A javelin graduated in centimeters is thrown downwards from the top of a tower of height h. Initially the centimeter marks pass an observer at the top of the tower with frequency f, speed c (not the speed of light of course) and "wavelength" λ (1cm):

f = c/λ

What are the frequency f', speed c' and "wavelength" λ' as measured by an observer on the ground? Newton's theory gives a straightforward answer (it is assumed that cc'-c):

f' = f(1+gh/c^2) = (c+v)/λ
c' = c(1+gh/c^2) = c+v
λ' = λ

where v=c'-c is the increase in speed.

Now let us answer David Morin's question:

http://www.personal.kent.edu/~fwilli...Relativity.pdf
David Morin: "A light source on top of a tower of height h emits flashes at time intervals Ts. A receiver on the ground receives the flashes at time intervals Tr. What is Tr in terms of Ts?"

If we compare the flashes to javelins thrown downwards at time intervals Ts (this comparison is allowed by Newton's emission theory of light), we see that each flash gets blue shifted because its speed at the receiver is c'=c(1+gh/c^2) (now c is the speed of light) and accordingly its frequency is f'=c'/λ=f(1+gh/c^2) - a result confirmed by the Pound-Rebka experiment.

On the other hand, the time intervals between the flashes remain unchanged: Tr=Ts. There is no gravitational time dilation.

Pentcho Valev