TRANSACTIONS OF THE SECTIONS. 133 



charging at once, or offering inducements for the retirement of men who are likely 

 to become invalids, a fictitious appearance of healthiness in a camp or army may be 

 produced, the diseases which arise there not being suffered to work out their 

 natural result, and their existence, or rather the existence of causes predisposing to 

 them, being thus effectively concealed. A third source of error, perhaps less 

 material, but still worth notice, is that of confounding in one class facts not identical. 

 For instance, the average death-rate of England is easily ascertained, but if you 

 wish to apply that practically to a particular place or class, you must not take the 

 figures as they stand. The death-rate of towns is not that of rural districts, the 

 mortality of infants greatly exceeds that of adults, the term of life is shorter in the 

 labouring than in the well-to-do classes. Age, class, and locality must be separately 

 determined before you can arrive at an even approximately accurate conclusion. 

 The figures, which serve for all collectively, precisely because they do so serve, are 

 illusory if applied to any one of these in particular. 



I mention such possible mistakes in order to show how utterly wide of the truth 

 is the idea that to work by and with statistical calculations is a merely mechanical 

 function, needing no ability beyond that of a careful clerk. They require common 

 sense and vigilance against errors, like every other method of inquiry. Figures do 

 not, indeed, deceive you, but, if you put them to a use they are not meant for, they 

 will let you deceive yourself. 



I suppose it is hardly necessary to remind you of the uses of the statistical method 

 as applied to national affairs. If a man in private life finds his money going too 

 fast, and wants to retrench, the first thing he does is to say, " I must keep regular 

 accounts." So individuals keep diaries of particular matters in which they feel 

 interested, not trusting to vague recollections, but setting down their notes day by 

 day. Every person of observant habits, or engaged in any pursuit which requires 

 accuracy, is in some sense a statistician ; and it is hardly possible to over- rate the 

 value of figures, partly as checking the universal tendency to exaggeration, not 

 wilful, but a kind of mental illusion which operates whenever we are deeply inter- 

 ested — partly as giving precision to ideas which would otherwise remain floating in 

 our minds in a vague, and therefore comparatively useless form. For instance, to 

 say generally that a given trade or employment — that of a grinder or soldier, in the 

 tropics — is unhealthy, conveys a very faint expression, and expresses our feelings but 

 slightly. But put it in this way — that the average length of human life in some 

 occupations is shorter by ten or fifteen years than that of an ordinary labourer — 

 and not only habitual calculators, but the man himself, however ignorant or thought- 

 less he may be, is able clearly to realize the sacrifice he is making in going into 

 that business. With regard to the question of drainage, of ventilation, of food, or 

 use or abuse of strong drinks, it is the statistical test employed on a large scale 

 which alone can be conclusive. I say on a large scale, for it is of the very essence 

 of statistical inquiry that by dealing with masses it eliminates individual peculi- 

 arities. We reason back from the mass to the individual. The unit of the statis- 

 tician, his typical or representative man, is the average man of many thousands. 

 We are familiar with the effect on public health of the establishment of sanitary 

 statistics. Let me point out one or two more instances in which figures form a 

 part, and a very important part, of the diary of our national life. Take the Post 

 Office returns, showing an increase of the correspondence as compared with the past, 

 and the difference which exists at the present day in the amount of letter-carrying 

 between one part of the country and another— say between the population of Ireland 

 and that of London. Take the Registrar-General's returns of marriages, and note 

 how the number of these, relatively to the population, rises or falls as the material 

 condition of the masses is for the time better or worse. Take the annual returns 

 of the Customs and Excise departments. Take our Census Abstract of the occu- 

 pations of the people. Take our Criminal Statistics, a comparatively new depart- 

 ment, and one probably still admitting of much improvement. You may read in 

 these collectively the social and economical history of the age in which we live ; 

 and note everywhere the absence of mere chance. We speak of chance — it is a 

 word we must use for convenience sake — but we really mean by it, not that the 

 result of the thing discussed is in itself uncertain, but that some or all of the deter- 

 mining causes of such result are to us unknown. We imply, not the absence of a 



