7/4/96 I noticed a misleading part in my explanation.  It is impossible to
split and then recombine colors with two IDENTICAL prisms.  If two
DIFFERENT SHAPES of prism are used, it is still impossible to combine the
colors back into a white ray, but there is a way to focus the colors all
into one spot, so they can be recombined upon a white screen. 

The prism stuff is a bit hard to discuss without graphics.  Look at this:

      WHITE            |\
 ---- LIGHT > ---->----|--\____
      RAY              |    \ ===-----_____
                       |      \   ---____  -----_____       
                       |________\        ----____
                         prism                   ----

The white ray goes through the prism and is converted into a spreading
cone of colored light.  In order to mix the colors together into a white
ray, their spreading must be reversed so that they concentrate to a point.
And then right at that point, the spreading must be cancelled out, so that
they all move together and form a white ray again.

But just one prism can SPREAD the colors, no?  Shouldn't it be able to be
run in reverse?  If it can fan them out, it should be able to un-fan them
again.  Yes, this is partly true.  If a second prism is placed opposite
the first, it reverses the spreading effects caused by the first.  But the
colors have been spreading for awhile.  After they have spread, a second
identical prism can change their angles back to zero again, but this
simply makes the colored rays travel parallel.  They are still separate
but no longer spreading. 

      WHITE            |\
 ---- LIGHT > ---->----|--\____                 _________
      RAY              |    \ ===-----_____     \        |
                       |      \   ---____  -----__\______|______Violet end
                       |________\        ----____   \    |
                         prism 1                 ---- \__|______Red end
                                                        \|
                                                     prism 2

The above is true of identical prisms.  What if we are allowed to use
different shapes of prism?  See below.

        |\                                                      ||
--->----|--\____                 _________________              ||
        |    \ ===-----_____     \                /    ____-----*|
        |      \   ---____  -----__\___________ /__----___---   ||
        |________\        ----____   \        /  ___---         ||
          prism 1                 ---- \____/----               ||
                                         \/                  WHITE
                                      prism 2                SCREEN

The first prism splits the white ray into spreading colors.  The second,
thicker prism not only stops the spreading, it reverses it entirely so
that the colored rays approach each other.  They come together at one
particlular place.  If a white screen is put right there, the colors will
combine to form white.  If the screen is placed too far or too close to
the prisms, the "white" spot will be missed and colors will appear on the
screen.

One final thing.  If we put two identical prisms together as below, we do
get a white light ray. But this is cheating, since it doesn't recombine
the colors. It prevents them from being separated in the first place.

                        _________ prism 2
      WHITE            |\        |
 ---- LIGHT > ---->----|--\------|------->-----
      RAY              |    \    |
                       |      \  |   
                       |________\|
                         prism 1

(((((((((((((((((( ( (  (   (    (O)    )   )  ) ) )))))))))))))))))))
William J. Beaty                            SCIENCE HOBBYIST website
billbeskimo.com                            http://amasci.com
EE/programmer/sci-exhibits   science projects, tesla, weird science
Seattle, WA  206-789-0775    sciclub-list freenrg-L vortex-L webhead-L