Hey Ed!
I read your attachment. Its not altogether clear what the writer was
describing but if its what it sounds like then I have seen this effect
often and its normal. But of course it still is usefull. If the writer
means that the polarity flipped on one of the magnets then this is easy
to demonstrate. Especially if you have a very strong magnet such as a
rare earth type and the other is a weaker ceramic type. One can take
lets say a small ceramic bar magnet and a samarium cobalt bar magnet off
the same size and bring the like poles together. They will repel until
they get close and then the polarity permanently flips in the ceramic
magnet and then they'll stick.
This is due to the magnetic property caled 'coercivity'. It is a measure
of the magnetic field density that is required to polarize a magnet. The
higher the coercivity is, the higher the field intensity required to
polarize the magnet to the magnetizing field. Rare earth magnets have a
higher coercivity than ceramic or alnico. Plus they have a very high
gauss magnetic field. So the high coercivity magnet will flip the
polarity of the low coercivity magnet. You can actually 'erase' or
demagnetize a magnet this way by carefully controlling the distance
between the magnets. A demagnetized magnet can be usefull for some
experimental work in that it is easier to manipulate weak fields within
it.
The other way to easily demagnetize a magnet is to heat it to its 'curie
point". This is by definition the temperature at which a magnet will
lose its field. Contrary to popular belief, when a magnet reaches its
curie point the field loss is irreversable. That is, it will not regain
its magnetization when it cools, UNLESS it cools within a magnetic
field. Then it will take the polarization of that field. This is one
posible way to make complex field shapes in a magnet. You don't need a
high strength field then to polarize even a high coercivity magnet. Just
heat it above its curie point then allow it to cool within a field of
your choosing.
Another possibility might be that the writer meant that the METAL PLATE
flipped. In this case it could be that the iron plate had become
magnetized by the first magnet. Then bringing the second magnet near
caused it to become re-polarized since the iron is of lower coercivity
than either of the magnets.
Maybe you aready know all this and were talking about something else. If
so let me know.
-Gerald