FEB. 19, 1921 sosman: distribution of scientific information 81 



maps and annual reports. ^° The size of the editions ^^ varies from a 

 few hundred to over a million, but most of them were of the order 

 of magnitude of 1000 to 3000. The Department of Agriculture dis- 

 tributed a total of 62.2 million documents in 1918. I have no data 

 at hand concerning bulletins of State bureaus, university bulletins, 

 or the undoubtedly numerous publications not listed in Ayer's Annual, 

 but their circulations are of the same order as for the scientific periodi- 

 cals. 



As I indicated at the beginning of this section by the example of 

 the Chemical Society, the scientific-periodical type of printed matter 

 is overwhelming in its bulk with reference to the producers of infor- 

 mation, among whom it finds a large part of its circulation. At 

 the same time, the extreme heterogeneity of the product and the 

 extreme subdivision of its units of publication are its most outstanding 

 characteristics. The result is a very ineffective distribution of the 

 information even among its producers. 



Praiseworthy efforts have been made to offset these disadvantages 

 by means of abstracting periodicals, indexes, and guides. But 

 so great is the bulk of material that even the indexes and guides 

 themselves are lost in the stream. 



So much for the producing class. With respect to the general 

 public the philanthropic-scientific-periodical type of literature is 

 an almost negligible channel of distribution. The usual circulation 

 of a scientific periodical is about half a dozen copies per million in- 

 habitants, a relatively minor ripple in the great current of wood- 

 pulp. Yet it is not quite as bad as it looks, thanks to the existence 

 of abstracting periodicals and of public libraries. The more im- 

 portant results of a research published in the Journal of Physical 

 Chemistry, for example, with a circulation of only a few hundred, 

 are reproduced in Chemical Abstracts, which, by means of its circula- 

 tion of over 15,000 and its appearance in every important library 

 of the country, may place the data before many thousands of persons. ^^ 



The foregoing remarks concerning the bulk and heterogeneity of 



*" Estimated from data in W. I. Swanton's Guide to United States Government publica- 

 tions. Bur. Education Bull. 1918, No. 2. See also E. GuERRiER, The Federal executive 

 departments as sources of information for libraries. Bur. Education Bull. 1919, No. 74. 



" No published summarized data are available on this point. An act of 1895 limits 

 the initial edition of any departmental publication to 1000. Special acts provide for special 

 publications or particular bureaus. To supply large demands reprintings are permitted. 



'- During the War a Library Information Service was maintained in Washington to 

 keep the librarians of the country in touch with Federal publications. Its future mainte- 

 nance is dependent on the passage of a bill now before Congress (H. R. 6870 and S. 2457). 



