766 



SCIENCE 



[N. S. Vol. XLIII. No. 1118 



true, and we must consequently agree that 

 for industries to retain their position and 

 make progress they must earnestly devote 

 time and money to the investigation of the 

 fundamental theory underlying the subject 

 in which they are interested. 



Research work of this fundamental kind 

 involves a laboratory very different from 

 the usual works laboratory, and also inves- 

 tigators of a different type from those em- 

 ployed in a purely industrial laboratory. 

 It means a large, elaborately equipped, and 

 heavily staffed laboratory engaged largely 

 on work which for many years will be unre- 

 munerative and which, for a considerable 

 time after its foundation, will obtain no re- 

 sults at all which can be applied by the 

 manufacturer. 



The value of a research laboratory is 

 essentially cumulative; in the beginning it 

 may be of service as bringing a new point 

 of view to bear on many problems; later, 

 accumulated information will be more and 

 more available; but most men acquainted 

 with industrial research work consider that 

 five years is the earliest date at which any 

 considerable results can be expected from 

 a newly established research laboratory and 

 that the development of really new material 

 in considerable quantities so that it will 

 have an effect upon the industry as a whole 

 can not be looked for in less than ten years' 

 consecutive work. This does not mean that 

 a laboratory is useless during the initial 

 period, since it will be of considerable serv- 

 ice in many other directions than in that of 

 its main work on the fundamental prob- 

 lems, but when this main line of research be- 

 gins to bear fruit it will absorb the energies 

 both of the laboratory and of the factory. 



It is often suggested that the problem of 

 the organization of scientific industrial re- 

 search is really the problem of obtaining 

 satisfactory cooperation between the manu- 

 facturers and the universities, possibly with 



small research laboratories in the factories 

 themselves acting as intermediaries. Vari- 

 ous schemes have been suggested for en- 

 abling the universities to carry out research 

 work of value to the manufacturers, but 

 if it is believed that the work chiefiy re- 

 quired for the development and mainte- 

 nance of industry deals with the funda- 

 mental theory of the subject, it will be seen 

 that this can not possibly be carried on to 

 any large extent in collaboration with a uni- 

 versity; it requires a continuity of applica- 

 tion by the same investigators over long 

 periods with special apparatus and with 

 the development of special methods which 

 can not be expected from any university. 

 This necessity for continuous work along 

 the same line is, indeed, the greatest diffi- 

 culty in making use of the universities for 

 industrial research. The conditions of a 

 university laboratory necessarily make it 

 almost impossible to obtain the continuous 

 application to one problem required for 

 success in industrial research and, indeed, 

 in the interests of teaching, which is the 

 primary business of a university, such devo- 

 tion to one problem is undesirable, as 'tend- 

 ing to one-sidedness. 



There are also difficulties in obtaining 

 the cooperation of manufacturers with uni- 

 versities and in the application of univer- 

 sity work to industry, which I see no hope 

 whatever of overcoming; the universities 

 do not understand the requirements of the 

 manufacturer and the manufacturer dis- 

 trusts, because he does not understand, the 

 language of the professor. Moreover, it is 

 quite essential that any investigator who 

 has worked out a new process or material 

 should be able to apply his work on a semi- 

 manufacturing scale so that it can be trans- 

 ferred to the factory by skilled men who 

 have already met the general difficulties 

 which would be encountered in factory ap- 

 plication. This development on a semi- 



