Pyrex: The Mixing Bowl That Was Never Actually Pyrex Glass

By Matthew Partsch · Owl and Twig Vintage · October 1, 2026

A 1919 patent protects a glass formula for both the lab and the kitchen. The colorful mixing bowl most collectors actually chase was never made from it.

What the 1919 patent actually covers

Corning research chemists Eugene C. Sullivan and William C. Taylor filed U.S. Patent No. 1,304,623 on June 24, 1915, with a divided continuation filed September 18, 1916. The patent granted May 27, 1919, describing a glass composition with a linear coefficient of thermal expansion under .000004, more than 75 percent silica, and specific ratios of silica to base and acid to basic oxides that gave the finished glass unusual resistance to sudden temperature change and chemical attack.

The patent's own language names both uses the resulting glass was meant for: laboratory apparatus and baking dishes, in the same set of claims. The kitchen application was not a later spinoff of a lab formula filed years afterward. Sullivan and Taylor wrote cookware into the patent from the start, which matches the condensed timeline of the company's own account: the glass existed as a laboratory material first, and Corning moved quickly once it saw a kitchen use for it. The patent has been in the public domain for decades; its 17-year term under the law in force when it was granted expired in 1936.

The 1914 kitchen accident behind the brand

The company's standard account, repeated by the Smithsonian and by Corning's own museum, begins in 1914 with Bessie Littleton, whose earthenware casserole dish cracked in her oven. Her husband Jesse Littleton was a physicist at Corning Glass Works, and when she asked whether the company's heat-resistant glass might hold up better, he brought home a glass battery jar used in industrial testing, sawed off the metal top, and gave it to her. Bessie baked a cake in it. The experiment worked well enough that it helped push Corning toward a consumer cooking line, and the company credits her test as the direct push behind what became Pyrex.

That is the version every institutional source repeats, and this guide repeats it the same way, with one honest caveat: it is Corning's own telling of its own origin, not an outside account that can be checked against independent witnesses the way a court record can. One source on Corning's internal history mentions that heat-resistant glass had already seen some internal testing for cooking uses as early as 1913, a year before the celebrated cake. Whether that earlier testing and the Littleton cake are the same thread of research or two separate sparks that both happened to point the company the same direction is not something the available sources resolve.

Corning launched Pyrex ovenware commercially in 1915, with the first retail sale reported at the Jordan Marsh department store in Boston. The name itself, according to the company's own later trademark filings, is an invented word with no Greek or Latin root: an "r" inserted between "pie" and "ex" purely for the sound of it, not a reference to the Greek "pyro" for fire, whatever folk etymology has grown up around it since.

Why the trademark paper trail doesn't match 1915

Secondary sources routinely state that Corning trademarked the Pyrex name in 1915, alongside the patent filing. A federal trademark search run for this guide could not confirm a registration anywhere near that date for cookware. The earliest confirmed PYREX federal registration this search located is No. 0195029, filed June 23, 1924, claiming first use in commerce of November 3, 1920, and registered February 17, 1925. Its goods description covers condenser lenses, light filters, mirror blanks for scientific apparatus, navy sight glasses, and vacuum pumps for laboratory use. That is a registration for scientific optical glass, not for bakeware, and its first-use date of late 1920 is five years after Corning's first retail sale of Pyrex ovenware.

This does not mean no earlier cookware trademark filing exists. Early 20th-century registrations lapse, go unrenewed, and get refiled per class across the decades, which is the ordinary mechanical reason a company's oldest mark can be hard to locate in later federal records, the same pattern this series has already documented for Crayola's trademark history. A records search can confirm only this: the commonly repeated "trademarked in 1915" claim does not match any registration this search could locate, and the one Pyrex trademark record with a dated paper trail from that general era covers telescopes and lab glass, not casserole dishes.

The bowl you collect was never borosilicate

Pyrex's reputation rests on one property: a borosilicate dish can go from a cold refrigerator into a hot oven without cracking from the temperature swing, something an ordinary soda-lime glass cannot reliably do. Collectors commonly repeat a simple rule for telling which kind of glass they are holding: an all-capitals PYREX stamp means old borosilicate, and a lowercase pyrex stamp means new soda-lime. America's Test Kitchen has published exactly that rule in print.

The clear glass side of the business is a separate, more complicated story. Corning used true borosilicate for most of its clear ovenware and measuring cups for decades, alongside other formulations including aluminosilicate glass at some plants, before gradually shifting toward tempered soda-lime glass across its tableware line. By the time Corning sold its consumer products division in 1998, the switch to soda-lime was largely complete for everything except a handful of holdouts, reportedly including the classic 13-by-9-by-2-inch baking pan. The buyer, later renamed World Kitchen and then Corelle Brands, kept making Pyrex-branded glass in the United States from soda-lime formulations going forward. Pyrex made today in France, by contrast, under Arc International at a plant in Chateauroux, is still borosilicate, and still stamped in capital letters, which means a brand-new French import can legitimately wear the same all-caps mark collectors associate with a seventy-year-old American bowl.

The soda-lime shift was not a simple downgrade. The U.S. Consumer Product Safety Commission fielded 66 complaints about shattering Pyrex over roughly a decade and concluded the glassware did not present an actionable safety defect. Consumer Reports testing in January 2011 confirmed that borosilicate does resist thermal shock better than soda-lime, but researchers studying the broader injury picture have noted that soda-lime's better resistance to breaking from an ordinary drop may represent a real net safety benefit, since dropped and broken glass causes far more injuries in practice than thermal-shock failures do.

How to read the backstamp and date a bowl

Pyrex changed its backstamp several times across the 20th century, and the stamp itself is still the fastest way to place a bowl within a rough date range.

Era	What the backstamp looks like
1915 to late 1930s	A simple circular mark reading PYREX alongside a Corning monogram, with no model number.
1940s	PYREX in a circle with "T.M. REG." arched above the name and "U.S. PAT. OFF." below it.
Around 1950	Adds a registered-trademark symbol, the words "TRADE MARK" above and "MADE IN U.S.A." below, plus a model number.
Late 1960s	Shifts from the circular format to a straight-line text stamp.
Mid-1970s onward	Adds a metric capacity figure alongside the U.S. measurement, and later substitutes "by CORNING" for "MADE IN U.S.A."

Printed decoration gives a second, independent date floor. Pyrex did not put a decorative pattern on its opalware line at all until 1956, so any pattern-printed piece postdates that year by definition. The nesting mixing-bowl shape with a spout and a small handle on each side, the shape collectors call Cinderella, was designed by John Philip Johnson and introduced in 1958, which puts a floor under every Cinderella-shaped bowl regardless of its pattern. The Corning Museum of Glass's own pattern library dates Butterprint, one of the first printed opalware patterns, to a 1957-to-1968 production run, starting with turquoise-on-white and white-on-turquoise in its original 1957 release before later orange-and-pink and white-on-pink promotional colorways followed.

Mold numbers add a third check. Any four-quart #404 bowl in a color other than yellow was made no earlier than 1949, since yellow was the only color offered in that size before then. The earliest #402 bowls, the one-and-a-half-quart size, were made in an orange-red shade advertisers called "Chinese Red," a hue distinct enough from later reds that it can place a bowl in the line's earliest production run on its own.

How to tell a real one

• Match the backstamp era to the claimed pattern. A bowl stamped in the late-1960s straight-line format cannot be wearing a pattern the Corning Museum of Glass dates to 1957, and the reverse is just as telling: treat a mismatch between stamp era and pattern date as a reason to look closer, not a minor detail.
• Check that an entire nesting set wears consistently. A genuine Cinderella set that has lived together for decades usually shows comparable wear, comparable fading, and a matching mold-number sequence across all four bowls. A set assembled from mismatched sources to look complete is common in the resale market and is not dishonest on its own, but it is a different, lower-value thing than an original boxed set.
• Confirm the pattern name against a dated reference before paying a rare-pattern price. The Corning Museum of Glass maintains an online pattern library specifically so a pattern name and colorway can be checked against a documented production run rather than a seller's own label.
• Treat an unusually crisp, bright pattern on a famously rare design with suspicion. The rarest test patterns were rare because so few left the factory; a pristine example of one of the hardest patterns to find is worth more scrutiny, not less.

The look-alikes: reissues and other brands

Pyrex has officially reissued at least one of its own rarest vintage patterns, which creates a specific, well-documented confusion risk rather than a vague one. The 1959 test pattern known as Lucky in Love, or Lucky in Clover, was a limited factory run so small that no production count survives; Corning's own museum calls it one of the most elusive Pyrex designs ever made, and surviving examples show a known factory flaw, with the clover and heart motifs bleeding into each other, that marks them as a run that never passed the company's own quality standard for a full release. Pyrex brought the pattern back as an official, modern, limited reissue in 2018. The reissue is new glass, clearly differentiated in the secondary market, and sells today for a small fraction of what an original 1959 piece brings, which is exactly the gap a seller hoping to pass off the newer piece as the older one is counting on.

A second look-alike risk runs the other direction: country of origin rather than age. Pyrex made today in France under Arc International is genuine borosilicate glass, stamped in capital letters, the same combination collectors use as shorthand for "old and American." A current French import is neither old nor American, and nothing about its stamp alone will tell you that.

A third source of confusion is brand, not Pyrex at all. Fire-King, made by Anchor Hocking, produced its own line of opaque colored kitchen glass, most famously the pale green Jadeite line, in the same postwar decades and often turns up grouped with Pyrex in a single "vintage kitchen glass" listing or booth display. Fire-King never carried a Pyrex trademark and is a different glass family from either Pyrex formulation described above; Owl and Twig currently has a Fire-King Jadeite batter bowl in the shop as an example of the real thing, and our guide to vintage coffee mugs covers Fire-King identification in more detail.

Where the records disagree

Two separate gaps in the record are worth naming rather than smoothing over. First, the timing of the brand's own origin story: the celebrated 1914 cracked-casserole account is Corning's standard telling, repeated by the Smithsonian and by Corning's own museum, but at least one account of the company's internal history describes heat-resistant glass already being tested for cooking uses as early as 1913. Whether Bessie Littleton's cake was the spark that started that testing or a dramatic confirmation of work already underway is not something the available sources settle one way or the other.

Second, the name itself. Corning's own later trademark filings describe PYREX as a purely invented word, formed by inserting an "r" between "pie" and "ex" for the sound of it, with no Greek or Latin derivation intended. A persistent popular etymology nonetheless credits the name to the Greek "pyro," for fire, treating "ex" as an arbitrary commercial suffix layered on top. The company's own filed explanation and the popular derivation cannot both be the actual reason the name was chosen, and this guide has no way to adjudicate between a 1920s trademark lawyer's stated rationale and almost a century of popular assumption; it simply notes that the company's own account does not support the fire-root story usually attached to the name.

What a real one sells for now

Values across Pyrex patterns span an unusually wide range, and the rarest figures reported in the press deserve more scrutiny than they usually get. An original 1959 Lucky in Love bowl sold on eBay in 2015 for a confirmed $4,000. A different, far more widely reported Lucky in Love listing drew a winning bid of $22,100 in July 2022, and an earlier bid on the same listing had reached $45,000; both bids were retracted before the sale completed, the final bidder telling the seller he had bid by mistake. Despite wide press coverage treating the $22,100 figure as a sale, it was never a completed transaction and should not be read as a realized price. A Turquoise Golden Birds bowl sold on eBay in May 2025 for a confirmed $5,793.02 CAD.

Common patterns trade at far more modest, asking-price levels: Butterfly Gold five-bowl sets are commonly listed around $250; Flamingo Pink four-bowl sets around $375; Butterprint sets between roughly $350 and $900 depending on colorway; Gooseberry sets between roughly $350 and $750; and Starburst sets with their original metal stand between roughly $500 and $1,400. The 2018 Lucky in Love reissue, by contrast, trades for $20 to $60, a useful benchmark for how much of the original's value comes specifically from its scarcity rather than its pattern. None of these mid-range figures is a confirmed sold price; they are current asking-price ranges from retail and marketplace listings, offered here as a general guide rather than a record of an actual transaction.

For a specific, confirmed example at the modest end of the market: Owl and Twig currently has a 1.5-quart Pyrex #402 nesting bowl in the Foraging Fancies mushroom pattern listed at $35, an asking price on an active listing rather than a sold figure, but a real, current data point from outside the rare-pattern end of the market most price coverage focuses on.

For another guide where a company's trademark record does not match the popular origin date, see Bakelite: the trademark filed before the patent. For another vintage kitchen glass guide, see vintage coffee mugs: four things to check before you buy. See more Field Guides