STATE OF SCIENCE—COMPTON 405 
Engineering developments can usually be carried through in 
accordance with a plan carefully prepared in advance, and often this 
can be done most effectively by a competent self-contained group like 
a company or a bureau. But scientific discovery, in its very nature 
and as proved by experience, does not progress according to pre- 
conceived plan and is stified if attempts are made to control the free 
initiative of the research workers or to limit the freedom of com- 
munication between them. This is one reason why most of the 
fundamental new scientific discoveries have originated in free environ- 
ment of the universities rather than in the quite properly more con- 
trolled atmosphere of industrial or governmental laboratories. When, 
however, it comes to practical applications and engineering develop- 
ments, then thorough planning and control are essential to efficiency. 
Thus the third lesson which I would draw is this: to the extent that 
we wish fundamental science to advance, we must maintain the 
maximum of opportunity for competent scientists to follow their 
own bent and to communicate freely with each other. 
The fourth lesson is, at first sight, in apparent contradiction with 
the last, but actually it is not. It is that teamwork has proven 
extraordinarily effective in producing results. To a certain extent, 
of course, teamwork implies control, which I have just decried. But 
what I mean by a team is a group of competent and imaginative 
project leaders whose skills and knowledge supplement each other 
and are supported by the technical assistance required to carry out 
their ideas. Such groups actually provide the maximum opportunity 
for quick initiative and for stimulating interchange of ideas. As 
science becomes more complex, or as its practical applications come 
more to the fore, the advantages of such team organization become 
more pronounced. 
The fifth lesson, which needs no amplification, is the increasing ex- 
tent to which a basic advance in theory or technique in one branch of 
science is likely to provide new concepts or new tools which can open 
up new frontiers for exploration and exploitation in other fields of 
science orart. Thisisnotanewidea. It was for this reason, for exam- 
ple, that the Rockefeller Foundation established, under the National 
Research Council, the great program of National Research Fellowships 
which were largely effective within a decade or two in raising the 
United States from a third-rate to a first-rate world position in science. 
The Rockefeller Foundation hoped, by thus stimulating advance in 
the fundamental sciences, to uncover new avenues of approach to 
medical science—a hope that has been brilliantly justified. And 
another lesson which can be drawn comes from the realization that an 
astonishing proportion of today’s leaders in American science, and of 
the project leaders who were the key men in our great scientific pro- 
