Well, I don't want to confuse you, but the electrons don't actually "flow". That's just a convenient mental model. In actuality, each electron nudges nearby electrons with its electric field, and the *effect* is similar to one electron entering one end of a wire nudging one electron out the far end. Imagine a a car in a 44 million car traffic jam on Interstate 95. As on car leaves Maine, another car enters Florida, but each individual car moves very slowly
To illustrate: Let's say for the sake of the math that 1m of 10 AWG copper wire used in household wiring contains 100g of copper (sorry I don't have suitable wire charts handy to give the correct figures, all further numbers will be more accurate), and the wire run from your light switch, up the wall across the ceiling and into your light fixture is 10m. That means 1kg of copper separates the switch from the light. Since copper has an atomic weight of 63.5, 1 kg of Cu is 15.75 moles; since copper is atomic number 29, and there are 6.02x10^23 atoms in a mole, 15.75 moles of copper contains 2.75x10^26 electrons. Let's say the fixture hold two 60W light bulbs @ 120V. That's one amp of current or one coulomb/second of electrons flowing in the wire. 1 Coulomb is 6.2x10^18 electrons. so it would take (2.75x10^26) / (6.2x10^18)= 4.4 x 10^7 sec (1.4 years) on average for an electron that leaves the switch to enter the light fixture. even if we only count the 11 electrons in copper's outer shell, instead of all 29 of copper's electrons, it'd still be over 6 months -- but I bet your light turns a lot faster than that!
In fact, household wiring is AC (alternating current), with the direction changing 60 times a second (in the US; 50 times a second in many other countries). In other words, the electrons barely have a chance to move at all, before they have to turn around and move in the other direction. In a 120V DC (direct current) lighting circuit, the electrons would be moving 10m/1.4 years (0.22 microns/second), but with the AC power most of us have at home, the electrons don't have any net motion at all.
This is a greatly simplified illustration, but with it in mind, here's how a capacitor works:
If we place a single free electron on one plate of a capacitor, it will be an unbalanced negative charge in a very small spot -- an intense electric field. This field will tend to repel other nearby electrons: a force that is absolutely palpable -- in fact, the repulsion of the electrons in the atoms of your skin against the atoms of all other atoms in all other objects are responsible for the "solidity" of every hard object you've ever felt. Next time you stand up and bang your head on an open cabinet, try to disregard it as "mere electron repulsion". I haven't had much luck with that, personally.
Those nearby electrons, which are directly repelled, will in turn repel electrons that are further away, and so on. Of course, all those electrons will also repel back. In addition to repelling other electrons, it will tend to attract positive charges. In actual wires, these "positive charges" tend not to be actual positive charges (like protons, which are nicely shielded by their atom's shell of electrons) but "virtual charges" called holes. A hole ia a place where a negative electron Would be, in a perfectly neutral mass of atoms, but *isn't* (perhaps due to the diffuse repulsion of distant surplus electrons)
I don't want to dwell on that, but it'll be useful when you get further in your study of electronics, such as understanding semiconductors like transistors: the free electron we added creates an "induced charge" on the opposite plate. Beginner texts often call these positive charges, but they are actually just "virtual charges" or holes cause by the electron's repulsion. Holes *act* like positive charges but aren't physical particles.
Okay, so we have this capacitor, charged with a single free electron. The effect of the one negative charge almost instantly spreads across the entire capacitor. This is a less repulsive (lower energy) state than having all the effect in one tiny spot. In fact, the energy is so low that it's easy (requires little energy) to add a single electron.
You may also be interested to know that the charge effect diffuses (equilibrates) over the entire system at "the speed of light" but that's not the speed of light (c or the ideal speed of light in a vacuum) that we often talk about. It's the speed of light in copper -- which can be described as the time it takes all those electric field interactions to interact and spread. On a subatomic level, the pure EM field of an electron moves at c, but on a large scale, electricity in a wire travels somewhat slower [ but not nearly as slowly as the actual electric flow
If there is a [conductive] defect in the capacitor's insulation. our one free electron will nudge (indirectly) some electron from the negative plate into the hole [virtual charge] on the other side. This same indirect nudging will equalize the charges between sides if we use a wire to connect the two sides, instead of a defect in the insulation. Because the charge is now more spread out, the electric field has less stored energy.
Now, as I mentioned up top, real electric currents are *entirely* nudging. Any actual movements of physical electrons is tiny. The nudging starts at the power plant's generators, and is performed by magnetic fields which are moved by engines, like steam or water turbine, driven by things like fossil fuels or water pressure.
So in a real capacitor, there aren't really physical electrons entering the capacitor, or physical positive charges on the opposite plate. It's all electromagnetic nudging. On a subatomic scale, you can think of the repulsion as tiny springs between the electrons, on a slightly larger scale, you can think of changes in water pressure in a pipe. The "induced charge" on the other plate is the same thing -repulsion- just harder to imagine. There are no actual physical point-like positive charge involved, just some places that are less negative than others, and therefore *relatively* positive compared to the overall circuit. [You could trace the circuit back to the power plant, and reference it to various local grounds, as required by the Building Code -- but trust me, that won't help you understand capacitors any better, and I know some electricians who don't "get it"]
On the scale that we humans live on, we can pretend the electrons are actually flowing in pipe-like wires. It's really not such a bad model, but don't take it literally.
