395 NATURAL SCIENCE. June. 



to classify the material. Classify — yes — but how classify ? Here 

 doubtless there were as many opinions as there were men who had 

 considered the question. They discussed .... and they postponed 

 the discussion to a later date ; for one must further discuss and 

 further reflect, and again reflect, and discuss and re-discuss, for this 

 is a question of principle. 



The question of classification of scientific papers is a difficult 

 one ; it is a question which each can decide in a purely conventional 

 manner, but which perhaps will never be decided rationally and in 

 a really scientific way. 



The decimal system ' of classification is only a method, a 

 notation, not a classification strictly speaking. One can apply such 

 a notation to any classification. But the classification of subjects, 

 and of sciences and their subdivisions, ought to be quite independent 

 of any preconceived system. The decimal classification, already in 

 use for science, is based on a purely conventional system which may 

 satisfy a librarian, but should not satisfy a man of science. Therefore 

 the London Congress rejected both the good (of the notation) and 

 the bad (of the classification), and called for a further pause before 

 classifying. 



In order to understand the difficulty better let us take an example : 



Knudsen, Martin. De 1' influence du plankton sur les quantites 

 d'oxygene et d'acide carbonique dissous dans I'eau de mer. 

 Comptes Rendus, cxxiii., p. 1088. 1896. 



How shall we classify this ? Is it zoology, botany, or physiology ? 

 No, it is rather chemistry, or perhaps physico-chemistry ? It is 

 certain that the work belongs to science, thus the first cipher of the 

 decimal classification is obtained ; but the second cipher is still 

 doubtful. Having plenty of time, we write to Copenhager^ to ask 

 the author. He replies " Oceanography." Oceanography ! that is 

 not included in the decimal classification. Never mind, let us admit 

 it as a new subject. But oceanography is " Physical Geography," 

 and Physical Geography is Geography, and thus our first cipher for 

 classification becomes inexact. The work falls outside the domain 

 of science, into the same great division as history. I refer here to 

 the classification as emended by the Brussels Institute, and not to 

 the American system in which Physical Geography is comprised 

 under Geology, a subject with which this paper of Knudsen's has 

 nothing in common. But any geographer of the modern school will 

 tell you that geography forms part of the domain of science, that 

 oceanography is a distinct science, a portion of Physical Geography, 

 which is nowadays divided into mathematical geography, morphology 

 of the terrestrial surface, oceanography, climatology, etc., etc. Now, 

 to classify lliis work in its proper place so as to be able to find it 

 again at nee il, so that its catalogue-slip may be of use to the original 



^Mr. Dewey's system was described in Natural Science for July 1896, 

 vol. ix., pp. 43-52 



