Skip to content
Topics

Search History Omakase

Start typing to search.

    ← Back

    HO Guides

    Toba–Fushimi: When the Road to Kyoto Became a Front Line

    Crossings, occupied buildings, and competing orders in the fighting that opened the Boshin War.

    Opening

    The troops advancing toward Kyoto carried a denunciation of Satsuma. After fighting began on 27 January 1868, a senior Tokugawa official still described the advance to British diplomats as the former shogun's retinue, intended to prepare for his return to the capital. In that explanation, the men blocking the roads were obstructing a journey that should have been permitted.[1]

    The new government in Kyoto would describe the same movement as an attack on the imperial court.[7] Between those accounts lay actual crossings, streets, and occupied compounds. Soldiers had to pass opponents who refused them passage, while commanders and territorial lords had to decide whose instructions they would follow.

    Toba–Fushimi is consequently best understood as a connected battlefield rather than one open field. One front formed along the Toba approach near Jōnangū and the old Koedabashi crossing. Another ran through Fushimi, where a shrine and a magistrate's establishment became opposing positions.[2] Following both makes it possible to see how a military advance changed into a struggle over who could act in the emperor's name—and why a declaration alone did not settle that struggle.

    1. A resignation had left an army in place

    Tokugawa Yoshinobu had offered to return governing authority to the court in 1867. The restoration proclamation of 3 January 1868 then abolished the shogunal office within a new governmental arrangement. These were separate steps, and neither instantly disposed of the Tokugawa house's territory, soldiers, officials, or role in foreign affairs. Yoshinobu withdrew from Kyoto to Osaka, where the next stage of the confrontation developed.[3]

    The distance between a proclaimed change and an operating government appears clearly in Ernest Satow's record of the weeks before the battle. Satow, a British diplomat then in Osaka, described conversations in which different parties disputed how the new authorities should notify the foreign representatives and how Yoshinobu's continuing administrative role should end. Even people anticipating an imperial government had practical questions about how it would take over.[4]

    For soldiers approaching Kyoto, those unresolved relationships became a geographical problem. The court was in the city; Yoshinobu and substantial forces associated with him were to the south. A return required movement through places where opposing forces were already stationed. The same road could be described as an approach for a political return or defended as an avenue of attack.

    Yoshinobu himself remained in Osaka during the fighting. His absence matters because a common shorthand—“the shogun marched on Kyoto”—can place him at a crossing where he was not. His forces were also not one homogeneous unit. Tokugawa troops operated with contingents from Aizu and Kuwana, while their opponents included forces from Satsuma and Chōshū, domains whose leaders were prominent in the new government.[1][2]

    A domain was a territorial lordship with its own retainers and military organization. The political alignment of its leadership could therefore affect the availability of an entire contingent. There was no single national chain of command accepted by everyone assembled south of Kyoto. The approaching fighting would alter those relationships, but it began before they had been resolved.

    2. Two approaches, two kinds of obstruction

    On the Toba side, the confrontation centered on a road and crossing. Jōnangū's historical account places Satsuma troops in the shrine precinct and guns directed toward the old Koedabashi approach. The northbound force was stopped near Akaike. On 27 January—the third day of the first month in the Japanese calendar—the standoff became fighting.[2]

    The crossing gave the encounter a form that a broad map of territorial allegiance cannot convey. An army intending to enter a city had to bring its leading units through a particular approach. Men farther back could not make that passage open simply by adding to the force's total strength. A roadblock and guns ahead turned movement into a local contest over access.

    The second front was farther east, within Fushimi's built-up district. There, Tokugawa-associated forces occupied the magistrate's establishment south of Gokōnomiya Shrine. Satsuma held the shrine; artillery also operated from the higher ground around Ryūunji to the east. The main east–west street, Ōtesuji, lay within this closely spaced military setting.[5][6]

    A shrine and a magistracy were not interchangeable patches of ground. The first brought walls, precinct space, and a position on the northern side of the confrontation. The second was an established governmental compound, now accommodating armed forces. Their occupation pulled the town's institutional geography into the battle. Fushimi did not have to be converted into an isolated fortress before warfare could make use of its buildings.

    The two fronts also posed different problems. On the Toba road, an approach was blocked. In Fushimi, opposing forces were already lodged within the same town. Success on one route could not be read directly from the situation on the other. The name “Toba–Fushimi” preserves that conjunction: separate encounters connected by the attempt to enter Kyoto and by the movements that followed the initial repulses.

    3. The magistracy became a target

    Fushimi's magistrate had once represented the shogunate through an office from which business could be conducted and orders issued. Occupied by troops, the compound acquired a different exposure. Its walls and buildings concentrated men in a recognizable position. Once artillery was directed toward it, a place associated with administering the town became one of the places through which the town was damaged.[5]

    The magistracy burned during the battle. Accounts of the Fushimi fighting include both firearm exchanges and close assaults. The familiar presence of the Shinsengumi—an armed corps celebrated for sword fighting—should not turn the entire encounter into a contest between swords on one side and guns on the other. The Tokugawa-associated forces also used artillery and firearms.[6]

    The distinction is practical. A modern weapon did not by itself provide secure ground, coordinated support, or an open retreat. Conversely, the presence of swordsmen in a force did not describe every unit's equipment. The battle involved different formations trying to use streets and compounds while their neighbors and routes were under pressure.

    News reaching Osaka carried some of this uncertainty. Satow recorded that officials initially reported repulses on both approaches and discussed trying another road. Later reports concerned withdrawal and wounded men coming downriver. He could see distant fires, but he was not standing at either front. His narrative combines observations in Osaka with information supplied by officials, travelers, and other intermediaries.[1]

    Read in that way, his account reveals something an orderly campaign diagram tends to conceal. The people directing or explaining the advance did not receive the finished story all at once. Information about blocked roads, burning premises, and retreating units arrived as the situation changed. The physical links between Osaka and Kyoto carried reports and wounded men in the opposite direction from the advance they had initially supported.

    Meanwhile, the occupied magistracy ceased to offer the establishment it had supplied at the beginning. Losing a governmental compound was not simply losing a symbol. It meant that a particular concentration of rooms, enclosure, and local administrative presence was no longer available in the same form. The fighting was dismantling part of the material setting from which Tokugawa authority had operated in Fushimi.

    4. The banner moved; the order followed

    As the fighting continued, the Kyoto government placed its campaign under the imperial prince of Ninnaji. Jōnangū's account describes him moving toward the fighting on 29 January, after his appointment as commander, accompanied by the imperial banner. His route led from Tōji along the Toba road and onward toward the Yodo–Fushimi area.[2]

    That movement made a claim of authorization visible in the contested landscape. Domain troops could now fight under a command presented as imperial rather than solely under the name of their own lord. Yet the banner's appearance was not the same event as the subsequent general order against Yoshinobu. Keeping those moments apart helps explain how the claim was extended beyond the men already present.

    The ensuing general proclamation declared Yoshinobu subject to punitive action. It characterized his earlier conduct as deceitful and his forces' advance as an attack on the court. Those accusations belong to the issuing government's case against him; the document is not an independent window into his private intentions.[7]

    Its instructions are more revealing than the denunciation alone. The order offered acceptance to people who now entered the government's service, including hereditary retainers who had followed the opposing side. It allowed that future petitions concerning the Tokugawa house might receive consideration according to service rendered. Conversely, people who communicated with or sheltered those designated enemies were threatened with punishment as enemies of the court themselves.[7]

    The proclamation was thus addressing men whose positions were not already settled. It combined a route into service with a warning about continued assistance to the other side. Hereditary attachment did not make a person permanently ineligible to change allegiance; providing shelter could make an apparently less active person liable to punishment.

    This was a concrete attempt to organize the consequences of the fighting. The banner identified a command in the field. The order told a wider range of people how the government proposed to distinguish service from opposition. Neither operation can be reduced to the sight of a piece of cloth causing every soldier to stop. Authority was being asserted through appointments, movement, written instructions, and the choices of those receiving them.

    5. A line of retreat depended on other people's decisions

    The fighting moved south. Jōnangū's account distinguishes the retreat through Tominomori on the Toba side from the fighting around the Senryōmatsu embankment on the Fushimi side. Both led toward Yodo and the connections back to Osaka.[2] Retreat brought troops into a landscape in which the position of other domains mattered as well as the position of the enemy immediately behind them.

    Tsu illustrates that problem. In Itō Yuriko's account for Mie Prefecture, there had initially been supporters of the Tokugawa side within Tsu. The domain followed an order from the court and took the new government's side, defending the strategically situated Yamazaki area. Her comparison with Kameyama identifies a different sequence: internal disagreement continued until the developing military situation favored joining the government.[8]

    These are distinct accounts of decision-making, not examples of one automatic reaction. An order from the court and information about the course of the fighting could both matter, but their weight and timing were not identical. Moreover, a domain's decision changed who could rely on its forces at a particular place. Political realignment had geographical consequences for an army trying to move through or retreat across the region.

    The men losing that support could describe it differently. Satow's memoir records conversations with wounded Aizu soldiers after the withdrawal. They blamed failures of support and denounced the Tōdō house of Tsu for changing sides at Yamazaki. They also commented on their opponents' skirmishing and weapons. These were participants explaining defeat, not an impartial assessment of every unit's performance.[9]

    The accounts describe opposite ends of a broken alliance. Itō follows the domain's response to the court; the wounded men describe the loss of support on which their defense had depended. The place held by that ally was what made the change militarily consequential.

    Kuwana complicates the picture further. Itō describes the domain dividing between those who continued with its lord and those who surrendered the castle to the new government. The lord's party continued fighting elsewhere.[8] “Kuwana” therefore cannot stand for one uninterrupted decision shared by every retainer. Even after the roads south of Kyoto had been lost, the political choices associated with the battle continued to divide people within the same territorial establishment.

    6. Calling someone an enemy did not make the description uncontested

    A Shōnai-domain compiler writing in 1869 questioned whether Yoshinobu, Aizu, and Kuwana had truly intended hostility toward the emperor. His compilation argued that accusations and responsibility should have been investigated, and treated the continuing war as a source of suffering. The passage survives through a later copy described and excerpted by the University of Tokyo's Historiographical Institute.[10]

    This was a viewpoint associated with the side that had resisted the new government. It does not disprove the proclamation's accusations simply by disagreeing with them. It does show that the category “enemy of the court” was not accepted as a complete explanation by everyone writing about the conflict soon afterward.

    An official notice reproduced in Kyoto Tachibana University's digital archive shows how the government's interpretation traveled beyond the armies. Issued under Chōshū's name later in the same Japanese calendar month, it named Fushimi in its opening as the place where Tokugawa forces had committed rebellion. The text moved from that charge to the dispatch of a commander and the seizure of the Tokugawa house's and its retainers' lands.[15]

    The battlefield's name had entered the justification for extending control beyond the fighting positions. The Shōnai compiler questioned the description of the defeated side's allegiance; the notice instructed people on the consequences of accepting the government's description. Defeat, the designation of an enemy, and the organization of succeeding rule were connected processes, rather than one universally accepted event.

    Toba–Fushimi opened the Boshin War rather than ending it. Fighting continued beyond the Kyoto–Osaka region and into 1869.[11] The outcome here changed the military and political situation profoundly, but it did not make all later resistance, negotiation, and submission unnecessary. The ruined magistracy and the roads of withdrawal belong to the opening transformation, not to a single moment when the whole country changed its mind.

    7. What the town kept

    The local traces lead back from declarations to the premises in which people lived and worked. Gekkeikan's historical collection account identifies a sake bottle marked with its predecessor's trading name, Kasagiya, among objects excavated from the former magistracy in 2008. Its surface had been damaged by heat.[12]

    The bottle connects the compound to ordinary trade. Heat damage records the destruction of goods as well as military positions; it does not identify the person who last used them.

    Damage was uneven. The Ōkura family's 1828 residence survived, according to the company's history, while nearby premises burned.[12]

    Gokōnomiya contains another, later rearrangement. Garden stones from the former magistracy site were transferred to the shrine, where Nakane Kinsaku created a new garden in 1961. The Kyoto City Greenery Association records the move in connection with the redevelopment of the old site.[13] Material from one side's establishment came to rest at the position occupied by its opponents, in a design made almost a century after the battle.

    That garden is not a diagram of the firing lines. It belongs to the later preservation and reuse of a town whose earlier institutions had been damaged and replaced. At Toba, too, the crossing's historical significance is carried partly by a modern commemorative marker: Kyoto City's record dates the Koedabashi-area battle monument to 1998.[14] Remembering a location and retaining its entire nineteenth-century configuration are different achievements.

    Taken together, the two fronts explain why the struggle over government cannot be separated from movement through particular places. An advance required passage. Defenders occupied crossings and compounds. Retreat required routes whose security depended on other domains. An imperial command advanced along a road, while a written order attempted to turn uncertain affiliations into enforceable obligations. The battle changed who could enter Kyoto and who could issue demands from it. Fushimi's damaged premises show where people encountered that change before it became a settled chapter title.

    Next topic · HO Guides

    Mibu and Ikedaya: A Police Force in Other People's Houses

    How a neighborhood accommodated the Shinsengumi—and how the use of domestic space changed when lodging became armed service.

    Continue reading
    Sources and NotesView sourcesHide sources

    Materials were consulted on 16 September 2026. Full dates in the narrative use the Gregorian calendar: the first month's third and fifth days correspond here to 27 and 29 January 1868. Original Japanese dates are retained in source descriptions. No modern map is treated as an unchanged battlefield survey. Partisan explanations are attributed, and statements about motives in proclamations or memoirs are not adopted as independent findings.

    1. [1]

      Ernest Satow, A Diplomat in Japan (London: Seeley, Service & Co., 1921), chapter XXV, “Hostilities begun at Yedo and Fushimi,” printed pp. 310–312, especially the reports of Itakura and Nagai and the subsequent withdrawal reports. Public-domain text, Project Gutenberg ebook 43541: https://www.gutenberg.org/cache/epub/43541/pg43541-images.html . See chapter XXV and the relevant preceding sections of chapter XXIV; the entire book was not read for this essay. Satow was in Osaka, not at the two fronts. His account distinguishes observations from reported information, though not every reported detail can be independently verified. The official characterization of the advance as a retinue, its anti-Satsuma document, and its intended destination are attributed explanations, not this essay's verdict on the advance. His force totals and imperfect road-direction description are not used.

      Return to the reference ↑
    2. [2]

      Jōnangū, 「鳥羽伏見の戦い」, one-sheet historical map, sections 「開戦前夜」「鳥羽の戦い」「伏見の戦い」「錦の御旗 翻る」. https://www.jonangu.com/img/file3.pdf . Used for the two fronts, the Toba roadblock and gun positions, the prince's appointment and subsequent route, and the distinct Tominomori and Senryōmatsu retreat areas. The publication combines a historical account with a modern excursion map; its suggested walks are not treated as troop itineraries. Precise troop counts, dialogue, and the map's visitor logistics are not used. The historical Takase channel is not silently identified with the present East Takase River.

      Return to the reference ↑
    3. [3]

      Ibaraki Prefectural Museum of History, 「Ⅳ 徳川慶喜大政奉還する」, sections 1–2 on the submission and acceptance of the 1867 proposal: https://www.rekishikan.museum.ibk.ed.jp/06_jiten/tokugawa/houkan.htm . The page’s evaluations of political skill, motives, and the “Eight Proposals” tradition are not used. National Archives of Japan, 「慶応3年(1867)12月 王政復古の大号令が発せられる」, full short introduction: https://www.archives.go.jp/ayumi/kobetsu/k03_1867_01.html . Used for abolition and replacement of offices. Satow, chapter XXIV, printed pp. 297–302, independently supplies the Western January 3 dating and reports Yoshinobu’s withdrawal to Osaka; pp. 306–308 show unresolved administrative transfer, URL in note 1. The traditional-calendar restoration date, Keiō 3, twelfth month, ninth day, is not rendered as a Western December 1867 date. Resignation, proclamation, and practical transfer remain separate events.

      Return to the reference ↑
    4. [4]

      Satow, A Diplomat in Japan, chapter XXIV, printed pp. 306–308, discussion of a proposed imperial notification, territorial surrender, and exchanges with Mutsu, URL in note 1. Used for the existence and substance of the recorded administrative and diplomatic questions, not Satow's conjectures about private stratagems or his assessments of individual competence. This brief Osaka context explains what remained unresolved for the northward advance; it is not a separate account of Osaka Castle.

      Return to the reference ↑
    5. [5]

      Fushimi Ward Federation of Senior Citizens' Clubs, 「伏見の戦跡(御香宮神社)」, ふしみ昔紀行 36, March 2004, hosted by Kyoto City, webpage dated 12 February 2009. https://www.city.kyoto.lg.jp/fushimi/page/0000014952.html . Used narrowly for the positions of the shrine, southern magistracy, eastern Ryūunji artillery, and burning of the magistracy. This is a local-history contribution hosted by the city, not a military operations record. Its compressed phrasing about Yoshinobu leading the advance and its generalizations about the banner are not adopted.

      Return to the reference ↑
    6. [6]

      Gekkeikan, 「京・伏見で、明治からの日本を決した戦闘 鳥羽伏見の戦い(1868年)」, section 「大手通り(大手筋)を挟んで市街戦」. https://www.gekkeikan.co.jp/enjoy/kyotofushimi/fushimi/fushimi11.html . Used for the Ōtesuji relationship and its explicit description of firearms on both sides. Exact distances and force totals, speculative counterfactuals about victory, and the article's separate late date for the banner are not adopted. The accompanying eighteenth-century image of the magistracy is not used to reconstruct battlefield geometry.

      Return to the reference ↑
    7. [7]

      「德川慶喜征討ノ大號令ヲ發シ諸侯ヲシテ去就ヲ決セシム」, Hōrei zensho, no. 11, as identified and transcribed on Japanese Wikisource, version 168786: https://ja.wikisource.org/w/index.php?title=德川慶喜征討ノ大號令ヲ發シ諸侯ヲシテ去就ヲ決セシム&oldid=168786 . See both complete paragraphs of this public-domain text. The main paragraph offers acceptance to former adherents and hereditary retainers, allows conditional future petitions, and threatens those communicating with or sheltering designated enemies. The appendix explicitly places the commander’s appointment before the general proclamation. The transcription’s header and appendix give the first month’s seventh day. A related National Archives catalogue entry, 「徳川慶喜征討大号令」, 太00211100-00600, JACAR A15071486300, is indexed under the ninth day: https://www.jacar.archives.go.jp/das/meta/A15071486300 . That catalogue identification was retrieved, but its images were not obtainable. The differing dates have not been explained by inventing separate issue and receipt stages. Accordingly the narrative uses the sequence and provisions, not a precise day for this general proclamation. This is a reading of the identified transcription, not a claimed collation with the archival original. The government’s accusations remain attributed.

      Return to the reference ↑
    8. [8]

      Itō Yuriko (伊東由里子), 「戊辰戦争、津藩と桑名藩」, January 1988, Mie Prefecture historical archives web edition. https://www.bunka.pref.mie.lg.jp/rekishi/kenshi/asp/arekore/detail20.html . It cites 三重県史 資料編 近代1(政治・行政I) (1987); that volume was not read independently. Used with attribution for Tsu, Kameyama, and the division within Kuwana. The article's explanatory framework is not extended into an assertion of one motive shared by all retainers. Its account of divination within Kuwana is not needed here.

      Return to the reference ↑
    9. [9]

      Satow, A Diplomat in Japan, chapter XXV, printed pp. 314–315, conversations with wounded Aizu soldiers on 1 February 1868, URL in note 1. The soldiers' complaints about support, Tsu, formations, and weapons are identified as retrospective participant explanations relayed by Satow. They are not used to authenticate every battlefield accusation, to assign all retreat to betrayal, or to evaluate an individual commander's ability.

      Return to the reference ↑
    10. [10]

      University of Tokyo, Historiographical Institute, 「鶴岡市郷土資料館所蔵幕末維新関係史料の調査・撮影」, 所報 42, entry HL22.510-ア-1~6, 「日本變新録」, descriptive paragraph and extract from the opening of book 8. https://www.hi.u-tokyo.ac.jp/publication/syoho/42/saiho_42_006/ . The relevant collection description and reproduced passages were read, not the entire collection or all surveyed documents. The institute identifies compilation in 1869 by the Shōnai retainer 秋保親愛 and warns that the later historical-society copy contains transcription errors. Used for the limited proposition that a near-contemporary writer questioned the enemy-of-the-court designation and called for examination of responsibility. His belief about others' intentions is not promoted into a finding that no participant intended hostility.

      Return to the reference ↑
    11. [11]

      National Archives of Japan, 「近代国家 日本の登場」, section 1, 「戊辰戦争」. https://www.archives.go.jp/exhibition/digital/modean_state/contents/boshin-war/index.html . It places Toba–Fushimi at the beginning of the war and the end of the fighting in 1869. Its featured flag drawings were made in 1888–1890; they are not treated here as illustrations made at the opening battle, nor was their complete image series inspected.

      Return to the reference ↑
    12. [12]

      Gekkeikan, article in note 6, 「幕末維新の激動を越えた笠置屋の酒」 and captions for the Ōkura residence and sake bottle. Used as the collection holder’s account of the 1828 residence, the bottle found in the 2008 investigation, and heat damage. The underlying 2010 excavation report was identified but not obtained; a separate 2015 report could not be read and is not substituted for it. Speculation about particular soldiers drinking from the bottle or a fire pump saving the house is not used. No claim that this is the only surviving Edo-period brewery structure is made.

      Return to the reference ↑
    13. [13]

      Kyoto City Greenery Association, 「御香宮神社」, garden date and 「歴史・いわれ」, paragraph describing the magistracy site's later redevelopment, stone transfer, and Nakane Kinsaku's 1961 work. https://www.kyoto-ga.jp/greenery/kyononiwa/2016/01/post_19.html . Used for the modern transfer and arrangement, not for its traditional claims about Kobori Enshū's rewards or its imaginative reconstruction of the modern designer's thoughts. The present garden is not represented as the original magistracy garden in its original layout.

      Return to the reference ↑
    14. [14]

      Kyoto City Historical Museum, marker record HU059, 「鳥羽伏見の戦勃発の地小枝橋」, location, erection date, and inscription fields. https://www2.city.kyoto.lg.jp/somu/rekishi/fm/ishibumi/html/hu059.html . Record consulted for the Koedabashi-area monument, erected in 1998. Its commemorative account is not used as independent confirmation of precise dialogue, gun types, or troop positions. The essay does not identify the present bridge structure or alignment as the unchanged 1868 crossing.

      Return to the reference ↑
    15. [15]

      Kyoto Tachibana University, Faculty of Letters Digital Archive, 「王政復古に関する高札」, dated Keiō 4, first month, twenty-fourth day (1868), complete transcription and accompanying interpretation. https://www.tachibana-u.ac.jp/documents_and_works/page10.html . See the full transcribed notice and all three explanatory paragraphs. Issued under the designation of Chōshū as the government army’s advance force, the text links Fushimi to its accusation of rebellion, confiscation, continued land taxes under existing rules, and instructions to await court orders. It is an independently hosted primary text showing a subsequent administrative use of the battlefield claim, not evidence that this particular board stood in Fushimi or that every announced measure was successfully enforced. The physical board and an archival provenance dossier were not inspected.

      Return to the reference ↑

    Table of Contents

    1. 00Opening
    2. 011. A resignation had left an army in place
    3. 022. Two approaches, two kinds of obstruction
    4. 033. The magistracy became a target
    5. 044. The banner moved; the order followed
    6. 055. A line of retreat depended on other people's decisions
    7. 066. Calling someone an enemy did not make the description uncontested
    8. 077. What the town kept
    Sources and notes

    Reading preferences

    Text size
    Glossary intensity

    Choose how much explanation appears as you read.

    Saved on this device. All terms remain available in the glossary.

    Browse the glossary →
    History Omakase

    Opening reading…