Geology Reference
In-Depth Information
surface is a problem. Freeze-thaw action involving soil moisture is another possibility in
some areas.
Leonard (1929) considered the cracking to be related to the formation of (orthogonal) joints, and
though such explanation is germane to those polygonal sets that are developed on structural planes,
it cannot apply to those many occurrences on surfaces due to weathering and erosion. Netoff (1971)
thought that polygonal cracks in sandstone in Colorado (USA) are due to shrinkage produced by the
desiccation of the smectite-bearing bedrock, and this may be valid for the particular occurrence, but
the presence of smectite does not appear to be essential to the formation of polygonal cracking.
Sosman (1916), on the other hand, concluded that such cracking is due to expansion, to extension of
the outer layer of rock. Schulke (1973) identified two types of cracking. He attributed littoral occur-
rences to the weathering of what he termed exhumed joint planes with which are associated veins of
quartz. He did not explain the patterns of fractures, and did not provide evidence for exhumation. In
any case, quartz veins are not necessarily or even commonly associated with polygonal cracking,
either on the coast or inland. The inland type Schulke (1973) related to the occurrence of duricrusts,
their desiccation and cracking as a result of insolation. Robinson and Williams (1989) also favour
development on crusted or case-hardened surfaces. The association with case-hardened surfaces is
real, but, as argued earlier, cracking is probably not related to insolation, and the encrustation may
have developed beneath the land surface, at the weathering front.
11.4.3
Evidence
Some features of polygonal cracking are especially characteristic and diagnostic. For example,
though the cracking is frequently restricted to the outer shell developed on a boulder or platform, it
also occurs on each of several superposed concentric shell layers. The plates are everywhere
in situ
:
nowhere have plates been observed displaced (joggled) or collapsed, as would be the case if
stretching were involved in their development.
Polygonal cracking is not evenly distributed. Although examples can be found on most outcrops,
prolific and well-developed examples are confined to isolated sites. Thus, on northwestern Eyre
Peninsula, examples can be located on most of the granite exposures, and well-developed examples
have been noted on summit platforms on Corrobinnie Hill and Wallala Hill, but it is only on
Tcharkuldu Hill that cracking is abundantly developed and preserved not only on platforms and
large residual boulders but also on structural planes. The most critical lines of evidence, however,
are first, that spalling is widely developed around boulders, and the resultant shells could form the
basis of the eventual plates of polygonal form. Second, weathering produces concentrations of min-
erals at the weathering front. Third, polygonal cracking can develop beneath the land surface (
Fig.
11.26)
and fourth, orthogonal patterns of fractures are associated with fault planes and tectonically
folded surfaces (
Fig. 11.27).
11.4.4
Explanations
Polygonal cracking can be divided into two morphological types, each with a separate origin. There
are first those sets of cracks, essentially orthogonal in pattern, that are developed on plane, gently
arcuate, or warped structural surfaces which are almost certainly planes of dislocation by shearing
(Vidal Romaní, 1991). Second there are those many patterns of polygonal cracking that define
five- or six-sided plates that are preserved on surfaces due to weathering and erosion, and typically
on boulders or platforms.
Orthogonal patterns of cracks are found on essentially plane surfaces in granite, and in sand-
stone and quartzite. Shearing along the plane of dislocation may have produced not only the
orthogonal systems of fractures but also the thin plates on which they are based. That such differ-
ential movement has taken place is suggested first by the character of the surfaces which are fre-
quently coated by recrystallised bedrock, and are polished and striated (slickensides). Such
orthogonal cracking has been observed also on faults exposed in quarry walls. At some sites,
e.g. The Granites, near Mt Magnet, Western Australia, cracking is developed on several layers, and
