We started the last chapter by formally deriving the Galilean transformation in an attempt to try and identify any wrong assumptions in the argument. We came to the conclusion that the only part of the argument that we could throw away was that time was the same in all frames of reference. This means that in order to find the correct transformation we have to find an alternative assumption to this one. In fact we have two alternative choices; either we can take the velocity of light or the interval as an invariant. Clearly there is no conflict between these two statements, since we derived the second from the first. Which we take is a matter of personal choice, so let us take the invariance of the velocity of light, simply on the grounds that it is the experimentally observed invariant.