Song dynasty incense registers: which ones survive

Moonlit Osmanthus opens on sweet osmanthus, then warm wood and the soft, vanilla-like resin of benzoin. Its materials are golden osmanthus, an agarwood accord, sandalwood and benzoin, and it lasts for years. The oldest recipe we know for osmanthus beads comes from an incense register finished in 1322, and that register is also the reason most of the older ones are remembered at all.

A register, pu 譜, set out what was known about one kind of thing: its varieties, how it was made and used, and what had been written about it. From the eleventh century, Song scholars wrote incense registers, xiangpu 香譜. They list the materials and where they came from, give recipes for blends, and collect stories, poems and the names of tools. At least eleven were written. Two have come down whole. This is a guide to who wrote them, what was lost and where to read what is left.

Why a register was needed

The 1322 preface to Chen’s register says why the job was hard: “Incense is one of the five smells, and people delight in wearing it. But a register of incense could only be made, in full, by someone from a family of officials, widely read, who had sailed on the merchant ships across the sea.”

The materials were the problem. In the classics, the preface goes on, incense meant millet, mugwort and fat on the altar, orchids at the belt and fragrant herbs in wine. “Since the Han and the Tang, anyone who spoke of incense had to take it from the products of the southern seas, and so there had to be registers.” Agarwood came from Hainan and Cambodia, frankincense and ambergris from the Arab lands, sandalwood by way of Srivijaya. A reader in Kaifeng or Hangzhou needed a book to tell the grades apart and to know what to do with them.

The book printed under Hong Chu’s name shows the usual plan. It has four parts: the materials, 香之品; marvels, 香之異, mostly from old tales; affairs, 香之事, the history and lore; and methods, 香之法, the recipes.

Shen Li: the first, now lost

The earliest Song incense register we know by its author was compiled by Shen Li 沈立, a high official whose family, a later writer says, “was once said to own more books than anyone”. It is lost. What we have of it is a handful of entries that Chen’s register marks “from Shen’s register”, and one note in Shen’s own words.

That note, dated the second month of 1074, sits in Chen’s second volume. Shen, then prefect of Xuancheng, writes that he had once compiled an incense register but had not described the hundred-ke seal in detail. A man named Wu Zhengzhong of Guangde had since sent him the carved seal and its incense recipe, “and only a fine mind could have made it”, so Shen had it cut in stone for anyone who cared to copy it. The seal itself is described in How the Song burned incense.

Some decades later, the poet Zhou Zizhi compared Shen’s book with Hong Chu’s. Shen’s register “was thought to have everything the various writers had recorded about incense,” he wrote. “Set beside what Hong Jufu collected, it does not have one or two parts in ten.”

Hong Chu: the collector

Hong Chu 洪芻, styled Jufu, passed the palace examination in 1094. His mother was the sister of Huang Tingjian, the great Song poet, who admitted to an incense habit. The nephew was a collector.

Chen’s register keeps a note by Huang that shows him at work. Huang and a monk named Hong were moored for the night on the river outside Tanzhou when two paintings of plum blossom in ink arrived by boat from the painter Zhongren. They looked at them by lamplight. “All that is missing is incense,” Huang said. The monk laughed, took a stick from his bag of curios and burned it, and it was “like walking along the hedges of Solitary Hill at dawn, in the first cold”: Solitary Hill, on West Lake, was famous for its plum trees. The recipe had come from Su Shi, who had it from the family of the statesman Han Qi. “Hong Jufu, who collected incense recipes old and new, thought none surpassed it,” the note goes on, and since its name did not say what it was, he renamed it Plum Called Back to Life.

Hong’s life ended badly. In 1127, after the Jin took the capital, he was punished for helping the Jin strip the city of its wealth, and sent for life to Shamen Island, off the Shandong coast, where he died.

His register is lost too, apart from extracts and quotations. One of those, in a Song commentary on Huang’s poems, records the Southern Tang ruler’s bed-curtain incense, the source of Goose-Pear Canopy.

The book under Hong’s name

A two-volume Incense Register has long been printed under Hong Chu’s name, and it is one of the two that survive whole. The Qing scholars who catalogued it for the imperial library in 1781 doubted it was his. The Song bibliographer Chao Gongwu had described Hong’s register as full of old recipes, among them a Han palace incense recorded by Zheng Xuan and a Lesser Zong incense from the History of the Southern Dynasties, and as missing the agarwood that the Tongdian says was burned for Heaven. The printed book has none of the recipes and does have the agarwood. A 2019 study came to the same view: it is an anonymous compilation of the Southern Song, drawn partly from Shen’s and Hong’s registers.

It is still worth reading. Its first part describes forty-two materials. The part on affairs keeps the fifth-century historian Fan Ye’s short preface on blending, quoted in Hexiang: blended incense, worn as beads, and a set of older writings on incense. Its methods include the Southern Tang ruler’s recipe in a few lines: a liang of agarwood, finely chopped, with the juice of ten goose pears, steamed three times in a silver vessel, “and when the pear juice has dried, use it”.

Its preface, a single paragraph, explains the whole enterprise. The classics, it says, matched good scent with good conduct. “The gentleman washes himself as clean as snow and scents himself with what is right, so that his good name, like a fragrance, never fades. My register has the same intent.”

Ding Wei on Hainan agarwood

The longest piece in that book is not a recipe. It is the Account of Heavenly Incense, Tianxiang zhuan 天香傳, by Ding Wei 丁謂 (966–1037), a chief minister who knew the palace incense from years of court rites. Dismissed in 1022, he was soon banished to Hainan. “In the midst of my troubles I had no worldly cares,” he writes, “and the pleasure of the censer only grew.”

He had heard that Hainan produced a great deal of incense, and he bought it in the villages: “not one piece in ten was fake”. A local officer explained the trade. The best agarwood came from the mountains in the middle of the island, but the Li people who lived there farmed for a living and only went up to gather incense in winter, when the ships from Hangzhou arrived to buy it.

Ding sorts the island’s incense into four names and twelve shapes, from crow-black wood to the “partridge-speckled” kind that Huang Tingjian later named in a poem. He tells of an Arab ship blown onto the island, whose captain held a lavish banquet to show off his wealth. The islanders noticed that his censer smoke hung low and his incense was dry, light and scorched. So they burned incense from across the strait at the same table, and its smoke rose high, “and the longer it lasted, the better it smelled”. The ship’s men were put to shame. He passes on what the ships said about frankincense: it came from a land of much rock and little rain, where one mountain “has had no rain for thirty years”. Grit in frankincense, he adds, is not a sign of fakery: the land there has no soil.

He also quotes an old Daoist method for scented beads. The materials are powdered as fine as they can be, pounded thirty thousand times, sealed in a good vessel, steamed and blended, “divided into pieces the size of beans and rolled into pills, strung like pearls and dried in the sun”. More on how beads like these were worn is in Did people in ancient China wear scent?

Ye Tinggui at the port

Ye Tinggui 葉廷珪 passed the palace examination in 1115 and from 1148 governed Quanzhou, the great port of the southern sea trade. He wrote a book on the incenses of the southern lands, the Nanfan xianglu 南蕃香錄. It is lost, but Chen’s register quotes him more than twenty times, usually at the head of an entry, where he says where a material comes from and how good it is.

His notes read like a buyer’s guide. Agarwood from Cambodia is the best and Champa’s comes second. Sandalwood comes from Srivijaya, “its scent clear and strong, but it escapes easily”. Benzoin is no good burned alone, but it brings out the scents of other materials. Rose water is the “flower dew” of the Arab lands. His account of ambergris is in Elegant Orchid & Ambergris.

Chen’s register: the one that kept the others

The second register to survive whole, and the fullest, is Chen’s. Its compiler, Chen Jing 陳敬, styled Zizhong, came from Henan, and the book took two generations: it was finished by the next one, Haoqing. The preface names the older registers it absorbed by their authors’ surnames, “Hong, Yan, Shen and Ye”, and says “all of them are in this book”. Chen kept two short pieces by a Yan Bowen, and Yan is usually taken to be him. The Qing editors counted eleven older registers in all.

Haoqing took the finished book to Xiong Penglai, a scholar who called himself the Pengli fisherman, after the old name of Poyang Lake, and asked him for a preface. Xiong protested that surely someone grander should write it, then told some old stories against incense. A Jin official named Liu Jihe loved it so much that he would stand over the censer to scent himself, until his secretary told him it was vulgar. Liu laughed and took the point. Xiong ends: “But a study with a zither by the window cannot tend its censer without this register.” He signed it in the autumn of 1322.

There are four volumes:

Volume 1: well over a hundred materials, among them borneol, ambergris and rose water, then how to prepare them, refine honey, fire charcoal, blend, store, mature and burn. The section on preparing materials and honey is followed step by step in Hexiang.
Volume 2: seal incense, with Shen Li’s hundred-ke seal, then blends: palace incenses, the bed-curtain incense of the Southern Tang ruler, and two dozen recipes named for ambergris, most of which contain none. Dragon’s saliva incense looks at them.
Volume 3: more blends, many named for plum blossom, orchid or osmanthus, including the four incenses of Huang Tingjian, then scents for pouches and clothes.
Volume 4: scented beads, among them Commissioner Sun’s osmanthus beads, which “kept this way, will hold their scent for years”. Then the stories and lore, censers, and a long anthology of prefaces, rhapsodies, poems and song lyrics.

The notes on Huang Tingjian’s four incenses are some of the best writing in it. Of one, called Deep and Still, made for him by a friend in Jingzhou, he says: “This incense is quiet, plain and lonely, not the kind the world goes for. Now and then I let the curtain down and burn a stick, and it is as if I see him.”

The Qing editors grumbled that Chen padded the book with quotations from the classics, but they gave him his due. “The eleven registers do not all survive today. Chen gathered their words into one, and many lost passages and forgotten stories are kept because of him.” Moonlit Osmanthus owes its oldest recipe to that.

Outside the registers

Not every Song writer on incense wrote a register. Fan Chengda governed Guilin in the far south and, in 1175, wrote an account of the region with a chapter on incense. Hainan incense, he says, has a clear, pure scent “like lotus, plum blossom, goose pear and honeycomb”. Burn a piece the size of a gaming die and the whole room is filled. Even its ash does not smell scorched. The people of the coast bought it from the Li of the hills at one ox for a load, and when they sorted the load at home, not one or two parts in ten were true agarwood. Ship-borne incense, he adds, “is often rank and fierce”.

Questions

What is the oldest Chinese incense register? Of the Song ones, Shen Li’s, from before 1074, is the earliest we know by its author, and it is lost. The two-volume book under Hong Chu’s name and Chen’s register, finished in 1322, are the ones that survive whole. Older writing on blending exists only in fragments, such as Fan Ye’s fifth-century preface.

Did Hong Chu write the Xiangpu printed under his name? Probably not. The Qing editors doubted it in 1781, and a 2019 study agrees. Hong’s own register survives only in extracts and quotations.

Do the registers have recipes for scented beads? Yes. The book under Hong’s name and Ding Wei’s account both describe beads pounded, rolled, strung and dried. Chen’s fourth volume opens with bead recipes, including osmanthus beads and ambergris beads; Elegant Orchid & Ambergris describes the second.

Can the old recipes be made today? Not as written. Many call for musk, real ambergris or wild agarwood, now protected or restricted, and some for lead white or saltpetre. Our formulas use materials the registers knew, with accords in place of agarwood and ambergris. The materials register lists them.

Where can I read them? Both surviving registers are in the Siku quanshu, the Qing imperial collection, and on Chinese Wikisource, linked below. The translations in this article are ours.

Sources

Chen Jing, Chenshi xiangpu 陳氏香譜, preface and Siku quanshu notice (熊朋來序, 1322; 提要); volume 1 (香品, 脩製諸香, quoting 葉庭珪 and 沈譜); volume 2 (百刻香印, with Shen Li’s note of 熙寧甲寅, 1074; 江南李主帳中香); volume 3 (韓魏公濃梅香, with 黄太史䟦; 黄太史四香, 深静); volume 4 (香珠; 孫廉訪木犀香珠; 覔香, 顔博文).
Xiangpu 香譜, two volumes, printed under Hong Chu’s name: Siku quanshu text and notice (提要, 1781, quoting 周紫芝 and 晁公武); punctuated text (序; 香之品; 香序, Fan Ye; 天香傳, Ding Wei; 江南李王帳中香法).
Fan Chengda, Guihai yuheng zhi 桂海虞衡志, 志香 (1175).
On Hong Chu’s life and the authorship of the printed Xiangpu: 洪芻, Chinese Wikipedia, citing 董岑仕, 〈今傳本《香譜》非洪芻作辨〉, 中國四庫學 4 (2019). On Ye Tinggui: 葉廷珪 (宋朝), Chinese Wikipedia.