^74 The America,, Oi,,/ot/,'sf. December, 1863 
alone. To those who accepl this explanation of parallel and 
intersecting joints the more concise expression, rectangular 
joints, may seem equally appropriate and descriptive. 
One important consideration affecting the validity of the 
earthquake theory that appears to have been overlooked 
hitherto is that, jusl as a fountain cannot rise above its 
source and the brightness of sunlight cannot exceed that of 
the solar photosphere, so the energy of an earthquake must 
always be greatest at its focus. Most earthquakes, except, 
possibly, in volcanic districts, may be assumed to originate 
in the formation of a fissure bj 7- a tensile or shearing strain, 
or in the slipping of the rocks along a pre-existing fissure, 
that is, in the formation of a fault. The rupture or move- 
ment in which the earth-waves take their rise will occur just 
so soon as the slowly accumulating strain overcomes the co- 
hesion of the rocks or the friction of the opposing walls. In 
other words, the energy of the earthquake is barely sufficient 
at its source to break the rocks. But as the vibrations 
spread from the source this energy is rapidly dissipated. 
Hence, we can scarcely expect that the earthquake, unaided 
by strains due to other causes, will break the rocks, except in 
the immediate vicinity of the focus, unless, indeed, the vibra- 
tions chance to pass into some formation that is much less 
elastic and more easily broken than the one in which they 
originate, as when the shock emerges from the earth and 
enters the walls of buildings. The idea, therefore, that an 
earthquake shock will give rise to a series of fractures paral- 
lel with the initial rupture — a system of joints — is essentially 
untenable, or at least does not admit of general application. 
Torsion and plication, on the other hand, although equal to 
the task of breaking the rocks, are powerless to produce the 
highly regular fractures observed in many formations. The 
arguments thus point very distinctly to a fusion or co-opera- 
tion of the two causes. The parallel joints observed in un- 
consolidated clays may, apparently, be best explained by the 
passage of strong vibrations into the clay from adjacent or 
underl} T ing hard rocks, the influence of torsion being reduced 
to a minimum here. 
When performing Daubree's experiment — breaking strips 
of glass by simple torsion — for the first time I was aston- 
