Parallel and Intersecting Joints. — Crosby. 369 
have never doubted that it is a valid explanation admitting of 
a wide application. 
As a rule, where the rocks are free, or nearly so, from shrink- 
age cracks, or joints due to contraction, and have not been ex- 
posed to severe crushing or shearing strains, they are traversed 
by two similar systems of nearly vertical joints crossing at 
large angles, and with, usually, no decisive indications that 
the two systems are not of the same age. Now Daubree's ex- 
periments show that while simple plication yields a single sys 
tern of longitudinal or strike joints, and the development of 
a transverse system would require a shifting of the axis of 
bending, torsion gives simultaneously two systems intersect- 
ing at large angles, but both oblique to the axis of torsion. 
These appear to exhibit all the normal characteristics of ac- 
tual joints, including irregular examples and the not infrequent 
instances where one joint ends squarely against another with- 
out cutting it. 
According to the earthquake theory, on the other hand, the 
dilferent systems of joints dividing any given mass of rock were 
developed in succession, the normal direction of each system 
being parallel with the earth-waves producing them. It has 
been shown,* however, that this will be true of the second s} r s- 
tem only in case the direction of the second series of earth- 
waves is approximately at right angles to the first series, or 
they have a high velocity of shock. For, while the iirst earth- 
quake of sufficient energy to break the rocks will develop 
a system of joints at right angles to its path, the sheeting 
of the rocks thus resulting will cause the second system 
to cut the first at large angles, regardless of the direction 
of the vibrations, in obedience to the principle that oblique 
strains would, if not developed instantaneously, be relieved 
by slipping along the joints of the first system and the forma- 
tion of a series normal to them. 
The fact that an occasional joi nt term i na t es against another 
without intersection is shown by experiment to be consistent 
with the torsion theory; but it is also readily explained by the 
earthquake theory, since each sheel determined by the iirst 
system of joints would, when subsequently exposed toendwise 
or oblique shocks, naturally break somewhat independently of 
*Proc. Boston Soc. Nat. Hist., xxn, 80; xxm, -J44. 
