Many thanks to all who replied to my query. As many of you suggested, I’m also inclined to interpret these sentences as strings of “flag words” (of bits of words) marking sections of the (RV?) text, although at the moment I’m not able to find a clue to the pattern they follow.
I add some information about the manuscript, as well as a reference to a similar weird sentence quoted by Winternitz:
— the manuscript (02366, Cambridge UL collection) is on palm leaves, in Grantha script, without date
- — Vedic accents are marked, the separation between elements in a compound (puroḥ-hitam, su-upāyanaḥ, etc.) is marked with a special sign, anunāsika is marked with a special sign as well
— the division of the text follows mainly the aṣṭaka/adhyāya/varga(/sūkta/ṛc) system; it seems that the weird sentences follow the end of each varga
— Winternitz (A Catalogue of South Indian Sanskrit Manuscript, Especially those in the Whish Collection […], 1905) quotes a similar sentence in his transcription of the explicit of a manuscript of the Ṛgveda padapāṭha (No. 166 [Whish No. 177], p. 223-224; see attachment). He couldn’t make any sense out of that sentence, in fact he marks it with two question marks: gatirnnādhadhāmaṣṭama nassanna sanūs sanam (??). His manuscript is also in Grantha script, on palm leaves, but it contains aṣṭakas 5 to 8, while “my" manuscript has only aṣṭakas 1-2: thus it is not possible to compare the sentence he quotes with the corresponding one in my manuscript.
Next week I’m going to transcribe more sentences from my manuscript and send them to the list, together with the transcription of the colophon(s). For the time being, it is perhaps worth noting that some words recur in the sentences: gatir, ddhvan, aṣṭama; possibly they are not quotations from the text, but kind of “metadata’ instead.
Best wishes,
Marco
---