60 H. C. Lmvlor on 



river for between a quarter and half a mile, into Gillhall demense 

 one notices that the enclosing sides of the valley converge into a 

 narrow defile where at some prehistoric period was the outflow of 

 the ancient lake, the waters of which wore into a soft section of the 

 rock, gradually wearing down what is now a deep cut in the river's 

 course, and by nature's action drained off the lake. This process 

 may have taken thousands of years to complete, and the fact that 

 there being two sepulchral monuments on the site of what was once 

 the bottom of the lake, of an age of possibly 2,000 years, shows 

 that the lake had disappeared at least as long ago. 



It would be quite easy to-day for engineers to erect a bank 

 across this narrow place, and we can at once see what the conse- 

 quence would be. The lake would reappear and extend right 

 up to aboiit the town bridge (;f Dromore. It would also branch 

 ofiF north up the flat valley of which the stream which flows into 

 the Lagan at Phil's Fort forms the middle, extending the lake for 

 a long way also in this direction. I think this stream is known as 

 the Coolsala Burn. Thus for military purposes the lake would 

 form an impregnable barrier for a mile-and-a-hall in a line due 

 west from Dromore, and possibly half a-mile north from Phil's 

 Fort. The only weak point in the defence would be at the place 

 where the Coolsala stream joins the Lagan, where the lake would 

 narrow down to, I think, perhaps 30 yards. Can we doubt that 

 when the northmen, under Tergesius, steadily advanced in the 

 ninth century, they saw this advantage, and built the dyke re-form- 

 ing the lake ? At the narrow of Coolsala, or the corner of the 

 Salleys, they saw the weak spot and threw up a fort to guard it ; 

 at the upper end of their newly-made lake they built two other 

 forts, one where the mount of Dromore now is, the other quite 

 near, guarding the next narrowing of the Lagan. It can be also 

 readily inferred that at this narrow defile another dyke was built, 

 flooding the valley east of Dromore for a considerable distance, 

 how far I have n(/t investigated on the spot, but at intervals of 

 from a quarter to half-a mile eastwards along the river are a num- 

 ber of other ring forts, all evidently forming with the dyked-up 

 river a chain of fortified defences four or five miles long, which 



