Siphon Experiment for Kids: Water Climbs, Then Falls

Water going uphill on its own looks like a rule being broken. Fill a tube, drop one end into a full jug and let the other end dangle lower, and water climbs over the rim and keeps flowing until the jug is empty. The usual explanation, that air pressure pushes the water up, is so widely believed that the Oxford English Dictionary printed it for ninety-nine years before a physicist in Australia pointed out the error in 2010. A siphon experiment for kids is worth running for the demonstration, and worth reading about for the correction: what drives a siphon is gravity, and atmospheric pressure has a different job entirely.

Key Takeaways

• You’ll start a siphon between two containers, measure how the flow rate changes with height difference, and find the point where it stops.
• It teaches that flow is driven by the weight difference between two water columns, that speed follows roughly the square root of the drop, and what atmospheric pressure actually contributes.
• Best fit is ages 8 to 14. The measurement half needs a kid who can read a stopwatch and a measuring jug.
• The usual failure is air in the tube. One bubble at the high point will stop everything.
• Never start a siphon with your mouth, and never siphon anything other than clean water.

What You’ll Need for This Siphon Experiment for Kids

Item	How much	Where to get it	If you don’t have it
Clear flexible tubing, 8–10 mm inside diameter	1.5 m	Hardware shop or aquarium shop, about $4	Aquarium airline is too narrow; look for the thicker stuff
Jugs or buckets, clear if possible	2	Kitchen	One can be the sink; you need a measurable one at the outlet
Measuring jug, 1 litre with markings	1	Kitchen	A container you have calibrated with a known cup
Stopwatch or phone timer	1	Pocket	Count seconds aloud
Books or a crate to change heights	A stack	Around the house	A chair and a worktop
Tape measure	1	Drawer	A metre rule
Food colouring	2 drops	Baking cupboard	Optional, but it makes the flow visible in the tube
Towel	1	Kitchen	This will spill at least once

Time: prep 10 min, experiment 35 min, cleanup 10 min Cost: around $4 for tubing; everything else is in the house Mess level: do this at the sink or outside

Do the Experiment

Step 1 — Start the siphon without using your mouth

Fill the upper jug with water and two drops of colouring. Submerge the whole tube in it so it fills completely with no bubbles, then pinch one end shut with a finger, lift that end out, keep it below the water level in the jug, and release it over the lower container.

What you should see: water runs out immediately and keeps running. If it stops after a second, there was air in the tube. The submerge-and-pinch method works first time far more often than any other, which is why it is Step 1.

Step 2 — Check the rule about levels

With the siphon running, lift the lower container slowly upward until its water surface is level with the upper jug’s surface.

What you should see: the flow slows as the surfaces get closer, stops dead when they are level, and reverses if you keep lifting. Nothing about the tube changed. Only the difference between the two water surfaces changed, and that difference is the whole engine.

Step 3 — Measure flow rate against height difference

Set the outlet 10 cm below the upper surface and time how long it takes to collect 500 ml. Repeat at 20, 30 and 40 cm. Keep the tube and the upper water level the same each time.

What you should see: faster at bigger drops, but not proportionally. Going from 10 cm to 40 cm, a fourfold increase in drop, roughly doubles the flow rate rather than quadrupling it. That square-root behaviour is the measurable result and it has a name.

Step 4 — Raise the top of the loop and see what happens

Keeping the inlet and outlet where they are, lift the middle of the tube higher and higher so the water has further to climb before it falls.

What you should see: the flow barely changes. The height of the loop is almost irrelevant, which surprises nearly everyone. What matters is the inlet surface and the outlet, not the route in between. Push it high enough, metres rather than centimetres, and it will eventually fail, which is explained below.

Step 5 — Break it on purpose, three ways

Lift the inlet end out of the water. Then restart and pinch the tube flat at the top. Then restart and let the outlet end rise above the upper surface.

What you should see: three different stoppages. Removing the inlet lets air in and the column drains out both ways. Pinching stops flow and it resumes when you release, provided no air got in. Raising the outlet reverses the flow. Each failure identifies a different part of the mechanism, which is more informative than watching it work.

What’s Actually Going On

Think of the water in the tube as two columns joined at the top. On the inlet side, the column reaches from the upper water surface to the top of the loop. On the outlet side, it reaches from the top of the loop down to the outlet. The outlet column is longer, so it weighs more, so the water as a whole is pulled toward the outlet. That net weight difference is the driving force, and it depends only on the vertical distance between the upper surface and the outlet. This is why the loop height in Step 4 made almost no difference: it adds the same amount to both columns.

The flow speed follows from energy. Water starting at rest at the top surface and arriving at the outlet a height h lower has converted gravitational potential into kinetic energy, so the ideal exit speed is v = √(2gh), which is Torricelli’s result. Square-root dependence is exactly what Step 3 should show: quadruple the drop and you double the speed. Real tubing loses some of that to friction along its length, so your numbers will undershoot the ideal, and that gap is itself worth noticing. HyperPhysics’s fluid pages carry the derivation.

So what is atmospheric pressure doing? Not driving the flow. It pushes down on both water surfaces, and since those pushes very nearly cancel, it cannot be the engine. Its job is to keep the water column intact. At the top of the loop the pressure is lower than atmospheric, by roughly the weight of the water column beneath it, and if the loop is high enough that the absolute pressure there falls to water’s vapour pressure, the water boils at room temperature, a bubble forms and the siphon breaks. At sea level that limit sits near 10.3 metres, which is the same barometric height that sets how tall a mercury barometer must be. NOAA’s JetStream page on air pressure gives the sea-level reference figure, and the NIST Chemistry WebBook gives the vapour pressure that sets the cutoff.

This distinction is not pedantry, and it has a history. In 2010 Stephen Hughes published a note in Physics Education pointing out that the Oxford English Dictionary had been defining a siphon as operating by atmospheric pressure since 1911, and that the definition was wrong. Richert and Binder followed up in The Physics Teacher in 2011, comparing the “chain of water” picture with the pressure picture and setting out where each is useful. Then in 2015 Boatwright, Hughes and Barry published a genuinely startling result in Scientific Reports: using degassed water in a reduced-pressure chamber, they ran a siphon 15 metres high, well above the barometric limit, because without dissolved gas to nucleate bubbles the water’s own cohesion held the column together under tension. So the honest version is layered. Gravity drives it. Atmospheric pressure sets the practical height limit for ordinary tap water. Liquid cohesion can beat that limit if you remove the dissolved gas.

How to Explain It to Your Kid

Ages 5–8: A chain hanging over a bar

Imagine a chain lying over a broom handle, with more chain hanging down one side than the other. The longer side is heavier, so it pulls the whole chain over. Water in the tube does the same thing, because the water is stuck to itself and behaves like a chain.

Then let them do Step 2 with their own hands. Raising and lowering the end to speed it up, slow it down, stop it and reverse it is the most satisfying part of this for a small child, and it teaches the rule without a word of theory.

Ages 9–12: It is the difference that matters

Make the point carefully: the only number that controls the flow is how far the outlet sits below the water surface in the jug. Not the length of the tube, not the height of the loop, not how much water is in the jug. Get them to predict Step 4 before doing it; most kids expect a tall loop to slow things down, and being wrong on paper is the useful bit. Our note on three holes in a bottle measures the same square-root law a different way.

Ages 13+: Test Torricelli and then look up the 2015 paper

Give them v = √(2gh) and have them convert flow rate in millilitres per second into exit speed using the tube’s cross-sectional area. Plot measured speed against √h and fit a line; the slope should come out near √(2g), and the shortfall is friction. Then set the reading: Hughes’s 2010 note, and the Boatwright 2015 result. A 15 metre siphon that works because water can be put under tension is a good antidote to the idea that school physics is settled and dull. Pair it with building a barometer, which measures the very pressure that sets the limit.

The question to ask: “If we did this with a liquid twice as dense as water, would the siphon run faster, slower, or the same?”

Why Teaching This Early Pays Off

Siphons are everywhere once you look: the trap under every sink, the flush mechanism in many toilets, aquarium cleaning, draining a pond, irrigation channels, the way fuel is moved between tanks and the way a hummingbird does not drink. Understanding that the controlling variable is a height difference, not a tube shape, makes all of those predictable rather than magical.

The deeper payoff is about the status of textbook explanations. This is a case where a reference work the size and age of the OED carried a wrong physical explanation for a century, was corrected by a short paper, and then the whole question was reopened by a result in 2015. That is what live science looks like, and a child who has met one example of it is better placed to understand that “it’s in a book” is not the end of an argument.

Keep the learning claim honest. One session does not change a trajectory, and a demonstration that looks effortless can be deceptive: Deslauriers and colleagues found in 2019 that learners consistently feel they have learned less from active methods while actually learning more, so a session where your kid spends twenty minutes failing to get the bubble out of a tube may be doing more than it appears. The National Research Council’s framework names planning and carrying out investigations as a core practice, and Step 5, breaking the thing deliberately, is a real example.

Before You Start: What Parents Should Know

Best age	8 to 14. Younger children enjoy Step 2 enormously but cannot manage the timing and the arithmetic.
Supervision	Light, but an adult should be the one who says the rule about mouths and about what may be siphoned.
Real hazards	None beyond spills and a slippery floor. Never start a siphon by sucking on the tube: it is how people swallow whatever is in the container, and it is the single reason this activity has an adult rule attached. Clean water only.
Don’t	Don’t siphon fuel, antifreeze, cleaning products, pool chemicals or anything from an unlabelled container, now or ever. Don’t use the kitchen’s drinking-water tubing afterwards for food. Don’t leave a running siphon unattended near electrical sockets.

When It Doesn’t Work

• Flow stops within a second. Air in the tube at the high point. Refill by fully submerging the tube and pinching both ends, and check the tube is not kinked.
• It never starts at all. The outlet is not below the upper water surface. This is the most common error and it is geometric, not technical. Measure it.
• Flow is a dribble. Your tube is too narrow, or it is squeezed somewhere along its length, or the height difference is tiny. An 8 mm bore and a 30 cm drop gives a satisfying stream.
• It works, then slows and stops on its own. Normal and correct: the upper level is falling while the lower one rises, so the driving height difference is shrinking. It will stop when they are level or when the inlet breaks surface.
• A bubble appears at the top and grows. Either the tube is leaking at a joint, or the loop is too high for your water. Lower the loop and check for pinholes by running a tap through the tube.

Take It Further

Build the Torricelli graph. Measure flow rate at six drop heights from 5 cm to 60 cm, three trials each. Convert to exit speed, plot against the square root of height, and fit a straight line. The deviation at the smallest heights is where tube friction dominates, and spotting that is a genuine piece of data interpretation.

Change the tube, keep the heights. Run the same 30 cm drop with a 6 mm tube, an 8 mm tube and a 12 mm tube. Flow rate should scale faster than the cross-sectional area, because wider tubes lose proportionally less to wall friction. That result surprises people and it is the reason plumbers care about pipe diameter so much.

What to Watch For Over the Next 3 Months

• Week 4: they explain the U-bend under the sink, or spot a siphon in a toilet cistern or an aquarium.
• Month 2: “it depends on the height difference” appears in an unrelated context, often about water pressure in a shower on an upper floor.
• Month 3: they ask how high a siphon can go, which is the question that leads to barometric height, vapour pressure and the 2015 experiment. That is a very good place to arrive.

Frequently Asked Questions

Does a siphon work because of air pressure?

Not as the driving force. Air pressure pushes down on both water surfaces and very nearly cancels out. The flow is driven by the weight difference between the two columns of water in the tube. Air pressure does set a practical height limit, around 10.3 metres at sea level, by determining when the water column breaks.

Would a siphon work in a vacuum?

With ordinary tap water, no, because dissolved gas comes out of solution and breaks the column. With carefully degassed water, yes: Boatwright, Hughes and Barry demonstrated exactly this in 2015, running a siphon above the barometric limit because the water’s cohesion held it together under tension.

Why doesn’t the height of the loop matter?

Because raising the loop adds the same length of water to the rising side and the falling side, so the weight difference between them is unchanged. Only the vertical distance from the source surface to the outlet matters, which Step 4 lets you verify in about two minutes.

Can my child start it by sucking the tube?

No, and this is the one firm rule here. It is the standard way people end up swallowing whatever was in the container, and the habit transfers to situations where the liquid is fuel or a cleaning product. Fill the tube by submerging it instead; it is also more reliable.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Hughes, S. W. (2010). “A practical example of a siphon at work.” Physics Education, 45(2), 162–166. doi:10.1088/0031-9120/45/2/006
2. Richert, A., & Binder, P.-M. (2011). “Siphons, revisited.” The Physics Teacher, 49(2), 78–80. doi:10.1119/1.3543576
3. Boatwright, A., Hughes, S., & Barry, J. (2015). “The height limit of a siphon.” Scientific Reports, 5, 16790. doi:10.1038/srep16790
4. National Oceanic and Atmospheric Administration, National Weather Service. “Air Pressure.” JetStream. https://www.noaa.gov/jetstream/atmosphere/air-pressure
5. Georgia State University. “Fluid Concepts.” HyperPhysics. http://hyperphysics.phy-astr.gsu.edu/hbase/fluid.html
6. National Institute of Standards and Technology. “Thermophysical Properties of Fluid Systems.” NIST Chemistry WebBook. https://webbook.nist.gov/chemistry/fluid/
7. Deslauriers, L., McCarty, L. S., Miller, K., Callaghan, K., & Kestin, G. (2019). “Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom.” PNAS, 116(39), 19251–19257. doi:10.1073/pnas.1821936116
8. National Research Council. (2012). A Framework for K-12 Science Education. National Academies Press. doi:10.17226/13165