In case nobody has answered this question yet, I will
give it a shot.

        First, let's review the practical knowledge you need to
know to understand the answer to this question and to correctly
answer test questions on the subject.

        The one thing to remember is that capacitance is measured
in Farads.  In order to correctly answer test questions, you must
remember that capacitance in parallel is simply found by adding
the values of all the individual capacitors that are in parallel
in the circuit.  Capacitors are measured in Farads and parts of
Farads such as microfarads and pikofarads and it is fair to add
caps in parallel.

        Now for your actual question.  Let's make this as simple
as possible.

        Anyone who has seen a classic air variable capacitor
knows that there is a shaft where the knob or dial cord pulley
goes and that shaft has a bunch of roughly half-moon shaped metal
leaves or plates attached to it so they rotate and fit neatly
down into a matching set of stationary plates that seem to be
part of the body or frame of the variable capacitor.

        The movable plates are called the roter and the fixed
plates are the stator.

        The shaft and its roter plates are electrically separated
from the stator plates.

        The value of an air-dielectric capacitor is determined by
the surface area of both the roter and stator plates, the
distance between them and by the dialect Ric constant of air.  The
smaller the spacing between the roter and stator, the higher the
capacitance until the plates touch and then it is a short.  If
that happens, the capacitor is no good any more.

        By turning the shaft, you move the roter so that it is
out in the open, sticking above the body of the capacitor.

        Since the roter and stator are no longer facing each
other, the capacitance is much, much less.  As you turn the shaft
and the rotors mesh in with the stator s, the capacitance rises.

        We say capacitance is greatest or highest when the plates
are meshed.

        An ordinary air variable capacitor is just ten to fifteen
1-plate variable capacitors in parallel, all tied to one central
shaft.

        They do it that way to increase the capacitance range.
Let's say that each roter and stator pair has a capacitance of 1
picofarad when meshed and half that value when open or turned so
that the moving plate sticks up above the stator.

        Now imagine a second pair of plates on the shaft that
make a parallel capacitor also.  When meshed, the two capacitors
add together to make 2 pikofarads and when open, they add
together to make 1 PF.

        Now, imagine a normal-looking air variable with fifteen
pairs of plates.

        When meshed, they have a value of 15 pikofarads and when
open, they equal 7 pikofarads.  Instead of only half a picofarad
difference between the meshed and unmeshed state, we can now vary
the capacitance by 7 whole pikofarads.

        I hope that answers your question.

        To avoid confusion, here are a few other bits of variable
capacitor wisdom.

        While a capacitor with closely-spaced plates can produce
a lot of capacitance and a fairly wide range of variation at
that, it can't handle very high voltages between the roter and
stator without arcing and ruining itself.

        That's why the plate tuning capacitor in a transmitter
may have big gaps between the plates compared with a tuning cap
out of an AM radio.

        Since the bigger gap or spacing means less capacitance,
the engineers may build the plates larger in size or put more in
the stack to still get a wide variation.

        You may also run across air variables whose plates are
kind of odd-shaped like tree leaves rather than half-moons.
These capacitors are designed to vary uniformly in capacitance as
you turn the knob.

        Finally, a bit of trivia.  The dielectric constant of air
isn't totally constant.  Changes in humidity, for example make
small changes in this constant so there are actually humidity
sensors which are basically fixed-value capacitors with air as
the dielectric.  They are supposed to be a certain value when the
air is dry and a slightly different value when the humidity is
high.  Since humidity is relative,electronic humidity sensors
have to know the temperature and compare that with a table of
values of what capacitance should be present at a certain
relative humidity.

Martin McCormick WB5AGZ  Stillwater, OK
OSU Center for Computing and Information Services Network Operations Group