AN ambitious plan for scientific bibliography and publications is described in a memorandum issued by Science Service, Washington. The plan is designed to eliminate some of the defects in our present system, such as the difficulty of publishing research results promptly or completely owing to the financial burden, and the inadequacy of much bibliographic work owing to lack of access to original papers, etc. It is proposed accordingly to centralise all scientific publication, abstracting and similar bibliographic services, and to substitute a photographic type of duplication for printed reproduction of scientific papers or abstracts. Under this scheme a research report, for example, submitted and accepted for publication, would be reproduced from the standard typescript form by some suitable method other than printing, and full copies of the report or paper would only be supplied to order. The author would, however, also provide a summary abstract, say, two hundred words in length, which after editing, if required, would be reproduced by the most suitable means and the abstract would be included in a weekly or monthly journal issued to all Scientific Workers desiring information in that particular field.