Richard Clark wrote:

On Fri, 29 Jul 2005 16:58:29 -0700, Jim Kelley
wrote:

It is quite evident that through the actions of the first interface,
that there is less energy incident upon the second interface.
Further, given that both interfaces operate with identical reflective
and transmissive properties, it follows the second interface could not
reflect enough to totally negate the reflections of the first.

True for any one reflection.


Hi Jim,

And true for ALL accumulated reflections there after. Reflections do
not add any energy to the cup when the first interface is draining it
more quickly.


Yes, that was my point. You seemed to have momentarily overlooked that
detail.

My analysis allowed ALL of the energy in the reflection from the
second interface ( 0.098X) to combine with the first reflection
(0.11X).

Single (first) reflection only.

This total superposition was both more than generous, and at
the same time very unlikely; and yet with this generous allowance
there is still excess reflection from the first interface.

Your generosity is hardly the issue. Previously reflections are more to
the point.

Hence for
something less than total superposition of ALL energies, it hardly
bodes a better yield in total cancellation - the energy just isn't
there in the first place. 0.098X 0.11X is the simple economics of
the balance.

Sounds good. It's wrong, but the sound of it is good nevertheless.

As an optical engineer, I've dealt with the harsh reality of this myth
of total reflection cancellation.

Failure can tend to make some people bitter. ;-)

Basically these claims are for first year students where demanding too
much inquiry would push them into switching majors to Business school.
Simple optics with simple, ordinary glasses exhibit quite useful
results, but they do not embody a proof. To anyone following the math
of my presentation, it is quite obvious what WOULD tend towards a more
complete cancellation - and such a subtle shift in the formula
diverges only slightly from the choir book hymn. It's not that hard
when the interface ratios drive the answer.

Another good sounding thing.

73, ac6xg