ARCHETYPE: U-Tubes, Manometers & Barometers

In 1643, Evangelista Torricelli filled a metre-long glass tube with mercury, capped it with a thumb, inverted it into a dish, and let go. The mercury slumped — but only to a height of about 76 centimetres, leaving an empty space above it that scandalized the philosophers (nature was supposed to abhor a vacuum). Torricelli's reading of the experiment created two instruments at once: the column height measures the weight of the atmosphere pressing on the dish — a barometer — and, more generally, a column of liquid in a bent tube can weigh any pressure difference against gravity. Bend the tube into a U, and you have the manometer: the simplest pressure gauge ever built, and the archetype this article drills. Every U-tube problem — and exams pose them in a dozen costumes — is solved by a single master move: in a connected body of one fluid at rest, all points at the same horizontal level have the same pressure. Start at an end where you know the pressure, walk through the tube adding $\rho g h$ going down and subtracting it going up, and read off the unknown at the far end. That's the whole algorithm. The costumes: open manometers reading gas pressures above or below atmospheric, two immiscible liquids balancing in a U, barometers healthy and faulty, differential gauges between two vessels. CBSE and NEET love the direct readings and the two-liquid balance; JEE-Main tilts the tube and traps air in barometers; AP's fluids unit makes gauge-versus-absolute the explicit learning objective. Board fills first, then the competitive escalations.

Key takeaways

Master move: same level + same connected fluid at rest ⇒ same pressure. Walk the tube: down adds $\rho gh$ , up subtracts it.
Open manometer: $P_{\text{gas}} = P_0 + \rho g h$ if the open arm rides higher (gas beats atmosphere); $P_{\text{gas}} = P_0 - \rho gh$ if it rides lower.
The manometer directly reads gauge pressure ( $\rho gh$ ); add $P_0$ for absolute — AP's favourite distinction.
Two immiscible liquids balanced on a common interface level: $\rho_1 h_1 = \rho_2 h_2$ — heights inversely proportional to densities.
Barometer: $P_0 = \rho_{\text{Hg}}\, g\, h$ with vacuum on top; 760 mm Hg = $1.013\times10^5$ Pa. Trapped air makes it read LOW by the trapped air's pressure.
Tube width, tilt, and shape are irrelevant — only vertical heights count (a tilted column reads longer along the glass, same vertical height).
Key idea: a liquid column converts a pressure difference into centimetres — walk the tube level by level and the answer falls out.

The core insight

Every U-tube, manometer, and barometer problem is the equal-pressure-at-equal-level rule executed as a walk through the tube — add $\rho gh$ descending, subtract it ascending, equate both routes at a common level in the same continuous fluid.

Where this lives today

Medicine still speaks this article's language. Blood pressure — 120/80 mm Hg — is quoted in Torricelli's column units a century after mercury cuffs began retiring (modern digital cuffs are calibrated to agree with the mercury column); ICU central-venous lines are measured in centimetres of water, because a low-density column amplifies faint pressures, exactly the gauge-liquid logic above. Aviation flies on the barometer: altimeters convert outside pressure to altitude, pilots reset the reference ( $P_0$ dialled in as "QNH") before landing, and cabin pressurization targets are set in column-derived units. HVAC engineers balance cleanroom and operating-theatre pressures with inclined water manometers — tilting the tube stretches millimetres of height into centimetres of scale, the $h/\cos\theta$ geometry used for sensitivity instead of against you. And weather maps' isobars are barometer readings, gridded: Torricelli's flickering column, now assimilated into supercomputer forecasts every six hours.

The setup

The instrument: a U-shaped tube holding a dense liquid (mercury for strong pressures, water or oil for gentle ones), one arm connected to the pressure you want (a gas bulb, a pipe, the atmosphere), the other arm open, closed, or connected somewhere else. The liquid shifts until the pressures at its two surfaces are held up consistently — and the height difference $h$ between the two arms becomes your reading.

Figure

The open-tube manometer — the height difference IS the gauge pressure

Why the master move works is last article's physics: within one continuous fluid at rest, two points at one level must share one pressure, else the fluid between them would accelerate. The two warnings that generate every trick question: the rule holds only within the same fluid — you may not equate levels across an interface where the liquid changes — and only at rest (the accelerating-container archetype breaks it deliberately).

The master solution pattern

Formalize the walk. Start at a point of known pressure (an open surface: $P_0$ ; a labelled gas: $P_{\text{gas}}$ ). Move through the liquid to the target:

P_{\text{end}} = P_{\text{start}} + \sum \rho g\,(\text{descents}) - \sum \rho g\,(\text{ascents}) \qquad (1)

where: each term uses the density of the liquid you are walking through and the vertical drop or climb — never distances along a slanted tube. Horizontal moves cost nothing. The manometer reading follows for the standard case (open arm higher by $h$ ):

P_{\text{gas}} = P_0 + \rho g h \quad\Leftrightarrow\quad P_{\text{gauge}} = \rho g h \qquad (2)

where: $\rho gh$ — the gauge pressure, which is what the instrument physically displays; if the open arm rides lower, the gas is below atmospheric and $P_{\text{gauge}} = -\rho g h$ . A manometer cannot tell you $P_0$ ; it weighs only the difference — the reason gas-law problems must add the barometer's $P_0$ by hand, and the reason AP writes gauge-vs-absolute as its own learning objective.

Derivation AThe two-liquid U-tube — heights inversely proportional to densities

We derive the balance rule for a U-tube holding two immiscible liquids (say water over one arm, oil in the other), open at both ends. This is the archetype's second-most-tested costume and the cleanest demonstration of where the reference level MUST be placed: at the interface between the liquids — the lowest level up to which a single continuous fluid fills both arms. Strategy: equate pressure at the interface level in both arms, then cancel. Watch what happens if you naively pick a higher level: the two arms hold different liquids there, and the rule is void.

Step 1. Set the reference at the lower interface — the level where liquid 2 (denser, filling the U's bottom) meets liquid 1 in its arm. Below this level, one continuous liquid (liquid 2) connects both arms: the rule applies.

Step 2. Left arm (holding column of liquid 1, height $h_1$ above the reference): walking down from the open top,

P_{\text{ref,left}} = P_0 + \rho_1 g h_1

Step 3. Right arm (liquid 2 standing $h_2$ above the reference): walking down,

P_{\text{ref,right}} = P_0 + \rho_2 g h_2

Step 4. Equate (same level, same connected fluid below):

\rho_1 h_1 = \rho_2 h_2 \quad\Rightarrow\quad \frac{h_1}{h_2} = \frac{\rho_2}{\rho_1}

Why the lighter liquid stands taller: it takes a longer column of the light liquid to press as hard as a short column of the dense one — 13.6 cm of water balances 1.0 cm of mercury. *Why $P_0$ and $g$ vanished: both arms are open to the same atmosphere and share the same $g$ — the balance is pure density-times-height. (If one arm is closed or pressurized, keep those terms; the walk, not the memorized ratio, is the safe tool.) Verified: units cancel to a pure ratio ✓; limit $\rho_1 = \rho_2$ gives $h_1 = h_2$ — one liquid, level arms ✓; anchor: water vs mercury = 13.6:1 by inverse densities ✓. Standard: NCERT Ch. 9 examples; ISC board sets.*

The barometer is the manometer with the atmosphere on one arm and nothing on the other. Fill a tube, invert it into a dish: mercury falls until the column's weight matches what the atmosphere can hold up, leaving Torricellian vacuum above (a trace of mercury vapour — negligible). Walking from the dish's surface ( $P_0$ ) to the vacuum ( $\approx 0$ ):

P_0 = \rho_{\text{Hg}}\, g\, h \qquad (3)

where: $h \approx 0.760$ m at sea level: $13600\times9.8\times0.760 = 1.013\times10^5$ Pa. Mercury's 13.6 is the whole point — a water barometer would need a tube over 10 m tall (Pascal actually built one with wine, to a neighbourhood's delight). Two immunities make barometer MCQs: widening the tube changes nothing (heights, not volumes), and tilting the tube changes nothing — mercury slides along the glass until its vertical height is again $h$ ; the column along the tube gets longer, the reading doesn't. The classic fault: air trapped above the mercury. That air pushes down with its own pressure $P_{\text{air}}$ , so the column stands short: the instrument reads $h' = (P_0 - P_{\text{air}})/\rho g$ — always low, by exactly the trapped air's pressure (worked example 4).

Figure

Two-liquid balance and the barometer — the other two costumes

Limiting cases

case	result	meaning
both arms open, one liquid	$h = 0$ : levels equal	no pressure difference to weigh
open arm higher by $h$	gas above atmospheric by $\rho gh$	the standard reading
open arm lower by $h$	gas below atmospheric	negative gauge; suction side
lighter gauge liquid	bigger $h$ per Pa	water manometers for gentle pressures (breath ~ cm of water)
barometer tilted	same vertical $h$ , longer mercury thread	vertical height is the only currency
trapped air in barometer	reads low by $P_{\text{air}}/\rho g$	the "faulty barometer" JEE family
container accelerating	equal-level rule FAILS	handled in the accelerating-container archetype

Common trap“Same level means same pressure — so I can equate the two arms anywhere”

Only within one continuous fluid. The equal-level rule is legal exactly up to the highest level at which a single unbroken liquid connects the two points. In a two-liquid U-tube, that ceiling is the lower interface: one centimetre higher, the left arm holds oil and the right arm holds water, and equating pressures there silently assumes $\rho_{\text{oil}} = \rho_{\text{water}}$ — the exact error the question is engineered to catch. The same discipline kills the "hop across the vacuum" slip in barometer problems and the "equate across the gas bubble" slip in trapped-gas tubes: a gas pocket transmits one pressure to both its ends (its columns are weightless at exam precision) but breaks the liquid's continuity. Rule of thumb: draw the reference line at the lowest interface, and walk — never leap.

Worked examples

1 · Board level — reading an open manometer (CBSE/AP pattern, 3 marks)

A mercury manometer is connected to a gas cylinder. The mercury in the open arm stands $20\ \mathrm{cm}$ higher than in the arm connected to the gas. Atmospheric pressure is $1.01\times10^5\ \mathrm{Pa}$ ; $\rho_{\text{Hg}} = 13600\ \mathrm{kg/m^3}$ , $g = 9.8\ \mathrm{m/s^2}$ . Find the gauge pressure and the absolute pressure of the gas, in Pa and in mm of mercury.

Step 1. Open arm higher ⇒ the gas is pushing mercury up that side ⇒ gas above atmospheric.

Step 2. Gauge pressure: $\rho gh = 13600\times9.8\times0.20 = \boxed{2.67\times10^4\ \mathrm{Pa}}$ — equivalently just $200\ \mathrm{mm\ Hg}$ (the height is the reading in column units).

Step 3. Absolute: $P = P_0 + \rho gh = 1.01\times10^5 + 0.267\times10^5 = \boxed{1.28\times10^5\ \mathrm{Pa}}$ .

Step 4. In column units: $760 + 200 = \boxed{960\ \mathrm{mm\ Hg}}$ . State which number answers which question — "the manometer reads 200 mm Hg (gauge); the gas is at 960 mm Hg (absolute)."

2 · Board level — water balances mercury (CBSE/ISC pattern, 3 marks)

A U-tube contains mercury. Water is poured into one arm until the water column is $13.6\ \mathrm{cm}$ tall. By how much does the mercury level in the two arms now differ? ( $\rho_w = 1000$ , $\rho_{\text{Hg}} = 13600\ \mathrm{kg/m^3}$ .)

Step 1. Reference level: the water–mercury interface. Below it, mercury is continuous through the U.

Step 2. Left arm at the interface: $P_0 + \rho_w g\,(0.136)$ . Right arm at the same level: $P_0 + \rho_{\text{Hg}}\, g\, h$ , where $h$ is the height of mercury in the right arm above that level — which is exactly the difference between the mercury surfaces.

Step 3. Equate and cancel: $\rho_w (0.136) = \rho_{\text{Hg}}\, h \Rightarrow h = \dfrac{1000\times0.136}{13600} = \boxed{1.0\ \mathrm{cm}}$ .

Step 4. Read the moral: 13.6 cm of water = 1 cm of mercury — the density ratio, inverted, every time.

3 · Competitive level — differential manometer between two vessels (JEE-Main pattern, ~2.5 minutes)

A mercury U-tube connects gas vessel A (left) to gas vessel B (right). The mercury on B's side stands $12\ \mathrm{cm}$ higher than on A's side. (a) Which gas is at higher pressure, and by how much (take $g = 10\ \mathrm{m/s^2}$ )? (b) If vessel B is now opened to the atmosphere ( $P_0 = 1.0\times10^5$ Pa) and the difference settles at $8.0\ \mathrm{cm}$ with A's side still lower, find A's absolute pressure.

Step 1. (a) Mercury is pushed down on A's side ⇒ A presses harder: $P_A > P_B$ . Walk from A's surface down to the reference (A's mercury level) then up $0.12$ m to B's surface: $P_A = P_B + \rho g(0.12)$ .

Step 2. $P_A - P_B = 13600\times10\times0.12 = \boxed{1.63\times10^4\ \mathrm{Pa}}$ (= 120 mm Hg — column units for free).

Step 3. (b) Same walk with $P_B \to P_0$ and $h = 0.08$ m: $P_A = 1.0\times10^5 + 13600\times10\times0.08 = 1.0\times10^5 + 1.09\times10^4 = \boxed{1.11\times10^5\ \mathrm{Pa}}$ . Timing note: differential manometers are direction-reading tests. Thirty of the ninety seconds belong to the sentence "mercury depressed on the side that pushes harder" — write the inequality BEFORE the arithmetic and the sign never flips on you. Answers in mm Hg are read straight off the column; convert only if the paper demands pascals.

4 · Competitive level — the faulty barometer (JEE-Main/NEET pattern, ~3 minutes)

On a day when a correct barometer reads $76.0\ \mathrm{cm}$ , a faulty barometer (some air trapped above the mercury) reads only $73.0\ \mathrm{cm}$ . (a) Find the pressure of the trapped air, in cm of Hg and in pascals ( $\rho_{\text{Hg}} = 13600$ , $g = 9.8$ ). (b) The faulty tube is now tilted at $30°$ to the vertical (true pressure unchanged, trapped air's pressure assumed unchanged for this part). What length of mercury thread stands along the tube?

Step 1. (a) Walk the faulty tube from the dish surface ( $P_0$ , i.e. 76 cm Hg) up to the trapped air at the top of the 73 cm column: $P_{\text{air}} = P_0 - \rho g(0.73)$ , i.e. in column units $P_{\text{air}} = 76.0 - 73.0 = \boxed{3.0\ \mathrm{cm\ Hg}}$ .

Step 2. Convert: $13600\times9.8\times0.030 = \boxed{4.0\times10^3\ \mathrm{Pa}}$ . The faulty barometer under-reads by exactly the trapped air's pressure — the general theorem of this family.

Step 3. (b) The balance fixes the vertical height at 73.0 cm; along a tube tilted $30°$ from vertical the thread length is $\ell = \dfrac{h}{\cos 30°} = \dfrac{73.0}{0.866} = \boxed{84.3\ \mathrm{cm}}$ .

Step 4. Sanity: a healthy tilted barometer plays the same game — vertical height loyal to the pressure, thread stretching along the glass. (Strictly, tilting also shrinks the trapped-air space and compresses that air — that refinement needs Boyle's law from the heat track; exam items either ignore it, as stated here, or become gas-law hybrids.) Timing note: part (a) is a 45-second read once you know the theorem (reads low by $P_{\text{air}}$ ). Part (b) is the tilt immunity in reverse: examiners alternate "reading unchanged" (vertical height) with "thread length" ( $h/\cos\theta$ ) — parse which is asked.

Exam traps

Down adds, up subtracts — with the local liquid's density. The walk (equation 1) is the only safe general tool. Each segment uses the density of the fluid you're in; horizontal legs are free.

Which arm is higher tells you who's winning. The pushed-down side belongs to the higher pressure. Write the inequality in words before the equation; sign errors in manometry are direction errors.

Gauge vs absolute. The column shows gauge ( $\rho gh$ ); gas laws and most "find the pressure of the gas" questions want absolute ( $+P_0$ ). AP marks the distinction explicitly; JEE hides it in the options.

Never equate levels across an interface. Reference line at the lowest interface of a multi-liquid tube — the warning box's rule. Also never through a gas pocket: the pocket transmits one pressure but breaks liquid continuity.

mm Hg is a pressure unit. $1\ \mathrm{mm\ Hg} = 13600\times9.8\times10^{-3} \approx 133\ \mathrm{Pa}$ ; 760 mm Hg = $1.013\times10^5$ Pa. Blood pressure's 120/80 lives in these units. Column-unit arithmetic (76 − 73 = 3 cm Hg) is faster and safer than converting early.

Tilt and width immunity. Barometer readings depend on vertical height only: fatter tube, same reading; tilted tube, same vertical height (longer thread, $h/\cos\theta$ along the glass). Both stated forms are one-mark MCQs.

Faulty barometer reads LOW, by the trapped air's pressure. And it becomes less accurate on high-pressure days (air compresses, but never to zero). The quantitative day-to-day tracking needs Boyle's law — flag it, don't fake it.

Choice of gauge liquid. Mercury for atmospheres, water/oil for whispers: a pressure of 1 cm Hg is 13.6 cm of water — low-density manometers amplify small differences. "Why mercury in barometers?" = a 10 m water tube is impractical (and water's vapour pressure pollutes the vacuum).

Origins · 1 minEvangelista Torricelli and how this was discovered

Evangelista Torricelli

1608–1647 · Florence

Evangelista Torricelli was Galileo's last assistant — he arrived in Arcetri three months before the old man died, then succeeded him as mathematician to the Grand Duke of Tuscany. Pump-makers had a practical mystery: suction pumps refused to lift water past about 10 metres. Torricelli's 1643 inversion of the problem was revolutionary — replace water with mercury, fourteen times denser, and the limit shrinks to a tabletop 76 centimetres. His mercury column simultaneously produced the first sustained laboratory vacuum (heresy to Aristotelians) and the correct explanation: "we live submerged at the bottom of an ocean of air," whose weight holds the column up. He noticed the height flickered day to day — the barometer was born forecasting weather before anyone knew it. He died at 39; the torr (1 mm Hg) carries his name through every vacuum lab on Earth.

History

1643

— Torricelli's mercury column: the barometer, the first laboratory vacuum, and "an ocean of air" in one experiment.

1648

— Pascal's Puy-de-Dôme climb: the column falls with altitude — the atmosphere really is a fluid with weight.

1654

— Otto von Guericke's Magdeburg hemispheres make atmospheric pressure a public spectacle: sixteen horses versus a vacuum.

1849

— Bourdon's mechanical gauge begins replacing liquid columns in industry; the U-tube stays as the calibration standard.

1896

— Riva-Rocci's mercury sphygmomanometer: blood pressure becomes a number in mm Hg, a unit medicine still refuses to drop.

Going further

Texts. NCERT Physics Class 11, Ch. 9 · HC Verma Vol. 1, Ch. 13 (the U-tube worked set) · Resnick–Halliday–Walker Ch. 14.

Provenance. All results standard. A verifier should recompute: example 1's $2.67\times10^4$ Pa and 960 mm Hg; example 2's 1.0 cm; example 3's $1.63\times10^4$ Pa and $1.11\times10^5$ Pa; example 4's 3.0 cm Hg = $4.0\times10^3$ Pa and $73.0/\cos30° = 84.3$ cm.

Prerequisites. Fluids at Rest: Pressure & Pascal's Law. Leads to. Floating-Body Problems (archetype) · Liquid in an Accelerating Container (archetype) · Option Elimination & Answer Checking (track capstone).

Why does the manometer read gauge pressure and not absolute?

Because the open arm has the atmosphere pressing on it. The column's height settles where the two arms' pressures balance at the reference level: gas pressure on one side versus atmosphere-plus-column (or minus) on the other. The atmosphere appears on both sides of that balance except through the height difference — so $h$ weighs only $P_{\text{gas}} - P_0$ . To make an absolute gauge you must remove the atmosphere from the second arm: seal it with vacuum on top, which is precisely what a barometer is (and why closed-tube manometers read absolute).

Why doesn't the tube's width matter? Surely a fat column of mercury weighs more.

It does weigh more — and it has proportionally more area to be pushed on. Pressure balance is weight per area: $\rho gh$ carries no $A$ . This is the hydrostatic paradox from the pressure article wearing its instrument costume. Practical fine print: very narrow tubes do err, but from capillarity (surface tension pulling the meniscus up or down), not from weight — mercury barometers use bores wide enough (~1 cm) to keep the capillary correction below the reading precision.

What is really in the "vacuum" above the barometer's mercury?

Mercury vapour at its room-temperature vapour pressure — about $0.2$ Pa, versus the $10^5$ Pa being measured: a part-per-million error, honestly ignorable. That's why mercury (and not water) makes a good barometer twice over: density 13.6 keeps the tube short, and its feeble vapour pressure keeps the top space effectively empty. A water barometer's top space would hold ~2300 Pa of water vapour at 20 °C, stealing over 2% of the reading — and the tube would be 10 m tall.

How do I handle a manometer where the gas arm also has some liquid column above the mercury?

Walk it — the master pattern absorbs any stack. Start at the gas: going down through the light liquid adds $\rho_{\text{light}}\,g\,h_{\text{light}}$ ; reaching mercury and continuing to the reference level adds mercury terms; then climb the open arm subtracting. The memorized special cases ( $P_0 \pm \rho gh$ , $\rho_1h_1 = \rho_2h_2$ ) are just walks somebody did for you. On any tube with three or more fluids, trust only the walk.

Why do doctors still say "millimetres of mercury" when nobody uses mercury cuffs anymore?

Unit inertia with a good excuse: a century of clinical thresholds (hypertension at 140/90, etc.) is written in mm Hg, so redefining to kilopascals would re-normalize every guideline and every doctor's intuition. The digital replacements are calibrated against mercury-column standards, so the unit remains physically honest: your 120 mm Hg systolic really would hold up 12 cm of mercury. A few countries chart kPa (120 mm Hg ≈ 16 kPa), but the torr's medical monopoly endures — Torricelli's column outliving Torricelli's instrument.

Can a suction pump really not lift water beyond about 10 m? Why exactly?

Correct — and it's this article's physics, inverted. A suction pump doesn't pull water up; it removes air above, letting atmospheric pressure push the water up. The atmosphere's entire budget is $P_0 = 1.013\times10^5$ Pa, which supports at most $h = P_0/\rho_w g \approx 10.3$ m of water — even with a perfect vacuum overhead. Torricelli's mercury version is the same statement at 76 cm. Deep wells therefore use submersible pumps that push from below (Pascal's law has no such ceiling) — a design decision made by equation (3).