G. F. Becker — Impact Friction and Faulting. 207 



above referred to is partially elastic, motion and permanent de- 

 formation would always occur on them also were the coefficient 

 of restitution and therefore also the coefficient of friction inde- 

 pendent of velocity. Motion and permanent deformation will 

 actually occur on such planes unless the strain falls short of 

 the elastic limit or, practically unless the strain is extremely 

 slight. The frictional resistance depends upon the geometrical 

 form and absolute dimensions of the minute projections on the 

 rubbing surfaces, the extent to which they interlock, and upon 

 their coefficient of restitution. To a certain extent surfaces 

 rubbing will polish one another and friction will progressively 

 diminish, but the dust resulting from attrition will always 

 suffice to produce new inequalities, while the smoother the sur- 

 faces become the greater must be the influence of the welding 

 which according to Messrs. Thomson and Tait occurs even in 

 J;he mere impact of polished spheres. A tearing of the surface 

 must follow cohesion at any point, so that this cause also must 

 continually interfere with the reduction of the surfaces beyond 

 a certain limit of smoothness. The character of the surface of 

 minimum friction resulting from the action of friction itself is 

 probably open to investigation. It will evidently be closely 

 connected with several of the physical properties of the material. 

 In any inquiry into friction it should not be forgotten that the 

 extent to which the inequalities overlap is in part determined 

 by the elasticity of the whole masses in contact, which must 

 therefore be taken into consideration as well as the coefficient of 

 restitution of the inequalities of the surfaces. 



Geological application. — Suppose :i mass of rock in place is 

 divided by any means into parallel sheets : let the country ad- 

 joining one side of this system of sheets be solid ; and let the 

 whole system be held together by a pressure normal to the 

 fissure planes. If the solid mass adjoining the system of sheets 

 is forced to rise, the conditions discussed at such length in the 

 foregoing pages are present. The energy of the moving mass 

 will be distributed through the system logarithmically. If the 

 s\ 7 stern is infinite, or contains a considerable number of sheets, 

 and if the coefficient of restitution is constant, the distribution 

 of energy will be represented by the simple logarithmic curve 

 however the friction may vary. If the friction is the same on 

 all the contacts, the geometrical result will be represented by 

 the same curve. If the friction remaining constant the system 

 is regarded as including a finite number of members the geo- 

 metrical results will be represented by the difference of two 

 simple logarithmic curves. These are the three fundamental 

 propositions asserted, and so far as I can see rigidly proved in 

 my former paper on faulting, but reached by a wholly different 

 method from that here adopted. New experiments as well as 



