their respective acquisition.

FOOTNOTES:

[37] Wheatley, _Pepys and the World he Lived in_, p. 84.

[38] I believe these rules were originally drawn up by Mr. B. R.
Wheatley.

_Classification of Books._

The classification of books, according to any set system, or according
to subjects upon the shelves of a library, is not easy, and for many
reasons it is not worth attempting. Unless the library is a very large
one, say, ten to twenty thousand volumes, with ample and adaptable
shelving, it is not to be desired. The main difficulty in shelf
classification lies in the fact that books on similar and kindred
subjects are issued in all sizes. There are books on Furniture, for
instance, in folio, in quarto, and in octavo. When shelf classification
is imperative, the folios are all put together, the quartos together,
and the octavos together. This is the nearest realisation of a shelf
classification, and by this method the folios may be far separated from
the quartos, and the quartos from the octavos. Moreover, if appearance
count for anything, as indeed it should in the most modest library, it
will be impossible to carry out any plan of shelf classification and
preserve at the same time an appearance of method and fitness. In
planning out how your books are to be placed, a great consideration is
the placing of them, so that books likely to be frequently referred to
shall be easy of access, and books less likely to be in request shall be
housed higher up.[39] Reference books should, as far as possible, be
placed together, and all easy of access. The main divisions into which a
private library classes itself are History and Biography, Fiction,
Poetry and Drama, Theology, Travel, Art, Belles lettres; but there are
so many considerations besides those of subject in any general
classification which should determine the position of a volume that I
must emphasise what has already been said about actual personal
convenience being first studied, and the library as arranged on the
shelves should be the result of personal convenience and graceful
effect. This is more particularly necessary when a library is in course
of expansion. The subjects which will expand quickest, and the space
they will require, can never be accurately gauged, and frequent
upheavals and readjustments will be necessary if any rigid plan is aimed
at. I would suggest that a separate shelf--or, if necessary, a separate
case--be reserved for unbound p