V O L U M E 5 , D O C U M E N T 2 5 7 a 7

Vol. 5, 257a. To Vladimir Varic;ak

Zurich, 3 March 1911

Highly esteemed Colleague!

I thank you very much for your letter and the

paper.[1]

I have now read the beau-

tiful study by Lewis and

Tolman,[2]

but I cannot understand how you can draw from

this an support endorsement of your opinion. I want to justify my opposite opin-

ion

explicitly.[3]

Let S be a nonaccelerated frame of reference, in which there are clocks of the

same kind at rest with it. Let these be synchronized, e.g. by means of light rays, so

that they show the time of S. Let the rod AB be in uniform motion relative to S. Its

“real” length, i.e., the length measured by itself, be l. Then it follows from the rel.

theory in the well-known way that its length with respect to S is . This

means: if one determines those clocks in S, which show the same state of hands,

when the points A and B are passing by them, then the distance of these points mea-

sured in S is . The contraction is observable by measurement, hence

“real.” In order that you see that the contraction is not simply affected by the defi-

nition of simultaneity in S, i.e. of a purely conventional nature, I add: it is impossi-

ble, to reset the clocks in such a way, that even after this resetting the rod always

has the length with respect to S, if it has the velocity ±v measured by means of

the clocks. From this one can conclude with Ehrenfest that a rotation without elas-

tic deformation is excluded in the theory of relativity, if you assume in addition that

a transversal contraction does not take

place.[4]

One cannot ask whether one has to

conceive of the contraction as a consequence of the modification of the molecular

forces or as a kinematic consequence from the foundations of the theory of

relativity.[5]

Both points of view are justified side by side. The latter point of view

corresponds roughly to the one of Boltzmann, who treats the dissociation of gases

in a molecular-theoretic manner; this is completely justified, although one can de-

rive the laws of dissociation from the second law without

kinetics.[6]

A principal

A B

U1

U2 U3

S

l 1

v2

c2

---- -–

l 1

v2

c2

---- -–

l'