926 THE AMERICAN NATURALIST. [VOL. XXXII. 
In the house with the main offices is a specially equipped 
composing room, where at present three typesetters and a 
head typographer give their entire time to the Bureau. Ina 
neighboring building the large cylinder press has been set up, 
as well as a paper-cutting machine in charge of a special 
machinist. With this staff of employees, we are able to print 
and issue nearly 100 different cards a day. For sorting the 
cards a double check system is used, which makes errors almost 
impossible. 
In regard to classing the cards, we were at first inclined to 
consider this of secondary importance, the arrangement being 
the concern of the user. We have found, however, that such a 
course would be simply disastrous. The entire bibliography is 
a structure growing by internal additions, as an animal or a 
plant does. Consequently, every element must have a definite 
destination. For a catalogue of current literature in pamphlet 
form such a chapter heading as Fauna of the Celebees would 
be far too detailed ; not so for a great catalogue destined to 
receive the contributions of many years. Indeed, we go much 
further and arrange the papers on the Celebees according to the 
animals dealt with. Thus there is a place reserved for the 
Lizards of Celebees ; possibly there are already cards at that 
place, possibly some will be added in the coming year. 
It is obvious that our central catalogue can be arranged with 
such detail ; but how can we duplicate such a catalogue in the 
hands of our subscribers? With such an overwhelming number 
of divisions, how can one find Celebees? Ina book one can 
have an index with references to the pages; can one not apply a 
similar system to cards by numbering the places in which cards 
may be entered? It is evident that such a course is perfectly 
practicable, and it is this that we have done with the single modifi- 
cation that we have preferred to choose such numbers as would 
permit us to enlarge our scheme at will. It is needless to 
explain how this last requisite has been reached. In the past 
such explanations have been given and have only served to ren- 
der complicated a very simple matter. It is not at all necessary 
to understand how logarithms are calculated to use a table for a 
definite mathematical problem, and so it is with our numbers. 
