Ghee Hidden in Milk: Bhavaprakasha on the Body's Last Reserve

📖 26 min read · Classical text study · Bhavaprakasha, Purva Khanda

Where This Passage Sits in the Bhavaprakasha Purva Khanda

The Bhavaprakasha is the youngest of the three books that Ayurvedic tradition groups as the Laghu Trayi, the "lesser triad". Bhavamisra compiled it in the sixteenth century, several hundred years after the works it draws on, and its character is compiler's character: it collects, it compares, it quotes its predecessors by name, and when two authorities disagree it usually says so rather than smoothing the seam over. The Purva Khanda, the first of its three divisions, is where the general theory lives — the elements, the doshas, the dhatus, the structure of the body — before the book turns to its famous materia medica.

The pages this article follows come from the stretch of the Purva Khanda that deals with the seventh and final dhatu. In the classical scheme the dhatus are the seven bodily constituents that food is progressively refined into, each one made from the one before it. Rasa becomes rakta, rakta becomes mamsa, and so on down the line until the last one. That last one is shukra: in classical usage the reproductive tissue, and simultaneously something broader, since the same texts describe it as the final refinement of everything eaten and the body's ultimate reserve. We have covered the chain itself elsewhere and will not re-run it here. What this passage does is different, and it is the reason the pages are worth an article of their own.

Bhavamisra has a problem. Most terms in his sequence correspond to something a reader can point at — mamsa is flesh, asthi is bone. The last term is not like that. It is not visible as a structure, not localised in an organ you can name, and in a large part of the population it does not appear to be present at all. A compiler with less nerve would have stated the doctrine and moved on. Bhavamisra instead lets the objection be raised, in the formal style Sanskrit scholastic writing keeps for the purpose, and then answers it.

The result is a short run of verses that is really a piece of reasoning about evidence. How do you argue that something exists when the argument cannot rest on showing it? The classical answer, repeated in four different images across these pages, is that you argue from things everyone already knows: from milk, from sugarcane, from an unopened bud, from crystal. This is what makes the passage more interesting than its subject matter suggests. It is a first-principles argument conducted entirely in kitchen objects.

One editorial note, because it affects how the citations below are written. The edition used here is the Hindi-commentary Bhavaprakasha, in which each Sanskrit verse is followed by a bhasha gloss restating it. The scan is old and its Devanagari verse numerals are in places genuinely ambiguous. Where a number is legible it is given; otherwise the citation is to the section and the transliterated line instead. Nothing here is cited to a verse that could not be read on the page.

Something crossed the water, but no water crossed it.

Shabda, Archi and Jala: Three Ways a Thing Can Travel

Before the passage reaches the final dhatu it makes a claim about movement, and the claim is more precise than it first looks. The line reads rasah sharire tridha sancharati — rasa moves through the body in three ways — and then names them by comparison: shabda-santana-vat, archih-santana-vat, jala-santana-vat. Like the propagation of sound. Like the propagation of a flame. Like the running of water.

The Sanskrit word doing the work is santana, which means a continuous succession, an unbroken series, one thing following another without a gap. It is a word about continuity rather than about transport. That distinction matters for reading the three comparisons correctly, and it is easy to miss.

Consider each case. When a sound crosses a courtyard, nothing physically crosses it: air near the source is disturbed, that disturbance disturbs the air beside it, and the pattern reaches the far wall although no parcel of air made the journey. When a flame runs along a wick, the flame at the far end is not the near flame moved along; it is a new flame lit by its neighbour. Only in the third case — water in a channel — does the same stuff actually arrive somewhere else.

So the three comparisons are not three synonyms for "flows". They are three genuinely different mechanisms, and a text that lists all three is being careful about something. The classical authors did not have a theory of wave propagation, but they had noticed the difference between a disturbance travelling and a substance travelling, and they had noticed that both look like movement from outside. Putting all three on the same list is an admission that they did not know which of the three the body was doing, and that all three were live possibilities.

There is a second reason the triad sits here rather than anywhere else. The verses that follow claim a substance is everywhere in the body at once, and a model in which nourishment moves like a stream makes that hard — a stream is somewhere in particular, and it has a front end. Sound and flame make "everywhere at once" far less strange, because in those cases there is no single place where the thing is. Intended or inherited, the triad prepares the ground for the pervasion argument.

Tikshna, Madhyama and Manda: Three Timetables for One Process

Immediately after the three santanas the passage does something a modern reader will recognise instantly: it says that the same process runs at different speeds in different people, and it ties the speed to a named variable.

The variable is agni, the digestive and metabolic fire, and the passage names three grades of it. Tikshnagni, sharp fire. Madhyamagni, middling fire. Mandagni, slow or dull fire. Each grade is then matched to one of the three propagation images from the verses just before. In a person of sharp agni, the text says, rasa moves shabda-santana-vat — quickly, like sound. In middling agni it moves like the succession of a flame, at a middle pace. In slow agni it moves jala-santana-vat, slowly, like water finding its way.

The pairing is elegant, and the Hindi commentary spells it out without embarrassment: sharp fire, quick passage; middling fire, middling passage; dull fire, slow passage. Then the passage draws the consequence that a practical reader would want, which is a duration. In the normal case, it says, rasa becomes shukra in about a month. In sharp agni it takes somewhat less than a month; in dull agni, somewhat more.

That hedge is the most interesting sentence in the section and it is easy to read past. A text that wanted to sound authoritative would have given three numbers. This one gives one number and then explicitly declines to fix the other two. It is telling the reader which part of the claim it stands behind and which part it does not, which is a distinction most premodern technical writing does not bother to draw.

It is also worth being clear about what the month is and is not. It is a claim internal to a historical model of physiology, made without any of the instruments that would be needed to check it. It is not a finding, it is not a measurement, and it does not describe anything that has been established about how the human body works. Its interest is in what it shows about the model — that the model was quantitative enough to produce a number, and honest enough to flag which numbers were soft. Where the same book handles the fire itself in detail, we have followed that separately in the Bhavaprakasha account of jatharagni.

The fragrance is already in there. What is missing is the opening.

Mukula-Gandha: The Objection About Children, and the Answer From a Bud

Here the passage stops moving forward and turns to face a difficulty. The edition marks it in the way Sanskrit scholastic texts conventionally do, with a bracketed prose line introducing the objection before the verse that answers it. The line reads: nanu balanam katham shukram na drishyata ityaha — "But how is it that shukra is not seen in children? To this he says…"

The word nanu is a formal marker. It introduces a purvapaksha, the position of an opponent, which the author states in its strongest form precisely so that answering it will count for something. That the compiler included it at all is the point. He could have stated the doctrine flatly. Instead he lets a reader interrupt.

The objection is a good one, and it is empirical. The doctrine says every person is built by a chain of seven transformations ending in this substance. Children are people. In children the substance is not in evidence. Either the doctrine is wrong, or something is missing from the account.

The answer comes in two verses and it does not dispute the observation at all. It concedes it and then reframes what the observation is evidence of. The first line runs balanam shukram asty eva kila saukshmyan na drishyate — in children it certainly does exist; by reason of its subtlety it is not seen. Then the image: mukule gandho yatha sann api nopalabhyate — as the fragrance in a bud, though present, is not perceived. And the completion in the following verse: kusumanam praphullanam gandhah pradurbhaved yatha — as the fragrance of flowers that have opened becomes manifest.

What makes this hold up as reasoning, rather than as decoration, is that the analogy is checkable. Anyone can test the bud claim. Take an unopened jasmine bud, crush it, and the scent is there. It was there before the flower opened; the opening did not create it, it released it. So the objector is offered not an assertion but a procedure: here is a case you can verify for yourself in which a thing is genuinely present and genuinely imperceptible, and the reason for the imperceptibility is saukshmya, fineness — the same reason being offered in the disputed case.

That is a real argument. It does not prove the doctrine; an analogy never does, and a modern reader should not pretend otherwise. What it does is shift the burden. Once one uncontroversial case of present-but-imperceptible is on the table, "I cannot see it" stops being decisive on its own. The objector has to say more.

The choice of image is worth a moment too. Of everything available to a sixteenth-century compiler that is present-but-hidden — a seed in the ground, a fire in the flint, salt dissolved in water — he picks the one that is about timing. A bud is not merely a container with something concealed inside it. It is a thing on a schedule, which will open when it is ready and not before. Applied to the case at hand, the analogy carries a claim about development, not just about concealment: what is absent from the child's account is not the substance but the season.

The verse just before this run gives the other half of the practical picture — a short list of substances the tradition associated with generation and with discharge: dugdham (milk), masha (black gram), majja (marrow), phala-majja (the pulp or kernel of fruits). It is a list of the richest, heaviest, most unctuous things in a premodern larder, and it belongs to the internal logic of a system in which like builds like. We are not reproducing that list as anything to act on, for reasons set out at the end of this article.

One of these was in the other, and no amount of looking would have found it.

Yatha Payasi Sarpih: Ghee Hidden in Milk

A few verses later the text does something that looks at first like a small housekeeping move and is in fact the centre of the whole passage. The bracketed prose heading announces the topic: atha shukrasya sthanam aha — "now he states the location". The reader is being promised a place. An organ, a vessel, a region.

What arrives instead is this:

The question was "where is it". The answer is sakale kaye — in the whole body. Which is, strictly speaking, a refusal to answer the question as asked, and the refusal is the doctrine. There is no place, because being in a place is the wrong category for this kind of thing.

Both halves of the comparison are chosen with care. Ghee in milk is not ghee sitting in a layer at the top of the milk waiting to be lifted off; that would be cream, and the analogy would collapse into simple concealment. Ghee is present in milk as a potential distributed through the whole of it. Every spoonful of the milk has it, no spoonful is the ghee, and nothing you can do by looking will separate them. The same holds for the cane: the sweetness is not stored in a cavity somewhere in the stalk. It is in the fibre, everywhere along it, and the only way to get at it is to destroy the stalk as a stalk.

There is a nice textual wrinkle in the second half. The Sanskrit word is gudha, "hidden" — gudhash cekshu-raso yatha, "as the juice hidden [in the sugarcane]". The Hindi commentator, reading the same syllables, glosses the line with guda, jaggery: as jaggery is latent in cane juice. The two readings are almost the same sound and they make the same point one step apart, one about juice hidden in the cane and one about the solid sugar hidden in the juice. Whether the commentator was punning deliberately or reading a variant, the doubled image strengthens rather than muddles the argument: at every stage of the cane, something is present that you cannot see and can only get out by work.

And that is the second thing the verse is doing. Both similes are about substances revealed by process. Milk does not yield ghee by standing still; it is soured, churned, separated, clarified over heat. Cane does not yield jaggery by being looked at. In both cases the hidden thing appears only at the end of a sequence of operations performed on the entire quantity — not on a part of it, because no part of it is the thing.

The verses immediately following push exactly this point, using the churn. In a person with a great deal of shukra, the passage says, a little churning yields ghee; in a person with little of it, much churning yields only cane juice. Read as physiology it is unremarkable. Read as a piece of reasoning about evidence it is rather sharp: the proposal is that you can infer how much of a distributed, invisible thing is present by how much work it takes to bring any of it out. That is an inference from yield — a way of measuring something you cannot see by measuring what it costs to get at it.

Asked what it looked like, the text reached for quartz.

Sphatikabham: Describing What You Can Only Compare

Just before the pervasion verse the text tries a different tactic, and the failure of that tactic is instructive. It attempts a straight description.

The verse gives five qualities: sphatikabham (crystal-like in appearance), dravam (liquid, flowing), snigdham (unctuous), madhuram (sweet), madhu-gandhi ca (and honey-scented). Then it does something unusual for a doctrinal text: it reports disagreement. Shukram icchanti kecit tu taila-kshaudra-nibham ca tat — but some hold it to be like taila (oil) and kshaudra (honey).

Look at what all five terms have in common. Every single one is a comparison. Crystal-like. Oil-like. Honey-like. Honey-scented. Even snigdha, which is the standard Ayurvedic quality-term for unctuousness, is defined across the literature by reference to oil and ghee. There is not one property in the list that is stated in its own right rather than borrowed from something else, and the "some say" clause shows that the borrowing was contested even among the authorities the compiler had in front of him.

This is the same intellectual situation as the bud and the milk, appearing in a different dress. When a tradition cannot inspect a thing directly, its descriptive vocabulary has to be assembled out of things it can inspect. Crystal for the look of it, oil for the feel, honey for the smell and the consistency. What you end up with is not a description so much as a set of coordinates in a space of familiar materials — here, in the region bounded by quartz, sesame oil and honey.

Notice too what is missing: no quantity, no location, no mechanism. A description built entirely from comparison can convey texture but cannot convey measure, and this one does not pretend to.

Preserving the disagreement rather than choosing a side is characteristic of Bhavamisra, and it is one of the reasons the book is useful to historians — it carries a partial record of what was still contested in the sixteenth century. Readers who want the wider picture will find it in our overview of the Bhavaprakasha and its author.

A layer that covers the whole of a thing has no particular place on it.

Vriddha Vagbhata's Saptami Kala: A Layer With No Address

Having made the pervasion claim in his own voice, the compiler does what compilers do: he calls a witness. The edition marks the quotation plainly — vriddha-vagbhato 'py aha, "the elder Vagbhata also says" — and the attribution is worth getting right. "Vriddha Vagbhata", the elder Vagbhata, conventionally refers to the Ashtanga Sangraha rather than to the more widely read Ashtanga Hridaya. Bhavamisra is reaching for the older and more detailed of the two works.

The quoted passage concerns the kalas. In classical anatomy a kala is a membrane or investing layer, a boundary between a structure and what it contains; the standard scheme counts seven of them, each named for what it holds. The seventh is the one at issue here: saptami shukra-dhara, the seventh, the shukra-holding kala.

What the quotation then says about it is the strangest sentence in the whole passage, and it is strange in a productive way. It gives the kala a location — a finger's measure on the right side, below the opening of the bladder, associated with the urinary passage — and in the same breath calls it sakala-sharira-vyapini: pervading the entire body.

A modern reader's first instinct is to call this sloppy. It repays a second look. The two clauses are answering two different questions that the classical framework does not regard as one question. Where does the thing emerge, where can it be found leaving? That has an answer, and it is anatomical. Where does the thing exist? That, on this model, has no anatomical answer at all, because existence and emergence are not the same relation. Compare the ghee: the place where ghee comes out of the churn is a particular place; the place where ghee is in the milk is nowhere in particular, which is to say everywhere.

Seen that way the two clauses are not in tension. They are the pervasion doctrine and the outlet, side by side, and the passage is being precise rather than careless. It is the same distinction the ghee simile drew three verses earlier, restated in the vocabulary of the kalas and backed by an older authority.

There is a small point of scholarly hygiene here too. Bhavamisra could have paraphrased the elder Vagbhata and let the claim stand as his own. He quotes instead, names his source, and marks where the quotation ends — which is what lets a reader five centuries later see that this idea did not originate with him.

The same bowl. Only drier, and it took years.

Vardhake: The One Verse About Growing Old

Set among these verses about latency and pervasion is a single line about age, and it is the most quietly consequential sentence in the passage. It appears as the answer to another short objection — if food produces rasa and rasa builds the dhatus, why does an old person eating the same food not build the same tissue?

Three things are packed into two lines. First, an observation that needed no theory to make: older bodies do not build tissue the way younger ones do, on identical food. Second, a proposed mechanism, entirely internal to the classical framework — vayu increases with age, vayu is characterised as dry, and its increase dries out the rasa, so there is less of the moist raw material the chain runs on. Third, and this is the move worth noticing, a conclusion drawn in the opposite direction from the obvious one.

The obvious conclusion, given a doctrine that says the chain ends in a reserve, would be to attend to the end of the chain. The verse does not say that. It says tatra anilam jayet — there, conquer the vayu. The recommendation is aimed at the thing doing the drying, which sits at the beginning of the causal story, not at the depleted item at the end of it. The Hindi commentary follows the same line, glossing it as attending to the vayu.

Whether or not one accepts the physiology — and it is a historical model of the body, not a description of how ageing works — the reasoning has a recognisable shape. Identify what is being lost, then ask what is causing the loss, and address that rather than the loss. It is the difference between refilling a vessel and looking for the crack.

It also closes the arc the passage has been drawing all along. The bud verses said the substance is latent in childhood and manifest in youth. This verse says the conditions that made it manifest do not hold indefinitely, and names what changes. Latent, manifest, and then a slow drying of the medium the whole chain depends on: childhood, youth and age, each given its own line. For the classical account of vayu in its own right, we have covered it separately in the five pranas and vayus and in the tridosha guide.

Vapuh Saro, Jivasya Ashraya: The Claim Underneath

The definition verse in this run is short and it is doing more than defining. It reads: shukram saumyam sitam snigdham bala-pushti-karam smritam | garbha-bijam vapuh saro jivasya ashraya uttamah || — it is saumya, white, unctuous, held to produce bala and pushti; it is the seed of the embryo, the sara of the body, and the supreme ashraya of the jiva.

Four separate claims are stacked in a single line, and they are of quite different kinds. That it is white and unctuous is description. That it produces bala (strength) and pushti (nourishment, in the sense of being well-made) is functional. That it is the seed of the embryo is reproductive. And then vapuh saro — the sara of the body.

Sara is a rich word and translating it as "essence" loses most of it. It is the same word used for the heartwood of a tree, the strongest and densest part left when the soft outer wood is stripped away; for the cream of milk; for the pith of anything. It carries the sense of the best part, the concentrated part, what remains when everything removable has been removed. To call this dhatu the sara of the body is to say it is what the body comes to when it has finished refining.

The last clause goes further still. Jivasya ashraya uttamah — the supreme seat, or support, or resting place of the jiva, the living principle. And the following verse qualifies it in a way that keeps the doctrine coherent: the jiva dwells in the whole body, it says, but especially there.

It is worth pausing on how large a claim this is, and how far it travels from anything that could be checked. The passage began with a question about timing and speed, moved to a question about evidence, and has arrived at a statement about where life is seated in a body. Nothing in the argument to this point supports the last step. The bud analogy establishes that imperceptible things can be present; it says nothing about the jiva. This is where the passage stops reasoning and starts asserting the framework it inherited, and a reader is entitled to mark the transition. The interesting thing is that the transition is visible at all — the text does not smuggle it in, it simply moves from one register to the other and carries on.

How to Read a Chapter That Argues From the Kitchen

Everything load-bearing in these pages is a household object. A bud. A bowl of milk. A stalk of cane. A churn. A piece of crystal. A lamp with a flame running along the wick. Water in a channel. There is not a single instrument anywhere in the passage, and no observation that required one.

That is not a limitation the text is hiding. It is the method. Without magnification, without chemistry and without any way to isolate what is under discussion, the available move is to find something in ordinary experience that has the property you want to argue for, show the property is real there, and propose it holds in the case you cannot inspect. The bud does that for imperceptibility, the milk for pervasion, the churn for inference from yield.

This is worth saying plainly because it cuts both ways. It is a genuine intellectual method and it produced arguments with real structure — an objection stated fairly, an analogy that can be checked, a hedge on the figures the author was not sure of, a source named rather than absorbed. It is also a method with a hard ceiling. An analogy can show that a claim is not absurd. It cannot show that the claim is true. Milk really does hold ghee throughout; that fact establishes nothing whatsoever about human tissue, and no accumulation of well-chosen similes would.

A word on what this article deliberately leaves alone. These pages sit next to material we have written up separately, and the neighbouring subjects are not re-run here. The seven-dhatu sequence itself is set out in our guide to the saptadhatu, and the middle of that chain in the Bhavaprakasha account of mamsa, meda and asthi. The classical vajikarana category — the preparations these verses gesture at, which we do not reproduce — is taken up as a textual subject in our study of Charaka's Chikitsa Sthana chapter 2. The Bhavaprakasha material on the monthly cycle is in rajasvala paricharya, and the embryological verses that follow this passage in garbhotpatti.

So the honest way to read these eight pages is as intellectual history: a record of how careful people reasoned about the invisible with the tools they had. This is a historical model of the body and not a description of human physiology, which is why nothing in this article is medical advice, nothing here describes anything a reader should do, and anyone with a question about their own health should consult a qualified doctor rather than a sixteenth-century compendium. Read that way the passage is genuinely impressive. Read as physiology it is simply out of date, and pretending otherwise does the text no favours.

What Is on Our Shelf, and Why It Is Here

A note on why there is a shelf section at all on an article about a sixteenth-century text. We sell none of what these verses describe, and we are not going to pretend the passage recommends anything we stock. What connects the two is one structural idea, and only that idea.

The verse at the centre of this passage is about a property belonging to the whole of a quantity rather than to any part of it you could point at. Ghee is not in a corner of the milk. Sweetness is not in a cavity in the cane. Three things we make have the same plain physical character, and that is a fact about how they are mixed, nothing more. None of these is a medicine, and none is offered as one. Each is described below only by its physical state, its size and how its material is distributed — and every one of those is something you can check rather than take on trust, by cutting, scooping or pouring.

More to Read on This Topic

Three short visual stories on related ground — all of them about what materials actually are, which is the question this chapter keeps circling:

• Earth Day: India's 5,000-Year Zero-Waste Beauty
• Going Chemical-Free Is Impossible (But This Isn't)
• Your Toner Has 12 Chemicals

Questions People Ask About This Passage