ON THE THEORIES OP ELEVATION AND EARTHQUAKES. 59 
The points at which the formation of fissures would begin will frequently 
not be determinable by calculation, because there might be an imiefinito 
number of points at which (under the assumed condition of the perfect 
homogeneity of the mass) the maximum tension might become equal to the 
cohesive power at the same instant, in which case a fissure would be just as 
likely to begin at one of these points as another. In actual oases, where the 
condition of homogeneity could only be approximately satisfieil, the points 
of incipient disJocation would be jrartly detorniiiied by accidental circum¬ 
stances. The planes of such dislocations at any assigned point would, as 
above explained, be strictly determinate in position. 
The direction of prop^ation is the next point to be considered in the 
formation of figures. When a fissure hud li^un to be formed tiicre would 
be a very rapid modification of the state of tension in its inimediaie neigh¬ 
bourhood, but if tiie fissure should be propagated through portions of the 
mass in which the maximum tensions at every point had become m-arly equal 
to the cohesive power, previously to the coiiutienceaient of any dislocation, 
it is easily shown that the position of the plane of the fissure at every point 
(^)> through which it would be propagated os well as at tlm point (P,) at 
which it would commence, would be at least approximately pcr|ifudiciilar to 
ths direction of maximum tension at (P), at the instant before the formation 
of the &sure begun at P,. Hence the direction of inuxiiunni tension at 
each point being known at the instant previous to the commencement of 
fracture, the position of the fissure subsequently forine<I could be very ap¬ 
proximately determined. The velocity with which the fissure would be pro¬ 
pagated iu such case woulil be very great. 
The intensity of the instantaneous tension nt any point of the mass would 
TO ditt'erent for different diroctiniis through that point, and would vary con- 
tinuougly with the variation of direction, from its maximum to its minimum 
value. If there should bo little difference between the greatest ami least 
tensions, the tendency to form a fuwure would not be much greater in one 
direction than in another, and therefore the direction of a fissure might, in 
sue case, be easily affected by accidental dreutnatanoes, which, on tho con¬ 
trary, would have comparatively littlu effect if the maximum were much 
greater than the inininium tension at each point. In the former ease the 
permanence of direction in tho fissure wouhl bo much smaller than in the 
latter. 
cohesive power of the mass should vary continuously in passing 
thrni '^^1' shouid Tcmain the same in dl directions 
point, the position of a fissure would he the sanio as if the 
be atf « same for every point. Neither would tho position 
TipnilJpiii* L variation of cohesive power along planes per- 
of inaxiiniim tousinu. Thus, if the tissuru were 
i*® affected by the discontimiou^ varia- 
from one horizontal bed to another. If there 
\YGfB ATl ififamxnl ......a, _<• ^ a a ^ « 
dial Iv «r.’ . * I ® .“urface of discontinuity, but would imme- 
^ rosuniD its original diivction, especially if the maximum 
u nuich greater rhaei tlic mininiunu If tJicre were a 
. J- plniy-s and surlace# uf discontinuity without any 
u io,, .'”***'^ diroctmn, they would give a xigxaff character to the fissure, but 
slinniui Its general direction. If, on the eontrar)-. they 
svstpm ‘direction, as in the case for instance of a single 
: or parallel joints, they would produce a deviation in the general 
