Many books say that entropy measures energy dispersal or energy spreading. > Key point 1.4: Thermodynamic processes entail **spatial redistributions** of internal energies, namely, the **spatial spreading** of energy. Thermal equilibrium is reached when energy has spread maximally; i.e., energy is distributed equitably and entropy is maximized. Thus, entropy can be viewed as a spreading function, with its symbol S standing for spreading. Although not Clausius’ motivation for using S, this can serve as a mnemonic device. Energy spreading can entail energy exchanges among molecules, electromagnetic radiation, neutrinos, and the like. [link>>>][1] Spreading *where*? Spreading in space? What kind of space? I mean, that in my mind "the spread of energy" is some kind of "density of energy in some *volume*". Why then in $\mathrm dS=Q/T$ the $Q$ factor gets divided by $T$ and not by some "abstract volume measure"? Why is *energy spreading* measured in units $\mathrm{J/K}$ and not in $\mathrm{J/m^3}$ ? Or, maybe, the temperature in $\mathrm S=Q/T$ is treated exactly like **some kind** of volume. I can certainly think in this way: The greater the $T$, the greater the number of "microstates" consistent with that $T$. P.S. I'm aware about Boltzmann definition of entropy. I understand it. I just can't handle entropy in phenomenological level (classical thermo level). [1]: http://energyandentropy.com/resources/Leff2012Phys.-Teach-1.pdf