March 19, 1920] 



SCIENCE 



281 



assistance, travel, etc., and such other es- 

 sential features as might enable an executive 

 committee, or other similar iboard, to make a 

 rational decision. After a set of preliminary 

 plans had received the approval of a division 

 of the council, the most promising course to 

 follow might be to establish a special prelimi- 

 nary committee on the given project. This 

 committee might be instructed to proceed to 

 get the work started if funds are already avail- 

 able, or to attempt to procure these if they are 

 not. It can not be too strongly emphasized 

 that some funds are necessary, even for the ac- 

 tivities of such a preliminary committee, for 

 it is neither safe nor desirable to ask research 

 workers to donate money, as well as time and 

 energy, to this sort of endeavor. After the 

 needed funds have become available the pre- 

 liminary committee may proceed to consult or 

 correspond with all probable cooperators, ask- 

 ing first their aid in completing and elabora- 

 ting the details of the preliminary plan in such 

 a way that it may be feasible. As this work 

 goes on it should gradually become apparent 

 how many cooperators may be hoped for, and 

 when the preliminary committee judges that 

 the project has reached a feasible stage the 

 committee may enlarge itself so as to include 

 all those of its correspondents who are willing 

 to cooperate. This enlarged committee (which 

 would be the oi^anization mentioned above, as 

 needed before a cooperative project may be 

 actually started) may then reconsider the de- 

 tailed plan and divide the work up among its 

 members. A project may fail at any stage, 

 even after the enlarged committee has been 

 formed, but it seems probable that a good 

 measure of success may be regarded as fairly 

 assured when this stage shall have been 

 reached. Haste is not desirable, to do good 

 work much time must be allowed, but the pre- 

 liminary committee would report the project 

 as imx)ossible at present, if it were found im- 

 practicable to obtain a reasonable number of 

 cooperators, or the necessary funds for its 

 work. 



A cooperative organization started in some 

 such way as this would almost surely be 

 successful, but the contemplated measure of 



its success must not be too large. It must 

 be remembered that this sort of cooperation, 

 if begun, would tread on new ground and 

 would surely encounter unexpected difficul- 

 ties. No considerable concrete results need 

 be looked for at the end of a single year 

 and the financial support available at the 

 start ought to give promise of remaining 

 available for several years at least. Never- 

 theless, the very idea of such cooperative en- 

 deavor in research fields is so extremely novel 

 that much discussion and publicity in the 

 proper circles would be needed before it might 

 be realized, and each preliminary plan sub- 

 mitted, each preliminary committee appointed 

 and each letter or publication or conference 

 produced by such a committee, would help to 

 build up the spirit of cooperation. It should 

 be recognized that the fact of cooperation itself 

 is vastly more important than cooperation on 

 any special project; if one project should fail 

 others should be attempted, the work must be 

 regarded as experimental. It would make 

 little difference just what particular problems 

 were undertaken in this way, but it seems 

 highly desirable that some problems might be 

 so attacked. Once applied in a concrete case 

 or two, the general idea would surely spread 

 more rapidly than ever could be the case if it 

 were held indefinitely in the phase of a priori 

 discussion. As in the prosecution of experi- 

 mental research itself, it is only by actual 

 trials that it can be found out what degree of 

 success might attend such cooperative organi- 

 zations as are here suggested. 



SOME SPECUL FEATURES OF COOPERATION IN EX- 

 PERIMENTAL RESEARCH 



Several features of cooperative research have 

 been impressed on the writer during a number 

 of years' experience with this sort of attempt. 

 First, there appear to be a large number of 

 good experimenters who do not have well-se- 

 lected problems in mind, who work on that 

 which lies close to them rather than on that 

 which seems to be most fundamental, most far- 

 reaching or most imperatively needed for the 

 growing structure of knowledge. These work- 

 ers are generally the younger men, and they 



