Śaṅkaraon Brahma Sūtra 3.3.45

. Your supposition, the purvapakshin objects, as to those fire-altars being not supplementary to the sacrificial act, but altogether independent of it, is untenable. The in- fluence of the leading subject-matter rather compels us to conclude that the instruction given by the text about the agni made of mind and so on, enjoins some particular mode of the same agni which the preceding sections describe as — the outcome of a real act But arc not indicatory marks — stronger than the leading subject-matter ? True in general but indicatory marks such as those contained in the pas- sages quoted above are by no means stronger than the general subject-matter. For as those passages are of the nature of glorifications of the fanciful fire-altars, the liiigas (have no proving power in themselves but) merely illustrate some other matter (viz. the injunction to which those passages are arthavadas) ; and as they are of that nature they may, there being no other proof, be taken as mere guwavadas, and as such are not able to sublate the influence of the prakara;i?a. On the ground of the latter, therefore, all those fanciful agnis must be viewed as forming parts of the sacrificial action. The case is analogous to that of the ‘ mental ’ (cup). On the tenth day of the Soma sacrifices occupying twelve days —which day is termed avivakya—a soma cup is offered mentally, the earth being viewed as the cup, the sea as the Soma and Pra^apati as the divinity to which the offering is made. All rites connected with that cup, viz. taking it up, putting it down in its place, offering the liquid in it, taking up the remaining liquid, the priests inviting one another to drink the remainder, and the drinking, all these riles the text declares to be mental only, i.e. to be done in thought only^. Yet this mental quasi-cup, as standing under the heading of a sacrificial act, forms part of that act. — The same then holds good with regard to the quasi-agnis made of mind and so on.