nd forming a library,
the arrangement of which he himself superintended. But the final
subversion of the Eastern Empire, and the capture of Constantinople in
1453, dispersed the literati of Greece over western Europe, and placed
the literary remains of that capital at the mercy of the conqueror. The
imperial library, however, was preserved by the express command of
Mohammed, and continued, it is said, to be kept in some apartments of
the seraglio; but whether it was sacrificed in a fit of devotion by
Amurath IV., as is commonly supposed, or whether it was suffered to fall
into decay from ignorance and neglect, it is now certain that the
library of the sultan contains only Turkish and Arabic writings, and not
a single Greek or Latin manuscript of any importance.

Such is a brief survey of the most celebrated libraries of ancient
times. Before we proceed to describe those of modern days, we shall
offer a few remarks on the extent of ancient as compared with modern
collections of books. The National Library of Paris contains upwards of
824,000 volumes, and is the largest in existence. It will be easy to
prove that it is the largest that ever has existed.

The number of writers, and consequently of books, in the bright days of
Egypt, Greece, and Rome, could not have been very great. It must, on the
contrary, have been limited by various causes, which contributed
powerfully to retard the composition of new works, and prevent the
multiplication of new editions. In fact, the histories of cities and of
nations, together with descriptions of the earth, which have become
exhaustless sources for the writers of modern times, must have been but
sterile themes at a period in which history was confined within the
limits of a few centuries, and hardly a sixth part of the world now
known had been discovered. Add to these considerations the difficulties
of communication, by which the inhabitants of different countries, and
often those of different sections of the same country, were kept apart,
together with the number of arts and sciences which were either wholly
unknown, or confined within very narrow bounds, and it will become
evident, that for every thirty or forty authors of the present day,
ancient Europe could hardly have supported one or two.

Another circumstance which may be adduced in support of our proposition,
is the fact, that an increase in the number of readers leads to a
proportionate augmentation in the number of works