516 THE BELL SYSTEM TErilNir'AL JOT-IJXAL, AIARril 1057 



spacings but with \'arious horizoiilal positions. 'I'o the right of each word 

 was a short vertical line. The distances x„ of these lines from the vertical 

 centerline of the paper were obtained from a list of random numbers 

 with a gaussian distribution such that for the list x- = 1 inch. Of course, 

 X- for each list would depart from this value. As the words were read, the 

 reader used a pencil to make a dot as near as possible to the correspond- 

 ing vertical line. For each sheet, the departures (!„ from the vertical 

 lines in inches were measured and (P was computed. The number of bits 

 M for pointing for that sheet were then taken as 



M = f log, (^1 + V^ (2.3) 



REFERENCES 



1. E. E. David, Naturalness and Distortion in Speech Processing Devices, Jour. 



Acoustical Soc. Am., July, 1956. 



2. J. C. R. Licklider, K. N. Stevens, J. R. M. Haj'es, Studies in Speech, Hearing 



and Communication, Technical Report, Acoustics Laboratory, MIT, Sept. 

 30, 1954. 



3. H. Quastler et al, Human Performance in Information Transmission, Report 



No. R-62, Control Systems Laboratory, University of Illinois, March, 1955. 



4. C. E. Shannon and W. Weaver, The Mathematical Theorj^ of Communication, 



University of Illinois Press, 1949. 



5. G. Dewey, Relative Frequency of English Speech Sounds, Harvard University 



Press, '1923. 



6. E. L. Thorndike, A Teacher's Word Book of the Twenty Thousand Words 



Found Most Frequentlj^, Teachers College, Columbia University, 1932. 



7. C. E. Shannon, Prediction and Entropy of Printed English, B.S.T.J., 30, pp. 



50-64, Jan., 1951. 



8. E. B. Newman and C. J. Gerstman, A New Method for Analyzing Printed 



English, J. Exp. Psychol., 44, pp. 114-125, 1952. 



9. Keith Hennej-, Radio Engineers Handbook, 3rd Ed., p. 568, McGraw-Hill, 1941. 



