In the autumn of 1893, Edmund Schilder, a 61-year-old master glassmaker in Cologne, wrote a single cryptic sentence in the margin of a production ledger: “The depth of the blue is not in the sand. ”
That ledger still survives in a private family collection in the Rhine Valley. The marginal note appears on the page recording the final batch of glass produced at Schilder’s workshop before the building was acquired and the furnaces dismantled, ending 43 years of family production in a deal whose terms were never disclosed. Schilder spent his entire career producing colored glass for the restoration and construction of Gothic Revival cathedrals across central Europe.
His clients included the Cologne Cathedral chapter, the restoration committee for Reims Cathedral in France, and the building committee for the new Votivkirche in Vienna. His glass was installed in church windows across three countries and was praised by architects and approved by restoration engineers overseeing some of the most important church construction projects of the 19th century. When his workshop was absorbed into a unified glass manufacturing company backed by a Frankfurt investment house, the formula he had spent four decades developing and refining retired with him. It was never written down in any form usable by anyone outside the family.
The world in which Schilder worked was defined by one of the great cultural ambitions of the 19th century: the Gothic Revival. This movement sought to restore and reinterpret medieval church architecture and had gained momentum across Europe since the 1790s. By the time Schilder became a skilled craftsman in the 1850s, the movement had become an industry. Cathedrals that had stood for centuries in various states of ruin or incompleteness were now being finished, restored, and sometimes substantially rebuilt.
Cologne Cathedral, the largest Gothic church in northern Europe, had kept an unfinished nave and medieval crane above its south tower since construction stopped in the 16th century. Work resumed in 1842 and continued for four decades. The demand for craftsmen able to produce materials matching medieval originals was enormous. In nearly every case, the most sought-after and most difficult material to supply was the windows.
Medieval stained glass was not the product of industrial chemistry. It was the product of workshop traditions developed over centuries through the accumulation of precise, secret technical knowledge passed from masters to apprentices under conditions never designed for documentation. The colors, especially the deep blues and reds that distinguished the most famous medieval windows, resulted from techniques whose exact standards were never systematically recorded. The craftsmen who produced the great windows of Chartres and the Sainte-Chapelle in the 12th and 13th centuries knew what they were doing.
But they never explained it in writing in a way that survived the following 700 years. By the time the Gothic Revival created steady, well-funded demand for glass rivaling medieval originals, the knowledge needed to produce it existed only in fragments. There were published accounts, notes from chemists who analyzed surviving medieval glass and attempted reverse engineering, and practical experiments by master craftsmen who spent their careers trying to imitate the originals. The results were inconsistent.
Some workshops produced acceptable glass. Far fewer produced glass truly comparable to medieval work in depth of color and clarity. Schilder’s workshop, by the testimony of the architects and restoration committees that commissioned work from it over four decades, belonged to the latter category. Published literature of the movement, architectural journals, and cathedral chapter reports from the 1850s through the 1880s returned repeatedly to the same difficulty.
Buildings could be reconstructed according to original plans and structural principles. Stone could be cut to the correct shapes, proportions respected, and buttresses placed where medieval builders placed them. But glass was another matter. Every major architect involved in Gothic Revival work understood that windows were the instrument through which the spatial and theological logic of the building was completed.
A Gothic cathedral with the wrong glass was, in the assessment of the period’s most serious practitioners, a building whose central function had not been realized. The stone envelope existed. What the stone envelope had been built to contain did not. Many explanations were offered for what made medieval glass different.
Some attributed the difference to the manganese content in medieval mixes, which was higher than in contemporary production. Others pointed to medieval wood-firing techniques, which produced thermal profiles and reducing atmospheres in furnaces different from 19th-century coal furnaces. Others cited the potash content of the mix, which varied with the source of wood ash used as a flux, affecting the refractive index of the finished glass in ways not fully understood. None of these explanations was sufficient to explain the visual behavior of the finest medieval examples.
Chemists who analyzed medieval glass from the great French and German cathedrals could identify its components. They could not reproduce the result. The difference between identifying what a thing contains and understanding what it does is precisely the gap revealed by Schilder’s production records. The supplementary metallic preparation was not listed in standard analysis of glass composition.
It was the thing analysis did not find because no one had yet named what they were looking for. Johann Schilder, Edmund’s father, established the workshop in Cologne in 1850 as an unremarkable family business. Johann trained in the traditional manner of the era, working under a series of workshop masters in the Rhine Valley before setting out on his own with a small furnace and a client base drawn from simple parish restoration projects. The work was competent.
It met the standards of the age. It did not distinguish itself. What changed the workshop’s trajectory was a journey Edmund Schilder made in the summer of 1862, at age 29, to the monastery of Sankt Blasien in the Black Forest, where a community of Benedictine monks had maintained a glass workshop in continuous operation since the 15th century. Monastery records show Schilder visited twice, first in the summer of 1862 and again in the spring of 1863.
What he was shown, told, or observed during those visits exists in no document available to the public. What is clear from the workshop’s production records is that something changed after the second visit. The glass produced after 1863 was tangibly different from what had come before. Architects and restoration committees who received the new work noticed the difference in their correspondence.
Their language was consistent. They described the color as deeper, the clarity as more uniform, and the surface texture as closer to medieval originals than anything they had previously seen from a contemporary manufacturer. Schilder’s reputation grew rapidly through the 1860s and 1870s. Commissions became larger and clients more significant.
By the 1880s, he was supplying glass to some of the most prominent Gothic Revival projects in central Europe, with an order backlog stretching years into the future. The workshop’s production records are detailed in some respects and completely silent in others. They record quantities, dimensions, color specifications, and delivery dates. They record furnace workers and cutters.
They record raw material costs by category, showing expenditures on silica, metal oxides, and firewood, along with a category consistently listed as “supplementary metallic preparation,” with no further explanation. Expenditures in that last category are not trivial. They appear consistently across decades of peak production as a routine and apparently essential cost of doing business. They appear in no document explaining what the supplementary metallic preparation was, where it came from, or how it was used.
The only person recorded as handling the supplementary metallic preparation category was Edmund Schilder himself. The furnace workers and cutters, skilled men who worked at the Schilder facility for years, could not, in accounts given after the workshop closed, describe what happened in the preparation stage. Schilder carried it out alone in the side room he converted from a storeroom into a private work space sometime in the late 1860s. The door was closed.
The room had no window. The work that happened there could not be reached by any account from the furnace floor. At peak production in the 1870s and 1880s, 14 men worked at Schilder’s furnaces. Many spent their entire careers there.
None was ever permitted into the preparation room while Schilder worked. The architectural record from 1863 to 1893 records what Schilder’s formulation produced. Windows installed in the nave of Cologne Cathedral during the great completion campaign of the 1880s include panels from the Schilder workshop that restoration specialists still count among the finest examples of Gothic Revival glass of the 19th century. The blue panels in particular, and the deep cobalt passages that anchor the color structure of the nave windows, are described in restoration reports from the 20th century in terms that seem to embarrass the specialists themselves.
The reports note that Schilder’s panels show optical properties that distinguish them from surrounding glass produced by other manufacturers of the period, and that those properties correspond more closely to characteristics found in medieval glass than to those of contemporary industrial production. Restoration specialists who examined the panels analyzed them using methods available in the 20th and 21st centuries but could not fully explain the glass’s optical behavior chemically. The composition, as far as non-destructive analysis could determine, does not directly differ from other glass of that period. Silica content falls within the normal range.
The metal oxide concentrations that produce color are not significantly unusual. Yet the glass behaves differently. The light passing through it distributes differently. The depth of color is greater than the chemical composition alone seems to justify.
The phrase “optical anomaly” appears in more than one report. The phrase “requires further research” appears in more than one conclusion. Further research, as far as published records show, has not produced a resolution. The consolidation of the European glass manufacturing sector in the early 1890s was one of the least documented industrial reorganizations of the era.
Large manufacturers, backed by capital from Frankfurt and Berlin investment houses restructuring German industry in the years following unification, moved to acquire the small independent workshops that formed the artisanal side of the market. The process was not dramatic. It involved offers, most negotiated privately, and the gradual absorption of small operations into larger companies. The Schilder workshop was acquired in the autumn of 1893 by a glass manufacturer whose major financial backers included investment interests linked to the Disconto-Gesellschaft, one of the largest Frankfurt banks of the time.
The terms of the acquisition do not exist in public records. The deal appears in the Cologne commercial register as a direct transfer of business assets. The workshop buildings, furnaces, tools, client relationships, and raw material inventory all passed to the acquiring company. What the acquiring company received in technical knowledge was not recorded, because that knowledge resided in Edmund Schilder’s head.
Architects and restoration committees who had relied on the Schilder workshop discovered the practical consequences of the acquisition in the years that followed. Orders placed with the new consolidated manufacturer for glass to specifications the Schilder workshop had met precisely for decades came back non-compliant. The colors were close but not right. The blues in particular differed in a way architects found difficult to describe precisely but impossible to ignore once the two versions were seen side by side.
The Cologne Cathedral restoration committee wrote to the consolidated manufacturer in 1895, 1896, and 1897, demanding that the glass supplied for ongoing restoration work match the standards established by Schilder’s previously installed panels. The manufacturer’s responses, preserved in the cathedral administration archive, were consistent. They confirmed that the glass supplied met all specified technical standards, and that any perceived difference was a matter of personal judgment. The correspondence ended without resolution.
Panels installed in restoration campaigns after 1893 are visibly different from Schilder’s panels to any observer standing in the nave and looking carefully. Conservation records acknowledge this. Restoration documents from the period acknowledge it. Official accounts attribute the cause to normal variation in manufacturing techniques between suppliers.
No commission was assigned to investigate whether the difference had a remediable cause. No one contacted Edmund Schilder to ask. Schilder retired to a house in the Eifel, a region of low hills south of Cologne, and lived there in reasonable comfort until his death in 1911. He gave no public interviews.
He published no memoirs. He did not attempt to transfer his technical knowledge to any public institution or professional body. The only documents recording his voice in any sense during the last two decades of his life are a series of letters to his son Karl Schilder, who did not enter the glass trade and worked as a surveyor in the Ruhr region. Those letters remain in the Schilder family’s possession and have not been made public.
What is known of their content comes from a single account given by Margarete Schilder, Karl’s daughter, to a German journalist named Wilhelm Wester in 1938. While researching a series of articles on traditional crafts in the Rhine Valley, Wester was told by Margarete that her grandfather’s letters to her father dealt with ordinary matters: management of his retirement property, family news, and his views on local affairs. She said that in the last three years of his life, the letters became shorter and more focused. In the final letter before his death, which she read as a young woman and described to Wester from memory, Edmund Schilder told his son that the preparation he used in the glass was not something he had invented himself.
It was something that had been shown to him, at Sankt Blasien in 1863, by a monk who had himself received it in the same manner from an older source. He wrote that the preparation involved a metallic compound whose origin and processing he was unwilling to describe in writing, because he had given the monk his word that he would not. He wrote that the compound changes the way light moves through the glass rather than the color the glass displays, and that this is why the analyses of chemists who examined his glass could not explain what they were seeing. He wrote that he had not passed the knowledge on because the monk’s instructions required that it go to a craftsman who would use it as it had always been used: in the service of buildings where light through windows is the purpose of the whole undertaking.
He stated he had not encountered, in the final decades of his career, a craftsman matching that description within an industry reorganized around the production of uniform glass. He did not name the metal. He did not describe the method of preparation. He closed the letter by telling his son that the marginal note in the production ledger, the one saying “the depth of the blue is not in the sand,” had said everything he intended to say, and that anyone who understood the craft well enough to work in it would understand what it meant.
Wester’s articles on Rhine Valley crafts were published in a regional newspaper in 1938 and never collected or reprinted. They were not widely read. The reference to the Schilder letter appeared in the third article of the series, within three paragraphs, and was never followed up. Wester died in 1943.
His research notes, which may or may not have contained a fuller account of the conversation with Margarete Schilder, were in his Cologne apartment at the time of his death. The apartment was destroyed during the Allied bombing campaign in the summer of 1943. The research notes did not survive. Sankt Blasien has a complicated history.
The original Benedictine institution was dissolved in 1806 during the Napoleonic secularization campaign, and the community dispersed. The monastery buildings were repurposed by the Baden state administration for various uses throughout the 19th century. A Benedictine community returned to Sankt Blasien in the 1860s, although the building that housed the glass workshop before the dissolution had been converted to other uses. The monastery’s records prior to dissolution, including any documents the glass workshop may have produced, were transferred to state custody at the time of the dissolution.
Those records are now held in the Generallandesarchiv in Karlsruhe. The portion of the archive relating to the Sankt Blasien glass workshop has been indexed. The index entries indicate the relevant records consist mainly of financial accounts and correspondence with clients. Technical records of the workshop, if they ever existed in written form, do not appear to be among the surviving materials.
A researcher named Hans-Jürgen Bühler, working at the University of Freiburg in the 1960s, spent several years examining the Sankt Blasien holdings as part of a broader study of monastic craft industries in southwest Germany. He published a study in 1968 that included a chapter on the glass workshop. The chapter is meticulous and well-documented with regard to the financial and commercial aspects of the workshop’s history. On the question of technical methods, the chapter acknowledges that no systematic record of the workshop’s glassmaking operations appears to have survived the dissolution.
In a footnote that received no further attention in the main text, Bühler notes that one of the monks present at the dissolution later gave a brief account to a leader in the dispersed community, indicating that some material from the workshop had been removed and transported before state officials arrived to take control of the building. The monk’s account does not describe what that material was, where it was moved, or whether it reached any destination where someone could use it. That footnote is the last mention of the subject in Bühler’s study. The glass Edmund Schilder produced between 1863 and 1893 is not everywhere.
It exists in specific buildings in Cologne, Aachen, numerous small churches in the Rhine Valley, and panels supplied for the Votivkirche restoration project in Vienna. The quantity is not vast. The Schilder workshop was a small operation throughout its history, never employing more than 14 workers at its peak. Its total glass production over four decades, though significant, represents a small fraction of the glass installed in Gothic Revival buildings in central Europe during that period.
What exists is enough for examination. Conservation scientists working on Cologne Cathedral in the late 20th and early 21st centuries produced a body of technical literature on the optical properties of the nave windows notable for what they could not resolve. Spectral analysis of Schilder’s blue panels shows absorption and transmission properties that are reproducible in measurement but not fully explained by the specific chemical composition. The panels transmit light in a way multiple analyses have confirmed, but none has been able to attribute to a specific component of the glass batch.
The language of the conservation literature on this point becomes wary to the point of circumspection. The architectural context matters here. Gothic Revival architecture was not merely a stylistic preference. The architects and theologians who led the movement understood the medieval cathedral as a theological instrument, a building designed to produce a specific experience of light as a medium of the sacred.
The windows of the great medieval churches were not decorative. They were functional in the precise sense: they transformed natural light into something the liturgical tradition understood as participation in the divine illumination described in the theology of the period. The quality of that transformation depended on the quality of the glass. Medieval glass craftsmen, working within the theological and architectural tradition of the High Gothic period, were not producing a commercial product.
They were producing a component of a larger instrument, and they understood their component’s function within that instrument. The knowledge they accumulated over generations of workshop practice was knowledge oriented toward a specific goal: producing colored glass that transformed light in ways that served the purposes of the buildings in which it was installed. The 19th-century Gothic Revival attempted to rebuild that tradition with the resources of the industrial age. It succeeded in many respects.
The structural ambitions of the original cathedrals were matched and sometimes exceeded. Stone carving was accomplished at a high level. The overall architectural effect, in the best examples, was genuinely impressive. Glass was the persistent problem.
The 19th-century industrial glass sector could produce colored glass in quantity and with consistent specifications. It could not, except in isolated and poorly understood cases, produce glass that behaved optically the way medieval originals did. It was isolated workshops like Schilder’s that maintained a thin thread of connection to an old way of working that the industrial model had not absorbed or standardized. By the end of the 1890s, most of those workshops had disappeared.
The industrial consolidation absorbed their commercial operations. The monastic workshops that had preserved certain craft traditions through the pre-industrial period had been damaged by the early 19th-century secularizations and never fully recovered. The thread was cut, and its severance coincided precisely with the period when the most ambitious Gothic Revival projects were entering their final phases, with demand for glass matching medieval standards at its peak. The consolidated manufacturers, who supplied most of the glass installed in Gothic Revival churches from the 1890s onward, produced a technically acceptable product.
The light that passes through it is colored. The windows are legible as images. The overall visual effect of a Gothic Revival church interior lit by such glass is the visual effect that era wanted. The specific optical quality that distinguished the medieval originals, and that distinguished the glass produced by the handful of workshops that had preserved some of the old knowledge, is not present.
Conservation specialists who work regularly in medieval churches and Gothic Revival buildings built after the 1890s recognize the difference. They describe it in different ways. Some formulate it in purely technical terms as a measurable difference in spectral transmission properties. Others formulate it in terms that are difficult to quantify, describing the experience of light in a medieval church as differing in quality from the experience of light in a late-19th-century Revival building even when the architectural forms are largely similar.
Schilder’s panels in the nave of Cologne Cathedral sit within a larger field of glass comprising works by other manufacturers of the period. A visitor walking the nave and looking carefully will notice, under certain lighting conditions, that the panels are not uniform in their optical character. The difference appears most clearly in the blue sections, where Schilder’s panels maintain a depth and uniformity of color unmatched by the surrounding glass produced by other workshops. It is the kind of difference that does not announce itself but accumulates in perception over time.
Conservation specialists who have spent years working in the nave describe their sensitivity to this distinction in ways they find difficult to explain to people who have not spent comparable time in the space. Under controlled light, the surfaces of the panels show what the conservation literature calls a distinctive fine structure that does not correspond to surface features produced by any documented 19th-century manufacturing technique. The panels made by Schilder are consistent in texture across different production years, meaning they were not a random result but deliberately reproduced batch after batch for 30 years. That texture is absent from every panel produced by any other manufacturer of the period.
The marginal note in Edmund Schilder’s production ledger, saying the depth of the blue is not in the sand, is a statement about this distinction. Sand, specifically silica sand, is the primary raw material of glass production. To say the depth of the blue is not in the sand is to say that the optical quality of Schilder’s glass did not depend on the base material or the traditional coloring agents that formed the standard glass batch of the time. It points to the supplementary metallic preparation, the category of expenditure that appears consistently in Schilder’s production records without detail, which Edmund Schilder handled personally without assistance in a private work space kept away from the furnace floor.
What that supplementary metallic preparation was is unknown. Analyses of the surviving glass have not identified any component that would explain the optical behavior in terms the current analytical framework can fully interpret. The monastic connection, the two visits to Sankt Blasien in 1862 and 1863, points to a source of technical knowledge predating the industrial period, transmitted through an old monastic workshop tradition. According to surviving monastery records, the Benedictine glass workshop at Sankt Blasien had been in continuous operation since at least the 15th century.
Four hundred years of workshop practice accumulated by craftsmen working within a specific theological and architectural tradition cannot be reduced to a published formula. It exists in the kind of knowledge found in practice and in the hands of practitioners, knowledge passed through demonstration and observation rather than documentation, and it disappears when the chain of practitioners is broken. That chain was broken at Sankt Blasien in 1806, when the dissolution order dispersed the community. The material a monk was said to have removed before state officials arrived represents, if the account is accurate, an attempt to preserve something that could not be written down.
Whether that material reached anyone capable of using it is unknown. What can be said is that Edmund Schilder visited the re-formed community at Sankt Blasien in 1862 and 1863, in the early years of that community’s return to a building that had housed a centuries-old craft tradition, and that his glass changed after those visits in ways contemporary analysis cannot fully explain by reference to materials and methods analysis can identify. The Schilder family’s possession of the production ledger bearing the marginal note is the closest thing to a direct statement from Edmund Schilder about the nature of what he had been shown. His son Karl did not enter the glass trade.
His granddaughter Margarete gave one account to a journalist in 1938, published in a regional newspaper and never reprinted. The journalist’s notes were destroyed in the bombing of 1943. The ledger is in private hands. The panels remain in the nave of Cologne Cathedral, periodically examined by conservation specialists who note the optical anomaly in their technical reports and refer to it for further investigation that published records do not show was completed.
The workshop building no longer stands. The site in Cologne where Schilder’s furnaces worked for four decades was rebuilt in the early 20th century, bombed in 1943, and rebuilt again. The furnace room and the private work space where supplementary metals were processed are gone. Sankt Blasien is now occupied by a Jesuit educational institution that took over the buildings in the 1970s.
The glass workshop has not operated for more than a century. The public archives in Karlsruhe hold the pre-dissolution records. The gap in technical documentation noted by Bühler in his 1968 study has not been addressed in later research. In the archives of Karlsruhe, a researcher with appropriate qualifications and sufficient patience can sit with the bound volumes of Sankt Blasien’s financial accounts and read the details of the workshop in the bureaucratic handwriting of the 19th century.
The financial accounts record what the workshop charged for its glass. They do not record what the glass was made of. What remains is the light. On a clear autumn morning, when the sun strikes the south-facing nave windows of Cologne Cathedral at the angle the medieval builders designed the building to receive, the light passing through Schilder’s blue panels behaves the way Edmund Schilder spent 40 years learning to produce, the way the monk at Sankt Blasien showed him how to achieve in 1863, and the way a chain of craftsmen reaching back to the 15th century had transmitted by practice rather than documentation.
It ended, as such things end, not in a dramatic moment of destruction but in the quiet absorption of a small workshop into a unified industrial enterprise whose investors had no interest in the theological function of colored light in Gothic architecture. The ledger marginal note survived because ledgers are ordinary things, and ordinary things are not searched for when a workshop is acquired and dismantled. The panels survived because they were set into stone walls, and stone walls are durable. The knowledge that produced them did not survive because it was confined to one man’s mind, and that man retired to the Eifel and died in 1911 without finding, by his assessment, a craftsman matching the description given to him by the monk at Sankt Blasien as someone worthy of receiving the secrets of the craft.
The depth of the blue is not in the sand. A 2004 restoration report on the Schilder panels notes the optical anomaly and recommends additional spectral analysis. The follow-up was listed as pending.