There is, however, another matter which must be taken into account. The palace of Ukhaiḍir could not have satisfied the needs of any but a very primitive society. It contains no bath, that indispensable requisite of existence in hot climates, nor any sanitary arrangements whatsoever. Moreover the seclusion of the ḥaram courts is very imperfect, a fact which points to a primitive stage of Islâm. It is true that the ḥaram courts are separated from each other and from the central court of honour, but they are overlooked by the windows of the two upper stories of the northern block, which must have belonged to the public part of the palace. Doorways open from the first floor on to a roof which is continuous with the roof of the ḥiram lîwâns, and even if low walls divided the roof spaces, the guests or guards who were lodged in the upper story had an uninterrupted view into all the courts below. When I first visited Ukhaiḍir I found it inhabited by some Arabs from Djôf. The wives and families of the shaikhs had taken possession of the rooms on the first floor, where none of my servants were allowed to penetrate. They dwelt there because, if they had occupied the lower courts, their movements could have been observed from above. All these observations point to, or can be reconciled with, a date in the eighth century for the building of the palace, but whether it belongs to the late Umayyad or to the early Abbâsid period cannot be decided from internal evidence. The sister buildings on the western side of the desert are Umayyad, but on the other hand Ya’qûbi, writing towards the close of the ninth century, mentions the fact that the castles of the Abbâsid khalifs were situated on or near the road to Mekkah. ‘He who wishes to travel from Kûfah to the Ḥidjâz goes out along the southern road by stations which are built and halting-places which are kept in repair, among which are the castles of the Hâshimid khalifs. The first station is Qâdisiyyeh.’[446] The Arabic word which I have translated ‘castles’ is _quṣûr_; it is the word which is applied to-day to the mud-walled palm gardens of the Baḥr Nedjef. Whether in this passage it should be taken to denote palm gardens or ḥîrahs situated along the Ḥadjdj road I do not know, but it is significant that, with the exception of Ukhaiḍir, no trace of any such ḥîrahs has remained to our day. Ukhaiḍir is not upon the road that runs from Kûfah to the Ḥidjâz, but neither is it more than two days’ journey removed from it. That the khalif Hârûn al-Rashîd carried his hunting expeditions into the region near Kûfah seems probable from the fact that it was on one of these occasions that he is said to have found the grave of the khalif ‘Ali at the spot which is now occupied by the city of Nedjef.[447] The story of the finding of the grave bears every sign of having been a legend invented by the Shî’ahs, but it lends additional colour to the supposition that the early Abbâsids frequented the eastern deserts in