The following equation(subset of Maxwell's equations of electromagnetic wave(s)) is said held in free space.

$$\underbrace{ \color{fuchsia}{\nabla^2\mathbf{H}_{}=\sigma\mu{\partial\mathbf{H}_{}\over\partial\mathrm{t}}+\epsilon\mu{\partial^2\mathbf{H}_{}\over\partial\mathrm{t^2}}} }_{\text{I want to derive this equation}}\tag{1}$$

It is said that above equation can be derived from the following equation but still I am in dought that this statement is true.


$$\underbrace{\nabla\times\left(\nabla\times\mathbf{A}_{}\right)=\nabla\left(\nabla\cdot\mathbf A\right)-\nabla^2\mathbf A}_{\text{General formula for any vector}}\tag{3}$$

$$\begin{cases}\nabla\cdot\mathbf E=0\\\nabla\cdot\mathbf H=0\end{cases}~~\leftarrow~~ \text{Holds at freespace} \tag{4}$$

$$ \nabla\times \left(\nabla\times \mathbf{E}_{} \right)= \nabla \left( \nabla\cdot \mathbf{E}_{} \right) -\nabla^2 \mathbf{E}_{} $$

$$ \nabla\times \left(\nabla\times \mathbf{E}_{} \right)= -\nabla^2 \mathbf{E}_{} $$

$$ ~~\iff~~ \nabla^2 \mathbf{E}_{} = -\nabla\times \left(\nabla\times \mathbf{E}_{} \right) $$

$$ = \underbrace{\sigma\mu{\partial\mathbf{E}_{}\over\partial\mathrm{t}}+\epsilon\mu{\partial^2\mathbf{E}_{}\over\partial\mathrm{t^2}}}_{\text{I've already derived it} } $$

Intuitively I think that it is dangerous to simply substituting $~ \mathbf{H}_{} ~$ for $~ \mathbf{E}_{} ~$ to obtain the first equation(pink-marked).

How eqn1 can be derived actually?


2 Answers 2


I don't know where you've seen this, but that source is wrong. This is the equation in a conducting medium, which means you have current $$ \vec j = \sigma \vec E $$ (Ohm's law) so no vacuum. Furthermore, if $\epsilon,\mu\neq\epsilon_0,\mu_0$ this means you are also assuming an isotropic dielectric and magnetic medium. In general, if you have to appeal to $\vec D,\vec H$, it's because you are not in vacuum.

For the derivation of your formula, take the curl of Maxwell-Ampere's law, with the current replaced by Ohm's law, and substitute the curl of $\vec E$ by Faraday's law.

Hope this helps and tell me if you need more details.


$\nabla × \vec{B} = \mu (\sigma \vec{E}) + \mu \epsilon \frac{\partial \vec{E} }{\partial t} $

Take the curl and sub gauss law for magnetism

$-\nabla^2 \vec{B} = \mu \sigma \nabla × \vec{E} + \mu \epsilon \frac{\partial \nabla × \vec{E} }{\partial t} $

$-\nabla^2 \vec{B} = -\mu \sigma \frac{\partial \vec{B}}{\partial t} - \mu \epsilon \frac{\partial^2 \vec{B}}{\partial t^2} $

$\nabla^2 \vec{B} = \mu \sigma \frac{\partial \vec{B}}{\partial t} + \mu \epsilon \frac{\partial^2 \vec{B}}{\partial t^2} $

  • $\begingroup$ I didn't know that curl operator is a linear operator. $\endgroup$ Apr 29, 2022 at 12:05
  • $\begingroup$ I think it also best to use faradays law to solve for B, as that way you can identify the correct orientation and magnitude $\endgroup$ Apr 29, 2022 at 12:21

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.