DURING the course of an analysis of some 1,600 periodical references, borrowed by the library of Butterwick Research Laboratories from outside libraries, the results of the analysis were compared with studies on the scatter of articles in periodicals carried out at the Science Library. These were originally published in the journal Engineering, vol. cxxxvii, p. 85, 26 Jan. 1934, and were recently summarized by S. C. Bradford in his book, Documentation (London: Crosby Lockwood and Son, Ltd., 1948). A certain misunderstanding became evident in the algebraic analysis carried out by Bradford and his collaborators, further consideration of which has proved of interest.