254 Joty—On the Origin of the Canals of Mars. 
The gradual slope of the curve near the point of maximum stress, rather 
suggests 60° as marking out the circle of effective stress if there is relative motion 
of planet and satellite. The anchoring effects on the crust of inertia and viscosity 
(supposing the planet’s crust to be underlaid by a viscous magma) will be less effec- 
tive at this distance. After the 60° point the stress increases but little. In this 
case action at the radius of 110 miles will require the satellite’s centre to be placed 
at a vertical distance of 63°5 miles. The stress varies inversely as the vertical dis- 
tance of the satellite; and hence it will be found that the horizontal tearing stress 
per square centimetre will have diminished to 2062 grammes, the density of 
Phobos being 4. A satellite having a diameter double that of Phobos (a diameter 
still much less than that of the larger asteroids) will at the distance of 63 miles 
exert a stress of from 15 to 30 tons per square foot, according to what assumption 
we make as to the density of the body, tending to rend the crust or surface 
material of the planet in a circle 220 miles in diameter. 
We have next to consider the probable effects of such a stress when the satellite 
is moving relatively to the surface of the planet, but not at such speeds as to 
seriously interfere with the establishment of static stresses. 
The satellite must be considered as carrying the circle of stress along with it 
over the planet’s crust. Tangents drawn to this circle, parallel to the direction of 
the line of motion, define two lines 220 miles apart. Along these lines the stress 
is advancing, increasing from zero at points very remote from the satellite up to 
the maximum amount, and again dying away. Directly in the line of advance of 
the satellite itself, the rate of change of stress is much greater. An action goes on, 
too, along the tangents which does not go on in the central line: a ¢earing action 
if a crack is once started. ‘Thus, suppose such cracks started when the satellite 
is passing over a weak spot in the planet’s crust. The advancing extremities of 
the cracks are now continually the site of concentrated forces, coming in for the 
stress acting on the already gaping walls of the crack. For the same reasons that 
we go to the edge of a piece of cloth in order to tear it, and do not waste our exer- 
tions upon the middle of it, the stress upon the edge of the crust will tear it, 
although possibly not great enough to be effective when applied to a sound part of 
the crust. 
The tearing action does not act in the same manner across the direct line of 
advance of the satellite; a line of weakness at right angles to the line of advance 
may open, perhaps, owing to the stress, but the effect of this is only to relieve the 
parts further on in the line of motion from intrinsic stresses and from those due 
to the satellite. 
Hence the motion of the satellite is accompanied by two parallel lines of 
maximum stress carried over the planet’s surface with each return of the satellite 
over the same ground. How long this repetition of the stress will continue is a 
