Unless you already know what's going on, it's probably hard to understand those words. So here are some illustrations.rjlittlefield wrote: It is usually more convenient to stop down the rear lens, but that practically guarantees that the light rays forming the image have been selected from the worst sections of the front lens.
First, here is an overview diagram of the lens system I'll be using. From left to right, we have an object, a triplet acting as the "Front Lens", a singlet acting as the "Rear Lens", and an image. The focal lengths and distances are set up to give 2:1 magnification.

Now, zooming in on just the lenses, let's consider what happens when we stop down at various positions. Nothing changes except the stop position.


Comparing the two images above, it should be apparent that stopping down at the rear lens selects rays that have gone through more peripheral areas of the front lens. (Note especially the strong concave element in the middle of the triplet.) For the on-axis ray bundle, the stop position makes no difference, but off-axis bundles are pushed into more peripheral sections of the front lens. If the stop is placed even farther back, as it would be with real rear lenses, the effect becomes even more pronounced.
Conversely, stopping down inside the front lens causes off-axis ray bundles to pass through more peripheral parts of the rear lens.
An optimal stop position would compromise between these effects, placing the stop in some intermediate position so as to avoid the extreme periphery of all lenses.

Unfortunately, the stop position that would give the best image quality is also the least convenient because commercial lenses certainly don't have a diaphragm at that position and in many cases the optimal position is not accessible to add one.
--Rik
Edit: July 21, 2010, to change title and move this topic to become a FAQ.