Gutenberg's patrices, according to Enschede, were made like bookbinders' stamps, of yellow copper, i. e., brass. With such patrices only lead matrices could be made, but the latter could be produced in two ways. Molten lead could be poured over the patrices or the patrices could be pressed into cold lead. The first mode is somewhat complex, but the matrix would have a smooth surface and need no further adjustment. The second mode is more simple, but required great force, although lead is a soft metal. Moreover the surface of the matrix would have to be trimmed, as the impression forces the metal downwards and sidewards, which makes the surface uneven, though by this pressure the lead becomes firmer and more compact, to the advantage of the type-founder. Enschede thinks that Gutenberg obtained his matrices by the second mode. He arrives at this conclusion by reason of the fact that Gutenberg's types were sharper in their impressions than Coster's. Developing this theory, he believes that Gutenberg had each letter engraved on a brass plate 2 mm. thick, therefore a mere letter without anything underneath it. This brass letter patrix was pressed, by means of a small flat plate, so far into the lead that its back formed an unbroken plane with the top surface of the lead, and was then removed. After the matrix had been made this way, the type were cast, which was done, not by pouring metal into the matrix, but by pressing the latter into semi-fused metal. In this way a great many letters could be cast from one matrix without any injury to it. Gutenberg's method was to cast in two tempos, according to Enschede, that is, the character was cast first and the shank was cast by another operation joining it to the character. Enschede warns us, however, that his theories are simply those of a practical founder and not a bibliographer's. But since no tools used by those early printers and type-founders have come to light to prove or disprove him, his theory is as valuable as any others advanced as to the methods used for casting type in those primitive days of printing. The shape of the type used as early as 1470 does not seem to differ materially from those of the present day. This is evident from old type which were discovered in 1878 in the bed of the river Saone, near Lyons, opposite the site of one of the fifteenth century printing-houses of that city. Also a page in Joh. Neider's "Lepra Moralis" printed by Conrad Homburch in Cologne in 1476 shows the accidental impression of a type pulled up from its place in the course of printing by the ink-ball, and laid at length on the face of the form, leaving its exact profile indented upon the page. This accidental imprint shows a small circle, and it is presumed that the type were pierced latterly by a circular hole, which did not penetrate the whole thickness of the letter, and served, like the nick in modern type, to enable the compositor to tell by touch which way to set the letter in his stick.