Does anyone here know a source where I can find Boltzmann’s original derivation (using primarily thermodynamic arguments) to the Stefan–Boltzmann law (the radiant power emitted by a body in thermal equilibrium is proportional to the fourth power of the absolute temperature)? I have done some research on Google, but I have not found anything much enlightening about this.


1 Answer 1


Boltzmann built on two known properties of cavity radiation.

(1) The energy density, $u$, defined as $u = U/V$, depends only on temperature, $T$.

(2) The radiation pressure, $p$ is given by $p= u/3$. Radiation pressure was given a firm basis c1862 by Maxwell. The factor of 1/3 arises because of the three-dimensionality of the cavity, in which radiation is propagating in all possible directions. [It's easy for us now to derive this equation by considering the cavity as containing a photon gas.]

Boltzmann (1884) used a thought-experiment in which a cavity is fitted with a piston, and we take the radiation inside it through a Carnot cycle. On a $p–V$ diagram the isothermals are just horizontal lines, because $u$ is constant so $p$ is constant. The heat input along the top (temperature $T$) isothermal is $\Delta U + p \Delta V$. This works out to be $4 p \Delta V$. If the lower temperature isothermal is only slightly lower, at temperature ($T - d T$) then the cycle appears as a thin horizontal box, and the net work done during the cycle is simply $dp \Delta V$ in which $dp$ is the infinitesimal pressure difference between the two isothermals. We can then apply the definition of thermodynamic temperature in the form

$$\frac{dT}{T} = \frac{\text{work}}{\text{heat input}} = \frac{dp \Delta V}{4 p \Delta V} = \frac{dp}{4 p}~.$$

This integrates up easily to give the Stefan-Boltzmann law.

A modern thermodynamic derivation would probably start from the fundamental equation (embodying first and second laws of thermodynamics)… $$\text {d}U = T \text{d}S - P \text{d}V. $$ Therefore for an isothermal change $$\left(\frac{\partial U}{\partial V} \right)_T = T \left(\frac{\partial S}{\partial V} \right)_T - p .$$ Using one of the Maxwell relations, this becomes $$\left(\frac{\partial U}{\partial V} \right)_T = T \left(\frac{\partial p}{\partial T} \right)_V - p .$$ But for the radiation, $$\left(\frac{\partial U}{\partial V} \right)_T = \left(\frac{\partial (uV)}{\partial V} \right)_T = u = 3p \ \ \ \ \ \text{and} \ \ \ \ \ \left(\frac{\partial p}{\partial T} \right)_V = \frac {\text{d} p}{\text{d}T}. $$ Making these substitutions and tidying, $$ 4p = T \frac {\text{d} p}{\text{d}T}.$$ As before, separation of variables and integration gives the Stefan-Boltzmann law.

  • $\begingroup$ How exactly have you written dT/T = Work/Heat input?I get it intuitively, but mathematically? $\endgroup$ Jan 29, 2022 at 13:03
  • $\begingroup$ @lazearoundall day. Kelvin temperature can be defined by $$\frac {T_1}{T_2}=\frac {Q_1}{Q_2}$$ in which $Q_1$ is the heat input to the Carnot engine at temperature $T_1$ and $Q_2$ is the heat excreted at temperature $T_2$. Subtracting 1 from both sides of the equation, we can write it as $$\frac {T_1-T_2}{T_2}=\frac {Q_1-Q_2}{Q_2}.$$ Using energy conservation (or the first law of thermodynamics, noting that there is no change in internal energy over a complete cycle), $(Q_1-Q_2)$ is the work output. $\endgroup$ Jan 29, 2022 at 13:19
  • $\begingroup$ @lazearoundallday Please see above comment. $\endgroup$ Jan 29, 2022 at 14:09
  • $\begingroup$ @PhilipWood Could you please provide me any source containing the derivation of the relation between radiation pressure and energy density of radiatons? $\endgroup$ Apr 30, 2022 at 8:40
  • $\begingroup$ @AYM Shahriar Rahman Reif: Fundamentals of Statistical and Thermal Physics (End of scection 9.13), Mandl: Statistical Physics (Section 10.4) $\endgroup$ Apr 30, 2022 at 18:53

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.