By Jordan Gold and Valery Gold
Research cutoff: November 7, 2026, CST. Historical documents and studies have different observation periods; the latest directly checked global production table reports 2024, not 2026.
A yellow-green glass cup can carry two histories at once. One is the history of an attractive household object: a glassmaker’s formula, a family table, an antique-shop shelf. The other is the history of uranium—a material that moved from decorative coloring into strategic stockpiles, nuclear fuel and a waste-management burden measured on a very different clock.
A black light makes the cup’s color spectacular. It does not reveal what its maker inhaled, whether a former factory was surveyed before redevelopment, or how much newly extracted uranium eventually becomes fuel rather than depleted material. Those questions require records, not a Geiger-counter video.
The records tell a more complicated story than either radioactive nostalgia or radioactive panic. Uranium compounds gave glass and pottery distinctive colors. Uranium can also damage kidneys through chemical toxicity, apart from its radioactivity. Mining studies establish serious hazards from radon exposure; they do not establish the risk from displaying an intact bowl. And the afterlives of consumer factories range from continued ceramics production to demolition and a new school; those histories cannot be inferred from uranium’s later strategic uses. 1 2 8 14 22 25
The public-interest question is not whether to fear everything that glows. It is whether we can follow the material, the people who worked with it and the obligations left behind without substituting either reassurance or suspicion for evidence.
A colorant before it was a strategic material
Commercial uranium-colored glassmaking emerged in the 1830s. Uranium compounds supplied yellow and yellow-green colors; additional ingredients and different recipes produced other appearances, including opaque custard and Burmese glass. “Vaseline glass” describes a narrower transparent yellow-to-yellow-green category, not every object containing uranium. The ingredients were added for their optical properties. Their use long preceded the early twentieth-century enthusiasm for radioactive remedies. 1 4
The green response under ultraviolet illumination is fluorescence: the material responds to incoming light. It is not a visible stream of radioactive emissions, and it should not be confused with the self-luminous radium paints used on historical dials. Radioactivity and fluorescence are different properties; a dramatic glow is not a calibrated measure of dose. Some uranium-containing glass does not fluoresce as expected. 1 17
Color also helped sell mass-market ceramics. Homer Laughlin introduced Fiesta in 1936, giving buyers a bright, coordinated alternative to conventional tableware. NIST’s historical account credits ceramic scientist Albert Bleininger with developing the glazes and helping modernize manufacturing, while designer Frederick Hurten Rhead developed the recognizable forms. NIST records the company describing its uranium glazes as “harmlessly radioactive.” That is evidence of historical sales language, not an assurance this report adopts. 3
Fiesta’s orange-red glaze contained uranium; ivory also used it. This was uranium in a ceramic surface glaze, not the same construction as uranium incorporated into a glass object. The distinction matters because composition and surface chemistry affect what can escape into food or liquid. The popularity of a bright plate does not establish that its buyer sought radiation therapy. 2 13
There really was a market for radioactive health claims. Radium-containing preparations and devices intended to put radon into drinking water promised vitality and relief. ORAU’s museum documents products such as radioactive water crocks and Radithor, as well as the eventual recognition of severe harm. Radium is a different element from uranium; radon is a radioactive gas in decay chains. Their histories intersect with uranium, but the remedies must not be folded indiscriminately into the history of decorative glass. 4
The interruption was real. A universal timetable is not.
Wartime demand changed access to uranium. The most firmly specified consumer chronology here is Fiesta red: ORAU documents natural-uranium production in 1936–1943, a return using depleted uranium in 1959–1969, and depleted-uranium red Fiesta Ironstone in 1969–1973. Its account attributes the wartime interruption to government acquisition of the company’s uranium stocks. 2
That chronology does not prove every glassworks in every country stopped and restarted on those dates. ORAU describes a roughly fifteen-year interruption in uranium-glass production beginning during World War II, but its own discussion includes production estimates spanning 1958 onward. A plant-specific history is firmer than a universal claim about an entire international industry. 1
Nor did the end of a particular product necessarily mean a radiation ban. Homer Laughlin’s company timeline describes a 1950s surge of low-cost imported dinnerware and shifting color trends; it records Fiesta’s 1973 retirement and its 1986 reintroduction in a lead-safe vitrified body, but it does not state a reason for the 1973 retirement. For Fenton, later reporting quotes company historian James Measell citing a changing giftware market and the costs of natural gas, labor and running a company, and says the decision to bring the company to a close was made in 2007. These are attributed explanations; they are not proof that every closure had the same cause. 5 23
Uranium glass has not become exclusively an antique. On pages checked for this report, Czech maker Zbynek Votocek explicitly describes hand-blown, hand-cut uranium-oxide glass reproductions “made today.” Preciosa Ornela currently presents a uranium-glass collection including rods, casting material and decorative products. These are manufacturers’ representations, not independent observations of a furnace operating on the research date. They do establish more than a vintage resale listing. Conversely, a museum’s list of American producers dated 2004 is not a reliable directory of firms operating in 2026. 6 7 1
What reaches the body matters more than the spectacle
Uranium presents two hazards that should be separated rather than allowed to blur together. It is a radioactive element. It is also a chemically toxic heavy metal, with the kidney—particularly the proximal tubules that help regulate what is retained or excreted—a principal target of chemical injury. 8 9
An intact finished object, soluble uranium in drinking water and fine workplace dust are not equivalent exposures. Solubility affects movement into blood and organs. Less-soluble inhaled particles can remain in the lungs longer; more-soluble compounds can move more readily into circulation. Particle size, route, duration, amount and isotopic composition all matter. Enrichment changes the isotopic mixture and radioactivity per unit mass; depleted uranium remains radioactive. 8 9
This is why a drinking-water concentration cannot simply be converted into a collector’s risk, and why a miner’s cancer history cannot be assigned to a glass decorator without exposure evidence. It is also why compliance limits are not themselves measurements of disease. A workplace air limit, a radiological release criterion and a study’s observed health outcome answer different questions.
Kidney studies: detectable changes are not the same as kidney failure
In a 2002 Finnish study, Päivi Kurttio and colleagues examined 325 people who drank from drilled wells. Median uranium in their drinking water was 28 micrograms per liter (interquartile range 6–135; maximum 1,920). Uranium in urine was associated with increased fractional excretion of calcium and phosphate; uranium in water and daily intake were statistically associated with calcium fractional excretion, but not with phosphate or glucose excretion. Uranium exposure was not associated with creatinine clearance or urinary albumin, indicators of glomerular function. The authors described a weak association with altered proximal-tubule function and wrote that its clinical and public-health relevance was “not easily established.” These are biomarker findings, not 325 diagnoses of kidney disease. 10
There is an important correction attached to that paper: a 2003 erratum changes the urine-uranium unit in Table 3 from micrograms per liter to nanograms per liter. It does not change the drinking-water concentration quoted above. A thousandfold unit error is precisely the kind of detail an exposure story cannot afford to overlook. 11
A 2006 study by Kurttio and colleagues examined 193 adults (95 men and 98 women) who had used drilled-well water for an average of 16 years. Median water uranium was 25 micrograms per liter (maximum 1,500). Ten kidney-toxicity indicators showed no statistically significant associations with uranium in urine, water, hair or toenails, and the authors concluded that uranium intake “was not found to have cytotoxic effects on kidneys in humans.” The same abstract reports that uranium exposure was associated with higher diastolic and systolic blood pressure and that cumulative intake was associated with increased urinary glucose. The two Finnish papers therefore point in somewhat different directions on kidney markers; they measured different indicators in different samples, and neither should be forced into the other’s conclusion. This report read the 2006 abstract through EPA’s HERO database; it did not obtain the full paper. 12
Both studies are observational. They cannot randomly assign lifelong exposure, exclude every correlated constituent of well water or prove that a biomarker change will progress to clinical illness. Conversely, failure to detect an association for selected endpoints is not a certificate of safety. ATSDR’s broader synthesis identifies kidney toxicity as a central concern while noting substantial limitations in human evidence for other outcomes. 9
For reproductive and developmental effects, the human evidence is less settled than the basic kidney toxicology. ATSDR describes limited human data and effects in some animal experiments. Those experiments identify hazards and research questions; they do not establish that living with uranium glass causes infertility or birth defects. 9
Miners and processors: strong warnings, different exposures
The 2023 Pooled Uranium Miners Analysis, led by Kaitlin Kelly-Reif and colleagues, combined seven cohorts comprising 119,709 male underground miners, with 7,754 lung-cancer deaths. The authors report that the analysis supports a linear exposure-response relationship between cumulative radon-progeny exposure and lung-cancer mortality, modified by age, time since exposure and exposure rate. 14
This is substantial occupational evidence, but the measured exposure is not “being near uranium.” Historical monitoring was less complete in early periods, and the paper states that neither occupational co-exposures (such as silica, arsenic, gamma radiation and diesel exhaust) nor smoking were assessed in this analysis. The authors discuss those limitations and differences among cohorts. None makes it appropriate to dismiss the radon findings; none makes the findings a dose estimate for an antique shelf. 14
Nuclear-material processors experienced another mixture. Sharon Silver and colleagues’ 2013 Fernald study followed 6,409 workers through 2004. Hourly male workers had elevated lung-cancer mortality relative to the U.S. population: a standardized mortality ratio of 1.25, with a 95% confidence interval of 1.09–1.42. Most evaluated cancers, however, did not show a convincing radiation dose-response. Intestinal cancer did show an association with estimated internal organ dose. 15
The investigators reconstructed exposure from urine measurements, job histories and other records, and considered several radiation sources and nonradiological hazards. Smoking data were incomplete. Their discussion warns that uranium measurements can also mark exposure to accompanying agents: an association cannot automatically be attributed solely to radiation. A comparison with national mortality rates and an internal dose-response analysis are different tests, not interchangeable answers. 15
Cato Milder and colleagues’ 2024 follow-up of 2,514 White male Mallinckrodt uranium-processing workers, observed through 2019, found a kidney-cancer association with reconstructed organ radiation dose. In the continuous Cox model, the hazard ratio at 100 milligray was 2.07, with a 95% confidence interval of 1.12–3.79. There were 16 kidney-cancer deaths in that analysis. The authors write that a causal radiation–kidney cancer association is “plausible,” but also that the dose-response “is driven by three cases with doses over 100 mGy” and carries considerable uncertainty. Categorical and other model results were less conclusive. 16
The exposure reconstruction incorporated external radiation, medical X-rays and internal uranium, radium and radon-related exposure. Silica dust was also assessed—high cumulative dust exposure was separately associated with non-malignant kidney disease—and smoking was unmeasured. The finding warrants attention, but calling it proof that uranium alone doubled workers’ kidney-cancer risk would discard both the exposure mixture and the statistical uncertainty. The same study did not find a lung-cancer radiation dose-response. 16
These are records of actual workforces, not interchangeable verdicts about every uranium product. They make the lack of comparable follow-up for some consumer-product workers more consequential—not a license to invent their outcomes.
What can leave a plate or glass?
Edward Landa and Terry Councell’s 1992 laboratory study tested 33 glass objects and two ceramic objects, using sequential exposures to deionized water, dilute acetic acid and nitric acid. The abstract reports a maximum of about 30 micrograms per liter leached from the glasses and about 300,000 micrograms per liter from the ceramic-glazed items—showing that uranium could leach and that the tested glazes could release far more under the extraction program than the tested glasses. Two ceramics are not a representative census of historical dinnerware. 13
ORAU’s account of their 24-hour glass tests reports average concentrations of 0.052 micrograms per liter in water and 5.9 micrograms per liter in acetic acid. The experiment is evidence about migration under specified conditions, not an ordinary meal’s dose. The very large ceramic maximum summarized by USGS came from the study’s extraction program, which included strong nitric acid; it should not be presented as what somebody routinely drank from a cup. 1 2 13
External radiation, dissolved uranium and swallowed fragments are separate exposure pathways. Historical NRC modeling summarized by ORAU also distinguishes whole-body gamma exposure from localized beta exposure at the skin; an on-contact reading should not be mislabeled a whole-body annual dose. This report does not claim to have measured any reader’s collection. 1 2
EPA’s practical advice is more useful than either bravado or alarm: radioactive antiques are usually not a health risk when intact and in good condition, but uranium glass and antique radioactive ceramics should not be used for food or drink. Chipping can introduce particles into food. EPA directs owners of broken radioactive antiques to state or local radiation-control programs for disposal advice. Display is different from eating, grinding or polishing. 17
Following newly extracted uranium without losing the mass
The latest official production compilation directly retrieved for this report is Uranium 2026, the joint NEA/IAEA “Red Book.” Its Table 1.18 reports 61,924 tonnes of uranium in 2024, compared with a revised 54,563 tonnes in 2023. These are uranium quantities, not the mass of rock moved, yellowcake shipped or uranium hexafluoride stored. The table includes Secretariat estimates and small amounts recovered through mine rehabilitation. Its 2025 column (66,074 tonnes) is a forecast, not completed production. 18
| Leading producer | 2024 production, tonnes uranium |
|---|---|
| Kazakhstan | 23,270 |
| Canada | 14,327 |
| Namibia | 7,332 |
| Uzbekistan | 5,200 |
| Australia | 4,656 |
| Russia | 2,738 |
| World, including other producers | 61,924 |
Source: Red Book 2026, Table 1.18, printed pages 77–78. Country figures include the report’s estimates; Namibia is marked as a Secretariat estimate. The same report lists 2024 reactor-related uranium requirements of 64,584 tonnes. Crucially, it defines requirements as anticipated acquisitions, not necessarily consumption. The report states that 2024 production met approximately 96% of world reactor requirements, with the remainder from secondary sources such as inventories, reprocessing, underfeeding, re-enrichment of tails and down-blended highly enriched uranium. That ratio compares two annual totals; it does not trace the fate of every atom mined that year. 18
The principal commercial pathway is uranium recovery, concentration into a product commonly called yellowcake, conversion and—where the intended fuel requires it—enrichment, followed by fabrication. In the common enriched-fuel route, concentrate is converted to uranium hexafluoride for enrichment and subsequently into a fuel material such as uranium dioxide. Some reactors—such as Canada’s CANDU units and two Argentine reactors described in the Red Book—use natural-uranium fuel and bypass enrichment. 20 18
Natural uranium contains roughly 0.7% uranium-235. NRC describes enrichment producing 3 to 5 percent uranium-235 for nuclear fuel and a depleted stream of about 0.3 percent. Enrichment separates material into a stream richer in uranium-235 and a depleted stream; it does not transform all incoming uranium into enriched fuel. 8
An illustrative mass balance makes the point. Starting with natural uranium at 0.711% uranium-235, producing 4% material and leaving tails at 0.25% yields roughly 12% enriched product and 88% depleted tails by uranium mass. This is arithmetic for those specified assumptions—not a measured global end-use percentage, and not applicable to every fuel route. Different product and tails concentrations change the result.
Inventories further complicate any annual accounting. EIA’s report released in 2026 puts year-end 2025 U.S. commercial uranium inventories at about 170 million pounds of U₃O₈-equivalent, up from about 156 million in 2024. It describes ownership of material at different stages of processing, including material at facilities outside the United States. An inventory is neither new mine production nor material already consumed. Secondary supply can include existing stocks and recycled material. 19 18
Depleted uranium has applications in counterweights, armor and armor-piercing ammunition. NRC says commercial demand is much smaller than the amounts generated. Storage, deconversion to more stable chemical forms and eventual disposition are therefore part of the story, not an accounting footnote. 8 20
Nor should fresh mining be equated with weapons demand. Research reactors, naval fuel and weapons inventories are distinct categories. Y-12 describes supplying U.S. naval-reactor feedstock from dismantled weapons and says U.S. enrichment for naval reactors stopped in 1992. That is a documented U.S. pathway, not a global military allocation. This investigation establishes no precise worldwide civilian-versus-military percentage. 21
Finally, three materials commonly blurred together are different: mine and mill tailings are residues of ore recovery; enrichment tails are uranium depleted in uranium-235; spent fuel is irradiated fuel containing remaining uranium, plutonium and fission products. Reprocessing can recover materials for reuse; NRC’s fuel-cycle outline still ends with final disposition of used fuel or high-level waste. NRC says no recycling or reprocessing facility or federal waste repository is currently licensed in the United States, and spent fuel is in interim storage. 18 20
The factories did not all have the same afterlife
Fenton: the factory was demolished, not turned into classrooms
Corning’s Rakow Research Library documents that Fenton ceased production of its traditional handmade glass products in July 2011—after the 2007 closure decision described by the company historian. In 2017 the library acquired a 290-box company archive, including formulas, trade catalogs, and material relevant to labor and industry history. An archive exists; this report has not inspected those boxes or established that they contain medical surveillance files. 22
Contemporaneous reporting documents the school board’s purchase signing in June 2017. Later reporting records factory demolition in 2018, and September 2020 opening coverage places the new Williamstown Elementary School on the former factory site. The school’s own website confirms its Caroline Avenue address. This is replacement construction on former industrial land, not reuse of the old glassmaking building as a school. 24 23 25 27
The regulatory history deserves equally careful language. A July 2016 News and Sentinel report, drawing on NRC records in the ADAMS database, described an August 2010 inspection with three reported violations: possessing depleted uranium oxide above the authorized limit, failing to perform the required periodic review of the radiation-protection program, and failing to provide annual training. After the license expired in 2015, Fenton sought termination, which required a final status survey. The paper quoted NRC health physicist Dennis Lawyer, from an ADAMS phone record, saying an earlier submitted survey “did not appear to show that the area could be released” and “actually showed areas where it appeared that the location did not meet requirements.” The report said a revised Final Status Survey Plan was approved on June 7, 2016, and quoted George Fenton saying “preliminary results look good.” That is a company characterization of preliminary results, not an NRC release finding. 26
Direct retrieval of the NRC license-termination document and the complete final survey did not succeed in this review: the NRC document server returned HTTP 403 to every request. Search-engine index entries describe an NRC record titled as Fenton’s license Amendment No. 11, but this report could not open it and treats that description as an unverified lead, not a confirmed termination date or release finding. Accordingly, this report does not claim to have independently verified the final radiological release determination, its measurements or its full scope. The earlier deficient survey and the access gap are not evidence that clearance was ultimately absent, that cleanup was omitted or that the school is unsafe. Nor do the reported violations establish a worker disease rate.
Newell: pottery continued, while the product changed
Homer Laughlin’s Newell complex illustrates continuity rather than abandonment. The company timeline records modernization of plants, Fiesta’s 1986 revival, and the 2020 sale of foodservice divisions and brand names followed by renaming as the Fiesta Tableware Company. NIST’s history identifies Newell as the manufacturing location associated with Fiesta. These records do not justify treating contemporary products as though they used the historical uranium glaze. They also do not establish the medical outcomes of the earlier glaze workforce. 5 3
A state-held record gives a more concrete view of later workplace hazards. Homer Laughlin’s January 2016 air-permit application includes a May 27, 2015 material safety data sheet for Arctic White Glaze listing crystalline silica (about 15.9%) and cobalt, noting that IARC, OSHA and NTP consider them carcinogens or probable carcinogens, and recommending a NIOSH-approved respirator when spraying without adequate ventilation. The same sheet also calls the glaze “a non-hazardous material” in its health-hazard section—an internal inconsistency the document does not resolve. This is evidence of hazard communication and specified controls—not proof of historical uranium exposure, actual compliance or medical surveillance. A later formulation cannot establish what the 1930s workforce breathed. 36
Cambridge: a preservation proposal did not save the plant
The National Cambridge Collectors’ archive identifies seven uranium-containing color formulas using historical batch-book material. Its health commentary is not used here as medical evidence. A separate October 1990 newspaper account reprinted by the organization says the company ceased operations in June 1954, reopened for several years under a worker-owned stock program until 1958, saw occasional storage and warehouse use afterward, and was being razed in 1990 after mid-1980s efforts to buy and convert the plant did not go forward. 28 29
The demolition contractor projected at least a year of work; the article does not establish a completion date. This report has not established the parcel’s current use or a radiological release history. A museum preserving the company’s products should not be mistaken for proof that its original factory survived.
Boyd: a glassworks passed to another glassmaker
A specialist historical account traces Boyd’s 1978 start to the purchase of the former Degenhart factory and reports the owners’ retirement and sale of remaining stock in 2014. ORAU’s explicitly dated 2004 account lists Boyd among uranium-glass producers. Boyd’s company page supplies family chronology (opening October 10, 1978) but still says Boyd Glass “still manufactures” collectibles; that undated page alone cannot establish current production, and it conflicts with the specialist account of a 2014 retirement and stock sale. Subsequent land use and environmental disposition were not verified. 30 1 31
A published 2014 interview provides a narrower glimpse of the work itself. John Boyd, who described adding uranium to batches, recalled wearing coveralls and a respirator, positioning ventilation to avoid dust, and leaving coveralls at the factory. He also discussed silica, cobalt and arsenic. These are a participant’s recollections reported by Collectors Weekly—not exposure measurements or proof that controls always worked. They are useful precisely because they describe a workplace mixture rather than reducing every hazard to uranium. This report does not adopt that article’s broad health reassurances. 35
Across these four consumer-product businesses, targeted searches did not locate a uranium-specific longitudinal health study establishing workers’ eventual outcomes. Searches included the biomedical literature, CDC/NIOSH-indexed material, NRC records, institutional histories and published company/archival descriptions. We did not examine private employment medical files, conduct interviews or exhaust all state and local records. “Not located” is a description of this investigation’s coverage, not a declaration that no surveillance ever occurred.
A new use does not erase an old obligation
Other industrial hazards complicate cleanup and responsibility. A West Virginia Brownfields account of Fostoria’s former site describes arsenic, metals and inorganic contamination, a brownfield grant and preparations for redevelopment. It does not establish uranium as the contaminant responsible. Assigning that cause simply because the business made colored glass would manufacture a connection the record does not supply. 32
Fernald provides a separate, clearly identified nuclear-industrial comparison—not a proxy for a dinnerware plant. DOE says uranium processing for the weapons production cycle ran from 1951 to 1989. Cleanup and ecological restoration were completed in 2006 except for groundwater, at $4.4 billion. A former warehouse became the visitor center; groundwater treatment and monitoring of the on-site disposal facility continue. The remedy was designed for an agreed future land use. A preserve with ongoing stewardship is not equivalent to an unrestricted parcel with no remaining obligations. 33
Worker compensation also has its own evidentiary meaning. A Department of Labor circular documents special exposure-cohort classes for some Fernald workers after NIOSH determined certain historical doses could not be reconstructed adequately. Eligibility rules protect qualifying people amid uncertainty; they are not an epidemiological finding that every eligible worker’s illness was caused by uranium. 34
There is room to appreciate the craft in a glowing cup and still ask who handled the powders, who kept the exposure records, and who carries the cost when a product or factory disappears. Workers and neighbors deserve more than an invented cancer story. They also deserve more than a comforting absence of data.
The responsible ledger keeps its categories intact: attraction is not therapy, fluorescence is not dose, a biomarker is not a diagnosis, a license violation is not a disease finding, and a new use is not a complete cleanup record. Following uranium honestly means keeping people visible while refusing to fill the gaps in their histories with claims the evidence cannot bear.
Reporting method and references
This is document-based reporting, not a systematic review, a collection-specific radiation survey or a site safety assessment. Six peer-reviewed studies were examined at different levels of access: the 2002 water study in full text (an OCR text of the scanned paper retrieved through NCBI’s public Entrez programming interface) plus its published 2003 correction; the 2006 water study as an abstract only; the leaching study as an abstract only, with additional methodological detail from ORAU’s account; the miners and Mallinckrodt papers in full text through the same NCBI interface; and the Fernald paper as a live abstract, with its methods and limitations read from a full-text copy retained earlier on November 7 because the live full-text page returned a bot challenge during this revision. Bibliographic details for all six were checked against Crossref DOI records. Government toxicology, fuel-cycle records, the 2026 Red Book, museum archives, manufacturer pages and explicitly identified secondary reporting supplied context. No interviews or outside contacts were conducted. On November 7, 2026, CST, 35 of 36 references were reopened and read at the scope stated in the companion source-verification file; the Fernald full text is the one reference with a limitation. NRC’s Fenton termination record and final survey could not be retrieved and are not cited as verified sources.
Anyone referenced in this report who wishes to respond or point out an error may do so through our website contact form. No one was contacted for this report.
[1] ORAU Museum, Vaseline and Uranium Glass. Read source
[2] ORAU Museum, Fiestaware (ca. 1930s). Read source
[3] NIST, Dinner Is on Us: NIST and the Creation of Fiestaware (2017). Read source
[4] ORAU Museum, Radioactive Curative Devices and Spas. Read source
[5] Fiesta Tableware Company, Timeline History. Read source
[6] Zbynek Votocek Art Nouveau, Uranium glass. Read source
[7] Preciosa Ornela, Uranium glass. Read source
[8] NRC, Background Information on Depleted Uranium. Read source
[9] ATSDR, Toxicological Profile for Uranium (2013). Read source
[10] Kurttio et al. (2002), Renal effects of uranium in drinking water. Environmental Health Perspectives 110(4):337–342. Read source DOI: 10.1289/ehp.02110337.
[11] Kurttio correction (2003), Environmental Health Perspectives 111(12):A632. Read source
[12] Kurttio et al. (2006), Kidney toxicity of ingested uranium from drinking water. American Journal of Kidney Diseases 47(6):972–982. Read source DOI: 10.1053/j.ajkd.2006.03.002.
[13] Landa and Councell (1992), Leaching of uranium from glass and ceramic foodware and decorative items. Health Physics 63(3):343–348. Read source DOI: 10.1097/00004032-199209000-00012.
[14] Kelly-Reif et al. (2023), Radon and lung cancer in the pooled uranium miners analysis (PUMA): highly exposed early miners and all miners. Occupational and Environmental Medicine 80(7):385–391. Read source DOI: 10.1136/oemed-2022-108532.
[15] Silver et al. (2013), Mortality and ionising radiation exposures among workers employed at the Fernald Feed Materials Production Center (1951–1985). Occupational and Environmental Medicine 70(7):453–463. Read source DOI: 10.1136/oemed-2012-100768.
[16] Milder et al. (2024; online 2023), Third mortality follow-up of the Mallinckrodt uranium processing workers, 1942–2019. International Journal of Radiation Biology 100(2):161–175. Read source DOI: 10.1080/09553002.2023.2267640.
[17] EPA, Radioactivity in Antiques. Read source
[18] NEA/IAEA, Uranium 2026: Resources, Production and Demand. Read source
[19] EIA, Uranium Marketing Annual Report, 2025 data released 2026. Read source
[20] NRC, Stages of the Nuclear Fuel Cycle. Read source
[21] Y-12 National Security Complex, Naval Reactors. Read source
[22] Corning Museum of Glass, Recent Acquisition: Fenton Art Glass Company Archive (2017). Read source
[23] James Dobbs, Foundations: Fenton Art Glass lives on through its legacy (2023). Read source
[24] News and Sentinel, Wood County Board of Education signs paperwork to buy Fenton site (2017). Read source
[25] Brett Dunlap, Officials cut ribbon to open new Williamstown Elementary School (2020). Read source
[26] News and Sentinel, Fenton to decommission art glass site in Williamstown (2016). Read source
[27] Wood County Schools, Williamstown Elementary home page. Read source
[28] Les Hansen, Uranium Glass, National Cambridge Collectors (2004). Read source
[29] National Cambridge Collectors, Cambridge Glass plant building is coming down (1990). Read source
[30] Glass Bottle Marks, Boyd’s Crystal Art Glass. Read source
[31] Boyd’s Crystal Art Glass, Our history. Read source
[32] West Virginia Brownfields, Work Progressing at Fostoria Site (2014). Read source
[33] DOE Office of Legacy Management, Fernald Preserve, Ohio, Site Fact Sheet (PDF). Read source
[34] Department of Labor, EEOICPA Circular14-02 (2013). Read source
[35] Ben Marks, These People Love to Collect Radioactive Glass. Are They Nuts? (2014). Read source
[36] Homer Laughlin China Company, January 2016 air-permit application, Attachment H (West Virginia DEP repository). Read source
