Re: Why called diffraction?

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

On 04/23/2003 11:48 AM, Roger Haar wrote:
> 'Why the word "diffraction" rather
> than "interference?" '

>         My first guess is it is historical with the
> patterns seem experimentally first.  These
> patterns appear to be some kind of a
> generalization of the pattern from  diffraction
> gratings.  Thus the pattern seen by directing x
> rays through crystals was called diffraction.

Maybe, maybe not.
 -- Yes, crystal diffraction patterns have some
    properties in common with two-slit diffraction
    patterns.  Notably they both differ wildly from
    the ray-optics result.  They have maxima and
    minima as a function of angle.
 -- OTOH crystal diffraction patterns do not in the
    superficial sense resemble two-slit diffraction
    patterns.
 -- There's much similarity in the physical process.
    This is probably a better basis for choosing names.

>         In my mind, the concepts of interference and
> diffraction are deeply intertwined.  Is it a
> single slit interference pattern or a diffraction
> pattern?

We're getting into serious hair-splitting here.

You could say that _all_ physical optics is
interference, by invoking (informally) Huygens'
construction or (formally) Feynman's path-integral
formulation.

I call the one-slit pattern a diffraction pattern,
whenever the pattern is nontrivial (non-ray-optics).

Interesting bit of physics:  Consider a slit with
some apodization, i.e. at the edge it gradually
goes from clear via progressive shades of gray
before becoming completely absorbent.

*) In the near field, near the slit, you don't get
anything that looks like a diffraction pattern.
You just get a shadow (the ray-optics result).

*) In the far field, there is no shadow, just a
diffraction pattern.

*) It is quite fascinating to look at the result
as a function of distance from the slit, to see how
the shadow morphs into the diffraction pattern.

Back when I was taking first-year quantum mechanics,
this was assigned as a homework problem.  Alas,
the equation that describes the intensity pattern
(as a function of distance downrange from the
slit) is a mess.  I wasn't going to learn
anything by staring at it.  So I plotted it out
on the computer.  Contour plot.  Thing of beauty.
Wrote my name on it and handed it in.  The look
on the professor's face was priceless.

>         I cannot think of a case of diffraction without
> interference.

Agreed.

Perhaps the following definition will summarize
what's been said:

 diffraction is defined to be:
    interference of scattered waves,
    excluding the trivial case of raylike
    coherent forward scattering à la Huygens.

> But there is interference without diffraction.

Sure, just arrange a non-scattering situation.

> Examples include a water ripple
> thank with two oscillating objects in the water;
> twin radio antennae, and maybe thin film
> interference.

Agreed.

Other examples include
 -- VLBI
 -- Fabry-Perot interferometry

===================

The passage that Bob S. quoted from Lipson et al.
seemed unhelpful.

IMHO the standard usage of terms such as scattering,
diffraction, and interference are quite reasonable,
far more consistent than Lipson et al. give them
credit for.