is the same at both ends of a siphon, the rate of flow only depends on gravity and not the difference in atmospheric pressure.
    If the above experiment is repeated at a lower atmospheric pressure (e.g. by going up a mountain), the same effect will be observed. There will be a pressure that neither enhances nor impedes flow, which will always be very close to the ambient pressure. This experiment can be repeated all the way up to the top of a mountain and on in to space, i.e. down to the zero pressure of a vacuum, hence reports in the scientific literature describing vacuum siphons.
    It might be argued that although molecular cohesion is operational in a vacuum siphon, this is not the case for siphons operating at normal atmospheric pressure. However, it is very difficult to see why molecular cohesion should enable siphons to operate in a vacuum but not at normal atmospheric pressure. What causes the links between water molecules to break at higher pressure so there is a need to invoke atmospheric pressure to push water into a siphon?
    Since publication of the original siphon paper I have received numerous emails from around the world with information about the definition of the siphon in non-English language dictionaries. Thus far is seems that every non-English dictionary has an incorrect definition of the siphon with the possible exception of Dutch dictionaries.
    A valid question is does it matter if dictionaries contain an incorrect definition of the siphon? Maybe it does. Correcting dictionary definitions of the siphon could be a first step in making people more aware of how useful siphons can be in everyday life – e.g. for draining floodwaters, relieving the pressure on dam walls, irrigation, transporting drinking water etc. The fact that gravity is being used means that no powered pumps are required. Power is often not available in flooded regions, which is another advantage of siphons. Increased worldwide knowledge of siphons work could assist individuals and communities mitigate the effects of climate change.
    To summarise, the possible role of atmospheric pressure in priming a siphon is not a sufficient reason for saying that siphons operate through atmospheric pressure rather than gravity. It is also incorrect to say that both atmospheric pressure and gravity are required for a siphon to operate. Therefore to avoid confusion only gravity should be referred to in the definition of a siphon. I hope that the OED as clearly most prestigious and influential dictionary in the world can lead the way in correcting the definition of the siphon.

  11. F Mahony says:

    Unfortunately, Dr Hughes’s siphon paper was not adequately researched and is lacking in appropriate tests. This resulted in a number of wrong conclusions drawn.

    A more thorough research by Dr Hughes, including reviewing technical manuals on hydrostatics, hydraulics and centrifugal pump suction design and performance would have provided him with a far better understanding of siphons. Studying the history of the work and tests conducted by Torricelli and Pascal would have been beneficial.

    For those that make a comparison between atmospheric or fluid pressure at the entrance and outlet, then concluding that they are either the same, or slightly higher at the outlet, yes the conclusion is correct. Except, it is the wrong comparison to make. Of more interest is the pressure comparison between the inlet, and the highest point of the siphon.

    Gauge pressure at the inlet is usually close to zero. This means an absolute pressure of approx 10 metres head. Vacuum gauge pressure at the top of the siphon can be up to approx negative 10 metres. This is equal to an absolute pressure of close to zero. Provided the difference in the absolute pressures at the inlet and the top of the siphon, less the static head from inlet water level to the top of the siphon is greater than zero, then the water will flow from the inlet to the top of the siphon.

    Dr Hughes claims that the maximum height of a water siphon actually depends on the tensile strength of water, and that it is not dependent on atmospheric pressure. Yet no tests were conducted to prove this point. And a simple test would have been required to prove his theory incorrect, that being determine the maximum height of a siphon at altitude.

    Extracts From PIA Australian Pipe Friction Handbook
    2.15 Discharge of Siphon Lines

    Theoretically, the limit of possible flow up the inlet leg of a siphon is reached when the pressure Ps at the highest point in the system is absolute zero (-10.35 m of water). In practice, however, this limit would be less because when the absolute pressure Ps at the summit becomes equal to the vapour pressure of the liquid, cavitation begins. Consequently, vapour collects at the bend forming so called vapour locks and the flow stops. Hence, for satisfactory operation of a siphon, the maximum static suction lift allowable can be expressed by the following equation:

    Static suction lift in metres < 0.102 x (Pa – Pv) – v squared/2g x (1 + Ks) – S

    Pa = atmospheric pressure in kpa
    Pv = vapour pressure of liquid in kpa
    Ks = the sum of resistance coefficients for the loss of head from inlet to the highest point of the siphon
    S = a safety margin – a deduction of 1 to 2 m depending of the pipeline installation.
    v = liquid velocity in m/s
    g = gravitational constant = 9.8
    When a siphon is nearing it’s maximum height, the flow rate will be lower, along with the velocity and friction losses.

    At sea level, International standard atmospheric pressure is 101.325 kpa. Water at 20 degrees has a vapour pressure of 2.337kpa.

    That means we can simplify the above formula to:

    * Sea Level: Static suction lift in metres < 0.102 x (101.325 – 2.337) < 10.09 metres
    1000m ASL: Static suction lift in metres < 0.102 x (89.88 – 2.337) < 8.92 metres
    2000m ASL: Static suction lift in metres < 0.102 x (79.50 – 2.337) < 7.87 metres
    * 2250m ASL: Static suction lift in metres < 0.102 x (77.07 – 2.337) < 7.62 metres
    3000m ASL: Static suction lift in metres < 0.102 x (70.12 – 2.337) < 6.91 metres
    3500m ASL: Static suction lift in metres < 0.102 x (65.78 – 2.337) < 6.47 metres

    So, if a test to see the maximum siphon level was conducted at sea level, and another at the highest point in Australia, Mt Kosciusko, the results would have been approx 10 and 7.6 metres respectively. Why the difference? Bloody atmospheric pressure. We need it to survive, but it gets in the way of what we thought was a breakthrough theory.

    All that is required is a few listeners who live at altitude, to test a siphon to determine the maximum height they can achieve. And report their findings.

    The result: you will find indeed that atmospheric pressure plays a part. Yes, so too does gravity. However, gravity should not replace atmospheric pressure, as Dr Hughes would claim, rather gravity should be added as well as atmospheric pressure. Of course, we could talk hydrostatics, but that is not a layman’s language that will easily help understand how siphons work.

    In relation to how a siphon works, there is only one compelling model, and a separate misguided theory proposed by Dr Hughes that involves tensile strength and hydrogen bonds. Yes, gravity and the height difference determines flow rate, except of course when the attempted rise exceeds atmospheric pressure, and you get no flow at all.

    Water droplets: Notice how they can only support about 4mm of their own weight. And slowly overfill a glass, it will rise about 4mm over the top before overflowing. Mmm, 4mm, so much for tensile strength!

    Cavitation: When the local pressure falls below the vapour pressure of the liquid, it turns to vapour. When pressure is restored, the vapour bubble collapses, and water surrounding the bubble causes the damage. It has nothing to do with ripping water apart.

    As to why it is difficult to pull two panes of glass apart, you are trying to create a vacuum. It has nothing to do with tensile strength of water. Do some research on Otto von Guericke and the vacuumed sphere that the horses couldn’t pull apart.

  12. Stephen Hughes says:

    Since my last posting on this site I have been doing further work on resolving the mystery of the siphon. My most recent publication on the subject is called ‘The Secret Siphon’ Stephen W Hughes 2011 Phys. Educ. 46 298 doi: 10.1088/0031-9120/46/3/007 (I chose the title in view of the fact that nature is proving elusive in revealing the modus operandi of the siphon). The paper can be downloaded either from the Institute of Physics website (using the doi above) or from the Queensland University of Technology eprints site (http://eprint.qut.edu.au/41923/.)

    In this paper I demonstrate that a regular everyday siphon can be constructed with the upper reservoir level held constant. In the siphon illustrated in the paper I do this by feeding water into the upper reservoir of a siphon from a laboratory tap. If the siphon outflow is slightly higher than the tap inflow, the water level in the upper reservoir will drop, reducing the siphon height and therefore increasing siphon flow until it becomes equal to the inflow. A similar argument applies when the inflow is lower than the outflow. There is a natural feedback mechanism ensuring that the siphon outflow closely matches the inflow. This sort of siphon can easily be constructed in the kitchen or classroom. Another way of constructing a constant level siphon is to connect a pump between the lower and upper reservoirs. As long as the initial siphon and pump flows are similar, the feedback mechanism will ensure the upper reservoir level is held steady (assuming the pump flow is steady).

    In this scenario, the surface of the water in the upper reservoir is analogous to the front face of a cylinder in an internal combustion engine. If a combustion engine seizes up for any reason, hot gas produced by exploding fuel cannot move the cylinders and the car will not move. In a constant level siphon, if the upper reservoir level is stationary it cannot be that the atmosphere is pushing water into the siphon. If it is not pushing then it cannot be imparting any energy to the siphon, and therefore the propulsive force in a siphon is gravity.

    I would be particularly interested to hear from anyone who can demonstrate that the atmosphere is pushing water onto a siphon even when the upper reservoir is held at a constant level. If no atmospheric explanation is forthcoming it means that all dictionary definitions of the siphon that say that atmospheric pressure pushes water into a siphon should be corrected. However, it could be argued that in siphons where the upper reservoir level falls, the atmosphere is pushing water into the siphon, but when the upper reservoir is held at a constant level the siphon reverts to cohesion. How do the hydrogen bonds know when to break and reconnect?

    Intriguingly, in the same issue of the journal Physics Education a paper has published on trees (Water transport in trees – an artificial laboratory tree, K Susman, N Razpet and M Cepi, 2011 Phys. Educ. 46 340 doi: 10.1088/0031-9120/46/3/015 ), suggesting that cohesion between water molecules is the only way of explaining how water can travel from the roots to the leaves of trees taller than 10 m – the height of a column of water than can be supported by the atmosphere. However, from the literature cited within the trees paper it is apparent that this is also controversial topic within the field of plant physiology.

    I am currently doing further experiments in an attempt to resolve the issue of whether atmospheric pressure is required to support the ascending column of water in a siphon, in other words is a siphon effectively like two back-to-back barometers with one side longer than the other? Another issue is whether the atmosphere presses on both sides of the water in a siphon so that it acts cohesively (an analogy to this is an elastic band holding a line of wooden blocks together so that they move as a whole), or whether water is held together via hydrogen bonds (in analogy like a stack of lego blocks) and that atmospheric pressure prevents water boiling and the chain of water molecules breaking. I hope to report on this work (via the peer review literature) by the end of the year.

  13. F Mahony says:

    Dr Hughes: You wrote:

    “If the siphon outflow is slightly higher than the tap inflow, the water level in the upper reservoir will drop, reducing the siphon height and therefore increasing siphon flow until it becomes equal to the inflow.”

    What exactly are you on? The siphon flow would do the exact opposite to what you claim. You claim it will increase. The siphon flow will actually decrease as the water level drops.

    And the changing of the water level in the upper reservoir has no relevance to the principles of why a siphon works.

    At sea level, the upper reservoir is simply in effect, a pressurised reservoir of water. Pressurised to approx 10 metres head pressure. Give that water an opportunity to escape, like to a lower pressure area, and it will do so.

    Sucking on a straw creates a lower pressure area, and what happens, the water goes from the higher pressure area to the lower pressure area. Same thing with a siphon.

    Water flowing on the downleg, reduces the pressure at the top of the siphon. Fit a vacuum gauge to the top of the siphon and you will see it with your own eyes. And thus in a siphon, like sucking on a straw, water moves from the higher pressure area to the lower one. And not a hydrogen bond in site!

  14. Stephen Hughes says:

    I made a mistake in my previous comment in saying that siphon flow would increase when the siphon flow was greater than the inflow. I should have said that the flow would decrease until siphon the flow became equal to the inflow. (I checked my publication in Physics Education ‘The Secret Siphon’ and thankfully this does not have the same mistake!). I agree that the atmosphere acts like a pressure vessel. The water in the reservoir has a pressured fluid above it, i.e. the air, at a pressure of 100,000 Nm-2 at sea level. However, it must be noted that this pressure is applied to both sides of the siphon. This is easier to appreciate if the siphon outflow is immersed in water. In this case siphon flow depends on the difference in height between the upper and lower reservoir levels, and is independent of the depth of immersion at either end.
    The pressure along the length of the siphon tube varies and has a hydrostatic component and a hydrodynamic component. The hydrostatic component is dependent on the height (h) above the upper reservoir and is given by the expression dgh where d is the density of water and g the acceleration due to gravity (9.8 ms-2). The hydrodynamic component is the same at all points along the length of the siphon tube and is given by the expression 1/2dv2 where v is the average velocity in the tube, assuming that the flow rate is constant and cross sectional area is also constant along the length of the tube. In a siphon, the minimum pressure will always be at the highest point and the pressure above the upper and lower reservoirs is at atmospheric pressure. Therefore the pressure profile along the length of a static and flowing siphon is symmetric with equal atmospheric pressure on either side. Since air pushes equally on both sides of the siphon, atmosphere pressure does not push water into a siphon.

  15. F Mahony says:


    It is good to see you acknowledge that water at both ends of the siphon are, in effect, pressurised reservoirs of water. (Now give some thought to how you can have water under compression, then be claiming that it is held together in a siphon by being under tension. i.e if we are pushing something together, how can we be pulling it apart at the same time.)

    I assume you also realize that siphon flow does NOT “just depend on the difference in height between the upper and lower reservoir levels”. The siphon pipe diameter and interior pipe roughness will also have an affect. So too will atmospheric pressure. (For atmospheric pressure, the key issue is whether we get flow, the actual flow rate formula is a seperate issue, however if you don’t have flow, then you obviously have an issue with flow rate)

    Build 2 siphons. One with a vertical height from upper reservoir to siphon crest of 5 metres, the second with a vertical height from upper reservoir to siphon crest of 15 metres. At sea level, the 5 metre one can operate, the 15 metre one will not.

    Now put both of them into a sealed chamber, and reduce the air pressure to 1 metre absolute. Neither will operate.

    Leaving them both still in the sealed chamber, return the pressure to atmospheric pressure, then pressurise the chamber to 20 metres absolute. Both will operate.

    What has changed? Gravity? No. Siphon pipe size? No. Pipe roughness? No. Air pressure above the water reservoirs? Yes.

    What is critical in getting the siphon to operate in the above tests is that air pressure must be greater than the siphon upper reservoir to crest height. In fact, I can build a siphon with a upper reservoir to crest height of 100 metres, and get that siphon to operate if I place it in a chamber, and increase the air pressure to greater than 100 metres absolute.

    In all the above tests, we have the same air pressure on both the upper and lower reservoirs (ignoring the very minor difference due to the reservoir heights).

    So does atmospheric pressure provide a source of free energy, i.e. like a perpetual motion energy source that operates a siphon. Of course not.

    As noted in my earlier post, water simply moves from the higher pressure area to the lower one. And both gravity and atmospheric pressure are key ingredients in the operation of a siphon.

Leave a Reply