Page:Elementary Text-book of Physics (Anthony, 1897).djvu/413

§ 325] to interference as explained, light should be restored by suppressing the interfering waves. If a second screen be placed at $$m'n'$$ so as to cut off the waves proceeding from points above $$b,$$ waves from points between $$a$$ and $$b$$ will no longer be neutralized, and light should fall at $$P.$$ To test this conclusion the edge of a flat flame may be observed through a narrow slit in a screen. Instead of the narrow edge of the flame, a broad luminous surface is seen, in which the brightness gradually diminishes from the centre towards the edges. If we consider the wave-front just entering the slit, it will be seen that elementary waves proceed from all points of it, and the slit being very narrow it is only in very oblique directions that pairs of these waves can meet in opposite phases. Hence light proceeds in oblique lines behind the screen, and from our habit of locating visible objects back along the line of light entering the eye, the flame appears as a broad surface. It will be seen by reference to Fig. 93 that the elementary wave that first reaches $$P$$ is the one to which the disturbance there is principally due. Other waves arriving later find there the opposite phase of some wave that has preceded them. When the velocity in all directions is the same, the first wave to reach $$P$$ is the one that starts from the foot of a perpendicular let fall from $$P$$ on the wave-front. Hence light is said to travel in straight lines perpendicular to the wave-front. If, however, light does not move with equal velocities in all directions, the last statement is no longer true, as will be seen from Fig. 95. Here $$mn$$ represents a wave-front, proceeding towards $$P$$ in a medium in which the velocities in different directions are such that the elementary wave-surfaces are ellipsoids. The ellipses in the figure may be taken as sections of these ellipsoids. The wave first to reach $$P$$ is not the one that starts from A at the foot of the perpendicular, but from $$A'.$$ It is from $$A'$$ that $$P$$ derives its light, and the line of propagation is no longer perpendicular to the wave-front. The demonstration here given fails when applied to a plane