________________
APRIL, 1921 ]
THE ANCIENT GEOGRAPHY OF INDIA
125
II.-TRADITIONAL. Local traditions, when properly sifted and corroborated by other sources, give some geographical information. Thus Dr. Fleet's identification of Sakala with Sialkot is due to the local tradition recorded in Cunningham's ASR., XIV, pp. 44-46, that it was founded by Salya and that it was originally called Sākala.
But tradition sometimes turns out to be wrong. Thus Gaibandhā (Rangpur district claims to be the country of the Matsyas ; Badnagar (Patna district) to be Kundinapura (the capital of Vidarbha) ; but the Epics show otherwise. Hence uncorroborated tradition has little value. The literary sources also sometimes mention names which cannot be located. Again tribes die out and disappear ; towns decay and are deserted; seaside emporia sometimes shift ; "the names of countries” (cities, etc.) "change" [Bhattotpala's Commentary on the BỊ hatsanhita] though the places themselves survive. All these facts make indentifications of sites mentioned in foreign and indigenous literary sources difficult. Hence we have to turn to
III.-ARCHÆOLOGY with its three branches (i) Monumental, (ii) Numismatic, and (ii) Epigraphical.
(0) The monumental remains of a place enables one to compare its present ruins with those described in a foreign or indigenous source. Thus Mahāban was long taken as the site of Aornos; but Dr. Stein's survey has proved beyond all question that the natural features of that mountain are totally dissimilar from those of Aornos as described by the historians of Alexander. (Ann. Rep. of 4. 8. I., 1904-5, p. 42.] The existence of a double-chambered oavo answering to the description of Sudatta's proves the identification of Po-lu-sha with Shahbazgarhl. (Cunningham's ASR., V., 9-15.)
(ii) The discovery of coins sometimes enables one to locate a particular nation or tribe. Thus at Nägari, a small town 11 miles north of Chitor, have been found seven copper coins (found nowhere else) with the legend "Majhamikāya s(l)bijanapadasa" showing that the Madhyamikas should he located there. (Cun. ASR., VI, pp. 196 205.) But coins pase from one country to another and so identifications based on their places oi discovery may be wrong. Monuments themselves cannot enable us to indicate the real site, unless (a) an ancient description of the monument is found or (b) it speaks through an inscription. Hence for ancient geography, as for everything else connected with the past of India, we are really dependent on the (ii) Epigraphio records which regulate everything that we can learn from tradition, literature, coins, architecture or any other source. Thus when find a we pillar in situ bearing the inscription that" here was born the Sakya sage" we make an identification of which there can be no doubt.
Dr. Fleet classified the epigraphic records according to their topics thus -
A. Records making a plain statement of events : the Häthigumpha inscription, the Allahabad Pillar inscription of Samudragupta and a few other are due to an historical instinct.
B. Records due to Religious motives: the Piprawa relic vase; the Rock and Pillar edicts of Asoka, etc. To these are to be added a new group :
C Records of Religious endowment : Barabar cave inscription of Asoka, Bhitari Pillar, etc.
D. Records of Secular donation. To these are to be added a new group: E. Literary inscriptions (to preserve Kāvyas and Natakas).]