staple is sunk into the wood.

[Illustration: Fig. 105. A book in the Medicean Library, to shew
attachment of chain.]

The chains are made of fine iron bars about one-eighth of an inch wide,
but not quite so thick, flattened at the end of each link, and rounded in
the centre, where a piece of the same iron is lapped round, but not
soldered. Each chain (fig. 106) is 2 ft. 3 in. long. So far as I could
judge all the chains in the library are of the same length. There is a
ring at the end of the chain next to the bar, but no swivel.

[Illustration: Fig. 106. Piece of chain in the Medicean Library, of the
actual size.]

The ironwork by which these chains are attached to the desk is somewhat
complicated. By the kindness of the librarian, Signor Guido Biagi, I have
been allowed to study it at my leisure, and to draw a diagrammatic sketch
(fig. 107) which I hope will make it clear to my readers. The lock is sunk
in the central support of the desk. The bar passes through a ring on each
side of this support, and also through a ring near each end of the desk.
These rings are fixed to the lower edge of the desk just under the
molding. A flat piece of iron is forged on to the bar near the centre.
This iron is pierced near the key-hole with an oblong slit through which
passes a moveable piece of iron, here shewn in outline of its actual size
(fig. 108). The bolt of the lock passes through a hole in this piece, and
holds the bar firmly in its place.

[Illustration: Fig. 107. Diagram to explain the ironwork at the Medicean
Library.]

[Illustration: Fig. 108. Outline of bolt forming part of ironwork.]

The bar is not quite so long as the desk; consequently, when it has been
unlocked, and the iron bolt sketched above withdrawn, it can be turned
round by taking hold of the central iron, and pushed to the right or to
the left, past the terminal rings. The chains can then be readily
unstrung, or another strung upon the bar.

In the next chapter I shall describe the changes in Library arrangements
adopted during the period which succeeded the Middle Ages; but, before
ending this present chapter, there are a few points affecting the older
libraries and their organization to which I should like to draw attention.

In the first place all medieval libraries were practically public. I do
not mean that strangers were let in, but even in those of the monasteries,
books were let out on the deposit of a sufficient caution; and in Houses
su