You now try observing horizontal rays of light but they seem to bend; that is, a beam of sunshine entering a pinhole on one side of your _camera obscura_ will not strike the wall at a spot exactly opposite but a little below it, if you have instruments sufficiently delicate to show this. You try vertical rays of light in this fashion: You examine with the spectroscope rays of light coming from two sources below (behind) your instrument, one at a distance and the other nearer. Now since you are moving away with increasing speed, the light from the farther source will have to take longer strides to catch up. Or in other words, its frequency will be reduced and it will be shoved toward the red end of the spectrum where the longer waves are. You will have noticed that when a whistling train rushes past the train you are on, the whistle as it comes toward you is raised in pitch (decreased wave-length) and as it recedes from you is lowered in pitch (increased wave-length). Now, says Einstein to himself, if my Principle of Equivalence is correct and there is no difference between (1) weight and (2) the accelerated upward movement of an observer, then all the optical effects that I have thought out in the second case must apply to the first, that is, to gravitation. It must follow that a ray of light passing through a gravitational field will be bent out of its course as though it were attracted by the heavy body. This prediction has been verified. It must further follow that light proceeding from a heavy body like the sun or a star will be held back or slowed up by the attraction of gravitation, and the spectral lines will be displaced toward the left as compared with the same lines in the spectrum of an earthly light. Now such displacement has been observed in stellar spectra but it does not seem to be of the right value to satisfy Einstein’s equation and it has not been observed in sunlight. The remarkable thing about it is that Einstein, by following a line of reasoning somewhat like that which I have crudely outlined, not merely supplied an explanation for phenomena that had been observed but could not be explained (such as the discrepancy in the orbit of Mercury) but he provided in advance the explanation for phenomena that had never been observed until he directed attention to it (such as the deflection of starlight by the sun). Sir Oliver Lodge says of this:[5] Before Einstein’s prediction nothing of the kind had been seen, nothing of the kind had been looked for, nor, so far as it is known, had such an amount of deflection been suspected. Whatever may ultimately be thought of the validity of Einstein’s views as a whole it is evident that he has worked out a mathematical method of unprecedented power and wide usefulness. Professor Bumstead of Yale says: