The Gilbert Atomic Energy Lab: the toy the government never banned, and what a set brings now

By Matthew Partsch · Owl and Twig Vintage · September 19, 2026

The Hazard Files, Case No. 002. In 1950 the company that made the Erector Set put four jars of uranium ore, three radioactive source disks and a Geiger counter into a red suitcase and sold it to children for $49.50. No regulator ever banned it. No injury from it has ever been recorded. It sold fewer than 5,000 copies, disappeared inside two years, and in December 2024 a complete one brought $16,500 at auction. Case No. 001 was lawn darts. This is its mirror image.

What the Atomic Energy Lab was

The Gilbert No. U-238 Atomic Energy Laboratory is a suitcase about 25 inches wide, 16 or 17 inches deep and 5 inches high, and the lid opens to a printed picture of a boy at a table running the experiments inside. The Museum of Radiation and Radioactivity at Oak Ridge Associated Universities, which is the reference collection for this object, lists the contents as four types of uranium ore, a beta-alpha source of lead-210, a pure beta source it records as ruthenium-106 with a question mark, a gamma source of zinc-65, a spinthariscope, a Wilson cloud chamber with its own short-lived polonium-210 alpha source, an electroscope, a battery-powered Geiger-Mueller counter, an instruction manual, the comic book Learn How Dagwood Splits the Atom, and a government booklet, Prospecting for Uranium. The four ores, named on every auction catalog and on the Smithsonian's photograph, are carnotite, autunite, torbernite and uraninite, dug from the Colorado Plateau. Add a set of colored nuclear spheres for building atom models, a Dri-Electric power pack, a deionizer, a compression bulb and the glass viewing chamber, and that is the set.

The price was $49.50. The Department of Energy, IEEE Spectrum and the standard reference works all give that figure; the Oak Ridge museum rounds it to about $50 and the A.C. Gilbert Heritage Society writes $50.00. The Department of Energy puts the 2021 equivalent at roughly $530, and IEEE Spectrum at about $500. Porter Chemical Company, Gilbert's rival in chemistry sets, was selling uranium-ore kits at $10 and $25 in the same years. Gilbert's was five times the price of the next most expensive atomic toy in America, and the box promised over 150 experiments, the sight of alpha particles moving at 12,500 miles per second, and a shot at a government reward for finding uranium.

The set that is most often photographed is the 1951 version, with a red textured case. The Oak Ridge museum records an earlier and rarer 1950 version in a tan case, described in the 1950 Gilbert catalog, with the contents arranged differently. The earliest dated image of the red case the museum could find is in the Rochester Democrat and Chronicle of November 23, 1950, so the second version was already in stores for that first Christmas.

Why Gilbert built it, and who helped

Alfred Carlton Gilbert was a Yale-trained physician, an Olympic pole vault champion, and from 1913 the manufacturer of the Erector Set. His company had bought American Flyer trains in 1938 and rebuilt them on two-rail S gauge after the war, and by 1950 it was the largest toy company in Connecticut with a settled line of chemistry sets, microscopes and telescopes sold as education rather than play. The Atomic Energy Lab was the top of that line. In his 1954 autobiography, The Man Who Lives in Paradise, Gilbert called it the most spectacular of the company's educational toys, wrote that some of the country's best nuclear physicists had worked on it, and that the company had worked closely with M.I.T. in its development.

The 60-page manual was written by Ralph E. Lapp, a physicist who had worked on the Manhattan Project. The comic in the box, Learn How Dagwood Splits the Atom, was drawn by Joe Musial for King Features in 1949 and, according to IEEE Spectrum, prepared with the consultation of Leslie R. Groves, the general who had run the Manhattan Project, and John R. Dunning, the Columbia physicist. The Smithsonian's National Museum of American History catalogs its own set with the note that the booklet was produced by the noted cartoonist Joe Musial. A newspaper column by Dr. Grace Langdon in the Bend Bulletin of October 21, 1950 praised the set in terms that explain the entire marketing position: it stresses the peace-time role of atomic energy, and there is not a hint of the A-bomb in the whole kit.

That was the point. Five years after Hiroshima, and one year after the Soviet Union tested its first bomb, the federal government wanted Americans to think of the atom as a power source and a prospecting opportunity. Gilbert wrote in the same autobiography that the government encouraged the set's development because it believed the lab would aid public understanding of the constructive side of atomic energy. That sentence is the whole inversion of this series. The lawn dart was a toy the government spent eighteen years trying to keep out of toy stores and then banned. The Atomic Energy Lab was a toy the government helped put in them.

What the government actually did

In 1948 the Atomic Energy Commission published a guaranteed ten-year minimum price for uranium ore delivered to it, and in a second circular a $10,000 bonus for producers who found and delivered ore from a new domestic deposit. The Court of Claims later summarized the program in Gay v. United States: Circular No. 1 for the guaranteed price, Circular No. 2 for the $10,000 bonus, Circular No. 5 for the low-grade carnotite ores of the Colorado Plateau delivered to the AEC depot at Monticello, Utah, and Circular No. 6 for a sliding-scale bonus by grade. In 1949 the AEC and the U.S. Geological Survey published a pocket guidebook, Prospecting for Uranium, to send citizens out to find it. The Library of Congress, which holds the booklet, notes that one Geiger counter maker reported a 500 percent increase in sales in a single year of the rush that followed.

That booklet is in the Gilbert box. So is a coupon for additional radioactive samples, recorded on the Griffin Museum of Science and Industry's set in Chicago. So is the Geiger counter. The lid art and the catalog copy told a boy he could prospect for uranium and collect the government's reward. The Atomic Energy Lab was, among other things, an official prospecting kit for the AEC's own program, packaged by a toy company and priced like a bicycle.

How radioactive it was, then and now

The manual contained one safety instruction that every modern account quotes, and it is about mess rather than dose. The Department of Energy's Office of Legacy Management reproduces it: users were advised not to take the ore samples out of their jars, because they tend to flake and crumble, and you would run the risk of having radioactive ore spread out in your laboratory. IEEE Spectrum, in Allison Marsh's 2020 history of the set, assessed the likely exposure from a set used as directed as minimal, about the equivalent of a day's UV exposure from the sun. The Bulletin of the Atomic Scientists, presenting the Griffin museum's set in its Turn Back the Clock exhibition, concluded the kit failed to sell because of cost and not because of any safety concern of the era.

The sealed sources are where the real physics sits, and they answer an authentication question nobody in the auction trade seems to ask. Polonium-210 has a half-life of 138 days. The polonium wire that made the cloud chamber work was effectively inert within a few years of purchase; after 75 years it has passed through roughly 200 half-lives and contains no measurable polonium at all. Zinc-65, the gamma source, has a half-life of 244 days and was gone by the early 1960s. Ruthenium-106, the beta source, has a half-life of 374 days and followed it. Lead-210, the beta-alpha source, has a half-life of 22 years and retains about a tenth of its 1950 activity. The uranium ore, with a half-life of four and a half billion years, is exactly as radioactive as the day it was dug, and it will still make a working Geiger counter click.

This matters at the table. A seller who demonstrates a cloud chamber showing alpha tracks is demonstrating a replacement source. A set whose zinc or ruthenium disk reads strongly on a meter has been re-sourced. Neither is a crime, but neither is original, and the price difference between an original set and a re-sourced one is the entire subject of the values section below. The one test that requires no paperwork is to hold a Geiger counter to the sealed ore jars. Genuine Colorado Plateau carnotite and uraninite read well above background. Painted gravel does not.

Why it disappeared after 1951

Three explanations are on the record, and they come from three different kinds of source. The first is the maker's own. Gilbert wrote in 1954 that the set had to be priced very high, that the company managed to lose a little money on every one sold, and that it was more appropriate for boys with some educational background than for the company's usual customer; you could not, he wrote, manufacture such a thing as a beginner's atomic energy lab. The Bulletin of the Atomic Scientists and the Oak Ridge museum both accept this. The museum records that by 1952 the U-238 lab had been replaced by the No. 11 Gilbert Chemistry Atomic Energy Set, a cheaper kit with uranium ore, a spinthariscope and an Atomic Energy manual, and that the advertisements which linger into 1952 and 1953 are discounted old stock.

The second explanation comes from a document inside the $16,500 set. RR Auction's catalog for its December 2024 sale describes a letter from the A.C. Gilbert Company dated September 25, 1953, telling a customer the set had been discontinued due to Government Restrictions. The letter has not been read for this guide, so that phrase is the auction house's summary of it, and it contradicts the man who signed the company's checks. Either Gilbert's sales department gave customers a tidier reason than the accountants had, or the AEC's rules on the distribution of reactor-made isotopes tightened around a product that used three of them, or both. No regulatory record naming the set has surfaced, and the standing account, from the maker, from the museums and from the trade press, is price.

The third is the popular one, and it is wrong in a specific way. Radar Magazine's December 2006 list of the ten most dangerous toys of all time placed the Atomic Energy Lab at number two, directly below lawn darts at number one, and every headline since has called it the most dangerous toy in history. The ranking is where the reputation comes from. Nothing in the 1950s record supports it: no injury, no complaint, no agency action, and a manual whose only warning was about crumbs.

How to tell a real one

Fewer than 5,000 were sold, so the base rate matters: a complete set appearing at an estate sale is a once-a-decade event, and the parts trade is where most of the risk is. Nine checks, in the order they cost you nothing.

• Read the case. The common 1951 set is a red, textured, paper-covered suitcase with brass hardware, which Hake's describes as textured paper and the Griffin museum as faux snakeskin; RR Auction calls it a reed suitcase. The underside of the lid carries a printed illustration of a boy using the lab against atomic imagery. A tan case is the rarer 1950 first version, described in the 1950 catalog, with a different internal layout.
• Measure it. The records span roughly 25 to 25.5 inches wide, 15.5 to 17.25 inches deep and 4.7 to 5 inches high. A Sotheby's listing gave 22.5 inches wide, which no museum record matches; treat it as a measuring error or a different object until proven otherwise.
• Count the ore jars and read the labels. Four: carnotite, autunite, torbernite, uraninite. The National Museum of Nuclear Science and History records three in its own set. Three jars is a set with one missing, and the missing one is the first thing to ask about.
• Find the source disks and their holders. The lead-210, zinc-65 and ruthenium-106 disks sit in holders; the polonium wire belongs to the cloud chamber. RR Auction's December 2025 catalog states that two disks in that set were replicas. If an auction house says it, assume a private seller will not.
• Look for the Dri-Electric power pack. The A.C. Gilbert Heritage Society singles it out as the part most often missing. The deionizer, compression bulb, glass viewing chamber, tubing, power leads, stand and legs go with the cloud chamber and go missing together.
• Count the nuclear spheres. The National Museum of Nuclear Science and History records eight of each color for the atom models. Loose colored balls are the easiest part of the set to fake and the hardest to prove.
• Check the paper. Gilbert Atomic Energy Manual, 60 pages, by Ralph Lapp. Learn How Dagwood Splits the Atom, King Features, 1949. Prospecting for Uranium, AEC and USGS, 1949. An inspection slip, a coupon for further samples and a Gilbert catalog turn up in the best sets. Each of these also circulates loose, so their presence proves the paper and not the case.
• Put a Geiger counter on the sealed jars. Real ore reads. This is the only test that needs no reference photo, and it is the reason a working counter belongs in the kit of anyone who buys atomic-age toys.
• Distrust a working cloud chamber. Polonium-210 does not survive two years, let alone seventy-five. Tracks mean a replacement source, and a replacement source means the set has been through a restorer's hands.

The look-alikes and the parts market

The Gilbert lab has few direct copies because nobody else was foolish enough to build a $49.50 toy. What it has instead is a crowd of cheaper cousins from the same three years, all of which contain real uranium ore, all of which get listed with the words atomic energy lab, and none of which are the U-238.

What it is	How it fools you	How to place it
Porter Chemcraft Atomic Energy kits, No. 10 and No. 25, from 1947	Earlier than Gilbert, contain real uranium ore and a spinthariscope, and sellers call them atomic energy labs	Chemistry-set format in a Chemcraft box at $10 and $25 when new; the spinthariscope uses a radium screen, and the booklet is The Story of Atomic Energy. The Oak Ridge museum has both
Porter Chemcraft Model No. 100, early 1950s	Two glass vials, one of ore and one of a uranium chemical, plus a radioactive screen	No cloud chamber, no Geiger counter, no electroscope. A vial of uranium chemical is a Porter feature; Gilbert shipped only ore
Gilbert No. 11 Chemistry Atomic Energy Set, 1952 onward	Same company, same words on the box, sold as the replacement	Uranium ore, a spinthariscope and an Atomic Energy manual inside a chemistry set. No U-238 number, no suitcase, no Geiger counter
The Kix Atomic Bomb Ring, 1947	A polonium spinthariscope in a cereal premium, often described as the same technology	It is the same technology, in a ring, for fifteen cents and a box top. Covered in the cereal premiums guide
Replica source disks and holders	Made to fill the two holders most often empty, and now appearing in auction catalogs as acknowledged replicas	A disk that reads strongly on a meter is either a replica with a modern source or an original that has been re-loaded. Either way it is not 1950
Loose manuals, Dagwood comics and Prospecting for Uranium booklets	Sold as from an Atomic Energy Lab, which is often true and never proves a case exists	All three were printed separately and circulated separately. Price them as paper
Advertisement reprints and catalog page reproductions	Listed under the set's name, sometimes at prices that suggest the object	Modern reprints on modern stock; the 1950 and 1951 Gilbert catalogs are archived by the Eli Whitney Museum and the Heritage Society for comparison

What it sells for

Five realized prices exist, which makes this the best-documented market in the Hazard Files and one of the best on this blog. Four are RR Auction results with the buyer's premium included, and they all landed within twenty months of each other.

Date and house	What was in the case	Estimate	Result
December 11, 2024, RR Auction sale 705	Complete, with manual, Prospecting for Uranium, Dagwood comic and a September 25, 1953 Gilbert Company letter; minor case scuffing	$4,000 and up	$16,500
June 11, 2025, RR Auction sale 719	Near complete; missing only the flashlight bulbs and two batteries	$4,000 and up	$14,196
August 12, 2026, RR Auction sale 747	Geiger counter, electroscope, spinthariscope, cloud chamber, deionizer, power pack, spheres, three batteries, four ores and all paper; several radioactive sources and holders missing; partial separation at the case base	$10,000 and up	$12,499
December 10, 2025, RR Auction sale 731	Near complete; two source disks replaced with replicas; adds an Adventures Inside the Atom comic and toy catalogs	$4,000 and up	$8,860
July 11, 2019, Hake's	Not complete; cloud chamber, spinthariscope, electroscope, power pack, spheres, ores, manual, deionizer and inspection slip; contents largely unused	none published	$2,360
August 20, 2022, George Cole Auctions	Described only as a vintage set, four photographs	$1,200 to $1,500	result not published
Sotheby's Buy Now, 2025, reported	Described as complete and original, very good, minimal case wear	asking, per press reports	$4,800 asking
1950 retail	New, in a Gilbert catalog		$49.50

Two things in that table are unusual. The first is the direction of the estimates. RR Auction opened three consecutive sales at $4,000 and up and watched each one close at two to four times that, and only on the fourth, in August 2026, did it raise the estimate to $10,000. The market moved faster than the house that was making it. The second is what the $8,860 result says. That set was near complete and came with extra comics and catalogs, and it was the cheapest of the four modern results by a wide margin, because the catalog disclosed that two of its source disks were replicas. The difference between the December 2024 set and the December 2025 set, roughly $7,600, is mostly the price of two sealed disks of decayed isotope that do nothing and cannot be seen to do nothing, plus one typed letter. That is what a collector is paying for: the original object, all of it, including the parts that stopped working during the Eisenhower administration.

The wider retail market runs on asking prices and rumor. The $4,800 Sotheby's figure was reported by Snopes and by hobbyDB in 2025; the listing page itself no longer shows a price. Loose Dagwood comics, manuals and spinthariscopes trade on the general marketplaces and no verified sold figures for them were found for this guide, so any number a seller quotes for a part should be treated as that seller's opinion.

Where the records disagree

Five museum-grade records were consulted, and they do not agree, which is usual for a mass-produced object and unusual to see written down. The Smithsonian's National Museum of American History, object 1990.0534.01, dates its set 1950 to 1951 and measures the case at 25.5 by 17 by 4.69 inches. The Griffin Museum of Science and Industry in Chicago, object 2017.4.1, a 2017 museum purchase, dates its set to 1950 and measures 25.5 by 17.25 by 4.875 inches. The National Museum of Nuclear Science and History, object 13.42.2 from the Jay Shelton collection, dates its set precisely to October 1950, measures it 25 by 17 by 5, and records three jars of uranium ore where every other record has four. RR Auction gives 25 by 16.5 by 5 for four different sets. The Sotheby's listing gives 22.5 by 16.5 by 4.75. Nobody has measured the same suitcase.

The dates disagree at both ends. The Oak Ridge museum, the Smithsonian, the Department of Energy and IEEE Spectrum say 1950 and 1951, and the museum has a newspaper image from November 1950 to prove the first year. The A.C. Gilbert Heritage Society, the collector organization for the company, writes that the set was introduced in 1951 and discontinued in 1952, one year only. Hake's cataloged its 2019 set as a 1952 set and wrote that the lab sold from 1950 through 1952. All three positions can be partly true if 1950 was a soft launch of the tan version, 1951 the red version's year, and 1952 the sell-off of stock, which is what the Oak Ridge account describes. The Heritage Society's date is the one to be careful with, because it is the one collectors repeat.

The isotope list disagrees in one place. The Oak Ridge museum, whose curators are health physicists, records the pure beta source as ruthenium-106 with a question mark. The Department of Energy and the nuclear museum drop the question mark. The Heritage Society drops it too. The question mark is the honest version, and it is repeated here.

And the reason for discontinuation disagrees between the company and the company, as described above: price in the founder's 1954 book, Government Restrictions in a 1953 customer letter as summarized by RR Auction. A guide that picked one would be guessing. The $16,500 set includes that letter, and whoever owns it now could settle the question by publishing the page.

Is it legal to own and sell one

Yes, under federal rules as they stand, and the contrast with lawn darts is the point of pairing the two. Lawn darts remain illegal to sell in the United States under 16 CFR Part 1306, and eBay prohibits the listings outright. The Atomic Energy Lab has never been the subject of a federal rule. Its uranium ore falls under 10 CFR 40.13(b), which exempts any person who receives, possesses, uses or transfers unrefined and unprocessed ore containing source material from licensing, provided the ore is not refined or processed. Four sealed jars of Colorado Plateau ore in a toy are the textbook case. The three sealed sources were reactor products distributed under the AEC's isotope program and are now decayed to trace quantities or to nothing. RR Auction has sold four sets openly in twenty months and Sotheby's has listed one, which is as much practical confirmation as a collector needs.

Two cautions, and neither is legal advice. Shipping is governed by carrier rules on radioactive material rather than by the NRC, and a seller should read those before boxing one. And state or local rules can differ, so a dealer who trades in these regularly should ask a lawyer once rather than a forum repeatedly. The Consumer Product Safety Commission, for the record, was created in 1972, twenty years after the last set was sold, and has never had occasion to look at it.

The toy that was never banned

The Hazard Files exist to look at the objects history swears were fine, and the Atomic Energy Lab breaks the pattern by being an object history swears was lethal. It was a $49.50 physics course in a suitcase, built with a Manhattan Project physicist's manual and a Manhattan Project general's comic book, sold at a loss to a few thousand families who could afford it, and abandoned because it was too expensive and too hard. The ore in it is exempt from licensing today, and no account of the set records a buyer needing a license then. Nobody was hurt by one. Its reputation was assigned to it in 2006 by a magazine list, and its price was assigned to it after that by a market that wants the reputation and the object in the same case.

Lawn darts were the opposite in every column. Cheap, made by the million, sold in the sporting goods aisle next to the toys, classified a hazard eighteen years before they were banned, and banned only after a seven-year-old died. They are worth a couple of hundred dollars because there are millions of them and it is illegal to sell one. The Atomic Energy Lab is worth five figures because there were never many and it has always been legal. The government's judgment and the market's judgment ran in opposite directions on both, and both times the market was following the law's shadow rather than the danger.

Case No. 003 will invert this one again. The Field Guides index has every guide on the site, the 1950s entry in the decade series covers the same Christmas mornings from the other side of the tree, and the Mechanical Martin is what the same boy was more likely to actually get.