190 



Proceeding^ of t lie lioijal Im^h Acudeuiij. 



sides left jagged in the process of cliipping, leading to the type 

 of polished celt of flat-oval section with flatted sides — a form condi- 

 tioned by chipping — and compare with the latter the forms of the celts 

 produced by a different method of manufacture — that of battering 

 or pecking the stone into shape — a process which does not appear to 

 have been employed till an advanced stage of Neolithic culture had 

 been reached, and which necessarily eliminates the flatted sides, 

 producing a round-bodied celt, contracting towards the cutting edge, 

 and pointed at the butt-end — we cannot reasonably doubt that the 

 series representing development represents also progress in time, and 

 that these roughly chipped celts, ground only at the cutting edge, 

 although they may, in some cases, have been made in later times, 

 are, as a type, to be placed early in the IS'eolithic period of Ireland 

 rather than towards the middle or end of that period. 



If, then, the Larne celts are to be regarded as the roughed-out 

 pieces, or blanks, from which the celts of the kitchen-midden type 

 and the narrow forms with ground edges only were formed, the laying 

 down of the Larne gravels cannot, as far as this evidence can be relied 

 on, be brought down to a later period than the earlier stages of 

 Neolithic times. 



The Sandhill Sites. 



We may now approach the problem on a different and more direct 

 line of evidence. In the sandhills of the north coast of Ireland several 

 occupation sites of Neolithic man have been discovered. The most 

 notable of these are at Dundrum Eay, County Down, AVhitepark Eay, 

 County Antrim, and the mouth of the River Bann, near Portstewart, 

 County Londonderry. The sandhills overlie the raised beach ; and in 

 the old surfaces laid bare, in the wind-blown hollows among the sand- 

 hills, worked flints and pottery are found in great abundance. 



The industry at these sites is quite different in character from that 

 at the Larne beach. Long flakes of the Larne type are not found in 

 the sandhills ; indeed, flakes of any considerable length are rare at 

 those sites. Ilouglied-out pieces, such as the Larne celts and the Uitchen- 

 midden and narrow-axe types, are also wanting. Yet the w orking of 

 flint was carried on extensively in the sandhills. Cores and flakes are 

 abundant, but the nodules worked were, as a rule, small ; and the 

 flakes for the most part are short and broad, or irregular in shape. 

 Scrapers occur- in great numbers, and this was one of the chief im- 

 plements manuiactured. At Dundrum Lay and Portstewart well- 



