The four editions of the _Speculum_ are, of themselves, presumptive evidence that each edition was printed from types. It is improbable that the printer would re-engrave blocks for a second edition when those of the first were in existence. If the first edition had been printed from types, and the types had been distributed, as is customary, the printer was obliged to reset them in order to make the second edition. These four editions were certainly the work of the same printing office, and, without doubt, of the same printer, for [p275] the engravings are the same, and the types, ink, paper, and workmanship have similar defects and peculiarities. The first edition shows pages of types only; the next edition has types and blocks, but the types are like those of the first; then comes a third edition in the same types, but with two pages of types differing somewhat as to body and face; lastly an edition entirely in the old types, in a worn condition. Each edition has more or less connection with the others.[137] [Illustration: English. Two-line Brevier.] The body or dimension of the types used in the _Speculum_ approximates the size known to all British and American printers as English; but it is rather larger than any of the modern standards. It is really intermediate between the body English and the little-used body of Two-line brevier or Columbian.[138] The appearance of twenty engraved pages in the second edition of the _Speculum_ cannot be explained with satisfaction. Bernard thinks that these pages are the relics of an earlier edition engraved, or at least attempted, on wood, which, for some unknown reason, were temporarily substituted for types. [p276] No trace of this imaginary edition has been discovered. It has been claimed that the engraver of these xylographic blocks was the probable inventor of typography. It is supposed that he matured the ideas he had cherished about movable types when he was engraving and printing the first edition of the book; that when he became fully convinced of their feasibility, he stopped the engraving of the blocks, and finished the work with types which were made for the purpose. This hypothesis is not reasonable. If the printer of the book suddenly abandoned blocks for types, the change would be abruptly marked in his work. The twenty pages at the beginning of the book would be xylographic, and all following would be typographic. But it will be perceived that the twenty pages are scattered, without any order, throughout the book. Instead of being the relics of an earlier edition, it is demonstrable that these xylographic blocks were cut from transfers obtained from a typographic edition. A traced drawing upon transparent paper, taken with accuracy from the first edition of the _Speculum_, and carefully laid over a corresponding xylographic page in the second edition, will show an agreement in the length of lines, in the abbreviation of words, and in the copying of little errors or