Sengan-en: Experiments at the Edge of a Garden

In 1857, an experiment at Sengan-en connected a garden ornament to a room near the bathhouse. Gas produced there was carried through pipes to a lantern shaped like a crane opening its wings. Matsuki Kōan, working on the project under Shimazu Nariakira, was among those investigating how coal gas could provide illumination. The experiment succeeded: the garden had become a place where a new arrangement of fuel, pipes and light could be tried.^[1]

A furnace foundation and a stone factory building nearby point to much larger undertakings. Yet the small lantern provides an especially useful introduction. Industrial experimentation at Iso, the coastal district containing Sengan-en, did not begin on an empty site cut off from the household of its patron. It developed beside his residence, using land, supplies and expertise already connected with the domain. Those connections made new work possible. They also required existing resources to do unfamiliar jobs.

Sengan-en began in 1658 as a retreat for Shimazu Mitsuhisa, the lord of Satsuma, the Shimazu-governed domain in southern Kyushu. Across the bay stood Sakurajima, the volcanic mountain that remains central to the garden's composition. Water and mountain supplied a scale no enclosure within the grounds could reproduce. The house faced a landscape extending far beyond its own land.^[2]

The garden nevertheless had to be made locally. Planting, water supply and changes to the shoreline shaped the setting from which the distant view was appreciated. The prefecture's cultural-property account records successive additions rather than one design completed in 1658: a winding-water poetry garden, moso bamboo brought from China through Ryukyu in 1736, and reclamation of the shore under Narioki in the nineteenth century. Sengan-en's apparently settled landscape had a history of acquiring plants and enlarging usable ground.

These changes belonged to the household's enjoyment and representation of its position. A poetry garden organized a particular kind of gathering. Imported bamboo altered the planting. Reclamation changed the estate itself, not merely what could be seen from it. Long before a factory stood nearby, maintaining the retreat involved bringing resources to Iso and deciding how they should be used.

Nariakira became lord in 1851. Under him, bamboo woodland adjoining the garden was cleared for the industrial undertaking that became known as Shūseikan, a complex concerned with cannon, ships and other manufactures.^[3] The cleared land should not automatically be identified with the famous imported bamboo grove. What matters is the documented change of use: ground beside a lordly retreat became a place for industrial construction.

The household and the industrial complex retained different purposes. They did not need to become one institution to share a setting. The practical connection is clearest in something that had already been brought to the garden: water.

The waterway from Sekiyoshi carried water from the upper Inari River toward Iso. Shōko Shūseikan's historical account dates its beginnings to around 1722, when it served Sengan-en, and identifies a later adaptation under Nariakira. The industrial works used waterpower, including water-driven bellows for a blast furnace.^[4]

A garden supply and a source of mechanical power asked different things of the same resource. In one use, water could be delivered to a designed landscape. In the other, its movement could drive equipment. The important connection was the existing ability to bring it across the intervening terrain. Redirecting that supply could serve a new purpose without first inventing the entire route by which water reached Iso.

The inlet at Sekiyoshi lies well beyond the estate. Its relevance is as part of the supply system, not as another feature tucked inside the garden. Seen through this connection, the industrial site extended into a hinterland of channels and sources even though its buildings were concentrated beside the residence.

Fuel created another outward connection. At Terayama, charcoal was made for the Shūseikan undertaking. The official history describes efforts to obtain stronger-burning charcoal by adopting techniques associated with Kishu, in present-day Wakayama, and applying them within the domain's production network.^[5] The factory's demands thus reached beyond the footprint of its furnaces to the making of fuel itself.

Water and charcoal were not interchangeable solutions. One supplied mechanical power; the other sustained heat. Their separate histories show what industrial construction required before a finished machine could be useful. Iso needed access to resources in forms suited to the work, not just possession of a suitable-looking building.

The reverberatory furnace was intended to melt iron for cannon. In this type of furnace, heat from the fire passes into the working chamber and is directed over the metal by the form of its interior. Descriptions and illustrations in a Dutch technical book supplied a starting point for construction. Small experimental work elsewhere in Kagoshima preceded the larger undertaking at Iso.^[6]

The first large furnace was begun in 1852 and failed. A second was completed in 1857 and produced cannon. That short sequence conceals several years of difficulty. Research published by Reimeikan, Kagoshima's prefectural history museum, identifies problems with the first furnace's structure and materials, including refractory bricks that deteriorated in the heat and contaminated the iron.^[7]

The material intended to contain the process could become part of what was being processed. A diagram could communicate the arrangement of a furnace, but it could not make an unsuitable brick survive inside one. The builders had to find a combination of design, construction and locally obtainable materials that would operate together.

This is a more precise problem than a general lack of knowledge about Western machinery. The technical information had arrived. The question was how to give it a working material form. A copied outline could still conceal differences in the properties of the stone, clay and fired components used to realize it.

The foundation now preserved at Sengan-en marks the larger enterprise, but the missing lining is crucial to explaining its history. Success depended partly on an object far less conspicuous than the furnace: a brick capable of enduring conditions for which ordinary building material was unsuitable. Making that object drew the project into the domain's ceramic world.

Reimeikan's study follows an unusually concrete supply decision. A 1854 instruction assigned the procurement of Amakusa pottery stone to an official whom the study identifies with porcelain production at Naeshirogawa. The author connects this order with the established supply of that stone for porcelain. Material from outside Satsuma could reach the industrial project through a connection already serving ceramic work.^[7]

The study also links responsibility for the bricks with an official of the domain's kiln. Experiments involved different materials, although their precise proportions remain uncertain. The order and the proposed identification together recover a practical response to the furnace's difficulty: approach people who already dealt with the acquisition and firing of ceramic substances.

The useful inheritance was not a ready-made furnace. It was a means of obtaining and working materials from which a suitable component might be developed. The procurement argument is strongest at that organizational level; it does not establish that the finished bricks simply reproduced a porcelain recipe.

The Iso kiln served a wider range of experiments. Kagoshima's industrial-heritage record identifies both refractory-brick production and improvements in Satsuma ware among its purposes. Work included richly gilded decoration, while European-style tableware was also reportedly attempted. The kiln occupied prepared ground on the hillside east of the industrial complex; parts of its supporting stonework survive, but the kiln itself does not.^[8]

Decorative ceramics and furnace bricks presented different tests of success. A vessel's finish and a lining's resistance to heat could not be judged by the same criteria. Their presence in one production setting nevertheless brought different demands into contact with the same broad field of material expertise.

This is where the garden and the works belong together most convincingly. An elite household's world of objects, suppliers and crafts did not simply disappear when military manufacturing arrived. Some of its connections became useful to that manufacturing, while ceramic work continued to include objects intended to be handled, displayed or used at a table. Adaptation was selective: an existing organization could help solve a new problem without every old technique automatically fitting the new task.

Nariakira died in 1858, and the program was reduced. British naval bombardment in 1863 destroyed much of the industrial complex. The reverberatory furnace and Iso kiln consequently belong to a phase that ended through more than technical obsolescence: patronage had changed, and war damaged the place where production occurred.^[3][6][8]

The next major factory was built under Nariakira's successor, Tadayoshi, and Tadayoshi's father, Hisamitsu. Completed in 1865, the stone machine shop contained imported equipment for work that included repairing ships and engines. Kagoshima's industrial record identifies Takeshita Seiemon with its construction and points to Nagasaki buildings as references for the undertaking.^[9]

The surviving building brings another kind of adaptation into view. Its masonry used locally available volcanic stone. Heavy roof beams and the curved stone treatment at the base drew on building practices familiar in Japan. Imported machinery thus occupied a structure made through a different combination of materials and construction habits than a factory copied complete from abroad.

A repair shop also addressed a different requirement from an experimental furnace. Once a vessel or engine existed, keeping it serviceable required a place where its components could be worked on. The purpose was no longer confined to demonstrating that a design could be built. It included maintaining equipment through use, damage and replacement.

The changes did not form a clean succession in which local experiment ended and foreign equipment solved everything. The building and its contents already resist that account: machinery could be imported while the structure enclosing it was locally designed and assembled. Different parts of an industrial system could be acquired in different ways.

Nor was the adjacent garden simply cleared away to make room for a uniform factory district. The modern site plan still makes the relationship legible: the house and principal garden occupy one part of the coastal estate, while the furnace remains, kiln site and machine shop mark neighboring areas.^[10] The plan is a present-day guide, not a survey of every nineteenth-century building. It nevertheless helps locate the separate settings through which the history must be followed.

The cotton-spinning mill introduced another form of technical contact. Its British engineers required accommodation in Kagoshima. The residence built for them in 1867 survives near Sengan-en as the Ijinkan, a two-storey timber house with verandas. The city's architectural account notes the combination of a Western-style arrangement with Japanese units of measurement.^[11]

The prefectural cultural-property record associates negotiations for the mill's British machinery and engineers with Satsuma representatives sent to Britain, including Godai Tomoatsu.^[15] The acquisition involved arrangements for specialists as well as equipment. Their work at Iso required somewhere for them to live, bringing the overseas negotiation back to a practical problem on the coast.

This was a different relationship from studying illustrations in a technical book. The foreign connection now included people present beside the works. The engineers' house supplied accommodation for that engagement, while its verandas and domestic rooms gave it a form quite different from the stone machine shop. Residence and production required their own buildings even when both arose from the same undertaking.

It was also distinct from the Shimazu residence within Sengan-en. The latter remained a setting for the family and its guests; the engineers' house existed because of a particular industrial engagement. Sharing the Iso district did not make their occupants members of one household.

The architecture in the principal residence belongs to a later domestic chapter as well as to the garden's longer history. The estate's own interpretation presents the house through Tadayoshi's life there in the late nineteenth century and through its reception of important visitors. Its rooms combine Japanese, Chinese and Western elements.^[10] The house should therefore not be imagined as a sealed interior left unchanged from Mitsuhisa's retreat or Nariakira's experiments.

The engineers' residence had its own subsequent travels. In 1884, it was moved to the former Tsurumaru Castle grounds to form part of the Zōshikan school. It returned to Iso in 1936.^[12] A building erected to accommodate industrial specialists had become useful to a school, then returned as a surviving reminder of its earlier purpose.

That journey supplies an unusually literal connection between the coastal works and Kagoshima's former governing center. Buildings could move between institutions even when their original occupants and functions did not. The Ijinkan's presence beside Sengan-en records both the arrival of engineers and later decisions about where their accommodation belonged.

The stone machine shop likewise passed through several kinds of ownership and use. It entered army control in 1871 and naval control in 1874. During the rebellion of 1877, the industrial area was occupied by Saigō's forces and suffered renewed destruction, although the machine-shop building survived. The property returned to Shimazu ownership in 1889; industrial operations ended in 1915. After alterations, the building opened as Shōko Shūseikan Museum in 1923.^[13]

These changes separate the survival of a structure from the continuation of a program. The family could recover a building without recreating Nariakira's undertaking. A museum could preserve its history after production had ceased. The factory's stonework remained useful, but the work carried out inside it changed.

The gas experiment offers a smaller version of that distinction. The estate's account records its abandonment after Nariakira's death.^[14] Producing light on one occasion did not automatically create a service that would continue beyond its sponsor. The crane lantern remembers an achieved experiment and a line of development that was not sustained in that form.

Sengan-en consequently preserves several outcomes together: a garden still organized around a distant view, the base of a destroyed furnace, an industrial building reused for history, and a residence that left and returned. None is an adequate substitute for the others.

Their connections explain how the experiments became possible. Land beside the retreat could be reassigned; an established water supply could take on mechanical work; ceramic procurement could serve furnace construction; a new industrial engagement could require another household to be accommodated. Their differing later histories explain why those possibilities did not amount to one uninterrupted march of progress. At Iso, an estate supplied a setting in which new work could begin. Keeping any particular undertaking alive remained a separate task.