5
$\begingroup$

On Section 3.2 of Carroll's Spacetime and Geometry, he works on the proof of how Christoffel symbols transform under a change of coordinate system, in order to the covariant derivative of a tensor continue to be a tensor. On equation (3.9), he writes:

$$\Gamma^{\nu'}_{\mu' \lambda'}\dfrac{\partial x^{\lambda'}}{\partial x^\lambda}V^\lambda + \dfrac{\partial x^\mu}{\partial x^{\mu'}}V^\lambda\dfrac{\partial}{\partial x^\mu}\dfrac{\partial x^{\nu'}}{\partial x^\lambda} = \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda}V^\lambda. \tag{3.9}$$ This equation must be true for any vector $V^\lambda$, so we can eliminate that on both sides. Then the connection coefficients in the primed coordinates may be isolated by multiplying by $\dfrac{\partial x^\lambda}{\partial x^{\sigma'}}$ and relabeling $\sigma' \rightarrow \lambda'$. The result is $$\Gamma^{\nu'}_{\mu' \lambda'} = \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} + \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial^2 x^{\nu'}}{\partial x^\mu\partial x^\lambda}.\tag{3.10}$$

However, as far as I can see, there should be a minus sign in the middle of the expression, not a plus. Was that a typo by Carroll?

Besides, if I choose to isolate the unprimed Christoffel symbol instead of the primed one, I obtain

$$\Gamma^{\nu}_{\mu \lambda} = \dfrac{\partial x^{\mu'}}{\partial x^{\mu}}\dfrac{\partial x^{\lambda'}}{\partial x^{\lambda}}\dfrac{\partial x^{\nu}}{\partial x^{\nu'}}\Gamma^{\nu'}_{\mu' \lambda'} + \dfrac{\partial x^\nu}{\partial x^{\nu'}}\dfrac{\partial^2 x^{\nu'}}{\partial x^\mu\partial x^\lambda}$$

Shouldn't there be some kind of symmetry between the transformation formulas? In the sense that if in this last equation I just prime what is unprimed and unprime what is primed, shouldn't I obtain the transformation law from the unprimed Christoffel symbols to the primed ones? (I really don't know if that's true, but intuitively it seems so, because the indices shouldn't matter).

$\endgroup$
3
  • 1
    $\begingroup$ Related : Why do the Christoffel symbols not transform as a tensor?. $\endgroup$
    – Voulkos
    Commented Jan 28, 2022 at 14:02
  • $\begingroup$ Yes. you are right. In the version reissued by Cambridge University Press at 2019, (3.10) is wrong, the sign should be minus. It is interesting that in the version published by Pearson Education at 2004, (3.10) is right. $\endgroup$
    – Sean
    Commented Jan 28, 2022 at 14:50
  • $\begingroup$ Oh, good to know it is indeed a typo, I thought I was missing something here. Thanks for the clarification! Furthermore, can you say something regarding the symmetry I discuss in the end of the question? $\endgroup$ Commented Jan 28, 2022 at 14:52

1 Answer 1

9
$\begingroup$

Yes. You are right. The sign should be minus.

It is interesting that in the original version published by Addison Wesley in 2004, the sign in (3.10) is right. But in the following versions reissued by Pearson in 2014, and Cambridge at 2019, the signs are wrong.

So, the correct one is $$ \Gamma^{\nu'}_{\mu' \lambda'} = \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} - \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial^2 x^{\nu'}}{\partial x^\mu\partial x^\lambda}.\tag{3.10} $$ It can also be rewritten as $$ \begin{split} \Gamma^{\nu'}_{\mu' \lambda'} &= \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} - \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial^2 x^{\nu'}}{\partial x^\mu\partial x^\lambda} \\ &= \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} - \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial}{\partial x^\lambda}\left(\dfrac{\partial x^{\nu'}}{\partial x^\mu}\right) \\ &= \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} - \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial}{\partial x^{\lambda'}}\left(\dfrac{\partial x^{\nu'}}{\partial x^\mu}\right) \\ &= \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} - \dfrac{\partial}{\partial x^{\lambda'}}\left(\dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^{\nu'}}{\partial x^\mu}\right) + \dfrac{\partial^2 x^\mu}{\partial x^{\lambda'}\partial x^{\mu'}}\dfrac{\partial x^{\nu'}}{\partial x^{\mu}} \\ &= \dfrac{\partial x^\mu}{\partial x^{\mu'}}\dfrac{\partial x^\lambda}{\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^\nu}\Gamma^{\nu}_{\mu \lambda} + \dfrac{\partial^2 x^\nu}{\partial x^{\mu'}\partial x^{\lambda'}}\dfrac{\partial x^{\nu'}}{\partial x^{\nu}}. \end{split} $$

$\endgroup$
2
  • $\begingroup$ That's precisely what I was looking for, altogether with the rewritten expression which shows the symmetry I mentioned in the end of my question. Thanks a lot for your help! $\endgroup$ Commented Jan 28, 2022 at 15:44
  • $\begingroup$ Glad to hear that I can help! $\endgroup$
    – Sean
    Commented Jan 29, 2022 at 2:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.