have three scenes of a dramatic performance, which takes place at a Soma sacrifice to celebrate the victory of Indra over the serpent Vṛtra, ending with a dance of the Maruts, represented by youths fully armed. This weapon dance is a relic of old vegetation ritual, the driving out of the old year, winter, or death, which is the foundation of the dances of the Roman Salii, the Greek Kouretes, the Phrygian Korybantes, and the German sword dancers. How can it be justifiable to spin theories thus in order to explain hymns which are taken by themselves without serious difficulty save in detail?
It is equally impossible to find any cogency in Dr. Hertel's arguments from the necessity of assuming two sets of performers, since the hymns were sung and a single voice in singing could not distinguish the interlocutors. Doubtless, if we accepted this necessity, we would be inclined to admit a priori that the song would tend to be accompanied by action and by the dance, so that drama would be on the way to development. But we do not know that the hymns of the Ṛgveda were always sung; on the contrary we do know with absolute certainty that, while the verses of the Sāmaveda were sung (gai), the verses of the Ṛgveda were recited (çaṅs). True, we do not have precise information of the exact character of the recitation, but there is not the slightest ground to suppose that a reciter could not have conveyed by differences in his mode of recitation the distinction between two different interlocutors, and the fact that this point is ignored in the argument is fatal to it. Moreover, we must admit that we are wholly ignorant as to the degree in which it was desired by the authors or reciters of these hymns to convey these differences of person. We do not know, and the ritual text-books did not know, exactly in what way these hymns were used. We find in the Ṛgveda a number of philosophic hymns; why should we not admit that a philosophic dialogue such as that of Yama and Yamī is possible without demanding that it should be a fragment of ritual? We have historical hymns in Maṇḍala vii; why should we turn the dialogue of Viçvāmitra and the rivers into a drama? Why should we insist that all hymns were composed for ritual use, when we know that ancient tales were among the things used to pass the period immediately following the disposal of the dead, and
