I am a bit confused as to what I should use to derive the equations of motions from the lagrange equation.

Suppose I have a lagrange function:

$$L(x(t), \dot{x}(t)) = \frac{1}{2}m\dot{x}^2-\frac{1}{2}k(\sqrt{x^2+a^2}-a)$$

Method 1: Principle of least action

$$\delta L = \delta \dot{x}(m\dot{x})-\delta x \frac{kx(\sqrt{x^2+a^2}-a)}{\sqrt(x^2+a^2)}$$

$$\delta W = \int_{t_0}^{t_1} \delta L \ dt$$

After doing integration by parts, i obtain:

$$\delta W= -\int_{t_0}^{t_1} \delta x \biggl[m \ddot{x}+\frac{kx(\sqrt{x^2+a^2}-a)}{\sqrt(x^2+a^2)} \biggl] dt$$

for stationary points, $\delta W = 0$

Hence, inside the integral, $$m\ddot{x}+\frac{kx(\sqrt{x^2+a^2}-a)}{\sqrt(x^2+a^2)} = 0$$

and this is the equation of motion.

Method 2: Euler-Lagrange Equation

Alternatively, we can consider the euler lagrange equation:

$$\frac{\partial L}{\partial x} - \frac{d}{dt}\biggl(\frac{\partial L}{\partial \dot{x}} \biggl) = 0$$

By substituting $L$ into the euler-lagrange equation, we get the same equation of motion:

$$m\ddot{x}+\frac{kx(\sqrt{x^2+a^2}-a)}{\sqrt(x^2+a^2)} = 0$$

So method 2 is a lot easier than method 1, but why do we arrive at the same answer? I have a hunch both methods are essentially calculating the same thing, but I am not sure if this hunch is right because the euler lagrange equations seems a bit too simple as compared to principle of least action. Is there something i'm missing here?

  • 2
    $\begingroup$ You can find the answer in every book for variational calculus. Action is a functional and Euler-Lagrange equation is the criterion for extremum point of this functional in functional space. E-L equation simply means that $\delta S/\delta x=0$. When you vary action directly, you just rederive E-L equation $\endgroup$ Commented Nov 16, 2019 at 16:43

2 Answers 2


Granted appropriate boundary conditions, the stationary action principle and the Euler-Lagrange (EL) equations are both precisely the condition that the functional/variational derivative $$\frac{\delta S}{\delta x^j (t)} \tag{1} $$ vanishes, so they better agree!


First, I think there is something wrong with your partial derivative of the Lagrangian with respect to $x$.

Second, the Euler-Lagrange equations are nothing more than the process that you performed in Method 1, done without committing to a specific form for $L$ but leaving it generic. In your first step you took partial derivatives of $L$ with respect to its position and velocity terms, in your second step you took the velocity derivative and involved it in an integration by parts, where you took a total time derivative and then added a minus sign. If your Lagrangian also involved $\ddot x$ you would then have a $+\frac{\mathrm d^2~}{\mathrm dt^2}\frac{\partial L}{\partial\ddot x}$ term from two integrations by parts, for example.


Your Answer

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

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