FEBRUARY 25, 1912] 
preted intelligibly only in terms of some 
such medium. The abandonment of this 
hypothesis reminds one of Baron Miin- 
chausen’s feat performed while he was ma- 
king his escape from prison. Since your 
historical reading may not have extended 
to the autobiography of this famous man, 
I may be permitted to relate that the 
Baron was letting himself down from the 
windows of a high tower by a rope, and 
when he reached the end of it he found 
that he still had a long distance to go. 
The last part of the descent was particu- 
larly difficult, so to get rope enough he 
ingeniously spliced on an additional piece, 
which he obtained by cutting off the part 
above him. 
The principle of relativity in its meta- 
physical form ignores the accelerations of 
bodies. It is true that the experimental 
results to which the principle has been ap- 
pled with such success are such that the 
study of acceleration in terms of the theory 
of relativity has not become necessary. 
But is it not reasonable to suppose that 
when suitable experiments have been in- 
vented and tried to test the effect of the 
acceleration of a system on the progress of 
light in it, it may be found that an effect 
can be detected? _Some effect may be de- 
tected, for example, due to the rotation of 
a body. I have never been able to perceive 
any sound objection to Newton’s assertion 
that we have evidence of absolute rotation 
by the observation of centrifugal force, and 
if a fixed direction of an axis and an abso- 
lute velocity of rotation can be determined 
in a mechanical system when accelerations 
are taken into consideration, why should 
the principle of relativity be treated as 
having universal validity ? 
But, after all, these questions raised by 
the development of the principle of rela- 
tivity are of secondary importance. The 
central question is whether or not this prin- 
SCIENCE 
291 
ciple can ever furnish a satisfactory ex- 
planation of natural phenomena. The for- 
mulas derived from it are evidently merely 
descriptive. This may be said with truth 
about all the formulas in which the general 
theories of physics have been embodied. 
Kirchhoff designates, as the task of the sci- 
ence of mechanics, the description of the 
motions which occur in nature completely 
and in the simplest possible way. This 
assertion that the task of the theoretical 
physicist is done when he has reduced the 
phenomena with which he is dealing to a 
set of formulas, or, as we may say, when he 
has constructed an ideal model which will 
reproduce the phenomena, is one to which 
we would all assent in general. At the same 
time most of us would reserve the right to 
criticize each model thus presented, and to 
give to one or the other a preference based 
on considerations which are not necessarily 
limited to the simplicity of the model or to 
the completeness with which it reproduces 
the phenomena. Surely an additional test 
of the value of the model will be the intelli- 
gibility of the elements of which it is com- 
posed. 
This last test has been generally met in 
the models which have been proposed as 
descriptions of natural phenomena. We 
can understand from what we see and feel 
what is meant by the motions of elastic 
spheres, and the model which uses them to 
represent the behavior of a gas is not only 
competent to reproduce the behavior of a 
gas, but is intelligible in the elements of 
which it is composed. The model of the 
elastic solid ether, incomplete and objec- 
tionable as it became when the subject of 
optics was enlarged and developed, was in- 
telligible in its elements. The model of 
electromagnetic operations embodied in 
Maxwell’s formulas is also one which is 
thus intelligible in its elements. When I 
say this I do not mean that we know all 
