Newton's Third Law & Interaction Pairs

A horse is arguing with you. "Whatever force I put on the cart," it says, "the cart puts exactly the same force back on me — your own Newton says so. Equal and opposite. The forces cancel. Therefore I cannot move the cart, and I am going back to my stable." Every year, some version of this argument appears on every board's paper on Earth, and every year students lose marks to a law that fits in one sentence: when one body exerts a force on a second, the second exerts an equal and opposite force on the first. The escape from the horse's trap is a single phrase with an entire module inside it: the two forces act on different bodies. The cart's backward pull acts on the horse; the horse's forward pull acts on the cart. No free-body diagram ever contains both, so nothing cancels. What actually decides whether the cart moves is the net force on the cart — and the horse's pull is unopposed there until friction says otherwise. This is the third and final law of motion, and in some ways the deepest. It says forces never come alone: every push is half of a handshake, every pull a tug-of-war with both ends live. It explains how you walk (you shove the Earth backward; the Earth shoves you forward), how rockets fly in vacuum, why a rifle kicks, and why the Earth itself lurches — immeasurably — toward every falling apple. Master the pair test in this module and the Free-Body Diagrams module that follows becomes almost mechanical: every force in every diagram will have a partner you can name, living in someone else's diagram.

Key takeaways

If body A exerts a force on body B, then B exerts a force on A that is equal in magnitude and opposite in direction: $F_{AB} = -F_{BA}$ , at every instant.
The two forces of a pair act on different bodies — they never meet in one free-body diagram and can never cancel each other.
A genuine pair passes a strict test: same interaction, same type of force, equal magnitudes, opposite directions, bodies swapped.
The normal force and the weight on a resting book are not a pair — the reaction to the book's weight is the book's gravitational pull on the Earth.
Equal forces, unequal effects: $a = F/m$ , so the lighter partner accelerates more (bullet vs rifle, apple vs Earth).
Key idea: forces come in equal-and-opposite pairs acting on different bodies — so they explain motion, never forbid it.

The core insight

Whenever body A exerts a force on body B, body B simultaneously exerts a force of equal magnitude and opposite direction on body A; the two forces are the same type, belong to the same interaction, and act on different bodies — which is why they never cancel and why nothing in the universe can push without being pushed back.

Where this lives today

Every spacecraft is a third-law machine. The Hall-effect thrusters steering thousands of Starlink satellites, the ion engines that carried missions to the asteroid belt on thrust weaker than a coin's weight, the launch abort motors on crewed capsules — all of them accelerate by hurling mass or ions one way and pocketing the reaction. Attitude control runs the rotational version: spin a reaction wheel clockwise and the telescope turns counterclockwise, which is how JWST points at a galaxy for days without burning a gram of fuel. The 1920 editorial's error is now a trillion-dollar industry's operating principle. On the ground the law is measurement infrastructure. Force plates in sports-science labs read the ground's push on a sprinter — the reaction to her backward shove — to tune starting technique; strain gauges in tow couplings and robot joints read one interface force with two addresses; and planet hunters detect worlds they cannot see by the reflex wobble a planet's gravity imposes on its star, the gravity pair at stellar scale doing exactly what the book does to the Earth.

The physical picture

Push on a wall. You feel the wall push back on your palm — harder as you press harder, vanishing the instant you stop. You cannot touch anything without being touched, with exactly equal strength, at exactly the same moment. That mutual grip is the third law made tangible, and it holds for every force in your syllabus: contact pushes, friction, tension, gravity, and later the electric and magnetic forces too. Now watch yourself walk. Your foot pushes backward on the ground; the ground's friction pushes forward on you — and that forward push is the only external horizontal force available to accelerate your body. You move because of the force the planet returns, not the force you exert (that one acts on the planet). On smooth ice the trick dies: push straight down all you like, there is no backward grip to return a forward push, so you stay put. Swimmers, sprinters, rockets and jet engines all run the same trade: throw something one way, get pushed the other. The whole art is bookkeeping. Every force belongs to a body; its partner always belongs to the other body. The classic sorting exercise is a book resting on a table — three bodies (book, table, Earth), two genuine pairs, and one famous impostor:

Figure

A book on a table — two true interaction pairs and one impostor balance

Read the figure left to right. The left panel shows the two forces on the book: normal force up, weight down. They happen to balance — but only because the book happens to be in equilibrium, and they are not even the same kind of force. The middle and right panels show the genuine pairs: the contact pair (table pushes book up; book pushes table down) and the gravity pair (Earth pulls book down; book pulls Earth up). Each pair straddles two bodies. No pair ever fits inside one diagram — that is the whole point of it.

The law and the pair test

Fix a naming convention worth internalizing for the rest of mechanics: write $F_{AB}$ for "the force that A exerts on B" — agent first, receiver second. With names attached, the entire third law is one line:

F_{AB} = -F_{BA} \qquad (1)

where: $F_{AB}$ — force of body A on body B (it acts on B); $F_{BA}$ — force of B on A (it acts on A); the minus sign — exactly opposite directions along the same line. Taking magnitudes, $|F_{AB}| = |F_{BA}|$ : neither body is "stronger," ever — not truck versus mosquito, not Earth versus apple. Notice what the equation refuses to do: it never adds the two forces, because sums of forces only make sense on a single body, and these two act on different ones. Because "equal and opposite" also describes ordinary balanced forces, examiners love impostors. Run every candidate pair through the full test — all five checks, not just the first two:

check	genuine third-law pair	the impostor (N and W on a book)
equal magnitudes?	always, at every instant	only while $a = 0$
opposite directions?	always	yes — coincidence of equilibrium
same type of force?	always (gravity ↔ gravity, contact ↔ contact)	no — contact vs gravity
act on different bodies?	always — that is the law	no — both act on the book
swap the names and it reads sensibly?	"A on B" ↔ "B on A"	fails — you cannot swap "table on book" into "Earth on book"

The last row is the fastest field test: state the force as "X on Y," and its partner must be "Y on X." The reaction to "Earth pulls book down (gravity)" is "book pulls Earth up (gravity)" — not the table's push, which belongs to a completely different interaction. Put the book in a falling lift and the impostor is exposed instantly: $N$ collapses toward zero while the weight does not budge, yet each true pair stays perfectly matched throughout the fall.

Derivation AWhy equal forces produce unequal motion — the recoil ratio

We derive the signature prediction of the third law: when two bodies interact only with each other, starting from rest, they move apart with speeds in the inverse ratio of their masses — and we get momentum conservation as a free gift. The strategy is to apply the second law to each body separately, feed in the third law, and compare. Watch one subtlety: the pair forces are equal at every instant and last for exactly the same time (you cannot touch without being touched), which is what makes the comparison exact rather than approximate.

Step 1. Let the two bodies be a rifle (mass $M$ ) and a bullet (mass $m$ ). During firing, the rifle pushes the bullet forward with some force $F(t)$ — spiky and unknown — and by equation (1) the bullet pushes the rifle backward with exactly $F(t)$ at every instant.

Step 2. Use the momentum form of the second law on each body over the firing time $\Delta t$ : the bullet gains momentum $\Delta p_b = +F\,\Delta t$ and the rifle gains $\Delta p_r = -F\,\Delta t$ , where $F$ is the average force. Same average force, same duration — the two momentum changes are equal and opposite whatever the force profile looked like. Why: this is the deep reason recoil problems never need the messy force history — the third law guarantees the totals match even when nobody can measure the spikes.

Step 3. Starting from rest, final momenta are equal and opposite: $mv = MV$ , so

\frac{v}{V} = \frac{M}{m}

Why: the lighter body carries the same momentum with more speed — a 100-times-lighter bullet leaves 100 times faster than the rifle recoils.

Step 4. Add the two momentum changes: $\Delta p_b + \Delta p_r = 0$ . The total momentum of the pair never changed. Internal forces — forces the system's parts exert on each other — always come in cancelling pairs, so they can redistribute momentum but never create or destroy it. That single sentence becomes the Conservation of Linear Momentum module. Verified: limits behave — equal masses give equal speeds; M enormous gives V nearly zero (fire from a battleship and the ship barely notices). Numbers from Worked Example 2 check: 0.02 times 300 equals 4 times 1.5 — both 6 kg m/s.

Chains of pairs: blocks in contact

The third law earns its exam keep in contact chains — several bodies touching, each interface carrying its own pair. The cleanest laboratory: two blocks side by side on a smooth table, pushed from the left.

Figure

Two blocks pushed in contact — the interface carries one pair

Treat the pair of blocks as one system first: internal pair forces cancel within the system's total, so the whole $6\ \mathrm{kg}$ obeys $a = F/(m_1 + m_2)$ . Then cut the system at the interface to expose the pair: the contact force on the rear face of the front block is whatever the front block alone needs to keep up.

N = m_2\,a = \frac{m_2\,F}{m_1 + m_2} \qquad (2)

where: $N$ — magnitude of the contact-pair force at the interface; $m_2$ — the block in front (the one being pushed only by contact); $F$ — the external push on the rear block $m_1$ . Read it physically: the interface transmits only what the downstream cargo requires. Push from the other side and the roles swap — the contact force becomes $m_1 F/(m_1+m_2)$ , bigger if $m_1$ is the heavier block. Same setup, different answer: JEE-Main has asked exactly this reversal, and equation (2) with the correct "who is downstream?" reading answers it in one line.

Go deeper · Derivation B · 1 minThe contact force between two pushed blocks — both directions

Derivation BThe contact force between two pushed blocks — both directions

We derive equation (2) and its mirror image for the two-block contact problem, because the pair of answers — 8 N one way, 16 N the other for the same numbers — is the single most-tested consequence of the third law in JEE-Main archives. The strategy is the two-pass method that runs all of Newtonian problem-solving: system pass for the acceleration, then a cut at the interface for the internal pair. The thing to watch is which block is "downstream" of the push — the contact force always equals the downstream mass times the common acceleration.

Step 1. System pass. Blocks $m_1 = 4\ \mathrm{kg}$ and $m_2 = 2\ \mathrm{kg}$ touch on a smooth table; a horizontal $F = 24\ \mathrm{N}$ pushes $m_1$ from the left. They move together, so one acceleration serves both: $a = \dfrac{F}{m_1+m_2} = \dfrac{24}{6} = 4\ \mathrm{m/s^2}$ . Why the internal forces vanish here: the interface pair acts once forward on m2 and once backward on m1 — summed over the system they cancel by equation (1), leaving only the external push.

Step 2. Cut at the interface; isolate the front block. The only horizontal force on $m_2$ is the contact push $N$ from $m_1$ : $N = m_2 a = 2 \times 4 = 8\ \mathrm{N}$ .

Step 3. Check with the rear block's diagram, where the third law places the partner force $N$ backward: $F - N = m_1 a \Rightarrow 24 - 8 = 16 = 4 \times 4$ ✓. One interaction, two addresses, one magnitude.

Step 4. Now push the same $24\ \mathrm{N}$ from the right, on $m_2$ . The acceleration is unchanged ( $4\ \mathrm{m/s^2}$ ), but the downstream cargo is now the $4\ \mathrm{kg}$ block: $N' = m_1 a = 4 \times 4 = 16\ \mathrm{N}$ . Why the asymmetry: the interface only ever carries what the block ahead of it needs. Heavier cargo, bigger contact force — even though F, total mass and acceleration are identical. *Verified: both answers satisfy each block's second law separately; 8 N and 16 N sum-check against F = 24 N with the correct signs, and the m2 F/(m1+m2) formula reproduces 24 times 2/6 = 8 and 24 times 4/6 = 16.*

Limiting cases

case	result	meaning
masses equal ( $m_1 = m_2$ )	contact force $= F/2$ , recoil speeds equal	perfect symmetry — each partner shares alike
one mass enormous (rifle, Earth, wall)	heavy partner's motion imperceptible	the pair is intact; $a = F/M$ merely vanishes as $M$ grows
bullet vs rifle ( $m \ll M$ )	$v/V = M/m \gg 1$	equal momenta, wildly unequal speeds — and KE splits unevenly too
statics ( $a = 0$ everywhere)	each interface transmits the full load above it	a stack's bottom face carries the total weight — via one interface
system pass (blocks glued mentally)	internal pairs cancel in the sum	why "treat them as one body" is ever legal

Worked examples

1 · Board level — skater pushes off a wall (CBSE/IGCSE pattern, 3 marks)

A skater of mass $50\ \mathrm{kg}$ , at rest on smooth ice, pushes on a rigid wall with a force of $120\ \mathrm{N}$ . (a) State the force the wall exerts on the skater. (b) Find the skater's acceleration while pushing. (c) Explain why the wall does not accelerate.

Step 1. By the third law, the wall pushes back on the skater with $\boxed{120\ \mathrm{N}}$ , horizontally, away from the wall — equal magnitude, opposite direction, no delay.

Step 2. The skater's own push acts on the wall, so it does not appear in her diagram. The only horizontal force on her is the wall's $120\ \mathrm{N}$ : $a = F/m = 120/50 = \boxed{2.4\ \mathrm{m/s^2}}$ away from the wall.

Step 3. The wall receives a genuine $120\ \mathrm{N}$ — but it is bolted to a building bolted to the Earth. With $M$ effectively planetary, $a = F/M$ is immeasurably small. The pair is perfectly intact; the response differs because the masses do.

Step 4. Write the sentence that earns the reasoning mark: the two forces are equal and opposite but act on different bodies — the skater's motion is decided only by the forces on the skater.

2 · Board level — rifle recoil (ISC/A-Level pattern)

A $4.0\ \mathrm{kg}$ rifle fires a $20\ \mathrm{g}$ bullet at $300\ \mathrm{m/s}$ . Find (a) the rifle's recoil speed, and (b) the average force on the shooter's shoulder if the shoulder stops the rifle in $0.10\ \mathrm{s}$ .

Step 1. During firing, rifle-on-bullet and bullet-on-rifle are a pair: equal average force, equal duration, so equal and opposite momentum changes. Bullet's momentum: $p = mv = 0.020 \times 300 = 6.0\ \mathrm{kg\,m/s}$ forward.

Step 2. The rifle therefore carries $6.0\ \mathrm{kg\,m/s}$ backward: $V = p/M = 6.0/4.0 = \boxed{1.5\ \mathrm{m/s}}$ . Check the ratio: $v/V = 300/1.5 = 200 = M/m = 4.0/0.020$ ✓.

Step 3. The shoulder must drain the rifle's $6.0\ \mathrm{kg\,m/s}$ in $0.10\ \mathrm{s}$ . Average force $= \Delta p/\Delta t = 6.0/0.10 = \boxed{60\ \mathrm{N}}$ — and by the third law the rifle presses on the shoulder with the same $60\ \mathrm{N}$ : the kick you brace for.

Step 4. Note the energy asymmetry for the reasoning mark: equal momenta, but $KE = p^2/2m$ , so the bullet takes $6^2/(2\times0.020) = 900\ \mathrm{J}$ while the rifle takes only $6^2/(2\times4.0) = 4.5\ \mathrm{J}$ . The third law equalizes momentum, never energy.

3 · Competitive level — the contact-force reversal (JEE-Main pattern, ~2.5 minutes)

Blocks of $4\ \mathrm{kg}$ and $2\ \mathrm{kg}$ rest in contact on a smooth table. A horizontal $24\ \mathrm{N}$ force can be applied either to the $4\ \mathrm{kg}$ block (case I) or, from the other side, to the $2\ \mathrm{kg}$ block (case II). Find the contact force between the blocks in each case.

Step 1. Both cases share the system pass: $a = \dfrac{24}{4+2} = 4\ \mathrm{m/s^2}$ , whichever side is pushed.

Step 2. Case I — downstream cargo is the $2\ \mathrm{kg}$ block: $N = m_2 a = 2 \times 4 = \boxed{8\ \mathrm{N}}$ .

Step 3. Case II — downstream cargo is the $4\ \mathrm{kg}$ block: $N' = m_1 a = 4 \times 4 = \boxed{16\ \mathrm{N}}$ .

Step 4. Cross-check case I from the rear block: $24 - 8 = 16 = 4 \times 4$ ✓; and case II from its rear block: $24 - 16 = 8 = 2 \times 4$ ✓. The interface carries exactly what the block ahead of it needs — nothing more. Timing note: ~2.5 min, and the system-then-cut routine should feel like one move. The planted distractor is "same both ways, 12 N" — symmetric-looking setups tempt symmetric answers. The contact force follows the cargo, not the geometry.

4 · Competitive level — three blocks and a statement audit (JEE-Main/NEET pattern, ~3 minutes)

Blocks of $1\ \mathrm{kg}$ , $2\ \mathrm{kg}$ and $3\ \mathrm{kg}$ sit in a row (lightest first) on a smooth table; $60\ \mathrm{N}$ pushes the $1\ \mathrm{kg}$ block. Find both contact forces. Then judge the statement: "the force the 2 kg block exerts on the 3 kg block is the reaction to the force the 1 kg block exerts on the 2 kg block."

Step 1. System pass: $a = \dfrac{60}{1+2+3} = 10\ \mathrm{m/s^2}$ .

Step 2. Interface between 1 kg and 2 kg: downstream cargo is $2+3 = 5\ \mathrm{kg}$ , so $N_{12} = 5 \times 10 = \boxed{50\ \mathrm{N}}$ .

Step 3. Interface between 2 kg and 3 kg: downstream cargo is $3\ \mathrm{kg}$ , so $N_{23} = 3 \times 10 = \boxed{30\ \mathrm{N}}$ .

Step 4. The statement is false. The reaction to "1 kg pushes 2 kg (50 N forward)" is "2 kg pushes 1 kg (50 N backward)" — same interface, bodies swapped. The 30 N at the next interface is a different interaction entirely; it is not even equal in magnitude. Audit check on the middle block: $50 - 30 = 20 = 2 \times 10$ ✓. Timing note: ~3 min. NEET phrases this as multi-statement conceptual MCQs; the reflex "reaction = same interface, names swapped" kills every impostor option in seconds.

Exam traps

Pairing weight with the normal force. The reaction to the Earth pulling the book down is the book pulling the Earth up — gravity pairs with gravity, on the swapped body. $N$ is a different interaction (contact), whose partner is the book pressing the table. The falling-lift test exposes the impostor: $N$ changes, $W$ does not.

Adding forces across diagrams. "Equal and opposite, so they cancel" is only true for forces on one body. A third-law pair never shares a diagram; if your equation contains both members, you have mixed two bodies' physics.

Assuming the heavier body exerts the bigger force. Truck hits scooter: forces are exactly equal at every instant. What differs is $a = F/m$ and what the structures survive. Boards plant "the truck exerts more force" as an option every single year.

Reaction of the wrong type. A pair is always the same kind of force. The reaction to a gravitational pull is never a contact push; the reaction to friction on the tyre is friction on the road.

Contact-force symmetry. In pushed-block chains the interface force is not $F/2$ by default — it equals downstream mass × acceleration, and it changes when the push swaps sides (8 N vs 16 N above).

Believing the pair implies equal effects. Equal momenta delivered, yes ( $mv = MV$ ); equal speeds or kinetic energies, no. $KE = p^2/2m$ punishes the heavy partner.

Which g? When weights enter (stacked blocks, lift problems), use the paper's convention: CBSE/ISC $9.8\ \mathrm{m/s^2}$ , JEE/NEET as stated (usually $10$ ), A-Level $9.81$ .

Origins · 1 minErnst Mach and how this was discovered

Ernst Mach

1838–1916 · Vienna and Prague

Mach — physicist, philosopher, the man the speed-of-sound ratio is named after — did more than anyone to expose how deep the third law runs. In The Science of Mechanics (1883) he argued that Newton's "mass" was circular as defined, and rebuilt it operationally: let two bodies interact and measure their accelerations; the ratio $m_2/m_1 = a_1/a_2$ defines their mass ratio. Mass, in Mach's construction, is not a substance but a score extracted from the third law's mutual pushes — which makes the law less a fact about forces than the definition of how much matter resists them. His insistence that mechanics rest only on observable interactions became famous as "Mach's principle" and directly shaped a young Einstein, who credited Mach's critique as a seed of general relativity.

History

1687

— Newton's Principia states Lex III, illustrated with a finger pressing a stone and a horse drawing a rope: the finger is also pressed by the stone.

1883

— Ernst Mach's The Science of Mechanics rebuilds the definition of mass from third-law acceleration ratios, turning the law into the operational bedrock of mechanics.

1920

— The New York Times mocks Robert Goddard for believing rockets could work in vacuum, insisting thrust needs air to push against.

1926

— Goddard flies the first liquid-fuelled rocket: exhaust pushed one way, rocket the other, no air required.

1969

— Three days after Apollo 11 launches, the Times prints a correction: "it is now definitely established that a rocket can function in a vacuum."

Going further

Texts. NCERT Physics Class 11, Ch. 5 (Laws of Motion, third-law section and the horse-and-cart discussion) · HC Verma, Concepts of Physics Vol. 1, Ch. 4–5 · Halliday, Resnick & Walker, Fundamentals of Physics, Ch. 5.

Provenance. Standard textbook physics throughout. A verifier should recompute: the skater ( $a = 120/50 = 2.4\ \mathrm{m/s^2}$ ); the rifle ( $p = 6.0\ \mathrm{kg\,m/s}$ , $V = 1.5\ \mathrm{m/s}$ , ratio $300/1.5 = 200 = 4.0/0.020$ , shoulder force $60\ \mathrm{N}$ , energies $900\ \mathrm{J}$ vs $4.5\ \mathrm{J}$ ); the two-block pair ( $a = 4\ \mathrm{m/s^2}$ , $N = 8\ \mathrm{N}$ vs $N' = 16\ \mathrm{N}$ , checks $24-8 = 16$ and $24-16 = 8$ ); the three-block chain ( $a = 10\ \mathrm{m/s^2}$ , $N_{12} = 50\ \mathrm{N}$ , $N_{23} = 30\ \mathrm{N}$ , middle-block check $50-30 = 20 = 2\times10$ ); Earth's response to a 1 kg apple ( $9.8/6\times10^{24} \approx 1.6\times10^{-24}\ \mathrm{m/s^2}$ ).

Prerequisites. Force, Inertia & Newton's First Law · Newton's Second Law: F = ma. Leads to. Free-Body Diagrams & Problem-Solving Strategy · Connected Blocks, Strings & Constraints (the Atwood Machine's sibling archetype) · Conservation of Linear Momentum — and, for the deep dive, the university-tier treatment in Newton's Third Law (Advanced Mechanics).

If every force has an equal and opposite partner, how does anything ever start moving?

Because motion is decided per body, and the partner force is never on the same body. To predict what the cart does you sum forces on the cart — the horse's forward pull, friction backward — and the cart's own backward pull on the horse simply is not in that list; it acts on the horse. Each body can perfectly well have an unbalanced net force even though every individual force has a partner elsewhere. The universe's forces all come in pairs, and yet things accelerate everywhere, precisely because no pair ever shares a victim.

Why are weight and the normal force not a third-law pair when they are equal and opposite?

Run the full test, not the first two lines. Same type? No — one is gravity, the other a contact force. Different bodies? No — both act on the book. Names swappable? No — "Earth on book" swapped is "book on Earth," which is not "table on book." And the giveaway: they are equal only while the book is in equilibrium. Ride a lift accelerating downward and $N$ drops below $mg$ instantly — a real pair can never drift apart like that. The genuine partners are: book pulls Earth up (gravity pair), and book presses table down (contact pair).

When a truck hits a small car, the car is destroyed and the truck barely dented. Doesn't that prove the forces were unequal?

It proves the responses were unequal. At every instant of contact the two forces are exactly equal — equation (1) has no loophole for size or speed. But acceleration is $F/m$ : the same force flings the light car violently while barely slowing the truck, and the car's structure must absorb far more violent deformation. Equal causes, unequal consequences, because the bodies differ. The rifle makes it quantitative: bullet and rifle swap equal momenta, yet the bullet carries 900 J to the rifle's 4.5 J. Damage tracks energy and acceleration, never the force symmetry.

Does the Earth really accelerate upward toward a falling apple?

Yes — by exactly the pair's demand. While the apple gains momentum downward, the Earth gains the equal amount upward; with $M \approx 6 \times 10^{24}\ \mathrm{kg}$ , a $1\ \mathrm{kg}$ apple's $9.8\ \mathrm{N}$ pull produces an Earthly acceleration of about $1.6 \times 10^{-24}\ \mathrm{m/s^2}$ — real, and hopelessly unmeasurable. The effect stops being academic when masses are comparable: the Moon and Earth visibly orbit their shared centre of mass, and astronomers discover unseen exoplanets from the reflex wobble their pull imposes on their stars — the same pair, scaled up.

How does a rocket push on anything in empty space?

On its own exhaust. The engine hurls gas backward — that is the rocket exerting a force on the gas — and the gas, in the very act of being hurled, pushes the rocket forward with equal magnitude. The pair is complete in itself; no air, road or launch pad participates. Air actually hurts (drag), which is why rockets perform best in vacuum — the fact the 1920 New York Times editorial got backwards. The quantitative version, thrust = (mass thrown per second) × (exhaust speed), lives in the Momentum & the General Second Law module and comes to full power in Rocket Propulsion.

In the two-block problem, why is the contact force different when the same push comes from the other side?

Because the interface only ever carries what the downstream block needs. The common acceleration is fixed by the total ( $a = F/6$ for 24 N on 6 kg), but the interface's job depends on who is ahead of it: pushing the 4 kg first, the interface must drag only 2 kg (8 N); pushing the 2 kg first, it must drag 4 kg (16 N). Both answers respect the third law exactly — each interface carries one pair, felt backward by the pusher and forward by the pushed. The symmetric-looking "12 N" option is the trap: contact forces follow cargo, not fairness.