I have a question, more related to a mathematical aspect of physics, which seems I am not understanding very well.

So, by applying Galilean transformation between two reference frames, which move at the speed $\epsilon$ relative to each other, the Lagrangians of a free particle looked from these two systems differ by

$$\Delta L=\frac{\partial L}{\partial (v^2)}2\vec{v}\cdot\vec{\epsilon}, \qquad \vec{v}=\frac{d \vec{x}}{d t},\qquad v:=|\vec{v}|. $$

On the other hand, it has to be

$$\Delta L=\frac{d F}{d t}.$$

Now in many texts, I see the argument that this is true only if $\frac{\partial L}{\partial (v^2)}$ is independent of $v$. It might be due to the lack of math skills, but this is not obvious for me.

Example. On contrary, let's assume that $$\frac{\partial L}{\partial (v^2)}=a v$$ and 1D motion, then we have the condition

$$\frac{d F}{d t}=2\epsilon a\frac{d x}{d t}|\frac{d x}{d t}|$$


$$\frac{d F}{d t}=\frac{\partial F}{\partial t}+\sum_i\frac{\partial F}{\partial u_i}\frac{d u_i}{d t}$$

Where $u_i$ are all possible time dependent functions, on which $F$ is dependent.

Could anyone help me, and explain why can't we express any of the partial derivatives in $\frac{d F}{d t}$ via $\frac{d x}{d t}|\frac{d x}{d t}|$?


  1. Landau & Lifshitz, Mechanics, $\S$3.

That's a very good question. I have explicitly done the analysis suggested by OP in Section 2 of my Phys.SE answer here. The conclusion is, unsurprisingly, that imposing that all Galilean transformations should be quasi-symmetries implies that the Lagrangian $L(v^2)$ must be an affine function of $v^2$.

  • $\begingroup$ Thanks Qmechanic, your reply on the topic you linked was so far giving me the most satisfied answer. Also, in the approach I took in the beginning, to show that the partial derivatives of $\frac{dF}{dt}$ can be expressed through $\frac{\partial x}{\partial t}\frac{\partial x}{\partial t}$, what I get in the end is that the function $F$ is expressed as an integral, which depend on the path taken by the particle. This is in the clear contradiction with the fact that variation of the action $S$, must be the same for all observers along the unique path. $\endgroup$ – 600nebo Nov 27 '18 at 8:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.