But even this is not all. It is a maxim of this strenuous age that all things are good or bad according to the results to which they lead, not in the narrow sense that “the end justifies the means,” but in the broader sense that we must know things by their fruits. The man who said “I go, sir,” and went not, was judged by his acts, not by his words; and no matter how much knowledge we store up and how many tables of data we collect we shall be derelict in our duty if we regard this as an end in itself. The state of mind in which the Mahatma spends his life in impassivity, contemplating inward things and making no outward motion, may have certain advantages, but it is not consonant with the spirit of this age and this land. By which I mean that when we have found out something from our statistics we must do something with it. More; we must so direct our statistical investigations that they bear directly on a possible course of action. This is done by the great manufacturing concerns that maintain statistical departments; but we all use statistics in this way. If a boy wants to go to the circus he first looks through his pockets to see whether he has enough cash. Here is the germ of a statistical investigation conducted for the specific purpose of getting information on which future action is to be based. Here sometimes, where the opportunity of collecting statistics is very great, and expense is no object, is a good excuse for gathering a great deal that would seem to be useless, with the expectation that some of it may turn out to be interesting and may suggest some line of work that had not previously been thought of. To go as far as this, the institution must be large and rich. But how many of us do anything with our statistics? How many collect statistics along special lines to assist in deciding what we shall do along those lines? How many of us, rather, consider that, when our statistics have been collected a disagreeable task has been done, and put them behind us till the year rolls round again? Perhaps we have had enough now of the philosophy of statistics. Let us see what concrete kinds of statistics are necessary and in what order of importance.