One of the oddest things about this state of affairs is that empirical evidence for or against it is possible, and that there is actually some slight evidence in its favor. If all “clocks” are slowed down at a great distance from the observer, this will apply to the periodic motions of atoms, and therefore to the light which they emit. Consequently all rays of light emitted by distant objects ought, when they reach us, to look rather more red or less violet than when they started. This can be tested by the spectroscope. We can compare a known line, as it appears in the spectrum of a spiral nebula, with the same line as it appears in a terrestrial laboratory. We find, as a matter of fact, that in a large majority of spiral nebulæ there is a considerable displacement of spectral lines towards the red. The spiral nebulæ are the most distant objects we can see: Eddington states that their distances “may perhaps be of the order of a million light-years.” (A light-year is the distance light travels in a year.) The usual interpretation of a shifting of spectral lines towards the red is that it is a “Doppler effect,” due to the fact that the source of light is moving away from us. But one would expect to find the nebulæ just as often moving towards us as moving away from us, if nothing operated but the law of chances. If the world is such as De Sitter says it is, the spectral lines of the spiral nebulæ will be displaced towards the red owing to the slowing down of distant clocks, even if in fact they are not moving away from us. This, for what it is worth, is an argument in favor of De Sitter. The same facts afford another argument in favor of De Sitter, for another reason. If, at a given moment, a body is at rest relatively to the observer, and at a distance from him, it will (in the absence of counteracting causes) not remain at rest from his point of view, but will begin to move away from him, and will continue to move away faster and faster; the further it is from him, the more its retreat will be accelerated. For bodies which are not too distant from each other, gravitation may overcome this tendency; but as this tendency increases with the distance, while gravitation diminishes, we should expect to find very distant bodies receding from us if De Sitter’s theory is right. Thus we have two reasons for the displacement of spectral lines in spiral nebulæ: one, the slowing down of time; the other, the movement away from us which we should expect at distances too great for gravitation to be sensible. However, it cannot be said that the argument, on either ground, is very strong. Eddington gives a list of forty-one spiral nebulæ, of which five have their spectral lines shifted towards the violet, not towards the red. Thus the material is neither very copious nor quite harmonious.