Five, seven or nine ponds: how Lefortovo Park is laid out

Sources give different counts for the ponds in Lefortovo Park. An account of the work carried out in 1722–1724 mentions five new ponds, a published reconstruction of the water system and a 1733 inventory of the Annenhof Garden give nine, while the current official page lists seven. These records refer to different periods and areas. It is therefore impossible to locate ‘Bidloo’s two vanished ponds’ on a modern plan by simple subtraction: the boundaries of the ponds, their names and what was being counted all changed.

The layout of the water garden explains why these figures cannot simply be compared. Water flowed from the upper Golovin Pond towards the structures below and the Yauza River; dams, canals and islands linked parts of the ensemble at different elevations. Fragments of this layout and its hydraulic structures survive in the park, but its present appearance does not entirely reflect Bidloo’s design of the 1720s.

How did the water system in the Golovin Garden come about?

Work on laying out the Golovin Garden began in 1703, two decades before Bidloo’s alterations. The main pond existed before the redesign. Sources differ on when Peter I acquired the estate: the Nekrasov Library gives 1721, while the park’s official page and DutchCulture give 1722. In a letter dated 12 May 1722, quoted in historical and cultural expert assessment, Peter already referred to the Golovin family’s house as purchased and gave instructions for the garden. The purchase had therefore taken place by 12 May 1722; these accounts do not establish the day and month of the transaction.

The historical and cultural assessment report dates the main work on redesigning the garden to 1722–1724. The expert quotes a letter from Peter I with instructions for laying out the garden — its water network, grotto, planting and other elements — citing an archival file in the Russian State Archive of Ancient Documents. By 1723, Krestovy Pond and an island surrounded by a canal had been created; further down, near the Yauza, a system of ponds and parterres was taking shape at an angle to the estate’s earlier axis. The number of ponds created under Bidloo therefore does not account for the whole of the earlier Golovin Garden.

Why did Peter I’s physician design the Golovin Garden?

Nikolai Bidloo headed the Military Hospital beside the Golovin estate and had laid out his own garden at the hospital before receiving the tsar’s commission. DutchCulture reports that Peter I often visited it. The physician was working on familiar ground when he extended the garden towards the Yauza at the tsar’s request.

This biographical detail also explains the geography of the project. The Military Hospital, the Golovin estate and Lefort’s residence stood near the Yauza, but did not form a single, level site. Bidloo had to connect the upper terrace with the water beside the river, the view from the palace with the islands below, and the walking avenues with the dams. The designer’s Dutch origins alone do not prove that any particular garden was copied directly in Moscow. The landscape architects H+N+S, who studied the park for a cultural and historical atlas, emphasise the combination of the Dutch water-garden tradition with the local terrain.

DutchCulture states that two dams were built for the fountains, and that five ponds were dug under Bidloo and connected by canals. The same account mentions cascades, grottoes, pavilions and sculptures. The 1733 inventory quoted in the assessment report already lists a statue of Venus on a dam and other garden ornaments. It records a later state of the ensemble: this list alone cannot establish which feature was installed in 1723 and which was added under Anna Ioannovna.

What exactly does the figure ‘five’ mean?

Five is the number of new ponds in the account of the work carried out in 1722–1724, quoted in the assessment report. The caption to a plan from the cultural and historical atlas, published in a report on its presentation, dates their creation to 1723. Five is not the total number of ponds: the old Golovin Pond existed before this phase of work, while other parts of the garden were created and remodelled later.

The difference in elevation between the upper terrace and the lower basins allowed water to be channelled downhill. DutchCulture mentions two dams built for the fountains; Peter’s letter of 1722 calls for canals to be dug and a grotto to be built beside the mill. These documents confirm the programme of water features and engineering works, but do not allow every surviving spillway to be confidently attributed to Bidloo’s original design. Comparing such a system solely by the names of individual ponds is not enough.

How did the water garden connect the terraces and the banks of the Yauza?

The account of the garden’s redesign quoted in the assessment report distinguishes between the Golovin estate’s earlier axis and the lower system along the Yauza, laid out at an angle to it. By 1723, parterres, Krestovy Pond and an octagonal island surrounded by a canal had appeared on the upper level; other ponds, canals and parterres were being created below. The water garden linked both terraces to the river. This plan cannot establish that all the ponds were created at the same time or had their present outlines.

The 1733 inventory also provides a broader spatial context. After listing nine fish ponds in the Annenhof Garden, it separately names a pond in the Lefortovo Garden beside the Yauza and a pond in the garden across the river near Lefort’s house. The list mentions rafts for crossing the river and swing bridges over the canals. This is evidence of connections between the gardens on the two banks, but it cannot simply be turned into a map of Bidloo’s work: the inventory was compiled after the Annenhof redesign.

The terrain linked the upper Golovin Pond, the intermediate levels and the low ground beside the Yauza. The landscape architects H+N+S see the terraces and water structures as the foundation of the garden, adapted to the site. Annenhof and the Catherine Palace were later superimposed on this structure; the history of each level can be traced through different documents and archaeological remains. The changes in elevation help explain the movement of water, but the visible lines of the avenues are not an unchanged plan from 1723.

Why does the historical account mention nine ponds?

There are two accounts from different periods supporting this figure. The caption to a plan from the cultural and historical atlas, published in a report on its presentation, mentions nine ponds in the entire water system, five of them created by Bidloo in 1723. The second account is the 1733 inventory quoted in the assessment report: it records the same number of fish ponds in the Annenhof Garden. Without comparing the plans, we cannot assume that these two counts of nine refer to the same set of water features: the garden was remodelled between the periods they describe.

The 1733 inventory defines the scope of the count more clearly than the atlas’s brief caption. It assigns nine ponds to the Annenhof Garden, then separately lists a pond in the Lefortovo Garden beside the Yauza and another across the river. The text does not provide a list of modern names with an exact match for each feature on the park map. The difference between the historical and present-day counts therefore does not prove that two ponds have been lost: first, the same territorial boundaries and a consistent method of counting divided or connected basins must be established.

The published formula ‘nine in total, five by Bidloo’ has a specific source: the caption to a plan from the cultural and historical atlas of Lefortovo Park, reproduced in a report on the atlas’s presentation in 2014. Separately, the assessment report quotes Peter’s letter giving instructions for the work and the 1733 inventory. Archival documents and a scholarly reconstruction answer different questions. It is particularly important not to present a modern caption to a plan as a verbatim eighteenth-century record, or to collapse the Petrine and Annenhof phases into a single moment.

The archival letter describes the design and progress of the work, the later inventory records the contents of the already enlarged garden, and the atlas brings evidence from several periods together into one picture. Between these stages, questions remain about which basins were newly dug, which incorporated existing water features, and how the authors counted the canals around the islands. The records of five and nine are themselves confirmed, but they do not establish an exact correspondence of ‘one historical pond — one modern name’. That requires dated plans with comparable boundaries and excavation findings for each area.

Which seven ponds does the park’s official page list?

The official page for Lefortovo Park lists seven ponds: Sapozhok, Banny, Golovin, Vosmigranny, Krestovy, Bolshoy and Ovalny. This list refers to the park’s present-day grounds; the canal around Rectangular Island is mentioned separately alongside it. The historical count of five has a different basis: the eighteenth-century water system extended beyond open bodies of water to include dams and stretches of the Yauza.

Differences arise even between modern descriptions of the grounds. An account of the palace and park ensemble prepared by the Moscow State Integrated Museum-Reserve mentions Annenhof Pond, divided into two bodies of water by an earth dam, as well as Vosmigranny and Bolshoy Ponds, which are connected to one another; it also separately names Severny Pond, part of Krestovy Pond and the canals around Rectangular Island. These details show why simply counting names across different publications can produce different results even for the surviving system.

Why do the names of the surviving ponds fail to yield an exact count?

A name does not always correspond to a separate expanse of water on another plan. The park’s official page lists Sapozhok and Banny as separate ponds. The museum-reserve, describing the former Annenhof Pond, states that it was divided in two by an earth dam. These texts use different descriptive frameworks; without a dated plan showing the boundaries of the area, each name cannot confidently be matched to a unit in the historical count.

The same museum account describes Vosmigranny and Bolshoy Ponds as connected to one another; it also separately mentions Severny Pond, part of Krestovy Pond and the canals around Rectangular Island. The official list therefore differs from the features of the ensemble described in this source. This difference alone does not prove that either institution is mistaken: they describe different aspects of the system and possibly areas that do not entirely coincide.

The shoreline itself has no fixed historical outline. The museum-reserve notes that Annenhof Pond is now divided in two by an earth dam, while Vosmigranny and Bolshoy Ponds are connected. Archaeological evidence adds another layer: the present shoreline of Ovalny Pond took shape after infilling in the twentieth century. Counting by names and counting separate expanses of water therefore produce different answers, even when all the authors describe the same park in good faith.

There are strict limits to what can be concluded here. The historical figure of ‘nine’ is known from the atlas’s reconstruction and, separately, from the 1733 Annenhof inventory, while ‘seven’ comes from the park’s current list. Without overlaying dated plans and pond boundaries, it is impossible to prove that a particular modern pond is one of the five created under Bidloo. Establishing such a correspondence requires each feature to be checked individually against documents and excavation findings.

How did Bidloo’s water garden work?

The water garden made use of the difference in elevation between the Golovin estate and the Yauza. Peter I’s letter, quoted in the assessment report, instructed that canals be dug and a grotto built beside the mill; DutchCulture mentions two dams built for the fountains. The upper Golovin Pond and the lower basins together formed an engineered water system. However, the available descriptions do not allow its historical hydraulic head to be calculated or the original layout of each spillway to be established without measured surveys.

The decorative and engineering functions were intertwined. Dams directed the water, while crossings connected different parts of the garden. The 1733 inventory lists five swing bridges across the canals and a raft crossing over the Yauza. This confirms that water was used for getting around the estate at the time. Such a historical description does not establish the practical conditions for visiting the park today.

Finds made in 2008 shed light on the practical use of Annenhof’s water network. The report by the Capital Archaeological Bureau on the 2008 excavations lists fragments of oars and rowlocks from archaeological layers beside the Annenhof Canal. The authors associate them with boating in the first half of the 18th century, but the small number of finds leads them to conclude that such outings were infrequent. This is physical evidence of the use of a particular canal; it does not prove that boats travelled across all of Bidloo’s ponds or that boating is possible on them today.

What does the Venus Dam reveal about the garden’s engineering?

The museum-reserve identifies the Venus Bridge-Dam as one of the park’s surviving hydraulic structures. The 1733 inventory, quoted in the expert assessment report, mentions a stone statue standing on it and decorative features below. Geophysical surveys in 1999–2003 found that the earth forming the dam’s body was well preserved, along with an anomaly at its right abutment associated with a spillway. Together, the documents and excavations confirm the existence of a historic junction in the water system, although they do not automatically establish its entire original appearance.

The inventory records the dam’s decoration in 1733, while the archaeological evidence describes the physical condition of its foundations at the time of the investigations. Alterations took place between those dates, so the present masonry cannot all be attributed to Bidloo’s work of 1723. Assessing the structure requires consideration of its position between the ponds, the course of the water and the surviving body of the dam; a single statue mentioned in an old inventory does not explain how the dam worked.

Two kinds of movement can be distinguished here: water flowing down between the terraces and people moving between parts of the garden. The very term ‘bridge-dam’, used in the museum’s description, recalls this dual role. The 1733 record shows that crossings and dams were part of the layout for strolling, as well as serving an engineering purpose. Without dating individual features, the structure’s present appearance cannot reveal the exact route or water level of the 18th century.

What survives in the park from the 18th century?

The garden’s topography and underlying layout can still be traced in the park today: its upper and lower sections, the cascade of ponds and the connections provided by crossings and canals. The museum-reserve lists the Venus Bridge-Dam, the remains of a terrace beside Annenhof Pond and the grotto among the surviving structures that served the pond system. These are tangible landmarks, but their present appearance reflects alterations, losses and restorations carried out over the years.

In a letter dated 12 May 1722, Peter I proposed building a grotto beside the mill; the expert assessment report records the reconstruction of the retaining wall beside Krestovy Pond in 1731 under Rastrelli. The later decorative scheme can be recognised in the niches and columns of the structure’s ornamental section. The park’s official page dates the Rastrelli Grotto to 1731. Peter’s original proposal and the surviving architectural form therefore cannot be treated without qualification as the result of a single building campaign.

The neighbouring palace complex also changed the character of the space. The museum-reserve dates the construction of Annenhof Palace, with its gardens, canals, fountains and greenhouses, to 1731–1735 and attributes the design to Francesco Rastrelli. Excavations of the ‘cascade’ uncovered white-stone elements from stairways, embankments and decorative features; finds made in 1973–2004 were displayed at an exhibition in 2025–2026. Some of the architecture can only be understood through documents and excavated fragments: visitors can no longer see it in its entirety.

Which outlines of the old ponds have archaeologists established?

The excavations showed that the present shoreline is an unreliable yardstick for measuring the 18th-century garden. The historical and cultural expert assessment brings together the results of investigations in 1999–2003: archaeologists recorded bank reinforcements, drainage, building foundations, bridge supports and a landing stage. This physical evidence complements the historical plans; each find still needs to be dated before it can be attributed to Bidloo’s work.

Investigations in 1999–2003 revealed reinforcements along the banks of ponds and islands, a bridge support on the Annenhof Canal, and the foundations of a greenhouse, the Annenhof ‘cascade’, the pavilion on Octagonal Island and the fountain beside the Venus Dam. A wooden boat landing stage was discovered beside Bolshoy Pond. These finds show that crossings and pavilions existed between the ponds, features that a simple count of open bodies of water cannot reveal. Yet the location of a single building or landing stage does not, by itself, establish the shape of the entire shoreline in any particular year.

On Octagonal Island, archaeologists recorded the foundations of a pavilion; the 1733 inventory also describes a pavilion on the ‘octagonal island’ where small birds were kept. This is a rare instance in which a written inventory and physical remains can be linked to the same place, although individual structures still require careful dating. The island’s historical name alone does not prove that its octagonal outline remained unchanged in every subsequent decade.

At Krestovy Pond, researchers discovered bank reinforcements made from piles; the report gives the pond’s 18th-century length as 82–83 metres. At Golovinsky Pond, they traced wooden embankments and established its former dimensions. Beside Ovalny Pond, archaeologists found that the present shoreline took shape after infilling in the 20th century; coins from the 19th and 20th centuries were also found there. This evidence explains why the present outline cannot simply be superimposed on the 1723 drawing without allowing for later infilling.

In 2008 the Capital Archaeological Bureau investigated not only the ponds, but also the Annenhof ‘cascade’, Rectangular Island and sections of the paths. The archaeologists describe structural foundations, the courses of canals and the sequence of layers of fill. This allows the visible ponds to be related to the spaces between them: building sites, passageways and former shorelines.

The 2008 excavations provide a clear example of how the outline of a water feature can disappear. On Rectangular Island, archaeologists excavated 715 square metres. By the 18th century, canals surrounded it on all four sides, but by the time of the investigations only the western canal remained visible at the surface. The other courses were identified through buried banks and excavation trenches. Parts of the historic layout lay one to two metres beneath the present surface. If only open bodies of water are counted, the infilled canals around Rectangular Island will be left off the list.

The same excavations revealed paths radiating from the central pavilion, two circular routes and a path around the perimeter. On the island’s eastern side, a surviving support for a bridge across the canal measured approximately 3 by 2.8 metres. These finds also show how water formed part of the layout for strolling: a crossing provided access to the island, while the paths guided movement around the pavilion. However, without comparing dates, the entire excavated layout cannot be attributed to Bidloo: this area continued to be used and altered after his time.

The stratigraphy explains this caution. According to the archaeologists’ report, layers of fill from the 19th and 20th centuries lay above the remains of the 18th-century layout on Rectangular Island; several buried soil horizons show that earth was added in stages. The authors link this raising of the ground to rising pond levels, but present it as a probable explanation. Changes in surface height mean that a location on an old plan coinciding with a present-day path does not in itself confirm that the area’s form and function have remained unchanged.

How did Annenhof and Catherine Palace change Peter I’s design?

After Bidloo’s time, the park continued to be remodelled as an imperial residence. In the description provided by the Kolomenskoye — Izmaylovo Museum-Reserve Annenhof Palace, with its gardens, canals, fountains and greenhouses, is dated to 1731–1735; the project is associated with Francesco Rastrelli and Empress Anna Ioannovna. Thus, even the surviving formal geometry does not necessarily date entirely to 1723. New compositional axes, decorative features and engineering alterations were layered onto it.

The Annenhof ‘cascade’ was a stone retaining wall with grand stairways, spillways and a balustrade between the palace complex and the upper park. The museum explains that the structure is now known primarily through archaeological finds: white-stone elements from stairways, canal embankments and decorative features. The excavated elements reveal the scale of architecture that can no longer be seen in its former entirety during a walk. These fragments help explain how Rastrelli used the changes in elevation to give the terraces a grand architectural setting.

Catherine Palace belongs to the next major phase. An official article on Moscow’s transport portal dates its construction to 1773–1796 and associates the main façade with Giacomo Quarenghi; DutchCulture gives 1797 as the completion date. The sources differ on the year of completion, so it is more reliable to specify a period: from 1773 to the end of the 18th century. The later palace changed the composition of the site; its present axis cannot be transferred to Bidloo’s plan of the 1720s without verification.

The cultural and historical atlas presents the history of Lefortovo Park in chronological layers: H+N+S researchers compared the design intentions, their implementation and the surviving elements of each period. This approach is especially useful for Lefortovo: the ponds of Peter I’s garden, the Baroque remodelling under Anna Ioannovna and the later stone palace coexist in the same space. Removing one layer to create a ‘pure’ picture of the 1720s would conceal the true history of the ensemble.

What does the 1733 inventory say about the boundaries within which the nine ponds were counted?

The 1733 inventory lists nine ponds containing fish within the Annenhof Garden. The historical and cultural expert assessment report quotes this passage from historian Ivan Zabelin’s 1873 publication. The inventory then separately names the Lefortovo Garden beside the Yauza, the garden across the river near Lefort’s house, and a pond beyond the Lefortovo settlement. Nine is therefore the number within the territory named in the document, rather than a total for all the neighbouring gardens. This distinction is especially important when reading plans: ponds connected by direct sightlines or walking routes could appear on opposite sides of a garden boundary in the written inventory.

The document also provides details that help identify parts of the water system. Four white-stone statues stood beside the large pond, meaning Golovinsky Pond; a stone Venus stood on the dam. Krestovy Pond had a separate stone dam, sculptures and a stairway; beside Ovalny Pond, the inventory mentions an enclosure for birds. It lists islands with pavilions on the large pond near the Yauza, and another pavilion and two stone statues on the octagonal island. The document does not give the present-day names of all nine ponds; it describes the garden’s property and therefore cannot serve as a ready-made key to a modern map.

The record of five swing bridges across the canals clarifies the water system’s function. Artificial waterways divided the grounds, but crossings maintained connections for strolling; the four rafts mentioned between the Lefortovo Garden and the house beyond the Yauza indicate a means of crossing the river. This is neither a guide to walking in the park today nor proof that every canal was built in the same year. The inventory was compiled after the works under Peter I and during the new construction at Annenhof, when the garden was already changing. Finds of bridge supports, a wooden landing stage and bank reinforcements confirm that such connections existed, although individual structures require their own dating.

The published caption to the atlas diagram lists nine ponds for the entire water system, attributing five to Bidloo; the 1733 inventory counts nine specifically within the Annenhof Garden. Matching numbers do not mean matching lists: after 1723, palace canals were added, banks were remodelled and the estate’s boundaries changed. Neither record establishes a simple formula whereby two of nine specific ponds disappeared, leaving precisely the seven named on the present-day park’s official page. The evidence confirms only that the number depends on the date and the area being counted.

Why does the word ‘surviving’ need qualification?

The museum-reserve identifies the Venus Bridge-Dam, the remains of the terrace beside Annenhof Pond and the Rastrelli Grotto as surviving elements of the hydraulic ensemble. The 1733 inventory lists sculptures and other decorative features on the dams. It confirms their presence in that year, but says nothing about their present state of preservation. The Annenhof ‘cascade’ is also known through archaeological fragments: the museum reports finds of white-stone elements from stairways and cladding. The word ‘surviving’ therefore needs clarification: does it refer to the topography, the material, individual fragments or an entire structure?

The rotunda pavilion illustrates the distinction between a place of historical remembrance and an authentic structure from Peter I’s time. The park’s official page dates it to 1904 and associates it with the commemoration of Peter I; the museum-reserve describes the rotunda as restored. Neither description provides grounds for considering the present pavilion an authentic structure built by Bidloo in 1723. Later commemorative buildings are also important to the garden’s history, but when reconstructing the design of the water system, they must be distinguished from the 18th-century remains.

The landscape architects H+N+S also highlight a change in the urban context: in the 20th century, a major road along the river severed the park’s connection with the Yauza. The researchers were already describing this as a problem in their 2013–2014 study; their assessment cannot be applied to every detail of the park’s condition in October 2026 without further verification. Yet this historical separation matters when interpreting the old plan: Bidloo’s water garden was designed as a system oriented towards the river and the opposite bank, whereas the modern park is experienced largely within its land boundaries.

How can we interpret today’s layout without getting the numbers wrong?

Start with the differences in elevation: the upper Golovinsky Pond, the Venus Dam and the basins below it explain the flow of water better than any single count. Then compare the distinctive outlines of the Octagonal and Cross Ponds with the straight lines of the avenues and canals. This helps reveal the formal garden design within the altered park, without assuming that every present-day shoreline follows its 1723 outline.

When reading a map, ask which period the record refers to, which area is being counted, and whether connected or divided bodies of water are included in the total. For Lefortovo Park, the answers vary: five new ponds are noted in the account of works under Peter the Great, nine appear in the atlas plan and the 1733 inventory of Annenhof, and seven are listed on the modern park’s official page. Matching or differing numbers alone do not establish what happened to each historic pond.

The principal surviving trace of Bidloo’s water garden is the relationship between the terrain, the bodies of water and the engineering structures. It reveals how Golovin’s former estate became part of a grand palace and park ensemble, and how it subsequently changed under new owners and architects. None of the three counts can tell this story on its own.

The park’s relationship with the German Quarter and neighbouring palaces can be explored on the ‘Window to Europe’ tour of Lefortovo: the bureau’s route runs from Baumanskaya to Gospitalny Val via Lefortovo Park and Catherine Palace.

Sources

1. ‘Historic Gardens and Parks of Russia’ Association. Moscow State Integrated Museum-Reserve ‘Kolomenskoye — Izmailovo’: the Lefortovo palace and park ensemble
2. Moscow State Integrated Museum-Reserve ‘Kolomenskoye — Izmailovo’. ‘The Cultural Layer. Architectural Archaeology. Lefortovo Park’
3. DutchCulture. Lefortovo Park in Moscow: Bidloo, the hospital and the water garden
4. H+N+S. Cultural and historical Park Lefortovo, Moscow
5. Rijksdienst voor het Cultureel Erfgoed. Cultuurhistorische Atlas Lefortovo, 2013
6. Moscow Transport. ‘Lefortovo — one of Moscow’s oldest districts’
7. Archi.ru. ‘Lefortovo Park: Past, Present, Future’ — presentation of a cultural and historical atlas
8. Metropolitan Archaeological Bureau. ‘Archaeological Excavations: Lefortovo Park’, 2008
9. Moscow Government. Historical and Cultural Expert Assessment Report for Proyezd No. 1329: Historical Overview of Lefortovo and Archaeology
10. Official Lefortovo Park page: its history and seven ponds
11. N. A. Nekrasov Library. Holland in Moscow
12. Mos Radius. ‘Window to Europe’ tour of Lefortovo and the German Quarter