________________
JULY, 1918]
THE HINDOLA MAHAL AT MANDU
177
This view receives further support from the historical evidence. MAndà was finally conquered by the Muhammadans in 1305, and was ruled by governors appointed from Delhi until Dilawar Khân declared himself independent in 1401. Like his predecessor, Dilê war Khân, although he spent several months of the year at Mândů, resided at Dhår, and Mandû only became the capital on the accession of Hûshang (1405-1434). Although Mândâ owed most of its splendour to Hûshang, we know from inscriptions that Dilâwâr Khân also erected buildings there, viz., the Târâpur Gates and the mosque named after him. This mosque, 34 as well as the Lat Masjid at Dhar, built by him in 1405, are both constructed, like the Hindola Mahal, of Hindu remains. I think, therefore, on historical and archeological grounds, that the Hindola Mahal was probably built in the first quarter of the fifteenth century, either by Dilâwar Khân or Hûshang, the chances being in favour of the former.
Before closing I give here (Plate III, B) an interesting example of an aiwan vaulted in this fashion.35 This building stands at Machi, near Hauzdâr on the trade route through Seistan, and was perhaps built about A. D. 1600.
Postscript.-In the first part of this article, when speaking of Kaşr Khâraneh, I said that Prof. M. van Berchem did not accept the view that it is a Muhammadan building. His reasons for doing so are as follows. Ķuşair Amrah is not fortified, and no one would expect it to be since the Umayyads, masters of a great part of Asia, had no need to fortify their residences when these lay well within the boundaries of their eni pire. On the other hand Kharaneh as well as Mashita, Taba, 'Amman and Abyad, are fortified buildings of Mesopotamian type adapted to the plan of the castra of the Roman limes, such as Qastal. He discusses the matter at considerable length, and concludes that the hypothesis which attributes these buildings to the Ghassanides, who were Syrians and Christians, is improbable owing to the complete absence of Christian symbols and the presence of fragments of images at Mashita, as well as the Dionysiae interpretation of the façade suggested by Clermont Ganneau. Further, Mashsta and Kharaneh are strongly influenced, if not entirely inspired, by Persia and Mesopotamia. Now the Lakhmids were of Mesopotamian origin, and the inscription of Nemara proves that at the commencement of the fourth century the Lakhmid king Imrulqais was allied with the Romans and Persians, and that his rule extended to the Roman boundary east of the Hauran. He may well have felt the need of frontier forts and Taba, Ammân, Mashita and Kharaneh form just such a line of advanced posts towards the west. Prof. van Berchem therefore places them in the fourth century. I submit that in the series of buildinga dealt with in this article, Kharaneh, by reason of the absence of windows in the bays, falls more naturally into the fourth than into the eighth century, and that support is thus lent to Prof. van Berchem's theory.
33 Zafar Hasan. The Inscriptions of Dhår and Manda: Epigraphia Indo-Moslemica, 1909-10, pp. 18-19.
4 Barnes (Capt. E.), Dhar and Manda: Journ. Bombay Branch of the Royal Asiatic Society. Vol. XXI, p. 384.
* Tate (G. P.) Seistan, plate to face, p. 136.