Sandhi Sharira in the Bhavaprakasha: 210 Joints, Eight Shapes, and the Sentence Where the Counting Stops

📖 24 min read · Classical Ayurveda texts · Bhavaprakasha series

Sandhi Sharira: What the Bhavaprakasha Counts After the Bones

If you have searched for sandhi sharira or for how many joints the body has according to Ayurveda, here is the direct answer before anything else: the Bhavaprakasha, Purva Khanda gives the number as two hundred and ten. It divides those into sixty-eight in the limbs, fifty-nine in the trunk, and eighty-three in the region above the neck. It then classifies every join in the body into eight named shapes. And it immediately adds that this figure covers the joins of the bones alone — that the joins of muscle, ligament and channel have no count at all.

Sandhi is a plain Sanskrit word for a junction. It is the same word used for the meeting of two seasons, and for the joining of sounds in grammar. Applied to the body it means the place where one part meets another and movement becomes possible. Sandhi Sharira — the anatomy of the joins — is the section of classical anatomy that deals with them.

The Bhavaprakasha is a sixteenth-century compendium by Bhava Mishra, and it is not an original work of dissection. It is a synthesis: Bhava Mishra collects the anatomy of the earlier Samhitas, arranges it, and has it followed by a Hindi bhasha commentary that unpacks each verse for a reader who is not going to sit with a teacher. We have written a full introduction to the text and its author in our guide to the Bhavaprakasha Samhita and Bhava Mishra’s place in the Laghu Trayi.

What makes this particular stretch of the Purva Khanda worth a slow read is not that it produces a number. Plenty of texts produce numbers. It is that within about a dozen verses the text produces a number, qualifies it, admits the limit of the method that produced it, and then switches to a completely different method for the next structure. That is a sequence of intellectual moves, not a table.

A join is the meeting of two distinct things.

Two Hundred and Ten: The Total, and Where It Comes From

The verse that carries the total is compact. Rendered in transliteration it reads kathita dehinam dehe sandhayo dve shate dasha — “in the bodies of embodied beings the joins are declared to be two hundred and ten.” The figure arrives as a Sanskrit number-phrase, dve shate dasha, literally “two hundreds, ten”.

That matters more than it looks: classical texts state quantities as words, not digits, and a word survives copying in a way a numeral does not. We come back to it in the section on how we are reading the numbers.

Two hundred and ten is not the Bhavaprakasha’s own discovery. It is the figure of the Sushruta Samhita, Sharira Sthana, chapter 5 — the chapter that enumerates the parts of the body — and Bhava Mishra is reproducing it, as he reproduces most of the anatomy, with the commentary doing the work of explanation. The Sushruta Samhita is the surgical tradition, and it is the tradition with the strongest reason to have counted: a surgeon needs to know what is under the blade. Our guide to the Sushruta Samhita as a surgical text covers how that knowledge was gathered.

The companion figure, the count of the bones themselves, is the more famous one and it belongs to a different article. The Sushruta Samhita reaches three hundred where the Charaka Samhita, Sharira Sthana, chapter 7 reaches a different total, and the reasons for the gap are genuinely interesting. We set them out in our article on why the classical enumeration counts three hundred bones. This article is about what happens between them.

Shakha, Kostha and Above the Griva: How the Total Divides

The text does not leave two hundred and ten as a lump. Three verses divide it across three regions of the body, and the three regional figures are given, again, as number-words:

Sixty-eight and fifty-nine and eighty-three come to two hundred and ten. The arithmetic closes, which is the first sign that we are reading a transmitted enumeration rather than a rounded impression.

Shakha means branch, the standard classical word for a limb. Kostha is the cavity. The third region is given not as a noun but as a boundary: everything urdhva jatru, above the collarbone. That three-way partition recurs across the anatomy wherever something is being counted.

The distribution is worth pausing on, because it is counter-intuitive. The region above the neck — by volume the smallest of the three — carries the largest share of the joins, more than the whole trunk. That is not an error. It falls out of a decision the classical anatomists made and modern anatomy makes differently: the sockets of the teeth are counted as joins, and there are thirty-two of them. The commentary says so explicitly, listing the roots of the teeth among the joins above the neck.

Kora: the class that swings on one plane.

Seventeen to a Limb: The One Sum You Can Check Yourself

The limbs are where the enumeration becomes fully transparent, and it is the one place a reader can verify the text’s arithmetic without trusting anybody. The commentary sets it out step by step.

Each of the digits carries three joins, except the thumb and great toe, which carry fewer — and the total across the digits of one limb comes to fourteen. Beyond the digits, in the leg, the text names the ankle (gulpha), the knee (janu) and the groin (vankshana), one each. Fourteen and three is seventeen. The commentary states the figure directly: seventeen in one limb.

Seventeen, taken four times, is sixty-eight. That is exactly the shakha figure the verse gave. The text is showing its working — and for the limbs, at least, the working is complete and it closes.

The arm is given by substitution rather than repetition: it is to be understood as the leg was described, the wrist standing where the ankle stood and the elbow where the knee stood. That is a compression device, not an omission — a text transmitted by memory does not repeat a list it can tell you to reuse.

Ashtavidha Sandhi: Eight Shapes, Named After Ordinary Objects

Counting the joins is the smaller half of the passage. The larger half classifies them. The text names eight kinds of sandhi, and the list is one of the most quietly brilliant pieces of teaching in classical anatomy.

In transliteration the verse runs: kora, ulukhala, samudga, pratara, tunnasevani, vayasatunda, mandala cha shankhavarta — ashta sandhayah. Eight joins.

Look at what those eight words actually are. Not one of them is an anatomical term. Every single one is the name of an ordinary object from a sixteenth-century household or workshop:

This is a taxonomy built entirely out of things a student could pick up. A mortar was in every kitchen. A casket was on every shelf. Everyone had seen a stitched seam and a crow. The text is not asking you to memorise eight abstractions; it is asking you to recognise eight shapes you already know, and then find them in a body.

A name like ulukhala carries its own definition. You cannot forget what a mortar joint looks like once you have been told the word, because the word is the picture.

Ulukhala. The name is the picture.

Kora, Ulukhala and Samudga: Hinge, Mortar and Casket

The commentary assigns locations to each of the eight, and the first three carry most of the body’s movement between them.

Kora — the hinge. The text places kora joins at the digits, the wrist (manibandha), the ankle (gulpha), the knee (janu) and the elbow (kurpara). Read that list again: it is precisely the set of joins that swing in one plane and nowhere else. The classification is doing real descriptive work — these are grouped together because they behave alike, not because they sit near each other.

Ulukhala — the mortar. Placed at the armpit (kaksha), the groin (vankshana) and the teeth (danta). The image is exact: a rounded end seated in a hollow that holds it while allowing it to turn. That the teeth appear in this group is the clearest possible statement that the socket of a tooth was understood as a join of the same family as the hip, differing in size and not in kind.

Samudga — the casket. Placed at the shoulder socket (amsapitha), the guda, the bhaga and the buttock region (nitamba). A samudga is a small box whose two halves close on one another, and the joins in this group are the ones where a shallow cup receives a rounded surface without the deep seating of the mortar.

Pratara, Tunnasevani and Vayasatunda: Plank, Seam and Beak

Pratara — the plank. Assigned to the neck (griva) and the spine (prishtha vamsha). A pratara is a flat board used for crossing, and the image is of broad flat surfaces sliding a little against one another. Each individual movement is small; the sum along a column is not. That is a fair description of what a spine does, arrived at by looking at the shape of the contact rather than at the motion.

Tunnasevani — the stitched seam. Assigned to the head (shira) and the pelvic bones (kati kapala). Tunnasevani means, quite literally, a tailor’s seam — the line where two pieces of cloth are sewn edge to edge. It is the name for the joins of the skull, and it is the one class in the eight defined by the fact that it does not move. A category of joins that hold rather than bend is a real analytical insight: it says the definition of sandhi is junction, and movement is a property some junctions have and others do not.

Vayasatunda — the crow’s beak. Assigned to the jaw (hanu), on both sides. The name is the most specific of the eight and the most visual: a beaked process that grips its opposite. Anyone who has watched a crow has the image.

Tunnasevani: a join whose whole job is to hold.

Mandala and Shankhavarta: The Ring and the Spiral

The last two classes are the ones that push the definition of sandhi furthest, and they are the reason the passage ends the way it does.

Mandala — the ring. Assigned to the throat (kantha), the hridaya, the kloma, and the channels: annular structures, rings of material closing on themselves. Note what has happened — the classification is now being applied to soft structures. The category has stretched past bone.

Shankhavarta — the conch-spiral. Assigned to the ear (shrotra) and the shringataka. A conch turns inward on itself in a coil, and the interior of the ear does the same. It is the one name in the list taken from a natural object rather than a manufactured one, and it is doing the hardest descriptive job: naming a shape that has no straight line in it anywhere.

By the time the list closes it has covered hinges and sockets and seams and rings and spirals, in bone and cartilage and channel wall. It is a comprehensive scheme. And the moment it finishes, the text turns round and tells you where the scheme runs out.

The Sentence Where the Counting Stops

This is the passage worth the whole article. Immediately after the eight classes and their locations, one verse does something that texts of this kind very rarely do. In transliteration:

asthnam tu sandhayo hyete kevalah samudahritah —
peshi snayu shiranam tu sandhi sankhya na vidyate.

“These that have been declared are the joins of the bones only. Of the muscle, the ligament and the channel, no count of joins exists.”

Read what that verse concedes. Everything preceding it — the two hundred and ten, the sixty-eight and fifty-nine and eighty-three, the seventeen to a limb, the eight classes with their locations — applies to one tissue. For the others, the text does not offer an estimate. It does not offer a range. It does not say “many”. It says the number is not known.

That is a statement about the limits of a method, made inside the text that is using the method, on the same page. The enumeration is not presented as the complete truth about the body. It is presented as what this particular way of looking can deliver, with the boundary marked.

There is a second reason the concession is intellectually serious. The tissues excluded — peshi, snayu, sira — are precisely the ones that are continuous. Bones are discrete objects: where two of them meet you can point at the meeting. Muscle grades into tendon and tendon into its attachment, and channels branch without ever fully separating. On a continuous structure the question “how many joins?” has no answer, because a join was defined as the meeting of two distinct things and there are no two distinct things. The text has not failed to count. It has noticed that the question changes shape.

These same tissues are enumerated elsewhere by quantity rather than by junction — the muscles have their own total in the same compendium, which we cover in our article on the Bhavaprakasha’s count of the muscles and the tissues of the trunk. Countable as objects; not countable as junctions. The distinction is the text’s, not ours.

Where counting stopped, the method changed to branching.

Sira Sharira: Seven Hundred Channels from Forty Roots

Having declared the channels uncountable as joins, the text turns to them directly — atha shiram aha, “now he states the siras” — and counts them a completely different way.

The figure is seven hundred. But the number is not reached by adding up regions, the way the joins were. It is reached by division from a root, and that change of method is the point.

The text begins with forty mula siras — root channels. Those forty are themselves four groups of ten, sorted by what they carry:

One hundred and seventy-five taken four times is seven hundred. The arithmetic closes again, and it closes on a structure that is branching rather than additive. You do not walk the body region by region adding as you go. You start at a root and follow it as it divides.

The same fourfold sorting — three doshas plus rakta — is the standard classical grouping for the channels. The Sushruta Samhita, Sharira Sthana, chapter 7 is the dedicated chapter on the siras and gives the same seven hundred with the same forty roots; the Bhavaprakasha is transmitting it. For the three doshas themselves as a framework, our guide to tridosha in the Ashtanga Hridaya sets out the categories this sorting depends on.

The Nabhi as Hub: How Bhavamisra Describes What He Cannot Count

Where do the forty roots begin? The text is unambiguous: at the nabhi, the navel. Nabhyam sarvani baddhah tah pratanvanti samantatah — “bound at the navel, they spread out in every direction.”

And here the passage does something that anyone who writes for a living should stop and admire. Having just admitted that this structure resists counting, the text reaches for images instead, one after another, in the space of a few verses.

The garden and the field. The whole body, the text says, is nourished by these channels pranalibhir iva aramah, kulyabhih kshetra dhanyavat — as a garden is by its water channels, as a field of grain is by its irrigation canals. The commentary spells the simile out: as a field is irrigated by a large channel and then by the small channels cut from it.

This is the same irrigation image the compendium uses elsewhere for how nourishment reaches the tissues in sequence — a comparison with a formal name in the classical literature, which we set out in full in our article on how the Bhavaprakasha describes the nourishment of the tissues. Here the same figure is being used for structure rather than for sequence.

The leaf. Yatha druma dale sakshat drishyante patatah shirah — as the veins are plainly seen in the leaf of a tree, so the channels are spread through the body of an embodied being. It is an observation anyone can repeat. Hold a leaf to the light and you have the diagram; no dissection required, and no authority to take on trust.

The wheel. Shirabhir avrita nabhih chakra nabhir iva arakaih — the navel is surrounded by its channels as the hub of a wheel is surrounded by its spokes. The pun is in the Sanskrit and survives translation: nabhi is the word for the navel and for the hub of a wheel. The image is not decorative; it is doing the work of a definition.

The text also states what the channels are for, in functional terms: prasarana akunchana adi kriyabhih — by the actions of extending and bending and so forth, continually, in the body, it is the channels that serve. Extension and flexion, named as the ordinary business of a living body.

One Hundred and Seventy-Five Each: The Arithmetic of the Vatavaha Siras

The text does not stop at seven hundred. It takes one of the four groups — the vatavaha channels — and distributes its hundred and seventy-five across the body. This is the passage that shows the enumeration is a genuine transmitted scheme rather than a round figure, because it, too, closes exactly.

One hundred, thirty-four and forty-one together make one hundred and seventy-five. And each of the two composite figures is itself broken down further. The forty-one above the collarbone is given as: fourteen in the neck, four in the ears, nine in the tongue, six in the nose, eight in the eyes. That sums to forty-one.

The thirty-four in the trunk is given as: eight in the guda and medhra region, two on each flank, six in the back, six in the abdomen, ten in the chest. That sums to thirty-four.

So the scheme closes at three levels: the sub-totals sum to the regional figures, the regional figures sum to a hundred and seventy-five, and a hundred and seventy-five taken four times gives the seven hundred. A figure invented for effect does not survive that kind of checking. This one does.

The largest single allocation above the collarbone goes to the tongue, at nine. For a tradition in which examination of the tongue is a routine part of assessment, an anatomy that gives the tongue the densest channel supply in the head is at least internally consistent with its own practice. The vatavaha grouping also connects to the classical account of the movements of vata, which we cover in our guide to the five vayus and their seats.

The comparison the text reaches for instead of a number.

A Note on the Numbers: Why We Quote Words, Not Digits

Every figure in this article comes from a Sanskrit number-word in the verse text, not from a digit printed on the page. That is a deliberate editorial decision and it is worth explaining, because it is the difference between a citation you can rely on and one you cannot.

The edition we are reading is a scan of a printed Devanagari text with a Hindi bhasha commentary. Printed Devanagari digits in a scan of that age are frequently ambiguous — a worn or thickened glyph can leave two readings genuinely open, and a single misread digit silently changes a claim.

Number-words do not have this problem. Dve shate dasha, ashtashashti, trishiti, saptashatani, chatvarimshat, panchavimshati — these are whole words with distinctive shapes and no near-neighbours. Where the verse states a quantity as a word, we have used it. Where a figure appears in this passage only as a marginal or inline numeral, we have left it out rather than transcribe a glyph we cannot read with certainty.

The same applies to verse numbering. We cite this passage at section level — Bhavaprakasha, Purva Khanda — and quote the transliterated line, rather than attaching a verse number we would be guessing at. A quoted line can be located by anyone with the text. A wrong verse number looks authoritative and sends a reader to the wrong place.

The Same Enumeration Elsewhere in the Classical Library

This passage is a node in a network, not an isolated curiosity. Four other places in the classical library carry parts of the same material, and reading them together is what turns a list of numbers into a picture of how the tradition thought.

The Sushruta Samhita, Sharira Sthana, chapter 5 is the source of the enumeration — the great counting chapter, which gives the totals for bone, join, muscle, ligament and channel in one place. Sharira Sthana, chapter 7 of the same text is the dedicated treatment of the siras, where the forty roots and the seven hundred are set out at length.

The Charaka Samhita, Sharira Sthana, chapter 7 is the parallel enumeration in the physicians’ tradition, and its totals do not all agree with Sushruta’s. The disagreement is substantive and well documented, and we deal with it where it belongs, in the article on the classical count of the bones.

The Ashtanga Hridaya, Sharira Sthana, chapter 3 carries Vagbhata’s compressed version of the body’s structural enumeration, in the characteristic style that made the Ashtanga Hridaya the most memorised of the three great compendia.

And the material immediately preceding this passage in the Bhavaprakasha itself is the account of the hundred and seven points where the structures of the body converge — a subject with its own classification and its own logic, which we cover separately in our article on the Bhavaprakasha’s treatment of the marma. The passage in this article picks up directly where that one ends.

Two more articles give the wider frame. The siras as carriers connect to the classical account of the great vessels and the seat of ojas in Charaka’s Sutrasthana 30; and the tissues these channels feed are the subject of our guide to the seven dhatus. For the philosophical frame that governs all of it — the body as a whole made of the same stuff as the world — see Sushruta’s opening chapter of the Sharira Sthana. And for where these books sit relative to each other, our overview of the Charaka Samhita’s structure and our account of how the Bhavaprakasha narrates the descent of the tradition both set the scene. The final chapter of the Sushruta Samhita makes a strikingly similar move to this one — producing an exact count and then stating its limits in the following verse.

Reading Classical Anatomy Honestly — and What We Leave Out

It is tempting to score this passage against a modern anatomy textbook and hand out marks. That is the least interesting thing to do with it, and it usually produces a wrong answer in both directions.

The figures differ from modern ones because the categories differ. A tradition that counts the socket of every tooth as a join will reach a higher total above the neck than one that does not. A tradition that includes cartilage where a modern count does not will differ again. These are decisions about what counts as a thing, and once you know the decisions, the numbers stop being mysterious. They are not a worse answer to the modern question; they are an answer to a different question.

What deserves respect here is the method, in three features. The enumeration is internally checkable — sub-totals sum to totals at three separate levels, which is not something a fabricated figure does. It is self-limiting. And where counting fails it reaches for images and says so, rather than manufacturing a number to fill the gap.

It is equally important to be clear about what this article is not doing. We have described a sixteenth-century compendium’s account of the structure of the body. We have not connected any of it to any modern condition, and we have not reproduced any therapeutic material, because none of that is what this passage is about and none of it belongs on a page like this one. The classical texts placed the management of a body in the hands of a trained physician, and that is where it stays.

Kalpana: Three Classical Forms on an Ordinary Shelf

There is no product anywhere in this passage. It is an enumeration and it recommends nothing at all. What the classical pharmacy does have, quite separately, is a vocabulary for the form a substance is prepared in — kalpana. Our shelf holds three ordinary daily-care items whose form carries one of those classical names. That is the entire connection, and we would rather state it narrowly than dress it up.

These are everyday personal-care items for teeth, hair and skin. Nothing below is a claim of any kind about any of them.

And the column we have to leave empty. The pages described above set out a classification and nothing else. They name no substance, no preparation and no practice, so there is nothing on our shelf that answers to them, and there should not be. Everything above is something applied externally. A classification is not a shopping list, and we are not going to make it into one.

Related Stories — More to Read

Frequently Asked Questions

Shop All Ayurveda Hub Products →