and moderately-simple printed descriptions of the
life-history of the animals should accompany the specimens; therefore,
as it was clearly impossible to describe every genus, it became
necessary to fix on some mode of associating in groups a number of
examples to which the descriptions might apply. Such divisions as
'classes' and 'orders' were manifestly too large, whilst 'families'
varied from a single genus, including a solitary species, to an army
of more than a thousand genera--e.g, the Linnaean families
Cerambycidae and Curculionidae in the Coleoptera. It was with some
regret that the idea of attaching a readable sketch to each division
of a given rank in recent systems of classification was relinquished;
but it was found to be impracticable, and the life-history sketch thus
became the foundation of the arrangement eventually adopted.

"Whether it might be a few species, or a genus, or a family, or an
order, that seemed to afford suitable scope for a page of readable and
instructive matter, it was decided that, throughout the entire
collection, such a group should be segregated, so as to form the unit
of the series. Eventually, in order that the sketches, which it was
proposed to print for that purpose on tablets, might all be in
positions where they could conveniently be read, it was found to be
expedient that each group or unit should occupy an equal space; and as
the blocks on which the table-cases rested were to be fitted up with
trays or drawers, twelve of which would occupy the table-case without
loss of room, these trays or drawers were adopted as the receptacles
and boundaries of the groups.

"The entire plan of the table-cases, and the limits of many of the
groups, were committed to writing before any considerable advance had
been made in procuring specimens. In one respect this circumstance was
found to be very advantageous--our desiderata were at once well
defined. It was an object that each of the groups should be
illustrated by carefully-selected specimens, and, until this could be
attained, other acquisitions need not be sought for. In making
purchases, such an object, steadily kept in view, exercises a powerful
influence against the seductive attractions of 'great bargains,' which
often turn out to be great misfortunes to a museum. Moreover, in
accepting donations, it is sometimes convenient to be able to refer to
a fixed plan. Where room is scanty, as in most museums, nothing is
more subversive of