Re: volume polarization vs. surface charge

["John S. Denker" <jsd@MONMOUTH.COM>]

Hi Folks --

Donald wrote:
> > The mobile charges are not *in* the dielectric but *on* it

Then at 10:21 PM 3/3/99 -0500, Sam disagreed, saying:
> Charges don't like to move around on insulators...
> [you] polarize the dielectric, you don't charge the dielectric.

I say *both* viewpoints are correct (although we can't take both viewpoints
at the same time).

Consider the following arrangement of induced dipoles:
          +- +- +- +- +- +- +- +- +- +- +- +-
(a)       +- +- +- +- +- +- +- +- +- +- +- +-
          +- +- +- +- +- +- +- +- +- +- +- +-

That is macroscopically indistinguishable from this arrangement of surface
charges:
          + 0  0  0  0  0  0  0  0  0  0  0 -
(b)       + 0  0  0  0  0  0  0  0  0  0  0 -
          + 0  0  0  0  0  0  0  0  0  0  0 -

If we reverse the applied field, the induced charges will move:
 * in viewpoint (a) they each move a submicroscopic distance, but the
number of charges involved is very large (proportional to the volume).
 * in viewpoint (b) they effectively move a macroscopic distance, from one
surface to the other, but the number of charges is much smaller
(proportional to surface area).

=========

(Of course, in addition to the "del dot P" charges discussed above, you can
inject real, unbalanced charge onto and into an insulating dielectric, as
we discussed a few weeks ago.)

Cheers --- jsd