Several times we find the first person aor. med. in the same form as in our passage, a vrikshi; see, for instance, the Nivid formula to the Visve devah, Sankhayana Srautasutra VIII, 21. In this Nivid, the text of which as given by Hillebrandt is not quite identical with that of Ludwig, we read according to Hillebrandt's edition: ma vo deva avisasa ma visasayur a vrikshi. This ma ... a vrikshi looks quite similar to our passage. The same may be said of Taittiriya Samhita I, 6, 6, 1. yat te tapas tasmai te mavrikshi. Considering such passages it is difficult not to believe that it is the verb a vrask which we have before us in our verse. It must be admitted indeed that the accusative

samsam does not agree with the construction of the later Vedic passages. Can the accusative stand in the ancient language of the Rig-veda in the same connection in which we have found the dative and the locative? So that a-vrask (in the middle or passive) with the accusative would mean: to be cut down in the direction towards another being, i. e. being forfeited to that being? In that case the translation of our passage would be: 'May I not, O gods, fall as a victim to the praise (or rather, to the curse) of my better.' If this explanation of the accusative is thought too bold, we should propose to correct the text so as to get a dative or, which would suit the metre better, a locative: ma' gya'yasah samsaya (or samse) a' vrikshi devah.