18 BRECCIA AT SERPENT MOUND CRATER Vol.

110

An Anomalous Breccia Associated with the Serpent Mound Impact Crater, Southern Ohio

Keith A. Milam1 and Adam Hester, Department of Geological Sciences, and Peter Malinski, Department of Chemistry and Biochemistry,
all of Ohio University, Athens, OH

Abstract. An anomalous carbonate breccia in the Serpent Mound impact crater in southwestern Ohio was examined and possible
depositional/emplacement mechanisms were evaluated in an effort to determine its origin. This breccia was likely formed by
sedimentary deposition and subsequent weathering during the Middle-Late Silurian prior to the Serpent Mound impact event.
This origin is supported by the lateral extent of the breccia, the elevation range over which it is exposed, its spatial association
with Middle-Upper Silurian strata, a mineral assemblage limited to dolomite, compositional homogeneity, and its similarity to
Middle-Upper Silurian geologic units. Field observations, mineralogy, and geochemical analyses do not support emplacement
by fault comminution, gravitational collapse of crater slopes, or ballistic/resurge deposition of ejecta.

OHIO J SCI 110 (2): 18-30, 2011

INTRODUCTION the marl associated with the structure. Others (Stout 1940, 1941;
The Serpent Mound Impact Crater Schmidt and others 1961) offered a contrasting view by attributing
The Serpent Mound impact structure (which shares its name this unit to extensive weathering of the upper few meters of the
with the well-known Paleo-Indian effigy; Willoughby 1919) is a Peebles Dolomite. Stout (1940) and Stout (1941) observed that
seven to eight km diameter (Reidel 1975; Reidel and others 1982) the breccia was confined to Adams, Highland, and Pike Counties
circular feature situated in southern Ohio at the junction of Adams, in southern Ohio. Stout (1941) further observed that it occurs in
Highland, and Pike Counties (Fig. 1). The structure lies within areas where strata between the Peebles Dolomite and Ohio Shale
the Bluegrass Section of the Interior Low Plateau physiographic were absent. An impact origin for the breccia was suggested by
province along the western edge of the Appalachian Plateau,
known as the Allegheny Escarpment, and to the east of the Central
Lowlands of western Ohio (Brockman 1998).
Although area bedrock consists of normally flat-lying
Ordovician-Mississippian sedimentary rocks (Fig. 2), the Serpent
Mound structure is a zone in which these strata have been disturbed.
This was first described by Locke (1838) and later interpreted
by Bucher (1933) and Reidel (1975) to have resulted from
“cryptovolcanism” or a “cryptoexplosion”. The absence of surface
or subsurface igneous rocks and the lack of a volcanic root at depth
discredit this notion (Bull and others 1967).
The Serpent Mound impact structure is the only known impact
crater in the state of Ohio (Fig. 1). Its impact origin was confirmed
with the discovery of shatter cones (Dietz 1960), shocked quartz
(Carlton and others 1998; Koeberl and others 1998), and coesite
(Cohen and others 1961) and is supported by the enrichment of
siderophile elements at the center of the structure (Carlton and
others 1998; Koeberl and others 1998). Serpent Mound represents
the eroded remnant of a complex impact crater. Initial mapping
efforts identified a centrally-uplifted area of Ordovician rocks
surrounded by a circular graben of downward-displaced Devonian
to Mississippian strata separated by a so-called “transition zone”
(Fig. 1; Reidel 1975; Reidel and others 1982). The transition zone
was characterized by Reidel (1975) and Reidel and others (1982)
as an area of folded and faulted (primarily Silurian) strata not
significantly displaced from normal positions. Carbonate breccia
in this zone is the focus of this study.

An Anomalous Breccia in the Crater

The first descriptions of this anomalous breccia were made circa
1916 by August Foerste and Raymond Lamborn who identified
this breccia as “marl” dividing the Silurian Bisher and Lilley
Members of the (then) West Union Formation (Schumacher
Figure 1. Oblique aerial views (looking to the NNE) towards the Serpent Mound
2002c). Their field investigations identified at least 17 locations of impact structure (Google Earth/USDA Farm Service Agency). The inset box in
(a) shows the location of the structure at the junction of Adams, Highland, and
Address correspondence to Keith A. Milam, 213 Clippinger Laboatories,
1 Pike Counties, while (b) shows the approximate outlines (dashed lines) of the
Athens, OH 45701. Email: [email protected] major features within the crater as well as the breccia type location (star symbol).
Ohio Journal of Science K.A. MILAM AND OTHERS 19

Dietz (1960) who reported “a quarry which revealed a large mass others 1961). As such, breccia should be found spatially associated
of what may be explosion breccia.” with Middle Silurian units and its mineralogy and bulk composition
The breccia was first observed by the authors at a location, should correlate most closely with Middle Silurian geologic units.
here designated the type location(Figs. 1, 3a, 4a) directly west of Breccias produced by fault movements are common to impact
the central peak of the impact crater (39° 2.036”N, 83° 25.373” craters (French 1998). The type location of the breccia is in the
W). Later the search was expanded to include other parts of the transition zone (Fig. 1) in an area where Reidel (1975) mapped
impact crater (see Figs. 3-4 and below) and outside the limits of the numerous converging faults (Fig. 3). Faults cut across sedimentary
disturbance as defined by Reidel (1975). The breccia ranges from strata ranging in age from the Lower Silurian Rochester (Estill)
matrix- to clast-supported and (apparently) polymict to monomict Shale to the Upper Devonian Ohio Shale (Fig. 2; Reidel 1975).
within and among exposures (Fig. 5). Clasts range from angular to Fault breccia, if present, should be exposed only in fault zones and
sub-angular coarse sand to boulders and typically consist of massive its composition should represent a mixture of Lower Silurian to
or cross-bedded dolomicrite and algal mats. Most commonly the Upper Devonian strata.
breccia is massively bedded, although poorly-developed bedding Central peaks in complex craters are comprised of densely
can be discerned at a few locales. At the type location (Fig. 3a) fractured and faulted crater floor strata that have been uplifted well
and at two other sites in the northern part of crater (Fig. 3, b and above their normal stratigraphic positions (Grieve and Thierriault
c), breccia exposures appear to display an overall fining upward 2004). Following the rise of a central peak, this weakened material
sequence. Physical properties (hardness and reactivity to acid) is particularly susceptible to gravitational collapse. The Serpent
indicated a carbonate composition for the breccia. Our initial Mound central peak is comprised of Upper Ordovician –Middle
observations and those of Stout (1940) and Stout (1941) support Silurian carbonates and shales (Reidel 1975). If the anomalous
the notion that this breccia is not common to the local stratigraphy breccia is the product of gravitational collapse, then it should
and may be confined to the Serpent Mound impact crater. Therefore be only locally exposed along the flanks of the central peak and
deposition/emplacement of the breccia may be related to the crater represent a compositional mixture of Upper Ordovician-Middle
landform and/or the impact event. Silurian target rocks (Fig. 2). However, if the collapse was more
We test the four most plausible mechanisms for emplacement or recent, then the breccia should represent a mixture of geologic
deposition of the anomalous breccia: (1) pre-impact sedimentary units to the nearby eastern drainage divide for the type location,
deposition (and subsequent weathering), (2) comminution during consisting of strata from the Lower Silurian Brassfield Limestone
fault movement, (3) gravitational collapse of the central peak, or (4) to the dolostones of the Middle Silurian (Fig. 2).
fallback/resurge of impact ejecta. As noted by earlier workers, the Breccias (lithic or melt) are commonly found in, near, or
breccia may represent either the Middle Silurian Lilley or Bisher associated with terrestrial impact craters (French 1998), deposited
Formations (Schumacher 2002c) or simply a highly-weathered following fragmentation and ejection of material (ejecta) from
horizon of the Middle Silurian Peebles Dolomite (Schmidt and an impact event. Most are deposited immediately adjacent to or
away from the crater rim, whereas some is entrained in the impact
plume and settles on the crater floor. In shallow marine settings,
where an impactor excavates and ejects both seawater and the
ocean floor, collapse of the transient cavity results in resurge of the
excavated water column. This resurge also entrains and deposits
ejecta on the crater floor (e.g. Ormö and Lindström 2000; Dypvik
and Jansa 2003).
If the anomalous breccia represents impact ejecta, then it would
represent a mixture of strata ejected during the Serpent Mound
impact event. The typical excavation depth for impact craters is
approximately one-tenth of the transient crater diameter (Melosh
1989), a parameter not directly measureable in complex craters
whose rims have characteristically collapsed. However, impact
modeling has shown that the transient crater diameter (Dt) is
approximately 50-60 percent of the final crater diameter (Df)
(Melosh 1989). With a maximum diameter of eight km (Reidel
1975), the Dt for Serpent Mound would have ranged from 4.0-4.8
km. Therefore the estimated excavation depth ranged from 400-480
m. A Late Mississippian impact would have excavated to a depth
that exceeds the total thickness of the Upper Ordovician-Lower
Mississippian carbonate-rich sedimentary strata still exposed near
the impact site. A Pennsylvanian-Permian impact (a possibility
offered by Watts 2004) would have excavated additional siliciclastic
sedimentary strata no longer preserved at the crater.

METHODS
Figure 2. Simplified stratigraphic column of rocks commonly exposed in the This study seeks to determine which of the mechanisms were
Serpent Mound area using thickness and elevation data from Rexroad and others responsible for the deposition or emplacement of the breccia.
(1965), Swinford (1985), Swinford (1991), Schumacher and Reidel (1997), Shrake
and others (2007), Schumacher and Reidel (2002a), Schumacher (2002b), and
Field observations, mineralogical, and geochemical analyses were
Baranoski and others (2003). employed to assess each of the four geologic processes.
20 BRECCIA AT SERPENT MOUND CRATER Vol. 110

In an effort to determine the occurrence and lateral distribution Planetary Geology Laboratory at Ohio University. All samples were
of the anomalous breccia in the Serpent Mound impact structure, analyzed using X-ray diffraction (XRD) to identify major minerals
we systematically revisited purported marl and breccia sites as present. XRD spectra were collected over one hour/sample at Ohio
first identified by Foerste and Lamborn (Schumacher 2002c) University using a Rigaku Geigerflex X-ray diffractometer (40 kV,
and Reidel (1975). Exposures were observed and described and 30 mA). XRD spectra of breccia samples were compared to those
samples were collected and compared to the type location to of known mineral phases and geologic formations in the vicinity
confirm the presence of the anomalous breccia elsewhere. The to determine whether the breccia consists of single or multiple
latitude, longitude, and elevation of each confirmed site were geologic source units. A representative subset (24 samples, Table
measured with a Garmin Colorado 400t GPS receiver [~5 m 1) was analyzed using X-ray fluorescence (XRF) to measure bulk
and ~3 m horizontal (WGS 84) and vertical (NAD27) accuracy chemistries. Samples were fused at 1050°C for two hours into disks
respectively,] and results were plotted on a geologic map of the and were analyzed by the heavy absorber fusion technique of Norrish
structure (Fig. 3). Our team also systematically searched both and Hutton (1969). XRF analyses were used to compare the major
inside and outside of the crater (outer limits defined by Reidel element oxide chemistry of five of the breccia samples (Table 1) to
1975) for additional breccia exposures. those of known undisturbed geologic units collected in this study
To assess which of the locally-exposed geologic units might and to those (Table 2) reported by Rogers (1936), Lamborn and
comprise the anomalous breccia, 22 samples of undisturbed geologic others (1938), Stout (1940), and Stout (1941). Comparisons were
units from outside of the crater and 13 samples of the anomalous made using geochemical variation diagrams and by performing a
breccia were collected, described, and processed for geochemical Spearman’s rank correlation coefficient analysis for SiO2, Al2O3,
analyses (Table 1). Between 10-20 g of each sample (including Fe2O3(+FeO), MnO, MgO, and CaO. These major oxides were
both clasts and matrix) was ground into powder (clay-size) using used because values were variable and all measured results were
an agate mortar and pestle, mixed thoroughly, and separated into above the average detectibility limit of the XRF analyses (<0.01
three aliquots. The first two aliquots (4-8 g each) were used for wt. percentage). Thresholds for the correlation coefficient analysis
geochemical analyses, while the remaining aliquot was stored at the were 90, 95, 99, 99.5 and 100 percent.

Figure 3. Geologic map of the Serpent Mound impact structure (after Reidel 1975), highlighting the type location (a) and two other sites where fining-upward sequences
are exposed (b and c).
Ohio Journal of Science K.A. MILAM AND OTHERS 21

RESULTS m above sea level. It is commonly situated adjacent to or near the

A total of 22 exposures of the breccia were located within the Middle Silurian Peebles Dolomite or the Middle Silurian Lilley
Serpent Mound impact structure (Fig. 3), all within the deformed and Bisher Formations. In a few exposures in the northern ring
area as defined by Reidel (1975) and only locally exposed as noted graben, the Ohio Shale seems to be stratigraphically superposed
by Stout (1940) and Stout (1941). Twelve of the sites are situated upon the breccia.
within the transition zone, while the remaining sites are exposed XRD spectra (2θ = 20-80°) of samples of undisturbed geologic
in the ring graben in the northern part of the crater. Near the type units outside of the crater demonstrate characteristic mineralogical
location, the breccia is exposed over an area of over 20,000 m2, but trends in the local stratigraphy (Table 3). Upper Ordovician-Lower
may be exposed over an area as large as 1.4 million m2. Measured Silurian rocks consist primarily of calcite and quartz, with lesser
sections reveal that the anomalous breccia ranges from 3-21 m in amounts of dolomite. Middle Silurian units are dominated by
minimum apparent thickness and ranges in elevation from 220-246 dolomite, whereas the Bisher and Lilley Formations have lesser
amounts of quartz. The Lilley Formation also contains calcite in
small amounts. Upper Devonian-Lower Mississippian units are all
dominated by quartz.
Breccia samples consist of dolomite. Examination of XRD
spectra (2θ = 20-80°) of the 13 breccia samples reveals primary
peaks at 2θ ≈ 31°, 41.1°, 51.1°, and 50.6°, which correspond to
dolomite. Sample SMB-9 is the only breccia sample that appears
to contain minor amounts of quartz. Representative examples of
breccia spectra are shown in Fig. 6.
XRF analyses show similar trends (Table 4). Upper Ordovician-
Lower Silurian units that have a significant quartz component
have characteristically higher weight percentage SiO2 than other
units dominated by carbonate minerals that have higher weight

Figure 5. Field images of the anomalous breccia showing an apparent range of

clast variety and clast/matrix ratios: (a) monomict, matrix-supported breccia, (b)
Figure 4. Breccia exposures at the type location (a) and in the northern part (b polymict, matrix-supported breccia, and (c) polymict, clast-supported breccia. All
and c) of the crater. Locations are labeled on Figure 3. three samples are from the type location shown in Figs. 3 and 4a.
22 BRECCIA AT SERPENT MOUND CRATER Vol. 110

percentages CaO and MgO (Table 4). Breccia samples also have impact structure suggest that the geologic process responsible for
very low weight percentages of SiO2, but have relatively high values its deposition or emplacement was not restricted to a small area,
for CaO and MgO (Table 4). Representative geochemical variation but was rather associated with the crater and may have extended
diagrams are shown in Fig. 7 for visual comparison of the bulk beyond its confines. The emplacement/depositional mechanism
compositions of breccia samples to known geologic units. Fig. either operated during the Middle-Upper Silurian or only affected
7 shows that weight percent oxides for the carbonate breccia are Middle-Upper Silurian carbonates. This is evidenced by the spatial
most closely associated with Silurian dolomitic formations, most association of the breccia with the Peebles Dolomite and the Lilley
specifically with the Middle Silurian Peebles and Upper Silurian Formation, dolomitic composition of the breccia, and its close
Greenfield Dolomites. Statistical analyses (Table 5) indicate that compositional correlation with Middle-Upper Silurian carbonates.
the anomalous breccia is 95-100 percent correlated with the Peebles The spatial, mineralogical, and compositional association of
Dolomite, the Greenfield Dolomite, Tymochtee Formation, Lilley the anomalous breccia with Middle-Upper Silurian carbonates is
Formation, and Bisher Formation. The Brassfield Limestone consistent with marine sediment deposition during the Middle-
and Drakes and Bull Fork Formations were correlated <90 to Late Silurian. The range of elevations over which the breccia is
99.5 percent, 90 percent, and <90 percent respectively. All other exposed (220-240 m above sea level) coincides with elevations
formations were correlated to the breccia at <90 percent. of Middle Silurian carbonates (Table 2), when the eastwardly
regional dip (Swinford 1985) is considered. Stratigraphic
DISCUSSION thickness, mineralogical, and compositional similarities between
The lateral extent and morphology of the breccia across the breccia and the Middle-Upper Silurian Tymochtee Formation
the transition zone and the ring graben of the Serpent Mound and Greenfield and Peebles Dolomites suggest that the breccia

Table 1

Samples collected for mineralogical and geochemical analyses. Unit descriptions may be found in Swinford (1985) and Shrake and others (2007).

Geologic Sample Approximate Sample XRD XRF Geologic Sample Approximate Sample XRD XRF
Unit No. Location/Elevation Unit No. Location/Elevation

Berea BS-1 39˚1'14.00''N 83˚16'41.37''W/357m x Brassfield BL-1 38˚55’58.27’’N 83˚27’53.81’’W/197m x x

Sandstone Formation

Bedford BD-1 39˚1’15.20’’N 83˚16’36.85’’W/375m x Drakes DK-1 38˚53’37.94’’N 83˚27’11.38’’W/201m x x

Formation
Ohio Shale OS-1 39˚1’7.90’’N 83˚17’24.62’’W/293m x x
BullFork BF-1 38˚56’1.36’’N 83˚28’41.75’’W/199m x x
OS-3 39˚1’7.90’’N 83˚17’24.62’’W/293m x x Formation
BF-1 38˚56’1.36’’N 83˚28’41.75’’W/199m x x
Olentangy OL-1 38˚56’8.58’’N 83˚21’21.76’’W/228m x x
Shale Breccia SMB-1 39˚2’2.34’’N 83˚25’22.41’’W/226m x

Tymochtee TY-1 38˚55’58.43’’N 83˚21’53.31’’W/244m x x SMB-1-1B 39˚2’2.34’’N 83˚25’22.41’’W/226m x x

Dolomite
TY-2 38˚26’25.55’’N 83˚21’32’’W/213m x x SMB-2 39˚1’45.20’’N 83˚25’48.22’’W/246m x x

Greenfield GN-1 38˚58’2.56’’N 83˚20’14.44’’W/227m x x SMB-3A 39˚1’59.82’’N 83˚25’47.94’’W/220m x x

Dolomite
SMB-4A 39˚2’12.52’’N 83˚25’17.84’’W/241m x x
Peebles PB-1 38˚56’10.75’’N 83˚26’46.70’’W/280m x x
Dolomite SMB-4B 39˚2’13.05’’N 83˚25’16.77’’W/235m x
PB-2 38˚56’10.75’’N 83˚26’46.70’’W/280m x x
SMB-4C 39˚2’14.01’’N 83˚25’16.12’’W/226m x
PB-3 38˚56’10.75’’N 83˚26’46.70’’W/280m x x
SMB-5A 39˚3’9.98’’N 83˚24’28.69’’W/220m x x
PB-4 38˚56’20.41’’N 83˚21’39.71’’W/ x x
SMB-5B 39˚3’9.98’’N 83˚24’28.69’’W/220m x
Lilley LY-1 38˚56’12.24’’N 83˚26’58.43’’W/274m x x
Formation SMB-6 39˚1’1.55’’N 83˚25’5.23’’W/219m x
LY-1 38˚56’12.24’’N 83˚26’58.43’’W/274m x x
SMB-7 39˚1’29.26’’N 83˚25’18.70’’W/236m x
Bischer BH-1 38˚56’12.88’’N 83˚27’3.37’’W/267m x x
Formation SMB-9 39˚2’4.02’’N 83˚25’57.97’’W/256m x

Estill Shale ES-1 38˚56’13.44’’N 83˚27’7.24’’W/267m x x SMB-10 39˚2’4.02’’N 83˚25’57.97’’W/258m x

ES-2 38˚56’13.44’’N 83˚27’7.24’’W/267m x x

ES-3 38˚56’13.44’’N 83˚27’7.24’’W/267m x x

Ohio Journal of Science K.A. MILAM AND OTHERS 23

Table 2

Previous geochemical analyses of Geological Units Exposed in Southern Ohio and Northern Kentucky from 1Lamborn and others (1938);
2
USGS Ohio Shale Analyses; 3Stout (1941); 4Rogers (1936); 5Stout (1940); and ✝unpublished data from K. Milam

Formation Sample SiO2 Al2O3 Fe2O3 FeS2 MgO CaO Na2O K2O CO2 TiO2 P2O5 SO3 MnO Ref.
(+FeO) (S)

Cuyahoga S1 64.2 15.7 7.1 1.6 0.5 0.5 3 1.1 0 1

Formation

Bedford S2 59.4 17.2 8.9 1.5 0.5 0.2 2.9 1.2 0 1

Shale
✝
S3 52.96 13.63 6.38 1.17 1.57 0.22 3.91 0.74 0.08 0.22

Ohio Shale S4 49.28 12.27 9.34 1.54 1.05 0.38 3.35 1.01 0.71 0.11 5.35 0.042 2

S5 63.03 16.56 5.62 1.56 1.1 0.4 4.05 0.88 0.1 1.16 0.06 1

Olentangy S6 63.11 16.57 5.12 1.17 0.98 0.52 4.44 0.96 0.11 1.47 0.01 1
Shale

S7 0.4 0.02 0.26 0.09 21.55 30.16 * * 47.25 0.01 0.08 0.05 0.03 3

S8 1.65 0.02 0.38 0.06 20.95 29.68 0.12 0.16 46.51 0.04 0.05 0.02 0.015 3,4

Greenfield S9 1.77 0.04 0.51 0.03 20.62 30.29 0.01 0.01 46.62 0.02 0.03 * 0.02 4
Dolomite

S10 3.1 0.03 0.43 0.19 20.51 29.2 0.1 0.11 45.65 0.06 0.17 0.03 0.03 4

S11 0.44 0.14 0.25 0.06 21.6 30.1 0.01 0.03 47.29 0.01 0.05 0.01 0.03 3

S12 0.92 0.12 0.9 0.15 21.15 29.55 0.02 0.07 46.74 0.02 0.06 0.01 0.06 3

S13 0.33 0.09 0.42 0.02 21.07 30.75 * * 47.07 0.02 0.05 0 0.08 3,4

Peebles S14 0.77 0.27 0.63 0.24 21.1 29.72 * * 46.81 0.03 0.03 * 0.045 4
Dolomite

S15 0.95 0.14 0.45 0.15 20.85 30.11 * * 46.84 0.02 0.09 * 0.03 3,4

S16 0.24 0.11 0.21 0.2 21.4 30.18 * * 47.34 0.01 0.02 0 0.01 4

S17 0.33 0.12 0.36 0.14 21.35 30.03 * * 47.15 0.01 0.05 0 0.01 3,5

S18 0.77 0.04 0.44 0.17 20.77 30.53 * * 47.1 0.02 0.02 * 0.02 3,4

S19 1.42 0.33 0.43 0.35 1.45 52.7 0.01 0.01 43.1 0.02 0.015 0.02 0.04 3,4

Lilley S20 7.01 1.64 1.09 0.14 19.08 27.17 0.16 0.12 42.47 0.09 0.04 0.14 0.05 3,4
Formation

S21 6.68 1.78 0.99 0.27 18.94 27.23 0.11 0.14 42.35 0.11 0.02 0.08 0.02 3,4

S22 13.4 1.8 2.88 0.63 16.52 24.47 0.1 0.11 38.5 0.14 0.05 0.08 0.06 3

S23 18.38 3.92 1.99 0.33 12.8 25.75 * * 35.4 0.22 0.07 0.04 0.07 3,4

Bisher S24 24.92 1.82 1.06 0.16 14.7 22.32 0.03 0.04 33.82 0.18 0.1 * 0.11 3,4
Formation

S25 7.34 1.59 0.91 0.47 18.92 27.2 0.1 0.11 42.25 0.09 0.05 0.04 0.04 3,4

S26 3.4 1.29 1.11 0.11 1.25 50.45 0.03 0.1 41.54 0.04 0.12 0.01 0.1 3

S27 3.48 1.35 1.71 0.23 3.63 46.92 0.03 0.11 41.7 0.04 0.12 0.02 0.11 3

Brassfield S28 7.41 1.55 1.9 0.25 2.7 45.44 0.15 0.25 39.27 0.14 0.16 0.02 0.15 3
Limestone
S29 4.5 1.29 2.69 0.3 2.07 47.45 0.03 0.04 40.38 0.07 0.21 * 0.22 3

Values for SrO, V2O5, BaO2 and ZnO were * or not reported; C, H, H20, and LOI values reported in most reference; * = values <0.01 weight percent
24 BRECCIA AT SERPENT MOUND CRATER Vol. 110

may represent one or more of these formations. This is partially exposures, which is in contrast to the XRD and XRF results.
in contrast with field observations that initially indicated that Collapse of the flanks of the central peak would result in mixtures
the breccia may be a previously unidentified unit. For example, of Upper Ordovician to Middle Silurian carbonates early on or
although the Peebles Dolomite is poorly bedded and does contain later collapses that mixed Lower to Middle Silurian carbonates.
<1m thick breccia layers, it is unlike the breccia in that it has a In either case, final mineral assemblages would consist of calcite,
characteristically vuggy texture, contains numerous macrofossils, dolomite, and quartz and breccia with higher weight percentages
and contains abundant asphalt (Swinford 1985; Baranoski and of CaO and SiO2 than observed. Similarly, slope collapse in the
others 2003). Although much like the anomalous breccia, the northern ring graben [where there are numerous breccia exposures,
Upper Silurian Greenfield Dolomite and Tymochtee Formation would have incorporated quartz-rich strata of the Upper Devonian-
are noticeably devoid of macrofossils (Swinford 1985); bedding Mississippian (Fig. 2), a notion not supported by our analyses
is well developed in the Greenfield and Tymochtee, but not the (Table 3; Fig. 6)].
breccia. However, the lack of primary and secondary sedimentary Deposition of brecciated material as ballistically-emplaced
features may be explained by extensive weathering, a notion that or resurged ejecta is also unlikely because of the compositional
led previous geologists (Stout 1940; Stout 1941; Schmidt and homogeneity. The concentration of the anomalous breccia within
others 1961) to identify the breccia as weathered Peebles Dolomite. the crater and apparent lack of breccia outside the crater is consistent
No evidence of fault movements such as strain indicators or with fallback of ejecta and its removal beyond the crater rim or
slickensides has been observed in any breccia exposure. Furthermore, deposition by resurge following a marine impact event. However,
in contrast to fault breccias, which are concentrated in fault zones, the lack of mineralogic diversity or compositional heterogeneity
the breccia is laterally extensive and exposed across the crater. The suggests that the breccia does not represent approximately 480
compositional homogeneity and lack of mineralogic diversity of the m of excavated Upper Ordovician-Lower Mississippian strata (as
anomalous breccia do not support fault movements that could have calculated above). Such ejecta would be comprised of a dolomite-
transported fragments from a wider range of geologic units, such calcite-quartz mineral assemblage representing a mixture of
as the Upper Devonian Ohio Shale. In addition, no evidence of carbonate and siliciclastic sedimentary rock. Even if the impact
fault movements such as strain indicators or slickensides has been event occurred as recently as the Late Permian as suggested by Watts
observed in any breccia exposure. (2004) and Schedl (2006) and did not excavate Ordovician strata,
Compositional homogeneity of the breccia is also inconsistent
with emplacement by post-impact gravitational collapse of over-
steepened slopes with the impact crater. Slope failure would result
in the entrainment and mixing of strata upslope from breccia

Table 3
Major minerals identified by XRD

Formation Name (Sample #’s) Dolomite Calcite Quartz

Berea Sandstone (BS-1) x

Bedford Shale (BD-1) x

Ohio Shale (OS-1, OS-3) x

Olentangy Shale (OL-1) x

Tymochtee Fm. (TY-1, TY-2) x x

Greenfield Dolomite (GN-1) x

Peebles Dolomite (PB-1 through 4) x

Lilley Formation (LY-1, LY-2) x x x

Bisher Formation (BH-1) x x

Estill Shale (ES-1 through 3) x x

Brassfield Limestone (BL-1) x x

Drakes Formation (DK-1) x x x

Bill Fork/Waynesville Fm. (BF 1,2) x x

Figure 6. Representative XRD spectra of selected geologic units exposed near
Anomalous Breccia SMB-1 through SMB-10) x ο
the Serpent Mound impact crater (lower) and the anomalous breccia (upper).
Sample names correspond to those presented in Table 1. Excerpt for BF-1, all
x = positive identification; spectra have been offset for clarity. Gray vertical lines show some of the primary
ο = minor amount in sample SMB-9 peaks for calcite (C), dolomite (D) and quartz (Q).
Ohio Journal of Science K.A. MILAM AND OTHERS 25
26 BRECCIA AT SERPENT MOUND CRATER Vol. 110

Figure 7. Geochemical variation diagrams of the anomalous breccia and geologic units exposed in Adams, Highland and Pike counties in soutwestern Ohio. Composi-
tional ranges for each geologic unit were constructed using data from Tables 2 and 4 (not including samples S19-S21 and S-25). Sample numbers correspond with those
in Tables 1, 2 and 4. The left column shows the full range of analyses for a given oxide and the right column show an expanded view. Error bars shown where larger than
the symbol for an individual sample.
Ohio Journal of Science K.A. MILAM AND OTHERS 27

ejecta from the event would have higher weight percentages SiO2 Shale in the northern part of the crater. Under this scenario, the
compared to our results. discrepancy in thicknesses might be explained by the erosional
If the carbonate breccia represents ejecta, then it would require unconformity between the Upper Silurian Tymochtee Formation
ejection of only up to 42 m (Swinford 1985) of Middle-Upper and the Upper Devonian Ohio Shale (Swinford 1985). Assuming
Silurian dolostones (Fig. 2), which is only one-tenth of that that Upper Silurian and Lower to Middle Devonian strata were
required for an eight km diameter impact crater. Such a notion deposited in the southwestern part of the state (as in eastern Ohio),
would require a pre-Late Devonian impact (i.e. ejecta would the combined thicknesses of these geologic units (Salina Group,
not include the detrital sedimentary rock of the Devonian and Onondaga Limestone, etc.) with Middle-Upper Silurian strata at
Mississippian) because of the observed superposition of the Ohio Serpent Mound could account for this discrepancy. However, the

Table 4

Major element (oxide) chemistry of geologic units and breccia samples (SMB) from X-ray fluorescence analyses.

Sample #* SiO2 Al2O3 Fe2O3 MnO MgO CaO Na2O K2O TiO2 P2O2 Cr2O3 LOI Total

OS-1 41.08 8.9 3.9 0.012 1 16.57 0.39 2.25 0.57 0.14 0.01 22.95 97.85

OS-2 52.96 13.63 6.38 0.22 1.17 1.57 0.22 3.91 0.74 0.08 0.01 18.08 98.77

OL-1 65.14 10.02 3.07 0.007 0.65 0.05 0.38 2.85 0.61 0.04 <0.01 17.42 100.2

TY-1 2.98 0.92 0.76 0.014 20.45 29.21 0.07 0.06 0.04 0.01 <0.01 45.77 100.3

TY-2 2.62 0.5 0.72 0.022 20.74 29.47 0.03 0.04 0.04 0.01 <0.01 46.22 100.4

GN-1 0.64 0.26 0.64 0.013 21.19 29.78 0.03 0.01 0.02 0.01 <0.01 46.51 99.11

PB-1 0.44 0.3 0.35 0.018 21.15 30.39 0.01 0.01 0.02 0.01 <0.01 47.57 100.3

PB-2 1.72 0.65 0.64 0.025 19.91 28.6 0.04 0.02 0.04 0.01 <0.01 48.51 100.2

PB-3 1.55 0.73 0.5 0.021 20.68 29.7 0.04 0.02 0.06 0.01 <0.01 46.79 100.1

PB-4 0.32 0.17 0.28 0.016 16.38 23.87 2.92 <0.01 <0.01 <0.01 <0.01 40.33 84.32

LY-1 24.29 7.02 2.65 0.018 13.26 18.96 0.2 1.54 0.4 0.02 <0.01 30.96 99.32

LY-2 14.02 0.93 1.99 0.035 13.94 29.67 0.07 0.09 0.1 0.04 <0.01 39.05 99.93

BH-1 16.74 1.22 3.02 0.055 14.35 26.42 0.05 0.13 0.18 0.05 <0.01 36.71 98.93

ES-1 65.31 8.39 4.99 0.046 2.42 5.54 0.14 2.63 0.76 0.02 0.01 9.27 99.69

ES-2 55.77 21.09 3.9 0.026 2.19 1.89 0.29 5.05 1.08 0.09 0.01 8.73 100.1

ES-3 56.47 5.89 8.68 0.03 1.97 9.57 0.16 2.05 0.59 0.13 <0.01 13.71 99.25

DK-1 22.57 3.94 4.77 0.132 9.15 25.32 0.31 1.07 0.33 0.12 <0.01 30.59 98.31

BL-1 34.43 1.64 1.62 0.07 0.37 33.57 0.01 0.44 0.1 0.03 <0.01 27.66 100

BF-1 2.34 0.81 1.39 0.096 0.74 52.48 0.06 0.06 0.05 0.11 <0.01 41.18 99.3

BF-2 8.46 2.43 1.97 0.101 0.98 46.93 0.12 0.38 0.15 0.19 <0.01 36.95 98.65

SMB-1-1B 0.18 0.19 0.73 0.016 21.15 30.01 0.02 0.01 0.01 0.01 <0.01 47.19 99.52

SMB-2 0.23 0.24 0.4 0.013 21.28 30.24 <0.01 0.01 0.01 0.01 <0.01 47.53 99.97

SMB-3A 1.58 0.61 0.45 0.018 20.42 29.02 0.31 0.01 0.02 0.01 <0.01 46.37 98.82

SMB-4A 0.27 0.28 0.26 0.016 21.34 30.27 0.01 <0.01 0.01 0.01 <0.01 47.7 100.2

SMB-5A 0.22 0.22 0.25 0.012 21.25 30.2 0.01 0.01 0.01 0.01 <0.01 47.65 99.85

* For geologic unit names, see Table 1.

28 BRECCIA AT SERPENT MOUND CRATER Vol. 110

Salina Group contains considerable halite, anhydrite, gypsum, and prior to the Late Devonian deposition of the Ohio Shale. Such
quartz (Pepper 1947; Frank 1963; Wolfe 2001), none of which shallow excavation can occur with impacts into deeper marine
were detected by XRD analyses of the anomalous breccia (Fig. 6). settings (Gault and Sonnett 1982; Ormö and Lindström 2000),
Therefore, if the breccia represents ejecta, it would represent the where larger amounts of water dissipate energy from the impact.
ejection and deposition of only Middle-Upper Silurian dolostones The erosional unconformity at the Upper Silurian-Upper Devonian

Table 5
Summary of Statistical Analysis Through Spearman’s Rank Coefficient (percent correlated threshold met: * denotes <90 percent)

Unit Sample SMB-1- SMB- SMB- SMB- SMB- Unit Sample SMB-1- SMB- SMB- SMB- SMB-
1B 2 3A 4A 5A 1B 2 3A 4A 5A

S18 95 95 100 99.5 95

Peebles Dolomite
Lilley Formation-
Shale Formation
Bedford Cuyahoga

S1 * * * * *
S19 95 95 99.5 95 99

S20 99.5 99.5 95 95 99.5

S2 * * * * *
LY-1 * * 95 90 *

OS-1 * * * * * LY-2 90 90 99 90 90
Ohio
Shale

Formation
S3 * * * * * S21 95 95 99.5 95 99

S4 * * * * * Lilley S22 95 95 99.5 95 99

Olentangy Shale

OL-1 * * * * * S23 95 95 100 99.5 95

S5 * * * * * S24 95 95 100 99.5 95

Lilley-
Bisher

S6 * * * * * S25 95 95 100 99.5 95

Tymochtee
Formation

TY-1 95 95 100 99.5 95 BH-1 90 90 99 90 90

TY-2 99 95 99.5 95 99 S26 95 95 99.5 95 99

Formation
Bisher

GN-1 99 99 99 95 99.5 S27 90 90 99.5 95 90

S7 90 90 95 90 95 S28 * * 95 90 *
Greenfield
Dolomite

S8 95 95 99.5 95 99 S29 95 95 100 99.5 95

S9 95 95 99.5 95 99 ES-1 * * * * *
Formation
Estill

S10 95 95 99 95 95 ES-2 * * * * *

PB-1 95 95 99.5 95 99 ES-3 * * * * *

PB-2 95 95 100 99.5 99 BL-1 * * * * *

PB-3 95 95 100 99.5 99

S30 * 90 95 95 *
Limestone
Brassfield

PB-4 95 95 99.5 90 99
S31 95 95 99.5 95 99
Peebles Dolomite

S11 95 95 99.5 95 99
S32 90 90 90 90 90
S12 95 95 99.5 95 99
S33 90 90 90 * 90
S13 99.5 99.5 95 95 99.5
Drakes

S14 95 95 99.5 95 99
Fm.

DK-1 90 90 90 90 90

S15 95 95 99.5 95 99
Formation
Bull Fork

S16 95 95 99.5 95 99 BF-1 * * 90 * *

S17 99.5 99.5 95 95 99.5 BF-2 * * 90 90 *

Ohio Journal of Science K.A. MILAM AND OTHERS 29

and the occurrence of evaporites and certain sedimentary features Brockman CS. 1998. Physiographic regions of Ohio. Ohio Division of Geological
in Upper Silurian carbonates (Treesh and Friedman 1974; Coogan, Survey, map with text.
Bull C, Corbató CE, Zahn JC. 1967. Gravity survey of the Serpent Mound area,
1996; Tomastik 1997) however, indicates a regression following southern Ohio. Ohio J Sci 67(6):359-371.
deposition of Upper Silurian carbonates in what is now southwestern Bucher WH. 1933. Über eine typische kryptovulaknische Störung im südlichen
Ohio. Ohio: Geologicsche Rundschau 23:65-80.
Carlton RW, Koeberl C, Baranoski MT, Schumacher GA. 1998. Discovery of
If the breccia represents ejecta from a dry impact during the microscopic evidence for shock metamorphism at the Serpent Mound structure,
Late Silurian, the shallowness of the excavation would require south-central Ohio: confirmation of an origin by impact. EPSL 162:177-185.
that the collision must have occurred at a highly oblique impact Cohen AJ, Bunch TE, Reid AM. 1961. Coesite discoveries establish cryptovolcanics
angle. The asymmetric shape of the Serpent Mound impact crater as fossil meteorite craters. Science 134(3490):1624-1625.
Coogan AH. 1996. Ohio’s surface rocks and sediments. In: Feldmann RM,
(Reidel 1975) is consistent with this possibility. At the most Hackathorn M editors. Fossils of Ohio, Ohio Div Geol Surv Bull 70. Ch. 3.
common angle of 45° (Gilbert 1893; Shoemaker 1962) however, Dietz RS. 1960. Meteorite impact suggested by shatter cones in rock, Science
the shallow (<42 m) excavation required by the compositional 131(3416):1781-1784.
Dypvik H, Jansa LF. 2003. Sedimentary signatures and processes during marine
homogeneity of the anomalous breccia would have been impossible. bolide impacts: a review. Sed Geol 161:309-337.
However, as impact angle decreases, so does crater depth (Gault Frank GW. 1963. Salt intrusion in northeastern Ohio. Ohio J of Sci 63(2):55-59.
and Wedekind 1978; Pierazzo and Melosh 2000). If the maximum French BM. 1998. Traces of Catastrophe: A Handbook of Shock-Metamorphic
possible excavation depth for the Serpent Mound impact event is Effects in Terrestrial Meteorite Impact Structures. LPI Contribution No. 954.
Lunar and Planetary Institute, Houston, Texas.
480 m and the anomalous breccia as potential ejecta requires only Gault DE, Wedekind JA. 1978. Experimental studies of oblique impact. Proceedings
42 m of excavation, then the amount of excavation represented of the Lunar and Planetary Science Conference IX:3843-3875.
by the breccia would be 8.75 percent of the maximum expected Gault DE, Sonett CP. 1982. Laboratory simulation of pelagic asteroidal impact:
for Serpent Mound. We used final and maximum crater depths Atmospheric injection, benthic topography, and the surface wave radiation
field. In: Silver LT, Schultz PH eds. Geological Implications of Impacts of
(Fig. 5 – along trajectory from Gault and Wedekind 1978) for Large Asteroids and Comets on the Earth, GSA Special Paper 190:69-92.
impacts into dry pumice dust as a first order approximation of Gilbert GK. 1893. The moon’s face, a study of the origin of its features. Bulletin
the percentage relationships between oblique and maximum (at a of the Philosophical Society of Washington 12:241-292.
Grieve RAF, Therriault AM. 2004. Observations at terrestrial impact structures:
90° impact angle) excavation depths. With this information, we Their utility in constraining crater formation. Meteoritics & Planetary Sci
calculated that the percentage of maximum possible crater depth 39(2):199-216.
(y) of an oblique impact is equal to: Koeberl C, Buchanan PC, Carlton RW. 1998. Petrography and geochemistry of
18.099 ln (x) + 13.3 drill core samples from the Serpent Mound structure, Ohio: confirmation of
impact origin, Lunar and Planetary Science Conference XXIX, Lunar and
where x represents the angle of impact. If y = 8.75%, then the Planetary Science Institute, Houston, Texas, abstract #1392.
required angle of impact for the Serpent Mound event would be Lamborn RE, Austin CR, Schaaf D. 1938. Shales and surface clays of Ohio.
approximately 0.78°, the probability of which is <3% (Gault and Geological Survey of Ohio, Fourth Series, Bulletin 39. Ohio Division of
Geological Survey, Columbus, Ohio. 281 p.
Wedekind 1978). Locke J. 1838. The Ohio Division of Geological Survey Second Annual
Report:201-286.
CONCLUSION Melosh HJ. 1989. Impact Cratering: A Geologic Process. New York. Oxford
University Press. 245 p.
Field, laboratory, and statistical analyses all suggest that the Norrish K, Hutton JT. 1969. An accurate X-ray spectrographic method for the
anomalous breccia is a Middle-Upper Silurian dolostone member of analysis of a wide range of geological samples. Geochimica et Cosmochimica
the Tymochtee Formation, Greenfield Dolomite, Peebles Dolomite, Acta 33(4):431-453.
or a mixture thereof. The lateral distribution and homogeneous Ormö J, Lindström M. 2000. When a cosmic impact strikes the sea bed. Geological
Magazine 137(1):67-80.
dolomitic composition all support the conclusion that the Pepper JF. 1947. Areal extent and thickness of the salt deposits of Ohio. Ohio J
anomalous breccia was deposited during the Middle-Late Silurian Sci 47(6):225-239.
and was present at the time of the Serpent Mound impact event. Pierazzo E, Melosh HJ. 2000. Understanding oblique impacts from experiments,
observations, and modeling. Annual Review of Earth and Planetary Sci
This breccia was not formed by fault movements, slope collapse, 28:141-167.
or ejecta from the impact event. Further sedimentological and Reidel SP. 1975. Bedrock geology of the Serpent Mound cryptoexplosion structure,
stratigraphic analysis, in the context of detailed geologic mapping, Adams, Highlands, and Pike Counties, Ohio, Ohio Div Geol Surv Report of
could provide insights into the formation of this pre-impact Invest No. 95, Ohio Division of Geological Survey, map with text.
Reidel SP, Koucky, FL, Stryker, JR. 1982. The Serpent Mound disturbance,
carbonate breccia and its stratigraphic significance. Southwest Ohio. American J Sci 282:1343-1377.
Rexroad CB, Branson ER, Smith MO, Summerson C, Boucot AJ. 1965. The
Acknowledgements. The authors would like to thank the reviewers for their Silurian formations of east-central Kentucky and adjacent Ohio. KY Geol
thorough reviews and thoughtful insights which improved the quality of this Surv Bull 2, Series X. 34 p.
manuscript. In addition, the authors would like to thank our collaborators, Mark Rogers JK. 1936. Geology of Highland County. Geological Survey of Ohio, Fourth
Baranoski and Greg Schumacher, at the Ohio Division of Geological Survey of Series, Bulletin 38. Ohio Division of Geologic Survey, Columbus, Ohio. 148 p.
the Ohio Department of Natural Resources for their thoughtful reviews of this Schedl A. 2006. Applications of twin analysis to studying meteorite impact
manuscript and their encouragement and insight into Serpent Mound and the local structures. EPSL 244:530-540.
stratigraphy. Mark first introduced author Keith Milam to the breccia during a field Schmidt RG, McFarlan AC, Nosow E, Bowman RS, Alberts R. 1961. Examination
trip in the fall of 2004. The authors further thank the many field assistants who of Ordovician through Devonian stratigraphy and the Serpent Mound chaotic
have contributed invaluable time and effort toward the completion of this project structure area. Field Trip 8 in Guidebook for Field Trips Cincinnati Meeting,
including: Tara Cardenas, Cory Frederick, and Lee Johnson of the Department of 1961. Geological Society of America Meeting.
Geological Sciences at Ohio University. We also thank the Martin family, Peebles, Schumacher GA, Reidel SP. 1997. Reconnaissance bedrock geology of the Sinking
Ohio for allowing us to use their farm as a base of operations. Spring, Ohio quadrangle. Digital Map Series BG-2. Ohio Department of
Natural Resources, Division of Geological Survey.
Schumacher GA, Reidel SP. 2002a. Reconnaissance bedrock geology of the
LITERATURE CITED Byington, Ohio quadrangle. Digital Map Series BG-2. Ohio Department of
Baranoski MT, Schumacher GA, Watts DR, Carlton RW, El-Saiti BM. 2003. Natural Resources, Division of Geological Survey.
Subsurface geology of the Serpent Mound disturbance, Adams, Highlands, Schumacher GA. 2002b. Reconnaissance bedrock geology of the Jaybird, Ohio
and Pike Counties, Ohio. Report of Investigations No. 146. Ohio Division quadrangle. Digital Map Series BG-2. Ohio Department of Natural Resources,
of Geological Survey. Columbus, Ohio. 60 p. Division of Geological Survey.
30 BRECCIA AT SERPENT MOUND CRATER Vol. 110

Schumacher GA. 2002c. August F. Foerste’s unpublished field notes for the area of Swinford EM. 1991. Reconnaissance bedrock geology of the Peebles quadrangle,
the Serpent Mound disturbance. Open-File Report 2001-01. Ohio Department Adams County, Ohio. Digital Map Series BG-2. Ohio Department of Natural
of Natural Resources, Division of Geological Survey. Resources, Division of Geological Survey.
Shoemaker EM. 1962. Interpretation of lunar craters. In: Kopal Z, editor. Physics Tomastik TE. 1997. The sedimentology of the Bass Islands and Salina groups in
and Astronomy of the Moon. New York/London: Academic. 538 p. 283-359. Ohio and its effect on salt-solution mining and underground storage, USA.
Shrake DL, Swinford EM, Schumacher GA, Larsen GE, Slucher ER. 2007. Carbonates and Evaporites 12(2):236-253.
Descriptions of geologic map units: A compendium to accompany Division Treesh MI, Friedman GM. 1974. Sabkha deposition of the Salina Group (Upper
of Geological Survey Open File 7.5 Minute Bedrock-Geology Maps. Open- Silurian) of New York state. In: Coogan AH, editor. Northern Ohio Geological
File Report 98-1. Ohio Division of Natural Resources, Division of Geological Society Symposium on Salt, Fourth Proceedings: 1:35-46.
Survey. 8 p. Watts DR. 2004. Paleomagnetic determination of the age of the Serpent Mound
Stout W. 1940. Marl, tufa rock, travertine, and bog ore in Ohio. Geological Survey structure. Ohio J Sci 104(4):101-108.
of Ohio, Fourth Series, Bulletin 41. Ohio Division of Geological Survey, Willoughby, CC. 1919. The Serpent Mound of Adams County, Ohio. Am
Columbus, Ohio. 56 p. Anthropologist 21(2):153-163.
Stout W. 1941. Dolomites and limestones of western Ohio. Geological Survey Wolfe ME. 2001. Gypsum in Ohio. In: Wolfe ME, editor. 2000 Report on Ohio
of Ohio, Fourth Series, Bulletin 42. Ohio Division of Geological Survey, Mineral Industries with Directories of Reporting Coal and Industrial Mineral
Columbus, Ohio. 468 p. Operators, Ohio Division of Geological Survey. p. 1-5.
Swinford EM. 1985. Geology of the Peebles quadrangle, Adams County, Ohio.
Ohio Journal of Science 85(5):218-230.