I could cite many more examples of ancient engraving which would yield impressions on paper, either by pressure or friction.  But our business is with printing rather than engraving; I will, therefore, go back to the subject, and cite a very early and interesting example of stamping engraved letters on clay.  I mean the Babylonian bricks, supposed to be four thousand years old, mostly sun-baked, but some apparently kiln-burnt almost to vitrification.  Of these there are now many examples in England, added to our stores by the indefatigable researches of Layard, Rawlinson, and others.  These bricks, which are about a foot square and three inches thick, are on one side covered with hieroglyphics, evidently impressed with a stamp, just as letters are now stamped on official papers.

Another evidence of the same kind, and of about the same age, is the famous Babylonian cylinder found in the ruins of Persepolis, and now preserved in the library of Trinity College, Cambridge.  It is about seven inches high, barrel-shaped, and covered with inscriptions in the cuneiform character, disposed in vertical lines, and affording a positive example of an indented surface produced by mechanical impression.  Such cylinders are supposed to have been memorials of matters of national or family importance, and were in early ages, as we know by tradition, very numerous.  Stamped or printed blocks of lead, bearing the names of Roman authorities, are to be found in the British Museum.

Printing on leather was practised by the Egyptians, as we discover from their mummies, which have bandages of leather round their heads, with the name of the deceased printed on them.  And in Pompeii a loaf was found on which the name of the baker and its quality were printed.  Among ancient testimonies, one of the most interesting is that afforded by Cicero in his de Natura Deorum.  He orders types to be made of metal, and calls them forma literarum—­the very words used by our first printers; and in another place he gives a hint of separate cut letters when he speaks of the impossibility of the most ingenious man throwing the twenty-four letters of the alphabet together by chance, and thus producing the famous Annals of Ennius.  He makes that observation in opposition to the atheistical argument of the creation of the world by chance.

We have besides, in what is generally classed as a manuscript, a reasonable although disputed evidence of an elementary stage of printing; I mean the Codex Argenteus (or Silver Book) of Upsala, which contains a portion of the gospels in Mesogothic, supposed to be of the fourth or fifth century, the work of Ulfilas.  In this codex the first lines of each gospel and of the Lord’s Prayer are in large gold letters, apparently printed by a stamp, in the manner of a bookbinder, as there are indentations on the back of the vellum.  The small letters are written in silver.  The whole is on a light purple or violet colored vellum.