Tachikawa Walk: Tamagawa Aqueduct and Sunagawa History Guide

A walk along the Tamagawa Aqueduct and old Sunagawa village in north Tachikawa, and the reason Edo’s 1653 waterway now dives under a river.

日本語で読む（Japanese version）

Edo’s water supply dives under a river. On the northern edge of Tachikawa, it still does.

Among the City of Tachikawa’s notes on its cultural heritage is a sentence listing the highlights of the Tamagawa Aqueduct, from upstream to downstream: a stone bridge memorial, an intake for a branch channel, the site of an old boat landing. And slipped into that list was a strange little phrase: “the fusekoshi under the Zanbori River.”

A fusekoshi (伏越) is a device that takes one waterway underneath another. The aqueduct passes under the Zanbori River. Somewhere in Tachikawa, in other words, a ditch dug by the people of Edo crosses a natural river at right angles, and it is the ditch that ducks beneath the river.

Isn’t that usually the other way round? The river was there first; the latecomer, the ditch, should politely go underneath. That, I had always thought, was the natural order of things. So why, in Tachikawa, does the aqueduct pass under the river? I set out to walk the northern edge of Tachikawa (立川) to find the reason.

Where to start a Tachikawa walk: Tamagawa-Josui Station and the green path

Ask people what lies at the northern end of Tachikawa and few will have an answer. Even residents may sum up everything north of the busy station area as “houses and fields.” But if you want to take a walk in Tachikawa — a city in the Tama area, the western half of Tokyo beyond the 23 central wards — this is where I would send you first.

The best place to set off is Tamagawa-Josui Station (玉川上水駅), an interchange between the Tama Toshi Monorail and the Seibu Line. The Tamagawa Josui — the Tamagawa Aqueduct, literally “the Tama River water supply” — flows east to west along the south side of the station. The monorail’s station guide says the land alongside the aqueduct is a green path where you can enjoy a walk in nature all year round. The station’s name tells you exactly where to go. You will rarely meet a station that needs less of a map.

Leave the station and the hum of the elevated track fades behind you, while the sound of earth and fallen leaves underfoot comes closer. The aqueduct is deeper than you might expect. Its banks are steep, and the water runs far below head height. Lean over and you can see the reddish-brown skin of the Kanto loam, the volcanic-ash soil of this region, laid bare.

The City of Tachikawa describes this stretch as “the earthen walls of both banks, soaked in more than 300 years.” Rather poetic, for a government text. But if you consider that the real hero of this channel is not the water but the earth that was dug out to make it, it is also the most accurate way of putting it.

The aqueduct runs for about 5.6 kilometres through the city, crossing Ichiban-cho, Kamisuna-cho, Sunagawa-cho, Kashiwa-cho and Saiwai-cho. On a map, a single man-made channel hugs the whole northern edge of Tachikawa.

Who dug this channel, and why? The story goes back about 370 years, to an Edo — the old name for Tokyo — that was short of water.

1653: the 43 kilometres dug by the Tamagawa brothers

The story of Edo’s water begins before the Tamagawa Aqueduct. According to the Tokyo Metropolitan Government Bureau of Waterworks, Tokugawa Ieyasu, who entered Edo in 1590, built the foundations of the city by filling in the lowland east of the castle; but in that low ground near the sea and on reclaimed land, digging wells did not yield good water. So aqueducts were planned early on, and by the end of the Kan’ei era, early in the Edo period, the Kanda Aqueduct, which drew from Inokashira Pond, is thought to have been completed. In the southwest, a reservoir called Tameike, near today’s Tameike-Sanno subway station, was also used as a water source.

Even so, it was not enough. Under the third shogun, Iemitsu, the system of sankin-kotai — under which feudal lords had to spend alternate years in Edo — was established, and lords, their families and retainers came to live in the city, so the population grew ever faster. In 1652 the shogunate drew up a plan to bring water from the Tama River into Edo. After a study of the designs and surveys on the ground, the contractors chosen for the work were two brothers, Shoemon and Seiemon. The senior councillor Matsudaira Izu-no-kami Nobutsuna and the magistrate Ina Hanjuro Tadaharu, among others, are said to have been involved.

Work began on the fourth day of the fourth month of 1653, and on the fifteenth day of the eleventh month, some 43 kilometres from the Hamura intake to Yotsuya Okido, the gate on the western edge of Edo, were complete, the Bureau of Waterworks writes. And then it conscientiously adds a note in brackets: “This year was a leap year with two sixth months, so it comes to eight months.” (The old lunar calendar sometimes added an extra month.) It is an answer prepared in advance for the reader counting on their fingers and thinking, “April to November — isn’t that seven?” I never expected to find such a thoughtful line in an official text.

The difference in height between Hamura and Yotsuya Okido is about 92 metres. Another bureau document restates this as a drop of only about 21 centimetres per 100 metres. That is like descending the height of one step of a station staircase over a hundred metres. I would lose track of the slope within half a day, and not a drop would ever reach Edo. Choosing the ridgelines of the Musashino Plateau, an unlined, simply dug channel carried water all the way to Edo by natural flow alone. In the sixth month of the following year, 1654, pipes of stone and wood had been laid beneath Edo, and water began to be supplied to Edo Castle and to the Yotsuya, Kojimachi, Akasaka, Shiba and Kyobashi areas.

As a reward, the brothers were granted the surname Tamagawa, given a stipend of 200 koku of rice, and appointed hereditary water officials, the bureau writes. The City of Tachikawa’s notes add that they were allowed to bear a surname and carry swords, privileges then reserved for samurai. Eight months of civil engineering, paid in a family name, swords and 200 koku. I tried to convert it into a modern salary and gave up. I have no idea what a sword goes for.

There is, however, another version of this story. According to the bureau, a report submitted to the magistrates in 1803 by Kojima Bunpei, one of the Hachioji Sennin Doshin (a body of semi-samurai guards based in Hachioji), which gathered his ancestors’ accounts and local traditions, says the work failed twice, and that it succeeded only after the intake was fixed at Hamura to a design by Yasumatsu Kin’emon, a retainer of Matsudaira Nobutsuna who had built the Nobidome Canal. No documents from the time of construction have been found, so nobody knows which is true. A different hero’s name is wedged into the hero’s tale. History usually survives in this shape.

As well as supplying Edo, the aqueduct became a tool for opening up the plateau along its route. The first branch channel was the Nobidome Canal of 1655. The bureau writes that it is said to have been granted to Matsudaira Nobutsuna, lord of Kawagoe Castle and the overall commissioner of the works, in recognition of his services. The commissioner’s reward, too, was water. The City of Tachikawa’s notes record that around the time of the Kyoho reforms (the early eighteenth century), many branch channels were drawn onto the Musashino Plateau, where the water table was too low for people to live, and many new farming villages were created. According to the Agency for Cultural Affairs’ Cultural Heritage Online, branch-channel facilities numbered 33 by the Kansei era in the mid-Edo period. Within Tachikawa alone, four were drawn off: the Sunagawa, Tonogaya, Shibasaki and Gengoemon branches.

Yet the story of Tachikawa’s northern half begins a little before the aqueduct — and with a name that is remarkably short of water.

A village called “Sand River”

The place name Sunagawa (砂川) is said to come from the Zanbori River, which flows through the city, writes the city’s newsletter on the compilation of its history. Sunagawa means “sand river.” In the old days, the Zanbori sometimes flooded in heavy rain, but most of the time it was a “river of sand,” bare sand and stones.

The lack of water became the place name itself. It was not a name chosen in the hope of plenty. It named the dry riverbed in front of people’s eyes, exactly as they saw it. An admirably honest piece of naming.

Please remember this dry river’s name. It is the other lead character in the mystery from the opening.

The Itsukaichi Kaido (五日市街道), the road that became the backbone of the village, is also older than the aqueduct. The city’s chronology records that in 1596 Ieyasu put the Ome and Itsukaichi roads in order, and the history newsletter writes that the Itsukaichi Kaido was developed as one of the routes for carrying materials and fuel from Tama to Edo for the rebuilding of Edo Castle and the expansion of the city. First came a road to carry things to Edo; later came a channel to carry water to Edo. Sunagawa was the land where those two lines crossed.

Settlement, too, is older than the aqueduct. According to the city’s chronology, from the Keicho era (1596–1615) Murano San’emon, a farmer from Kishi village (today’s Kishi in Musashimurayama City), began to clear land for Sunagawa Shinden (Sunagawa New Fields), and in the Kan’ei era people began to live there. The newsletter says early Edo-period development followed the Zanbori River. The Musashino Plateau was covered in volcanic soil that was hard to farm and had few water sources. Even so, around today’s Sunagawa Sanban, one of Sunagawa’s oldest wells was dug in a style called a maimaizu well — a spiral path winding down into a funnel-shaped pit to reach the water — and people gradually settled. The village began around a single well.

In 1629, Azusami Tenjin-sha Shrine (阿豆佐味天神社) was founded as the village’s guardian shrine. In 1633, Sunagawa Shinden became a falconry ground of the Owari domain. In 1654 the Tamagawa Aqueduct opened, and three years later, in 1657, the Sunagawa branch channel was opened. Here, the newsletter says, the development of Sunagawa began in earnest.

In the land survey of 1689, Sunagawa had 175 households. The breakdown preserved in the city’s chronology is more than 229 cho of dry fields, more than 73 cho of grassland for thatch, and more than 39 cho of woodland — and no rice paddies at all. The city’s cultural-heritage notes also say Sunagawa’s land, being volcanic Kanto loam, was unsuited to paddy fields.

A village with nothing but dry fields, thatch grass and woods received a branch channel. What shape did it grow into? The answer survives on the map, as lines that look as if they were drawn with a ruler.

Strip-shaped plots and the straight Sunagawa branch channel

Here the city’s walking-course notes contain a sentence you should not miss. Sunagawa village was laid out deliberately along the Itsukaichi Kaido and the Sunagawa branch channel, with plots lined up in long strips. The branch channel runs dead straight beside the road.

By contrast, the Shibasaki branch channel, drawn off near Matsunaka Bridge in 1737, had to deliver water to homes already standing in Shibasaki village, and so it is said to have wound along the boundaries.

A straight channel and a winding one. Shibasaki’s water bent to fit the lives already there. Sunagawa’s water was drawn onto a plateau where few people yet lived, and life lined up to fit the water. The channel came first, the houses after. That is why the line is straight.

As more people moved in, land along the Itsukaichi Kaido was allotted in order, and the settlement stretched long from east to west. Locals called this long, thin village “Niri-hatcho, the long new fields,” the newsletter writes. Niri-hatcho — two ri and eight cho — is roughly 9 kilometres. Borrowing soy sauce from next door took no time at all, but going from one end of the village to the other was a proper day out.

A publication of the city’s history and folklore museum describes the shape of a Sunagawa farm like this: houses stand in a row facing the Itsukaichi Kaido, which runs east to west, with fields and a homestead woodland behind them. A house facing the road, fields behind it, woodland further back still. A whole life fitted into a long, thin strip. The trees along the road have become the face of this route, too. One travel blog by someone who walked the Itsukaichi Kaido says that walking the leafy avenue in summer felt like walking through a mole’s tunnel. In summer, Sunagawa lies inside a green tunnel.

If you want to see what remains of all this, walk the Taka-no-michi (鷹の道), the “Hawk Road.” It was a path for the officials of the Owari Tokugawa family’s falconry ground to patrol the villages, and ordinary villagers were normally not allowed to use it, the city writes. Along it, you can still see the layout of the old village — house, homestead woodland, field — in its divisions. The city’s notes even say that the name “Taka-no-michi” was sometimes used as a byword for the arrogance of the falconry officials. I know of no other case where the name of a road became an insult.

The village was divided into groups, from Ichiban-gumi (Group One) to Juban-gumi (Group Ten), and each group managed its own roads and channels. The town names changed when addresses were reorganised after the merger, but the signs at junctions and bus stops still say “Sunagawa x-ban.” The authorities may change the names, but people’s mouths are slow to obey.

There is also a place where you can see the standing of a group leader in a building. At the Kawagoe-michi Ryokuchi Old Folk House Garden (川越道緑地古民家園), the house of the Kobayashi family, who served as head of Sunagawa’s Ninth Group, has been moved and rebuilt. It is a thatched farmhouse from the end of the Edo period, and the main building is of the “six-ken” type, rare in Sunagawa. The alcove, staggered shelves and built-in writing desk in the inner room at the northwest corner are as formal as those of samurai houses of the time, the city writes. By the end of the shogunate, the group leader of a village carved out of a waterless plateau had a reception room to match a samurai’s.

The Tamagawa Aqueduct and Sunagawa in numbers

• 43 — kilometres. From Hamura to Yotsuya Okido. It was dug in about eight months, they say.
• 92 — metres. The drop over those 43 kilometres. Go forward 100 metres and you descend about one stair.
• 33 — branch-channel facilities drawn off the aqueduct by the Kansei era. Edo’s water was generously shared out on the way.
• 5.6 — kilometres. The stretch flowing through Tachikawa. It crosses five neighbourhoods, dutifully, from west to east.
• 175 — households in Sunagawa in 1689. Not a single rice paddy appears in the survey.
• 10 — metres. The height of Konpira-yama, which appears later. It is the highest peak in Tachikawa.

A wayfarer’s quiz

In the early Meiji era, boats once sailed on the Tamagawa Aqueduct. How long did that last? (A) 2 years, (B) 20 years, (C) 200 years. Look for the answer as you walk on.

Once the strip village had taken shape, the villagers were handed one more important job: looking after the aqueduct itself.

The villagers who guarded the aqueduct: duty villages, water watchmen and patrol officers

A 43-kilometre unlined channel was not finished once it was dug. According to the Bureau of Waterworks, for a while after supply began, the descendants of the Tamagawa brothers handled management and repairs, but before long the shogunate took direct charge. Watch-houses were placed along the aqueduct, and townspeople and farmers called mizuban-nin, water watchmen, checked the volume of water and cleaned the debris screens known as akutadome.

The villages along the route were also given their own stretches. They looked after the noticeboards displaying regulations, cut the grass on the banks, cleaned the channel and helped when officials passed through. Such villages were called mochiba-mura, “duty villages,” the bureau writes. Edo’s water supply was kept polished by the hands of the villages it passed through.

There is an amusing record, too. Cherry and pine trees were planted along the aqueduct, but when they grew into big trees they dropped more leaves and got in the way of maintenance, so large trees were cut down. Create a famous blossom-viewing spot, then fell the trees that grow too big. When you see the ongoing work to deal with fallen trees along today’s green path, you realise the same headache has been around for 350 years.

The system for managing the aqueduct and the state of its facilities were recorded in the ten-volume Josui-ki (Records of the Aqueducts), compiled in 1791 by Ishino Hiromichi, the commissioner of works. The figure of 33 branch channels comes from this book.

In Sunagawa, there was a family who kept watch over the aqueduct. The New Edition of the History of Tachikawa City: Sources, Early Modern 2, published by the city, contains 309 old documents relating to the Sunagawa area. According to the city, among them is a wealth of material on the “aqueduct patrol officer” and “falconry-ground guide” roles held by the Sunagawa family, which shows how the Tamagawa Aqueduct was managed and how people lived. The village did not only receive the benefits of the aqueduct; it also kept watch over it.

The falconry ground overlapped with Sunagawa life as well. According to the city’s chronology, Sunagawa Shinden became an Owari-domain falconry ground in 1633, was abolished in 1693, and became a falconry ground again in 1717. Falcons and officials connected to the shogun’s family came and went through the village, and through it ran a Hawk Road the villagers could not use. Draw water from the branch channel, guard the aqueduct, guide the falconry officials: a Sunagawa farmer’s year was bound remarkably tightly to Edo’s needs.

The branch channel’s water had one more job besides drinking and farming. The bureau writes that branch channels also powered waterwheels, and the Tachikawa City History and Folklore Museum’s booklet Exploring Tachikawa’s Waterwheels introduces the many waterwheels built on the aqueduct’s branches from the Edo period onward. A straight channel makes it easy to set waterwheels in a row. Behind the strip village, waterwheels were turning.

The history of an aqueduct protected like this took a big turn in the Meiji era (1868–1912). The door to our mystery opens in Meiji.