Two notations, and ten orphans
On this page11
Every container here that records a neck finish uses the dash form
48-400, 53-400, 58-400, 63-400, 70-400, 75-400 and 89-400, plus one row reading 38DBJ. Not one container uses a millimeter notation.
Ten closures use the millimeter form
One at 38MM, one at 48MM, one at 53MM, three at 58MM, two at 70MM and two at 89MM.
And five of those six values exist on this shelf in the dash form as well
48, 53, 58, 70 and 89 all appear as a dash notation on the container side. So the same nominal size is written two ways on one shelf.
All ten of the millimeter closures are recorded as fitting nothing here
Which is the consequence. A match on the notation string cannot connect 58MM to 58-400, so ten closures corresponding to sizes well represented here sit unattached.
Nine of the ten are sealing discs
A sealing disc is a flat part that corresponds to a neck size rather than threading onto one, so the millimeter form is a reasonable way to describe it. The notation is not wrong; it is a different convention.
Of the fifteen closures that fit nothing here, ten are those rows
And the remaining five record no neck finish at all: a dome lid, a dairy lid, two lip balm tube caps and a deodorant lid. Those five do not thread onto anything.
So the fifteen split cleanly into two causes
Ten are a notation mismatch, which is a record problem with a fix. Five are parts that genuinely do not pair with a threaded neck, which is not a problem at all.
Which is the distinction worth carrying away
An unmatched row can mean the data uses two vocabularies or it can mean the relationship does not exist. Counting unmatched rows without separating those two overstates the gap by two thirds here.
And the rule above shows what the fix looks like
For compliance purposes, the English dimensions shall be used. The metric approximations are included only for convenience. One sentence naming the governing form and demoting the other to a synonym.
Nothing on this page asserts that any pair fits
A 58MM sealing disc and a 58-400 jar correspond in nominal size according to the record. That is not a statement that they have been put together, and no page in this library makes one.