1. Hydrocarbon `test' in Devonian chert from Selinsgrove, Pennsylvania.
Laser Raman microprobe analysis suggests that the rather well-ordered,
highly reflective hydrocarbon forming the 20-micron-thick wall of
this microfossil is graphitic, a surprise in a suite
of rocks showing few signs of metamorphism
(Mary I. Garland, pers.commun., 2003). Small angular carbonate
rhombs and tiny, bright pyrite grains are visible in the finely
granular, silica- and dolomite-rich groundmass of this black to
brownish-black chert.
Sample 12, 160X magnification, long-axis field-of-view 0.7 mm,
reflected, plane-polarized light.
Bioclasts and Pyrite in Lower Devonian Chert from Pennsylvania
Samples 12 and 40 were collected from two of seven
key localities in Snyder county, central Pennsylvania.
They represent E.N.E.-trending outcrops of cherty strata which
are exposed at numerous localities in a 900 km2
study area. The samples are part of a large provenance study (approximately 100
samples for petrography, 120 for multi-element geochemical analyses)
designed to test the utility of numerous tools for matching prehistoric
lithic artefacts to outcrops, including known quarry sites from
which artefacts, hammerstones and chert debitage have been recovered.
The laboratory section of the study focused to a large extent on
transmitted and reflected-light petrography, and whole-rock and trace-element
geochemistry, with the flexibility to test additional mineralogical
techniques such as electron microscope / microprobe analysis and laser
Raman microprobe spectroscopy. This document highlights the potential
for a classic palaeontological approach. Bioclasts, or their remnants, may
be identified as; 1) macrofossils such as
shell fragments: 2) spores or plankton, commonly 100-300
microns in diameter,
replaced by hydrocarbons:
3) spicule-like forms, siliceous with minute
inclusions of other minerals: 4) pyritized debris (framboids are common,
possibly nucleated on algal or other biogenic detritus).
Left:
2. The `spicules' and pyrite in chert from Walnut Acres,
sample 40.
80X magnification, long-axis field-of-view 1.4 mm,
transmitted, plane-polarized light.
3. Right:
Hydrocarbon test, rhombohedral carbonate and pyrite in sample 40.
160X magnification, long-axis field-of-view 0.7 mm,
reflected, plane-polarized light.
4. Left:
Hydrocarbon tests in chert sample "2-Float".
80X magnification, long-axis field-of-view 1.4 mm,
transmitted, plane-polarized light.
5. Right:
Framboidal pyrite, ovoid instead of the more-abundant spheroidal
form, sample "2-Float".
80X magnification, long-axis field-of-view 1.4 mm,
reflected, plane-polarized light.
Petrography
A non-technical summary of the bioclasts follows, modified
after Wilson (2002).
Macrofossils are extremely rare in a suite of 52 hand specimens.
In thin section, however,
fossil debris is plentiful. Roughly speaking, 10 thin sections in
50 contain macrofossils, most notably in the form of
brachiopod shell fragments, 30 display small rounded
tests of microfossils, and 35 have elongate siliceous entities
described, on simple morphological grounds, as spicules,
although they may possibly have alternative identities,
such as brachiopod spines. The
rounded tests are generally composed of hydrocarbons, while the
elongate spicules, which are most often noted in relatively fine
(micritic), ferruginous matrices, are polycrystalline aggregates of
granular silica with fine carbonate, hydrocarbon, oxides and other
minerals. The fine spatial resolution and sensitivity to chemical
bonding makes laser Raman spectroscopy the logical tool to examine these
features in more detail. The tests are most probably spores from
terrestrial plants, though some of the entities
in these rocks have been identified
as acritarchs and the remains of chitinozoans (B.E. Eley and
P.H. von Bitter, Royal Ontario Museum, pers.commun., 2003).
The fossil content does not immediately identify the cherts as
primary siliceous marine sediments: they lack the depositional
laminations noted in some Jurassic cherts and they
may well be largely replacements of host carbonate rocks, generated
during diagenesis and incorporating bioclasts, including spores,
plankton and spicules, that sank into the seafloor carbonate ooze
as the limestones and dolostones were accumulating. The fossils,
regardless of their exact nature, are potentially useful clues for
petrographic provenance.
It is presumed that the observed bioclasts are but a fraction of the
material available, a selection fortuitously exposed on the surfaces
of polished thin sections. Although some of the samples display
better than 90 volume percent silica replacement, 19 out of 50
key samples contain an estimated 51-96 volume percent carbonates,
such that standard recovery techniques used on small bulk samples
should provide ample microfossil suites.
Chert-bearing units are quite common in the Devonian of the region
(Harper, 1999; see also Laird, 1932; Parkins, 1977). Provenance
studies involving aspects of chert (macroscopic properties,
geology, mineralogy, palaeontology, geochemistry) are nothing
new, but the fine grain-size, apparent homogeneity and even the
physical toughness and chemical resilience of the material all conspire to
keep `chert sourcing' a definite challenge (Luedtke, 1992).
Some Biostratigraphic Lore
Richardson (1969) provided an overview of Devonian biostratigraphy
in Europe and north America. In the lower Devonian, the focus of
our present interest, the European Gedinnian, Siegenian and Emsian
stages are roughly the equivalent of the Helderberg, Deer Park and
the lower half of the Onesquethaw stages in the eastern USA, while
the middle Devonian Eifelian and Givetian are roughly equivalent to
the upper Onesquethaw and the Cazenovia and Tioughnioga stages
respectively. Gedinnian spores tend to be very small: a study of
500 specimens from the Welsh Borderland revealed a range of 8 to 62
microns with a mode of just 17.5 microns, while forms >80 µm
are not known. Larger forms appear in the Siegenian and Emsian.
Some megaspores (>200 µm) occur in the Siegenian, and in the
middle Devonian (Eifelian, Givetian) yet larger forms occur, often
>200 µm, quite often 300-500 µm.
Parkin's (1977) study found that
the distribution of acritarchs varies greatly in the Onondaga, even in
a single sample. All the spores are miospores (<200 µm and
often 25-50 µm in diameter). The most abundant were
Apiculiretusispora arenorugosa and Calamospora
tenuis. In paleoecological terms, Parkins determined that the
Bois Blanc sediments were deposited in a quiet, subtidal marine
setting, while the Edgecliffe and Clarence members of the Onondaga
were laid down respectively in reef and lagoonal settings.
Of particular note is the Shriver chert member of the Old Port
Formation, a lower Devonian unit lying between Ridgeley sandstone
and Mandata shale, and laterally correlative to Licking Creek
limestone. The Old Port is sometimes mapped alongside the Onondaga
chert (Bailey and Katz, 2002). The middle Devonian Onondaga
Formation outcrops north of the study area, and has been mapped
across parts of southwestern Ontario and western New York state.
The Onondaga chert may be distinguished from chert of the lower
Devonian Bois Blanc Formation based on petrology and microfossil
content, including a `preservation ratio' derived from palynomorphs
in the cherts (Parkins, 1977). The latter is the number of
well-preserved palynomorphs per 1,000 poorly-preserved
palynomorphs, organic fragments and sulphide spheres in a given
sample. Spherical sulphide grains up to 30 µm in diameter
often occur in the cell walls of palynomorphs, and are considered
to be the `remains of organic walled microfossils'. Parkins'
detailed study identified many species of microfossils (acritarchs
and spores) in both the Bois Blanc and Onondaga cherts.
Wall (2000) examined artefacts at the 36Un82 archaeological site in
Allenwood, Pennsylvania, which displays evidence of Middle Archaic
through Late Woodland occupations. Late Archaic sites in the
Susquehanna (West Branch) river valley show a wide range of lithic
resources, both primary (Shriver chert) and cobbles (Onondaga
chert). The most common artefact material is a grey to black,
medium to coarse chert (probably the local Shriver chert).
This particular study did not make extensive use
of micropalaeontological procedures, as it was shown that
thermal alteration and deformation have obliterated much of the
diagnostic value of the fossils. However, this very finding does help
to `fingerprint' the local cherts (see the "sequel", below).
BAILEY,DN and KATZ,G (2002) What's black and gray and found all
over? Chert in the Middle Creek and Penns Creek watersheds in
central Pennsylvania. Eastern States Archaeological Federation
Conference, manuscript, 11pp.
LAIRD,HC (1932) The Nature and Origin of Chert in the Lockport
and Onondaga Formations of Ontario. PhD Thesis, University of
Toronto, 134pp. plus 18 plates.
LUEDTKE,BE (1992) An Archaeologist's Guide to Chert and Flint.
Archaeological Research Tools 7, Institute of Archaeology,
University of California, Los Angeles, 172pp.
PARKINS,WG (1977) Onondaga Chert: Geological and Palynological
Studies as Applied to Archaeology. MSc Thesis, Brock University,
104pp.
RICHARDSON,JB (1969) Devonian spores. In `Aspects of Palynology'
(Tschudy,RH and Scott,RA, editors), Wiley-Interscience, New York,
510pp., 193-222.
WALL,RD (2000) A buried Lamoka occupation in stratified
contexts, West Branch valley of the Susquehanna River,
Pennsylvania. Pennsylvania Archaeologist 70 no.1, 1-44.
WILSON,GC (2002) Geochemistry of chert samples from Snyder and
Union counties, Central Pennsylvania. TGSL Report 2002-09 for
A.D. Marble and Company, Inc., Conshohocken, PA, 10+30pp.
Graham Wilson, posted 04 December 2002, updated 02 July 2003, html updated 09 January 2006
For further background information, visit the original
"Rock of the Month" Chert
![fossil test [124 kb]](chert12a.jpg)
![chert [102 kb]](chert40a.jpg)
![chert [38 kb]](chert40b.jpg)
![chert [84 kb]](chert02a.jpg)
![chert [54 kb]](chert02b.jpg)