254 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1944 
ernment, when finance is involved, may result in the essential help 
coming too late. One day we shall learn to trust our scientific advisers 
with a reasonable fraction of our money on a block-grant basis and 
ask no account except at longish intervals. 
RESEARCH ON A LARGE SCALE 
I come now to that class of research for aeronautics whose scale is 
such that success depends on planning of large experimental resources 
and on planning so that application to practice may meet the foreseen 
needs of design and its capacity to exploit new discoveries. 
We must, in my view, plan research for aeronautics in three phases. 
First, we must relate all our main effort to advances in basic theory. 
Odd pieces of information without a clear, strong framework are 
worth little. Second, we must provide the experimental information 
by which theory may be built up and its limitations recognized and 
reduced. Third, we must ensure that experimental application is 
made in such conditions that the practical value of the theory is 
confirmed. 
There are three chief parties to this undertaking: first, those who 
are by trade workers in the field of theory and those who have the 
flair for the associated exploration by experiment; second, those who 
make use of the results in the design and construction of aircraft; and 
third, those who use the aircraft and on whom we rely to exploit the 
product of the efforts of the first and second. The extent to which 
these should enter into planning of research can be illustrated by an 
example—the problem of reducing the cooling drag of power plants. 
That it is possible to reduce the power wasted in cooling an airplane 
power plant to 2 percent or less of the brake horsepower was estab- 
lished many years ago. Indeed, it was shown that at flight speeds that 
were then within sight and have now been passed the cooling could be 
made to help to propel the airplane. But the cooling of a power plant 
is a matter that goes far beyond broad conceptions of this kind. It 
involves complex flows of air and liquids, demanding regulation to 
meet the varying conditions of flight and high standards of reliability 
in functioning and of ease of maintenance, which are of the greatest 
concern to the user. 
It was not until other developments had reduced the rest of the drag 
so much that the power-plant drag was a dominating factor that the 
designer became convinced that the problem demanded his serious 
attention. He has finally succeeded in producing cooling systems that 
are no less reliable and have a much lower drag. The user accepts 
the slight additional embarrassment to maintenance in return for the 
higher speed and greater range. 
