RESEARCH IN THE EXACT SCIENCES. l8l 



and in doing so should act as the agent of, and be regarded as doing 

 the work of, the Carnegie Institution. He should be supplied with 

 such office, appliances, and assistants as are necessary to commence 

 work in that branch of the field with which he feels himself most 

 conversant, beginning work on a small scale, to be enlarged and 

 extended into neighboring fields as success became assured. The 

 opposite faults of beginning on too large a scale and of making no 

 provision for possible expansion should both be avoided. 



V. 



The head of the institute should be aided by a council comprising 

 the leading experts best qualified to advise as to the various depart- 

 ments of work. This council might be an international one, and, 

 if the work of the institute is sufficiently expanded to justify it, 

 should hold an annual meeting. 



In order to secure the advantages of mutual consultation, attri- 

 tion, and cooperation, it may eventually be desirable that the work the 

 Institution has already undertaken or is now promoting in the vari- 

 ous branches of exact science should be merged with the proposed 

 institute. 



VI. 



The institute should be started on a very modest scale. The case 

 is one in which everything depends on correct methods from the 

 beginning. By the adoption of these, results may be reached at 

 small expense which, without them, would never be reached with 

 any amount of labor. It seems to me that $10,000 or $15,000 would 

 be ample for the expenses of the first year, as the number of em- 

 ployees who could be successfully put to work would be small. The 

 principal appliances required would be books, but I think that three 

 or four office rooms would suffice for all the purposes of the first 

 year or two. 



The expenses of subsequent years would depend upon the ex- 

 pansion which it found desirable to give to the work. 



Appended hereto are letters on the subject from Prof. H. H. 

 Turner, of Oxford, and L,ord Rayleigh, to each of whom I pre- 

 sented the question of the desirableness of working up the great 

 mass of observations alluded to. 



Simon Newcomb. 



