94 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



from the place of junction of the two former. The 

 arrangement then is a triradiate one ; we shall call the 

 end of each cist which is nearest the common centre of 

 the tritaph " proximal,'' the remoter end " distal." The 

 tangential cists are composed each at the distal end of an 

 outer end stone about 3 feet wide, of two side stones or 

 monoliths placed on edge and measuring up to 8 feet in 

 length, and of an entrance at the proximal end. This 

 entrance consists of a pair of upright pillar stones one at 

 each side, measuring 18 to 24 inches across, and standing 

 within and partly overlapped by the large side stones. 

 Sometimes at the base of the pillar stones and stretching 

 between them is a flat slab, sill or step, and sometimes 

 traces of a low wall built of smaller stones. In no case is 

 this end closed by a single large end stone. 



Between the proximal ends of these tangential cists 

 and running out at right angles to them is the third or 

 radial cist of the tritaph. It its formed of two pairs of 

 small side stones each 1 to 2 feet across, and in all cases is 

 open at its distal end where in some cases there are a few 

 steps, formed of rough slabs, leading down from the 

 surface of the ground to the floor. No gateway is present 

 in the radial cists, which thus differ constantly in several 

 points of structure — of sides, and of both ends — from the 

 tangential cists. The average size of the tangential cist 

 is 5ft. 9in. by 2ft. 8in., and of the radial cists 7ft. bj* 

 2ft. 3in. The floor, at a depth of 18 to 24 inches below 

 the present surface, showed in all cases indications of a 

 pavement of flat stones, on the average about 1 foot across 

 and 1 to 2 inches thick. All these stones, from the large 

 monoliths to the small gate posts and floor stones, are 

 of the grey clay slate of which the mountain is formed, 

 and some of which crops out close at hand with highly 

 inclined cleavage planes, so that probably lar^e slabs could 



