GEOGRAPHICAL EXPLOKATION OF JIOAB. 229 



through it. Tlierc is an amphitlicatrc ; the pavement of a forum, surrounded 

 by the bases of cohimus, is entire, 41 paces by 38, and just beyond it tlie 

 eastern gate of the city, outside which is an interesting little Doric temple, 

 12 yards by 10, facing east, the niches being still in sitti. 



The streets here have been arcaded ; and we found iu some places the flat 

 slabs of stone which formed the flooring of the dwellings above the streets 

 still entire. By the side of these old streets the ancient Khan looked but a 

 work of yesterday. 



We followed the Eoman road from Um "Weleed to Um el Kuseir. There 

 is no ruined bridge as marked iu the maps ; but there is a long massive wall 

 across part of the plain, built for the purpose of guiding the ffoods into the 

 cisterns. Um el Kuseir is of the same type as the last named city, but not 

 so extensive. 



Hence we struck eastward to Ziza, where we spent a week. It is men- 

 tioned in the ' N'otitia' as the headquartc}-s of the Dalmatian Illyrican cavalry. 

 The remains of Ziza are very perfect. The tank is simply magnificent, 140 

 yards by 110 (see Photograph); many of the stones are G feet in length. 

 Much engineering ingenuity is shown in the mode by which the upper valley 

 has been banked, and a system of sluice-gates arranged for filling the pool 

 and letting off the superfluous rainfall. 



Above it is a strong Saracenic fort, still entire, and which was occupied by 

 Ibrahim Pasha. Tiie upper story has been fitted for engines of war, and 

 many stones taken from Christian chambers marked with plain symbols appear 

 in the walls. The ancient city is on a long ridge further up, occupying 

 several acres, and full of sculptured ruins. The whole hill is honeycombed 

 Avith cisterns. The principal remains seem not earlier than the Christian 

 period, comprising several churches. 



Sis miles east of Ziza we crossed the Hadj road, not far from the base of 

 the limestone range which forms the eastern limit of the highlands of Moab. 



A little beyond this, at the very base of the hiUs, but without any trace 

 of water, we discovered a palace which surpasses in interest any other of the 

 niins which this expedition has brought to light. Prom the eminences near 

 Ziza we had detected a pile of masonry in this direction ; the Beni Sakkr gave 

 it the name of Mashita, and spoke of it as being like the other ruinous 

 heaps which we were continually examining. 



A gazelle had beguiled our ride, and not a little were we startled when 

 we reined iu our horses in front of a facade of which only the photographs 

 can give the slightest idea. Two days were well spent in photographing and 

 measuring (see Plan and Photographs). We were in utter perplexity as to 

 the origin of these magnificent buildings ; nor was our difficulty lessened by 

 the long lines of inscriptions in an unrecognized character on the lower coriieri 

 outside the inner palace. One thing was plain, the palace had never been 

 finished, at least in its decorations ; and we have to thank Mr. Fergusson for 

 having given us the clue to the solution of the problem. Mr. Fergusson is 

 decidedly of opinion that it is the work of Chosroes II., the Sassanian king 

 of Persia, after his conquest of Syria, jS'orth Arabia, and Egypt in a.d. 611- 

 622. The builders seem to have been interrupted, for it is evident that the 

 decorations were never finished. This is explained by the advance of the 

 Emperor Ileraclius, who so brilliantlj- swept the Persian out of the whole of 

 his conquests, and recalled for a moment the glories of old Home. 



There are no more ancient remains of any kind in the neighbourhood, 

 and no Saracenic additions whatever. Mashita stands forth in absolute 

 solitude and isolation; unlike the cities of Moab, with their traces of many 



