308 Proceedings of the Royal Irish Academy. 



(Stonefield) which was open about forty years ago, and many people 

 Avho had been in it came to show where it was. Notwithstanding 

 their assistance we searched in vain all the morning, and it was well 

 on in the afternoon before the crowbar indicated a large flag, which 

 proved to be the top of the chamber. (Plate VII., fig. 4.) Admittance 

 was gained by removing some of the stones at one side of this flag, 

 and letting down a ladder into the chamber, on the floor of which 

 was a heap of stones which had evidently been thrown down through 

 an aperture in the roof since closed. 



The passage from this chamber was terminated by a wall built 

 across it at a distance of 11 ft., which formed a step. Here the roof 

 of the lower passage was discontinued, and one could get up into 

 the upper passage, which was here enlarged, and carried back 

 nearly to the wall of the circular chamber. This part of the upper 

 passage was so widened as to form a sort of rectangular chamber 

 above the lower passage, an arrangement I have not found in any 

 other cave, although there is something like it at Loughcrew. This 

 upper passage turns nearly due north, just beyond the step, but from 

 this point onwards it has been filled up with stones, which were for- 

 merly thrown in at what was probably the entrance. 



As anyone in the lower part of this cave would be at a disadvan- 

 tage should an enemy get in at the entrance, I surmise that the 

 arrangement is intended to give those inside the circular chamber 

 a chance of not being smothered should the foe make a fire at the 

 entrance or in the upper passage, for it would be an easy matter to 

 stop up the hole between the two passages so as to exclude smoke. 



Indeed, in nearly all the caves I have examined, the entrance is 

 higher than the chambers, probably with a view to making it quite or 

 nearly impossible to smother those inside should anyone attempt 

 to do so. 



All the caves I have examined are excavated in small hills or hill- 

 sides, no doubt for the sake of drainage, and the chambers are never 

 under the highest point of the hills, but almost always well on one 

 side, where there is a good slope on the surface. 



The entrances which one might expect to find at the sides of the 

 hills are generally close to the top, so that it is necessary to get down 

 into them. This position of the entrances may in part be accounted 

 for by the supposition that in many cases they were connected with 

 dwellings on the surface, which would be made on the highest ground 

 for defensive purposes, and partly from the difficulty if not impossi- 

 bility of making smoke pass down so as to get into the chambers. 



