Within the language of the Rigveda there appear four derivatives amaritr. '-, amf~r-, mf~r-, arn~ri-from an Old Indic root m? (ablaut-type pratar?tr. '-, suprat~r-, tSturi-:tf 'cross'), attested in seven passages in total, 1 and restricted in distribution to three hymn-cycles of the collection (IV ~, VIII 4, IX1). ~ Since the last occurrence of these formations is encountered at IX 61.24 in the Soma-man..dala, undoubtedly a compilation of hymns directed to that divinity which have been extracted from the family-books, we may at once try to be more exacting about the probable source of the hymn in question. Valuable evidence in this direction is provided by the mention of the name Turv~i~a Y~idu in verse 2, for elsewhere in the sam. hita -omitting five passages in the heterogeneous 1st cycle and one in the younger Xth cycle --this name (in one form or another) a appears only in the Vth through VIIIth mag.dalas, distributed in the following fashion: V ~, VI a, VII ~, VIIF. The manifestly predominant use of Turv~i~a in hymn-cycle VIII and its absence in hymn-cycle IV, the only other locus of the formations from m? under discussion, thus seem to clearly suggest that IX 61 originally stemmed from the same bardic circle as the hymns of the VIIIth ma~.dala. And in support of this view, we note that the anukramag~ attributes the composition of IX 61 to a poet, Amahiyu -~flgirasa, whose family name figures, second only to the K~.nvas, among the bards of cycle VIII (e.g., Purumedha .~fig., Priyamedha .~flg., Virfipa Afig., etc.), but never among those of cycle IV. We may therefore consider the set of forms under discussion to be part of the isolated and therefore characteristic vocabulary of the two aforementioned man..dalas.
Equal precision must also be reached with regard to the derivatives 1 Five notably are Indra verses.
2
Abhipramt~r-, it will turn out, belongs to a different root. Its appearance at X 115.2c seems to support this view. 3 Turwt~a, Turvfi~a Y~du, Turvfi.4a Y~dva.