In Patañjali, [51] the author of the Mahābhāṣya, whose date is certainly to be placed with reasonable assurance about 140 B.C., we find much more effective evidence bearing on the existence of drama. We learn from his criticism on a rule laid down by his predecessor Kātyāyana, as to the use of the imperfect tense of things which a person has himself seen, that it was normal to use in his time phrases describing a past event as if it had occurred before the eyes of the speaker; we can understand this only of a character in a dramatic performance of some kind, and it is significant that the phrase cited in illustration of the usage is ‘Vāsudeva has slain Kaṅsa’. The reference is to the famous legend of Kṛṣṇa, son of Vasudeva, and his wicked uncle Kaṅsa, who first sought to destroy him in his childhood, and afterward paid the penalty of his evil deeds by death at the hands of Kṛṣṇa. This notice receives further elucidation by a famous passage, first adduced by Weber, in which Patañjali explains the justification of the use of phrases such as ‘He causes the death of Kaṅsa’, and ‘He causes the binding of Bali’. [52] Both these deeds, the actual killing, the actual binding, are deeds of the remote past; how then can the present be in place? The answer, we learn, is that the events are described in the present because the sense is, not that they are being actually done, but that they are being described. Of the modes of description no less than three are then set out. In the first place we have the case of the Çaubhikas or Çobhanikas, who before the eyes of the spectators actually carry out—naturally in appearance only in the first case—the killing of Kaṅsa and the binding of Bali; they represent in fact by action, without words, so far as this passage formally tells us, the slaying of the wicked Kaṅsa, the binding of the evil Bali. Secondly, we have the painters; they describe by their paintings, for on the canvases themselves we see the blows rained on Kaṅsa and the dragging of him about; a painter, that is to say, kills Kaṅsa and has Bali bound by painting a scene describing these incidents. Thirdly, we have those who use words, and not action of the Çaubhika type, the Granthikas; they also, while relating the fortunes of their subjects from their birth to their death, make them real to the minds of their audience, for they divide themselves into two parties, one set adhering to Kṛṣṇa, and one to Kaṅsa, and they adopt different colours, the adherents of Kaṅsa black, and those of Kṛṣṇa red, though, by what is probably an erroneous correction, the colours are ascribed in the inverse order by many of the manuscripts.