It is a curious situation, especially in view of the fact that Einstein’s law, which has superseded Newton’s, is not very suitable for use as an astronomical clock, as has been pointed out by Larmor. Perhaps the physicist will soon be able to use the atom as the theoretical clock for physics, and we can go on using the corrected rotation of the earth as our practical standard. There is a faint chance, for instance, that if physics can invent some way of measuring the minute time intervals along the track of an electron, then electrons might be used as giving the fundamental measure of time. Thus if the velocity of an electron were first measured by some indirect method the electron itself might then be used as a clock. But in the meantime the astronomers should make a formal announcement to the Royal Society of what they have been up to. It then might be found necessary to appoint a commission to discover exactly what physics is now doing. For by using an astronomical clock of the new type it is assuming classical laws while researching on processes which are already known to undermine the absolute validity of these laws. Theoretical physics cannot hope to clear up its fundamental problems until it has considered exactly what is involved in this suspicious procedure. Like most professions, physics includes a good deal of bluff, but unlike the others physics is now occupied on a campaign to get rid of all pretence. For instance, physical text-books have been filled for twenty years with phrases of this kind: “an electron with a velocity of so many cms per sec.” Yet the professors omitted to tell their students the awful secret that this hypothesis of electron velocities is one that has never yet received direct experimental confirmation. To-day a reaction has set in and the demand is being made that physical theory shall not make use of conceptions that do not correspond to directly observed quantities. Thus the latest theories of the atom have eliminated the idea of electron orbits because it was realized that these were nothing more than a mathematical trick for calculating something quite different: the wave-length of the light an atom can emit. In place of the orbits it is hoped to substitute something which only makes use of the directly-observed features of the atom, but this new picture is not complete.