RESEARCH IN THE EXACT SCIENCES. 1 87 



existing material to the already overv'olumiuous scientific literature 

 of the day. The right man will be the man who has the courage to 

 "scrap" and to doit relentlessly. Science wants immensely the 

 courageous pruner to-day; but his is not an enviable task, and the 

 Carnegie Trustees would have to support their man pretty steadily 

 to enable him to be effectual. He will be sure to make some mis- 

 takes, and these will be at once seized on and trumpeted abroad. If 

 we suppose that the above three conditions can be fulfilled, may we 

 not question whether the man pictured would not be of such caliber 

 that he would do far better work for science if he were allowed to 

 use other people's observations where he chose, and to observe and 

 collect himself where he found them defective or incapable of throw- 

 ing light on the branches of science he was peculiarly interested in ? 

 In other words, the director would be reduced to an ordinary scien- 

 tific worker, placed in one sense under very favorable conditions, in 

 another under unfavorable conditions ; he would have ample mate- 

 rial and support, but he would differ from an academic teacher in 

 having no school wherein he might train his subordinates in his 

 methods. 



(5) On the whole, I doubt whether the founding of an institute 

 to ' ' scrap ' ' and codify existing observations and scientific material is 

 feasible if desirable. I am inclined to think that more might be 

 done by a Statistical and Compidating Institute. This institute should 

 have a competent director and a highly trained staff. It should be 

 prepared to report on anj^ data or material submitted to it at a mod- 

 erate fee. This fee might be remitted on the recommendation of 

 the director, or a committee, in the ca.se of first-class work from a 

 man of scientific repute but small means. It would have to be re- 

 tained, however, to prevent a flood of worthless material being sent 

 in to be reduced. The institute might also offer advice on the col- 

 lection of material on observational method and on statistical treat- 

 ment, again charging a slight fee to prevent the institute being u.sed 

 as a source for providing research work for those who were too idle 

 or too dull to discover such work for themselves. Besides, private 

 individuals, learned .societies — astronomical, meteorological, or bio- 

 logical —might and probably soon would use the institute to carry 

 out special investigations on the value of material already amassed 

 in some one or other branch of their special sciences. Finally, 

 Government departments would very .soon fall into the habit of ask- 

 ing for reports on the special material of their own spheres. The 

 like course would betaken by local bodies in the case of demographic 



