822 REPORT — 1900. 



only serve so badly, and virge its members, in the later discussions, to supply the 

 shortcomings of the occupant of the Chair. 



Of all statistical work the enumeration of the units of population must ever 

 take the foremost place, and on the eve of the census to be taken before many 

 more months have passed a reference to that great impending task could hardly 

 be omitted on this occasion. In common with all students of the machinery of 

 census-taking I am sure I echo the feelings of the Section — as I do those of the 

 Eoyal Statistical Society, vrho have long laboured in this direction — in deeply 

 regretting that the first census of the twentieth century is not to possess the 

 distinction many had hoped to see conferred upon it of being by preliminary 

 announcement — as I hope it may prove to be in ultimate fact — the first of a series 

 not of decennial but of quinquennial countings of the people. 



The growing complexity of social conditions and speed of life in all its functions 

 at the present date, contrasted with the leisurely movements of a hundred years 

 a^o, would alone and amply justify a more frequent stock-taking of the inhabitants 

 of Great Britain than has been the practice in the past. The practical wants of our 

 much-multiplied system of local government cannot fail, I believe, ere long to 

 bring about the granting of an intermediate numbering, even if for the moment 

 other considerations overrule the more academic pleas of statisticians for this 

 reform, or the arguments, soimd as I believe them to be, for a permanent Census 

 Office, a permanent Census Act, and a trained and continuous Census Staff', to 

 whom preparation of the machinery beforehand and detailed elaboration of the 

 results after the actual census year might with real economy be entrusted. 



Like probably many another student of practical statistical organisation, I 

 have to own to some modification of the demands for enquiry into the condition 

 as well as the numbers of the people, which I once believed might be properly 

 combined with the actual operations of enumeration. Some little experience in 

 measuring the extent and the value of the answers elicited by question and by 

 schedule have shown me that with due regard to the quality, if not even to the 

 quantity, of the replies extracted from the least instructed section of the popu- 

 lation, you must limit your curiosity unless you are to be landed in doubt, in 

 difficulty, and in misconception. 



Specific and parallel enquiries in point of time by one or another central 

 body may no doubt be devised and directed so as to bring out a definite and 

 limited series of facts, afi'ording matter to be compared with population totals. 

 But to load the census proper with side issues is not to help forward the best type 

 of statistical knowledge, and the attempt may well be pushed too far. I fancy 

 there is now some reason to believe that ten years ago we erred in this respect. For 

 these reasons I have never in recent years been able to go along with many active 

 and highly intelligent foreign colleagues, whose more sanguine aspirations as to 

 possibilities of what a census can tell it is always pleasing to witness, even if the 

 feasibility of their suggested developments may be questioned. 



Sound and reasonable advice on such a subject may be found in the timely 

 remarks of my colleague Mr. J. A. Baines, in his paper to the Royal Statistical 

 Society in February last on the ' limitations ' of census-taking. From no better or 

 more practical source could we hope to be instructed on what can and what can 

 not with advantage be got than from the able officer whose superintendence of 

 the vast Indian Census of 1891 brought him such widespread recognition. 



The mention I have made of the suggestions of foreign statisticians on census- 

 taking reminds me that although the proposal which has been before the Inter- 

 national Statistical Institute in one form or another for a synchronous ' world's 

 census,' at the moment of passing from one century to another, is hardly likely, 

 for administrative reasons and in view of the previous fixtures of the great census- 

 taking Governments of the etirth, to be literally realised, the dates of the great 

 countings of the nations will nevertheless come sufficiently close for all practical 

 comparisons. The great Russian enumeration, on the success of which M. Troinitsky 

 is so heartily to be congratulated, is not yet long accomplished. The twelfth 

 census of the United States is now being taken. The Scandinavian enquiry 



