46 THE OIL-FIELDS AND THE FIRE-TEMPLE OF BAKU
After this long, and, I fear, somewhat technical, disquisi- tion, I turn to the more attractive problem of determining the possible age of the temple and its buildings. I may state at once that I used to hold the generally current opinion that the sanctuary was of Gabr, or Parsi, origin — a Zoroastrian fire- temple.^ Further study of the subject has forced me to aban- don this view (certainly for the present temple) as the follow- ing paragraphs will show.
So far as my researches go, I have not been able to find any allusion to a temple on the site in the classic writers of Greece and Rome ; ^ nor in the early Armenian authors ; ^ nor do the medieval Arab-Persian geographers refer to it, as we might
��Kaspischen Meeres in den Jahren 1860- 1861, St. Petersburg, 1905. I have ■worked throughout from photographs and personal inspection of the tablets, having secured Dorn's work only after this chapter was set up.
1 See Jackson, Notes of a Journey to Persia, I, in JAGS. 25. 176-178.
2 The Byzantine writer Priscus (d. after 471 a.d.) incidentally refers to the region of the naphtha wells of Baku, but not the temple, in an allu- sion to ' the flame that comes out of the rock beneath the sea ' (ttjv Ik t^s ii4>d\ov TT^rpas dvacpepofiivTjv <f)\6ya) in Frag. 8 (ed. C. Miiller, FHG. 4. 90 b, and ed. Dindorf, 1. 313). The state- ment by Henry, Baku, 1. 25, who cites Gibbon (ch. 46) as his authority, to the effect that the Roman emperor Heraclius appears to have destroyed the temple at Baku along with others, is due to a misconception. The temple which Heraclius destroyed was at Shiz (now Takht-i Sulaiman), and was not the sanctuary at Baku ; see Jackson, Persia, pp. 141-142.
8 One might be inclined to see a reference to Baku in the allusion by the Armenian writer, Moses Khorene
��(5th century or later), to a fire altar in Bagavan that was fostered by the Sasanian king Ardashir. The refer- ence {Mos. Khor. 2. 77) reads in the German translation by Lauer, Ge- schichte Gross-Armeniens, p. 136, Re- gensburg, 1869, 'Auch vermehrt er [d. h. Artaschir] noch den Tempel- dienst und befiehlt, das Feuer des Ormist, welches auf dem Altar in Bagavan war, ohne erloschen leuchten zu lassen.' The modern Armenian author Alishan {Hin Havatk gam Hetanosagan gronk Hayots [i.e. An- cient Beliefs, or Pagan Religions of Armenia], pp. 50-51,302, 432, Venice, 1895) understands the passage as mean- ing Baku ; but both the German au- thorities, Htibschmann (Altarmenische Ortsnamen, in Indogerinanische For- schungen, 16. 411) and Marquart {Erdnsahr, p. 110, n. 2), seem to allow of no question that Bagavan (Bagvan) is a small village, now Turkish Uich- Kilise, near Diadin and Mount Npat (Ala-dagh), north-north-east of Lake Van, a hundred miles from Baku. So that idea (otherwise attractive) may be dismissed as misleading.
�� �
