The Palestine Plant Was Knox, Not Owens-Illinois

Smith’s 1989 history says that when Three Rivers closed, “many of the former employees of the company moved to Palestine in East Texas, where a large Owens-Illinois Glass Company had begun operation” [ev-0039].

There was no Owens-Illinois plant at Palestine. The plant there belonged to the Knox Glass Bottle Co. of Mississippi, and it did not exist when Three Rivers closed.

This matters to the expert for two reasons. It is the destination claim in the company’s own dispersal story, so getting it wrong misdescribes what happened to the workforce. And the correction runs the other way from the usual folklore: it removes a local competitor that never existed, and replaces it with a plant that opened three years later.

What the Palestine plant actually was

Chester Underwood — brother of Knox founder Roy Underwood — opened it as the second “Southern Knox” plant after Jackson, Mississippi, under the subsidiary name Knox Glass Bottle Co. of Mississippi, eighth in the Knox chain [ev-0052]. It opened July 5, 1941, after only 95 days of construction [ev-0052]. A Glass Industry reporter wrote in 1942: “On a plot of ground that at this time last year was a cotton field, now stands a modern Knox Glass Bottle Company plant” [ev-0052].

A 90-ton natural-gas furnace fed “four Lynch 10 machines, a Lynch JP machine, and a Lynch JPM machine through six forehearths and feeders” [ev-0052]. In 1942 it made containers for “soft drinks, beverages, foods, cosmetics, pharmaceuticals, chemicals, extracts, wines and liquors, fruit, and, in fact, glass containers of every description”; twelve years on it was reported making liquor, wine, and pharmaceutical bottles along with miscellaneous and wide-mouth ware [ev-0052]. Ground was broken in June 1957 for a dedicated amber-glass factory, with the potential to raise output by about 40 percent [ev-0052]. At its height it employed almost 500 people and was “the town’s most important industry and one of the few glass houses ever built in Texas” [ev-0052].

It carried the T-in-a-keystone mark from 1941 until 1952, when the whole Knox system went to K-in-a-keystone. The T probably stood for Texas, “since the ‘P’ was already in use by the Wightman Bottle & Glass Co., at Parker’s Landing, Pennsylvania” [ev-0052].

Closing date, minor variance: the BRG plant table gives the Palestine plant’s life as 1941–1985 [ev-0052], while the state marker records that “the final glass jars manufactured here” bear the closing date December 20, 1984 [ev-0013]. Nothing held resolves the year, and nothing in this article turns on it.

Why the Owens-Illinois attribution fails

The complete O-I plant roster excludes Palestine. The Bottle Research Group’s table of Owens-Illinois plant numbers was compiled against O-I corporate records with two former O-I employees. It enumerates every plant number the company ever used, and it contains exactly one Texas location in company history: Waco, Plant No. 15, 1944 to present, which inherited its number from Okmulgee, Oklahoma — idle after 1932 and sold in 1939 [ev-0054]. Palestine appears nowhere in it [ev-0054]. Because the roster is a complete enumeration rather than a sample, its silence is close to the strongest form negative evidence takes.

The dates do not permit it. Smith’s sentence places an operating O-I plant at Palestine when the Three Rivers workforce dispersed at the end of 1937 and into 1938 [ev-0039]. The Palestine ground was a cotton field a year before July 1941 [ev-0052], and O-I’s Waco plant dates from 1944 [ev-0054]. Whatever the Three Rivers workers moved toward in 1938, it was not either plant.

The State of Texas says Knox. Marker #13794, “Site of Knox Glass Company Plant,” tells the Knox founding story at the site [ev-0013], and the Palestine newspaper’s 2006 dedication account, built on plant historian Jimmy Odom’s records, tells it the same way [ev-0014].

Tips had it right; Smith did not

Tips’ 1971 letter names both plants and attributes them correctly: “Owens-Illinois Glass Company built a big factory in Waco[;] another National Company built a plant in Palestine, which was mostly operated by skilled workmen they got from the Three Rivers factory when it was closed down” [@ev-0041, p. 44].

Knox fits “another National Company” well enough. It was a Pennsylvania- headquartered combine operating in Texas under the geographically confusing name Knox Glass Bottle Co. of Mississippi [ev-0052]. Writing fifty years later, Smith appears to have filled that blank with the industry’s household name.

This is one of several places where Tips’ first-hand account survives contact with the record better than the published history that drew on it.

The worker migration: real, later, and documented in one case

Tips’ claim that the Palestine plant was “mostly operated by” ex-Three Rivers men is not corroborated by anything held from the Palestine side. Neither the state marker [ev-0013] nor the 2006 dedication retrospective [ev-0014] mentions Three Rivers at all; the local account emphasises that local men were hired to build the plant and many stayed on as employees [ev-0014]. That does not contradict Tips — construction labour could be local while skilled glassmen came through the company’s own network — but the broad claim rests on Tips alone.

One case is documented end to end, and it runs through Knox rather than directly to Texas. Mold designer Carl Stendebach, born December 26, 1910, “worked as a mold designer at the Three Rivers Glass Factory until it closed in 1937. He and Daisy then moved to Pennsylvania where he worked for a glass factory. Later, they moved to Palestine, where they lived the rest of their lives” [ev-0015].

His WWII draft card, registered October 16, 1940, names the Pennsylvania employer: the Wightman Bottle & Glass Co. of Parkers Landing, Armstrong County [ev-0021]. Wightman had been a Knox affiliate since 1921, bought and rebuilt by Roy Underwood after a 1929 fire and renamed Knox Glass Bottle Co. of Parker in 1952 [ev-0021]. Stendebach was therefore inside the Knox organisation nine months before Knox opened Palestine [ev-0021]. The 1950 census finds him at Palestine, Anderson County — age 39, shop foreman at a glass company, 62 hours in the prior week, with Daisy and Carl Jr. [ev-0021].

Two cautions on this chain. The draft card gives his birthplace as San Antonio where the family history says Three Rivers [@ev-0021; @ev-0015] — a variance in the record, not a break in the chain. And one man’s route is not a migration: it establishes that the pipeline existed, not how many travelled it.

The chronology also corrects the shape of the story. Since the plant did not exist before July 1941 [ev-0052], any Three Rivers to Palestine move dates from 1941 or later — three years and more after the closure, not the immediate transfer Smith’s compressed sentence implies [ev-0039].

There is a small irony in the marks. Palestine took the T-in-a-keystone because Wightman already held the P [ev-0052] — Wightman being the plant where the ex-Three Rivers mold designer was then working [ev-0021].

A loose end for the marks article

The Knox monograph cites a Mexi-Pep bottle base made by Three Rivers Glass Co. carrying a “31” date code for 1931 [ev-0052]. That sits against BRG’s statement elsewhere that Three Rivers did not use date codes [ev-0042]. The conflict is recorded here and belongs to the maker’s marks article; nothing in this one depends on it.