Will such a stress actually tear open the crust? We are not able to answer this question with any certainty. Much will depend upon the nature and condition of the crust. Thus, suppose that we are here (Fig. 12) looking down upon the satellite which is moving along slowly relatively to Mars' surface, in the direction of the arrow. The satellite has just passed over a weak and cracked part of the planet's crust. Here the stress has been sufficient to start two cracks. Now you know how easy it is to tear a piece of cloth when you go to the edge of it in order to make a beginning. Here the stress from the satellite has got to the edge of the crust. It is greatly concentrated just at the extremities of the cracks. It will, unler such circumstances probably carry on the 182 tear. If it does not do so this time, remember the satellite will some hours later be coming over the same place again, and then again for, at least, many hundreds of times. Then also we are not limited to the assumption that the {Fig. 12} satellite is as small as Phobos. Suppose we consider the case of a satellite approaching Mars which has a diameter double that of Phobos; a diameter still much less than that of the larger class of asteroids. Even at the distance 183 of 65 miles the stress will now amount to as much as from 15 to 30 tons per square foot. It is almost certain that such a stress repeated a comparatively few times over the same parts of the planet's surface would so rend the crust as to set up lines along which plutonic action would find a vent. That is, we might expect along these lines all the phenomena of upheaval and volcanic eruption which give rise to surface elevations. The probable effect of a satellite of this dimension travelling slowly relatively to the surface of Mars is, then, to leave a very conspicuous memorial of his presence behind him. You see from the diagram that this memorial will consist o: two parallel lines of disturbance. The linear character of the gravitational effects of the satellite is due entirely to the motion of the satellite relatively to the surface of the planet. If the satellite stood still above the surface the gravitational stress in the crust would, of course, be exerted radially outwards from the centre of the satellite. It would attain at the central point beneath the satellite its maximum vertical effect, and at some radial distance measured outwards from this point, which distance we can calculate, its maximum horizontal tearing effect. When the satellite moves relatively to the planet's crust, the horizontal tearing force acts differently according to whether it is directed in the line of motion or at right angles to this line. In the direction of motion we see that the satellite 184