The limitation $\sqrt{-g} = 1$ is not essential if we take account of the confusion of tensor-densities with tensors referred to at the end of \SecRef{49}. It will be remembered that the force $(X, Y, Z)$ occurs because we attribute to our mesh-system an abstract Galilean geometry which is not the natural geometry. Either inadvertently or deliberately we place ourselves in the position of an observer who has mistaken his non-Galilean mesh-system for rectangular coordinates and time. We therefore mistake the unit mesh for the unit of natural volume, and the density of the energy-tensor~$\mf{T}_{\mu}^{\nu}$ reckoned per unit mesh is mistaken for the energy-tensor itself~$T_{\mu}^{\nu}$ reckoned per unit natural volume. For this reason the conservation of the supposed energy-tensor should be expressed analytically by $\dd\mf{T}_{\mu}^{\nu}/\dd x_{\nu} = 0$ and when a field of force intervenes the equations of classical hydrodynamics should be written \[ \frac{\dd}{\dd x_{\nu}} \mf{T}_{\mu}^{\nu} = \mf{T}_{4}^{4}(-X, -Y, -Z, 0), \Tag{(55.51)} \] the supposed density~$\rho$ being really the ``density-density'' $\rho\sqrt{-g}$ or~$\mf{T}_{4}^{4}$\footnotemark.\footnotetext {It might seem preferable to avoid this confusion by immediately identifying the energy, momentum and stress with the components of~$\mf{T}_{\mu}^{\nu}$, instead of adopting the roundabout procedure of identifying them with~$T_{\mu}^{\nu}$ and noting that in practice $\mf{T}_{\mu}^{\nu}$~is inadvertently substituted. The inconvenience is that we do not always attribute abstract Galilean geometry to our coordinate-system. For example, if polar coordinates are used, there is no tendency to confuse the mesh $dr\, d\theta\, d\phi$ with the natural volume $r^{2} \sin\theta\, dr\, d\theta\, d\phi$; in such a case it is much more convenient to take $T_{\mu}^{\nu}$ as the measure of the density of energy, momentum and stress. It is when by our attitude of mind we attribute abstract Galilean geometry to coordinates whose natural geometry is not accurately Galilean, that the automatic substitution of~$\mf{T}_{\mu}^{\nu}$ for the quantity intended to represent~$T_{\mu}^{\nu}$ occurs.} Since \Eq{(55.1)}~is equivalent to \[ \frac{\dd}{\dd x_{\nu}} \mf{T}_{\mu}^{\nu} = \{\mu\nu, \alpha\} \mf{T}_{\alpha}^{\nu}, \Tag{(55.52)} \] the result~\Eq{(55.4)} follows irrespective of the value of~$\sqrt{-g}$.