Novoslobodskaya’s stained-glass windows: from a cathedral in Riga or commissioned for the Metro?

The stained glass at Novoslobodskaya station comprises 32 compositions created for the Moscow Metro to designs by Pavel Korin and made by craftspeople in Riga. The claim that finished windows from Riga’s cathedral were transferred to the underground hall appears even in a city guidebook. Records of the commission and production establish a different origin for the compositions themselves. However, the provenance of every original piece of coloured glass cannot yet be traced.

An existing church window would have retained its original design and dimensions. The 32 compositions for Novoslobodskaya were designed to fit the station’s pylons and to be artificially backlit. Documents allow us to trace the development of the designs and the assembly of the stained glass; the history of the individual sheets of coloured glass is less well understood.

Where did the story about existing cathedral windows come from?

The city guidebook ‘Moscow’s 100 Best Places’ explicitly states that the station’s stained-glass windows were originally made for the cathedral in Riga. The same page names the architects Alexey Dushkin and Alexander Strelkov and the artist Pavel Korin. The guidebook provides no transfer date, reference to a primary source or details of where these works were previously installed.

The phrase ‘stained glass from the cathedral’ could refer either to finished windows or to sheets of glass from old stocks; different documents are needed to verify each possibility. Tatyana Knyazhitskaya and Dmitry Vorobyov reportthat handmade sheets of coloured glass remained in Latvia after the war. Their 2023 article does not link this material to a particular cathedral or to a delivery for Novoslobodskaya. In a 2025 publication the same authors describe the story of pre-revolutionary stocks being stored in Riga’s cathedral as a legend.

Riga Dome Cathedral has its own stained-glass history: according to the cathedral’s website, its stained-glass windows were installed in the late 19th century, and some of the compositions on the south side were destroyed during the Second World War. This does not establish that any component or batch of glass reached Novoslobodskaya.

Why the word ‘cathedral’ is not proof

The guidebook does not specify which cathedral it means. Identifying it as Dome Cathedral without a transfer record would replace an unnamed building with a specific one: the history of its windows confirms their existence, but not their dispatch to Moscow.

A finished stained-glass window incorporates an image, a defined shape and a metal framework. The original sheets of coloured glass are a separate matter. Even if old material was used, the Riga workshop assembled compositions for Novoslobodskaya to Korin’s designs. Images or measured drawings of the earlier windows would be needed to support the claim that church windows were transferred.

Tatyana Fyodorova, who oversaw the station’s construction, mentioned another source of material: in her memoirs, she writes of coloured glass from Czechoslovakia. Her account concerns supplies for the construction work and does not establish the origin of every component in all 32 stained-glass panels.

The sources answer different questions. The guidebook repeats the claim about existing windows; E. A. Minukhin’s 1954 publication describes the ensemble’s production shortly after the station opened; a museum catalogue record names the Maksla production centre. Fyodorova’s memoirs concern the supply of glass. No production record linking each batch to each stained-glass panel has been found among the documents examined.

What confirms that the stained glass was made specifically for the Metro?

An article by E. A. Minukhin in the journal ‘Architecture of the USSR’ in 1954, written close to the time of construction, discusses the decision to install coloured compositions in Novoslobodskaya’s underground hall. According to him, a workshop was established in Riga and the first stained-glass panels designed for artificial lighting were commissioned for the station. This commission describes new work intended for a specific location and viewing conditions, rather than the transfer of existing cathedral windows.

The state historical and cultural assessment report of 2017 names A. N. Dushkin and A. F. Strelkov as the architects, attributes the designs for the coloured-glass panels to P. D. Korin, and states that they were made in the Riga workshops of the Latvian branch of the USSR Art Fund. The report also notes the use of gilded brass in the decorative finishes. The document distinguishes between the station’s architectural design, the artistic concept and production.

In their 2025 study, Korolko and Sokolova date Korin’s designs to 1951 and describe the ornamental patterns with medallions. Their introduction mentions thirteen stained-glass panels, but they later discuss the restoration of all 32. The figure of 32 is independently given by Minukhin in his 1954 publication and by Knyazhitskaya and Vorobyov in their 2023 article. Twenty-six of the upper medallions feature stylised flowers and stars; the other six are devoted to science, music, architecture, art, agriculture and industry. This decorative scheme relates to the design of a public station in the early 1950s.

How the commission fitted into the station’s construction

Novoslobodskaya opened on 30 January 1952 as part of the section of the Circle Line from Kurskaya to Belorusskaya; this date is given in the historical and cultural assessment report. Tatyana Viktorovna Fyodorova was appointed head of construction at Shaft No. 20 in December 1948. In her published memoirs, she describes how Dushkin and Korin visited the station while it was under construction and inspected the hall and the pylons intended to hold the stained-glass panels. Her account shows that the artist and architect were working with this particular interior before the station opened.

According to the assessment report, the architectural design was selected through a competition. In his memoirs, published in the illustrated volume ‘The Architecture of the Moscow Metro’, he wrote that after the design won the competition, he spent another two months refining the station with Dushkin’s involvement. His account offers a rare insight into how the stained-glass concept emerged: initially, they discussed relief decoration with Vera Mukhina and the use of uranium glass, before deciding to commission designs for ordinary coloured glass. The architect’s recollections reveal how different options were discussed, but do not replace production records.

Fyodorova recalls the first installed stained-glass panel being illuminated on 17 July: the Riga artists were at the station with Korin, the frame was adjusted on site, and the light was then switched on. In an entry dated 25 November, she writes that the coloured compositions were still on their way and installation was scheduled to begin on 1 December. Taken together, these two episodes allow for a trial installation before the entire series was fitted; they do not confirm that all 32 panels were already in place in July. In the published extract, the precise year of each diary entry is not given alongside the date, so we do not treat it as a precise production date in its own right.

The assessment report describes the soft light of the ‘grotto-like station’, while Minukhin writes that the stained glass was designed for artificial lighting. Here, the design, metalwork, lamps and niche dimensions are interdependent. Transferring an existing church window from a naturally lit building would have required alterations to its geometry and lighting; none of the sources examined contains records of such work.

Minukhin gives the dimensions of each composition: a metal framework measuring 1256 × 2988 mm and an area of 3.75 m². Each consisted of eight parts: three rectangular panels, two panels with cut-outs, two upper segments and a circular medallion 770 mm in diameter. This description shows how the design and structure were calculated to fit a single framework within a pylon. Dimensions alone cannot establish the provenance of the original sheets of glass; they describe the finished composition.

How the stained glass differs from the mosaic at the end of the hall

Korin also contributed to the design of ‘Peace Throughout the World’ at the end of the hall. The stained-glass panels are set into the side pylons and transmit light from behind; the panel on the far wall is assembled from opaque pieces. They use different materials and have undergone different changes. Tretyakov Gallery holds a 1951 design for the mosaic, with inventory number ZhS-6736, and describes the reworking of the composition in detail. This museum object confirms Korin’s work on the station, but says nothing about any previous location of the stained-glass panels.

This distinction also matters when reading repair reports. The historical and cultural assessment states that the stained glass, chandeliers and end-wall mosaic were restored in 2003. Combining these works into a single account could easily lead to changes to the wall mosaic being attributed to the side panels. For the question of the cathedral windows, it is specifically the information about the glass, frames and production in Riga that matters.

What roles did Korin and the Riga craftspeople play?

The stained glass made in Riga for Novoslobodskaya was based on Pavel Korin’s designs: the artist developed the compositions, and the craftspeople assembled them from glass and metal. A catalogue record from the All-Russian Decorative Art Museum attributes the production of the entire ensemble to Maksla. The production centre had workshops for assembling stained glass and facilities for cold-working glass. Knyazhitskaya and Vorobyov’s study adds that the enterprise was established in Riga in 1946, and that craftspeople skilled in traditional leaded and soldered stained-glass techniques were still working in Latvia at the time.

In 1954, Minukhin distinguished between the stages of the work: Pavel Korin painted eight original designs for Novoslobodskaya’s 32 stained-glass panels, while a team of Latvian craftspeople led by A. T. Vilkins produced the full-size working cartoons. In their 2023 article, Knyazhitskaya and Vorobyov attribute these drawings to the artist himself, without distinguishing between the stages of preparing the image. Both sources credit him with the artistic concept, but describe the production of the cartoons differently. These texts do not establish who prepared each working drawing. The Maksla craftspeople selected and cut the glass, assembled the pieces in lead cames and painted the details.

Published recollections of Novoslobodskaya’s construction describe the installation of the first stained-glass panel: Korin and young artists from Riga were present at the station, and the backlighting was switched on after installation. This account by a participant in the work describes the process of adjusting the panel on site. It shows how the artist and craftspeople worked together, although the provenance of each piece of glass remains unknown.

Why the commission went to Riga

By the late 1940s, Riga still had not only individual craftspeople but also an established production network. Knyazhitskaya and Vorobyov date the founding of Māksla and the department of artistic glassworking at the Riga Secondary School of Arts and Crafts to 1946. The enterprise had assembly premises and the necessary equipment; its craftspeople also worked with lead came. This was a tangible infrastructure without which assembling dozens of large compositions in a short time would have been difficult.

Researchers name Egons Cēsnieks, Jēkabs Bīne, Ernests Veilands and Arnolds Vilkins among those involved in the work. The expert assessment report gives a different short list of craftspeople working in the Riga workshops of the Latvian branch of the Art Fund. Neither list is presented as a complete roster for the commission; they cannot simply be combined into an exhaustive team list. What matters more to the reader is the documented division of labour: Korin produced the designs, and Māksla’s specialists executed the ensemble.

The damage described by conservators in 2003 reveals how the panels were assembled: individual pieces of glass are held by H-section lead came, while the joints and framework hold the entire composition together. Lead allows an intricate design to be assembled from small pieces of coloured glass, but requires precise fitting and responds to stress over time. Without this structure, the stained glass could not retain its shape within the station’s slightly curved pylon.

In a 1954 article, Minukhin listed other stained-glass commissions undertaken by the Riga workshop: for the Latvian SSR pavilion at the All-Union Agricultural Exhibition, the railway station in Sochi and public buildings. This is evidence of an active workshop producing new works. The list alone does not establish that their designs shared features with Korin’s sketches, or that the coloured glass sheets came from the same batch.

In their article on the Latvian SSR pavilion, Knyazhitskaya and Vorobyov trace this continuity through a specific example: the star motif, used in various ways in Novoslobodskaya’s medallions, also appeared in the pavilion’s central window. This reflects the development of a decorative device by the same artistic team, rather than the transfer of finished glass panels between sites. The comparison helps explain that Māksla was an active workshop creating new works, rather than a storehouse of dismantled church windows.

Why does the stained glass resemble church windows if it was made for the station?

Coloured glass and leaded construction bring church windows to mind because this technique was used in church interiors long before the metro was built. In 1954, Minukhin explicitly wrote that Novoslobodskaya helped overcome the perception of stained glass as something belonging to religious buildings. Similarity of technique tells us nothing about the previous purpose of any particular pieces of glass.

In his memoirs, Alexander Strelkov identified another source for the designs: according to the architect, Korin drew inspiration from the patterns on brocade garments displayed in the Palace of Facets while working on his sketches for the station. This is a later recollection by one of Novoslobodskaya’s architects, rather than a catalogue of every reference the artist studied. It explains the density of the stained glass’s ornamentation and helps distinguish the source of an artistic motif from the origin of the glass itself. Even a confirmed resemblance between a design and a historic textile tells us nothing about where the glass sheets were stored or what they were previously intended for.

At the station, stained glass works differently from glass in an external window opening. The compositions are set into the pylons and lit from within: the viewer sees their colours without daylight from outside. Minukhin particularly emphasised that the Riga commission was designed for artificial lighting. This dictated a requirement for the ensemble: the glass had to produce a clear design in an underground hall where the architecture itself provides the light.

The imagery also matters when assessing the theory. According to Korolko and Sokolova’s description, the upper medallions depict flowers, stars and six fields of human endeavour. Riga Cathedral’s webpage describes its historic windows in terms of events in church and city history, and the lost southern windows in terms of scenes from the Gospels. These are different artistic programmes; a shared technique does not mean that one ensemble was once the home of the other.

What the designs of the thirty-two compositions reveal

All the stained-glass panels share a common layout, with a decorative field and an upper medallion, yet Korolko and Sokolova find no two identical designs. Flowers, birds, vases and geometric shapes intertwine in the ornamentation. In twenty-six medallions, the design centres on stylised flowers and stars. This is more than incidental decoration for empty niches: the shared layout unifies the long hall, while variations in detail give each panel its own identity.

The remaining six medallions stand out for their figurative subjects. A 2025 study identifies these as science, music, architecture, art, agriculture and industry. In the painted medallions, the artist could refine figures and details where assembling coloured pieces alone was insufficient. The combination of ornament and an illuminated figurative inset gives passengers two levels of viewing: the overall rhythm of the pylons and an individual story in the upper section.

As Korolko and Sokolova point out, Novoslobodskaya’s stained-glass decoration was the first undertaking of its kind in the country’s metro systems. The approach was also a technical experiment in visual perception: a window in a pylon receives light from a lamp, whereas a conventional church window depends on changing daylight. In the underground hall, the colours could remain visible at any time of day. Minukhin’s observation about artificial lighting therefore has a direct bearing on the artistic design, rather than merely offering an evocative description of the station.

Can the origin of the glass itself be established?

The documents examined allow us to identify with confidence the artist who produced the designs and the place where the ensemble was made. They contain no delivery note for the original glass, inventory of the cathedral’s stocks or analysis of the composition of each fragment. It is therefore just as unfounded to claim that all the material used at Novoslobodskaya was new as it is to describe all 32 stained-glass panels as ready-made cathedral windows.

Knyazhitskaya and Vorobyov report that Latvian workshops held stocks of handmade coloured glass at a time when production of the required range in the USSR was still developing. This information leaves open the possibility that older material was used, but does not establish where the particular batch used for Novoslobodskaya had been stored. Such a conclusion would require delivery records or a physical examination of the glass.

Comparing the sources yields a precise, if limited, conclusion. The guidebook’s claim that the completed stained-glass panels were previously intended for a cathedral is inconsistent with the documented metro commission and Māksla’s work. The hypothesis concerning older Riga glass remains a separate question: the known information about the production facilities does not rule it out, but the origin of individual fragments has not been established here.

What is known about the supply of coloured glass

The most specific account of the material in the sources examined comes from Tatyana Fyodorova, who oversaw the station’s construction. In the chapter ‘Three Years in the “Sloboda”’, she writes that the coloured glass for the panels arrived from Czechoslovakia. In the same paragraph, Fyodorova attributes the chased ornamentation to Māksla’s craftspeople in Riga. As someone involved in the construction, she thus distinguished between the origin of the material and the place where the artistic work was carried out. Her account provides strong grounds for questioning the theory that Riga Cathedral was the sole source of the glass.

Fyodorova’s recollection, however, was published after construction and is not a surviving delivery note covering every batch. It does not tell us how much coloured glass came from Czechoslovakia, whether local stocks were also used or which panels incorporated glass from each delivery. The scholarly article on the Latvian workshop allows for the presence of older handmade material in the republic, but does not link a particular store of glass to these 32 compositions. Both accounts can be considered together without treating either as an exhaustive record of supplies.

Proving that the glass came from a cathedral would require a chain of specific evidence: a document recording its transfer from the named building, a record of its arrival at the workshop and an indication that it was used in particular works. The theory that completed windows were transferred would also require historic images or measured drawings matching the present compositions. No such chain appears in the documents examined. Its absence from the available sources does not prove that it never existed somewhere in an archive.

The known restoration history makes assessment ‘by eye’ still more difficult. During repairs ahead of the 1980 Olympics, some missing pieces were replaced with acrylic, while in 2003 missing elements were replaced with glass from various manufacturers. Even if one fragment in a present-day panel can be dated, that date cannot automatically be applied to all the others or to the original 1951 assembly. Establishing the material’s origin requires examining specific areas of a particular composition and the history of its repairs.

What changed during the 2003 restoration?

The restoration involved all 32 stained-glass panels, so the ensemble’s present appearance cannot be regarded without qualification as the untouched work of 1951. Korolko and Sokolova’s study draws in part on a 2024 interview with conservator Vladimir Mikhailov. The authors describe his workshop’s methods: damage was recorded on tracing paper, the coloured panels were examined on a light table, and removal from and reinstallation at the operating station took place at night.

Mikhailov’s craftspeople sought to preserve the historic fragments. Where appropriate, cracked glass was bonded, while missing pieces were replaced with glass closely matching their colour and texture; missing painted details were recreated. Distorted frameworks were straightened and resoldered. These measures do not mean that all the glass was replaced: the study carefully distinguishes between preservation, repair and the replacement of missing elements.

How the conservators worked

According to Korolko and Sokolova’s study, Vladimir Mikhailov’s workshop was invited in autumn 2002 to tender for the restoration of the glass compositions. The company Borey Techno was responsible for the metal surrounds, while Mikhailov and four other specialists worked on the stained glass. The project covered all 32 works and was due to be completed by March 2003. The workshop first received the panels from the central hall, followed by those facing the two platforms. This division helps explain why the researchers describe a single restoration of the ensemble, although each composition was treated individually.

The operating station could not be left without its decoration and protective barriers for long. According to the study, the panels were removed, taken to the workshop and returned at night. Empty niches were temporarily covered with white acrylic; installation began after the last passengers had left, at around 1 am, and finished at about 5.30 am. These details do not concern the original 1951 commission, but explain the constraints on the restoration and the sequence in which the works were treated.

Before repairs began, the craftspeople made a tracing of each stained-glass panel, marking cracks, missing pieces and proposed ways of accessing damaged areas. Fragments were examined on a light table using a magnifying glass and microscope. This sequence matters: immediately replacing a cracked section could destroy evidence of its technique and earlier interventions. The workshop prepared treatment plans, so decisions about dismantling were made for each individual panel, rather than following a single procedure for all thirty-two.

Lead came holds the glass along its edges. To remove a fragment from the centre, the conservators sometimes had to dismantle the panel from an edge or corner; another method involved cutting away part of the came on the reverse and then soldering it back into place. The article’s authors show why the phrase ‘replace one piece of glass’ conceals a risk to neighbouring historic elements. Mikhailov’s task was to preserve as much of the existing material as possible.

Where their condition allowed, cracked pieces were bonded with UV-curing adhesive rather than replaced entirely. The study notes that the workshop had already experimented with this material, although its use in stained-glass conservation was not common practice at the time. The choice between bonding and replacement depended on the condition of each fragment. Bonding does not in itself alter the artistic design, but becomes part of the panel’s material history and matters for its future monitoring.

To replace missing fragments, the craftspeople selected glass by colour, texture and place of manufacture. The article mentions glass from the Bryansk factory and glass made in the United States; these were restoration materials, rather than evidence of supplies in 1951. Missing painted elements were recreated: the researchers report the use of paints followed by firing at 600 degrees. The present ensemble thus combines older leaded glass, restored fragments and inserts from different periods.

While examining the original painted sections, Mikhailov identified several techniques: multilayer painting with firing, freehand drawing and application using a mask. The article distinguishes three different approaches to assembly; its authors note that panels assembled less securely suffered greater damage. These observations confirm the involvement of several craftspeople, although identifying the maker of each piece of glass would require further attribution research. Traces of the techniques used proved more informative than the later general description ‘Riga craftspeople’.

The curved installation within the pylons and vibration posed a separate problem. According to the conservators’ description, the lead came stretched, filler fell out, frameworks pulled apart and some pieces of glass were on the verge of falling out. Damage was more severe in the three lower rows closest to passengers. The conservators found steel strips on some panels that had reinforced the structure but had come loose over time. The craftspeople straightened and resoldered the frameworks so that the panels could be fitted into protective cases. These distortions explain why preserving the ensemble involved far more than removing surface dust.

The protective cases also have their own history. Mikhailov proposed enclosing the panels between sheets of toughened glass; according to the researchers, this was replaced with acrylic roughly a year later. The authors relay the conservator’s suggestion as to why the replacement occurred, but do not present it as a documented decision by the client. To diffuse the backlighting, he used Canson tracing paper on the rear glass of each case; this was later removed. These layers affect how viewers perceive the older coloured glass.

The 2017 expert assessment report briefly mentions the replacement of glass in the panels during the 2003 works. The scholarly article presents a more precise picture: some historic elements were preserved, others repaired, and missing pieces replaced. The report’s brief wording cannot be read as evidence that all 32 panels were replaced in their entirety. The same report states that, following restoration, the station was more brightly lit than originally intended. Comparing the two sources helps distinguish the preserved object from the changing conditions in which it is displayed.

How should we interpret Novoslobodskaya’s stained glass after examining the legend?

Consider the ensemble as a collaboration between architect, artist and workshop. Dushkin and Strelkov determined the position of the stained-glass panels within the pylons; Korin created the designs; and Riga’s Māksla craftspeople assembled the compositions for the station. A consistent arrangement unites the 32 works, yet their patterns and medallions differ. Backlighting makes the glass an integral part of the underground hall’s interior.

The story of ready-made cathedral windows is compelling, but the available documents confirm that the ensemble was commissioned and made for the metro. The origin of the sheets of coloured glass used remains an open question. This distinction does nothing to diminish the appeal of the stained glass: it brings the focus back to Korin’s vision, the skill of Riga’s craftspeople and the material history of works that restorers had to save.

The route the Underground Palaces tour includes Novoslobodskaya. In the hall, you can examine the backlighting, decorative patterns and figurative medallions of the stained-glass panels, then compare Korin’s work with the decoration of other stations along the route.

Sources

1. E. Minukhin. ‘The Art of Stained Glass’, ‘Architecture of the USSR’, No. 6, 1954
2. State Historical and Cultural Expert Assessment Report on Novoslobodskaya Station, 2017
3. All-Russian Museum of Decorative Art. The Maksla Art Workshops
4. O. V. Korolko, A. S. Sokolova. ‘The History of the Restoration of Pavel Korin’s Stained-Glass Windows’, 2025
5. City guide ‘Moscow’s 100 Best Places’
6. Riga Cathedral. Stained-glass windows
7. T. Knyazhitskaya, D. Vorobyov. ‘The Stained-Glass Windows of the Latvian SSR Pavilion at VDNKh’, 2023
8. ‘Three Years in the Sloboda’. Recollections of the Construction of Novoslobodskaya Station
9. Excursion "Palace Dungeons"
10. State Tretyakov Gallery. Sketch for the Mosaic at Novoslobodskaya Metro Station
11. ‘The Architecture of the Moscow Metro’. Architect A. F. Strelkov’s recollections of the decorative design of Novoslobodskaya station
12. T. V. Knyazhitskaya, D. E. Vorobyov. ‘The Stained-Glass Windows of the Lithuanian SSR Pavilion at VDNKh’, 2025