Kepler's Laws of Planetary Motion

Johannes Kepler spent eight years fighting an eight-arc-minute discrepancy. He had inherited the finest naked-eye observations ever made — Tycho Brahe's twenty-year ledger of planetary positions — and his circular models of Mars's orbit fit them almost perfectly. Almost: two slivers of a degree, about a quarter of the width of a full moon, refused to disappear. A lesser scientist would have blamed the data. Kepler trusted Tycho's arc-minutes more than two thousand years of circular tradition, threw the circles away, and in 1609 announced something scandalous: planets move on ellipses, sweeping out equal areas in equal times. Ten years later he added the strangest law of all — the square of a planet's year is proportional to the cube of its distance — a single rule binding Mercury to Saturn. Kepler never knew why his laws held. That answer arrived in 1687, when Newton showed all three fall out of universal gravitation — which is exactly the journey your syllabus retraces. You already own the orbit condition from the satellites article; this article turns it into Kepler's third law, explains the second law as angular-momentum bookkeeping, and gives you the ellipse vocabulary the first law demands. Exams treat the three laws differently, and it pays to know the pattern: the first law is asked as vocabulary (focus, not centre), the second as a speed-comparison and conservation argument (perihelion fast, aphelion slow), and the third as the workhorse of numericals — ratio problems so standard they have their own archetype module two doors down.

Key takeaways

First law (1609): each planet moves on an ellipse with the Sun at one focus — not at the centre.
Second law (1609): the Sun–planet line sweeps equal areas in equal times; the sweep rate is $\dfrac{dA}{dt} = \dfrac{L}{2m}$ , constant because gravity exerts no torque about the Sun. Consequence: fastest at perihelion, slowest at aphelion, $v_p r_p = v_a r_a$ .
Third law (1619): $T^2 \propto a^3$ — for circular orbits Newton's form is $T^2 = \dfrac{4\pi^2}{GM}\,r^3$ , with the constant fixed by the central body alone.
The third law is a scale: measure $T$ and $a$ of anything orbiting a body, and you have weighed that body.
In Sun units the law is friendly: $T$ in years, $a$ in AU gives $T^2 = a^3$ exactly.
Key idea: three empirical rules from Tycho's data, one explanation — the inverse-square law; Kepler describes, Newton explains.

The core insight

Kepler's laws state that planets orbit on ellipses with the Sun at one focus, sweep equal areas in equal times (angular-momentum conservation in disguise), and obey $T^2 \propto a^3$ — all three being consequences of the inverse-square law, with the third law's constant $4\pi^2/GM$ depending only on the central mass.

Where this lives today

Kepler's laws are the working grammar of modern astronomy. Exoplanet hunters use the third law daily: a transiting planet reveals its period $T$ directly, and $T^2 = (4\pi^2/GM_\star)a^3$ converts it to an orbital distance — which decides whether the world sits in its star's habitable zone. The star S2, tracked for decades on a 16-year ellipse around Sagittarius A*, weighed the Milky Way's central black hole at four million suns by exactly worked example 3's method — and its perihelion sprint (second law in action, at 2.6% of light speed) provided a test of general relativity. Spacecraft navigators live on the second law's timing: mission plans exploit the slow aphelion crawl for cheap course corrections. Even satellite operators quietly obey Kepler — every transfer between orbits rides half an ellipse whose period the third law sets, as the orbit-energetics archetype shows next.

The physical picture

An ellipse is a stretched circle with two special points inside — the foci. Historically it is drawn with two pins and a loop of string: the sum of distances from any point of the curve to the two pins is constant. Half the long axis is the semi-major axis $a$ — the ellipse's "average radius" and the $a$ in the third law. How stretched the ellipse is, is measured by the eccentricity $e$ : the focus sits a distance $ea$ from the centre, so $e = 0$ is a circle and $e \to 1$ is a cigar. The planet's closest approach (perihelion) is $r_p = a(1-e)$ ; its farthest retreat (aphelion) is $r_a = a(1+e)$ ; the two add up to $2a$ — a fact exam setters adore, because giving you $r_p$ and $r_a$ secretly gives you $a = (r_p + r_a)/2$ . The first law's real content is the word focus. The Sun does not sit at the ellipse's centre; it sits offset, at one focus, with nothing at the other. For Earth ( $e = 0.017$ ) the offset is tiny — our orbit drawn to scale looks circular — but it is why Earth is about 5 million km closer to the Sun in January than in July. (And no, that is not why summer is hot: the seasons belong to Earth's tilt. The southern hemisphere gets its summer at perihelion.)

Figure

The ellipse Kepler forced on astronomy — anatomy and the equal-area clock

The second law is the orbit's speedometer. The line from Sun to planet sweeps area at a constant rate — so where the arm is long (aphelion) the planet crawls, and where the arm is short (perihelion) it sprints, keeping the swept wedges equal. This is not a new force at work; it is conservation of angular momentum wearing geometric clothing, as the derivation below shows. Halley's comet spends about 76 years per lap but races through its perihelion months at 60 times its aphelion crawl. The third law is the pattern-binder: $T^2 \propto a^3$ across every planet, moon, comet and satellite of the same central body. Double a satellite's orbital radius and its period grows by $2^{3/2} \approx 2.8$ ; put Saturn ten times farther out than Earth and its year is $10^{3/2} \approx 32$ times longer (the true figure is 29.5 — Saturn's $a$ is 9.5 AU). The law's constant carries the central mass, which turns astronomy into weighing.

The math: the three laws, HS-honest

First law. At HS level the first law is a statement with vocabulary, not a derivation: closed orbits under inverse-square gravity are ellipses with the force centre at one focus. Proving that ellipses (and, for unbound objects, parabolas and hyperbolas) are the only shapes allowed requires machinery beyond algebra — the UG deep dive delivers it. What you must command is the anatomy:

r_p = a(1-e), \qquad r_a = a(1+e), \qquad r_p + r_a = 2a \qquad (1)

where: $a$ — semi-major axis (use it wherever the third law says "distance"); $e$ — eccentricity, $0 \le e < 1$ for bound orbits. Read it as a recipe: any question giving closest and farthest distances is handing you $a$ and $e$ in disguise — add and halve for $a$ ; subtract and divide by $2a$ for $e$ .

Second law. The sweep rate is constant, and its value is angular momentum in disguise:

\frac{\Delta A}{\Delta t} = \frac{L}{2m} = \frac{v_p r_p}{2} = \frac{v_a r_a}{2} = \text{constant} \qquad (2)

where: $L = mvr\sin\theta$ — angular momentum about the Sun; at the two extremes of the orbit the velocity is exactly perpendicular to the radius ( $\sin\theta = 1$ ), which is why the cleanest consequence lives there: $v_p r_p = v_a r_a$ . Speeds trade against distances, one-for-one.

Third law. For circular orbits — the version every board asks you to derive — the orbit condition from the satellites article delivers it in three lines (Derivation B):

T^2 = \frac{4\pi^2}{GM}\,r^3 \qquad\text{and for ellipses}\qquad T^2 = \frac{4\pi^2}{GM}\,a^3 \qquad (3)

where: $M$ — the mass of the central body; the orbiting mass has cancelled. The elliptical upgrade — replace $r$ by $a$ — is stated, not proved, at this level (Newton proved it; the UG deep dive shows how). Two working forms follow. For ratios: $\left(\dfrac{T_1}{T_2}\right)^2 = \left(\dfrac{a_1}{a_2}\right)^3$ , no constants needed. For the Sun's family in friendly units: $T$ in Earth-years and $a$ in astronomical units gives $T^2 = a^3$ exactly, because Earth itself calibrates the constant.

Figure

The third law as a straight line — T squared against a cubed, one line per family

the law	exam job	working tool
first — ellipse, Sun at focus	vocabulary + $r_p, r_a$ conversions	$r_p + r_a = 2a$ , eq. (1)
second — equal areas	compare speeds along one orbit	$v_p r_p = v_a r_a$ , eq. (2)
third — $T^2 \propto a^3$	numericals across orbits	eq. (3); years-and-AU for the Sun

Derivation AThe second law is conservation of angular momentum

We derive the equal-area law from the single fact that gravity points along the Sun–planet line. The strategy is geometric: compute the area of the thin triangle swept in a short time $\Delta t$ , and show the sweep rate equals $L/2m$ — which cannot change, because a force with no lever arm about the Sun exerts no torque. Watch how "equal areas" turns out to be nothing but rotational bookkeeping.

Step 1. The swept sliver is a thin triangle. In time $\Delta t$ the planet moves through $\vec v\,\Delta t$ ; the triangle has the radius $r$ as its base and only the perpendicular part of the displacement, $v\sin\theta\,\Delta t$ , as its height ( $\theta$ is the angle between $\vec r$ and $\vec v$ ):

\Delta A = \frac12\, r \,\big(v\sin\theta\,\Delta t\big)

Why only the perpendicular part: motion along the radius stretches the arm without sweeping any angle — a triangle needs base and height, and height means perpendicular.

Step 2. Divide by $\Delta t$ and recognise the numerator:

\frac{\Delta A}{\Delta t} = \frac{r\,v\sin\theta}{2} = \frac{L}{2m}

since angular momentum about the Sun is $L = m\,v\,r\sin\theta$ .

Step 3. Show $L$ is constant. Torque about the Sun is $\tau = rF\sin\phi$ , where $\phi$ is the angle between the radius and the force. Gravity points from planet to Sun — exactly along the radius — so $\phi = 180°$ and $\tau = 0$ . No torque, no change in $L$ . Why this is the deep content: the second law would hold for ANY central force — inverse-square or not. It tests only the direction of gravity, not its strength law. The first and third laws are the ones that fingerprint $1/r^2$ .

Step 4. Cash the consequence at the apsides. At perihelion and aphelion, $\vec v \perp \vec r$ , so $L = mv_pr_p = mv_ar_a$ :

v_p\,r_p = v_a\,r_a \quad\Longrightarrow\quad \frac{v_p}{v_a} = \frac{r_a}{r_p} = \frac{1+e}{1-e}

Verified: Earth ( $e = 0.017$ ): $v_p/v_a = 1.034/0.983 = 1.03$ — a 3% seasonal speed wobble ✓. Halley ( $r_a/r_p = 35.1/0.59$ ): ratio ≈ 60 ✓, used in worked example 2.

Go deeper · Derivation B · 1 minThe third law from the orbit condition

Derivation BThe third law from the orbit condition

We derive $T^2 = (4\pi^2/GM)\,r^3$ for circular orbits — the derivation CBSE, ISC and NEET ask for by name. The strategy is to reuse the satellites article's one condition (gravity = centripetal force) and eliminate the speed in favour of the period. Watch which mass survives: the law's constant knows only the body being orbited, which is what makes the law a weighing machine.

Step 1. The orbit condition at radius $r$ :

\frac{GMm}{r^2} = \frac{mv^2}{r} \quad\Longrightarrow\quad v^2 = \frac{GM}{r}

Step 2. Replace speed by period. One lap is $2\pi r$ in time $T$ , so $v = 2\pi r/T$ :

\frac{4\pi^2 r^2}{T^2} = \frac{GM}{r}

Step 3. Rearrange:

T^2 = \frac{4\pi^2}{GM}\,r^3

Why the orbiting mass is gone: it cancelled in Step 1 — Kepler's third law holds for a pebble and for Jupiter alike, around the same central body.

Step 4. Read the constant. $4\pi^2/GM$ is the same number for everything orbiting mass $M$ : one constant for the Sun's family, a different one for Earth's satellites, another for Jupiter's moons. Comparing two orbits of the same family kills the constant entirely: $(T_1/T_2)^2 = (r_1/r_2)^3$ . Verified: Moon check — $r = 3.84\times10^8\ \mathrm{m}$ , $GM_E = 3.986\times10^{14}$ : $T = 2\pi\sqrt{r^3/GM} = 2\pi\sqrt{1.42\times10^{11}} = 2.37\times10^6\ \mathrm{s} = 27.4$ days ✓ (the sidereal month). Earth–Sun: $a = 1$ AU, $T = 1$ yr calibrates $T^2 = a^3$ in Sun units ✓.

Limiting cases

case	result	meaning
$e = 0$	circle: $r_p = r_a = a$ , constant speed	both foci merge at the centre
$e \to 1$ (bound)	cigar orbit: $v_p/v_a = \frac{1+e}{1-e} \to$ huge	comets: months at perihelion, decades asleep
same $a$ , different $e$	same $T$	the third law is blind to shape
$a \to 4a$	$T \to 8T$	the ratio grammar: $T \propto a^{3/2}$
heavier central body, same $a$	$T \propto 1/\sqrt M$ : shorter year	the constant $4\pi^2/GM$ at work
two bodies, same central mass	$T^2/a^3$ identical	one family, one constant

Common trap“The Sun is at the centre of the ellipse” — and its deadly cousin, using the wrong distance in the third law

The Sun sits at a focus, offset from the centre by $ea$ — for eccentric orbits that offset is enormous (Halley's empty focus is beyond Neptune's distance from the Sun-side focus). Diagram questions and assertion–reason items test exactly this word. The numerical cousin kills more marks: the $a$ in $T^2 \propto a^3$ is the semi-major axis — for an ellipse, neither $r_p$ nor $r_a$ but their average $(r_p + r_a)/2$ . Given "closest approach 2 AU, farthest 6 AU," the law wants $a = 4$ , not 6 and not 2. And the law compares orbits only within the same family: Earth's moon and the Sun's planets have different constants — equating them is a classic trap option.

Worked examples

1 · Board level — a year from a distance (CBSE/ISC pattern, 2–3 marks)

Mars orbits the Sun with semi-major axis $1.52\ \mathrm{AU}$ . Find its orbital period in Earth years.

Step 1. Same central body as Earth — the Sun — so use the calibrated form: $T^2 = a^3$ with $T$ in years, $a$ in AU.

Step 2. Cube: $a^3 = 1.52^3 = 3.51$ .

Step 3. Root: $T = \sqrt{3.51} = \boxed{1.87\ \mathrm{years}}$ — the familiar Martian year of about 687 days.

Step 4. Sanity: farther than Earth ⇒ longer than one year ✓, and comfortably less than Jupiter's 11.9 ✓.

2 · Board level — the comet speedometer (ISC/NEET pattern)

Halley's comet has perihelion $r_p = 0.59\ \mathrm{AU}$ and aphelion $r_a = 35.1\ \mathrm{AU}$ . Find (a) its semi-major axis and period, (b) the ratio of its perihelion speed to its aphelion speed.

Step 1. Part (a): $a = \dfrac{r_p + r_a}{2} = \dfrac{0.59 + 35.1}{2} = 17.8\ \mathrm{AU}$ .

Step 2. Third law in Sun units: $T = a^{3/2} = 17.8^{1.5} = 17.8 \times \sqrt{17.8} = 17.8 \times 4.22 = \boxed{75\ \mathrm{years}}$ ✓ (observed: 75–76).

Step 3. Part (b): the second law at the apsides — $v_pr_p = v_ar_a$ :

\frac{v_p}{v_a} = \frac{r_a}{r_p} = \frac{35.1}{0.59} = \boxed{59.5 \approx 60}

Step 4. Read it: the comet does almost all its living in a few perihelion months and sleeps for seven decades — equal areas demand it.

3 · Competitive level — weighing the Sun (JEE-Main pattern, ~2.5 minutes)

Given Earth's orbital radius $a = 1.5\times10^{11}\ \mathrm{m}$ and period $T = 3.16\times10^7\ \mathrm{s}$ , estimate the mass of the Sun. ( $G = 6.67\times10^{-11}$ .)

Step 1. Invert the third law for the central mass: $M = \dfrac{4\pi^2 a^3}{G\,T^2}$ .

Step 2. Cube and square: $a^3 = 3.38\times10^{33}\ \mathrm{m^3}$ ; $T^2 = 9.99\times10^{14}\ \mathrm{s^2}$ .

Step 3. Assemble: $M = \dfrac{39.5 \times 3.38\times10^{33}}{6.67\times10^{-11} \times 9.99\times10^{14}} = \dfrac{1.33\times10^{35}}{6.66\times10^{4}} = \boxed{2.0\times10^{30}\ \mathrm{kg}}$ .

Step 4. Read the method: any orbit is a balance, and timing it weighs the body at the focus. This is how the Sun, Jupiter (via its moons), and the Milky Way's central black hole (via the star S2) are all weighed. Timing note: the arithmetic is three powers-of-ten juggles — do the mantissas and exponents separately and this is a 90-second question. The formula itself should come from memory as "third law solved for M," not from a fresh derivation.

4 · Competitive level — apsidal energy-and-areas combo (JEE-Advanced flavour, ~4 minutes)

A planet orbits a star with perihelion $r_p = 2\times10^{10}\ \mathrm{m}$ , where its speed is $v_p = 8\times10^4\ \mathrm{m/s}$ , and aphelion $r_a = 8\times10^{10}\ \mathrm{m}$ . Find (a) its aphelion speed, (b) its period, given the star's $GM = 1.6\times10^{20}\ \mathrm{m^3/s^2}$ .

Step 1. Part (a) is pure second law: $v_a = v_p\,\dfrac{r_p}{r_a} = 8\times10^4 \times \dfrac{2}{8} = \boxed{2\times10^4\ \mathrm{m/s}}$ .

Step 2. Part (b) needs the third law with the semi-major axis: $a = \dfrac{r_p + r_a}{2} = 5\times10^{10}\ \mathrm{m}$ .

Step 3. $T^2 = \dfrac{4\pi^2 a^3}{GM} = \dfrac{39.5 \times 1.25\times10^{32}}{1.6\times10^{20}} = 3.08\times10^{13}\ \mathrm{s^2}$ .

Step 4. $T = \boxed{5.6\times10^6\ \mathrm{s}}$ — about 65 days. Timing note: the examiner's bet is that you'll feed $r_a$ (or $r_p$ ) into the third law instead of $a$ — that error changes $T$ by a factor of $(8/5)^{3/2} \approx 2$ , and the wrong option is sitting there. Say "semi-major axis = average of the extremes" aloud in your head before touching the calculator.

Exam traps

Focus, not centre. One-word trap, one guaranteed mark. The Sun is at a focus; the other focus is empty; nothing physical sits at the ellipse's centre.

Feeding the third law the wrong distance. For ellipses the law runs on $a = (r_p + r_a)/2$ . Perihelion or aphelion alone give wrong periods by up to a factor of $(1\pm e)^{3/2}$ .

Mixing families. $T^2/a^3$ is one constant per central body. Comparing a Moon orbit with a Sun orbit in the same proportion is meaningless — check what each object orbits before writing the ratio.

Equal areas ≠ equal arc lengths. The planet does not cover equal distances in equal times — that would be constant speed. It covers equal areas: short-arm sectors are wide, long-arm sectors are narrow.

"Angular momentum and energy are both constant, so speed is constant." $L$ and $E$ are constant; speed is not — kinetic energy trades against potential energy along the ellipse. Only on a circle is speed constant.

Speeds by the wrong conservation. $v_pr_p = v_ar_a$ works only at the two apsides (velocity ⊥ radius there). At a general point, $L = mvr\sin\theta$ needs the angle — JEE-Advanced likes planting a general point and watching candidates drop $\sin\theta$ .

Years-and-AU outside the solar system. $T^2 = a^3$ is calibrated by Earth's orbit and works only for objects orbiting the Sun. For satellites of Earth or moons of Jupiter, return to $T^2 = 4\pi^2r^3/GM$ or use ratios within that family.

The seasons distractor. Perihelion in January does not cause northern summer — Earth's axial tilt does. Assertion–reason papers recycle this annually.

Origins · 1 minJohannes Kepler and how this was discovered

Johannes Kepler

1571–1630 · Graz, Prague & Linz

Kepler was a mathematics teacher with terrible eyesight who became astronomy's greatest pattern-hunter. Hired in 1600 by Tycho Brahe — and inheriting, on Tycho's death a year later, the most precise observations ever taken — he waged what he called his "war on Mars": eight years of failed circles, epicycles and ovals against Tycho's data. Two stubborn arc-minutes of error finally forced him to the ellipse, published with the equal-area law in Astronomia Nova (1609), a book whose honesty is startling — it narrates every dead end. The third law arrived in 1619's Harmonice Mundi, found by sheer numerical persistence in tables of periods and distances. Kepler did all this while casting horoscopes to pay rent, defending his mother in a witchcraft trial, and losing his position in the Counter-Reformation's upheavals. He believed the laws revealed divine geometry; he could not know they were the inverse-square law's fingerprints — the evidence Newton would use, seven decades later, to identify the culprit.

History

1543

— Copernicus puts the Sun at the centre but keeps perfect circles; the fit to data stays imperfect.

1600

— Kepler joins Tycho Brahe in Prague; a year later he inherits two decades of arc-minute observations.

1609

— Astronomia Nova: the ellipse law and the equal-area law, wrung from Mars's eight stubborn arc-minutes.

1619

— Harmonice Mundi: the third law, T² ∝ a³, binding the whole planetary family.

1687

— Newton's Principia derives all three laws from universal gravitation — description becomes explanation.

2009

— NASA's planet-hunting telescope is named Kepler; it finds thousands of worlds using his third law as a ruler.

Going further

Texts. NCERT Physics Class 11, Ch. 8 (Gravitation) · HC Verma, Concepts of Physics Vol. 1, Ch. 11 · Kepler's story told properly: Koestler, The Sleepwalkers.

Provenance. All results are standard canon. A verifier should recompute: Mars $1.52^{1.5} = 1.87$ yr; Halley $a = 17.8$ AU, $T = 75$ yr, $v_p/v_a = 59.5$ ; the Sun's mass $2.0\times10^{30}$ kg from $4\pi^2a^3/GT^2$ ; example 4's $v_a = 2\times10^4$ m/s, $a = 5\times10^{10}$ m, $T = 5.6\times10^6$ s; the Moon check $27.4$ days.

Prerequisites. Newton's Law of Universal Gravitation · Satellites: Speed, Period, Energy & Weightlessness. Leads to. Kepler Ratio Problems (archetype) · Kepler's Third Law (UG deep dive).

Why did it take Kepler eight years? Couldn't he just plot the points and see an ellipse?

No one could see the orbit directly — Tycho's data gave directions from a moving platform (Earth) whose own orbit was unknown. Kepler had to solve for two orbits simultaneously, using Mars's returns to the same point every 687 days to triangulate Earth's path first. Then a circle for Mars fit to within 8 arc-minutes — success by any standard before Tycho. Kepler's greatness was refusing that success: he trusted the instrument over the theory. Low-eccentricity ellipses look maddeningly like off-centre circles; only arc-minute data could tell them apart, which is why the ellipse waited for the one astronomer with both the data and the stubbornness.

Is Kepler's second law special to gravity? What if the force weren't inverse-square?

The second law is the least picky of the three: it holds for any central force — any force always aimed at the Sun, whatever its strength law — because such a force has no torque and cannot change angular momentum. If gravity were $1/r^3$ , equal areas would survive; the ellipse and $T^2 \propto a^3$ would not. That is why examiners phrase it as "which law is a consequence of angular-momentum conservation alone?" The first and third laws are the ones that fingerprint the inverse square specifically.

Where does the planet's own mass go? Shouldn't heavier planets orbit differently?

It cancels — the same cancellation that makes all masses fall alike. Gravity's pull on a planet is proportional to the planet's mass, and so is the inertia resisting it; the orbit that results depends only on the central mass. Jupiter and a pebble at 5.2 AU share the same year. The honest fine print: the Sun isn't infinitely heavy, so really both bodies orbit their common centre of mass, and the constant is $4\pi^2/G(M+m)$ . For planets, $m/M \le 10^{-3}$ and the correction hides beyond the third significant figure — but detecting exactly that correction (the star's answering wobble) is one way exoplanets are discovered.

If the speed keeps changing along the ellipse, what plays the role of "the" orbital speed in the third law?

None of the instantaneous speeds — the third law simply never mentions speed. It relates the period (a whole-orbit quantity) to the semi-major axis (a whole-orbit average of size). If you want a speed anyway, the useful one is the mean orbital speed $2\pi a/T$ roughly; but every exam quantity you need lives in the trio $(T, a, GM)$ plus the apsidal pair $(v_p, v_a)$ from the second law. Questions mixing "speed at a random point" into the third law are testing whether you know it doesn't belong.

Do Kepler's laws work for Earth's artificial satellites?

Perfectly — with Earth as the central body and Earth's $GM$ in the constant. Satellite orbits are ellipses with Earth's centre at one focus (a circular orbit is the $e=0$ special case); a satellite in an eccentric orbit moves fastest at perigee (the Earth-orbit word for perihelion); and $T^2/a^3$ is one constant for everything orbiting Earth — which is why the Moon (27.3 days at 60 Earth radii) and the ISS (92.5 minutes at 1.06 Earth radii) sit on the same straight line when you plot $T^2$ against $a^3$ . The Kepler-ratio archetype module drills exactly these.

What is at the second, empty focus?

Nothing — and that emptiness bothered astronomers for generations. It is a purely geometric point: the ellipse needs two foci to be drawn, but the physics singles out only the one holding the Sun. There's a pleasant consolation prize: the empty focus is where the planet's direction of motion, reflected in the ellipse, converges — and in optics, waves leaving one focus of an elliptical mirror really do converge on the other. For orbits, though, the second focus is scenery. Exam options that place "the centre of mass" or "a hidden companion" there are always wrong.