The volume prints two application-manuals from the Mantra-Mahodadhi back to back — one for a twelve-syllable Hanumān mantra, one for an eighteen-syllable one.
Read them in sequence and something becomes obvious that neither text says. The rendering says it:
"the two prayoga-vidhis are variations on one shared ritual template rather than independently invented procedures."1
⭐ Same maṇḍala. Same nine seat-powers. Same four widening circles. Same eight epithets, same eight commanders, same ten guardians. The mantras differ, the viniyoga differs, and the counts differ.
Everything else is one procedure with two formulas dropped into it.
It is worth setting out in full, because the build has no other complete maṇḍala procedure and because the shape is the point.12
A sarvatobhadra maṇḍala — "auspicious on every side" — is drawn, and the seat is built up inward: a tortoise and the cosmic serpent, then earth, then the milk-ocean, then the jewel-island, then the wish-tree, and finally a lotus seat.
On the lotus seat's four corners: dharma, knowledge, dispassion, lordship. On its four cardinal points: their opposites. At the centre: a further stack — bliss, consciousness, the three guṇas, self and supreme self.
Around it, nine seat-powers: Vimalā · Utkarṣiṇī · Jñānā · Kriyā · Yogā · Prahvī · Satyā · Īśānā · Anugrahā.
Then the image or yantra is bathed, installed at the centre, and worshipped in four widening circles:
One — the six body-parts at the lotus's central whorl: heart, head, crown, armour, three eyes, weapon. Two — eight epithets on the eight petals: Rāmabhakta · Mahāteja · Kapirāja · Mahābala · Droṇādriharaka · Merupīṭhācaraṇakāraka · Dakṣiṇāśābhāskara · Sarvavighnanivāraka. Three — eight monkey-generals at the petal-tips: Sugrīva, Aṅgada, Nīla, Jāmbavān, Nala, Suṣeṇa, Dvivida, Mayanda. Four — the ten guardians of the directions with their weapons at the outer square.
⚑ The construction runs from the cosmological to the personal and then back out to the geographical. Tortoise and serpent at the base; the deity's own body-parts at the centre; his companions at the petal-tips; the directions at the perimeter. You assemble a cosmos, install a body in it, and then populate the surroundings with his colleagues and the compass.
Three things differ, and the third is the interesting one.12
The mantra. The twelve-syllable is almost entirely bīja-syllables, closing on the single readable word namaḥ. The eighteen-syllable is plain, readable Sanskrit: Oṃ namo bhagavate āñjaneyāya mahābalāya svāhā.
The viniyoga. Twelve-syllable: ṛṣi Rāmacandra, metre Jagatī, bīja hasauṃ, śakti hasphreṃ. Eighteen-syllable: ṛṣi Īśvara — Śiva, not Rāma — metre Anuṣṭubh, bīja huṃ, śakti svāhā.
⭐ The seer changes. The same deity, the same maṇḍala, the same companions — and one mantra is seen by Rāma and the other by Śiva.
The count. Twelve-syllable: 12,000 repetitions constitute one full puraścaraṇa, then a tenth of that in fire-offerings, a tenth of that in libation, a tenth of that in feeding brahmins. Eighteen-syllable: a lakh — 100,000 — with a tenth offered again in fire.
⚠ A shorter mantra costing 12,000 and a longer one costing 100,000, in the same template, with no reason given for either figure.
The eighteen-syllable manual carries a sentence its neighbour does not, and it is a sequencing rule:
"Once the mantra is thus accomplished, only then should its practical application be undertaken."2
⭐⭐ Accomplishment first; use second. A hundred thousand repetitions before the mantra may be applied to anything.
⚑ And the twelve-syllable manual, sitting immediately before it in the same volume, from the same source text, in the same template — says nothing of the kind. Its 12,000 and the tenth-portion chain complete the puraścaraṇa, and the text moves on.
So a reader working from the first manual has a mantra he may presumably use at once. A reader working from the second must complete a lakh before touching it.
⭐ Which is exactly the class of instruction this build has recorded as chronically missing — a gate between practice and application — and here it exists, in one of two adjacent texts, with no indication whether it governs both.
⚠ The honest reading is that it probably does. Two vidhis from one source sharing one template are unlikely to differ on whether accomplishment precedes use. But the rule is printed once, and a practitioner who opened only the first page would never encounter it — which is the placement failure the build has been tracking, arriving here between two facing pages rather than across a hundred documents.
Look at what is actually worshipped on the eight petals, because the list is not decorative.1
Rāmabhakta — devoted servant of Rāma. Mahāteja — greatly radiant. Kapirāja — king of monkeys. Mahābala — immensely strong. Droṇādriharaka — bearer of the Droṇa mountain. Merupīṭhācaraṇakāraka — worshipper of Meru's own seat. Dakṣiṇāśābhāskara — illuminator of the southern quarter. Sarvavighnanivāraka — destroyer of every obstacle.
⭐ Four are qualities, three are deeds, and one is an office. The first four could describe a disposition; Droṇādriharaka and Dakṣiṇāśābhāskara are specific to two episodes — the Sañjīvanī errand and the southward search — and Sarvavighnanivāraka is what the worshipper wants.
⚑ So the petals run from what he is, through what he did, to what he does for you. Circling the diagram is a rehearsal of that progression, eight times over, in a fixed order — and the practitioner ends each circuit on the obstacle-remover.
Immediately after the twelve-syllable manual's maṇḍala mechanics — the nine powers, the four circles, the tenth-of-a-tenth chain — the volume prints four short verses, and the rendering flags what they are doing there.
They are the brahmins of Dharmāraṇya's praise of Hanumān, from the Skanda Purāṇa's Brahma Khaṇḍa, Dharmāraṇya Māhātmya 37.3–8 — and they are offered as "a reminder that underneath all the maṇḍala mechanics, Hanumān praise can still just be four lines of namaḥ."7
⭐⭐ Four salutations, placed deliberately after the most elaborate procedure in the appendix, and carrying a chapter-and-verse citation that the manuals themselves do not have.
⚑ Note the asymmetry. The maṇḍala runs for pages and is attributed to a named work with no locator. The coda is four lines and is cited to book, section and verse-range. ⚠ The volume's most precisely sourced item in this appendix is the one that says the apparatus is optional.
⭐ Which is an editorial judgement rather than an argument, and it is the only place in the ritual section where anyone makes one. Nothing in the manuals concedes anything; the editor concedes it for them, by what he prints next.
One detail is worth recording precisely because it is the kind of thing that only appears when several texts are read together.
The eight commanders here are Sugrīva, Aṅgada, Nīla, Jāmbavān, Nala, Suṣeṇa, Dvivida, Mayanda.1
The Nārada Purāṇa's mantra material gives the same eight with Mainda for Mayanda.3 Its lamp-offering rite, three pages later, gives a different eight — Sugrīva, Aṅgada, Suṣeṇa, Nala, Nīla, Jāmbavān, Prahasta, Suveṣa.3 The Tantrasāra kalpa gives a third — Sugrīva, Lakṣmaṇa, Aṅgada, Nala, Nīla, Jāmbavān, Kumuda, Kesarī, with Pavana and Añjanī at right and left.4
⭐⭐ Four rosters, four texts, one volume. The ritual frame is stable to the point of being copied verbatim between adjacent manuals; the personnel are not.
⚑ Which is a real finding about how this material was transmitted. The maṇḍala's architecture — tortoise, serpent, milk-ocean, wish-tree, nine powers, four circles — survives copying exactly, because it is structural and any deviation would be visible as an error. The list of who stands on the petals does not, because a name substituted for a name still yields a working diagram.
⚠ No contributor notices, and the volume prints all four.
When two procedures share a template, read them against each other before reading either alone. This pair is the clean case: the differences are three, they are locatable in a minute, and one of them is a rule that changes what the practice is for. ⚑ A rule stated once in a matched pair is almost certainly meant for both — and the only way to find it is to lay them side by side.
Distinguish what a text fixes from what it merely carries. The maṇḍala here is fixed: it is copied without variation, and a mistake in it would be structural. The companion roster is carried, and it drifts. ⭐ The general test is whether an error would be visible — a wrongly built seat has a hole in it, a wrongly named general does not — and the parts of any inherited procedure that drift are the parts where mistakes are invisible.
And notice which of the two costs more, because it is not the one you would guess. The longer, plainer, more readable mantra requires a hundred thousand repetitions; the shorter, opaque, seed-syllable one requires twelve thousand. ⚠ If you are choosing between practices on effort, the intelligible one is eight times the price — and neither manual says why, which means the reason is not in the text and any explanation you supply is your own.
Two application-manuals from the Mantra-Mahodadhi,12 and two of the volume's other ritual texts for the roster comparison.34
Bounded, and heavily. The rendering states that the maṇḍala mechanics are condensed to prose rather than transcribed "namaskāra by namaskāra", since "the two prayoga-vidhis share one template closely enough that reproducing both in full would be pure repetition."1 ⚠ So the identity of the two templates is the rendering's judgement, made while condensing. This page rests on it and cannot independently confirm it. [RENDERING — verify against Hindi]
⚠ Bounded again. The Mantra-Mahodadhi attribution is the volume's; this build has not opened it. The translator of the second manual is credited as Śrī Vaidehīśaraṇ Śāstrī; the first carries no translator credit. [PLAUSIBLE — needs corroboration]
🔴 A tension preserved. The accomplishment-before-application rule appears in one manual and not the other, in a matched pair from one source. ⚑ Either it governs both and was printed once, or the two mantras genuinely differ on it. The volume gives no basis for choosing, and this page asserts neither.
Open question, and it is the one a practitioner would need answered. Why does the eighteen-syllable cost a lakh and the twelve-syllable twelve thousand? ⭐ The lakh-system's own rule is a lakh per syllable — which would give twelve and eighteen lakhs respectively, and matches neither figure.5 ⚑ So these counts come from somewhere else, and the source is unstated. A count that follows no stated rule and differs eightfold between two otherwise identical procedures is either carrying information the text has lost, or is arbitrary — and the difference matters enormously to anyone deciding which to undertake.
The volume's ritual appendix holds two philosophies of what a manual is for, and this pair is the pure case of one of them.
These manuals are architectural. They specify a structure exhaustively — every layer of the seat, every power around it, every circle of worship, every name on every petal — and they specify the volume of practice exactly. The practitioner executes; nothing is left to judgement.
The Saubhāgya manual, four pages earlier, is the opposite. It names seven mudrās precisely and then leaves the count entirely open: "according to one's own strength, understanding, and fixed count."6
⭐ What each text pins down is what it believes cannot be got right by feel. These two believe efficacy is quantitative and structural, so both are fixed. The Saubhāgya believes it is structural only, so the shape is fixed and the volume is the user's.
⚑ And the disagreement is invisible to a reader working through the appendix in order, because nothing marks the transition — three manuals in eight pages, two of them handing you a number and one of them handing the number back.
⚠ The volume convenes the disagreement, as it convenes every other, and adjudicates nothing.
Puraścaraṇa: The Lakh-System of Mantra-Siddhi. That page sets out the counting contract's own arithmetic — one lakh per syllable of the mantra, so a five-syllable formula takes five hundred thousand — and defends the frank measurability as what makes the practice available to an ordinary person. These two manuals use the same vocabulary and neither obeys the rule. ⭐ A twelve-syllable mantra at 12,000 and an eighteen-syllable at 100,000 are not one-lakh-per-syllable, or any other consistent function of length. ⚑ What the pairing produces is that puraścaraṇa is a word with at least two incompatible specifications in circulation — a general rule keyed to syllable-count, and per-mantra figures handed down with the mantra — and a practitioner who has met only one of them will mis-price every practice he meets from the other tradition by an order of magnitude. ⚠ The volume prints both without noticing, and the lakh-rule is the one that sounds like a principle, which is why it is the one likely to be generalised wrongly.
Sampuṭīkaraṇa: Enclosing the Core. That page records a manual whose whole design is symmetry — placement, image, gesture, garland, count, garland, gesture, image, placement — a session assembled on the way in and disassembled on the way out. These manuals build outward and never come back. ⭐ The maṇḍala is constructed from a tortoise up to a lotus seat, populated in four widening circles, and the text simply ends — there is no reversal, no dissolution, no instruction for taking any of it down. ⚑ What the pairing yields is a genuine structural difference between two ritual grammars in one appendix: one treats a session as a state to be entered and exited, the other treats it as a construction to be completed. ⚠ And the second leaves an obvious question unanswered — what is the practitioner supposed to do with a fully assembled cosmos when he stands up — which the first grammar answers as a matter of course and this one does not raise.
Sharpest implication. Two texts, adjacent, from one source, sharing a template so closely that a translator declined to reproduce both — and the difference between them includes a rule about when a mantra may be used at all. ⚑ This is what the build's recurring answer filed away from its question pattern looks like at minimum distance: not a hundred documents apart, not a hundred pages, but the facing page. ⭐ Which suggests the mechanism is not forgetting or scattering at all. It is that neither text was written to be read beside the other — each was written as complete, and completeness in this genre evidently does not include the rules the author took for granted.
Generative questions.
Does the Mantra-Mahodadhi state the accomplishment-before-application rule generally, once, for all its mantras — with these two manuals inheriting it and only one restating it? ⭐ That would explain the asymmetry exactly, and would mean the volume has excerpted two chapters out of a book whose governing rules live in a preface nobody translated.
What happens to the maṇḍala afterwards? Nothing in either manual dissolves it, and the volume's other grammar dissolves everything symmetrically. ⚑ Either the construction is understood to persist — in which case a practitioner accumulates assembled diagrams — or the taking-down was too obvious to write, which is the same assumption that swallowed the eligibility rules.