6 4 8 



LOUIS VESSOT KING 



developed, the rock cylinders were caused to bulge out laterally 

 over the central portion, where the thickness of the nickel-steel 

 jacket was reduced to o . 25 centimeter and o . 33 centimeter, respec- 

 tively, in the two sets of experiments. The rock was thus sub- 

 jected to a continuous succession of breakdowns, so that it was 



possible from these observations to 

 I determine the relation between the end 



and lateral pressures required to keep 

 the rock in movement. 



Considering the central portion of 

 the rock cylinder throughout which the 

 flow takes place, we may reasonably 

 assume, when the bulge is small, that 

 the average pressure-intensity P along 

 the direction of the axis is given by 



W being the load on the steel piston and 

 b its radius. As the bulge becomes 

 sensible, it is necessary to make a cor- 

 rection to allow for the increasing area 

 over which the pressure is distributed. 

 Referring to Fig. 4, we denote by P the 

 average pressure-intensity across a plane 

 at right angles to the axis at the position 

 F IG 4 of maximum bulge where the radius of 



the cross section is b. We denote by p 



the resultant traction per unit area exerted by the nickel-steel jacket 



on the rock specimen in a direction making an angle e with the axis. 



Then, considering the equilibrium of one-half of the rock specimen, 



we may write 



vPM+lP cos € dS=*b 2 P, (18) 



the integral representing the total component of the tractions 

 between the rock specimen and the nickel-steel jacket in a direction 

 parallel to the axis of the cylinder. When an exactly similar jacket 

 is filled with tallow and deformed by the application of a load on 

 the steel pistons in the same way, we may consider the pressure in 



