482 MB. S. GEIEVE ON" THE DISCOVEET 



rience showed us that if it had. been raised over any thing, or was 

 the superstructure covering a place of interment, we could only 

 ascertain this by digging down to the living rock, which is about 

 3 feet below the original level of the sand at the outer edge of 

 the mound, and possibly is the same under the apex. As the 

 falling sand made this work very dangerous, we proceeded to re- 

 move about 12 feet of the upper part of the hillock, and then to 

 dig downwards, as circumstances permitted. By the end of 

 September fully one third of the apex had been dug off, and 

 every spadeful most carefully examined, so that not even the 

 minutest object could be passed. The same rule has been followed 

 all through ; and though the work went on so much more rapidly 

 during our first visit, it was entirely owing to the fact that we had 

 mostly pure sand to deal with : this contained not a vestige of 

 remains, and seemed as if deposited from the sand-pit. To give 

 some idea of the nature of the deposits as revealed by the sections 

 examined during the digging, we may state that the outside of the 

 cone is covered with grass, and beneath this with turf and blown 

 sand to a depth varying from one to five feet, the greatest 

 depth being at the north side of the apex * and gradually thinning 

 off all round to the outer edge. Below that is a series of strata, 

 composed principally of shells, which taper off from the apex 

 similar to the upper deposit, and underneath these is pure sand. 

 Where we began our excavations we found almost solid sand ; 

 then, after a few feet, we came upon a thin layer of shells near 

 the surface, which was at first only about an inch thick, but as 

 we worked inwards we found this line gradually getting thicker, 

 until near the summit it was composed of numerous layers which 

 were pretty clearly defined, though here and there they ran into 

 each other, and altogether were about 8 feet from top to bottom. 

 The greater part of the shells were those of the Limpet {Pa- 

 tella vulgata^ L.) ; however, others were intermixed ; and besides 

 these were a few bones, bone-implements, and oblong water-worn 

 stones of a slaty character, some of which, we suppose, have been 

 used as limpet-hammers, which we shall speak of presently. 

 Others have one end rubbed so as to form an edge, and are similar 

 in appearance to prehistoric implements from the Swiss lake- 

 dwellings, and also from Orkney, Shetland, and Wigtoushire. 

 There are also a few oval and nearly round stones that showed 



* The strong south winds have blown the sands northwards, causing the 

 accumulation on that side of the mound. 



