Wattenbach (_Schriftwesen_, _etc._) says that uncials date from the second century A.D. From examples still extant of the fifth and following centuries, it seems that while the Roman capitals were not uncommon, in Celtic MSS. the form generally adopted was the uncial. It was the form also usually chosen for ornamentation or imitation in those Visigothic, Merovingian, or Lombardic MSS., which made such remarkable use of fishes, birds, beasts, and plants for the construction of initial letters and principal words, of which we see so many examples in the elaborately illustrated Catalogue of the library at Laon by Ed. Fleury, and in that of Cambray, by M. Durieux. Most of these pre-Carolingian designs are barbarous in the extreme, dreadfully clumsy in execution, but they evince considerable ingenuity and a strong predilection for symbolism. Before concluding this chapter perhaps something should be said concerning the shape of books, though this is a matter somewhat outside the scope of our proper subject. Yet, as the brief digression will afford an opportunity for the explanation of certain terms used in MSS., we will avail ourselves of it. The ancient form of writing upon skins and papyrus was that of the roll. The Hebrew, Arabic, or Greek terms for this do not concern us, but its Latin name was _volumen_, "something rolled," and from this we obtain our word volume. Such words as "explicit liber primus" etc., which we often find in early MSS., refer to this roll-form; _explicare_ in Latin meaning to unroll; hence, apropos of a chapter or book, to finish. When transferred to the square form, or codex, it simply means, "here ends book first," etc. The modern book shape first came into use with the beginning of the Christian era under the name of codex. Here it will be necessary to explain that _caudex_, _codex_, in Latin, meant a block of wood, and had its humorous by-senses among the Roman dramatists, as the word block has among ourselves, such as blockhead.[5] So _caudicalis provincia_ was a jocular expression for the occupation of wood-splitting. [5] Terence, _Heautont._, 5. 1, 4. Whether the word had originally any connection with _cauda_, "a tail," is not here worth considering, as if so, it had long lost the connection; and when used to mean a book, had only the sense of a board, or a number of boards from two upwards, fastened together by means of rings passed through holes made in their edges.