From this it is clear that Egypt (_a_) can lay no claim to have invented this method, and (_b_) had travelled many stages on the long road of piscatorial invention. The complete absence in the Nile Valley of hooks of bone, flint, or shell which occur in so many neolithic centres in other parts of the world adds confirmatory evidence. In Egypt no records of the progenitor of this copper Hook survive. No family tree helps us, as elsewhere, to surmise whether the thorn, the flint, or the shell constituted the material of the first hook, for no non-metallic prototype has come to light. The numerous bone and ivory points, all more or less like the slender rod or pin of ivory shown in _El Amrah and Abydos_,[779] may, perhaps, indicate the gorges used by fishermen in pre-dynastic times. The absence, however, in the above example of any indentation in the middle, round which the line was frequently attached, tends (in my view) rather to negative the suggestion. The earliest hooks were of simple shape. The point was barbless. The head, which in all cases lay in the plane of the hook, was formed by doubling over the end of the shank against the outside of the latter, so as to form a stop or an eye, which might, or might not, have been an open one.[780] Their length (varying from 2 to 6 cms.), if contrasted with the bronze hooks of the Swiss Lakes, is short in proportion to their width from the outside of the point to the outside of the shank.[781] The XIIth Dynasty displays a few barbed hooks alongside barbless ones. One of the latter, belonging to Petrie, excites our interest, for the string of its attachment (some nine inches in length) is composed of double stout twist, while another proves itself the ancestor—in fact itself is—the Limerick hook with a single barb. By the XVIIIth Dynasty barbed hooks, usually of bronze, largely predominate. Instead of being headed up in the older fashion they show the end of the shank expanded, so as to form a small flange in a plane at right angles to that of the hook. A line bent on the shank below this flange (even if slight), and drawn hard up against it had the advantage of chafing less than when made fast to a hook of the earlier type. The New Kingdom hooks, which continue scarcely altered in Roman times, are well designed, but their barbs are less intelligently placed than are those of the Middle Kingdom.[782] But even in Roman times several types of hook, fairly well distributed in the Northern Mediterranean, seem unknown in Egypt; for instance, double hooks, barbed or barbless, of the Bronze Age in Switzerland, hooks with a split eye or an eye made by twisting the end of the shank round itself (as found in Crete) and many others are yet to seek.[783] The cluster or gang hook early confronts us in the tomb of Gem-Ni-Kai.[784] The fisherman here extends his index finger to feel the faintest bite: below the water the line ends in a cluster of five hooks, one of which holds a large fish.