AN ANCIENT CHINESE CAPITAL — BISHOP 573 



the terreplein behind it. The latter sloped gradually and on the 

 whole evenly downward to the general level of the area inside the 

 walls; there was no sudden change of profile to indicate where the 

 terreplein had terminated and the inner face of the rampart had 

 begun (cf. fig. 1). That this did not represent the original condition, 

 however, appeared likely, for the following reasons. 



We found, in the first place, that this long interior slope was for 

 the most part covered with uncompacted earth, in places to a depth 

 of 4 or 5 feet and obviously washed down from the parapet above. 

 Secondly, at certain points beneath this layer were to be seen portions 

 of what appeared to be a continuous stratum containing Han roofing- 

 tiles, bricks, and potsherds and resting directly upon the tamped earth 

 of the original agger. These facts suggested, first, that the top of the 

 parapet had once been somewhat more than 10 feet higher (see ante) 

 than the terreplein; and, second, that on the latter had once stood 

 buildings, erected presumably on a level, not a sloping, surface. Thus 

 we might reasonably suppose that the rampart as originally constucted 

 had had, in addition to a parapet, a true terreplein and a distinct inner 

 face — the latter doubtless much lower than the outer one and now 

 completely masked by detritus washed down from above. 



As already stated, the cart road from Hsi-an Fu entered the ancient 

 city enclosure through a wide gap at the eastern end of the southern 

 wall. This opening, we felt certain for several reasons, indicated the 

 spot where once had stood a city -gate. None of the old Chinese maps 

 of Ch'ang-an appears to show a gate at this spot; nor does it seem 

 very probable that one should have been placed there ; for gates do not 

 ordinarily occur at the corners of Chinese city walls. Possibly the 

 gap in question may have been cut through the rampart at some time 

 subsequent to the original building of the wall. 



Be that as it may, the track here moimted quite steeply and once 

 inside the enceinte turned abruptly to the left or west. This change 

 in direction brought it parallel to a low rise in the ground directly 

 across the opening and a short distance inside or north of it. Whether 

 this rise represented what had originally been an earthen wall forming 

 part of the defences of the gate, we were unable to determine by inspec- 

 tion alone. But had there been such a wall, its effect would have been 

 to compel an attacking force, once it had broken through the gate, to 

 turn sharply to the left and thus expose its right or unshielded side to 

 an enfilading fire from the defenders. 9 



For some 400 yards or so to the west of the above gap we found the 

 gigantic agger displaying in general much the same profile and dimen- 

 sions as those just described. Then came a second wide opening, 



• Shields formed part of the equipment of the Chinese warriors of the time, as shown, for example, on the 

 famous Han "reliefs" (in reality incised drawings on stone slabs) from the province of Shantung and by 

 numerous passages in the surviving historical records. 



