How to Read This Record

This is a map of the territory, not the territory. It says who the recurring names are, what the shape of the story is, where the sources fight, and which misreadings have caught people before — so that a question can be answered without flattening a distinction that took work to establish.

The one-paragraph version

Charles R. Tips founded the town of Three Rivers, Texas in 1913 and a glass factory there in 1922, on local sand and natural gas. The factory grew from a hand-blown shop into a mechanized plant and dominated the Texas commercial glassware market by the late 1920s. The Depression weakened it; the mortgage reached the competitor Ball Brothers Glass Company, which took the plant in late 1936, ran it for about a year filling existing contracts, and shut it down. Tips maintained to the end of his life that the closure was an act of anticompetitive destruction rather than business failure [@ev-0041; @ev-0039].

Nearly every clause in that paragraph is disputed somewhere below, and its shape — a prospering company undone by the Depression and then by a competitor — is itself contested by the federal antitrust record, which puts the first constraint on this company in October 1928 and in other hands [ev-0059].

The recurring names

Charles R. Tips, the founder. He founded the town, raised the capital, recruited the talent, ran the business, and lost the factory, and he left the most detailed first-person account of it in two letters written for the marker dedication [ev-0041]. His voice is the spine of this record, and it is recollection written roughly fifty years after the events.

Ball Brothers Glass Company of Muncie, Indiana, the largest commercial glass producer in the United States in this era. Ball took the Three Rivers plant in late 1936, ran it about a year, then closed it. Tips said the motive was to eliminate a price-cutting competitor [ev-0041]; evidence on Hartford-Empire licence restrictions and Ball’s demand notices bears on that motive independently of Tips [ev-0042].

Michael David Smith, author of Texas Glass (1989), the only book-length treatment of the company and the single most important secondary source here [ev-0039]. It is also the source most often in tension with the primary record, on dates, on names, and on the scope of the antitrust case. Treat Smith as authoritative on what he documents and check his dates against the contemporaneous material.

Kevin L. Mackey, collector and researcher, who assembled the article collection and annotations behind many of the primary clippings. His corrections to Smith and to Tips’ own memory are among the most valuable material held here, precisely because he flags where recollection diverges from the printed record [ev-0040].

Thelma Lindholm, the local historian who compiled the marker-research binder. She is not an author in the usual sense — she assembled and preserved the primary documents: Tips’ letters, county deed records, newspaper clippings [ev-0041]. The binder is a secondary artefact; what is inside it is primary, and the distinction matters when weighing it.

A. F. Tremblay, the engineer Tips recruited from a glass factory in Monterrey, Mexico to build the company’s own bottle-forming machinery. Every secondary source describes that machinery as getting around the Hartford-Empire patent monopoly. The federal antitrust record shows it did not: Tips wrote in May 1929 that Hartford had “notified us that the feeders we are now using infringe their patent rights. We have not contested this claim, but have agreed with them on a payment of release of damages” [@ev-0059, part 16, fol. 60713]. His name is spelled differently by nearly every source that mentions him — Tips writes “Trembley,” Smith “Harold Trembley,” the Bottle Research Group “Harold Trembly” [@ev-0041; @ev-0039; @ev-0042]. The variance is a useful measure of how carefully a given secondary source checked its work.

W. L. Moody III, the financier who bought $100,000 of company stock in 1929, the transaction that made Tips company president. He was of Galveston; Smith places him in Houston [ev-0039], which the primary material corrects [ev-0041]. His later bankruptcy during the Depression removed the company’s financial backstop.

William Church, the San Antonio attorney Tips says betrayed the company by tipping off Ball Brothers instead of securing the loan priority, named in both Tips’ account and Smith’s book [@ev-0041; @ev-0039].

Hartford Empire Company, holder of a near-total patent monopoly on glass-container machinery. It licensed Three Rivers in 1929, after serving an infringement notice the company did not contest [@ev-0059, part 16, fol. 60713], for beverage bottles, packers’ and prescription ware but not fruit jars — which Three Rivers was already making [ev-0042] — and, the exclusion that mattered most, not milk bottles, whose rights had been assigned to the Thatcher Manufacturing Company in 1920 [@ev-0059, part 09, fols. 10273-10274]. Three Rivers went on making milk bottles regardless, on one home-built feeder that Hartford left deliberately unlicensed and expressly reserved the right to sue over “at some time dictated by its own judgment” [@ev-0059, part 16, fols. 60862-60863]. That licensing system is the structural context for everything that follows, and it begins acting on this company in 1928, at the peak, not in the Depression.

Names to weigh lightly

Gutzon Borglum. Smith writes that the Mount Rushmore sculptor “joined the design staff” [ev-0039]. Nothing corroborates it, and the reading that fits the rest of the record is a conflation with his son Lincoln, whose Beeville ties date from 1938, after the factory closed. This should not be raised without new evidence.

James Kapp, also rendered Kepp, Capp, and Knapp, the first president of the company — a San Antonio furniture-store owner, not Tips, who became president only in 1929 [@ev-0041; @ev-0039; @ev-0042]. Four spellings across the sources is another name-reliability marker.

R. A. Ellerman, listed as agent for the receiver in the 1933 trade directory [ev-0042], and appearing in Tips’ account as an office and purchasing employee [ev-0041] — which suggests the receivership was run internally rather than by an outside trustee. That is an inference from two sources that do not mention each other.

The shape of the story

Knowing which phase a source is describing prevents most dating errors.

Founding and hand-blown era, 1921 to 1924. Tips advertises in a trade magazine, recruits glassmakers, incorporates in May 1922, begins production September 1922 [ev-0041]. The factory closes mid-1923 for about six months and reopens in January 1924 [ev-0042]. Glass is hand-blown; the plant is small and unprofitable.

Mechanization and boom, 1924 to 1931. Tremblay arrives and builds the company’s own machinery — which Hartford-Empire held, and Tips conceded, to infringe its patents [@ev-0059, part 16, fol. 60713]; automated equipment comes online in early 1925 [@ev-0041; @ev-0042]. Capacity roughly doubles yearly through 1928 — a second tank in December 1925, a third furnace in July 1926, a 125-ton furnace in 1928. A second plant opens at Bastrop, Louisiana, running from June 1929 to 1933. Sales offices open in Dallas, Houston, Oklahoma City, Little Rock, and New Orleans, and shipping shifts from rail to company trucks. The company claims over 75 per cent of Texas commercial glassware by 1929, and capital grows from $50,000 in 1922 to $600,000 in November 1930 [@ev-0041; @ev-0042].

The boom is also when the squeeze starts, which no secondary account records. From October 1928 Tips was negotiating to merge Three Rivers into the Thatcher Manufacturing Company; in November 1929 Thatcher’s chairman wrote privately that Hartford’s president “wanted to give them a franchise on milk bottles. I heard about it in time to talk with him and it was not done” [@ev-0059, part 16, fol. 60716]. Read this phase as a boom under a ceiling, not as an unobstructed rise.

Depression and receivership, 1931 to 1936. The company falls behind on wages — a March 1932 cullet pledge against $26,000 — and pays in scrip [ev-0041]. Bankruptcy proceedings begin in October 1932 and Ellerman is listed as the receiver’s agent by 1933 [ev-0042]. The Bastrop plant closes. Tips seeks a federal loan; Church carries word to Ball Brothers, which buys the mortgage notes [ev-0041]. The Three Rivers property is auctioned on 27 November 1936, with Ball winning at $80,000 [ev-0042].

Ball’s year and shutdown, 1936 to 1938. Ball restarts production on 8 December 1936, filling Tips’ existing contracts, and runs the plant through 1937 making soda bottles, prescription ware, packers’ ware, and fruit jars [ev-0042]. Winding down begins in late 1937 and the formal deed to Ball is dated 31 March 1938 [ev-0042]. The date of final cessation is disputed. The property was eventually sold in September 1954 [ev-0042].

Where the sources fight

When did the factory close? Four years are on offer. Smith gives 1937, apparently for the closing of the transaction rather than the last day of production [ev-0039]. The institutional accounts give 1938. Brantley gives 1939, which the Bottle Research Group judges most likely correct, and Toulouse gives 1947, which every other source rejects [ev-0042]. Neither Brantley nor Toulouse is held here; both reach this expert only through the Bottle Research Group’s account of them [ev-0042]. The December 1936 takeover and the 1937 contract-filling year are well documented; the date of final cessation is genuinely uncertain.

How did Ball get the factory? Tips describes a federal court foreclosure on the mortgage notes [ev-0041]. The Bottle Research Group describes a bankruptcy receivership ending in a November 1936 auction [ev-0042]. These may be two views of one process — a receivership and a mortgage foreclosure can coexist — but neither source mentions the other’s mechanism, and nothing held here establishes that they are the same event. Both are recorded.

How broad was the antitrust case? Tips names three defendants, a Toledo trial, and a Supreme Court appeal [ev-0041]. Smith names only Ball and mentions only a fine [ev-0039]. Both understate it. The case is Hartford-Empire Co. v. United States, and its printed transcript record on appeal to the Supreme Court, October Term 1943, is now held here as an OCR extract of four of its fifty-eight parts [ev-0059]. Ball Brothers is one defendant among many; the acquisition and closing of Three Rivers appears as one numbered item in the Government’s charge sheet, with a corresponding numbered demand for “An order directing Ball Brothers to divest itself of the plants and other assets acquired from Three Rivers Glass Company” [@ev-0059, part 18, fols. 64374-64375]. What is still unresolved is the outcome: neither the decision nor the decree is held.

Common misreadings

“Aqua glass.” Three Rivers made flint and amber. Aqua is the natural un-decolorized tint of iron-bearing sand glass, not a colour anyone chose; earlier suppliers’ versions of the same bottle designs are aqua and the Three Rivers versions are colourless [ev-0042].

“Green glass.” A Geology of Texas passage says the Carrizo sand could make “clear, green glass” [ev-0041]. It describes the sand’s potential, not the factory’s output. An “olive jar” in Smith’s price list is a jar shape, for olives [ev-0039].

“Borglum designed their bottles.” Almost certainly wrong, and it rests on a single sentence in Smith [ev-0039].

“First glass factory in Texas.” An accretion, and this expert can now trace how it grew. The 1973 state marker does not say it. Its inscription runs from “Opened 1922 by company headed by Charles R. Tips” to “had shown Texas potential in the industry,” and the word first appears nowhere in it [ev-0011]. Smith nonetheless reports that the Historical Commission erected the marker “commemorating it as the first glass manufacturer in Texas” [@ev-0038, p. 27; @ev-0039, p. 27], and elsewhere calls the plant “Texas’ first glass company” [@ev-0038, p. 16; @ev-0039, p. 16]. The Handbook of Texas then states it flat, with no attribution at all: “The first glass factory in Texas was built at Three Rivers in 1922” [ev-0004]. And the compiler who assembled the marker application cut the equivalent claim where she met it, correcting a reporter’s “first glass factory in the United States” to read “the first glass factory in Live Oak County” [@ev-0041, p. 42].

The chain is legible: the marker claims no primacy, a book attributes the claim to the marker, an encyclopedia repeats it as fact, and the person closest to the documents cut it to a county. What this expert cannot do is settle the underlying question. Nothing held bears on pre-1922 Texas glassmaking at all - no census of manufactures, no trade directory, no notice of an earlier Texas glass house, nothing. The claim is unsupported here, not refuted here, and the difference should be stated: this expert has never seen the evidence that would decide it.

“The plant closed in 1937.” This is the popular account and Smith’s [ev-0039]. The record points to December 1936 for the takeover and 1938 or later for the final shutdown [ev-0042]. 1937 is the year Ball ran the plant.

“The trouble started with the Depression.” Tips’ own account, and Smith’s after it, has a thriving plant undone after 1931 by hard times and then by one treacherous attorney. The federal antitrust record puts the first constraint in October 1928, at the company’s peak, and puts it in different hands: Hartford-Empire and Thatcher, not Ball Brothers [@ev-0059, part 16, fols. 60710-60716]. Ball is the last actor in this story rather than the first, and a source that starts the trouble in 1931 is describing the second half of it.

Treating agreeing sources as independent confirmations. This is the important one, and it is the reason the section below exists.

The source landscape

Primary material. Tips’ own letters of 11 October 1971 and 18 May 1973, the fullest first-person account, written roughly fifty years after the events and demonstrably imperfect on some details [@ev-0041; @ev-0040]. Period newspaper clippings preserved in Lindholm’s binder and Mackey’s collection, which are the best evidence for specific dates and figures [@ev-0041; @ev-0040]. Live Oak County deed records cited in the marker research, which anchor the legal timeline [ev-0041]. The American Glass Review yearbooks, which are annual snapshots of what the factory was producing but reach this expert only through the Bottle Research Group’s reproduction of them [ev-0042]. The Tremblay bill of sale of 15 March 1928, a physical document linking Tremblay, his patent, and the company [ev-0041].

Secondary material. Smith 1989, the most complete single narrative, held here in two independent transcriptions of the same book [@ev-0039; @ev-0038]. The Bottle Research Group chapter of 2019, which cites trade directories and Roller’s files and supplies the most precise dates, and which occasionally mischaracterizes Smith [ev-0042]. Mackey’s annotations, a correction layer over both Smith and Tips [ev-0040]. Lindholm’s binder, which is a collection rather than a narrative [ev-0041].

What the primary material physically is. The two bodies that carry almost all of the clippings above — Lindholm’s binder and Mackey’s collection — reach this expert as vision-OCR transcriptions. Behind them are ninety and eighteen TIFF page scans respectively, and this expert holds neither: both are recorded by reference to an operator volume that is not mounted and has not been located [@ev-0056; @ev-0057]. Nothing quoted from a clipping in this record has been checked against the image of the clipping. The custody entries note that a PDF of each compilation sits in this repository’s processed inbox, which is not admitted evidence and which this expert does not read — so the images are near at hand without being in hand [@ev-0056; @ev-0057].

There was a collection before this one. Writing to Lindholm on 11 October 1971, Tips opens with a sentence that is easy to read past: “I am afraid I do not have much beyond what I gave to Mrs. Adlof that I can tell you about the Three Rivers Glass Factory” [@ev-0041, p. 43]. Mrs. Arthur E. Adlof of Three Rivers had gathered material from him already, before Lindholm asked, and the survey committee’s own press notice names her as one of two people collectors should contact [@ev-0041, p. 2]. What she gathered is not in the binder and is not held here. The consequence is worth stating plainly: the fullest first-person account this expert has is explicitly a second telling, given by a man who believed he had already said most of it to somebody else.

The echo chamber. The secondary literature on this company is small and self-referential: the institutional entries, the marker text, the company’s present-day web presence, and the collector catalogues draw from Smith, from the marker text, or from each other [ev-0040]. They are useful for confirming what the conventional account says. They are not independent evidence for any claim, and counting them as such is the fastest way to manufacture confidence this record does not support.

What would settle the open questions

The closure date needs a document from 1938 or later that speaks to production rather than to ownership — a state factory inspection, a trade-press notice, a tax roll. The foreclosure mechanism needs the Live Oak County court file. Neither is held, and both are obtainable.

The antitrust scope was the third item on this list and is no longer open in the same way. The federal case record is now held [ev-0059], and it has already corrected the scope, the chronology and the cast of this story. What it does not close: fifty-four of its fifty-eight parts were never searched, no page images have been matched to the OCR, and the decision and decree are held at no level. Every locator in this record should be read as provisional until the images are matched.

A smaller gap would settle a great deal at once. Mounting the volume that holds the two sets of page scans would put an image behind every newspaper clipping this record quotes and every marginal note it reports, and would turn a machine reading nobody can check into one anybody can [@ev-0056; @ev-0057]. It is the only open item here that needs no archive, no purchase and no correspondence — only the disk.