434 



SCIENCE. 



[N. S. Vol. I. No. 16. 



serviceable index must, of necessity, be of 

 great extent, and undertaken upon a scale 

 of considerable magnificence. 



It may be that the time has not yet come 

 when the scientific men of all the world can 

 cooperate together in such a task as this, 

 but if cooperation is possible in anj^ field of 

 intellectual activity, surely it is in that of 

 science. Such cooperation is not only es- 

 sential to thorough work in indexing, but 

 would also have a most important influence 

 in promoting united efforts in other branches 

 of scientific activity. 



The considei'ations suggested are these : 



1. The catalogue should be international 

 in name and scope. This is essential in 

 order to secure the unreserved support of 

 all nations engaged in the production of 

 scientific literature. It should, therefore, 

 not bear the imprint of any society or organ- 

 ization, or derive its distinctive character 

 from any one nation. Since the titles will, 

 of necessity, be quoted exactly, it might be 

 well that all annotations and comments 

 should be in the same language as the title. 

 To insist that only English or French 

 should be used would be fatal to its general 

 adoption bj^ other countries. Titles in the 

 Scandinavian, Slavonic and Oriental Lan- 

 guages and dialects and others would, how- 

 ever, need to be translated into French, 

 German or English. 



2. It should be exhaustive within its own 

 limits, no latitude being given to the jiidg- 

 ment and taste of its editors, in the matter 

 of rejecting titles. 



3. It should be printed in annual in- 

 stallments, each installment including every 

 paper or work printed within a single year, 

 and each installment should be published in 

 not more than six (preferably not more 

 than three) months after the close of the 

 year. 



4. The publication should be in the 

 form of a bibliographical catalogue, with 

 the titles arranged alphabetically by au- 



thors, the papers by each author to be num- 

 bered, beginning with number one. This 

 would render it possible to identify anj^ 

 paper, either in an annual or a general 

 index, by simple reference to author, year 

 and number. 



In recommending that the catalogue 

 shall be published in book form, I am by 

 no means unmindful of the merits of the 

 card-catalogue system in work of this kind. 

 I use card-catalogues freely in my own 

 work, and in the National Museum there 

 are hundreds of thousands of cards by 

 means of which the vast collections of 

 specimens and papers are kept under con- 

 trol. The card-index has its limitations, 

 however, and these are nowhere more evi- 

 dent than in connection with such a 

 scheme as a universal scientific catalogue. 



The verj' bulk and unwieldiness of the 

 card system is an objection, which maj' be 

 partly appreciated if one can imagine the 

 contents of the ten volumes of the Royal 

 Society's Catalogue transformed into card 

 form and arranged in drawers.* 



In the volumes as they now stand, the 

 eye can sweep rapidly over page after page 

 in search of a given title, and thirty or 

 forty impressions pass to the mind at a 

 glance, instead of one, while the strain 

 upon the attention caused by turning over 

 the pages is much less than where each 

 title card is scrutinized singly. 



For finding a book or reference when 

 the name of the author or its title is 

 known, the card system is without rival. 

 It is less useful, however, when, as often 

 happens, one is ' looking up ' a subject in a 

 general way. A card-catalogue, after it has 

 attained to great bulk, requires much labor 



* Dr. Carrington Bolton prepared the copy for his 

 ' Select Bibliography of Chemistry ' on slips of 

 standard sizes, and it filled 7 standard trays or » 

 length of nearly 9 feet. Tlie slips were on thin paper 

 — if they had been of card the lengths would have 

 been nearly 20 feet. When printed the 12,000 titles 

 ■were presented in a light convenient octavo volume 

 of about 1,200 pages. 



