80 Prof. C. Gr. Knott on Reflexion, and Refraction of 



purely elastic vibrations. These, generally speaking, will be 

 transmitted with considerable speed, comparable to that of 

 sound in steel wires. Such high speeds have indeed been 

 observed, their existence depending upon a small compressi- 

 bility (or high rigidity) combined with a comparatively small 

 density. The destructive effects of earthquakes are, however, 

 due to the propagation of guasi-eiastic disturbances. In them 

 the material is distinctly strained beyond the limits of elasticity; 

 or, at all events, so strained as to bring about conditions in 

 which other strain- coefficients than the usual ones of rigidity 

 and compressibility play the important part. It is quite to 

 be expected that these quasi-elastic disturbances should travel 

 much more slowly than the purely elastic ones. The investi- 

 gation given above into the effect upon the type of elastic waves 

 as they suffer reflexion at the boundary of two isotropic elastic 

 media suggests the existence of analogous effects in the propa- 

 gation of all seismic disturbances. The aeolotropy and discon- 

 tinuity of the earth's crust will transform a disturbance of an 

 originally simple type into one or more of excessive complexity. 

 Furthermore, wherever a ^wcm'-elastic disturbance suffers 

 transformation at some region of discontinuity, it will give 

 rise to a new set of elastic disturbances. And again, as the 

 quasi-elastic disturbances lose energy per unit volume, partly 

 because of radiation, partly because of dissipation, they will 

 gradually lose their <^/as«-character, and become of a purely 

 elastic nature. It is quite conceivable, then, that under certain 

 circumstances the speed of a disturbance might undergo 

 strange variations, appearing even to be accelerated as its 

 intensity diminished. Such a phenomenon was observed by 

 Lieut.-Col. Abbot at Flood Rock explosion in 1885. Of 

 course a peculiar change of this kind might easily be due to 

 the different elastic properties of successive portions of rock 

 travelled through. It is quite possible, however, that the 

 other explanation is the true one. It is known that a very 

 intense sound travels faster in air than one of less intensity; 

 and the same will be true of vibrations in elastic solids. But 

 there must be a superior limit to the intensity, for intensities 

 above which this relation will cease to hold. Viscosity, 

 friction, and the little understood effects of permanent strain 

 will make themselves more and more strongly felt as the 

 strains increase beyond the limits of elasticity. I understand, 

 indeed., that in the case of cannon-reports the sound has been 

 observed to travel somewhat less rapidly during its early than 

 during its after stages. Here the very large initial aerial 

 distui bam_es bring in conditions, either thermodynamic or 

 elastic, under which the ordinary theory fails even in ap- 



