The second form, the _dunn_--or _schmalnackiges Beil_--may be compared to a long and flattened almond with a small part at the pointed end removed and a narrow strip cut off from each side. The flatter surfaces nearly meet at the end opposite the cutting edge, leaving this end thin. The surfaces have become much more nearly flat, and the cross-section a rectangle with somewhat short ends and slightly curved sides. These belong to the period of the earliest stone graves or still earlier. They could be easily fastened in a wooden handle. This form is very common in Scandinavia. The third form, the _breit_--or _dick_--_nackiges Beil_, has almost exactly the shape of a thick chisel-blade with broad and thick back opposite the edge, and is rectangular in cross-section. It appears in the later megalithic tombs and the underground stone vaults or cists. [Illustration: FORMS OF PREHISTORIC AXE Hammer axes--Late Neolithic. Thin-backed axe. Dunn-nackiges Beil--Early and Mid-Neolithic. Palæolithic hand-stones--"Coups-de-Poing."] Late in the Neolithic period, usually after the introduction of copper, we find an axe--or "hammer-axe"--shorter and much thicker, somewhat in the shape of a very light stonemason's hammer, and with a hole for the handle. These axes sometimes had two cutting edges, sometimes one edged and the other blunt for hammering. Many of them were exceedingly beautiful in form, design, and finish. But this method of fastening the head to the handle greatly weakened the brittle stone. Many of them were probably merely articles of luxury or adornment. The hole was made by twirling a stick or bone, with plenty of sand, water, and patience. We have thus in the axes and the megaliths a well-established sequence of forms, but no means of fixing dates except at the beginning and end of the whole period. Apparently there was a long time between the Scandinavian shell-heaps and the fully established Neolithic culture, of which we have practically no records. Peculiar types of axes (except the mattock), and the megaliths do not occur in the province of the banded pottery, which itself will probably some day give us the clew to a system of chronology. The pottery of Thessaly, Thrace, and certain parts of the Balkan Peninsula is being gradually synchronized with that of Mycenæan and pre-Mycenæan Greece. Important discoveries seem reasonably certain in a not distant future. We can only wait for them with what patience we can assume.