The Bhāgavata's world is a set of concentric islands separated by oceans of salt, sugarcane juice, liquor, clarified butter, milk, churned curds and sweet water. Each ring doubles the one inside it.
That much is geography, and it is the part usually quoted. The part that makes 5(20) worth a page is what the inhabitants do. Every ring worships a different deity, through a different practice, with its own mantra printed in the text.
Read as a set, the six assignments are supposed to look like a system. They do not quite hold together — and the way they fail is more informative than the tidy version would have been.
Beyond Jambūdvīpa lie six inhabited rings. Each has a ruling son of Priyavrata, seven sub-regions, seven boundary mountains, seven rivers and four named classes of people. And each has a god:1
| Dvīpa | Named feature | Worshipped as | The people's practice |
|---|---|---|---|
| Plakṣa | golden plakṣa tree | the sun | "follow the three Vedas" |
| Śālmali | śālmali tree, Garuḍa's abode | the moon | "using the Vedas" |
| Kuśa | clump of kuśa grass, blazing | fire | "accomplished in rites" |
| Krauñca | Krauñca mountain | water | cupped hands of river water |
| Śāka | fragrant śāka tree | wind | prāṇāyāma and "supreme meditation" |
| Puṣkara | lotus of a hundred million golden petals | Brahmā | "undertake rites" |
The mantras are given in each case and they are not interchangeable. Krauñca's addresses the substance directly — "O water! You are the valour of the supreme being… As we touch you, may we also be purified." Śāka's addresses a function — "He is the one who maintains the breath of life." Puṣkara's addresses a procedure — "His manifestation is in the form of the rites and the rituals."1
Each ring is also careful to state that the deity is a form of the illustrious one rather than a rival to him: they worship him "in the form of the fire god," "whose atman is the moon god." The scheme is a distribution of one being across six modes, not a survey of six religions.
The obvious reading — six rings, the five great elements plus their maker — does not survive contact with the list.
Fire, water and wind are elements. The sun and the moon are not. They are luminaries, and the classical set they belong to is a different one.
Earth and space are missing. Both are mahābhūtas, both are available, and neither is assigned to a dvīpa. Earth is not absent from the chapter — it appears elsewhere in the volume as a recipient of the sense of smell — so the omission is not a vocabulary gap.
And Brahmā is not an element at all. He is the creator, and the ring he governs is the one containing his lotus-seat.
So the set is: two luminaries, three elements, one deity. It has the shape of a system — one per ring, exhaustive, orderly — without being any system the tradition otherwise uses. A reader who arrives expecting the pañca-mahābhūta will supply the missing two and misreport the chapter, which is the specific error this page exists to prevent.
What the six do share is a weaker property: each is something that pervades and sustains from a distance. The sun and moon light and feed, fire carries oblations, water purifies, wind maintains breath, Brahmā manifests as rites. That is a coherent theme and it is not an elemental one.
Five rings run the template exactly. The sixth abandons it.
Plakṣa through Śāka each receive: seven varṣas named for seven sons, seven boundary mountains listed by name, seven rivers listed by name, and four classes of inhabitant listed by name. The chapter supplies all four lists, five times, without deviation.
Puṣkara receives none of it. Vītihotra has two sons, Ramaṇaka and Dhātaki, and they take two varṣas divided east and west by a single mountain, Manasottara. No seven rivers. No four classes.1
The break is total and it is not a lapse of attention — the chapter's remaining energy goes into Puṣkara at length, describing the lotus, the hundred million golden petals, the sun's chariot circling Manasottara, and the cities of the world-guardians. It has plenty to say. It simply stops saying the sevenfold thing.
And the ring that breaks the template is the one whose deity is not an element and whose feature is Brahmā's own seat. The outermost inhabited ring is structurally different in every respect at once — deity, division, and the absence of a named population.
Beyond it the chapter passes out of the inhabited world entirely: Lokāloka, the mountain dividing lit from unlit, and past that "a region made out of gold, like the surface of a mirror. If something is dropped into it, it can never be regained. Therefore, all living entities avoid this place."1
Read in sequence, Puṣkara is a transition rather than a member. Five rings of the system, one ring that is already leaving it, and then the boundary.
The class-lists deserve separate attention, because they make a claim the chapter never states.
Every one of the five templated dvīpas has exactly four classes, and not one of them uses the familiar names:1
The count is invariant; the names are entirely local. The implicit position is that fourfold social division is a property of inhabited worlds, and the names are a local accident — an unusually strong universalism to state by construction and never argue for.
One set is patterned in a way the others are not. Śāka's four all end in -vrata: they are graded by vow — truth, gift, observance. And Śāka is the ring whose practice is prāṇāyāma and meditation rather than Vedic recitation or rites. The one population named for vows is the one doing breath-work, which is either a nice piece of design or a coincidence in a chapter with thirty-odd name-lists. At n=1 it is worth recording and not worth building on.
The rulers do the same thing, repeatedly, and the chapter narrates it each time without ever remarking that it has happened before.
Idhmajihva divides his dvīpa among seven sons and, "retiring from a material existence, immersed himself in yoga." Hiraṇyaretas divides his and "himself resorted to austerities." Ghṛtapṛṣṭha divides his, "became a devotee of the illustrious one," and "obtained Hari's lotus feet." Medhātithi divides his and, "immersing his mind on Ananta, he himself became a devotee and entered a hermitage." Vītihotra installs his two sons and, "like his elder brothers, engaged himself in rites for the illustrious one."1
Five of the six. Only Yajñabāhu of Śālmali is left without a stated exit.
The last of the five is the only one that notices: "like his elder brothers." One backward glance in a sequence of five, and it comes at the ring where everything else about the template is breaking down.
Check a "system" against the system it resembles before repeating it. Six rings and six gods invites the five-element reading, and the list refuses it. Where a source's set is nearly a standard set, name the differences explicitly or the standard set will be reported instead.
Count the template's iterations and find where it stops. Five clean repetitions and one break is a structure. Summaries that describe the sevenfold pattern as holding across all the dvīpas are describing five of six.
Record what varies and what does not, separately. Here the count is invariant and the names vary — which is the chapter's whole implicit anthropology, carried entirely by the shape of its lists.
Treat a single self-aware instance in a repeated sequence as marked. Like his elder brothers is the only cross-reference in five parallel abdications, and it lands where the pattern is ending.
The set is almost a canonical set. Two members swapped, two missing, one from a different category. The near-miss is the sign that a scheme has been assembled rather than derived.
The last member of a series breaks its template. Fewer subdivisions, no population, a different kind of deity. A terminal member that stops conforming is usually a boundary rather than an instance.
A repeated structural count with fully local names. Four classes everywhere, no name shared. The invariant is the claim; the names are camouflage.
One instance in a repeated sequence refers to the others. In a chapter of parallel formulae, the single backward reference is doing something the others are not.
Strongest support. All six deity assignments, all six mantras, the five complete four-name class lists, Puṣkara's two-son division, and the five abdications are quoted or tabulated directly from one continuous passage at 5(20).
Tension the page preserves. Whether the six assignments were ever meant as a single scheme, or accumulated as six independent regional traditions later set in a row, is not decidable from the chapter. The page reports that they do not form a known system and does not assert that a lost one lies behind them.
Absence-claim discipline. Earth and space are not assigned is a claim about chapter 5(20) as read. No sweep for dvīpa-deity assignments elsewhere in the volume was run, and a different chapter could assign them.
The Śāka observation is n=1. One dvīpa whose class-names share a morphology and whose practice is interior. Recorded as an observation about one ring, not as evidence that practice and nomenclature track each other.
Open question. Do the other Purāṇas assign the same six deities to the same six dvīpas? If the assignments are stable across the genre, the near-miss with the five elements is inherited and old; if they vary, this chapter's particular set is a choice and worth reading as one. The vault holds Śiva Purāṇa material and could begin the check.
The chapter's geography is exhaustively regular and its theology is not, and the volume shows no sign of registering the difference.
That is a shape this build has met before. The volume executes structure with great consistency — a word-identical priestly payment printed twice, a citation formula deployed only for demotions, fifteen faculties returned in fixed order — and it does not comment on any of it. Here the same habit produces a cosmology in which the counts are rigid to the point of monotony while the contents answer to no scheme at all.
There is also a quiet convergence with this volume's treatment of renunciation. Five world-rulers divide their inheritance and walk away from it, and the text presents this as simply what a ruler does when the succession is arranged. Elsewhere in the volume the renunciate's claim is contested — Dakṣa attacks it, Yadu is vindicated for it, and the argument is never settled. In 5(20) the argument does not arise. Abdication is the default and it happens five times without a word of justification.
The volume's most contested question in one Skandha is its unexamined background assumption in another.
No load-bearing cross-domain handshake. This page's subject is the internal structure of one chapter's cosmography — which deities are assigned to which rings, where the template holds and where it breaks. Its productive connections are to the volume's other structural regularities and to its treatment of renunciation, all inside eastern-spirituality and carried in Connected Concepts.
The chapter's implicit universalism about fourfold social division is the one element with genuine cross-domain reach, and the vault holds no counterpart for it — nothing on invariant social structure under local naming. Recorded as a gap rather than attached to a page on a different mechanism.
Sharpest implication. Six rings, six gods, and it is not the five elements: two luminaries, three elements, one creator, with earth and space simply left out. The scheme has the shape of a system and the membership of an accumulation.
Generative questions.