kva ciyānujātā kva cic chilā-vāluka-nirggateyaṁ kva cit prakīrṇṇā śubha-śīta-toyā saṁprasrutā maṅgakarīva gaṅgā

kva ci yānujātā kva cit suśucy-ambu-ruhānujātā kva cit suśucy-ambu-ruhānujātā Iyantuśucyambu-ruhānujātā None of the available images support any reading with buruhā. saṁprasrutā saṁprasrātā maṅgakarīva meddhyakarīva Dominic Goodall suggests that perhaps we can take the faint head-mark attached to the left horn of ma as an unusual e, in which case we can read me ’ṅgakarīva, “who, as it were, makes up my body (/makes up a part of me)”. Given the way water spouts forth from numerous locations near the site of the inscription, we might also fancy that we are dealing with a metrically conditioned equivalent to mataṅga-karīva “like one who has an elephant's trunk”.

At times overgrown by , at times disappearing in pebbles and sand, at times spreading out its pure and cool water, the Ganges gushes forth here, like a maṅgakarī!

Gushing forth is this (stream), purifying as the Ganges, at some places bedecked with bright lilies and lotus flowers, at some places bubbling out from pebbles and sand, (and) at some places spreading out its pure and cool water.

Our estimate of the inscription's date follows the proposal first made by Krom () and subsequently endorsed by Damais (611; 101) as well as De Casparis (). pointed out that the inscription is inspired by Kālidāsa's , canto 13, verses 54–57.

First published by H. Kern (); re-edited by B. C. Chhabra (, ) based on photographs by N. van Erp (OD 2158–2162). Re-edited here by Arlo Griffiths, using the same photographs but taking into account also the photogrammetry made by Adeline Levivier in October 2025.

33-34 I13-14II 49, 94, 104, 138-139 26-27 102-103 28 43-45 23-24