to the planet's surface. If in a fluid or plastic state, Phobos, for instance, could not possibly exist near the planet's surface. The forces referred to would decide its fate. It may be shown by calculation, however, that if Phobos has the strength of basalt or glass there would remain a considerable coefficient of safety in favour of the satellite's stability; even when the surfaces of planet and satellite were separated by only five miles. We have now to consider some things which we expect will happen before the satellite takes its final plunge into the planet. 180 This diagram (Fig. 11) shows you the satellite travelling above the surface of the planet. The satellite is advancing towards, or away from, the spectator. The planet is supposed to show its solid crust in cross section, which may be a few miles in thickness. Below this is such a hot plastic magma as we have reason to believe underlies much of the solid crust of our own Earth. Now there is an attraction between the satellite and the crust of the planet; the same gravitational attraction which exists between every particle of matter in the universe. Let us consider how this attraction will affect the planet's crust. I have drawn little arrows to show how we may consider the attraction of the satellite pulling the crust of the planet not only upwards, but also pulling it inwards beneath the satellite. I have made these arrows longer where calculation shows the stress is greater. You see that the greatest lifting stress is just beneath the satellite, whereas the greatest stress pulling the crust in under the satellite is at a point which lies out from under the satellite, at a considerable distance. At each side of the satellite there is a point where the stress pulling on the crust is the greatest. Of the two stresses the lifting stress will tend to raise the crust a little; the pulling stress may in certain cases actually tear the crust across; as at A and B. It is possible to calculate the amount of the stress at the point at each side of the satellite where the stress is at its greatest. We must assume the satellite to be a certain size and density; we must also assume the crust of 181 Mars to be of some certain density. To fix our ideas on these points I take the case of the present satellite Phobos. What amount of stress will he exert upon the crust of Mars when he approaches within, say, 40 miles of the planet's surface? We know his size approximately--he is about 36 miles in diameter. We can guess his density to be between four times that of water and eight times that of water. We may assume the density of Mars' surface to be about the same as that of our Earth's surface, that is three times as dense as water. We now find that the greatest stress tending to rend open the surface crust of Mars will be between 4,000 and 8,000 pounds to the square foot according to the density we assign to Phobos.