Jar Supplier Transition — Three Rivers → Owens-Illinois

Why the H&H collection contains so many mismatched lid-and-jar pairs — and why the mismatch is documentary evidence, not collector contamination.

The pattern

The Crystalvac glass-jar line ran from June 1932 through at least 1947. Across that window H&H drew from at least two primary suppliers:

Three Rivers Glass Company — exclusive (or nearly so) supplier from the June 1932 launch (250,000-jar initial order) through the December 1936 Ball Brothers reorganization. Documented marks: "3 Rivers" star on heel or base, 601-1 / 601-3 / 601-4 / 601-5 / 601-7 patent-stack variants.
Owens-Illinois Glass Company — picks up the line after the Three Rivers physical plant winds down (post-late-1937). Owens-Illinois mold marks appear on later-period Crystalvac blanks per the Crystalvac Jars page.

A Ball Brothers overlap window exists (Dec 1936 → Nov 1937, when the Three Rivers physical plant ran under Ball ownership fulfilling existing orders). Some pieces from this window carry both the "3 Rivers" mark and the Ball logo — see Three Rivers Glass Company item #37 (Ideal Bottling Co.).

The continuity decision (oral history)

In the 28 May 2026 Draves listening session, the curator (Brett Elmendorf) explained the lid/jar mismatch pattern to Nancy and Tim Draves while walking through Nancy's mystery H&H jar:

"H and H knew Three Rivers were getting shut down. They switched manufacturers and they got the holes in the [lid]. They kept the threading the same, so that's why you'll find all kinds of [lids]. ... I don't know what they actually changed, [but I've] collected most of [them]."

Transcript: raw-archives/oral-history/2026-05-28_nancy-tim-draves-listening-session.transcript.turns.md (~02:25 timestamp range). The phrasing is Whisper-noisy in places ("leaves" → "lids," "cedar" → unresolved); the substantive claim is the deliberate threading-spec continuity.

The curator-side reading: H&H knew the Three Rivers shutdown was coming (consistent with the 1932 receivership, 1933–1936 multi-year closure, and 1936 Ball reorganization — all documented in the Three Rivers Glass Company timeline). Rather than re-tool the entire packaging line, H&H specified the new supplier (Owens-Illinois) to match Three Rivers' thread pitch and finish diameter. The hole pattern on the lid (likely the small air-relief or sealing perforations) changed; the mating surface did not. Old Three Rivers lids therefore continue to fit new Owens-Illinois jars and vice versa — which is exactly the mismatch pattern collectors find in the wild.

Why this matters for the collection

A jar with a "3 Rivers" star base and a lid stamped for the post-1937 hole pattern is not contamination — it is documentary evidence of H&H's transition-window inventory mixing. Catalog both pieces; do not assume the lid is from a later piece.
A jar with an Owens-Illinois mold mark and an older Three Rivers-era lid is similarly correct. The threading continuity means H&H likely consumed remaining lid stock against new jars rather than discarding it.
The deliberate-continuity finding strengthens the supply-chain professionalism narrative on the corporate page: H&H tracked its glass supplier's financial condition closely enough to engineer a no-disruption switch. This is consistent with G. P. Menger's 1920 forward-buying discipline after the Santos Fours crash — same operational mindset, different commodity.
It also constrains the dating of any individual jar: a 3-Rivers-marked body is a 1932–1937 piece; a 3-Rivers-marked lid could be 1932 through whenever the lid-stock was consumed, possibly into the early 1940s.

The bail-wire → wood-handle → straight-wire transition

A parallel transition runs on the handles of the larger pail-format jars, also discussed in the same session:

Early period (1932 onward): bail wire (curved heavy-gauge wire formed into a carrying loop, both ends pinned through the jar's neck or shoulder lugs).
Mid period: wood handle added on a metal core through the bail — a comfort/grip improvement.
Late period: back to straight wire without the wood handle.

Brett's note: "So some of the bail-wire handles got a flat spot because the [wood] handle would hold it, but at some point they got rid of [the wood]. But I don't know why." The flat spot on surviving bail wires is the diagnostic feature — a bare bail wire showing a flat impression where a wood handle once was is a mid-period piece that lost its handle, not an early-period piece.

Sequencing the wood-handle window precisely against the Three Rivers → Owens-Illinois supplier transition is open research; the two transitions may be coincident or independent.

Open research