or diverted to our eyes. In Fig. 78 I have tried to draw four rays of light like strips of cards; starting from the Sun in the middle and going out sideways from him, so that if they are not bent or reflected, they will never reach the Earth, which is supposed to be on the right of the picture. But I have supposed them to be reflected (by solid particles in the surrounding corona) at the places shown, and to come earthwards. Rays which are edgewise and cannot thus be reflected are left out of the picture. Now, on reaching the Earth, if we try to "polarize" these four rays by bending them downwards, the top and bottom rays will bend, but those at the sides are edgewise to this direction of bending and refuse. Hence in our "polarized" picture we should get light from the corona at the top and bottom of the picture, but not from the sides. Another picture with sideways bending would show the opposite.
The sunlight is originally a bundle of rays with their edges in all directions; but in the encounter with the solid particle those rays which are edgewise refuse to bend, leaving only the others, so that the ray is now "polarized" as it is called. If the corona is merely a mass of gas, the light will not be polarized. If the corona is partly gaseous and partly solid particles, we shall get some light from the solid particles which is polarized and some from the gas which is not. How much is there of each kind? Here again we come to the question of measurement, of finding out how much of a thing there is, and by arranging the particular photographs to be taken we can find it out by measure-
