ARCHETYPE: Charge Sharing & the Lost Energy

Take a charged capacitor, touch its terminals to an uncharged one, and listen: a tiny snap. Now audit the books. Every microcoulomb that left the first capacitor arrived at the second — charge balances to the last digit. The energy does not. A fixed fraction of the stored energy has simply vanished, and here is the scandal that examiners have milked for forty years: the missing fraction doesn't depend on the wire. Thick copper, thin nichrome, superconductor — same loss. The snap you heard was part of the evidence leaving the scene. This is the charge-sharing archetype, and it owns a permanent seat in CBSE and ISC three-mark questions, JEE-Main energy audits, NEET single-answer traps, and CIE 9702 Paper 4. Every version is the same five-line argument: after the wires connect, both capacitors must sit at one common potential; conserved charge pins that potential; and the energy difference — always a loss — follows from arithmetic. Dress it in opposite polarities and the loss gets brutal; dress it as two touching spheres and the same law hands out charge by radius. So this article treats the procedure itself as the concept: the master ledger (charge conserved, potential shared, energy taxed), the two formulas worth memorising, the polarity twist, the sphere variant, and the traps built from students who average voltages instead of conserving charge. One archetype, an entire family of exam questions.

Key takeaways

Joined capacitors form one parallel system at a single common potential: $V_f = \dfrac{C_1V_1 + C_2V_2}{C_1 + C_2}$ — a weighted average, never the simple average.
Charge is exactly conserved; stored energy is not. The loss is $\Delta E = \dfrac{C_1C_2\,(V_1 - V_2)^2}{2\,(C_1 + C_2)}$ — zero only when $V_1 = V_2$ .
The lost fraction contains no resistance and no time: the wire's quality changes how fast the energy leaves (heat and a radio snap), never how much.
Connect opposite plates and the charges partially cancel: replace $C_2V_2 \to -C_2V_2$ in the numerator, and the loss grows to $\dfrac{C_1C_2(V_1+V_2)^2}{2(C_1+C_2)}$ .
Two conducting spheres joined by a long wire are the same problem with $C = 4\pi\varepsilon_0 R$ : charge shares as $q \propto R$ , surface charge density as $\sigma \propto 1/R$ .
Key idea: after connection, one potential; through the process, one conserved charge; between the books, one inevitable loss — fixed by the capacitances alone.

The core insight

When charged conductors are connected, charge redistributes until all sit at a single common potential $V_f = \sum C_iV_i / \sum C_i$ ; total charge is conserved while stored energy always decreases, and the lost energy $C_1C_2(V_1-V_2)^2/[2(C_1+C_2)]$ is independent of how the connection is made.

Where this lives today

Charge sharing is a live engineering tax. Inside every phone, billions of logic gates charge tiny capacitances from a fixed rail each clock cycle — wasting energy by exactly this article's mechanism, a real slice of why chips heat up and why "adiabatic" charge-recycling logic is a research field. Supercapacitor packs in buses face the problem at scale: cells at different voltages must be balanced before hard-connecting, or the equalising surge burns real energy — battery-management chips exist largely to referee this handshake gently.

The setup

Geometry first, because in an archetype the geometry is the physics. Capacitor $C_1$ is charged to $V_1$ ; capacitor $C_2$ sits at $V_2$ (often zero). Close the switch connecting positive plate to positive plate, negative to negative. Charge floods from the higher-potential capacitor to the lower until the flow has no reason to continue — which is the moment both capacitors show the same voltage. Two plates joined by a wire are one conductor; one conductor, one potential.

Three readings of the figure earn marks before any algebra:

Connected plates form one conductor. Once the switch closes, the two top plates plus the wire are a single piece of metal, and so are the two bottom plates. Each piece settles at one potential — so the two capacitors end up with the same voltage across them. Same plates joined, same polarity: this is exactly the parallel arrangement from Capacitors in Series & Parallel, whether or not the circuit is drawn to look like it.

Charge has nowhere to go. The top island of metal is not connected to the bottom island; whatever total charge the top plates started with, they keep. That single sentence is the conservation law that solves every problem in this family.

Energy has somewhere to go. While charge flows, a current runs through wire that has some resistance, heating it; the brief current spike also radiates a burst of electromagnetic energy (that snap on a nearby AM radio). Stored field energy leaves the capacitors and never comes back. The surprise — proved below — is that the total leaked is fixed before the switch even closes.

The master ledger

Every charge-sharing question is the same three-line ledger. The system: $C_1$ at $V_1$ , $C_2$ at $V_2$ , joined like-plate-to-like-plate. First, conservation. The joined top plates hold the total charge, and it cannot change:

Q_{\text{tot}} = C_1V_1 + C_2V_2 \qquad (1)

where: $C_1V_1$ and $C_2V_2$ — the initial plate charges $Q_1$ and $Q_2$ . Read it physically: this number is decided before the wire touches; everything downstream is bookkeeping. If one capacitor is uncharged, its term is simply zero. Second, the shared state. Both capacitors now feel one potential $V_f$ , and together they behave as a single capacitance $C_1 + C_2$ holding $Q_{\text{tot}}$ :

V_f = \frac{C_1V_1 + C_2V_2}{C_1 + C_2} \qquad (2)

where: $V_f$ — the common final potential. It is a capacitance-weighted average of the initial voltages: always between $V_1$ and $V_2$ , dragged toward the voltage of the bigger capacitor. Setting $C_1 = C_2$ gives the simple average — the only case where averaging voltages is legal, and examiners know you'll be tempted. Third, the tax. Compare energies before and after:

\Delta E = E_i - E_f = \frac{C_1C_2\,(V_1 - V_2)^2}{2\,(C_1 + C_2)} \qquad (3)

where: $(V_1 - V_2)^2$ — the loss is driven by the disagreement between the capacitors and dies only when they already agree; $\dfrac{C_1C_2}{C_1+C_2}$ — the series-style combination of the two capacitances. Notice what is absent: no resistance, no time, no mechanism. That absence is the most examined sentence in this article. Connect the capacitors plus-to-minus instead and the top island now holds $C_1V_1 - C_2V_2$ — partial cancellation. Every formula survives with that one sign change:

V_f = \frac{C_1V_1 - C_2V_2}{C_1 + C_2}, \qquad \Delta E = \frac{C_1C_2\,(V_1 + V_2)^2}{2\,(C_1 + C_2)} \qquad (4)

where: the loss now grows with the sum of the voltages — reversed connections are savagely lossy. The special case $C_1V_1 = C_2V_2$ kills the final state entirely: $V_f = 0$ , both capacitors dead, one hundred percent of the stored energy dissipated. A favourite JEE single-line shock.

situation	final potential	energy lost
like plates joined	$V_f = \dfrac{C_1V_1 + C_2V_2}{C_1+C_2}$	$\dfrac{C_1C_2(V_1-V_2)^2}{2(C_1+C_2)}$
opposite plates joined	$V_f = \dfrac{C_1V_1 - C_2V_2}{C_1+C_2}$	$\dfrac{C_1C_2(V_1+V_2)^2}{2(C_1+C_2)}$
second capacitor uncharged	$V_f = \dfrac{C_1V_1}{C_1+C_2}$	fraction $\dfrac{C_2}{C_1+C_2}$ of $E_i$
two spheres, radii $R_1, R_2$	common $V$ ; $q_i \propto R_i$	same formulas with $C_i = 4\pi\varepsilon_0R_i$

Derivation ACommon potential and the lost energy

We derive equations (2) and (3) — the final potential and the energy tax — the way a marking scheme expects: conservation of charge on the isolated plate-island fixes the final state, then subtraction of energies reveals the loss. The strategy is to resist writing any current or resistance at all; the endpoints are pinned by charge alone, so the energy difference cannot depend on what happens in between. Watch one subtlety: we conserve charge, never energy — assuming energy conservation here is the classic self-inflicted wound.

Step 1. Initial charges: $Q_1 = C_1V_1$ , $Q_2 = C_2V_2$ . The joined top plates form an isolated island holding $Q_{\text{tot}} = Q_1 + Q_2$ — no path exists for this total to change.

Step 2. After redistribution both capacitors share one potential $V_f$ (joined plates are one conductor). Their charges are then $Q_1' = C_1V_f$ and $Q_2' = C_2V_f$ . Conservation:

C_1V_f + C_2V_f = C_1V_1 + C_2V_2 \;\;\Rightarrow\;\; V_f = \frac{C_1V_1 + C_2V_2}{C_1+C_2}

Why: this is a weighted average — multiply out and each initial voltage carries its capacitance as the weight. The bigger capacitor wins the argument about where to settle.

Step 3. Energy before and after:

E_i = \tfrac12 C_1V_1^2 + \tfrac12 C_2V_2^2, \qquad E_f = \tfrac12 (C_1+C_2)V_f^2

Step 4. Subtract and simplify. Substituting $V_f$ and putting everything over $(C_1+C_2)$ :

E_i - E_f = \frac{(C_1V_1^2 + C_2V_2^2)(C_1+C_2) - (C_1V_1+C_2V_2)^2}{2(C_1+C_2)} = \frac{C_1C_2(V_1-V_2)^2}{2(C_1+C_2)}

Why: the numerator collapses because the $C_1^2V_1^2$ and $C_2^2V_2^2$ terms cancel, leaving $C_1C_2(V_1^2 + V_2^2 - 2V_1V_2)$ — a perfect square. A perfect square is never negative: the process always loses energy, whichever way the charge flows.

Step 5. Read the result. $\Delta E$ contains only capacitances and initial voltages — the wire's resistance sets the timescale of the loss (and how it splits between heat and radiation), not its size. Verified: limits — $V_1 = V_2$ gives $\Delta E = 0$ (no disagreement, no flow, no tax); $C_2 \to 0$ gives $\Delta E \to 0$ (connecting to nothing). Numbers: $C_1 = 6\ \mu\mathrm{F}$ at $12$ V onto empty $C_2 = 3\ \mu\mathrm{F}$ : $V_f = 8.0$ V, $E_i = 432\ \mu\mathrm{J}$ , $E_f = 288\ \mu\mathrm{J}$ , loss $144\ \mu\mathrm{J} = E_i \times C_2/(C_1+C_2)$ ✓.

Go deeper · Derivation B · 1 minTwo spheres share charge by radius

Figure

Two spheres on a long wire — charge shares by radius, density by inverse radius

The variants examiners actually pose

The uncharged partner. $V_2 = 0$ collapses everything: $V_f = \dfrac{C_1V_1}{C_1+C_2}$ , and the fraction of energy lost is exactly $\dfrac{C_2}{C_1+C_2}$ — capacitances only. Equal capacitors lose exactly half. This is the single most reused version.

The polarity reversal. "Plus to minus" means subtract the charges — equation (4). Look for the sting: if $C_1V_1 = C_2V_2$ , everything dies and the loss is total. When the final potential matters, quote its magnitude and state which initial polarity survived.

The sphere handshake. Two charged spheres touched or wired: share by radius, then separate — each keeps its share. Chains of touches are this rule applied patiently, one handshake at a time; equal-radius handshakes split the total equally, the version NEET reuses.

The heat question. "How much heat is produced?" is equation (3) in thermodynamic clothes — quote $\Delta E$ (boards accept "dissipated as heat"). No resistance given? That's the hint, not an omission.

Limiting cases

case	result	meaning
$V_1 = V_2$	$\Delta E = 0$ , nothing flows	no disagreement, no tax — connection changes nothing
$C_2 \to 0$	$V_f \to V_1$ , $\Delta E \to 0$	connecting to nothing costs nothing
$C_2 \to \infty$ (uncharged)	$V_f \to 0$ , all of $E_i$ lost	an infinite empty tank swallows charge and energy alike
$C_1 = C_2$ , $V_2 = 0$	$V_f = V_1/2$ , exactly half of $E_i$ lost	the famous 50% tax between equal capacitors
$C_1V_1 = C_2V_2$ , reversed	$V_f = 0$ , loss = 100%	equal-and-opposite charges annihilate the stored state
spheres, $R_1 = R_2$	equal split, $\sigma_1 = \sigma_2$	the "touch and separate" NEET standard

Worked examples

Board level — charged meets uncharged (CBSE/ISC pattern, 3 marks)

A $6.0\ \mu\mathrm{F}$ capacitor is charged to $12\ \mathrm{V}$ and disconnected from the battery. It is then connected across an uncharged $3.0\ \mu\mathrm{F}$ capacitor. Find the common potential, the final charge on each capacitor, and the energy lost.

Step 1. Conserved charge: $Q_{\text{tot}} = C_1V_1 = (6.0)(12) = 72\ \mu\mathrm{C}$ .

Step 2. Common potential, equation (2): $V_f = \dfrac{72\ \mu\mathrm{C}}{9.0\ \mu\mathrm{F}} = \boxed{8.0\ \mathrm{V}}$ .

Step 3. Final charges: $Q_1' = C_1V_f = \boxed{48\ \mu\mathrm{C}}$ , $Q_2' = C_2V_f = \boxed{24\ \mu\mathrm{C}}$ . Check: $48 + 24 = 72\ \mu\mathrm{C}$ ✓.

Step 4. Energies: $E_i = \tfrac12(6.0\times10^{-6})(12)^2 = 432\ \mu\mathrm{J}$ ; $E_f = \tfrac12(9.0\times10^{-6})(8.0)^2 = 288\ \mu\mathrm{J}$ . Loss $= \boxed{144\ \mu\mathrm{J}}$ — exactly the fraction $C_2/(C_1+C_2) = 1/3$ of the initial store, and exactly what equation (3) predicts. State the conclusion in words for the last mark: charge is conserved; energy is dissipated as heat (and a little radiation) in the connecting wires, independent of their resistance.

Board level — both capacitors charged (ISC/A-Level 9702 pattern)

Capacitors of $2.0\ \mu\mathrm{F}$ charged to $100\ \mathrm{V}$ and $4.0\ \mu\mathrm{F}$ charged to $50\ \mathrm{V}$ are connected positive plate to positive plate. Find the common potential, the charge that flows through the wire, and the energy dissipated.

Step 1. Total charge: $Q_{\text{tot}} = (2.0)(100) + (4.0)(50) = 200 + 200 = 400\ \mu\mathrm{C}$ .

Step 2. Common potential: $V_f = \dfrac{400\ \mu\mathrm{C}}{6.0\ \mu\mathrm{F}} = \boxed{66.7\ \mathrm{V}}$ — between 50 and 100 V, pulled toward the larger capacitor's voltage ✓.

Step 3. Charge flow: $C_1$ falls from $200\ \mu\mathrm{C}$ to $C_1V_f = 133.3\ \mu\mathrm{C}$ , so $\boxed{66.7\ \mu\mathrm{C}}$ passed through the wire into $C_2$ (which rose from 200 to $266.7\ \mu\mathrm{C}$ ✓).

Step 4. Energy: $E_i = \tfrac12(2.0\times10^{-6})(100)^2 + \tfrac12(4.0\times10^{-6})(50)^2 = 10 + 5 = 15\ \mathrm{mJ}$ ; $E_f = \tfrac12(6.0\times10^{-6})(66.7)^2 = 13.3\ \mathrm{mJ}$ . Dissipated: $\boxed{1.67\ \mathrm{mJ}}$ . Cross-check with equation (3): $\dfrac{(2.0)(4.0)(50)^2}{2(6.0)}\ \mu\mathrm{J} = 1.67\ \mathrm{mJ}$ ✓.

Competitive level — the polarity reversal (JEE-Main pattern, ~2.5 minutes)

A $3.0\ \mu\mathrm{F}$ capacitor charged to $200\ \mathrm{V}$ is connected to a $6.0\ \mu\mathrm{F}$ capacitor charged to $50\ \mathrm{V}$ , positive plate to negative plate. Find the final potential difference and the fraction of the initial stored energy that is lost.

Step 1. Reversed connection ⇒ charges subtract on the joined islands: $Q_{\text{net}} = (3.0)(200) - (6.0)(50) = 600 - 300 = 300\ \mu\mathrm{C}$ .

Step 2. Final potential: $V_f = \dfrac{300\ \mu\mathrm{C}}{9.0\ \mu\mathrm{F}} = \boxed{33.3\ \mathrm{V}}$ , with the polarity of the $3.0\ \mu\mathrm{F}$ capacitor's original charge (its side had the larger $CV$ ).

Step 3. Energy before: $E_i = \tfrac12(3.0\times10^{-6})(200)^2 + \tfrac12(6.0\times10^{-6})(50)^2 = 60 + 7.5 = 67.5\ \mathrm{mJ}$ . After: $E_f = \tfrac12(9.0\times10^{-6})(33.3)^2 = 5.0\ \mathrm{mJ}$ .

Step 4. Loss $= 62.5\ \mathrm{mJ}$ , so the lost fraction is $\dfrac{62.5}{67.5} = \boxed{0.93}$ — reversal burns almost everything. Formula check: $\dfrac{C_1C_2(V_1+V_2)^2}{2(C_1+C_2)} = \dfrac{(3.0)(6.0)(250)^2}{2(9.0)}\ \mu\mathrm{J} = 62.5\ \mathrm{mJ}$ ✓. Timing note: the entire problem is one sign decision. Read "positive to negative", write $Q_{\text{net}}$ as a difference, and this is 90 seconds of arithmetic; miss the sign and every subsequent number is wrong with full confidence. Sanity gate: a reversed connection must lose more than the like-plates version would.

Competitive level — spheres on a wire (JEE-Advanced flavour, ~4 minutes)

Two isolated conducting spheres of radii $2.0\ \mathrm{cm}$ and $4.0\ \mathrm{cm}$ carry charges $+30\ \mathrm{nC}$ and $-12\ \mathrm{nC}$ . They are connected by a long fine wire. Find the final charge on each sphere, the common potential, the ratio of surface charge densities, and the energy dissipated. Take $k = 9.0\times10^9\ \mathrm{N\,m^2/C^2}$ .

Step 1. Total charge: $Q = 30 + (-12) = 18\ \mathrm{nC}$ — signs first, always.

Step 2. Share by radius: $q_1 = \dfrac{2}{2+4}(18) = \boxed{6.0\ \mathrm{nC}}$ , $q_2 = \dfrac{4}{6}(18) = \boxed{12\ \mathrm{nC}}$ .

Step 3. Common potential: $V = \dfrac{kq_1}{R_1} = \dfrac{(9.0\times10^9)(6.0\times10^{-9})}{0.020} = \boxed{2.7\times10^3\ \mathrm{V}}$ . Check from the other sphere: $kq_2/R_2 = (9.0\times10^9)(12\times10^{-9})/0.040 = 2.7\times10^3$ V ✓.

Step 4. Densities: $\dfrac{\sigma_1}{\sigma_2} = \dfrac{R_2}{R_1} = \boxed{2}$ — the small sphere is twice as densely coated despite holding half the charge.

Step 5. Energy audit with $U = \dfrac{kq^2}{2R}$ per sphere. Before: $\dfrac{k(30\times10^{-9})^2}{2(0.020)} + \dfrac{k(12\times10^{-9})^2}{2(0.040)} = 202.5 + 16.2 = 218.7\ \mu\mathrm{J}$ . After: $\dfrac{k(6.0\times10^{-9})^2}{2(0.020)} + \dfrac{k(12\times10^{-9})^2}{2(0.040)} = 8.1 + 16.2 = 24.3\ \mu\mathrm{J}$ . Dissipated: $\boxed{194\ \mu\mathrm{J}}$ — heat in the wire plus a spark's worth of radiation. Timing note: budget four minutes only if the energy audit is asked; parts one to four are a 90-second ladder. The examiner's favourite twist hides in Step 1 — a negative charge in the total. Add signed charges before sharing; students who share magnitudes die at the first hurdle.

Exam traps

Averaging the voltages. $V_f \ne \dfrac{V_1+V_2}{2}$ unless $C_1 = C_2$ . In Example 2 the naive average gives 75 V; the truth is 66.7 V — and 75 V will be sitting among the options.

Conserving energy instead of charge. Writing $\tfrac12C_1V_1^2 = \tfrac12(C_1+C_2)V_f^2$ feels virtuous and is simply false — energy is lost here. Charge is the conserved book. The reflex check: your $V_f$ from charge conservation, then energies computed from it.

Dropping the sign in reversed connections. "Positive to negative" means the net island charge is $C_1V_1 - C_2V_2$ , and it can even be zero. Magnitudes-only workings produce confident wrong answers.

Blaming (or crediting) the wire. Any answer for the heat that contains $R$ is wrong; any claim that a better wire saves energy is wrong. $R$ sets how fast, never how much. CIE 9702 and JEE both set distractors on exactly this.

Unit slips in the µ-world. $\mu\mathrm{F} \times \mathrm{V} = \mu\mathrm{C}$ and $\tfrac12\,\mu\mathrm{F}\times\mathrm{V}^2 = \mu\mathrm{J}$ : stay in micro-units throughout and convert once at the end. Mixing $10^{-6}$ into only half the calculation is the classic two-mark leak.

Sphere problems: sharing magnitudes. Add the signed charges first (Example 4's total is 18 nC, not 42 nC), then distribute by radius. After separation each sphere keeps its share.

Constants and rounding. All boards accept $k = 9.0\times10^9\ \mathrm{N\,m^2/C^2}$ . Keep $V_f$ unrounded when computing $E_f$ — early rounding shifts the loss enough to miss a three-significant-figures box.

Origins · 1 minJames Prescott Joule and how this was discovered

James Prescott Joule

1818–1889 · Manchester

Joule was a brewer's son who ran precision experiments between batches of ale, and he owned one conviction with total stubbornness: energy that disappears from one ledger must reappear in another. In 1840, aged twenty-two, he found that a current $I$ through a resistance $R$ delivers heat at the rate $I^2R$ — the law that names the heat in your connecting wire. Three years later his paddle-wheel experiments pinned the mechanical equivalent of heat, welding mechanics and thermodynamics into one currency and making "where did the energy go?" a question that always has an answer. The charge-sharing paradox is Joule's worldview in miniature: the missing $C_1C_2(V_1-V_2)^2/2(C_1+C_2)$ is not lost at all — it is exactly the heat and radiation his ledger demands, delivered by his own $I^2R$ law integrated over the transient, whatever $R$ happens to be. The unit of every energy in this article carries his name.

History

1745

— The Leyden jar gives electricity its first storage vessel; experimenters immediately begin sharing charge between jars and noticing sparks at every connection.

1785

— Coulomb's torsion balance quantifies the force law, making the potential of a charged sphere a calculable quantity.

1840

— Joule announces the

I^2R

heating law: current through resistance is a heat engine, including the humble connecting wire.

1843

— Joule's paddle-wheel measurements fix the mechanical equivalent of heat; energy bookkeeping becomes non-negotiable physics.

1853

— William Thomson analyses the discharge of a Leyden jar and shows the stored energy can oscillate through an inductance before dissipating — the fine print behind "where the lost energy goes".

Going further

Texts. NCERT Physics Class 12, Ch. 2 (Electrostatic Potential and Capacitance) · HC Verma, Concepts of Physics Vol. 2, Ch. 31 · CIE 9702 Paper 4 capacitance sets — common-potential questions appear almost yearly.

Provenance. All results are standard. A verifier should recompute: Example 1 ( $V_f = 8.0$ V, $48/24\ \mu\mathrm{C}$ , loss $144\ \mu\mathrm{J}$ ); Example 2 ( $V_f = 66.7$ V, $66.7\ \mu\mathrm{C}$ flowed, loss $1.67$ mJ); Example 3 ( $V_f = 33.3$ V, fraction $0.93$ ); Example 4 ( $6.0/12$ nC, $2.7\times10^3$ V, $\sigma$ ratio 2, loss $194\ \mu\mathrm{J}$ ); and the perfect-square collapse in Derivation A.

Prerequisites. Capacitors in Series & Parallel · Energy Stored in a Capacitor · Electric Potential & Potential Difference. Leads to. Sphere Potentials & Earthing · Capacitor Energy and Dielectrics (UG deep dive).

Why must the two capacitors end at the same potential?

Because after connection their joined plates are literally one piece of metal, and a conductor in equilibrium sits at a single potential — otherwise charge inside it would still be moving. Current flows exactly until the disagreement is gone: the flow stops when, and only when, both capacitors read the same voltage. It is the conductor doctrine from electrostatics, now running a circuit.

Charge is conserved, so why isn't energy conserved too?

Total energy is conserved — it just stops being stored electrostatic energy. During redistribution a real current runs through real wire: some energy leaves as Joule heat, some as a burst of electromagnetic radiation. Charge has no such exit: the plate islands are electrically isolated, so every microcoulomb stays on the books. The exam-safe sentence: charge is conserved because it cannot leave; field energy is not conserved because it can — as heat and radiation.

If the wire is superconducting, where does the energy go?

Not into heat — into radiation and sloshing. With zero resistance the current spike is enormous and fast, and accelerating charge radiates: the loss leaves as a radio-frequency snap. In the idealised circuit with inductance, the charge oscillates indefinitely and never settles at $V_f$ ; any whisper of resistance damps it to the textbook answer. The examinable core: the settled final state always shows the same missing $\Delta E$ .

Why do the capacitors count as being "in parallel" after connection? The circuit doesn't look like the textbook diagram.

Parallel means exactly one thing: both elements share the same pair of nodes, hence the same potential difference. Plus-joined-to-plus and minus-joined-to-minus creates precisely that — two capacitors between the same two islands of metal. Rotating the drawing or stretching the wires changes nothing. That's why the final state behaves as a single capacitor $C_1 + C_2$ holding the total charge, the rule you met in Capacitors in Series & Parallel.

Does the length or thickness of the connecting wire change anything at all?

It changes the story, not the ledger. A long thin wire: slower redistribution, loss delivered mostly as heat. A short fat wire: violent spike, more escaping as radiation. Final charges, final potential, total energy lost — identical in every case, because charge conservation pins the endpoints before the wire gets a vote.

In sphere problems, why does charge share by radius and not by surface area?

Because the sharing is refereed by potential, not by geometry's sense of fairness. Each sphere sits at $V = kq/R$ ; equal potentials force $q \propto R$ . Area scales as $R^2$ , so density falls as $\sigma \propto 1/R$ — big sphere, more charge, thinner coat. If charge shared by area, the potentials would disagree and current would still flow: not equilibrium.

After two touching spheres are separated, does the charge redistribute again?

No. Separation breaks the conducting path, and each sphere becomes an isolated conductor keeping whatever share it held at the moment of separation. That's why "touch, then separate, then touch the next sphere" chain problems are solvable step by step: each handshake is a fresh charge-sharing event using the charges left over from the previous one, and nothing changes between handshakes.