4 o6 A. A. MICHELSON 



continues at a much slower rate and ultimately comes to rest at 

 a point short of its original position. 



If the stress is too great, the slow yield may increase until 

 rupture occurs. 1 



The following may be considered as a provisional attempt to 

 formulate the behavior of substances under stress by the simplest 

 expressions which have been found to satisfy all the essential 

 requirements. 



The formulae which follow are, in fact, sufficiently general to 

 cover every case thus far examined, including materials of widely 

 different properties, such as lead, tin, copper, aluminum, zinc, 

 iron, steel, quartz, glass, calcite, limestone, slate, marble, wax, 

 pitch, gelatin, and rubber. It may, however, be expected that 

 a more thorough investigation will require modification in the 

 formulae which may be made to fit special cases with greater 

 accuracy. 



The type of strain selected for this investigation is the torsion 

 of cylindrical rods, as this is the only strain in which the form 

 remains unaltered. It is very probable that the laws governing 

 this special type may be made to include other distortions, such 

 as extension, compression, bending, etc. 



Very decided changes may be expected from the effects of 

 temperature and pressure, 2 but these may be taken into account 

 by an appropriate alteration in the value of the "constants" which 

 enter into the formulae. 



1 Rupture may occur in consequence of such slow yielding, or it may be prac- 

 tically instantaneous. In the former case the result is due to separation of the viscous 

 coupling; in the latter, to the snapping of the spring. 



2 A preliminary investigation of the effect of hydrostatic pressure on elasticity 

 and on viscosity was begun several years ago. It was hoped that this would show 

 results in conformity with those which maintain in the body of the earth — whose 

 enormous pressure produces an increase in both rigidity and viscosity sufficient to 

 make the body of the earth (which at its actual temperature under ordinary condi- 

 tions would certainly be in a molten state) as solid as steel. This expectation has 

 been partially realized for a number of materials, metallic and non-metallic, the results, 

 notwithstanding certain anomalies — traceable to the effects of previous history — 

 showing a perceptible increase in rigidity and a very marked increase in viscosity 

 even with the relatively small pressures obtainable in the laboratory. 



