built in 1506 for the
College de Navarre, Paris, now the Ecole Polytechnique[326]. My
illustration (fig. 64) is from a photograph taken shortly before its
destruction in 1867. I have calculated that it was about 108 ft. long by
30 ft. wide.

The library of the College d'Autun, Paris, was similarly arranged. An
inventory taken 29 July, 1462, records: "dix bancs doubles, a se seoir
d'une part et d'autre, et ung poupitre; esquelz bancs et poupitre out este
trouvez enchaisnez les livres qui s'ensuyvent, qui sont intitulez sur la
couverture d'iceulx[327]." The catalogue enumerates 174 volumes, or rather
more than 17 for each "banc" or lectern. The expression _bancs doubles_ is
interesting, as it seems to imply that there were at that time libraries
in which _bancs simples_ were used; that is to say, lecterns with only one
sloping surface instead of two.

A study of the catalogue drawn up in 1513 for the Augustinian House of S.
Victor, Paris, by Claude de Grandrue, one of the monks, shews that the
same system must have been in use there. Further, his catalogue is an
excellent specimen of the pains taken in a large monastery to describe the
books accurately, and to provide ready access to them. A brief prefatory
note informs us that the desks are arranged in three rows, and marked with
a triple series of letters. The first row is marked A, B, C, etc.; the
second AA, BB, etc.; the third AAA, BBB, etc. To each of these letters are
appended the numbers 1, 2, 3, 4 and so on, to shew the position of the
required volume. For instance--to take one at random--_Abaelardi confessio_
is marked P. 13: that is, it is the thirteenth book on the desk in the
first row marked P. When the catalogue proper--in which each manuscript is
carefully described--was finished, the author increased its usefulness by
the composition of an alphabetical index[328].

How, I shall be asked, can the form of the bookcase or desk (_pulpitum_)
be inferred from this catalogue? I reply: In the first place, because
there are no shelf-marks. The librarian notes the letter of the desk, and
the place of each book on it, but nothing more. Secondly, because the
number of manuscripts accommodated on each desk is so small. There are 50
desks, and 988 manuscripts--or, an average of little more than 19 to
each. At Zutphen the average is exactly 18. This piece of evidence,
however, is so important that I will give it in detail. The following
table, compiled by myself from