THE ANCIENT BURZIN MITRO FIRE
��remains only as to the precise localization of the Fire of Burzin Mitro, which it would be natural to place at the ruin-crowned site, described hereafter, five miles north of the village of Mihr on our road. Yet, owing to a slight obscurity in the Pahlavi text above translated, we cannot be exactly sure from what point the ' nine parasangs towards the west ' were reckoned ; and Houtum-Schindler, starting from the Gunabad region, is inclined to suggest the village of Burzinan, near the north- western limit of the Nishapur province.^
��again we might compare the mountain cavern of Pardah-i Rustam, 'Curtain of Ruatam's (Harem),' near Gunabad, which we visited later in company with Major Sykes (see p. 299, below). It is not to be confounded with the Diz Gumbadan, at Gird-Kuh, north of Damghan (p. 172, n. 3). It is, however, identical with the mountain in the Avesta called Spento-data (Yt. 19. 6), after the hero's name, and the same as the Pahlavi Spand-dat, situated in the 'enclosure,' or district (Var), of Revand (Bd. 12. 2, 23). The identi- fication of this general section of the long and broken Ridge of Vishtasp is made still more clear by the fact that Lake Sovar, or Sovbar, of the Zoroas- trian texts, situated on Mount Kon- drasp, which mountain is mentioned directly after Mount Spand-dat in the Bundahishn (Bd. 12. 24, cf . Bd. 7. 14 ; 22. 1, 3 ; Zsp. 6. 22 ; and cf. Yt. 19. 6), is the same as the shrunken sheet of water Chashmah-i Sabz, in the Nisha- pur Mountains, according to Houtum- Schindler {Acad. 29. 313) and Sykes (Geog. Joum. Zl. 3) ; compare like- wise (though with certain differences) Justi, Beitrdge, 2. 16, and Stackelberg, PersiscTie Sagengeschiche, WZKM. 12. 239-240; see also, on Chashmah-i Sabz, Le Strange, pp. 386-387 ; Yate, Khurasan, p. 363. The additional
��evidence given by the village Guna- bad, across the Binalud Kuh, north- east of Nishapur, has already been referred to.
1 See Houtum-Schindler, Academy, 29. 312-313. It is appropriate to pre- sent both sides of the question more in detail, at the same time drawing attention to the slight uncertainty in regard to the accuracy of the text and of the correct understanding of the passage in its connection. Nine para- sangs, or farsakhs, would be a distance of about thirty-six miles, if we are to use as a standard the Khurasan farsakhs, which are notoriously long (see Cur- zon, 1. 245, n, 1). In the manuscripts (with the exception of the Iran. Bd,, which uses the figure for '9') the number 'nine' is spelled out in full, ndv [ = (ishga]. though an error in the archetype codex may not be wholly excluded. But from what point are we to reckon ' nine parasangs toward the west ' in the connection in which it stands ? It would seem natural to say from some point in the Gunavat territory ; and this is what led Hou- tum-Schindler to suggest the village of Burzinan as a possibility, a suggestion which might be more easy to accept if we point to the fact that at least two of the Bundahishn manuscripts have not the word Rivand directly before
�� �
