It is further a noteworthy fact that the capital of Kosala is in the original Ramayana regularly called Ayodhya, while the Buddhists, Jains, Greeks, and Patanjali always give it the name of Saketa. Now in the last book of the Ramayana we are told that Rama's son, Lava, fixed the seat of his government at Çravasti, a city not mentioned at all in the old part of the epic; and in Buddha's time King Prasenajit of Kosala is known to have reigned at Çravasti. All this points to the conclusion that the original Ramayana was composed when the ancient Ayodhya had not yet been deserted, but was still the chief city of Kosala, when its new name of Saketa was still unknown, and before the seat of government was transferred to Çravasti.
Again, in the old part of Book I., Mithila and Viçala are spoken of as twin cities under separate rulers, while we know that by Buddha's time they had coalesced to the famous city of Vaiçali, which was then ruled by an oligarchy.
The political conditions described in the Ramayana indicate the patriarchal rule of kings possessing only a small territory, and never point to the existence of more complex states; while the references of the poets of the Mahabharata to the dominions in Eastern India ruled by a powerful king, Jarasandha, and embracing many lands besides Magadha, reflect the political conditions of the fourth century B.C. The cumulative evidence of the above arguments makes it difficult to avoid the conclusion that the kernel of the Ramayana was composed before 500 B.C., while the more recent portions were probably not added till the second century B.C. and later.
This conclusion does not at first sight seem to be borne out by the linguistic evidence of the Ramayana, For the epic (arsha) dialect of the Bombay recension, which is practically the same as that of the Mahabharata, both betrays a stage of development decidedly later than that of Panini, and is taken no notice of by that grammarian. But it is, for all that, not necessarily later in date. For Panini deals only with the refined Sanskrit of the cultured (çishta), that is to say, of the Brahmans, which would be more archaic than the popular dialect of wandering rhapsodists; and he would naturally have ignored the latter. Now at the time of the Açoka inscriptions, or hardly more than half a century later than Panini, Prakrit was the language of the people in the part of India where the Ramayana was composed. It is, therefore, not at all likely that the Ramayana, which aimed at popularity, should have been composed as late as the time of Panini, when it could not have been generally understood. If the language of the epic is later than Panini, it is difficult to see how it escaped the dominating influence of his grammar. It is more likely that the popular Sanskrit of the epics received general currency at a much earlier date by the composition of a poem like that of Valmiki. A searching comparative investigation of the classical Kavyas will probably show that they are linguistically more closely connected with the old epic poetry, and that they deviate more from the Paninean standard than is usually supposed.