If you hear someone claim that a food item was invented at the 1904 World's Fair in St. Louis, you can be almost certain that they're wrong. That particular event, officially named the Louisiana Purchase Exposition, has become a uniquely powerful magnet for bogus food origin stories—a phenomenon I covered in detail back in 2016. The hamburger, the hot dog, peanut butter, iced tea, the club sandwich, cotton candy: All have been alleged to have been invented on the fairgrounds in St. Louis in 1904, or at least popularized there. None of them actually were.
There is one classic American food item, however, whose origins truly are linked to the 1904 Fair: the ice cream cone. But it probably wasn't invented in a burst of inspiration right there on the fairgrounds, despite the many dramatic tales of quick-witted ice cream vendors running out of cups and rolling waffles into improvised cones. Instead, this is a story about good old-fashioned hustle, the ambition of newly arrived immigrants seeking their fortunes...and, of course, plenty of lawyers. How much more all-American could a food story be?
Before the Cone
Antecedents to the ice cream cone certainly existed long before 1904. People had been eating ice cream from various cone-like containers for decades; they just didn't eat the containers themselves.
Way back in 1807, Philibert-Louis Debucourt created an etching called The Interior of the Café Frascati, which shows, alongside elegant Parisians enjoying lemonade and punch, a woman lifting to her mouth an oddly spiral-shaped utensil, containing what might very well be ice cream. Culinary historian Robin Weir claims that this is the first pictorial evidence of an ice cream cone, and he might be right. It could also be a really ugly spoon.
As ice cream grew in popularity in the 19th century, roving vendors began selling it on city streets in a variety of cups and containers, including cone-shaped glass utensils and the notoriously unsanitary "penny licks"—tiny stemmed glasses in which ices were sold at the British seashore and on London streets.
This photograph, circa 1877, depicts a group of children swarming an ice cream vendor's cart in London; the boy in the center is eating a "halfpenny lick" of ice cream. At the time, vendors would serve portions of ice cream in glasses that would then be reused by the next customer without washing, potentially spreading disease.
But when did people start putting ice cream into edible cones? Many historians have pointed to the recipes of British author Agnes B. Marshall as precursors. Mrs. A.B. Marshall's Cookery Book (1887) includes instructions for making "Cornets with Cream," cone-shaped vessels made of a sweet paste of blanched almonds and flour, rolled around cornet molds, baked, and filled with sweetened vanilla-flavored whipped cream. "These cornets can also be filled with any [ice] cream or water ice," Marshall notes—the latter referring to frozen-water concoctions like granitas—"and served for a dinner, luncheon, or supper dish."
It's possible that future innovators were inspired by fancy confectionery like Mrs. Marshall's, but it seems more likely that the ice cream cone evolved out of efforts to help street vendors avoid the breakage and sanitation concerns that came with using dishes and spoons.
One of the earliest alternatives to the penny lick was "hokey-pokey," devised by immigrant street vendors in London in the 1870s. A British journal described hokey-pokey as "a coarse kind of Neapolitan ice" made with a blend of one part water to two parts milk, sweetened and thickened with sugar and cornstarch. The mixture was frozen, pressed into rectangular molds, then cut into slices and wrapped in white paper, to be sold from vendors' carts.
Within a few years, Italian immigrants had brought "hokey-pokey" to northeastern cities in the US, and it spread from there. In 1885, the Cincinnati Enquirer chronicled the appearance of this "novel luxury." "Just the thing for picnics," the reporter declared. "No spoons nor saucers needed, no washing of dishes." The emphasis here lends credence to the idea that the hokey-pokey's paper wrapper represented growing interest in single-use vessels and the resulting ease of cleanup.
In both England and the US, Italian immigrants dominated the urban ice cream trade, and innovations within it. By the 1890s, vendors had 86'ed the hokey-pokey's paper wrapper and created, as one observer quoted in a later newspaper article remembered it, "a half-inch slab of ice cream placed between two square pieces of sweetened wafer"—an early ice cream sandwich. Others experimented with various molds and devices to transform those wafers into edible cups.
In 1901, Antonio Valvona, an Italian citizen living in Manchester, England, filed a patent for an "Apparatus for Baking Biscuit Cups for Ice Cream." The device was designed for baking "dough or paste...composed of the same materials as are employed in the manufacture of biscuits [that is, cookies], and when baked the said cups or dishes may be filled with ice-cream, which can then be sold by the venders of ice-cream in public thoroughfares or other places."
The following year, Valvona teamed up with Frank Marchiony, an Italian immigrant in New York, to found the Valvona-Marchiony Company, which produced the patented cups and the ice cream sold in them. Valvona operated the firm's factory in the UK, while Marchiony ran the American operations, first on the Lower East Side of Manhattan and then in Brooklyn as trade grew.
By 1904, Marchiony had advanced from pushing a solo street cart to operating "two big wafer and ice cream factories" and a fleet of 200 carts, according to a newspaper profile at the time. But Marchiony's carts had plenty of competition on the streets of Manhattan, including from those of his cousin, Italo Marchiony, who had worked for the Valvona-Marchiony Company briefly, but by 1903 had established a rival ice cream firm. In September of that year, Italo Marchiony filed a patent for his own "molding apparatus used in the manufacture of ice cream cups and the like."
Italo's innovation, he claimed, was that his design made it possible to mold biscuit paste "in particular and unusual shapes" that were not possible before due to "the delicacy of the substance molded and the difficulty of forming and extracting the same from the molds." Unlike Valvona's two-part mold, which had a hinged top and bottom that closed like a clamshell, Italo Marchiony's device was composed of three pieces: a hinged top that could be raised vertically, plus a bottom consisting of two blocks that were hinged horizontally. "Upon the separation of the blocks," his patent application explained, "the withdrawal of the substance which has been formed in the mold is now readily permitted, notwithstanding the delicacy or crispness of the substances molded." The illustrations from Italo’s patent application show precisely the kind of unusual shapes he was trying to mold: tiny flat-bottomed teacups made of biscuit dough, complete with delicate handles.
And that's where things stood on the eve of the St. Louis World's Fair. People were eating ice cream from edible containers, but nothing that anyone today would consider an ice cream cone.
Meet Me in St. Louis
One thing we know for sure: Visitors at the World's Fair ate plenty of ice cream in cones—or cornucopias, as they were called at the time. Mrs. F. M. Hicklin wrote home to her family in South Carolina that the "waffle cornucopia filled with ice cream" was a treat widely enjoyed on "the Pike"—the Exposition’s version of a midway, a mile-long strip lined with cafes, amusements, and food concessions.
A photograph taken at the Fair shows Mrs. Montague Lyon and her three children standing outside the "New York to the North Pole" attraction, each holding an ice cream cone. (The eldest boy has managed to insert his cone almost completely inside his mouth.) The long cones have waffled sides and pointed bottoms, and they appear to resemble the molded type of waffle cone familiar to us today instead of just a rolled-up waffle. This suggests that the confection was neither an improvised creation from a waffle stand nor a product of Valvona's or Italo Marchiony's biscuit-cup mold.
What isn't clear is who the Lyons bought their cones from and how that vendor came up with the idea. I suspect that it wasn't the work of last-minute desperation, but rather a dodge to get around restrictive concessionaire licensing.
To sell any sort of item—from jewelry and souvenirs to food and beverages—vendors had to apply to and be accepted by the Exposition's Division of Concessions and Amusements. The competition was stiff, for there was a fortune to be made from selling snacks and drinks to daily crowds of 35,000 or more captive customers.
As Pamela J. Vaccaro captures in her definitive, and aptly named, history of the St. Louis World's Fair, Beyond the Ice Cream Cone, over 500 applicants sought the right to sell popcorn and peanuts at the event. Only one, C. A. Windmueller of St. Louis, won the contract. Another St. Louis firm, the Star Bottling Company, snagged the coveted "soft drink" concession, which gave it exclusive rights to sell flavored sodas, lemonade, root beer, ice creams, ices, and "all hot and cold drinks usually served at soda fountains."
A few weeks after the Fair closed on December 1, 1904, Star sued the organizers for $257,000 in damages for a range of alleged contract violations. Among the many items in dispute was which food and beverage items Star's exclusive franchise covered. The records of the court case singled out one in particular: "Whether ice cream cornucopias...pertained to an ice cream concession or were a food, because of the edible wafer wrapping the ice cream, and pertained to a restaurant or lunch stand concession."
Star Bottling was ultimately awarded $14,000, but court records leave this pertinent question unanswered: Which restaurant or lunch stand concessionaire came up with the idea of the ice cream cornucopia, thus horning in on Star's ice cream concession?
All we know for sure is that the ice cream cornucopia had been introduced to America, and it was about to become an even bigger hit than it had been at the Exposition. As soon as warm weather rolled back around in the spring, confectioners across the country had a brand-new treat to offer their customers.
By May 1904, newspapers from Florida to North Dakota were running ads and articles that referenced "Cornucopia Ice Cream Sandwiches." Many noted that the trendy new product had originated on the Pike in St. Louis, and dozens of state and county fairs granted "ice cream cone" concessions to vendors throughout the summer and fall of that year.
In August, the Macon Telegraph profiled the Cornucopia Waffle Oven Company of St. Louis, which had just been awarded the cone concession for the Tri-State Fair in Georgia. "This company is introducing a new novelty for serving ice cream," the paper reported. "They were first introduced at the great world’s fair at St. Louis and immediately, on account of its daintiness and neatness, became the most popular confection and have proven equally so in the East, and especially at the parks at Coney Island, Atlantic City, Chicago and various other famous resorts." The ice cream cone had arrived.
Marchiony Lawyers Up
Back in Brooklyn, Frank Marchiony's business at the Valvona-Marchiony Company was thriving. A profile that year in the Brooklyn Eagle reported that Marchiony was "constantly adding to the plant" and that Valvona-Marchiony had become "the largest manufacturer of the kind in the United States." Around the same time, Marchiony stopped using the name "biscuit cups" and started advertising "'Crispo' wafers and ice-cream cones," indicating that the company had fully jumped aboard the cone craze.
In the now-booming ice cream cone market, the Valvona-Marchiony Company had a valuable asset: Antonio Valvona’s patent for his "apparatus for baking biscuit cups." The firm landed a big victory in 1905 when a federal judge sustained the validity of Valvona's patent and ordered a competitor named D'Adamo to cease using his version of the device, which the judge declared to be "identical with the complainant’s, except in a trifling detail."
In 1910, the Valvona-Marchiony Company's lawyers started going after competitors selling cone-shaped containers, too, slapping patent infringement suits against companies from Missouri to Ohio to Indiana. Marchiony hired private detectives in Pittsburgh to infiltrate the Star Wafer Company's factory and identify what they claimed were copycat baking molds. On July 10, 1910, Frank Marchiony even filed suit against his own cousin, Italo, who had since moved out to Hoboken, New Jersey, and was operating a large factory with his own patented molds.
These suits met with varying degrees of success. In the Louisville Cone Company case, a Kentucky circuit court ruled that "we can see no difference in the applicable principles between a cup and a cone, inasmuch as the latter seems, in this connection, to be nothing more than the former in a different shape." But a judge in Boston examined two offending devices and denied a preliminary injunction, concluding that they "do not very closely resemble the device shown in the patent itself."
In a surprising twist, the federal court in New Jersey ruled that Italo Marchiony's patent was invalid and infringed upon his cousin's firm. "Structurally and functionally the devices are substantially alike," the judge determined, then added insult to injury by stating that "the defendant [Marchiony] never invented a single detail of the apparatus in question." Italo was ordered to discontinue production and pay damages, but his lawyers secured a stay as they appealed the decision.
The cases dragged on for years. In the process, the Trenton Evening Times observed in 1913, "the magnitude of ice cream cone manufacturing was brought out. The factories are running night and day turning out millions of cones each month."
The Federal Court of Appeals in Philadelphia settled the matter once and for all in 1914, in a case that subsumed the Italo Marchiony appeal and that of many other manufacturers. The judge took a narrow reading of Valvona's patent, concluding that his innovation was limited to details of the mold's design related to heat conduction and in no way prevented others from creating other types of baking molds.
"All Can Make Cones!" the International Confectioner trade journal jubilantly declared. The Valvona patent became essentially useless, and the American ice cream cone industry has never looked back.
Myth-Making
Curiously enough, none of the records from the many court cases make any reference to the St. Louis World's Fair, and none of the defendants purported to have invented the ice cream cone by rolling up a waffle. It wasn't until more than a decade after the Fair that people started making such claims. Almost all the tellings involve an element of drama—typically, an ice cream vendor runs out of cups, or guests start ruining their clothes as they eat melting ice cream with spoons, and some crafty person saves the day with a rolled-up waffle.
Many supposed inventors have been proposed over the years. One story attributes it to Ernest Hamwi, a Syrian-born concessionaire selling zalabia, a sort of thin Middle Eastern waffle. Others credit it to Abe Doumar, David Avayou, or two different sets of brothers, Nick and Albert Kabbaz and Charles and Frank Menches.
In the magazine Hoboken History, Italo Marchiony's own daughter claimed that he, too, had invented the ice cream cone at the World's Fair, though that account was written in 1992, almost nine decades after the events in question. The invention story recorded closest to the actual event was published by C. M. Egbert in the New York Produce Review and American Creamery in 1916, and it omits one important detail: the inventor's name.
"A certain young lady," Egbert wrote, had a fairground concession selling "a sweet cake which she baked flat" on a waffle iron–like device. Her brother sold ice cream at a nearby stand. One customer who had bought some cakes from the sister asked the brother to put a scoop of ice cream inside of them. The brother made it work by "rolling up the cakes while they were still hot in the shape of a cornucopia and pinching over the end.... He then joined forces with his sister, putting the two concessions together, and they soon did a rushing business in ice cream cones, as they were very promptly dubbed."
Egbert worked for an ice cream company in Spokane, Washington, and he had heard the story secondhand from his manager, who had attended the Fair. "This I believe is the true and correct story of the beginning of ice cream cones," he wrote, "although I have personally met at least fifty people, each of whom was the 'originator.'"
None of the many people later credited with inventing the cone appear in the Fair's records of official concessionaires, nor are any listed as residents in the 1904 St. Louis city directory. But that doesn't mean they weren't selling food at the Fair, since many concessionaires hired out-of-town workers to staff their booths or subcontracted to vendors from other cities. And, because most of these alleged inventors went into the ice cream cone trade in the heady years right after the Fair, each of them would have had a vested interest in claiming to have been the inventor.
Abe Doumar, for instance, set up a stand on Coney Island in 1905, where he sold ice cream cones made on custom-made waffle irons; he soon set up his brothers in Doumar's stands, including one in Norfolk, Virginia, where one of the original 1905-vintage machines is still used to make cones today. By 1913, Ernest Hamwi had established the American Cone Company, later the Missouri Cone Company, in St. Louis, but it was only much later—in his obituary in 1943—that he was credited with improvising the product on the Pike.
With so many competing after-the-fact stories and so little tangible evidence, it seems unlikely that we'll ever know who came up with the ice cream cone. Maybe the real lesson from this story is that the certainty and drama we crave in food histories—the idea that a single person had to have been the first, and that their invention was made in a flash of brilliance to solve an immediate emergency—are rarely confirmed in reality. But who wants to believe that something as delicious as the ice cream cone resulted from the dull, drawn-out process of trial and error, with multiple individuals working in parallel on mundane problems like the inconvenience and expense of washing the dishes?