Jump to ContentJump to Main Navigation
Hopewell Settlement Patterns, Subsistence, and Symbolic Landscapes$

A. Martin Byers and DeeAnne Wymer

Print publication date: 2010

Print ISBN-13: 9780813034553

Published to Florida Scholarship Online: September 2011

DOI: 10.5744/florida/9780813034553.001.0001

Show Summary Details
Page of

PRINTED FROM FLORIDA SCHOLARSHIP ONLINE (www.florida.universitypressscholarship.com). (c) Copyright University Press of Florida, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in FLASO for personal use (for details see http://www.florida.universitypressscholarship.com/page/privacy-policy). Subscriber: null; date: 21 June 2018

Hopewell Cosmography at Newark and Chillicothe, Ohio

Hopewell Cosmography at Newark and Chillicothe, Ohio

Chapter:
(p.128) 5 Hopewell Cosmography at Newark and Chillicothe, Ohio
Source:
Hopewell Settlement Patterns, Subsistence, and Symbolic Landscapes
Author(s):

Hively Ray

Horn Robert

Publisher:
University Press of Florida
DOI:10.5744/florida/9780813034553.003.0006

Abstract and Keywords

This chapter is an extension of pioneer work on the astronomical alignments of the geometrical earthworks of Newark and Chillicothe with a more global approach, not only situating the Newark Earthworks within the wider natural landscape of the Raccoon Valley but also directing their attention to the Chillicothe region, undertaking a very ambitious analysis of how the multiple, large-scale earthworks of this region may be deliberately aligned both with one another and with local natural-astronomical points of relevance.

Keywords:   cosmography, Ohio Hopewell, Newark, Chillicothe, Raccoon Valley, earthworks

The construction of large geometric earthworks is generally attributed to the Hopewell culture in Ohio during the period A.D. 1–400. The objective of our research for the past 25 years has been to test the hypothesis that astronomical knowledge played a significant role in the location and design of the large, geometrically sophisticated earthworks for which Ohio Hopewell is best known. Our work has established that some of these earthworks are aligned to significant rise or set points of both the sun and moon. However, the intentionality of such alignments remains a difficult and unresolved problem. In the absence of written sources, reliable oral tradition, or small art clearly related to the design of the earthworks, it is hard to justify compellingly either belief or doubt when features of these structures are interpreted as deliberate alignments to solar or lunar events visible on the horizon.

The maps of the earthworks published by Squier and Davis in 1848 afford a first approach to their design and scale, but well-known inaccuracies render them unreliable as a basis for a precise quantitative study of the geometry and orientation of the sites and, without corroboration, useless for astronomical analysis. Fortunately, beginning with the Smithsonian Bureau of Ethnology’s surveys in the 1880s, major geometric earthworks have been mapped with greater accuracy (Thomas 1894). Since the 1930s these surveys have been tested and supplemented by aerial photography, remote geophysical sensing, and scientific excavation.

In this chapter, we summarize our work during the past decade at Newark and Granville and in the Ohio Hopewell core near Chillicothe. In the early 1980s, we published papers on the geometry and astronomy of the earthworks at Newark, Ohio, and High Bank, near Chillicothe, Ohio. That work depended on the accuracy of James Middleton’s surveys of the sites for the Bureau of Ethnology. More recently we have been able to utilize the newly rediscovered 1862 Salisbury survey of the entire Newark site (Salisbury 1862) as well as to (p.129) draw on the results of aerial photography, pollen analysis, remote geophysical sensing, and excavation at Newark and other Ohio Hopewell sites.

The qualitative interpretations we outline here are analyzed in quantitative detail in a statistical study of five key lunar alignments along features of the Circle-Octagon earthworks at Newark (Hively and Horn 2006a) and in forthcoming papers. In these papers we will detail the possible role of astronomical observations made from high hilltops in the planning and construction of Newark-Granville and major sites in the Chillicothe region and more closely examine the role played by regional and celestial “topography” in the location and layout of the Newark-Granville and Chillicothe sites. The general theme of these studies is Hopewell cosmography and its landscape. We have found that the Hopewell builders were concerned not only with the geometry and astronomical alignment of some of their works but also with the place and location of those works in the local terrain. That place and those places appear to have been chosen not only for broadly economic suitability but also for their significance in the builders’ view of their world (Buikstra and Charles 1999: 216; Sahlqvist 2001: 79–102).

Statistical Analysis

We have recently completed a Monte Carlo statistical analysis of the most accurate and compelling lunar alignments associated with the Newark Octagon (Hively and Horn 2006a). In our original papers, we found at the Newark site 17 possible alignments to local horizon rise and set extremes of the moon over its 18.6-year cycle.1 Rise extremes are shown in (Figure 5.1). The plausibility of these alignments being deliberate depends upon the likelihood that the five well-defined alignments on the avenue axis and walls of the Observatory Circle-Octagon, shown in (Figure 5.2), are accidental. The structure is a unique combination of geometrical and empirical symmetries—the geometrical symmetry of the octagon and the astronomical symmetry of the long lunar cycle. Five of the eight lunar extremes are aligned with the octagon (the maximum possible on the walls and avenue axis of the nearly perfect Circle-Octagon). The remaining three lunar extremes are aligned with other features of the Circle-Octagon but are not aligned on the walls. The shape of the octagon conforms to a precise geometrical plan, and its size is related to the associated circle as shown in (Figure 5.3).

In most cases, statistical analysis cannot establish the probability that the builders intended chosen properties of a single site. At Newark, the scale, geometrical precision, and complexity of the Circle-Octagon make it possible to use Monte Carlo techniques to assess the statistical distribution of random (p.130)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.1. Range of movement of moonrise along the eastern horizon from the Newark Circle-Octagon.

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.2. The five lunar alignments embedded in the Newark Octagon.

(p.131)
Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.3. The OCD Geometry of the Newark Circle-Octagon.

astronomical alignments in such structures by building and analyzing billions of comparable sites by computer. With well-defined computer-generated data sets we can compute the probability that astronomical alignments comparable to those found at Newark are random accidents. We tested 72 models that made a range of assumptions, which we believe are persuasively indicated by data at the site, about intended geometrical plans, wall errors, alignment tolerances, exclusive lunar alignments, and deliberate geometrical distortion. Our analysis showed that for the most plausible models, the probability that the octagon design achieves five lunar alignments (comparable to those actually found) on its symmetry axis and four walls by chance varies between 4.1.6 ×10–8 and 2.8 .2 ×10–7.

The dependence of the probability of comparable chance alignments on the assumptions underlying the models is discussed in detail in the aforementioned work (Hively and Horn 2006a). Although the precise probability of comparable chance alignments varied with the modeling assumptions, the conclusion that the probability was less than .001 was quite robust and was true of all reasonable models. The Monte Carlo analysis does not by itself (p.132) prove that the Hopewell builders of the octagon intended these lunar alignments, but it does militate strongly against dismissing the hypothesis of deliberate astronomical alignment on statistical considerations alone. This leads us to conclude that the hypothesis of intentional astronomical alignment in Hopewell earthworks has sufficient inherent plausibility to be taken seriously as a basis for further study of Ohio Hopewell.

Horizon Altitudes

In the course of our statistical work on the Newark Circle-Octagon, we tested two hypotheses regarding the horizons over which the Hopewell builders viewed the lunar extremes at moonrise and moonset. In our original paper we judged that the lunar extremes were observed and presumably recorded at first gleam on moonrise and at last gleam on moonset as seen over local horizons by observers within the Circle-Octagon earthworks. This gave the best fit to the data. A statistically implausible error in one of the five core alignments led us to reconsider.

An error of 1°.7 in the alignment of wall EF with the north maximum moonset is four times the average error of the other four alignments. The 1°.7 error results from a local hill on the azimuth of wall EF 1.9 km from vertex E of the Newark Octagon. If the intent of the builders had been to mark the northern maximum moonset with high accuracy as observed from vertex E, it would seem strange to choose a location where the distant horizons were obscured by local obstructions. Moreover, the local hill forming the visible horizon does not appear to distinguish itself in any way as being particularly useful as an astronomical foresight. If a primary priority of the designers had been to conduct precise observations of the lunar maximum north set point along wall EF, it would appear to have been better to move the octagon farther south, away from obstructing hills north of Raccoon Creek. The alignment error of wall EF relative to the moonset as it would appear on a zero-altitude horizon for lower limb tangency is negligible (0°.2). These considerations led us to formulate a zero-altitude hypothesis which we now believe gives a better explanation of the astronomical alignments at the site.

The zero-altitude hypothesis asserts that the key lunar alignments found in the Circle-Octagon earthworks were designed to fall along lunar lines which had been determined by observations conducted from high points with distant horizons and not by observations made from the valley floor where the earthworks are located. If the intent of the builders was to accurately record the “true rise and set points” as they would be seen unobscured by local topography, then it is plausible that these astronomically significant lines would (p.133) be determined from observations made from high hilltops in the region rather than from positions within the octagon structure itself.

The question of how the lunar extremes might have been discovered and marked with an accuracy of about 0°.5 is also addressed by the zero-altitude hypothesis. Such a feat would have required determined observation over several generations. Consistent and precise observations would have been difficult to make from valley floors, where the directions to rise and set points appear to vary significantly due to obscuration by local topography and seasonal vegetation. The most logical place to make and record repeatable observations of lunar rise/set points would indeed be from high places with unobstructed views of distant horizons. Such high places could be dedicated to long-term observations, as they would not have been prime areas for other activities.

The zero-altitude hypothesis would predict that suitable elevations should be found in positions where they could function as long-distance backsight for the lunar alignments found in the Circle-Octagon structure. Skeptics will rightly observe, however, that the fulfillment of such a prediction is likely to be vacuously true by chance alone given the large number hills which surround the valley. Consequently, evidence for this hypothesis will only be compelling if high points can be found which have additional characteristics (beyond their functioning as a backsight for a single alignment) suggesting that they were deliberately chosen and used to make the observations which the hypothesis requires.

As we reflected on empirical problems involved in the initial planning and layout of the Circle-Octagon, we studied the local topography to locate vantage points that command a view of the earthwork location on the flood plain of Raccoon Creek. The terrain of the nearby Appalachian foothills affords many such points. We have used three criteria in looking for sites which may have played a role in the planning and layout of the earthworks: (1) the site can serve as a backsight which allows a zero-altitude horizon view of a lunar (or solar) extreme over the length of an octagon wall or along the symmetry axis of one or more geometrically regular constructions, such as the CircleOctagon, the Great (or Fairground) Circle, the Wright Square, the Salisbury Square; (2) the site is prominent among the many eligible hill sites, affording an optimum view of the earthworks; and (3) the site may show evidence of Hopewell or Early Woodland activity.

One site, the most prominent hill (H1) southwest of the earthworks, fits these criteria in an impressive fashion. From the highest elevation on H1 the azimuth to the north maximum moonrise passes through the long axis of the Observatory Circle-Octagon, and the azimuth to the north minimum moonrise (p.134)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.4. H1 alignments of the Newark Earthworks. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

passes through the centroid of the Great Circle, as shown in (Figure 5.4). Samuel Park, in fact, noted that in 1870 a “circumvalation” (enclosure) and mounds near the H1 location site afforded a clear view of Cherry Valley and the Newark Earthworks (Park, 1890).

Here, and throughout this paper, unless specified otherwise, the azimuths for rise/set events will be defined in terms that we believe offer the best fit to the data for the Circle-Octagon. The horizon over which the rise and set points are observed is assumed to be zero, consistent with the evidence that the astronomical observations that discovered and marked the positions were made from high elevations. We further assume the rise/set events were determined by lower limb tangency to the horizon. We also have assumed a date of A.D. 250 in computing lunar rise/set azimuths, although the maximum deviation of azimuth for any date in the interval A.D. 1–400 will be less than 0.05 degrees, that is, negligible. Finally, all the astronomical alignments proposed in this study have an accuracy of 1 degree or better. Given the observational difficulties associated with weather, daylight, and lunar phase, this is consistent with the best accuracy that could reasonably be expected for marking the lunar extremes.

(p.135) A second site may suggest one reason the Hopewell chose to build so extensively at Newark. Coffman’s Knob, southeast of the Newark Works (the highest hill overlooking the area), commands a view of the confluence of Raccoon Creek and the South and North Fork of the Licking River. The view from Coffman’s Knob also suggests a fourth criterion for sites which may have played a role in the layout of earthworks: a site from which lunar extremes appear aligned with both earthwork architecture and prominent features of the local terrain. Viewed from Coffman’s Knob, the moon at its north maximum extreme sets in Sharon Valley. Seen at its north minimum extreme, the moon sets in the valley of Raccoon Creek. Over the 9.3-year period between maximum and minimum standstills, the observer would see the extreme north set point of the moon cross slowly from one valley to the other, certainly a striking visual phenomenon (Figure 5.5).

Major features of the earthworks at Newark and Granville appear to conform to these views of moonset from Coffman’s Knob. The azimuth through Sharon Valley to the north maximum moonset passes through the center of the Salisbury Square parallel to the walls of the square (within 2 degrees) and passes through the approximate center of the elliptical enclosure to the north.2

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.5. Coffman’s Knob alignments of the Newark Earthworks. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

(p.136)
Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.6. Newark Earthworks showing Prominence H3 alignments. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

The azimuth through Raccoon Creek valley to the north minimum moonset passes near the entrance to the Great Circle of the Newark Earthworks and through a large, perhaps incomplete, circular embankment situated in the valley below the well-known “Alligator” (or panther) effigy east of Granville. No Hopewell construction is recorded near the peak of Coffman’s Knob, though there are Middle Woodland habitation sites, enclosures, and mounds nearby (Pacheco 1997: 51, (Figure 2.1)).

Zero-altitude horizon views of tangent moonrise and moonset from H1 and from Coffman’s Knob are along lines on which major elements of the Newark design reside—the long axes of the Newark Circle-Octagon and the Salisbury Square and Ellipse; the Great Circle; the Granville “Circle” and other features along the Raccoon Creek valley between Newark and Granville (Hooge 1993: 156–198).

Seen from a zero-altitude horizon vantage point, wall EF of the Newark Octagon is aligned accurately with the north maximum moonset. This suggests that this wall may have been carefully constructed to fall on that line as observed from high hilltops. When moonset is viewed over the local horizon (p.137) from vertex E of the octagon, wall EF misses the north maximum (similarly defined by lower limb tangency) by nearly four lunar diameters. Consistent with the roles we have suggested for H1 and Coffman’s Knob in the layout of Newark, it is pertinent to look for a hill on the bearing of wall EF which could serve as a backsight for wall EF and allow a zero-altitude horizon for moonset observation. Across the valley to the southeast, a small prominence (H3) in the first range of hills 4,300 meters distant from vertex E allows a zero-altitude view of the north maximum moonset on the line of octagon wall EF (Figure 5.6). The local topography affords comparable locations for longdistance backsights associated with the three remaining lunar oriented walls of the octagon. We examine the plausibility of these as vantage points, given the minor differences between the horizons seen along these walls from high points or from within the octagon, in a forthcoming paper.3

Comparison between Hopewell site planning at Newark and in the Chillicothe region is appropriate for many reasons. There are obvious similarities between the Newark Circle-Octagon and the High Bank Circle-Octagon, south of Chillicothe. It will also be necessary to test at Chillicothe the two hypotheses we have proposed in our recent work at Newark and Granville: (1) relatively high, zero-altitude horizon vantage points overlooking suitable second-terrace locations for earthworks were used in the planning and layout of ceremonial centers, and (2) some earthwork alignments to lunar extremes also incorporate alignments to major features of local topography, such as linear valleys and high prominent peaks.

Newark and High Bank

Design similarities between the circle-octagon components at Newark and High Bank are salient here, as is the fact that these are the only such constructions known in Hopewell or in American antiquity. The Newark Observatory Circle diameter (OCD unit) is accurately repeated at High Bank. At both Newark and High Bank, the dimensions of the octagons are derived from the circle diameters. Both octagons encode alignments to lunar extremes. The axis of the High Bank Circle-Octagon, however, does not replicate the Newark Circle-Octagon axis. As shown in (Figure 5.7), from a point on the perimeter of the High Bank Circle-Octagon (13) opposite the avenue, the observer does not see the moon rise at its north maximum extreme. Instead, one looks down the broad valley of the ancient, preglacial Teays River through which the present Scioto River flows.

In contrast to the Newark Octagon, only one of the eight alignments on lunar extremes that are achieved at the High Bank Circle-Octagon follows an octagon wall. Only one of the four solar alignments on the Circle-Octagon follows (p.138)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.7. Teays River valley alignment from point 13 on the perimeter of the High Bank Circle.

an octagon wall. At the High Bank Circle-Octagon the best fit with all of these alignments is to local rather than zero-altitude horizons. The High Bank alignments, with the lower limb of the moon or sun tangent to the horizon at rise and set, do, however, correspond with the zero-altitude interpretation of Newark.

In our Monte Carlo statistical analysis of randomly built equilateral octagons (with fourfold rotational symmetry) we discovered that there were only two precise octagon shapes which achieved the greatest number of alignments to lunar/solar extreme rise/set points along octagon walls or along an avenue axis (a maximum of five alignments is possible). These shapes were defined by the choice of a single vertex angle. Equilateral, symmetric octagons could have been built with a large variety of different shapes. Our analysis shows that at the latitude of Newark, the shape that aligns most accurately with the lunar extrema is defined by a maximum vertex angle (at the adjoining avenue) of 153°.6. The actual Newark Octagon has an average large vertex angle of 154°.8. (p.139) Similarly, the octagon shape (at the latitude of High Bank) which aligns with the most lunar/solar extrema is defined by a large vertex angle at the adjoining avenue of 165°.7, whereas the actual High Bank Octagon has an average large vertex angle of 165°.5. Thus the hypothesis of deliberate experimentation to determine the shapes which aligned with the most lunar/solar extrema “explains” the design of both octagons known to have been constructed by the Hopewell. In these cases, in the absence of other more compelling reasons for the choice of shape we believe this is a very notable fact.

The design of the High Bank Octagon is the optimum shape for aligning with a combination of solar and lunar extreme rise/set points at its latitude; yet compared with the Circle-Octagon at Newark, it fails to encode that astronomical design on the most prominent feature, its long axis. As the High Bank Circle-Octagon avenue is not aligned with a solar extreme, only two sides of the octagon rather than a possible four are aligned with lunar or solar extremes. The 0°.75 difference in latitude between the Newark Earthworks and the High Bank Earthworks makes it impossible to replicate the accurate match to lunar extremes achieved on the Newark Circle-Octagon merely by repeating the Newark design in (Figure 5.3) at Chillicothe. Some change in design is required.

High Bank in Context

In considering the differences between the two circle-octagon designs above, do we perhaps take up the problem from the wrong end by beginning with Newark, then finding High Bank anomalous? This finding reflects our initial approach. Because we could better document the Newark site, we began our study of the archives, surveying, and analysis there. When we reached the conclusion that there could be deliberate astronomical design at Newark, we tested that hypothesis against the available evidence at High Bank.

Let us reconsider, beginning with High Bank in the context of the Chillicothe Hopewell. What is the focus of the long axis of the High Bank design? From point 13 (Figure 5.7) on the circle opposite the avenue, features of the Circle-Octagon (symmetry axes, vertices, the single opening in the circle embankment other than the avenue) are aligned with (1) the southern maximum moonrise, (2) the winter solstice sunrise, and (3) the broad valley of the preglacial Teays River (Figure 5.8).4

The latter orientation deserves attention for four reasons: (1) south of High Bank, the Scioto abandons its meandering course to hug the west wall of the Teays Valley on a straight bearing of 143°.4 for more than 10 km; (2) the 143°.4 bearing from point 13 on the High Bank Circle bisects the large circle of the Liberty Works 6.3 km to the south (Figure 5.8); (3) the opposite 323°.4 bearing (p.140)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.8. Teays River valley with Liberty-High Bank-Works East alignment. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

from point 13 on the High Bank Circle bisects the large circle of Works East 3.5 km to the north;5 and (4) the north lunar maximum rise (53°.4, tangent to the local horizon) occurs at a 90-degree angle to the major axis of the CircleOctagon.

As is evident in (Figure 5.8) the siting of High Bank, relative to the Teays Valley and to both Liberty and Works East, suggests that these works belong to a (p.141) common plan. The design incorporates the empirical symmetries of the lunar and solar extremes, and the geometrical symmetry of the octagon, with the nearly unique linear topography of a long segment of the ancient river valley. Though High Bank appears anomalous to the viewer who approaches it with Newark as the norm, it not only fits its local context, but offers clues to a better understanding of that context, and perhaps many features of Newark as well.

We have seen the manner in which the design of Hopewell earthworks from Newark to Granville may have incorporated local topography with the geometrical symmetry of the octagon and the empirical symmetry of the long lunar cycle. At Chillicothe we find additional evidence of such design, supporting one hypothesis framed at Newark—some earthwork alignments that function to mark lunar extremes also incorporate alignments to major features of local topography. No geometric-astronomical feature of the Newark Earthworks is as demonstrably integrated with its topographic setting as is High Bank with this linear segment of the Teays-Scioto valley. Yet this coincidence of geometric, astronomical, and topographic features at High Bank is consistent with our findings at Newark.

The second hypothesis framed at Newark, that relatively high, zero-altitude vantage points overlooking suitable second-terrace locations for earthworks were used in the planning and layout of ceremonial centers, is not confirmed at High Bank.6 It remains to be seen whether it is supported by other earthen enclosures in the Hopewell core region near Chillicothe.

Paint Creek Valley

We have seen the alignment of Liberty, High Bank, and Works East along the bearing of the Teays Valley south of Chillicothe. Strong as this evidence of alignment is in the context of High Bank, it is important to ask if other regional sites are aligned both with local topography and with each other. The valley of the ancient Paint Creek is an obvious place to begin, as shown in the bedrock geology map of Ross County (Figure 5.9).

The course of the preglacial valley of Paint Creek bearing northeast is evident in (Figure 5.9), beginning at lower left. Although the valley walls are not as straight as the west wall of the Teays Valley below Chillicothe, the valley itself is prominent, as anyone approaching Chillicothe along U.S. Route 50 from the west can see. On the north side of the valley, three earthwork enclosures, Bourneville, Anderson, and Dunlap, are aligned with the approximate center of the large circular enclosure at Seip. On the south side of the valley, three enclosures, Spruce Hill, Mound City, and Cedar Bank, are aligned with the geometric center of the Baum Earthworks. The location and dimensions of these eight enclosures confine each of these “alignments” to a narrow range (p.142)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.9. Bedrock geology and distribution of major sites in the Chillicothe region. Contour interval 200 ft (Quinn and Goldthwait 1985: (Figure 3), p. 5, and with the authors’ additions). Courtesy of Ohio Geological Survey. Key: AN = Anderson; BA = Baum; BL = Blackwater; BO = Bourneville; CB = Cedar Bank; DU = Dunlap; FR = Frankfort; HB = High Bank; HO = Hopeton; HW = Hopewell; JU = Junction; LI = Liberty; MC = Mound City; SE = Seip; SH = Spruce Hill; SHR = Shriver; WE = Works East.

(p.143)
Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.10. Parallel alignments on Old Paint Creek valley and Teays River valley. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

within ± 0°.3 of 44°.3. The northern of the parallel 44°.3 alignments passes through the center of the enclosure at Anderson. The southern passes through Mound City at Mound 3 and through the eastern “gate” of the Mound City enclosure near the Scioto River. As with the Teays Valley alignments, suitable soils determine in part the location of large second terrace enclosures, but this does not explain the extension of the “alignments” beyond the preglacial Paint Creek valley to intersections with the Teays Valley near Dunlap and Cedar Bank (Figure 5.10). The 44°.3 bearing of the old valley aligns with no lunar or solar extreme.7

Two bearings from the Seip-Pricer Mound (near the center of the large circular embankment at the Seip enclosure) through other ceremonial centers do align with lunar extremes. As shown in (Figure 5.11), a 52°.2 bearing from the Seip-Pricer Mound on the north maximum moonrise passes close to the centroid of the Shriver enclosure (south of Mound City) and through the center of the southern small circle at the east wall of the Hopeton Polygon. A 66°.4 bearing from the same position on the Seip-Pricer Mound across the (p.144)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.11. Seip-Pricer Mound alignments. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions.

northern boundary of the Baum enclosure and through the estimated center of the Works East Square aligns with the north minimum moonrise. The situation here recalls H1 at Newark, where zero-altitude, lower limb tangent bearings taken from the same position pass along the symmetry axes of the Circle-Octagon and the Great Circle to the northern maximum and minimum moonrise. Yet the Seip enclosure is located in a valley, not on a prominence like H1.8 The corporate center at Seip appears to play a focal role among the Paint Creek sites, as it anchors alignments to lunar extremes and to other Hopewell enclosures, in some cases to both.

Some of these possible alignments may be accidental. Patterns consistent with astronomical alignment may result from choices based on properties of water and soil, sun, wind, and other factors that have not entered our reckoning. The question we must consider is whether the evidence we cite supports the conclusion that it is reasonable to believe the builders of these works were drawing patterns on a scale larger than the already extensive earthworks themselves. Were the Hopewell builders, and perhaps their Early Woodland (p.145) predecessors, thinking in terms that drew patterns in earth and sky on a common map?

Mound City and Shriver

The impression of singular local topography is strong where the northeastbearing valley of ancient Paint Creek intersects the Teays Valley near Mound City. From this point the Appalachian range (the Logan Range) east of the Scioto River appears in sharp relief, from Sugarloaf Mountain on the north to Mount Logan on the south. This range is the most prominent signature of Chillicothe topography, visible from high points in the region to the west for as much as 35 kilometers ((Figure 5.12) and (Figure 5.13)). This profile may have played a role in the siting of both Mound City and the Shriver “Circle” 728 meters south of Mound City.

Viewed from Mound 7, which forms the most prominent peak within the walls of Mound City, the Appalachian profile from the peak of Sugarloaf to the peak of Mount Logan defines the 48°.5 range of the monthly north-south movement of the moon during the season of minimum extremes which recurs every 18.6 years. During that “season,” which lasts about 18 months on either side of the extreme, the rising moon will repeatedly approach the limits of 66°.3 at the north and 114°.8 at the south, but it will appear to the observer as if bound within those limits. During the 9.3 years after the minimum extreme, the rising moon moves gradually toward the season of maximum extremes, where it travels monthly from 52°.1 at the north to 129°.3 at the south (see (Figure 5.1); the range of lunar extremes shown at Newark is slightly wider than at Chillicothe).

At the maximum extreme the rising moon travels 77°.2 from north to south each month, a visible increase of 28°.7, or 57 lunar diameters, over the minimum season. The profile of the Appalachian range to the east, viewed from Mound City, makes this difference, more than 28 lunar diameters at each boundary of the range, hard to miss. Skeptics have wondered why the minimum lunar extremes would be noticed (Aveni 2004). At Mound City the answer is in the mountains.9

The view of the Appalachian range from Mound 7 is salient for a second reason. From the same position where the minimum moonrise extremes occur over the peaks of Sugarloaf and Mount Logan, the observer will see the summer solstice sun rise at the northern base of Sugarloaf, and the winter solstice sun rise over the southern base of Mount Logan ((Figure 5.13) and (Figure 5.14)). Today trees obscure this view of the eastern horizon from Mound City. It would have been necessary to clear trees as far away as the location of Hopeton, (p.146)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.12. Eastern profile of the Appalachian range from Chillicothe. This photograph, taken from the intersection of Clinton Road and Pleasant Valley Road 9 km east of the Logan Range, shows the entire range from north to south. Today trees and construction obscure the view of the range from Mound City. Photo by the authors.

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.13. The Logan Range profile from Mound City. Photo by the authors. Key: ss = summer solstice sunrise; nm = north minimum moonrise; sm = south minimum moonrise; ws = winter solstice sunrise.

(p.147)
Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.14. Profile of Sugarloaf from Mounds 3 and 7, Mound City. Photo by the authors.

2.5 kilometers across the Scioto River, for the views described here to have been possible. Kendra McLaughlin’s recent findings at Fort Ancient are consistent with such land clearing, as are Brad Lepper’s and DeeAnne Wymer’s earlier findings at Newark (McLauchlan 2003: 557–566; Lepper 1996; Wymer 1997: 153–171).

Although this view of the solstices and of the minimum standstills of the moon over the Appalachian range from Sugarloaf to Mount Logan is possible anywhere within the embankment walls at Mound City, it is most accurate when viewed from Mound 7. When Squier and Davis excavated Mound 7, they found a large deposit of round, 10to 12-in mica sheets which they interpreted as a segment of a mica crescent 20 ft long. They assumed the rest of the figure was still hidden in their unfinished excavation (Squier and Davis 1848: 154). In 1920 Mills’s exploration of the Mound City group reduced the 20-ft crescent to an 8-by-4-ft mica deposit “made to conform to the rounded contour covering of the base of the small mound covering burial number 9” (Mills 1922: 493). It appears from this later excavation that Squier and Davis were wrong about the crescent. But Mound 7 does appear to have a lunar context, both in terms of its location and aspects of its design.10

The monthly swing of the moon from Sugarloaf to Mount Logan during the season of minimum extremes would have called attention to the 57 lunar diameter difference between the minimum and maximum moonrise extremes. In this context, some features of the Shriver “Circle” a short distance south of Mound City deserve attention.

One astronomical feature of the Shriver “Circle” may indicate a common element in the designs of Shriver and Mound City. As documented in the Squier and Davis survey of Shriver, they described a large, irregular circle wall, bounded by an exterior ditch, with a mounded area near the “center” of the circle. Unlike many Early Woodland circles, but comparable to Hopewell circles (p.148)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.15. The Shriver “Circle.” From Pederson and Burks 2002: 15, (Figure 1). Courtesy of the Midwestern Archaeological Center and with the authors’ additions.

at Liberty, Seip, and Baum, Shriver has many openings, or “gates.” Like the large circle at Seip, there is a mound near its center. When one looks east from the central mound at Shriver, two of the gates frame the maximum moonrise extremes (zero-altitude, tangent) as seen over other Hopewell sites, Hopeton at the northeast, Works East at the southeast (Figure 5.15).11

Since Squier and Davis surveyed six openings in the Shriver wall, with an average width of about 10 m, this coincidence of two such openings with lunar extremes may well be accidental. Yet the orientation of the remaining four openings is also consistent with an intent to indicate other Hopewell or Early (p.149) Woodland sites and prominent features of the local terrain. One additional opening may have an astronomical orientation.

Reading clockwise from the north (Figure 5.15), the breaks in the Shriver embankment, viewed from the central mound, open a 2-degree window on (1) the Scioto River opposite the eastern opening or “gate” in the Mound City wall, (2) the north maximum moonrise extreme (52°.1) and the southwest corner of the Hopeton “Square,” (3) the highest point on Mount Logan, recorded by Squier and Davis (1848: 92, and Plate II) as the site of a mound which “afforded the most extended view that can be obtained in the entire region,” and (4) the south maximum moonrise extreme (129°.3) and the southwest quarter of the Works East Square. Neither gate 5 nor gate 6 is aligned with a recorded Hopewell site. Numerous Early Woodland mounds, including the eponymous Adena Mound, were located near Lake Ellensmere at the northeastern base of Mount Prospect (see next section). Gate 6 is aligned with the zero-altitude horizon bearing of the winter solstice sunset (239°.0), but the observer would see the sun set south of gate 6 because the local horizon is not zero but 2°.5.

Each of these bearings considered individually is inexact, based at present on the Squier and Davis survey alone. Aerial photography and recent geophysical survey have confirmed the Squier and Davis location of the southern segment of the Shriver wall and ditch (Pederson and Burks 2002). With geophysical sensing and excavation it may still be possible to ascertain the location of the two western gates at Shriver. These findings should be compared, where possible, with the locations of enclosure openings in the large “circles” at Baum, Seip, and Liberty (see Appendix 2 below).

Shriver and Mount Prospect

We have sought evidence to test the zero-altitude hypothesis (originally formulated at Newark) in the Chillicothe region by asking two questions. First, is there an elevation in the region that seems to be best suited for observing the rise/set of the sun and moon at zero-altitude? Second, does the distribution of Hopewellian earthworks in the region seem to have any astronomical relation to this point? We have found an affirmative answer to both questions. As always one has to confront the issue of whether the evidence for intentionality is more than one would expect by chance alone. Nevertheless, given the astronomical context already established we believe the evidence bears further examination.

Shriver is on the second terrace of the Scioto, in a position where fire beacons lit up and down the Teays Valley from Liberty in the south, to Dunlap, perhaps Blackwater, in the north would have been visible. Horizons to the east average 0°.5. But 1 km to the southwest, local horizon altitudes are higher. A (p.150) 275-meter-high plateau 1.5 km southwest of Shriver, for which we will revive an old Chillicothe name, Mount Prospect, blocks any view of sites on Paint Creek. Anderson and Hopewell are near but hidden. Shriver is not a promising site for zero-altitude views of moonrise or moonset.

In this context Mount Prospect is promising. It is not the tallest peak in the region. But it is the highest land close to the broad terraces of the Scioto River where Mound City, Shriver, and Hopeton were built early in the Middle Woodland period. It is also the best known today. Thomas Worthington, one of the Chillicothe founders and the sixth governor of Ohio, built his home there. He first named his estate Mount Prospect Hall for the view of the valley and the Appalachian range to the east that it offered. Later he changed its name to Adena, whence it lent that name to the Adena culture.12 The view of sunrise over Mount Logan across the Scioto was placed on the Great Seal of Ohio in 1803.

Worthington’s Mount Prospect commands a view of the Teays Valley north and south, and Shriver, Mound City, Hopeton, and Cedar Bank are easily seen from Mount Prospect. Beacons at Dunlap and Blackwater to the north would have been visible, as also at Works East, High Bank, and Liberty toward the south. Mount Prospect also overlooks the North Fork of Paint Creek, bringing Anderson and Hopewell into the range of visible sites, assuming the use of beacons and some extensive clearing. Only the sites on Paint Creek, and Frankfort, distant on the North Fork, would have been beyond direct view.

Mount Prospect and the Zero-Altitude Hypothesis

Sketches in Squier and Davis indicate mounds on Mount Prospect, but no Hopewell enclosure is known there (Squier and Davis 1848: Plate II). Nevertheless, the view from Mount Prospect may have played a role in the site planning for Shriver, perhaps also for Hopeton and other corporate centers. If one stands on Mount Prospect above Shriver and looks toward the northeast, the moon at its north maximum will rise near a line that passes through the “center” of the Shriver “Circle” and passes through Hopeton. There is no place closer than Mount Prospect from which this zero-altitude view of the north maximum moonrise over Shriver and Hopeton is possible.

As shown in (Figure 5.16), eight major Hopewell earthworks fall on bearings to lunar extremes when viewed from a single position on Mount Prospect: (1) the north maximum moonrise (52°.1) aligns with the center of Shriver and the southwest corner of Hopeton,13 (2) the south minimum moonrise bearing (114°.8) passes near the estimated center of the large circular enclosure at Works East, (3) the south maximum moonrise bearing (129°.3) passes through the southern extension of High Bank and touches the northeast corner of (p.151)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.16. Mount Prospect Plateau alignments. Courtesy © 2002 DeLorme (http://www.delorme.com) 3-D TopoQuadsGand with the authors’ additions. Key: Shriver-Hopeton: north maximum rise, 52°.1; Works East: south minimum rise, 114°.8; High Bank-Liberty: south maximum rise, 129°.3; Seip: south maximum set, 230°.8; Anderson-Frankfort: north minimum set, 293°.6.

the Liberty Square, (4) the south maximum moonset bearing (230°.7) passes through the southeast corner of the Seip Square, and (5) the north minimum moonset bearing (293°.7) passes within 10 m of the northeast corner of the Anderson “Square” and misses the northeast corner of the Frankfort Square by perhaps 100 m, an “error” of less than .05°.14 Zero-altitude horizons and a tangent moon at rise and set define all of these bearings.

It is difficult to estimate the probability that these alignments are accidental. The earthwork “targets” are large, ranging from 7 degrees at Works East, the site closest to Mount Prospect, to 1 degree at Frankfort, the most distant site. Defining the earthwork target is also problematic. One could choose among circle centers large and small, corners and centers of squares, and geometric centers of entire enclosures. Here we have considered alignment as a bearing passing through or within 1 degree of an enclosure. Some insight about the (p.152) probability of accidental alignment could be gained from a Monte Carlo study of the likelihood that randomly positioned high points would have alignment relations to the Hopewell sites comparable to those actually found at Mount Prospect.

If this position on Mount Prospect, which we will call Prospect Point, served as the common backsight for lunar extremes seen over Shriver and Hopeton, Works East, High Bank and Liberty, Seip, and Anderson and Frankfort, its role in the large design of the Chillicothe core resembles the roles of Coffman’s Knob and H1 in our earlier interpretation of the smaller design of works at Newark and Granville. In that respect the zero-altitude hypothesis framed at Newark gains support.

Yet there are significant differences. At Newark, bearings on lunar extremes from both Coffman’s Knob and H1 bisect major earthworks and the Observatory Circle-Octagon and the Great Circle, and pass through centers at the Salisbury Square and the elliptical enclosure north of the Salisbury Square. At Chillicothe there is no consistently accurate pattern. At Newark the aligned sites are visible from H1 and Coffman’s Knob. At Chillicothe there is no direct line of sight from Prospect Point to Seip or Frankfort, or even to Anderson and Hopewell, unless the observer leaves Prospect Point and moves toward the western edge of the plateau.

Prospect Point could have served as a position from which to lay out or coordinate roughly the locations of as many as seven Hopewell enclosures: Shriver, Hopeton, Liberty, Works East, Seip, Anderson, and Frankfort. Observations of zero-altitude lunar rise/set points may have played a role in the placement of these enclosures.

A Regional Plan?

The suggestion of a larger regional plan calls our attention back to earlier discussion of the array of sites along the Teays Valley north and south of High Bank—Works East, High Bank, and Liberty aligned along the 143°.4 bearing of the old valley. If Mount Prospect played the role in the layout of the Shriver-Mound City-Hopeton complex that we have begun to suspect, there is reason to believe it may have played a further role in the layout of the Teays Valley array. The long axis of the High Bank Circle-Octagon and the bearing of the valley determine the line on which Liberty and Works East fall. As viewed from Mount Prospect, the intersections of the azimuths of the Seip south minimum and maximum moonrise extremes with the Teays Valley-High Bank axis could have determined the places where Liberty and Works East would (p.153) be built ((Figure 5.15) and (Figure 5.16)). It remains for experimental archaeology to determine the feasibility of laying out such a design on such a scale.15

On the preglacial Paint Creek valley we have established another array of sites along both sides of the northeast-bearing valley. We have seen already that bearings on the north maximum moonrise and the north minimum moonrise connect Seip with Shriver-Hopeton, and with Works East. Now we discover that the 52°.2 north maximum moonrise bearing from Seip through Shriver and Hopeton passes over Prospect Point, drawing the Seip enclosure into a possible regional plan centered on Mount Prospect. Seip, as we noticed earlier, is the focus of lunar bearings on Shriver-Hopeton and Works East. It is also the origin, with Baum, of the double array of sites ranged in parallel along the bearing of the old Paint Creek valley (see (Figure 5.10) and (Figure 5.11)).

If we look at Mount Prospect as the possible center for the development of a larger regional plan, we can utilize the astronomical evidence we have gathered to frame testable hypotheses about stages in the discovery and implementation of such a plan. We must remind ourselves that even where the alignments we have described are solid evidence, the question remains whether the alignments add up to evidence of intent. On the assumption that the alignments are intentional, the following scenario offers a rough guide for such a plan:16

  1. 1. Lay out the Shriver-Mound City-Hopeton complex, with moonrise and sunrise events seen over the Appalachian range to the east as the focus of many aspects of earthwork location and orientation.

  2. 2. Build the High Bank Circle-Octagon, emulating the style of Hopeton. The symmetry axis of the Circle-Octagon determines the bearing on which the future sites of Liberty and Works East will fall ((Figure 5.8), (Figure 5.16), and (Figure 5.17)).

  3. 3. Build Liberty at the intersection of the Teays Valley-High Bank axis with the south maximum moonrise extreme as seen from Mount Prospect.

  4. 4. Build Works East at the intersection of the Teays Valley-High Bank axis with the south minimum moonrise extreme as seen from Mount Prospect.

Given the attention of the Hopewell builders to the long, linear Appalachian range east of the Scioto, and to the long linear valley of the Teays south of Mount Prospect, it is reasonable to believe they saw something special too about the long, linear valley of the preglacial Paint Creek. Perhaps its major attraction was that it led back to the origin of the regional array at Mount Prospect, Shriver, Mound City, Hopeton, and Cedar Bank (see (Figure 5.10)). (p.154)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.17. The Hopeton and High Bank Circle-Polygons (Hopeton: Squier and Davis 1848, Plate XVII, corrected to true north; High Bank: Squier and Davis 1848: Plate XVI).

The sites along the ancient course of Paint Creek could be drawn into the larger plan in the following manner:
  1. 5. Build Seip on the intersection of the south maximum moonset bearing from Mount Prospect and the south minimum moonset from Works East. This would be a challenging task, a “sightline” of nearly 26 km from Works East to Seip, crossing three intervening ridges with an average elevation gain of 115 m (see (Figure 5.18)).

If Anderson, Cedar Bank, and components of Hopeton were already in place, as some archaeological evidence suggests (for Cedar Bank and Hopeton see Byers, 2004: 538–539; for Anderson see Pickard and Pahdopony 1995), the next stage in the implementation of the plan may have been:

  1. 6. Build Baum to the east of Seip on the intersection of a north minimum moonrise bearing from Seip to Works East, and the 44°.3 bearing of the old Paint Creek valley. That bearing following the south side of the valley passes through Mound City and Cedar Bank. It will also pass through the present or future site of Spruce Hill (Figure 5.10). This is another challenging task, 27 km over four intervening ridges.

  2. (p.155)

    Hopewell Cosmography at Newark and Chillicothe, Ohio

    Figure 5.18. Plan maps of Liberty, Works East, Seip, Baum, and Frankfort. All at same scale (Liberty: Squier and Davis 1848: Plate XX, corrected to true north; Works East, Seip, Baum, and Frankfort: Squier and Davis 1848: Plate XX, nos. 3, 2, 1 and 4)

    (p.156) 7. Draw attention to the parallel bearing of sites on the north side of the valley by building a new enclosure at Dunlap which completes an array of sites that begins at Seip, passes through the present or future site of Bourneville, and passes through Anderson. This 27-km “sightline” will cross four intervening ridges.

  3. 8. Design Dunlap so that it is clear that it is meant to complete the regional plan. Include an Anderson-like circle on the north side of the Dunlap polygon (Figure 5.19). Add long parallels that both recall the parallel alignment beginning at Seip and Baum, and resemble the parallel walls that connect Hopeton with the Shriver-Mound City complex ((Figure 5.10) and (Figure 5.16)). Align the parallels with the bearing of the old Teays Valley, recalling the Liberty-High Bank-Works East axis, and the second stage in the development of the regional plan (Figure 5.8)

    Hopewell Cosmography at Newark and Chillicothe, Ohio

    Figure 5.19. Anderson site and Dunlap site (Anderson: Pickard and Pahdopony 1995: 4, (Figure 1), mapped by J. C. Anderson, 1979, courtesy of Midwest Archaeological Center; Dunlap: Squier and Davis 1848: Plate XXIII. no. 2).

    .

(p.157) If we return now to Mount Prospect and the radial array of sites aligned from Prospect Point to lunar extremes, only one further step is required to complete the plan:

  1. 9. Build Frankfort, extending the alignment on the north minimum moonset from Anderson. The addition of Frankfort completes the regional plan by incorporating sites on the North Fork of Paint Creek beyond Anderson, most important in this respect, Hopewell (Figure 5.16).17

If anything like what is suggested here is true of Mount Prospect, and of Prospect Point, the situation is quite different from what we have found at Newark. As already noted, alignments made from Prospect Point most often do not indicate symmetry axes or other geometrically determined features of enclosures. Yet Mount Prospect, with its zero-altitude vantage point over perhaps 260 degrees of the horizon, could have played a major, perhaps determining, role in the location of ceremonial architecture in Chillicothe Hopewell.

Arguments for planning on this scale, with the demands on the economy, social organization, information storage and retrieval, and investment of time and labor required, place the burden of proof on those who accept the alignments as deliberate rather than accidental. The choice of Mount Prospect as the focus of an array of alignments on lunar extremes marked by sites 26 km distant is problematic, especially since there is no known Hopewell site at Prospect Point. Yet individual components of the argument are more secure:

  1. 1. Alignments to two lunar and two solar extremes seen from Mound City over prominent points on the Logan Range.

  2. 2. Alignments to four solar and eight lunar extremes on the Circle-Octagon at High Bank.

  3. 3. Three sites (Works East-Liberty-High Bank) aligned on the 143°.4 bearing of the Teays Valley, orthogonal to the minor axis of the High Bank Octagon and to the 53°.4 azimuth of the north maximum moonrise.

  4. 4. Alignments from a single point at Seip to Shriver and Hopeton on the north maximum moonrise, and to Baum and Works East on the north minimum moonrise.

  5. 5. Eight Paint Creek valley sites (Baum-Spruce Hill-Mound City-Cedar Bank, Seip-Bourneville-Anderson-Dunlap) aligned along opposite sides of the valley on 44°.3.

None of these alignments involves Mount Prospect. Yet if we assume that the Hopewell builders did utilize Mount Prospect as a regional focus, further elements fall into place:

  1. (p.158) 1. Eight of the sites which fall on the above more secure alignments also fall on five lunar extremes as viewed from Prospect Point (Shriver-Hopeton, Works East, High Bank-Liberty, Seip, Anderson-Frankfort).

  2. 2. Liberty and Works East are located at the intersection of lunar rise extremes, as viewed from Prospect Point, with the Teays Valley 143°.4 axis.

  3. 3. Seip is located at the intersection of two lunar set extremes, as viewed from Prospect Point and from Works East, and the 44°.3 line which passes through Bourneville, Anderson, and Dunlap.

  4. 4. Baum is located on the intersection of a lunar rise extreme as viewed from Seip and the 44°.3 line which passes through Spruce Hill, Mound City, and Cedar Bank.

If we assume that Mount Prospect played a central role in Hopewell planning at Chillicothe, the suggested scenario accounts for these phenomena:

  1. 1. The location of the Shriver-Mound City-Hopeton complex, in terms of its broadly economic suitability and in terms of its situation relative to the Logan range and the visible extremes of the rising sun and moon.

  2. 2. The locations of Liberty and Works East along the Teays Valley, by their relation to the long axis of High Bank and their position on lunar extremes as viewed from Prospect Point.

  3. 3. The location of Seip relative to lunar extremes from Works East and Prospect Point.

  4. 4. The location of Baum relative to lunar extremes and the old Paint Creek valley topography (44°.3 parallels).

  5. 5. The locations of Anderson and Frankfort on a lunar extreme as seen from Mount Prospect.

Certainly alternative scenarios can be constructed, but any scenario that assumes that all or most of the recorded alignments are not accidental must explain how so many archaeologically related sites could have been placed on these lunar extremes in another order of construction and without adopting Mount Prospect as a focal location. We believe the logic of the argument requires serious consideration.

Conclusion

On the basis of our recent work at Newark and Chillicothe, these hypotheses are plausible and deserve continued testing:

  1. (p.159) 1. The five lunar alignments defined by walls of the Newark Octagon and the connecting avenue are an intended part of the Newark design.

  2. 2. The location and design of major features of many earthworks at Newark and Granville are coordinated with lunar extremes and local topography as viewed from vantage points which afford a zero-altitude view of the distant horizon.

  3. 3. The location and design of major features of earthworks in the Hopewell core at Chillicothe are coordinated with local topography and with lunar (and solar) extremes as viewed from vantage points that afford a zeroaltitude horizon.18

It remains for further research to discover whether these alignments, which cannot be tested with the rigor possible on the perhaps uniquely precise and complex Newark Circle-Octagon, are more plausibly regarded as accidental or deliberate. Field testing of the plausibility of astronomically oriented construction on the scale suggested here is required. As we have seen, the third hypothesis appears to be disconfirmed at High Bank. There is no evidence on the Circle-Octagon that the alignments to lunar and solar extremes are made from high zero-altitude horizon locations. All of the alignments are along local horizons.19 There is no unambiguous evidence for alignments on zero-altitude horizons in the layout of individual Hopewell enclosures in the Chillicothe region.

At Newark and Granville the evidence for zero-altitude observations from high elevations is much stronger. The large design of the Great Circle, the Observatory Circle-Octagon, and the Salisbury Square and Ellipse conforms to lunar extreme azimuths viewed from H1 and Coffman’s Knob. In some cases the local horizons at individual enclosures are too low to make it possible to decide whether the builders aimed at the zero-altitude moonrise or moonset or built so that the local observer would see the rise and sets defined by the geometry over nearby, local, non-zero-altitude horizons.

To this point we have found no site in the Chillicothe region that displays the accuracy achieved on the Newark Circle-Octagon. Consequently, we have no other site where the distinction between local and zero-altitude horizon alignments can be as readily decided. One is tempted to believe that the builders at High Bank decided to go to Newark and “do it right,” to build a circleoctagon which displayed the symmetry of the octagon and the 18.6-year cycle of the moon unambiguously on its walls and long axis.

We do not yet have a comprehensive view of Hopewell at Newark or at Chillicothe. We understand too little about the economy, social structure, and (p.160) social-cultural education among these Middle Woodland peoples to test a priori the plausibility of planning on the scale suspected here. At present we know too little about the relative chronology of works at Newark-Granville and in the Hopewell core at Chillicothe to do more than guess at the scale and scope of design or the pace and pattern of building.20

Our current hypothesis is that Hopewell planners understood not only the yearly cycle of the sun but also the longer 18.6-year cycle of the moon, and that they sought to express this understanding in the siting and design of corporate ceremonial architecture. Where the natural environment itself suggested links, for example, between the paths of rivers, valleys, mountain ranges, and the long journey of the moon, they sought to build on that suggestion. They linked their corporate centers to salient features of the land, to one another, and in some cases to the heavens. Something like this cosmographic vision seems to us to be required in order to account for the ingenuity, the meticulous craft, and the labor invested in this remarkable architecture.

Appendix 1 : Motivations

We generally believe that there is considerable virtue in not speculating about the motives of a prehistoric culture for which there is no unambiguous evidence. Nevertheless, one must also recognize that some skepticism about the inherent plausibility of Hopewell astronomy arises because possible motives are not widely understood. As a result, we will include here a brief description of possible motivations.

First, it should be remembered that observing and understanding the moon almost certainly was a high priority for any society advanced enough to be interested in geometrical symmetry and regularity as well as the practical necessity of understanding seasonal and repetitive natural phenomena relevant to survival. Observing and understanding the moon would have been of high cultural priority for at least four broad reasons: (1) the moon would have appeared to be mysterious and powerful and to embody the fundamental rhythms of the natural environment through its predictable changes in position, shape, and brightness; (2) the cyclical change in the extreme rise/ set points of the moon would be easily noticed in a context where the local horizons provided many reference points (valleys and hills) which made the changes obvious, (3) at the latitude of Newark the amount of moonlight at north lunar maximum (15.25 hours) was much greater than that at south lunar maximum (8.75 hours), a striking phenomenon calling even more attention to the monthly north-south swing of the moon; and (4) the cycles of the moon (the month and 18.6 years) provide convenient calendrical time scales (p.161) for organizing human activity. The 18.6-year cycle roughly equals one human generation.

The impressive scale of the earthworks (especially the Newark Circle-Octagon earthworks, which enclose more than 60 acres) is difficult to understand as being required for any obvious utilitarian purpose. Indeed, the shape and precision cannot be clearly seen by any human observer on the ground. It can only be seen and appreciated from a vantage point high in the sky overlooking the earthworks, precisely the vantage point of the powerful and ever mysterious moon. The large scale of the earthworks also gives the encoded alignments an accuracy and permanence that has survived almost 2,000 years.

The geometrical earthworks themselves constitute undeniable evidence of a cultural passion and fascination for geometrical experimentation and construction on a large scale. The recognition that symmetrical geometrical structures could also encode an understanding of the moon’s motions would have been a marvelous and powerful but not implausible discovery. Even today when one stands inside the Newark Circle-Octagon earthworks (especially under the light of a low altitude full moon when the octagon walls cast dark bold shadows across the softly illuminated interior) one cannot help but be stirred by the stunning expenditure of physical, cultural, and intellectual energy involved in their construction. They stand today as impressive testimony to what a culture can achieve when motivated by ideas with sufficient power to inspire its collective imagination, intellect, and courage.

Appendix 2: The Baum Circle Segment

Perhaps, with the aid of aerial photography, geophysical sensing, and excavation, the plausibility of purported alignments at Shriver can be tested at the Baum, Seip, and Liberty circles. Here, tentative results are cited for the large circular component of the Baum Circle, based on Squier and Davis as well as Middleton’s resurvey of the site (Thomas 1894: 483).21 Although the Baum Circle has no central mound, the geometrical center of the Baum site is at the northeast corner of the square. The wall of the large circular segment, extended southwest toward the smaller circle (Figure 5.20), is broken by seven openings that average 10 m in width. Reading clockwise from north, a bearing from the geometrical center of Baum through these openings aligns to the following locations within 1 degree: (1) the entrance to the Bourneville oval, 1°.1; (2) the southwest corner of the Hopewell Square, 26°.2; (3) the southern slope of Spruce Hill, Shriver, and the northwest quarter of the Hopeton Polygon, 46°.8; (4) the point (13) on the High Bank Circle-Octagon wall opposite the avenue, 76°.3; (5) the southern quarter of the large circular component at (p.162)

Hopewell Cosmography at Newark and Chillicothe, Ohio

Figure 5.20. Baum: Rotation of enclosure altered to match Middleton’s resurvey (Squier and Davis 1848: Plate XXI, no. 2).

Liberty, 91°.0; (6) the western slope of Beath Ridge 6.5 km northwest of Baum, 307°.8, and (7) a narrow and prominent peak on Barger Ridge 6 km northwest of Baum, 329°.8. The 307°.8 bearing through enclosure “gate” 6 also coincides with the tangent zero-altitude azimuth of the north maximum moonset.

Seen from the geometrical center of the Baum enclosure, five of the seven openings in the circle walls at Baum align with other Hopewell or Adenan sites. Two align with nearby features of the Appalachian ridge that flanks the northern side of the old Paint Creek valley. In contrast to Shriver, only one of these bearings through gaps in the circle walls at Baum is aligned with a lunar or solar extreme.22 Beath Ridge and Barger Ridge stand out in the local terrain less than many other peaks and valleys.

Notes

(1) Unlike the sun, which travels each year between the same rise and set extremes defined by the summer and winter solstices (standstills), the rising moon makes the journey from its northeastern extreme to its southeastern extreme each month. The (p.163) lunar extremes themselves change slowly. For 9.3 years they move gradually from a narrow range on the horizon (minimum extremes) to a broad range (maximum extremes). The entire cycle, from maximum to maximum or minimum to minimum, repeats every 18.6 years. The local range of these extremes, like those of the sun, is determined by latitude. Between Newark and Chillicothe the range of difference is about 1°.

(2) James and Charles Salisbury first mapped this square, to the north of the confluence of Raccoon Creek and the south fork of the Licking River, in 1862. Neither the Salisbury Square nor the ellipse to the north of the confluence has survived (Salisbury and Salisbury 1862). According to the Salisbury survey, the walls of the “square” are not parallel. The northeast and southwest sides of the square differ by 4°; the northern and southern sides by 2°. The average of the azimuths of the northeast and southwest sides of the figure is 307°, within 1.3° of the north maximum moonset extreme as seen from Coffman’s Knob.

(3) Since the acceptance of this chapter for publication, we have found that the pattern described here for H1 and H3 is repeated at two comparable hill positions (H2 and H4) from which the north and south maximum and minimum moonsets are seen along octagon walls EF and BC, through the Great Circle or the Wright Square. Further, we were surprised to find that (1) the bearing between H1 and H4 aligns with the summer solstice sunrise and winter solstice sunset and (2) the bearing between H2 and H3 aligns with the winter solstice sunrise and the summer solstice sunset. It now seems plausible that the monumental, complex earthworks were planned using this solstitial “template” and built on the valley floor below the surrounding observation points to incorporate lines to the extreme rise and set points of the moon in their geometrical structure (Hively and Horn 2008).

(4) Greber’s recent excavation at the High Bank Circle indicates that the circle wall near Middleton’s survey station 13 (which corresponds to the location of the “Observatory Mound” at Newark) was a focus of activity at High Bank. Plausible evidence of more than 200 posts was found in a 2 m × 18 m excavated trench (Greber 2002: 2).

(5) This is the best current estimate of the location of the Works East enclosure (N'omi B. Greber personal communication, August 2004).

(6) Our work at High Bank since this paper was accepted for publication lends some support to the conjecture that high zero-altitude vantage points did play a role in the layout of the High Bank Works, and other Chillicothe sites. This will be discussed in forthcoming papers.

(7) See Waldron and Abrams (1999: 97–111) for a discussion of the intervisibility of Adena hamlets and mounds in the Hocking River valley. Few of the enclosures along the old Paint Creek valley would have been intervisible. A plausible test of the intentionality of these alignments would be to look for evidence of Early and Middle Woodland activity on the ridges along the both sides of the valley which these 44.3° lines cross.

(8) Jester Hill (390 m) lies on the same bearing as Seip and Hopeton 4.6 km southwest of the Seip enclosure, affording a view of the valley locations of Seip and Baum.

(p.164) (9) It should be noted that the peaks of Sugarloaf and Logan necessarily would have been marked with fires, as the profile of the range would seldom have been visible at moonrise.

(10) We had completed this research before we read Byers’s account of possible lunar symbolism at Mound 7. His interpretation of Mills’s account of the content of Mound 7 converges with our own findings regarding its lunar (and solar) context (Byers 2004: 354–358, and chap. 15). Romain’s discussion (Romain 2000: 152–157) of the orientation of the floor plans of some Mound City charnel houses is germane, as is his study of the diagonal solstice orientations of Mound City and other “squares” among Chillicothe region Hopewell enclosures (Romain 2004a, 2004b, 2004c).

(11) The Shriver enclosure may be deliberately oriented to the Logan Range, as the figure is oriented toward the peak of Bunker Hill, the most distinct prominence on the Appalachian profile visible from the Shriver location.

(12) National Historic Landmark Nomination, Adena (Thomas Worthington House), United States Department of Interior, National Park Service, 2003.

(13) The line on the same bearing can be extended backward to its beginning at the Seip enclosure, a distance of 26 km.

(14) The exact location and orientation of the Frankfort Square is uncertain.

(15) Lekson (1999: 117–118) contributes to early stages of a discussion of these problems.

(16) The scenario draws on both Greber’s essay on enclosures (Greber 2006) and Byers’s Hopewell Episode (Byers 2004). Neither is to be blamed for the obviously precipitate suggestions we make here. Many, perhaps most, of the Chillicothe sites are accretional. Any serious effort to work out a chronological sequence for specific features of individual sites will have to deal with this problem in detail. See Greber 2006.

(17) Ascertaining the place of Hopewell among the Chillicothe sites is a more complex issue than we can discuss here.

(18) As indicated in note 3 above, our recent work finds that the geometric earthworks in Newark’s Cherry Valley may have been planned and laid out from four sites in the surrounding hills that together form a pattern aligned to the summer and winter solstices. If so, both Newark and High Bank are lunar and solar sites.

(19) See note 6 above.

(20) Current evidence is reported and reviewed in Byers 2004.

(21) Thomas says that the resurvey of the circular portion of Baum agreed essentially with the Squier and Davis figure given in Ancient Monuments, Plate XXI, no. 1. Yet he continues, “The circular portions of the work are considerably worn and two sections of considerable length are so nearly obliterated that the line cannot be retraced with any certainty.”

(22) In 1990 Greber and Linsay, utilizing Middleton’s survey data and their own examination of the topography at Baum, found local horizon altitude alignments with the north maximum moonrise and moonset among the openings or gates in the square. They found no solar azimuths aligned on “gates” (N’omi B. Greber personal communication, 1991). Allowing for the local horizon altitude and lower limb tangency, two sides of the Baum Square are aligned with the winter solstice sunset.

Notes:

(1) Unlike the sun, which travels each year between the same rise and set extremes defined by the summer and winter solstices (standstills), the rising moon makes the journey from its northeastern extreme to its southeastern extreme each month. The (p.163) lunar extremes themselves change slowly. For 9.3 years they move gradually from a narrow range on the horizon (minimum extremes) to a broad range (maximum extremes). The entire cycle, from maximum to maximum or minimum to minimum, repeats every 18.6 years. The local range of these extremes, like those of the sun, is determined by latitude. Between Newark and Chillicothe the range of difference is about 1°.

(2) James and Charles Salisbury first mapped this square, to the north of the confluence of Raccoon Creek and the south fork of the Licking River, in 1862. Neither the Salisbury Square nor the ellipse to the north of the confluence has survived (Salisbury and Salisbury 1862). According to the Salisbury survey, the walls of the “square” are not parallel. The northeast and southwest sides of the square differ by 4°; the northern and southern sides by 2°. The average of the azimuths of the northeast and southwest sides of the figure is 307°, within 1.3° of the north maximum moonset extreme as seen from Coffman’s Knob.

(3) Since the acceptance of this chapter for publication, we have found that the pattern described here for H1 and H3 is repeated at two comparable hill positions (H2 and H4) from which the north and south maximum and minimum moonsets are seen along octagon walls EF and BC, through the Great Circle or the Wright Square. Further, we were surprised to find that (1) the bearing between H1 and H4 aligns with the summer solstice sunrise and winter solstice sunset and (2) the bearing between H2 and H3 aligns with the winter solstice sunrise and the summer solstice sunset. It now seems plausible that the monumental, complex earthworks were planned using this solstitial “template” and built on the valley floor below the surrounding observation points to incorporate lines to the extreme rise and set points of the moon in their geometrical structure (Hively and Horn 2008).

(4) Greber’s recent excavation at the High Bank Circle indicates that the circle wall near Middleton’s survey station 13 (which corresponds to the location of the “Observatory Mound” at Newark) was a focus of activity at High Bank. Plausible evidence of more than 200 posts was found in a 2 m × 18 m excavated trench (Greber 2002: 2).

(5) This is the best current estimate of the location of the Works East enclosure (N'omi B. Greber personal communication, August 2004).

(6) Our work at High Bank since this paper was accepted for publication lends some support to the conjecture that high zero-altitude vantage points did play a role in the layout of the High Bank Works, and other Chillicothe sites. This will be discussed in forthcoming papers.

(7) See Waldron and Abrams (1999: 97–111) for a discussion of the intervisibility of Adena hamlets and mounds in the Hocking River valley. Few of the enclosures along the old Paint Creek valley would have been intervisible. A plausible test of the intentionality of these alignments would be to look for evidence of Early and Middle Woodland activity on the ridges along the both sides of the valley which these 44.3° lines cross.

(8) Jester Hill (390 m) lies on the same bearing as Seip and Hopeton 4.6 km southwest of the Seip enclosure, affording a view of the valley locations of Seip and Baum.

(p.164) (9) It should be noted that the peaks of Sugarloaf and Logan necessarily would have been marked with fires, as the profile of the range would seldom have been visible at moonrise.

(10) We had completed this research before we read Byers’s account of possible lunar symbolism at Mound 7. His interpretation of Mills’s account of the content of Mound 7 converges with our own findings regarding its lunar (and solar) context (Byers 2004: 354–358, and chap. 15). Romain’s discussion (Romain 2000: 152–157) of the orientation of the floor plans of some Mound City charnel houses is germane, as is his study of the diagonal solstice orientations of Mound City and other “squares” among Chillicothe region Hopewell enclosures (Romain 2004a, 2004b, 2004c).

(11) The Shriver enclosure may be deliberately oriented to the Logan Range, as the figure is oriented toward the peak of Bunker Hill, the most distinct prominence on the Appalachian profile visible from the Shriver location.

(12) National Historic Landmark Nomination, Adena (Thomas Worthington House), United States Department of Interior, National Park Service, 2003.

(13) The line on the same bearing can be extended backward to its beginning at the Seip enclosure, a distance of 26 km.

(14) The exact location and orientation of the Frankfort Square is uncertain.

(15) Lekson (1999: 117–118) contributes to early stages of a discussion of these problems.

(16) The scenario draws on both Greber’s essay on enclosures (Greber 2006) and Byers’s Hopewell Episode (Byers 2004). Neither is to be blamed for the obviously precipitate suggestions we make here. Many, perhaps most, of the Chillicothe sites are accretional. Any serious effort to work out a chronological sequence for specific features of individual sites will have to deal with this problem in detail. See Greber 2006.

(17) Ascertaining the place of Hopewell among the Chillicothe sites is a more complex issue than we can discuss here.

(18) As indicated in note 3 above, our recent work finds that the geometric earthworks in Newark’s Cherry Valley may have been planned and laid out from four sites in the surrounding hills that together form a pattern aligned to the summer and winter solstices. If so, both Newark and High Bank are lunar and solar sites.

(19) See note 6 above.

(20) Current evidence is reported and reviewed in Byers 2004.

(21) Thomas says that the resurvey of the circular portion of Baum agreed essentially with the Squier and Davis figure given in Ancient Monuments, Plate XXI, no. 1. Yet he continues, “The circular portions of the work are considerably worn and two sections of considerable length are so nearly obliterated that the line cannot be retraced with any certainty.”

(22) In 1990 Greber and Linsay, utilizing Middleton’s survey data and their own examination of the topography at Baum, found local horizon altitude alignments with the north maximum moonrise and moonset among the openings or gates in the square. They found no solar azimuths aligned on “gates” (N’omi B. Greber personal communication, 1991). Allowing for the local horizon altitude and lower limb tangency, two sides of the Baum Square are aligned with the winter solstice sunset.