Direct answer: 柏子香 (bǎi zǐ xiāng) is the most unpretentious formula in the classical Chinese canon — the kernel of the oriental arborvitae (Platycladus orientalis, 侧柏), dropped briefly into boiling water, steeped in wine and honey, then dried out of the sun — and its entire significance rests on the fact that it could be made for almost nothing.
Key facts: The Ming manual 香乘 (Xiang Cheng) writes the formula down with no measurement at all: 柏子实不计多少 — “seeds, in whatever count.” The tree in question is 侧柏, a plant that shares no genus with the cypress and juniper of other incense cultures, and the blend sits as the tradition’s single-material (or near-single-material) answer to the many-layered court 合香.
What the Manual Actually Records
Direct answer: Four actions and a seal — scald, steep, honey-seal for a week, shade-dry — and not a single weight anywhere in the entry.
The source line reads 柏子实不计多少，带青色，未开破者: gather the seeds in any amount, still green, with the husk intact. Drop them into boiling water (沸汤焯), steep them in wine, seal them in honey for seven days (蜜封七日), remove them, and dry them away from direct sun before they are burned.
| Role | Ingredient | What it contributes |
|---|---|---|
| 君 (chief) | 柏子 — the seed itself | Carries the whole fragrance; nothing else is added for scent |
| 臣 (minister) | — | Empty — there is no secondary scent material |
| 佐 (assistant) | wine + honey | Takes the hard edge off the roasted kernel and lends sweetness |
| 使 (guide) | the drying and heating | The step that actually forms the aroma |
Look at the table and the point announces itself: the chief position is filled by a tree seed, the minister slot is blank, and the two lower roles are a soak and a drying step. No 沉香, no 檀香, no musk. A formula this spare is not a fragment that lost its other ingredients — the emptiness is the recipe.
Why “No Quantity” Is the Whole Point
Direct answer: The text supplies no weight because 柏子香 was never intended to be dosed precisely — “in whatever count” is the instruction, not a lacuna.
A formula that writes 沉香三两 wants to be reproduced with its ratios intact. A formula that writes 不计多少 wants to be worked with whatever the kitchen has. The difference matters because modern readers reach for a gram scale and feel cheated when the manual declines to cooperate. There is no 两/钱 proportion hiding here, and fabricating one would corrupt the record. The faithful rendering is simply: as many green, closed seeds as you possess, scalded, wine-soaked, honey-sealed, and dried.
Across folk retellings, it is the treatment that wobbles rather than the seed quantity — some versions steep in 黄酒 rice wine, some leave out the wine and keep only the honey. That looseness is the formula’s temperament, not a flaw.
For the formulas in this cluster that do weigh their materials, one 两 (liǎng) ran near 37 g in Song–Ming practice, breaking into ten 钱 (qián) of about 3.7 g, each 钱 into ten 分 (fēn) near 0.37 g. Because the 两’s gram value shifted dynasty to dynasty, the trustworthy number is always the ratio between items, not the converted grams. The recipe library carries the full note.
The Processing, Step by Step: Two Causal Paths
Direct answer: The blanch and the honey steep are not cosmetic touches — each one sets off a physical change that turns a bland green kernel into something that smolders fragrantly.
A seed fresh off the tree is green and sealed, its oils and starches locked behind an unbroken coat. Three steps undo that, one after another.
Path one — the blanch opens, the drying closes. Boiling water scalds the seed briefly → the heat relaxes and loosens the husk → the wine and honey that follow can now soak into the kernel instead of beading off its surface → the later shade-drying drives off water and wine while the honey’s sugars remain behind. The product is a seed sweetened through its interior, not just glazed on top. Skip the blanch and the husk stays sealed, so the soak runs off the way water rolls off a waxy leaf — almost nothing gets in.
Path two — honey browns under heat, so sweetness becomes roast. Honey is mostly sugars, and sugars do not evaporate cleanly — when the ember’s heat reaches them they caramelize → caramelization yields the toasted, nutty sweetness that marks the scent → the wine contributes its own residual sugars plus a mild astringency that trims the fresh kernel’s sharpness. It is the same chemistry behind roasted nuts and the crust of bread, and it is why a dried, treated seed smells so unlike a raw one: the honey does its work at the moment of burning, not while it sits in the jar.
Neither path is spelled out in the source as a mechanism; both are a present-day reader’s reconstruction of why the old steps run in the order they do. Traditional belief, not a measured claim.
柏子 Is Not “Cypress” — a Naming Trap
Direct answer: The English word “cypress” covers three distinct plants, and 柏子香 uses the one that is, botanically, not a cypress at all.
| Label | Plant | Role in incense |
|---|---|---|
| 侧柏 (Platycladus orientalis) | Oriental arborvitae, in the cypress family | The seed of 柏子香 |
| Cypress (Cupressus spp.) | True cypress, Mediterranean and N. American | Occasionally its wood |
| Juniper (Juniperus spp.) | Juniper, the source of gin’s berry note | Sometimes berry or wood |
| 柏木 (bǎimù) | A loose “cypress/cedar” wood label in Chinese | As a wood |
The snag is linguistic rather than botanical. Chinese 柏 sweeps up a loose cluster of conifers, while English “cypress” and “juniper” name separate genera. When a recipe specifies 柏子, it means the arborvitae kernel — and reading it as juniper berry (a wholly different, sharp, gin-like material) would produce a different incense altogether. This is a naming trap, not a substitution guide.
Reading a Batch of Seeds Before You Use Them
Direct answer: Three visible cues — hue, whether the husk is closed, and surface oil — separate a workable batch of 柏子 from one the classical method never assumed.
| Cue | What to look for | What it tells you |
|---|---|---|
| Hue | Still green, not browned | A pre-ripe kernel with its oils intact |
| Husk | Sealed, no split | The coat will hold the soak properly |
| Surface | A faint natural sheen | Healthy oil, not stale or hollow |
| Age | Firm, not crumbled | Old kernels dry out and burn flat |
The single rule linking the rows: a kernel that is already brown, split, or hollow has lost the exact oils and sugars the blanch-and-steep sequence is built to work on. A green, closed seed is what the recipe assumes; anything less is a substitution the source never sanctioned.
Common Misreadings
- “柏子香 is cypress-wood incense.” It is a seed, from the arborvitae — a different part of a different plant than the cypress and juniper woods an English reader pictures.
- “The missing quantity is a defect I should patch.” The missing quantity is the recipe’s character; adding a gram count invents data the source never wrote down.
- “Honey is just a sweetener.” Honey is a processing agent that caramelizes into the roasted note during the burn — its role is chemical, not decorative.
- “The blanch can be skipped.” Without it the husk stays sealed, and the wine and honey slide off rather than soaking in.
The Author’s View
Position: 柏子香 is the most instructive formula in this cluster for a reader today, because it demonstrates that the tradition was a practice a poor scholar could afford — not a luxury reserved for the court.
Reasoning: A scent assembled from a tree kernel, wine, and honey demands no rare material, only patience and process. It stands as the honest counterweight to the “agarwood or nothing” framing that dominates English incense writing. The same tradition that produced 花蕊夫人’s opulent musk-and-agarwood court blend also produced a man roasting seeds over a quiet stove — Su Shi’s line 铜炉烧柏子，石鼎煮山药 (a bronze burner of cypress seed, a stone pot of yam) sets the incense beside a bowl of boiled yam. Both registers belong equally to the record.
The boundary condition, though, runs the other way too: the simplicity that makes 柏子香 approachable is the same simplicity that leaves anyone chasing a “complex” scent unsatisfied. It does not layer, it barely shifts as it burns, and it will not impress a collector hunting for depth. Its virtue is honesty, not richness — and that is precisely where its worth lies. It does not mean the recipe is inferior; it means it answers a different question.
The author’s reading, stated plainly — a judgment call, not a documented claim.
Summary
- 柏子香 is a single-material incense built from the oriental arborvitae kernel, not a cypress or juniper wood.
- The 香乘 records no quantity — “in whatever count” is the formula’s honest temperament, not a gap to repair.
- The blanch-and-steep sequence works because blanching opens the husk so wine and honey can enter, and honey caramelizes into the roasted note during burning.
- The name 柏 is a linguistic trap that English “cypress” does not map cleanly onto.
- The formula’s entire value is its accessibility — a practice, not a luxury.
FAQ
Q: What exactly is cypress-seed incense?
A single-material formula from the Chinese tradition: kernels of the oriental arborvitae, washed in boiling water, steeped in wine, packed in honey for roughly a week, then air-dried away from sunlight. What you get is a humble roast-and-sweet scent rather than a layered court perfume — the tradition’s quiet answer to its own expensive agarwood blends, and the reason Su Shi’s verse places it next to a plain bowl of yam.
Q: Does anything else go into it?
Essentially nothing — the kernel is the entire scent. The wine and honey are a processing soak that sweetens the roasted-seed character, not co-formulated fragrance materials. That is what separates it from a compound 合香, which stacks several ingredients under a 君臣佐使 framework of chief, ministers, assistants, and guides. 柏子香 leaves the minister slot empty; the chief is the seed, and the supporting roles are the soak and the drying. The source’s refusal to give a fixed quantity flows from the same temperament — this is the canon’s least fussy formula.
Q: Why does the recipe carry no amount?
The source itself declines to give one — 柏子实不计多少 means “seeds, in whatever count,” and that open-endedness is deliberate. Weighted formulas exist to be recreated exactly; this one exists to be made from whatever a household has on hand. Supplying a gram value would be inventing data the text never recorded, so the truthful position is to read the quantity as open and let the wine-and-honey step absorb the regional variation instead.
Q: Is 柏子 the same thing as juniper or cypress?
No — and this is the classic snag. 柏子 is the kernel of the oriental arborvitae (侧柏, Platycladus orientalis), which sits in the cypress family yet is neither a true cypress (Cupressus) nor a juniper (Juniperus). The character 柏 sweeps up a loose set of conifers, while English “cypress” and “juniper” point to separate genera. A formula naming 柏子 means the arborvitae kernel specifically — and dropping in juniper berry would hand you a wholly different, sharper, gin-like material.
Q: Could I make it myself?
Yes, in principle. Blanch green, closed seeds, steep them in wine, honey-seal for about a week, and dry them out of the sun. Two caveats shape the project. One, the result is a loose roasted grain rather than a pressed stick, so it sits over charcoal or an electric heater instead of being lit on its own. Two, since the source fixes no measure and the treatment shifts by region, you are working with a method, not a precise spec. It is the friendliest on-ramp to the classical approach for anyone with access to the kernels.
Sources
- 香乘 (Xiang Cheng), 周嘉胄 (Ming) — the 柏子香 formula. Full text on Wikisource.
- 苏轼 (Su Shi), 《十月十四日以病在告独酌》 — “铜炉烧柏子，石鼎煮山药.”
- Compendium of Materia Medica (本草綱目) on ctext.org
- Platycladus orientalis in Flora of China
Related Resources
Natural next step: see the opulent counterpoint in 花蕊夫人衙香.