I definitely do (consider ray tracing to be "physics"). The more sophisticated ray methods (see, e.g., P. L. Marston, “Geometrical and catastrophe optics methods inscattering,” in Physical Acoustics, A. D. Pierce and R. N. Thurston,eds. (Academic, New York, 1992), Vol. 21, pp. 52–148 for an overview) explicitly state the conditions you mention. Sir Michael Berry is especially well known for addressing those issues.
/************************************
Down with categorical imperative!
flutzpah@yahoo.com
************************************/
----- Original Message ----
From: "LaMontagne, Bob" <RLAMONT@providence.edu>
To: Forum for Physics Educators <phys-l@carnot.physics.buffalo.edu>
Sent: Thursday, April 10, 2008 1:09:18 PM
Subject: Re: [Phys-l] paraxial approximation
Do most of you really consider ray tracing to be "physics"? The essence
of the functioning of an optical device seems to be to get the light
emitted from a point source and travelling through all possible paths
through the device to all arrive at some other point in space at the
same time. Ray tracing is a good aid to finding the point where the
image for a point object may appear, but it doesn't really seem to
demand that the light following various paths has to interfere
constructively at the image - at least not explicitly. In other words,
in the usual way it is implemented, ray tracing gives a necessary
condition for an image to appear, but not a sufficient condition. I
realize that it is possible to demonstrate that the paths indicated by
the rays are isotemporal, but I have never seen a textbook that
emphasizes that point (except through a mention in a few texts that
Fermat's principle leads to the law of equal incidence and relection
angles for a plane mirror.)