496 



SCIENCE 



[Vol. LVI, No. 1453 



our present methods reached a fundamentally 

 correct solution of this important problem. 



The ideal — I take it all are agreed here — is 

 to have all new information so recorded that 

 any part of it, major or minor, can be located 

 promptly and certainly. Classified bibliogra- 

 pihies and indexes, in book or card form, ab- 

 stracts with subject and author indexes in book 

 form, etc., are some of the means in use to 

 approximate this ideal. I should make it clear 

 that I am here discussing only the recording of 

 information, not criticism. 



A review of most of these agencies, at least 

 the biological ones, shows that they have been 

 very inadequately supported, have been able at 

 best to do the work incompletely, great as has 

 been the service tihey have rendered. And, per- 

 haps more important, continuity, because of 

 inadequate financial support, has not been 

 guaranteed, with the result that agencies have 

 arisen, grown to great usefulness, declined, and 

 disappeared, to be followed by gaps before 

 other agencies have got under way. The result 

 is that biological literature is not only inade- 

 quately recorded, but it is recorded in so many 

 places that the task of rather exhaustively con- 

 sulting the literature on many biological sub- 

 jects is indeed exhausting. Moreover, an enor- 

 mous amount of duplication exists, all agencies 

 in the same general field covering the literature 

 in well-known journals, but frequently being 

 forced to neglect that in the less well known. 



I believe it is worth while to inquire into 

 the fundamental causes of the difficulty and 

 just what sort of service the present worker 

 demands. As the number of research journals 

 and the amount of material published have 

 steadily increased it has become more and more 

 impossible for the individual worker personally 

 to subscribe for an appreciable amount of the 

 literatu!-e in his field. More and more he is 

 depending upon bibliographic and abstract 

 services to bring to his attention the contribu- 

 tions he must consult critically. It could not 

 be otherwise. But in general the workers 

 desire to own these bibliographies or abstracts, 

 especially the latter, as they constitute, if well 

 done, a sort of master key to the literature. I 

 believe I am correct in saying that usually 

 workers prefer a good abstract journal cover- 

 ing their field to a single or even several re- 



search journals. Granting that workers need 

 and want most of all adequate abstract and 

 index services, it is clear that if they are to 

 avail themselves fully of such agencies these 

 must be procurable at prices within their reach. 

 In other words, the cost of the service must be 

 relatively low — very low for most of us. Sec- 

 ondly, the service must approximate complete- 

 ness. It should be of such character that reli- 

 ance can be continuously placed on it to report 

 adequately practically all the literature. 

 Thirdly, the service must be prompt; the de- 

 mands of the worker on his abstract journal 

 are much like those on the newspaper. Fourth- 

 ly, there should be reasonable guarantees of 

 the continuity of the service. 



As regards completeness and adequacy, I be- 

 lieve only one type of service has been devel- 

 oped which is generally satisfactory and that is 

 the abstract journal with detailed indexes. 

 Classified bibliographies, even though based on 

 an examination of the texts, can at best but 

 incompletely record the new information, 

 though their utility has been and is very great, 

 esfiecially for libraries. In some instances, 

 notably in the case of libraries, classified bib- 

 liographies in card f oi-m are preferred ; but for 

 individuals the expense and labor involved in 

 filing and files have not made these very popu- 

 lar. Moi-eover, such bibliographies are cumu- 

 lative and not periodic. The same objections 

 presumably hold for abstracts in card form. 

 Another objection to the latter is that it is 

 difficult to accompany them by a usable de- 

 tailed subject index. 



The reason for the popularity of abstracts 

 with detailed indexes is obvious. Unlike bib- 

 liographies, abstracts, if well prepared, quickly 

 give the reader definite information as to the 

 exact content of contributions. In many cases 

 the titles can not do this even though formu- 

 lated ■with care; and many of them are not 

 formulated with care. Indeed, many are dis- 

 tinctly cryptic. From complete abstracts vir- 

 tually complete subject indexes can be pre- 

 pared, based, of course, on a careful analysis 

 of the abstract as well as of the title. By and 

 large, the complete- subject index is perhaps of 

 the greatest importance. But its preparation 

 necessitates complete abstracts, unless indeed 

 the indexing should be done wholly from the 



