The Hercules Tsuba

Eight capitals run round the plate of this iron guard. They are not engraved on its surface: they are the openwork itself, left standing in the iron with their serifs, and together they make the ring that holds the centre of the guard to its rim. Read clockwise from the top, they spell HERCULES, the Latin name of Hercules.

Enlarge

Enlarge

The plate is a square with softened corners. A broad, rounded rim stands above the thinner pierced ground, so that the weight of the guard lies at its edge and the letters seem to hang in light. The iron is dark and finely textured. Copper lines part of the opening for the tang, and the two smaller openings beside it are closed with dark plugs: the traces of the guard’s fitting to a sword.

Roman letters were printed in Japan in the last years of the sixteenth century. The press of the Jesuit mission, set up first at Kazusa and then at Amakusa, in Higo, issued books in Latin between 1591 and 1597, among them a Latin grammar in 1594 and a dictionary of Latin, Portuguese and Japanese in 1595; and Virgil’s Aeneid, in which the name of Hercules stands, was probably read in the mission’s seminary. It is the world in which a word like this one could be read in Japan, and it is offered here as that: the context of the guard, not the source of its design.¹

The guard can be dated by that world. Roman letters could be known and copied in Japan only in the Christian century: the mission brought its press to Japan in 1590, the expulsion edict of 1614 ended its printing, and the expulsion of the Portuguese in 1639 closed that world; in Higo, Roman-letter seals were given up in 1648. Its letters are of the kind printed in those years. The guard was made between about 1590 and 1640.²

Roman letters came to Japan with the Jesuits, whom Oda Nobunaga took under his protection in 1569 and kept there until his death in 1582. The guard belongs to the afterglow of that world, before the country closed.

The word itself had reached Japan. In the Jesuits’ letters of the 1580s, it appears, one Japanese Christian was known as “the Hercules of Bungo”: Shibata Reinō, confidant and spear-master of the Christian lord Ōtomo Sōrin. Among the missionaries the name appears to have praised a warrior’s strength; it belonged to no scripture.³

Its capitals are of the kind set on the title pages of the mission’s Latin books, and to the museum their precision suggests a pattern, laid out by someone who knew Roman letters.

Elsewhere on Japanese guards, it appears, Roman letters were chiefly Christian signs, kept for those who could read them: INRI, alpha and omega. Here the word stands in the open, the name of a pagan hero, not a line of scripture. That, Akiyama Kyūsaku would later argue, is how the guard came safely through the years when Christianity was forbidden.⁴

To the museum it reads as a commission, made for someone who wanted this word in iron, in the decades when nanban fashion, from Portuguese dress to painted screens, ran through the taste of the great. Few things survive in which that first encounter is written so plainly.

The written record of the guard begins on 19 December 1905. That day the connoisseur Akiyama Kyūsaku was shown a tsuba as “the treasured piece of a certain person”: oblong, with something square in its outline, a thick rim with the ground cut down inside it, and eight Roman letters pierced through the ground. A senior colleague present called it at once the work of an armourer. Akiyama did not agree, though he could not yet say whose work it was.⁵

Two years later, having seen a tsuba signed by Hōan, Akiyama came to think that the lettered guard was Hōan’s work too. He asked to see it again, set the two side by side, and found them of one hand.

Hōan was a family of tsuba makers. By the accounts that Wada Tsunashirō and Ogura Sōemon carry, its first master, Kawaguchi Saburōemon Hōan, lived in Owari and worked in the circle of Nobuie and Yamakichibei; the family followed the Asano lords to Kai and to Kii, and in 1619 to Hiroshima.⁶

By 1913 the guard was in the collection of Count Terauchi. That year Wada Tsunashirō, writing of Hōan, described an unsigned iron tsuba by him with the letters “Hercles” pierced in it, and added in small type 寺内伯爵所藏, “in the collection of Count Terauchi”. The print has seven letters, the U lost in setting the type; Akiyama, who had twice held the guard, counted eight.⁷

The Count Terauchi of 1913 was Terauchi Masatake (1852–1919), a soldier of Chōshū: Army Minister from 1902 to 1911 and Governor-General of Korea from 1910 to 1916, created count in 1911, field marshal in 1916, and prime minister from 1916 to 1918.

Enlarge

He was also a collector. The private library of the house, which he began and his son and heir Hisaichi opened in Yamaguchi in 1922 as the Ōho Terauchi Bunko, survives in large part at Yamaguchi Prefectural University. Of the house’s fittings, Ogura in 1925 closes his pages on Hōan by naming this guard once more: “a celebrated tsuba of the Terauchi collection”.⁸

In 1925 Ogura Sōemon, fifth head of the house of Amiya, wrote the chapter on Hōan in his Nihon tsuba tōkō-roku. As the masterpieces accepted as works of the first-generation Hōan he named two guards: “the English-letter openwork tsuba of the Terauchi family”, given in Wada’s account, and a signed round iron guard pierced as a wheel, formerly in the collection of Inukai Tsuyoshi (Bokudō). His words “of the Terauchi family” show the lettered guard still with the house after the count’s death in 1919.⁹

Ogura described the guard once more, it appears, in a lecture on sword fittings first printed in 1934–35: an exceedingly finely made tsuba, a square plate with softened corners and a raised rim, with the letters “Hercles” pierced in its ground. With the signed wheel guard, he wrote, it raised Hōan’s name the higher; and of both he added that they seemed to stand at a considerable distance from Hōan’s usual work.¹⁰

Over thirty years these accounts agree in every particular with the guard shown here: eight Roman letters pierced in the ground, the name of Hercules, a square plate with softened corners, and a thick rim standing above a lowered ground. The guard of the Terauchi family is this one.

Enlarge

Today the guard carries the papers of the NBTHK, the Society for the Preservation of Japanese Art Swords, as Tokubetsu Hozon Tōsōgu, issued in 2025. The paper names it 英字透鐔, an openwork tsuba of Western letters. Its paper records no history; it attributes the guard, unsigned, to Jingo 甚五, with the characters the NBTHK reserves for the first and second masters of that name. Its pre-war history is the museum’s own research.¹¹

Enlarge

This is the Terauchi guard, and two judgements stand side by side in its history: Hōan, as the connoisseurs who wrote of it between 1907 and 1935 judged it (the first generation, in Ogura’s words of 1925), and Jingo of the NBTHK paper today. The guard is unsigned. What follows is the museum’s own study of the work, weighed against both.

Enlarge

Hōan, the literature records, finished most of his guards with etched relief. “To work a design in low relief by etching is Hōan’s forte,” writes Owari to Mikawa no tsubakō (1983), a study of the tsuba makers of Owari and Mikawa, of plate guards and openwork guards alike; and it adds that “rarely there are pieces with no etched design.” The NBTHK says as much of a tsuba signed Hōan (Jūyō Tōsōgu, 63rd session, designated 17 November 2017): “making full use of his favoured yakite-kusarashi technique, he brings a subtle variation to the flat ground.” The acid eats back the softer iron and leaves the harder standing, in low ridges with softened edges.¹²

The method is seen at its finest on the other guard of Ogura’s pair. The signed wheel guard formerly in Inukai’s collection, which Ogura set beside this one as the twin unsurpassed pieces among Hōan’s masterpieces, carries a long-eared hare at the edge of its ground: “a long-eared hare in yakite-kusarashi set on the ground at the rim,” as Owari to Mikawa no tsubakō (1983) describes it. It is very probably the guard designated Jūyō Tōsōgu at the NBTHK’s 48th session (2002), whose description records that Hōan “brings out the long-eared hare in his favoured yakite-kusarashi technique.” Ears raised by etching stand as low lines, soft at the edges, a little proud of the iron.¹³

This guard carries no etched design: its letters are plain openwork in hammered iron. Along the rim run the fine linear ridges of tekkotsu, the harder strands of the forged iron that stand slightly proud of the surface. They are natural to the iron, and over most of the rim they are only that: short, broken and uneven.

Enlarge

Enlarge

One line is different. It stands higher, longer and more even than the rest, a single emphatic line along the rim. In the museum’s reading it was brought up by etching, the iron on either side eaten back as it is around a hare’s etched ears: the yakite-kusarashi recorded as Hōan’s forte, turned not to a design but to the iron itself.

Enlarge

Enlarge

In the museum’s reading, the two guards that Ogura paired share that handling. Another signed Hōan wheel guard with a long-eared hare shows it too: the ears are soft-edged ridges raised from the rim, with etched lines running along the rim beside them. Of the likenesses the museum finds, this is the closest, for it is the same method as on signed works.

Enlarge

Enlarge

Enlarge

Enlarge

A second likeness lies in the construction. The letters are anchored to the rim with a small recess where the two meet, so that each seems seated in the rim rather than set against it. On the signed Hōan wheel guard with a long-eared hare shown here, the spokes meet the rim in the same way.

Enlarge

Enlarge

The iron points the same way. It is of the first quality: dense and soft at once, as Akiyama found it, the quality for which the great masters of iron are known. Of the signed guard designated Jūyō Tōsōgu in 2017 the NBTHK writes: “Hōan’s iron is renowned for its tenacity, so much so that it is called uwabami-tetsu, python iron.” To the museum, the iron of this guard has that same dense, tenacious, almost viscous quality. Its colour is reddish, and darker than the iron of the guards known under Jingo’s name.¹⁴

Against these stands the paper’s Jingo. The paper’s form allows the first or the second master. The second, Jingorō, was born in 1620 and was twenty at the close of the years in which the guard was made; within those years the paper’s form points rather to the first master, Shimizu Jinbei (1599–1675), who settled in Higo in 1632.¹⁵

The guards known under Jingo’s name are bold too, but their boldness lies elsewhere. Again and again the NBTHK’s descriptions of them name one method as his own, unmatched among the makers of Higo: brass applied inlay (shinchū suemon-zōgan), a figure cut separately and set large on the face of a hammered, fired iron plate. The figure is above all an eagle or hawk on pine, the subject the descriptions bind most closely to his name; an owl, a crab or an octopus besides. The reverse is left open, with a crescent moon or the tip of a branch. Openwork is the art of the neighbouring Higo houses, the Hayashi and the Nishigaki, not of Jingo. The two signed guards of the school that the NBTHK has designated with their Higo mountings, both given to the second master, are inlaid too, one in gold and one in brass wire. This guard carries no inlay and no brass. Its boldness is in the iron itself, cut through so that the letters are the whole design.¹⁶

In the museum’s study, then, four points stand on Hōan’s side. Akiyama, who knew Hōan’s signed work, set the two guards side by side about 1907 and found them of one hand. The iron has the tenacity the NBTHK records as the mark of Hōan’s, and a reddish colour, darker than Jingo’s iron. Etching, the method recorded as his forte, has raised one line of the rim, as it raised the ears of the hare on the guard Ogura paired with this one. And the letters are seated in the rim as the spokes of a signed Hōan wheel guard are. On the other side stands the paper’s Jingo, which within the dates points to the first master.

Here the museum allows itself a thought, offered as no more than that. The guard was, as the museum reads it, made to order and at the height of the craft, and it was left unsigned. Yet its rim carries that one line, raised by the method its maker’s work is known for. It is tempting to read the line as a signature of a quieter kind: the maker’s mark left in the iron, where a name would have stood.

A paper records a judgement made at a single sitting, among very many works; at that level its attributions are, of necessity, cautious. The museum’s study goes further. On the evidence of the work itself, the museum attributes the guard to the first-generation Hōan.