About Iki Island

Gateways: 郷ノ浦港・芦辺港・壱岐空港 (Japanese) (source: city.iki.nagasaki.jp).

Routes and today's status are transcribed from the official boards in the "Iki today" ferry section above.

• See (見る, JP) 319
• Eat (食べる, JP) 188
• Stay (泊まる, JP) 93
• Do (遊ぶ・体験, JP) 104
• Buy (買う, JP) 47

The former lord of the Hirado domain ordered a gazetteer of Iki in 1850 (Kaei 3), and a shrine priest on the island finished it eleven years later in twenty-five volumes. Twenty-four of those volumes take one village each. We have transcribed 3 things out of them and made an index of each. Every index is in Japanese, and each describes the island as it was then, not as it is now.

Not every volume gives every figure. The count beside each item is the number of villages whose volume states it; where a volume is silent, our index says so rather than filling the gap.

• Area of paddy 22 of 24
• Yield of paddy, in koku 22 of 24
• Area of upland field (火田) 24 of 24
• Yield of upland field, in koku 23 of 24
• Number of households 24 of 24
• Number of people 20 of 24
• Number of shrines 24 of 24
• Number of temples and halls 20 of 24

By district: Iki District 11, Ishida District 10, 両郡 (Japanese) 2, district not stated 1.

A gazetteer does not always claim to know. Ours records which of 3 ways the book puts each origin, and the index keeps that distinction instead of flattening it into one voice.

• The record says the name is handed down so 45
• The record states the origin outright 28
• The record offers the origin as a supposition 13

Of the 86 names, 7 could be tied to a place on the island today; the rest are left untied, because tying them would be our guess and not the book's. 5 have a longer piece written about them.

Frontier guards and beacon fires were posted on this island, and Mongol war fleets reached this sea twice. What was left of those defences is scattered through the village volumes, a few lines at a time. We have read 44 of them out and set them in one list, across 20 villages. What stands in that list is what the record says, not what is on the ground now.

How we sort them. The sorting is ours, not the record's words.

• Castle site 32
• Beacon or lookout post 8
• Where a battle is said to have been 4

Our ledger keeps a separate column for whether the record says a beacon or a lookout station stood at a site: 9 of the 44 carry it. That column is not the sorting above, and we do not say how the two overlap.

Only 2 of the 44 give a year that can be put to a number. The rest name an era, or a happening such as a campaign, and we do not reckon a year out of that. 13 name someone as having held the site; where the record says the holder is not handed down, the column stays empty rather than being filled. For those 13, our ledger also keeps how the record puts it:

• The record states it outright 6
• The record says it is handed down so 6
• The record offers it as a supposition 1

Where the record gives two answers, both stand. Naming one of them in English would settle a question the record leaves open.

• 日高喜が島を取った年 (Japanese) The records give different years: 1564, written 永禄七年 (Japanese); 1569, written 永禄十二年 (Japanese). We have not judged which of them is right. How each reading is worded (Japanese)

None of this is on a map. What the record gives for a location is a span in the old measures and the marks on its boundary; turning that into a point would be our guess at where a site stood. The Japanese list names the volume each entry was read from, so a place can be followed up there.

The island's shrines run through several of our pages. Its temples stayed where they were, scattered through the village volumes. Here are 143 of them, read out by hand, across 24 villages. What stands in that list is what the record says, not what is on the ground now. The kind below is the last character of the name the record itself gives, so we romanise it rather than translate it: a translation would be our reading of the word.

• Ji (寺) 69
• In (院) 36
• An (庵) 30
• Bō (坊) 8

Every column is filled only where the record writes it. Where it is silent, our list says so instead of closing the gap:

• The image the temple is dedicated to (本尊) 131 of 143
• The head temple it is attached to (本寺) 63 of 143
• The person the record names as having opened it 92 of 143
• A year that can be put to a number 62 of 143

Where the record names someone as having opened a temple, our ledger also keeps how the record puts it:

• The record says it is handed down so 59
• The record states it outright 33

The years run from 997 to 1846. They are not all the same kind of year. Our ledger keeps the record's own word for what each year is the year of, and across the 62 years it holds there are 14 different words. Some mark the opening of a temple, others the passing of a priest, a restoration, the recasting of a bell, a change of name. So nothing here says that a temple was founded in a given year — gathering them under one heading would turn the year a priest died into the year a temple began. The words are the record's, so they stay in Japanese:

• 開基 (Japanese) 42
• 開闢 (Japanese) 4
• 中興 (Japanese) 3
• 寂 (Japanese) 2
• 遷化 (Japanese) 2
• 改号 (Japanese) 1
• 開山 (Japanese) 1
• 建立 (Japanese) 1
• 死 (Japanese) 1
• 草創 (Japanese) 1
• 造立 (Japanese) 1
• 鋳替 (Japanese) 1
• 入寂 (Japanese) 1
• 末山 (Japanese) 1

The record says of 3 of them that they had ceased. Those stay in the list, because the record kept them.

Where the record gives two answers, both stand. Naming one of them in English would settle a question the record leaves open. Names stay in Japanese: our ledger holds no reading for them, and supplying one would be our guess.

• ひらいた年 (Japanese) The records give different years: no year, only 慶長年間 (Japanese); 1492, written 明応元年 (Japanese). We have not judged which of them is right. How each reading is worded (Japanese)
• 開山が世を去った年 (Japanese) The records give different years: 1330, written 元徳二年 (Japanese); no year, only 正安 (Japanese). We have not judged which of them is right. How each reading is worded (Japanese)
• ひらいた人 (Japanese) The records name different people: 花陰妙栄大姉 (Japanese); 雪岩和尚 (Japanese). We have not judged which of them is right. How each reading is worded (Japanese)
• ひらいた人 (Japanese) The records name different people: 文翁永訓和尚 (Japanese); 古渓妙泉大禅定尼 (Japanese). We have not judged which of them is right. How each reading is worded (Japanese)

None of this is on a map either, and for the same reason: the record gives a span in the old measures, not a place. Halls are left out of the list — the same hall name repeats within one village and the record gives nothing to tell them apart.

Village by village, place name by place name, name by name — none of those let the island be read in time order. These 31 entries are the ones we could date, pulled out of what we have put into modern Japanese and set in order. The earliest is 646 and the latest 1741. 5 of the entries are tied to a place on the island, and between them they name 2 places. Where the record says only that something happened long ago, it is not here — we do not set such an entry against a year.

How we group them. The grouping is ours, not the record's words.

• Temples 15
• Government and administration 5
• Taxes and relief 5
• Disasters 4
• Bays and the shore 1
• War 1

This one stands beside the timeline of the shrines. What is already on that one is not repeated here, and some things sit there instead — the Mongol invasions among them. To follow the island straight through, read both.

What we have transcribed can be read by shrine, by year and by place name. It could not be read by person. This index gathers the 16 names that appear on more than one of our pages, and the 72 places where they appear. What stands under a name is what a record says there, not a life of the person. The names are in Japanese: our ledger records no reading for them, and supplying one would be our guess rather than the record's.

Where those 72 places sit:

• Notes on individual shrines 24
• Longer pieces from the records 22
• Notes taken from the 1926 shrine record 18
• Notes taken from the 1861 gazetteer 6
• Extra notes on the seats 2

Being listed as not decided is a provisional state, not a permanent one. One name stood in that list until evidence took it apart. We had printed the same name, 後藤正足 (Japanese), for the man who compiled the gazetteer of 1861 (Bunkyū 1) and for the man who compiled the shrine record of 1926 (Taishō 15). They were two men with different names: the gazetteer was compiled by 後藤正恒 (Japanese), and we had been giving him the later man's name. The mistake was ours, not a disagreement between the sources, which write the two names apart. The two are listed above as separate people and the pages that carried the wrong name have been corrected. Of the earlier compiler we still know no dates beyond the closing years of the Edo period, and we print none.

A name is not always one person. Where our sources use the same name for figures we cannot judge to be one, we neither merge them into a single life nor split them by guess: the records stay under one name, marked as not decided. Giving such a name a single title would settle a question we have not settled.

• 鎮信 (Japanese) The records gathered under this name come from Iki Meishō Zushi (壱岐名勝図誌), 1861 (Bunkyū 1); Nagasaki-ken Iki Ishida-gun Mura Yōran (長崎県壱岐石田郡村要覧), 1894 (Meiji 27); Iki Jinja Shi (壱岐神社誌), 1926 (Taishō 15); Iki no Kuni Jinja Shi (壱岐国神社誌), 1941 (Shōwa 16). The years those records state run from 1628 to 1676. We have not judged them to be one person. Our note on this name (Japanese)

The shrine and history pages here are not written from a guidebook. They are transcribed from old books. 7 of the 13 below are the books our pages are transcribed from. The other 6 are named only inside quotations in those books; we have not read them ourselves, so we give neither a year nor a compiler for them.

How each of the 13 stands in the National Diet Library's Digital Collections:

• Published on the internet 5
• Through the transmission service only 2
• We hold no copy of it 6

We do not print the identifier of a copy that can be opened only through the transmission service. An identifier for a restricted copy is a pointer into that copy, and handing it out sends people there. For those books we cite the work itself, by title and year. This English page prints no identifier for any of the books, the freely readable ones included, and builds no link into the collections; those links are on the Japanese page, and only for the books anyone can open.

A dash in the Compiled column means our ledger states no year we have verified. Titles are romanised, not translated — a translated title would be our reading of the book rather than its name — and a title left in Japanese is one for which our ledger holds no reading.

Our ledger attaches a note to 2 of these books, where what we hold about the book itself does not add up. Those notes, and the compiler recorded for each book, are on the Japanese page. They are also why the index of names on this page leaves a name undecided instead of settling it.

Where is Iki Island?

Iki Island is in Iki City, Nagasaki Prefecture, Japan: 玄界灘に面し、博多港の西北約 67 km、唐津東港の北約 42 km (Japanese).

How large is the island?

The total area is 139.42 km², and the highest point is 岳ノ辻 at 212.8 m.

What is the population?

The population is 22,856 as of 2026-09-30, published by Iki City.

How do you get there?

The gateways are 郷ノ浦港・芦辺港・壱岐空港 (Japanese). Ferries and flights connect the island with mainland Kyushu; check the day's ferry status on the sea page before you travel.

Answers are assembled from the ledger of sources.