138 
part, or topical classification, comprises 9 
heads designated by odd numbers extend- 
ing from 3 (or 03) to 35, the numbers 05, 
09, 138, 17, 21, 15, 29 and 33 being left va- 
cant. The entire symbol consists of four 
figures, the first two being taken from the 
systematic, the latter two from the topical 
classification. 
The Committee states: ‘‘The numbers 
used to designate the subjects are scattered 
over a series extending from 203 [0,203] to 
6,635 in such a way that any branch or any 
subject may be in future completely divided 
and numbers applied to the divisions with- 
out deranging the system previously in 
use.’’ This statement shows clearly that 
the Committee not merely has failed to study 
the decimal system, which they condemn ; 
but has not even examined critically their 
own system ; for I believe that every atten- 
tive reader has already detected, in the 
course of my brief exposition, that no modi- 
fications nor interpolations aré possible, for 
the simple reason that all the numbers are 
occupied. There are, it is true, eight va- 
cant numbers, but these are, indeed, ‘ scat- 
tered,’ so that no new sub-divisions can be 
introduced, for at least 2 places are needed 
to sub-divide. How necessary new divis- 
ions are may be gathered from the fact 
that one single division 0,407 would receive 
according to our experience, about 2,000 
cards, annually. In ten years there would 
be twenty Library Bureau drawers filled 
with cards, all with the some identical sym- 
bol. No one, of course, consults a bibliog- 
raphy to know what has been published in 
regard to the Morphology of Vertebrates ; 
he wishes to know what has appeared on 
the liver, the teeth, the skull; he desires in- 
formation regarding Sphingidee, not a ram- 
bling list of 8,000 papers on Lepidoptera (in 
10 years there would be 8,000). We have 
at present on stock 400,000 cards dealing 
- with Morphology (which is not sub-divided 
_by the:committee); several times a day we 
SCIENCE. 
[N. S. Von. X. No. 240. 
have occasion to seek out a relatively small 
number of cards from this great collection. 
Thus, if anyone desires to receive the cards 
relating to the liver 14 cards would be 
gathered together for him and he would be 
charged 14 cents, including postage and 
packing. The entire internal organization 
of the Bureau depends upon a precision of 
registration, such as an outsider can 
scarcely conceive to be necessary. Con- 
stantly one has to turn to the stock to re- 
place a torn or soiled card, to return a 
duplicate to its place, or for some similar 
operation. Were we to try to use the Royal 
Society’s schedules we should have to 
double our staff, and should even then live 
in a state of confusion. that would be in- 
tolerable. 
It seems astounding that the Committee 
should propose a system which admits of 
no modification, since Professor Armstrong 
stated at the first Conference that the 
‘¢ whole question resolves itself into whether 
we can adopt an inelastic system.’’ But 
never was a system so inelastic as that pro-- 
posed by the Committee. The Dewey Sys- 
tem, the Cutter System and the Halle Sys- 
tem may all be regarded as rivals. That of 
the Royal Society has not attained to this: 
eminence, for it cannot claim to have even 
attempted to solve the initial problem, that 
of combining permanence with expansi- 
bility. Indeed, the Committee seriously 
proposes that their system might be revised 
and modified at intervals of five years! So: 
far, then, from the great problem having 
been solved, it has not been even discovered 
what the nature of the problem is. 
Quite as important as expansibility is the: 
quality of adaptability to special needs. We 
have seen that the new schedules require 
that the systematic portion of the symbol 
should always be placed first. In other 
words, the symbols can not be adapted to 
the needs of those who wish to use topical 
divisions as their primary classification. 
