Page 8 - i1052-5173-32-10_FY22_new

A B

C D

Figure 4. Geometric similarity might, but need not, mean similarity of process. (A) Cross-bedded Jurassic aeolian sandstone near Boulder, Utah,
USA. Width of view ~15 m. (B) Cross-beds in fluvial Pleistocene basaltic sands near Mono Lake, California, USA. Width of view 50 cm. Bedding in
both A and B formed in turbulent, high-energy environments where the Reynolds number was likely >10 , and thus grain inertia dominated. (C)
4
Intersecting modal layering in Cretaceous granodiorite near Mack Lake, California, USA. Width of view 60 cm. How these features form is
unknown, but the extremely high viscosity of silicate liquids means that Reynolds numbers were likely 10 or less. Therefore, viscous forces
–6
were dominant, rendering impossible the sorts of grain interactions that produce crossbedding (Glazner, 2014). (D) Intersecting bands of diage-
netic iron oxide in sandstone from the Triassic Chinle Formation, Utah, USA. Oxide layers do not correspond to depositional layering. Width of
view 5 cm. Although these examples are geometrically similar, the erosive turbulence that truncated bedding in sedimentary rocks (A, B) cannot
happen in highly viscous granitic magmas (C) and is irrelevant to the chemical processes that produce diagenetic banding in sandstones (D). The
chemical processes that produced banding in the sandstone (D), however, may be relevant to banding in granodiorite (C).
settling-tube experiments may be mislead- melting basalt, turbulence in granitic mag- petrography, mineralogy, physical chemis-
ing. Their experimental design allowed mas) is inconsistent with results from try, or of any other ancillary discipline.
aggregate grains composed of micron-sized another discipline (e.g., thermodynamics, and
particles to grow to sand size with each cir- fluid mechanics), then other explanations The second suggestion, of deposition from
cuit of the flume. This more accurately repli- should be sought, regardless of the eyeball a liquid magma, is too little developed for
cated natural conditions and showed that test. Just as physical and chemical reason- critical consideration. To constitute a use-
thinly bedded mud can form under currents ing applied to geologic examples must fit ful working hypothesis it should be supple-
capable of transporting these aggregates in the geologic observations, field interpreta- mented by the suggestion of conditions
determining deposition and erosion.
ripples. Their work implies that muddy sedi- tions must satisfy fundamental physical and
ment can be eroded and transported laterally chemical principles. Field-based hypothe- The first quote is a highly restrictive
without showing obvious signs of distur- ses that are consistent with other relevant statement of “the rocks don’t lie” philoso-
bance; thus, a series of layers may contain information survive and are strengthened. phy and comes from Read’s (1948, p. 170)
cryptic lacunae, and any particular layer may Retaining an idea that fails valid tests sim- defense of granitization, a long-discredited
record environmental conditions from else- ply because an alternative model has not yet hypothesis for the origin of granites. The
where in the basin (e.g., Meyers and Sageman, been developed is unproductive. Rather, second, from Gilbert (1906, p. 324), refers
2004; Lazar et al., 2015). failed tests are opportunities to develop new to his own suggestion that banded granites
hypotheses and to look at the rocks from result from erosion and deposition in a
SUMMARY different perspectives. magma chamber. Gilbert knew that ascrib-
The statement “the rocks don’t lie” is Consider the following quotes: ing those features to a familiar process was
true, but their messages may be misinter- I see the granite problem as essentially one not even “a useful working hypothesis”
preted. If a field interpretation (e.g., rhyolite of field geology—it is not primarily one of without more definition and information. In

8 GSA TODAY | October 2022