My original comment about the finding of penicillin (marketed about 1950) in
a bottle that was made about 1920 led to some discussion about bottle reuse
and deposition lag.  Reuse of the bottle was only one possible explanation
of the time difference offered, and in this case, it was probably the wrong
explanation.  The penicillin was manufactured and sold by Charles Pfizer and
Co., a very large, reputable pharmaceutical firm, and it is unlikely the
company would have sold antibiotics in previously used bottles in 1950.  My
point was simply that one cannot assume that the date of manufacture of a
bottle corresponds closely to its marketing date.  I agree with Bill that
one should not try to infer too much from a single bottle.  Working in a
drug store museum with a few thousand paper-labeled bottles may make it
possible for me to find more cases where the paper label gives a  good date
on packaging to compare with the bottle manufacturing date range.
Incidentally, for those of you interested in reading more on bottle reuse in
the U.S. since 1800, I recommend the paper by Jane Busch, Historical
Archaeology, Vol. 21, No. 1, 1987, pp 67-80.