770 



SCIENCE. 



[N. S. Vol. IX. No. 231. 



in describing a single memoir. While 

 recognizing the usefulness of these, he 

 thought some limit would have to be con- 

 sidered. 



Professor Foster replied that for three 

 years past the Royal Society had requested 

 each author to give an analysis of his paper 

 in such form that it might serve as a sub- 

 ject-index, and that in a large majority of 

 cases it had been found possible to limit the 

 analysis to three subject- entries. 



Professor Riicker pointed out that signifi- 

 cant vFords would serve as a sort of tempo- 

 i-ary expedient where a sudden interest 

 sprang up in some new discovery, instancing 

 the Eontgen rays. After some further discus- 

 sion the resolution as to subject-entries was 

 carried unanimously (the Belgian dele- 

 gates abstaining from voting). 



Professor Armstrong then moved that 

 " registration symbols, in accordance with 

 those in the schedules of classification, shall 

 bs entered upon the slips in some conspic- 

 uous manner, and upon a uniform plan." 

 He explained that at the first Conference 

 schedules in accordance with the decimal 

 system had been prepared and submitted, 

 and that the Conference had decided against 

 them. The plan now proposed is distinctly 

 not the Dewey system. The figures given 

 have no absolute value, and are solely for 

 the purpose of enabling librarians to sort 

 the cards and arrange the material. 



This point was emphasized by Professor 

 Riicker, who stated that in a system in 

 which the numbers had an absolute value 

 the method was equivalent to starting a 

 new language, and he did not believe that 

 the average scientific man would learn a 

 language for such a purpose. 



Chevalier Descamps addressed himself to 

 thequestion of classification. He recognized 

 the serious attention which had been given 

 to the subject by the Royal Society, but said 

 that the Society was not the first to take up 

 the study which had been pursued by a large 



number of authors, men of science and 

 practical men. To provoke a general de- 

 bate on classification seemed inopportune. 

 He had pointed out in 1896 the pos- 

 sibility of a bibliographical classification 

 based on the decimal system. This did not 

 meet with favor, and the Royal Society had 

 endeavored to produce a purely scientific 

 classification. For its labors it merited the 

 most profound recognition, but he regretted 

 that the Royal Society had not explained 

 the ideas which underlay its schedules. To 

 be good a bibliographical classification 

 should be both stable and elastic. The 

 adoption of a mixed system of symbols, and 

 more especially the lack of identity of 

 meaning of the same symbols in the differ- 

 ent sciences, seemed regrettable. He saw no 

 objection to giving symbols a definite signifi- 

 cance. 



The statement of Chevalier Descamps (of 

 which the above is but a brief abstract) 

 brought from Professsor Riicker au argu- 

 ment which probably expressed the opinion 

 of most of the scientific men present, and is 

 according!}' given in full : 



" I think it would be desirable if I say a 

 few words with regard to the very interest- 

 ing remarks with which Chevalier Descamps 

 has favored us. I think we must all agree 

 that the questions he has raised are ques- 

 tions of the greatest interest to any one who 

 has attempted to take any share in a work 

 of this sort. But I very much regret, 

 speaking for mj^self, that I find mj^self at 

 variance with him on several fundamental 

 points. In the first place, he urges us to 

 adopt the scientific system of classification, 

 which shall not change from five years to 

 five years, or ten years to ten years, but 

 which shall hold good for all time, or for a 

 verj' long period of time. One of the very 

 great advantages of our system is that we 

 recognize that science is a growing subject. 

 The notation that fits it to-day will not fit 

 it next year, or ten years hence. Let us 



