Recognized HTML document

420   Life and Letters of Francis Galton


who are become the great men of science. Now the difficulty in social statistics is of exactly the same kind, but greater. Therefore by no straightforward and expeditious method can the problems in which you-and I may be permitted to add myself-are so much interested, be solved. Each is a separate and difficult undertaking requiring a vast deal of thought and planning, just like planning a campaign. Quetelet's own history is an example of this. His promises and hopes and his achievements in 1835-6 remained in state quo up to the last edition of his- work (Physique Sociale) in 1869. He achieved nothing hardly of real value in all those 33 years'. So again Buckle, who started with a flourish of trumpets in the first chapter of his History of Civilisation, did next to nothing beyond a few flashy applications that have rarely stood aftercriticism.

The way in which your object might best be attained requires, I think

(1) A man (or men) conversant with the methods, and especially the higher methods, of statistics.

(2) Conversant with the existing statistical data.

(3) With his heart directed towards the solution, one by one, of such parts of such of your problems as he can, after much thought, see his way to attack successfully.

(4) Proportioning his labour so as to stop short when he has reached a fairly near approximative result, and not to waste himself in figures in order to procure a slightly closer approximation. In short he must be the master and not the slave of his statistics. The waste of effort by statisticians seems appalling. (I know it is so in meteorological statistics.)

How to get all this? I gather that you have in view the establishment of a Professorship or Readership at Oxford. Before you fix your mind in that direction or in that of Cambridge, I should like to tell you by way of warning the experience our Geographical Society has had in doing the same for Geography in the two Universities. I happen to have been closely connected with the movement and am indeed going down to Cambridge next week to see if the dismal want of success of our Reader there can be obviated. The result of very much inquiry has been, that unless the subject on which a Professor lectures has a place in the examinations he will get no class at all. His position will be that of a salaried sinecurist, which is proverbially not conducive to activity. 'Still, he would have leisure and personally would have interest in his work, and if only a Reader, is removable after 5 years. A professor is permanent. He would live in much isolation at Oxford as far as his own subject is concerned, for all the main interests of the place are scholastic, and many of them are rather petty. It occurs to me that perhaps as good a way as any might be to found a professorship at the Royal Institution in London, and to require a yearly course of lectures. The Royal Institution audience is just the sort to stimulate on the one hand and to curb the vagaries of the inquirer on the other. It is a mixture of some of the ablest philosophers, of many persons of wide social interests and of the general public. The existing professors are all men of the highest ability in their several lines : Lord Rayleigh in Physics, Dewar in Chemistry, Victor Horsley in Biology. If a Professorship in Social Statistics could be established there on the same basis as those mentioned, it would have to be nominally renewed each year up to five years' (I think) tenure. Then the re-election is for another (practically) 5 years. The cost is, I think, about £300 to £500 a year, not more. Pray excuse my impertinence if you think it such, in venturing to suggest, but my only object is to show what seems to me to be the best direction of action. I think London would be by far the best residence for an inquirer into social statistics. Believe me, Very faithfully yours,

FRANCIS GALTON.

Looking back after thirty years one is compelled to think that Florence Nightingale's scheme, if it could have been carried out, was essentially better than Francis Galton's. How could a school of trained applied statisticians have been created by six lectures a year at the Royal Institution ; that institute has a most valuable platform for announcing in a popular way the results of recent research, but it is not an academic centre for training enthusiastic young

' I venture to think that this is far too sweeping, it overlooks not only what Quetelet achieved in organising official statistics in Belgium, but his great work in unifying international statistics.