Tobacco Documents Online

EDWIN BIDWELL WILSON i institution. In 1922 when the Harvard School of Public Health was organized,he became the first Professor and Head of the Department of Vital Statistics. He was also profes- sor in the Department of Economics in the College. In the year following his retirement in 1945, he was Steven- son Lecturer on Citizenship at the University of Glasgow. On his return to Boston, he joined the Office of NavalResearch where he remained until his death. For his' services he received the Navy's Distinguished Civilian Service Award. ' At the School of Public Health he had opportunity to come into contact with questions of a medical nature, an opportunity he realized fully. As a result of conversa- tions with Hans Zinsser on the effect of bacterial dissocia- tion and changes in virulence on the shape of the epi. • Fellowship Committee and on the scientific advisory com- demic curve, he became interested in mathematical mittee of the Tobacco Research Council. He was actively epidemiology and contributed many papers to this field. . interested in the cancer problem and was at one time Discussions with Howard Mueller. led to contributions to Chairman of the University's Cancer Commi§sion. He. the theory of dosage response. He early became interested early joined the American Society for the Control, of in factor analysis, a method which was then becoming Cancer, became its president and later a director of the popular with psychologists. The last of a series of papers 'succeeding American Cancer Society. in this area (Psychometrika 4:133-148, 1939) ended None who knew him could fail to appreciate his mind, with a typical sentence: "It would be- a pity if Wundt keen as 'ever at the time of his death, his fund of recollec- had taken Psychology away from her mother Philosophy tions of earlier days at Harvard and M.I.T.; his penetrat- and married her to Science only to have her desert to ing comments on places and personalities; his quiet cour- a paramour Mathematics." . , ." tesy and lack of ostentation. His contributions to statistics are notable not only for. .' " He leaves two 'daughters, the. Misses Doris and Enid Statistical Inference" (J.A.S.A 22:209-212,1927) he sug- gested what was later to become known as the confidence interval. He wrote a series of papers on the fourfold and larger contingency'tables and indeed was working on one at the time of his death. His insights greatly and a keeper of bees. His death deprives society of one 15:120-125, 1929) where he stated, "This illustrates the of the few men whose interests are catholic and whose principle that we must have a plurality of samples if intellect and curiosity lead them to contribute to many we wish to estimate the variability of some statistical fields. quantity, and that reliance on such formulas v/Vn is Edwin B. Wilson was born on April 25, 1879 in Hart- not scientifically satisfactory in practice, even for esti- ford, Connecticut, the son of a former Yale teacher and mating unreliability of means." And again in "Probable school superiritendent. He was graduated from Harvard Error of Correlation Results" (Proc. A.S.A. 1-4, 1929) in 1899 summa cum laude in mathematics. He continued he said, "The sampling error is indeed a notable experi- h Prof. Edwin Bidwell Wilson, former president of the American Statistical Association, died on December 28, 1964 after a brief illness. He had not had one career but seven with publications in mathematics, physics, aer- onautics, statistics both theoretical and applied, econom- ics. epidemiology and socioloay. He was also a farmer their number and depth, but also because of his realiza- tion that much of classical statistical theory does not " work in practice. In his address at the 125th anniversary of the American Statistical Association, which is included in this issue, he,refers to his "Note on C. S. Peirce's Experimental Discussion of the Law of Errors" (P.N.A.S. ment; it has accomplished much good and will accom- plish more, but there are conditions under which it simply does not work. I would not defy it, I would not abolish it, for to many it has become a great moral issue and the disturbance of their emotions may be very serious to our "own freedom of conduct to indulge in In 1910-11 he held a post at Harvard as Lecturer im'' " a moderate amouryt of discreet nullification." Mathematics. -However, his deepening interest in applied In "Probable Inference, the Law of Succession, and is studies at Yale where he came to Know Willard Gibbs. He received the Ph.D. degree from Yale in 1901 and then went to the l;cole Normale Superieure, the Sor- ' bonne and College de France in Paris. Wesleyan Univer- sity conferred on him its honorary LL.D. in 1955. He began his teaching career in mathematics at Yale. mathematics led him to move to M.I.T. where he held secretary of the administrative committee which ran the ' helped to clarify the logical framework of these important various positions, including the Chairmanship of the I Department of Physics. He helped develop a department of aeronautics there during World War I. For two years; i 1920-22, while M.I.T. was without a president, he was statistical tools. He was the author of two textbooks: Gibbs-Wilson Vector Analysis and The Advanced Calcu- lus, books well known to several generations of students in mathematics. These books were re-issued by Dover in 1960 and 1958,• respectively. E. B. Wilson's activities were recognized by his asso- ciates. He was past president of the Social Science Re- search Council and of the American Academy of Arts 'and Sciences. He was vice president of the National Aca- demy of Sciences and edited its Proceedings from 1915 until his death. He held a fellowship in the Royal Statisti- cal Society~of London and memberships in the Mathema- tical Society of Benares (India) and the American Philo- sophical Society. He received the John F. Lewis Prize from this Society in 1963. He was on the Guggenheim .. 52 The American Statistician, flpril; 1965.° :. ,

f s `~Vilson of Brookline; two sisters, the Rev. Eleanor WiI- son of Anahola, Hawaii, and Dr. Jane Wilson Hall of Calais, Nlaine; and two brothers, Dr. Perrin T. Wilson of Cambridbe and Dr. Theodore W. Wilson of Baltimore, Maryland. In a review of the Rukeyser biography of Willard Gibbs (Science 99:386-389, 1944)• Professor Wilson wrote: "It was my privilege as a young man to become ac- quainted with a considerable number of distinguished scholars of the generation of Willard Gibbs who seemed to me to be much alike in their simplicity, dignity and friend- lines,t-gentleman of the old school we youngsters called them. T hey did'not wish to be hero-worshipped, they were not patronizing; they did not proselytize, they were living examples of what the best in university life has been, is now, and will be so long as there are youth who are in- spired by such examples to try to become in all simplicity worthy successors to them:' Certainly, Professor Wilson was himself an example of the best in university life. Jane Worcester, Harvard University Alexander D. Langmuir, M. D., Chief, Epidemiology Branch, ; Kenichi Koyanagi The world of statistical methods in industry lost a great leader in the death of Kenichi Koyanagi on 16 January 1965. If any one man was responsible for the ad- vance of quality of Japanese product dating from the year 1950, that man was Koyanagi. Although he majored in German literature at the Uni- versity of Tokyo, he turned his attention, after the war, to the needs of his country and perceived how necessary it would be for Japan to depend henceforth on exports. He heard of the statistical control of quality and per- ceived that in its broad aspects, it was precisely what Japan needed. As manager of the Union of Japanese Scientists 'and C ~\, 'uE E R 11. OJIADPORTUNITY UNITED STATES PUBLIC HEALTH SERVICE . COMMUNICABLE DISEASE CENTER for MATHEMATICAL STATISTICIAN $19,000-$22,000 To serve as chief statistician for an extensive program . of investigation, surveillance, and control of communi- cable diseases in the nation; direct staff of 10 profes- sional statisticians; collect and analyze national mor- bidity and mortality data; conduct epidemiological surveys; teach biostatistics to Epidemic Intelligence Service Officers; apply epidemic theory to practical con- trol problems; guide development of computer services for the Communicable Disease Center. Requires Ph.D. or Sc.D. and 4 years of broad responsi- ble and relevant experience. Headquarters on campus Emory University, Atlanta, Georgia. Moderate travel in nation and overseas. Equal employment oppor- tnnity and full Federal civil service benefits. Interested applicants should write to: Communicable Disease Center, PHS, DHEW Atlanta, Georgia 30333 : Proceedings -o the 34th Session INTERNATIONAL STATISTICAL INSTITUTE Engineers, supported by industry, he invited lecturers ' , from the U. S. and elsewhere, conducted courses at all levels, including top management, put out books, and... commenced publication of books and journals at differ- ent levels of sophistication. His organization instituted the Deming Prize, which has been in large measure re= sponsible for the competition amongst Japanese com- panies, a reward given annually to a manufacturing con- . cern for the most outstanding improvement of quality . through statistical methods, and a sum of money to a•. Japanese statistician for work in theory and application. lie was recipient of the Edwards Medal from the American Society for Quality Control in 1963. He was a leader in forming the Tokyo Chapter of the American Society for Quality Control in '1952. This was an ex- ample of his vision to see the advantages of tying in with a large existing organization.. •(W. Edwards Deming) 100 scientific papers deal with problems of interest to statisticians and economists. They describe too the latest techniques and methods developed in various parts of the world. 2-.Volume set, $30.00. ' UNIVERSITY OF TORONTO PRESS-, Toronto 5, Canada Books for the United States shJpped Jrom our Brooklyn warehouse .