632 TRANSACTIONS OF SECTION F. 



Birmingham, and the nature of averages and index numbers is dealt with in 

 lectures and examinations. In Manchester statistics may be taken as one of a 

 long group of special subjects. In Liverpool elementary knowledge of statistics 

 is expected of students in economics, and an effort is now being made to introduce 

 a course of statistics. In Glasgow and in Edinburgh statistics has no formal 

 place, but an attempt is being made in the latter city to recognise it. In Dublin 

 piirt of the ordinary lecture course is devoted to elementary statistical methods, 

 and the subject ' Elements of Statistics ' counts as an essential part of the 

 examination for the annual prizes in political economy ; the questions are non- 

 mathematical. At Cambridge statistics is not distinguished as such in the 

 syllabus for the Tripos in Economics, though statistical methods are implied in 

 the Part II. papers on ' Advanced Economics, mainly Analytic' So far as I can 

 learn, there is no provision in lectures or examinations at Cambridge or Oxford 

 for the application of mathematics to statistics. More complete recognition is 

 given in the Faculty of Economics in London. There statistics up to the point 

 reached in Birmingham and Manchester is demanded of all Pass students, and a 

 considerable amount beyond is expected of those who have preferred mathematics 

 to logic in their first year. Regular courses of lectures are specially devoted to 

 the subject in two of the constituent colleges or schools of the university, and a 

 large number of students have passed through them. Though the mathematics 

 required stops just short of the infinitesimal calculus, there is no such limitation 

 in statistics as' one of eleven possible special subjects for honours. 



It may reasonably be held that the application of mathematical reasoning 

 to tabular information is so special a subject that it may safely be left to post- 

 graduate study and individual initiative for men who are working at so wide a 

 subject as political economy or taking a practical course in commerce ; but no 

 one can, I think, reflect seriously on the statements just made without coming to 

 the conclusion that, in view of the immense importance attached to statistical 

 reasoning in modern times — whether in trade disputes, or in proposals for social 

 reforms, or in political pamphlets and speeches, or in the public Press— men who 

 have not mastered the main criteria of the adequacy of such statements, who are 

 not acquainted with the possibilities of such measurement, and who do not know 

 the main statistical facts already common property, are not completely prepared 

 in their professed subjects, and that there is not much likelihood that they will 

 obtain this knowledge while the universities give so low a formal place to, and 

 so little organised teaching in, these subjects. 



If, however, it is admitted that the Pass student in economics cannot be 

 expected to have more mathematics than is required for matriculation, and that 

 there is not a sufficient demand for a course which shall apply more advanced 

 analysis to economics and statistics, we might expect that other means would be 

 taken to supply the country with those expert statisticians that so many public 

 departments, whether central or local, so obviously need. Those responsible for 

 the syllabus for First Division clerks in the Civil Service were not of this 

 opinion. Out of twenty subjects, political economy is one, and in its syllabus 

 the application of statistical methods to economic inquiries is named ; but only 

 two questions have been set in the last three years, and these do not involve 

 mathematics. Mathematics itself occupies a conspicuous place, but there is no 

 sign that its application to statistics or the theory of probability are included. It 

 may be said that the examination is intended to be a test of non-specialised 

 education, and that technical methods are best studied in the departments them- 

 selves. But though I find that the Record Office demands a knowledge of history 

 and of a language from its recruits, there seems no evidence that an adequate 

 knowledge of mathematics is required among those who are drafted to offices 

 where the public statistics are handled. It may happen that those who are 

 responsible for statistical analysis have gained high honours in mathematics ; but 

 in the existing routine it seems just as likely that they gained their distinction in 

 Latin verse or medijeval history. Should a department discover that the handling 

 of statistioe did not come naturally to an ordinary educated person, it can perhaps 

 fftU b^ck on the seventh clause of the Order iji CouiiciJ of 1870 : ' In case the 



