1 86 CARNEGIE INSTITUTION 



initiative and relation which the Roj'al Society has assumed to the 

 latter enterprise. 



(c) As to period : I have already indicated my opinion that for 

 the Carnegie Institution, created to promote research, the most service- 

 able contribution in bibliography will be that which exhibits the 

 recent and current literature rather than that which is retrospective. 

 The investigator who is to advance the boundaries of knowledge will 

 not, except as he is a bibliographer or historian of his subject, have 

 much occasion for retrospect. In so far as he ha,s occasion for such, 

 he will require not a mere list of titles, but the actual books them- 

 selves. For these he must have recourse to a particular library or 

 libraries. It is the duty of those libraries, through their catalogues, 

 to furnish him with a statement of their contents. In the field covered 

 by the International Catalogue, the Royal Society's " Catalogue of 

 Scientific Papers," already covering the period 1 800-1883, and pro- 

 posing to cover also 1884-1900, is so nearly comprehen.sive as to 

 render parallel attempts extravagant ; just as in the field of medicine, 

 for which the Index Medicus will cover the current literature, the 

 Catalogue of the Surgeon General's Office lyibrary forms for all prac- 

 tical purposes a comprehensive statement of the existing literature. 



In considering undertakings more special, within a narrower field 

 or a particular department of literature, the following considerations 

 should apply : 



I. In any subject in which there is active research, accompanied 

 by a continuing literary record, a bibliography to be serviceable must 

 also be continuing. A grant of a given amount will therefore, as a 

 rule, be more effective if applied to a continuing bibliography within 

 a narrow field than if exhausted upon a (periodically) limited bib- 

 liography within a larger field. 



II. A bibliography differs from a selected list of titles, on the one 

 hand, and from a catalogue of a particular collection, on the other. 

 It attempts to be a complete exhibit of the literature of the subject. 

 Such completeness exists in no single place or institution. A bibli- 

 ography compiled at second band can, however, be of but little 

 authority. A bibliography which consists merely of brief titles, 

 without explanation or analysis or an attempt to locate the material, 

 can be of bat meager utility. The preparation 01 a serviceable bib- 

 liography requires (i) direct use of the complctest existing collections 

 of material ; (2) the most efficient bibliographic tools ; (3) expert 

 bibliographers, not merely specialists in the subject matter ; (4) 

 promptness and frequency of issue ; (5) a form of publication which 



