Owing to the perishable nature of the shaft, the discovery of a perfect arrow is exceedingly rare. There are, we believe, only two recorded instances in Ireland--one (here figured) in the King’s County, in Ballykellan bog, and the other near Glenarm, Co. Antrim. The head seems to have been fastened into the split of the shaft by a kind of cement, and then tied with sinew.[91] Much superstition has been attached to arrow-heads in Great Britain and Ireland. They have been regarded as ‘elf-shots’ or ‘elf-stones’ cast by the fairies at men or cattle: hence they have been often worn as charms, or used as such in curing bewitched cattle by putting them into their drink, which was supposed to remove the evil. Knives of flint are also found; the most finished examples are triangular in shape and pointed, with straight back, bevelled edge, and tang for insertion into a handle. Saws are numerous, and the serrated edges were either originally so chipped, or they became so by use. These are widely distributed, and have been found in Egypt, Greece, and through Europe to Denmark and the British Isles. [Illustration: Flint Knife.] Flaking tools or ‘fabricators,’ as they are called, were used in the making of flint instruments, examples of which may be seen in the National Museum collection. These are narrow pieces of flint 4 or 5 inches long, with blunt ends and a ridge on one or both faces. A deer’s horn or other pointed bone was also used for flaking; the finer work was probably usually done with such tools. The method was by pressure, and the finest flaking can thus be executed. Though this can be imitated with some practice, yet the best work on the faces of arrow- and spear-heads baffles imitation, showing the perfection of workmanship to which primitive man reached. Among other objects are ‘bracers’: these are pieces of stone rubbed down to a smooth surface and pierced with a hole at each end. They are supposed to have been worn on the wrist for protection against the bowstring when let home. * * * * * _Stone Celts._--There is a general similarity in type among the stone implements known as celts, or chisels and axes. As no well-defined line can be drawn between these, as the same object could be used as an instrument or weapon, Sir John Evans, in his exhaustive work on _Ancient Stone Implements_, uses the term ‘celt’ to include both, and classifies them under three heads: (1) chipped and not ground or polished; (2) chipped, with ground edge only; (3) those ground and polished completely. The last group is further sub-divided into four classes, according to shape. Some Irish archæologists consider that many of our implements present features which make this classification unsatisfactory; but as they are not yet agreed on any classification sufficiently clear and comprehensive, we shall retain the threefold division in our brief description.