Aveust 4, 1899.] 
use letters rather than numbers, the Com- 
mittee calls upon the scientific world to aid 
it in crushing out the work that has already 
been done. For this it must have weighty 
reasons, or it would be wrong for any govern- 
ment to favor such action. 
The Committee writes: ‘‘ As between a 
letter and a number no single final reason 
ean be alleged, but there are a number of 
considerations which led the Committee to 
prefer the letter. In the first place, the 
Committee divided Science into more than — 
two principal parts, and the tendency of the 
recent Conference was to add to the number 
of divisions. Hence a decimal system is 
inapplicable to the Primary Divisions, un- 
less by grouping together several of the 
Principal Divisions under one or more 
heads. Science is arbitrarily made to fit a 
system which does not naturally fitit. In 
the second place, it is convenient to have a 
single symbol for each Principal Science, 
whereas, if numbers were employed, two 
figures would be required, owing to the 
number of divisions. Lastly, the Com- 
mittee believe that fewer mistakes would 
be made in sorting the slips and cards if 
attention had to be paid to a letter and 
number rather than to a single but longer 
number. On the whole, then, the Com- 
mittee, decided to represent the Principal 
Sciences by arbitrary letters.’’ 
These are the arguments, but I refuse to 
believe that it will be generally felt that 
they justify the work of destruction which 
is proposed. But even the little weight 
these arguments may seem to have disap- 
pears when they are closely examined. The 
letters required for the sciences thus far se- 
lected for treatment run from A to R, thus 
permitting for the time being, it is true, the 
use of a single symbol ; but it is evident that 
this is possible simply because the proposed 
catalogue is to contain only certain sciences. 
Professor Carus has already pointed out the 
dangers of thus disregarding a general sys- 
“SCIENCE. 
135 
tem of notation compatible with the ad- 
junction of new branches. With the addi- 
tion of a few subjects, the 26 letters of the 
alphabet will be exhausted, just as the ten 
digits would be. This eventuality destroys 
the whole value of the first two arguments. 
The difficulty contemplated in the third 
argument has long since been met in the 
Concilium - Bibliographicum by separating 
the two figures indicating a Principal 
Science from those representing sub-divis- 
ions of that science, and by printing these 
figures by themselves as ‘the signature’ of 
the set. The sample card given below will 
illustrate this feature. 
This is the entire argument relating to 
the Principal Sciences. For the sub- 
divisions of a science the Committee pro- 
posed the use of numbers, but preferred 
other numbers than those already employed. 
‘* As to the system by which these numbers — 
should be chosen, the Committee had before 
them the deliberate decision of the Confer- 
ence of 1896, that the Conference was ‘ un- 
able to accept any of the systems of classifi- 
cation recently proposed,’ among which the 
decimal system was, of course, included. 
This system had, therefore, not been ac- 
cepted, and the Committee agreed, after 
further discussion, that it was not desirable 
again in any way to propose a decimal sys- 
tem. Such a system is open to two objec- 
tions. In the first place, it assumes that 
each subject is to be divided into not more 
than ten divisions, each of which may be 
divided into not more than ten, and so on. 
This arbitrary use of the number ten is in 
practice extremely inconvenient, since it 
has no relation to the rational divisions of 
the sciences ; and, in so far as it assumes 
that the subjects indicated by the figures in 
a certain place of decimals are subordinate 
to those which occur in a higher place, it 
involves theoretical considerations, the 
validity of which may not in all cases be 
admitted in the future. In the second 
