Direct answer: 寿阳公主梅花香 is a classical plum-blossom 合香 (compound incense) that tradition attaches to 寿阳公主, a Southern Dynasties princess, and its plum character is assembled from sandalwood, herbal middles, and white plum (白梅) rather than taken from the flower. The 香乘 records the formula in 卷十八 as a list of weights followed by a three-step method: roast, grind coarse, seal for a month.
Key facts:
- The formula comes from 香乘 (Xiang Cheng), compiled by 周嘉胄 in the Ming dynasty, 卷十八.
- Nine materials: 甘松, 白芷, 牡丹皮, and 藁本 at 半两 each; 茴香, 丁皮, and 檀香 at 一两 each; 降真香 at 二钱; 白梅 at 一百枚.
- Method: dry-roast every material except the clove bark, reduce the rest to a coarse powder, then rest it (窨) in a closed jar for a month.
- The roles follow 君臣佐使 — chief, minister, assistant, guide — with sandalwood as chief and the fruit and bark carrying the plum signature.
- One 两 ran roughly 37 g in the Song–Ming period, within a range of about 31–41 g; a 两 divides into ten 钱 of about 3.7 g.
- The sandalwood is marked 不见火 (“does not see fire”), which sits oddly against the instruction to roast everything except the clove bark.
- The princess attribution is legend, tied to the 梅花妆 (plum-blossom makeup) story in 《太平御览》, not documented authorship.
- 香乘 keeps 梅花香一, 二, and 三 alongside this recipe, and the versions do not agree on a single formula.
The Formula as the Manual Records It
Direct answer: The record is nine weights and three verbs, and the verbs are where most of the craft sits.
甘松半两、白芷半两、牡丹皮半两、藁本半两、茴香一两、丁皮一两、檀香一两、降真香二钱、白梅一百枚
Read it as a list of materials first: spikenard (甘松), angelica (白芷), peony bark (牡丹皮), and lovage (藁本) at half a liǎng apiece; fennel (茴香), clove bark (丁皮), and sandalwood (檀香) at a full liǎng; rosewood or lakawood (降真香) at two qián; and one hundred white plums (白梅). The instruction that follows is short enough to miss: everything except the clove bark is roasted dry, ground to a coarse powder, and left in a sealed vessel for a month before use.
Three verbs — roast, grind, seal — and two of them are the ones a modern maker is most likely to compress. Roasting removes water before the powder goes into a jar; sealing gives the ground materials time to stop smelling like separate ingredients. Neither step adds a material, which is why a formula this short still takes a month to finish.
Four Roles, One Structure
Direct answer: The nine materials are arranged on the 君臣佐使 ladder borrowed from pharmacy, and the ladder explains why the weights fall where they do.
| Role | Materials | What the role does in the blend |
|---|---|---|
| 君 (chief) | 檀香 sandalwood, one liǎng | Sets the warm woody floor the whole blend stands on |
| 臣 (minister) | 甘松, 白芷, 藁本 — half a liǎng each | Supplies the herbal, faintly medicinal middle register |
| 佐 (assistant) | 白梅 and 牡丹皮 | Carries the tart-sweet signature the nose reads as plum |
| 使 (guide) | 茴香, 丁皮, 降真香 | Binds the registers and lifts the top of the scent |
The structure is not decoration. A chief at a full liǎng against ministers at half that weight tells you the intended balance before you smell anything: the wood is the body, the herbs are a shading layer, and the fruit is a signature that has to be loud enough to be recognized without displacing the wood. The guides sit at the edges — fennel and clove bark in full-liǎng doses, the rosewood at a fifth of that — where they act on the blend’s surface rather than its center.
A reader who knows the ladder can predict what a change will do. Raise the sandalwood and the blend gets heavier and slower; raise the plum and the tart note sharpens but the wood recedes; cut the guides and the whole thing smells flatter at the top. That predictive use is the reason the classical manuals bother to name the roles at all.
What the Weights Say About the Blend
Direct answer: Read as ratios, the formula is a four-part herbal group at half weight, a three-part backbone at full weight, one small resin accent, and a fruit measured by count instead of mass.
| Material | Recorded quantity | Group |
|---|---|---|
| 甘松 spikenard | 半两 (0.5 liǎng) | Herbal middle |
| 白芷 angelica | 半两 (0.5 liǎng) | Herbal middle |
| 牡丹皮 peony bark | 半两 (0.5 liǎng) | Herbal middle, plum signature |
| 藁本 lovage | 半两 (0.5 liǎng) | Herbal middle |
| 茴香 fennel | 一两 (1 liǎng) | Spice backbone |
| 丁皮 clove bark | 一两 (1 liǎng) | Spice backbone, unroasted |
| 檀香 sandalwood | 一两 (1 liǎng, 不见火) | Wood backbone |
| 降真香 rosewood | 二钱 (2 qián) | Resin accent |
| 白梅 white plum | 一百枚 (100 pieces) | Fruit signature |
A liǎng in the Song–Ming period was on the order of 37 grams, though the figure drifted with the dynasty across a band of roughly 31 to 41 grams; each liǎng held ten qián of about 3.7 grams, and each qián ten fēn of about 0.37 grams. Because the gram value moved, a modern maker who converts to grams inherits a false precision. The ratios are the part that travels: four at 0.5, three at 1.0, one at 0.2.
The fruit is the exception to the whole system. A hundred plums is a count, not a weight, which means the formula’s largest single ingredient by volume is also the one the text declines to weigh. That is not an oversight so much as a different kind of measurement — the maker judges the fruit by size and condition, and the ratio adjusts around it.
Building the Plum Note
Direct answer: The scent called “plum” here is a construction with two moving parts: a stable fruit note standing in for a flower that cannot survive heat, and a tart-over-warm contrast that the nose reads as blossom.
Chain one — why the flower gives way to the fruit. Fresh plum blossom carries a faint, heat-labile fragrance → nothing in that fragrance survives a smoldering ember → a recipe built on petals would smell of nothing by the time it reached the room → the manual reaches instead for 白梅, a dried or pickled plum whose tart, slightly savory character holds through combustion → a stable fruit note becomes the recognizable plum signal. The fruit is not a compromise for lack of flowers; it is the version of “plum” that can be burned.
Chain two — how a tart note becomes a blossom. Sandalwood lays a warm, slightly creamy floor → the herbal ministers add a green, faintly bitter layer that keeps the wood from turning sweet → the plum and peony bark place a cool, tart accent above that floor → a cool note set against a warm ground reads to the nose as freshness rather than as fruit, the same trick a painter uses with colour → fennel and clove bark lift the surface so the contrast arrives immediately rather than unfolding slowly. The “plum” is a relationship between two registers, not a substance.
That second chain explains a detail a modern reader might otherwise skip. Peony bark is not a plum material and does not smell of plum; it supplies a dry, slightly astringent edge that makes the fruit read as flower rather than as pickle. Take it out and the blend still smells of plum, but of stewed plum — heavier, sweeter, less like something growing.
Fire Before Grinding, and the Sandalwood Exception
Direct answer: Roasting dries the herbs so they grind into powder instead of paste, and the one material the text pulls away from direct heat is the sandalwood — which is also the material the method seems to include.
The mechanics are straightforward. Raw botanicals hold water and oils → grinding wet material produces a clogging paste rather than a powder → dry heat drives off the water and part of the volatile load → the result grinds cleanly and burns at an even rate. Roasting also tames the harshest edges of 甘松 and 藁本, both of which are pungent enough to dominate a blend if they go in raw.
The clove bark is the stated exception, and the reason is not mysterious: clove bark is already aromatic and dry, and heat drives off the very eugenol-type volatiles that make it useful as a guide. Roast it and you spend the material for nothing. That is the logic the method follows.
The sandalwood is the harder case. The weight table marks 檀香 as 不见火 — “does not see fire” — while the method says to roast everything except the clove bark. The two instructions do not sit easily together, and I would not pretend the text resolves them. My reading, as an interpretation rather than a fact, is that 不见火 governs direct flame contact: the sandalwood is dried gently, away from an open fire, while the herbs take the harsher roast. A maker working from the text has to choose, and the choice shows up in the result — a roasted sandalwood reads drier and more resinous, an unroasted one keeps more of its creamy top. The author’s view, not a verified fact.
The Month in the Jar
Direct answer: 窨 — sealing the ground blend for a month — is the step that turns a pile of powders into one material, and it is the step most often skipped.
The rationale is equilibration rather than chemistry. Freshly ground particles each release their own volatiles at their own rate → in a closed jar those volatiles distribute through the headspace and re-adsorb across the whole mass → the sharper materials lose their isolation and the blend begins to smell like one thing → after a month, the first impression on opening the jar is the accord rather than the individual herbs. Grinding coarsely, as the text specifies, leaves enough particle structure for that exchange to happen without turning the blend into dust.
In practice the step costs nothing but patience and a labeled jar. Keep it closed, keep it out of heat and light, and resist opening it weekly to check progress — each opening vents the headspace the blend is trying to build. The one-month figure is a rule of thumb from the manual rather than a threshold where something switches on; a blend left longer will not spoil, and one opened early will simply smell more separated.
One Idea, Several Recipes
Direct answer: 梅花香 is a theme rather than a single formula, and the manual preserves several settings of it — including this one — without treating any of them as definitive.
香乘 lists 梅花香一, 梅花香二, 梅花香三, and 寿阳公主梅花香 as separate entries. Each assembles a different set of woods, herbs, and plum materials to reach the same impression, which is what a named incense was in this tradition: a recognizable idea that a maker adapted to the materials at hand. The variation is not sloppiness in the sources; it is evidence that the name described a target rather than a recipe card.
That has a practical consequence for anyone reading the manual today. Comparing two 梅花香 entries tells you which materials the tradition considered interchangeable for the plum effect — which barks, which fruits, which woods — and that comparison is more useful than hunting for the one “correct” version.
The Legend and the Craft
Direct answer: 寿阳公主 lends the recipe a name and an atmosphere; she does not lend it an author, and the distinction is worth keeping.
The story is well known. A plum blossom fell on the princess’s forehead and left a mark that became a fashion, the 梅花妆, recorded in the Song-era encyclopedia 《太平御览》. The tale explains why a plum incense would want her name attached: it places the scent at court, in winter, among people with leisure and taste. What it does not do is establish that she composed the blend, and none of the sources cited here makes that claim. The attribution is a dedication, not a byline.
My position is that the legend is the least interesting thing about the recipe and the most seductive. The interesting part is the construction — a floral impression built from fruit, bark, and wood, long before the vocabulary of accords entered Western perfumery. Treating the princess as the author would be mistaking a title for a person, and it would also hide the actual craft, which is a maker’s decision about how to make a flower out of what survives a fire. The author’s view, not a verified fact.
Accord Versus Flower: A Confusion Table
Direct answer: The recipe invites one specific mix-up — reading “plum incense” as “incense made from plum blossoms” — and the confusion resolves cleanly once you separate the note from the material.
| Question | This recipe’s answer | What a flower-only reading assumes |
|---|---|---|
| Where does the plum note come from? | 白梅 fruit plus peony bark over sandalwood | Plum blossom petals |
| Does the source material survive burning? | The fruit note does | The blossom’s volatiles do not |
| What is the note’s character? | Tart, cool, slightly savory over warm wood | Sweet, light, fleeting |
| What does the recipe prove? | A floral impression can be built | A flower is a raw material |
| What fails if you substitute petals? | Nothing recognizable reaches the room | The blend loses its signature |
The last row is the useful one. Substituting petals for the fruit does not make a more authentic plum incense; it makes a blend whose signature evaporates before the smoke reaches anyone. The word “plum” in the title describes a target impression, and the recipe reaches that target with the materials that can carry it.
FAQ
Q: What is 寿阳公主梅花香?
A classical plum-blossom compound incense from 香乘 卷十八, attributed by tradition to a Southern Dynasties princess. Nine materials go in: four herbs at half a liǎng, three at a full liǎng, a small measure of rosewood, and a hundred white plums. The method is roast, grind coarse, and seal for a month. Its plum character is an accord of fruit, bark, and wood rather than a flower extract.
Q: Does the recipe use real plum flowers?
No, and the omission is deliberate. Plum blossom fragrance is faint even on the branch and does not survive a smoldering ember, so the manual uses 白梅 — dried or pickled white plum — for the tart note and peony bark for the dry edge that makes it read as blossom. The result smells of plum because the contrast is engineered, not because a flower was included.
Q: Why does the manual keep several 梅花香 recipes?
Because the name described a target rather than a fixed product. 香乘 records 梅花香一, 二, and 三 alongside this one, each arranging different woods, herbs, and plum materials toward the same impression. Comparing them shows which materials the tradition treated as interchangeable for the effect — the practical information a maker actually needs, and something a single authoritative formula would hide.
Q: Did Princess Shouyang compose the blend?
There is no evidence that she did. 寿阳公主 is remembered for the 梅花妆 story in 《太平御览》, in which a fallen blossom left a mark that became a court fashion. The recipe borrows her name for its atmosphere, and the attribution should be read as a dedication rather than a claim of authorship. The craft in the formula stands on its own without the legend.
Q: How would I scale the formula down for a test batch?
Divide the weighed materials by the same factor and keep every ratio intact — the four half-liǎng herbs stay at half the weight of the three full-liǎng ones, and the rosewood stays at a fifth of that. The plums are the awkward part, since the text counts them rather than weighing them; halve the count, then adjust by eye until the fruit reads clearly against the wood. Scale the aging time only loosely: a smaller batch equilibrates faster, and a month remains a safe upper bound rather than a requirement.
Sources
- 香乘 (Xiang Cheng), 周嘉胄 (Ming), 卷十八 — the formula translated above. Full text at Wikisource.
- 《太平御览》 — the source of the 梅花妆 story attached to 寿阳公主.
- 本草綱目 — the materia medica describing sandalwood, angelica, peony bark, and the other materials in the blend. Full text at ctext.org.
Related Resources
- Chinese Incense Recipes (香方) — the wider recipe library and unit conventions.
- The 合香 Blending System — how 君臣佐使 shapes a compound formula.
- Sandalwood Incense Guide — the chief material in this blend, and how grades differ.
For the full library of translated classical formulas, start with Chinese Incense Recipes (香方).
Summary
- 寿阳公主梅花香 is a 香乘 formula built on sandalwood, four herbal middles, three spice-and-wood materials, a small rosewood accent, and a hundred white plums.
- The plum note is an accord — fruit plus peony bark over warm wood — not an extract of plum blossom.
- Weights follow the 君臣佐使 ladder: chief at a full liǎng, ministers at half, guides at the edges.
- Roasting dries the herbs for grinding; the sandalwood’s 不见火 note complicates that instruction and deserves a maker’s judgment.
- A month sealed in a jar lets the blend stop smelling like separate ingredients.
- The princess attribution is legend; the construction is the real subject.