Place, then, the globe at the south side of the rim of the table, with its axis inclining towards its surface, and its poles always pointing in the same general direction, not following the circuit of the orbit, and set it in motion towards the east, revolving rapidly on its axis as it moves. While directly south of the light, it would be found that the north pole would be illuminated, while no revolution on the axis would bring the south pole within the circle of the light. This is when a line drawn from the axis of the globe would cut the lamp, were the inclination brought as low as the surface of the table. Next set the globe in motion, following the rim of the table, and proceeding to the east or right hand, keeping its axis always looking in the same general direction, or in an attitude that would be parallel to a north and south line drawn through the sun, were the inclination as low as the surface of the table. This movement would be, in one sense, sideways, the circle of light gradually lessening around the north pole, and extending towards the south, as the globe proceeded east and north, diminishing the length of the days in the northern hemisphere, and increasing them in the southern. When at east, the most direct rays of the light would fall on the equator, and the light would cut the two poles, rendering the days and nights equal. As the globe moved north, the circle of light would be found to increase around the _south_ pole, while none at all touched the _north_. When on the north side of the table, the _northern_ pole of the globe would incline so far from the sun as to leave a space around it in shadow that would be of precisely the same size as had been the space of light when it was placed on the opposite side of the table. Going round the circle west, the same phenomena would be seen, until coming directly south of the lamp, the north pole would again come into light altogether, and the south equally into shadow.