MONDMENT OF JOSHUa's CONQUEST OF CANAAN." 247 



that it had none, and in any case it must have been quite 

 a recent boring. Again, south-eastward and south-west- 

 ward ak)ng the hills we had crossed there was no spring as 

 far as Ain Fakrun, eleven miles off one way, and probably 

 as far as Signs, ten miles oif the other way, while southward 

 there was not one up to the muddy well of Ain el Atash, 

 i'^h miles distant; and the little stream that flowed forth at 

 ■our feet spent itself in a couple of miles on the thirsty 

 ground. The hills near at hand looked arid, but especially 

 the further of two that encroached to the south upon the 

 plain of Behira el Twila ; from top to bottom it showed not 

 the slightest sign of vegetation, and was consequently known 

 ■as the Bald Mountain. 



Thus at Ain el Bordj were two of the special features of 

 the well of Tigisis most clearly displayed. The third was 

 <xlso visible. A large fragment of the city wall, about 

 20 yards long and 18 feet high, crowned the slope on the 

 east side of our pass, while fragments stretched along the 

 descending brow of the cliff until they almost touched 

 the Roman fountain. Right easy must it have been to pro- 

 tect the fountain from all intruders by means of archers 

 posted on the wall. And how substantial tliat wall was! 

 It had a facing of squared stones botli inside and outside 

 18 inches thick; Avhile between the facings lay 4 feet of 

 rubble stones, mostly of great size. The squared stones 

 that had once covered the rest of the wall Avere strewn over 

 the ruined site of the town in vast confusion. 



That site itself measured about a third of a mile from 

 south-east to north-west by a quarter of a mile from south- 

 west to north-east, and the series of Arab huts upon it and 

 as many more dismantled huts were walled with its frag- 

 ments and sometimes upheld by its pillars. 



The Arabs were eager to show me inscriptions, and I 

 copied in all seven Latin ones, which I show you to-day ; 

 but I could not hear of a single one in any characters but 

 Roman, although I carefully drew both Phoenician and cunei- 

 form characters and showed them to our guides as patterns. 



Our guides were tolerably exacting, for they made me 

 pay in all 4 francs 50 centimes for the privilege of looking at 

 and copying the stone writings ; but a man brought us a bowl 

 of sheep's milk and charged us nothing therefor. The 

 Arabs in Algeria regularly milk their ewes. It Avas the 

 second time in my life that I had tasted sheep's milk ; and 

 this time I found it delicious. 



