Graha Chikitsa in the Sushruta Samhita: One Protocol, Repeated Four Times (Uttara Tantra 28-31)

📖 27 min read

Where these four chapters sit in the Uttara Tantra

The Uttara Tantra is the last and longest division of the Sushruta Samhita, and it is the part of the compendium that moves through the specialities. It opens with the eye. It works through the ear, the nose, the head. Then, at chapter 27, it turns to the illnesses of very young children and to the nine figures the tradition calls the Bala Grahas.

Chapter 27 is a description chapter. It names the nine Grahas and sets down the signs by which each was said to be recognised, and we read it closely in an earlier article on the nine Bala Grahas of chapter 27. It ends on the printed page with the line that closes every chapter of the text: Thus ends the twenty-seventh chapter of the Uttara Tantra in the Sushruta Samhita which deals with the specific features of an attack by the nine Grahas.

What follows is what this article is about. Chapters 28, 29, 30 and 31 are the Pratishedha chapters — the therapeutics. One chapter each, in order, for Skanda-Graha, Skandapasmara, Shakuni and Revati. They are short. All four fit inside eight pages of Kaviraj Kunja Lal Bhishagratna's English translation, published as Volume III of his three-volume Sushruta Samhita in Calcutta in 1916, which is the edition quoted throughout here. Chapter 32 then begins again with Putana, and the sequence carries on through the remaining Grahas.

It is worth knowing how the Uttara Tantra is put together, because the placement is not accidental. The division opens with the eye and gives it a very large number of chapters — the eye section alone runs longer than most whole divisions of other classical works. Ear, nose and head follow, and that whole block is the Shalakya Tantra, the speciality of the parts above the collarbone. Only when Shalakya closes does the text turn to children. So the Graha chapters do not sit in a corner reserved for the supernatural. They sit in the ordinary run of specialities, immediately after the head chapters and immediately before the remaining Grahas, treated with the same apparatus of chapter headings, numbered verses and closing formulae as the surgical material.

So chapter 27 tells you what each Graha was believed to look like, and chapters 28 to 31 tell you what was to be done. Splitting description from therapeutics like this is the standing habit of the Samhita — the eye section does it, the head section does it, the throat section does it. It is worth saying plainly at the outset that Graha Chikitsa, the treatment of the Grahas, belongs to the branch of classical Ayurveda called Kaumarabhritya, the care of children, which sits alongside Bhuta Vidya in the eightfold scheme of Ashtanga Ayurveda. It is one of the eight, not a fringe.

Every chapter opens the same way: a decoction, and a sprinkling.

Graha Chikitsa: the protocol that repeats four times

Read chapter 28 on its own and it reads as a single elaborate procedure. Read chapter 29 straight after it and something starts to feel familiar. By chapter 31 the pattern is unmistakable, and it is the most interesting thing in these eight pages.

The four chapters are not four different therapies. They are one sequence, written out four times, in the same order every time. Here it is, with the slot names the text itself uses:

Six of the seven steps appear in all four chapters. The seventh, the topical paste, appears in three: chapter 28 has no Pradeha slot filled at all. And twice the text declines to write out a fresh formula and simply points back — chapter 30 says the fumigations are those mentioned in the chapter on the treatment of Skanda-graha, and adds that the medicated ghrita prescribed there may also be advantageously given in the present instance.

That detail is worth pausing on, because it tells you which half of the procedure the compiler thought was doing the work of distinguishing one case from another. He is willing to reuse a smoke recipe and a ghee recipe wholesale across two different Grahas. He is not willing, anywhere in these four chapters, to reuse a place of worship, a set of offerings or a mantra. Those are written out fresh every single time, at length, with no cross-references at all.

None of this is stated in the source. It only becomes visible when the four chapters are set out next to one another, which is not how anyone reads a printed Samhita — you read chapter 28, then chapter 29, several pages apart, and the repetition registers as style rather than as design. Laid side by side it stops looking like style.

There is a practical reason a text might be built this way. The Samhitas were composed to be memorised and recited long before they were books anyone owned, and a fixed sequence is far easier to carry in the head than four unrelated procedures. Once a student holds the order — sprinkle, anoint, give the ghrita, apply the paste, fumigate, tie, worship — every new chapter reduces to a list of substitutions. The structure becomes the mnemonic and the ingredients become the content. That is exactly the design you would expect from a teaching tradition that had to transmit a large body of detail accurately without relying on anyone reading it.

It also explains the one thing the table above cannot show, which is length. The pharmacological steps are terse — a decoction is given in a single clause, an oil in two. The worship and the mantra take up more than half of each chapter. If the sequence is a memory frame, the parts that get the most words are the parts the compiler thought could least afford to be approximated.

Sechana: the sprinkling that opens every chapter

Every one of the four chapters begins the same way, with a decoction and a sprinkling. Sechana is the sprinkling or affusion of the body with a liquid, and it is always the first instruction after the chapter's opening line.

The plants change each time. For Skanda-Graha it is a decoction of the leaves of the Vilvadi trees, which the text specifies as possessing Vayu-subduing virtue. For Skandapasmara it is Bilva, Shirisha, Golomi — glossed in the translation as Duralabha — and the drugs of the Surasadi group. For Shakuni it is Vetasa, Amra and Kapittha, that is, cane, mango and wood-apple. For Revati it is Ashvagandha, Sariva, Ajashringi, Punarnava, the two kinds of Saha, and Vidari.

Read that last list again and notice how ordinary it is. Ashvagandha and Punarnava are among the most-used plants in the entire classical materia medica, named across the Charaka Samhita and the Bhavaprakasha in contexts that have nothing to do with Grahas at all. They are not exotic ritual substances. They are the everyday shelf.

The qualifier attached to the very first one is worth keeping. The Vilvadi decoction of chapter 28 is specified as possessing Vayu-subduing virtue — that is, it is chosen for its action on Vata, one of the three doshas of classical physiology. This is the only place in these four chapters where a material is justified in the ordinary technical language of Ayurvedic pharmacology rather than simply named. For one clause, the Graha chapters speak the same dialect as the rest of the Samhita, and then they move on.

It is also worth being clear about what Sechana physically involves, because the English word sprinkling undersells it. A decoction is made by boiling plant material in water and reducing it, so what is being applied is a warm, strongly coloured, strongly scented liquid, and it is applied to the whole body. In a household with a distressed infant, the first instruction in all four chapters is to warm something aromatic and pour it over the child.

One structural point about the group names. The text repeatedly prescribes not individual plants but ganas — standing groups with fixed membership, like the Surasadi group or the Kakolyadi group. A gana is a shorthand: the reader is expected to already know which plants it contains. That is a sign of a text written for people inside a training tradition, not for a general reader, and it is the same convention the Samhita uses in its surgical and dietetic sections.

Cooked, not merely infused. The text is precise about which.

Abhyanga and the Sarvagandha taila

The second step is always an oil, and always an oil that has been cooked rather than merely infused. The phrase in the translation is consistent: oil duly cooked and prepared with a decoction and a Kalka, the Kalka being a wet paste of ground plant material. This is the classical method of oil-making — a measured quantity of oil simmered with a decoction and a paste until the water has gone — and the same method underlies every medicated taila used in Abhyanga across the classical literature.

For Skanda-Graha the oil is cooked with the decoction of the roots of those same Vilvadi trees, with the drugs of the Sarvagandha group, and with Kaitaryya and Surananda as Kalka. For Skandapasmara it is simply the Sarvagandha group again. For Shakuni it is oils cooked with the drugs of the sweet and astringent groups. For Revati it is Kushtha and Sarjarasa, the resin.

The Sarvagandha group appears in three of the four chapters, which makes it the single most repeated material in this whole section. A footnote attached to the end of chapter 27, on the facing page, tells us exactly what it is: The drugs of the Eladi group are known as Sarva-gandha. Eladi is the cardamom-headed group of aromatics. So the most-repeated ingredient across the Graha therapeutics is, in plain terms, the aromatic shelf — the fragrant plants.

That is a small thing but it changes the texture of the reading. The oil being rubbed into a distressed child is a warm, strongly aromatic one. Whatever else is going on in these chapters, the sensory character of the procedure is deliberate and it is consistent.

The ghrita draught: Madhura and Kakolyadi

Third comes the only internal preparation in the sequence: a draught of ghrita, clarified butter, cooked with plants and usually with milk.

Chapter 28 gives it as clarified butter duly cooked with Devadaru, Rasna, the drugs of the Madhura group and with milk. Chapter 29 gives clarified butter cooked with milk, with the decoction of the barks of the Kshiri trees, and with the Kakolyadi group. Chapter 31 gives ghrita cooked with a decoction of Dhava, Ashvakarna, Kakubha, Dhataki and Tinduka, again with the Kakolyadi group. Chapter 30, as noted, borrows chapter 28's.

Look at which groups keep coming up. Madhura means sweet. The Kakolyadi group is the classical set of sweet, cooling, nourishing plants. The Kshiri trees are the milk-exuding trees. Add actual milk, and actual clarified butter, and the internal preparation given in every one of these chapters is a sweet, rich, nourishing one.

Pradeha and Utsadana: the slot that is sometimes empty

The fourth slot is a paste worked onto the skin, and it is the one slot in the protocol that is not always filled.

Chapter 28 skips it entirely. Chapter 29 has Utsadana, a rubbing, with a paste of Vacha and Hingu taken together — sweet flag and asafoetida, two of the sharpest-smelling substances in the classical kitchen. Chapter 30 has a full Pradeha, a plaster, of ten named things: Madhuka, Ushira, Hrivera, Sariva, Utpala, Padmaka, Rodhra, Priyangu, Manjishtha and Gairika. Chapter 31 has a plaster of Kulattha, burnt and powdered Shankha — conch-shell — and the Sarvagandha group once more.

Chapter 30's list is the longest single ingredient list in these eight pages, and it is worth noticing what kind of list it is. Ushira, Sariva, Utpala, Padmaka, Rodhra, Priyangu, Manjishtha: this is very close to the standing classical set of plants named for their effect on the complexion and the surface of the skin. Gairika is red ochre, a mineral. So the one Graha whose chapter gets the most elaborate topical preparation gets a preparation drawn from the cosmetic-adjacent end of the materia medica rather than the drastic end.

Chapter 30 also does something none of the others do: it sends the reader out of the section entirely. The various medicinal powders, the diet and the remedial measures in general, it says, are those described in connection with the medical treatment of ulcers — a reference to the Chikitsa Sthana, the therapeutics division, and its opening chapter on wounds. A chapter about a bird-formed figure ends by pointing at the surgical curriculum.

Dhupana sits mid-sequence, with no seam marking it off.

Dhupana: the fumigation shelf

Dhupana is fumigation, and it is where these chapters are furthest from anything a modern reader will find familiar. The materials are not plants.

Chapter 28 fumigates with mustard seeds, the cast-off skin of a snake, Vacha, Kakadani, clarified butter, and the hairs of a camel, a goat, a sheep or a cow, mixed together and burnt. Chapter 29 uses the dung of an owl and a vulture, human hair, the nails of an elephant, clarified butter and the hairs of a bull. Chapter 31 uses barley, bamboo, and the dung of vultures and owls mixed with clarified butter, morning and evening. Chapter 30 borrows chapter 28's.

There is no way to read this list as a pharmacological one and it would be dishonest to try. Hair, horn, nail, slough and dung burnt together produce an acrid, unmistakable smoke, and the materials are chosen from a category — the discarded outer coverings of animals, the leavings of night birds — that carries obvious symbolic weight in the ritual world these chapters belong to. What the smoke does is fill a room with something the household cannot ignore.

It is also the step where the two halves of these chapters are hardest to separate. Fumigation sits in the middle of the sequence, between the oil and the amulet, with no seam marking it off from the plant-based steps on either side. The compiler does not appear to have regarded it as belonging to a different order of thing.

Strung, tied at the neck, and legible to the whole household.

Dharana: the garland tied at the neck

The sixth step is a physical object made and worn. In every chapter, named materials are strung together and fastened to the child, usually at the neck.

For Skanda-Graha the string carries twigs of Soma-balli, which the translation glosses as Guduchi, along with Indra-balli, Shami, the thorns of Bilva and the roots of Mrigadani. For Skandapasmara it is Ajanid, Vandi, Markati and Kukkuti. For Shakuni it is a longer list: Shatavari, Mrigadani, Ervaru, Nagadanti, Nidigdhika, Lakshmana, Sahadeva and Vrihati. For Revati the text calls it a neck-ornament and specifies that it is to be made of Varuna, Arishta — that is, Nimba, the neem — Ruchaka, Sinduka, which is Nirgundi, and Putranjivaka, and that the child should always wear it.

Two of those names deserve a second look. Putranjivaka, the tree named at the head of Revati's amulet, carries the word putra, son, inside it. Revati is described in her own mantra as the mother of many sons. And Shatavari, at the head of Shakuni's list, is a plant whose name and classical reputation are bound up with women and childbearing throughout the literature. The amulets are not arbitrary bundles of wood; at least some of their components have been chosen so that the name of the material echoes the name or nature of the figure being addressed.

Two more of the amulet materials are among the most heavily loaded plants in the Indian tradition. Shami, on Skanda's string, is the tree of the Mahabharata episode in which the Pandavas hide their weapons, and it is worshipped in its own right at Dussehra. Bilva, whose thorns go on the same string, is the bel tree, sacred to Shiva — who is Skanda's father. So the garland tied around the neck of a child in chapter 28 carries, among other things, a piece of the tree associated with the god being petitioned and a piece of the tree associated with his father. Whatever else these strings were, they were legible objects: anyone in the household would have recognised what was on them.

A ditch, a bower, a cow-shed: the site matches the addressee.

A ditch, a bower and a cow-shed

Now to the part of the protocol the text never abbreviates, and the place where the four chapters diverge most sharply.

Every chapter ends with worship at a named location, and no two locations are the same:

• Skanda-Graha — in the inner quadrangle of the child's own house, or at the crossing of roads, on three successive nights.
• Skandapasmara — in a ditch dug for the purpose, with the child afterwards bathed at the crossing of roads.
• Shakuni — inside a Karanja bower, with the child bathed inside the bower.
• Revati — in a cow-shed, with the bathing done at the junction of rivers.

Set them out like that and the logic is immediate. Each site belongs to the figure being addressed.

One location appears twice, and it is the only overlap in the group: the crossing of roads. Chapter 28 offers it as an alternative to the courtyard for the worship of Skanda, and chapter 29 names it as the place where the child is to be bathed. A crossroads is the classic in-between place — belonging to no house, owned by nobody, where paths that were separate meet — and it recurs in exactly the two chapters whose figures are described as companions of one another. Everything else in this part of the protocol is unique to its chapter.

Skanda is a god of the household and of the army, and he is given the inner courtyard of the house. Skandapasmara is described in his own mantra as Skanda's trusted and beloved friend, an ugly-faced one known by the epithet Vishakha, and the offerings made to him are cooked and uncooked meat, fresh blood, and food prepared for the ghosts — so he is worshipped in a pit in the ground. Shakuni is a bird: the mantra calls her sharp-beaked, brown-eyed, and says she traverses the ethereal sky in her flight. She is worshipped in a bower, in the branches. Revati's mantra calls her the mother of many sons, and she is worshipped in the shed where the cattle are kept and where the milk is.

The offerings follow the same rule as the sites. Skanda gets garlands, flags, perfumes, cooked food, newly harvested barley and Shali rice. Shakuni gets huskless sesamum, flower garlands, and two mineral pigments, Haritala and Manahshila — yellow orpiment and realgar. Revati gets white flowers, milk, parched paddy and boiled Shali rice.

Red for Skanda, white for Revati. Never accidental.

Red for Skanda, white for Revati

There is a further layer inside the offerings, and it is the tightest piece of internal design in these four chapters.

Chapter 28 specifies, for Skanda, garlands of red flowers, red flags, and red perfumes such as Kumkuma. Then, a few lines later, the mantra to be recited describes Skanda himself: effulgent with the red rays of his own divine person smeared with the paste of red sandal-wood and decked with the garland of red flowers.

The offerings are not merely red. They are the specific items the mantra says the god is already wearing. The rite hands the figure an image of himself.

Chapter 31 runs white all the way through on the offerings side — white flowers, milk, parched paddy, boiled white rice — and here the correspondence is not with the goddess's own appearance, because Revati's mantra describes her as dark-complexioned, clad in parti-coloured garments and painted with anointments of various kinds. The white belongs to the cow-shed and the milk, to the place rather than the person.

So the colour scheme is doing two different jobs in two different chapters: matching the figure in one, matching the setting in the other. What it is never doing is arriving by accident. Across eight printed pages every colour word in the offering lists can be traced either to the mantra's description of the figure or to the location named for the rite.

The pharmacy and the mantra, side by side

This is the observation that matters most for anyone trying to understand what kind of document they are reading.

It is tempting to imagine that these chapters have a rational medical layer and a magical layer, and that a careful reader could separate them — keep the plants, set aside the mantras. The text gives no support at all to that idea. The two run together in single paragraphs, in the same voice, with the same instructional verbs, and with no seam anywhere.

Within a few lines of one another in chapter 31 you get a decoction of Ashvagandha and Punarnava, a cooked oil, a ghee draught, a conch-shell plaster, a neem and Nirgundi neck-ornament, a cow-shed, white flowers, and a verse addressed to a tall and terrible-looking goddess. There is no shift of register. Chapter 30 does something even more striking: it sends the reader to the ulcer chapter of the Chikitsa Sthana for the powders and the diet, and then, on the same page, describes the offerings of orpiment and realgar to be made inside a Karanja bower.

A cross-reference to the surgical curriculum and an instruction about a bower sit in the same short chapter, written by someone who evidently saw no contradiction. The honest description of these chapters is not "medicine with some ritual attached" and not "ritual with some medicine attached". It is a single procedure in which both were load-bearing, produced by a tradition that had not drawn the line modern readers are used to.

The chapter that bathes the nurse too

Chapter 31 contains one clause that does not appear in any of the other three, and it is easy to read straight past.

After the offerings in the cow-shed, the instruction is that the nurse as well as the child should be bathed in a junction of rivers.

The nurse. Not the child alone. Every other chapter in this group directs its entire procedure at the infant — sprinkle the child, anoint the child, fumigate the child, tie the charm on the child, bathe the child. Chapter 31 is the only one that widens to include the person doing the caring, and it puts her in the same river.

Something similar happens on the practitioner's side. Chapter 28 requires that the physician, or a person acting on his behalf, bathe at night for three successive nights. Chapter 29 requires the child to be bathed by the physician observing the necessary fast. Chapters 30 and 31 both specify a physician of self-control. In four short chapters the text three times makes a demand of the practitioner's own body and conduct, and once makes a demand that includes the caregiver.

Strip the supernatural frame away for a moment and look at what the household is actually being asked to do over these three nights. Warm liquid on the skin. Oil worked in by hand. Something rich to swallow. A room thick with smoke. An object made and tied on. A bath. And every adult in the house stopping what they were doing, fasting, washing, staying up, and giving one small sick person their entire attention for three days running.

Whatever anyone concludes about the aetiology, that is a recognisable description of intensive care by touch, warmth, feeding and undivided attention — and the text thought to extend it to the exhausted person holding the baby.

Three centuries arguing in the footnotes

One last feature of these eight pages, and it is the kind of thing that only shows up in a heavily annotated edition.

The footnotes to these four short chapters contain three separate interventions, from three different points in time, and they disagree with each other.

First, chapter 28 uses the word Kaitaryya, and a footnote explains: Kaitaryya, according to Dallana, means Nimba or Guduchi. Dallana is the twelfth-century commentator on the Sushruta Samhita, writing roughly a millennium after the text he is glossing, and he is being cited to settle what a plant name means — and even then he offers two possibilities rather than one.

Second, in chapter 29, the translator openly overrules him. The note reads: Although Dallana prescribes the drugs of the Kakolyadi group to be used as Kalka, we are inclined to take the sentence to mean that the decoction of those drugs should be used. That is a 1916 editor in Calcutta disagreeing, in print, with a twelfth-century commentator about whether a group of plants goes into the preparation as a wet paste or as a decoction.

Third, in chapter 31, a note records an addition from a completely different work: There is an additional text in the Bhavaprakasha which says that Guggulu, Nalada and Giri-kadamba should also be used in the preparation of the medicated oil. The Bhavaprakasha is a sixteenth-century compendium. It has added three ingredients to a recipe in a text older than itself, and the nineteenth-century edition has preserved the addition as a note rather than folding it into the line.

It is also a useful corrective to the way classical Ayurveda is sometimes presented — as a sealed body of knowledge handed down unchanged. The footnotes of four minor chapters show three separate people changing it.

What these chapters are not

Some plain statements are needed here, because this is material about sick children and about a category of illness the text explains in terms no one uses now.

This is a historical document, not a set of instructions. Nothing described above is something to carry out. Nothing on this page tells anyone what to do about an unwell child. A child who is unwell needs a qualified doctor, promptly, and nothing in a two-thousand-year-old chapter changes that in any way.

The chapters record ritual practices of their own period. Chapter 28's account of the worship of Skanda includes, among the offerings, the sacrifice of a cock, along with the ringing of bells, the lighting of a sacrificial fire with libations, and the consecration of the bathing water by recitation of the Gayatri Mantra. We note it because leaving it out while quoting the rest of the chapter would misrepresent the source. It belongs to the ritual world of the text's own period. It forms no part of Ayurvedic practice today, and nothing we make or sell has any connection to it whatsoever.

The aetiology is not one we share. These chapters attribute infant illness to the action of named non-human agents. That framework is not an early version of a modern one and we are not going to dress it up as such. It is what people who were trying hard to help sick children believed, in a period when the causes of those illnesses were genuinely unknown, and it produced a body of careful, systematic, well-organised writing.

On our own shelf: three ways of making a thing

These four chapters are, among other things, a long record of how things were made. Something is boiled down into a liquid. Something is simmered into an oil. Something is ground to a paste. Those three verbs recur far more often than any single ingredient does, and the text is consistently precise about which one applies.

That interest in how a material is made is the only thread we draw from these pages to our own shelf, and it carries nothing else with it. Here are three things we make, described purely by their method.

Related stories — more to read

• Coconut oil, in short
• Oil pulling as a morning ritual
• Akshaya Tritiya: a ritual of the calendar

And elsewhere on the blog: the nine Bala Grahas of chapter 27, which is the companion piece to this one; the head chapters that come immediately before; the ear section of the same Uttara Tantra; and the eightfold framework that gives Kaumarabhritya its place.

Questions people ask