For an isolated electron at rest in Galilean coordinates $\kappa_{4} = e/r$, so that $\kappa_{\alpha} \kappa^{\alpha} = e^{2}/r^{2}$. On integrating throughout infinite space the result is apparently \PageSep{212} infinite; but taking account of the finite radius of space, the result is of order~$e^{2}R$. By~\Eq{(90.62)} this represents the part of the (negative) mass of the electron\footnote {This must not be confused with mass of the energy of the electromagnetic field. The present discussion relates to \emph{invariant mass} to which the field contributes nothing.} which is not concentrated within the nucleus. The actual mass was found in \SecRef{80} to be of order~$e^{2}/a$ where $a$~is the radius of the nucleus. The two masses $e^{2}R$ and~$e^{2}/a$ are not immediately comparable since they are expressed in different units, the connection being made by Weyl's constant~$\beta$ whose value is left undecided. But since they differ in dimensions of length, they would presumably become comparable if the natural unit of length were adopted, viz.\ the radius of the world; in that case $e^{2}/a$~is at least $10^{36}$~times $e^{2}R$, so that the portion of the mass outside the nucleus is quite insignificant. The action-principle here followed out is obviously speculative. Whether the results are such as to encourage belief in this or some similar law, or whether they tend to dispose of it by something like a \Foreign{reductio ad absurdum}, I will leave to the judgment of the reader. There are, however, two points which seem to call for special notice--- (1) When we compare the forms of the two principal energy-tensors \begin{align*} T_{\mu}^{\nu} &= -\frac{1}{8\pi} \bigl\{G_{\mu}^{\nu} - \tfrac{1}{2} g_{\mu}^{\nu} (G - 2\lambda)\bigr\}, \\ E_{\mu}^{\nu} &= -F_{\mu\sigma} F^{\nu\sigma} + \tfrac{1}{4} g_{\mu}^{\nu} F_{\alpha\beta} F^{\alpha\beta}, \end{align*} it is rather a mystery how the second can be contained in the first, since they seem to be anything but homologous. The connection is simplified by observing that the difference between them occurs in $\Ham A/\Ham g_{\mu\nu}$ \Eq{(90.51)} accompanied only by a term which would presumably be insensible except inside the electrons. But the connection though reduced to simpler terms is not in any way explained by Weyl's action-principle. It is obvious that his action as it stands has no deep significance; it is a mere stringing together of two in-invariants of different forms. To subtract $F_{\mu\nu}F^{\mu\nu}$ from~$\Star{G}^{2}$ is a fantastic procedure which has no more theoretical justification than subtracting~$E_{\mu}^{\nu}$ from~$T_{\mu}^{\nu}$. At the most we can only regard the assumed form of action~$A$ as a step towards some more natural combination of electromagnetic and gravitational variables.