In article <1996Oct3.105034@alder.cc.kcl.ac.uk>, udee785@alder.cc.kcl.ac.uk says...
[other posts quoted are deleted]
>It is possible, just possible, that the lenses
>are tested on a body slightly misaligned at the
>mount or mirror box or film plane. If many
>lenses show uneven unsharpness at different
>corners and the pattern persists (like all
>lenses worse on the left upper than the lower
>right) the body may need to be checked. On the
>other hand, on such a body all lenses are tilt
>lenses which may be useful to some degree so
>one may not want to rectify the problem. Such a
>defect also will cause more trouble with wide
>angles than teles.

All of the above is true, but I have never seen such
misalignment in a Nikon body, and I have checked MANY.
The machined parallelism in the body is accurate, and
much more easily achieved than in a lens where
parallelism depends on MANY different factors, such as
correct centering of all elements, correct parallel
mounting of all elements, correct parallel mounting
of the lens cell(s), correct parallel extension of the
focus mount, correct mounting of the bayonette, etc.
It is a wonder that most lenses don't show obvious
misalignments, but most Nikkor MF primes don't, the
majority of Nikkor MF zooms don't (though a fair
number do...), but a high percentage of AF zooms do
show some easily detected misalignment (I have not
checked a lot of AF primes, but most of the ones I
have checked were OK).
Hope This Helps