megalithic plan underlying Canterbury Cathedral

darla. Megalithic Plan Underlying Canterbury Cathedral Author(s): Lyle B. Borst Source: Science, New Series, Vol. 163, No. 3867 (Feb. 7, 1969), pp. 567-570 Published by: American Association for the Advancement of Science Stable URL: http://www .jstor.org/stable/1726286 Accessed: 07/08/2011 11:02 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at htip:/www._jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science. http:/Avww.jstor.org

sediments in this core do not record Megalithic Plan Underlying Canterbury Cathedral continuous deposition. Core RC 11-256 the Abstract. Woodhenge and the Trinity chapel, Canterbury, are strikingly similar upper boundary of the event. Therefore, in outline. One is megalithic, the other Norman Christian over Saxon Christian. a rate of deposition based on the sedi- An analysis of the geometry shows that both are based on Pythagorean triangles: contains an apparent hiatus at ments of the X zone in cores RC 11-248 Wocdhenge with sides, 6, and RC 11-256 is meaningless, because in megalithic yards. The structurally more recent eastern end of Canterbury Cathea complete record of the event is not dral may have been built over and around an older megalithic site. The longipresent in either case. 17.5, and 18.5, and Canterbury with sides 12, 72, and 73 tudinal axes of the composite cathedral differ by 2°, and these, if aligned on The lower boundary of the X zone in RC 7-4 is in flow-in, and the sedimen- Betelgeuse, would indicate buried megalithic structures dating from 2300, 1900, and 1500 B.C, tation rate for this core is not reliable. The core RC 7-2 was taken from the The visitor to Canterbury Cathedral ter a; (11) the arcs at the small (minor) Blake Outer Ridge in a zone of interwho is fortunate enough to fall in with end have a common center f; (111) the mittent nondeposition or mild erosion, the retired clergyman, serving as volundistance between the centers aß is 6 or both (7) which tends to preclude a teer guide, whose hobby and pride are megalithic yards of 2.72 feet or 0.829 dependable sedimentation rate. the history, m; (iv) each minor radius is 1 mega- In contrast to the cores mentioned clerical and architectural, of the great edifice, will have his atlithic yard smaller than the major raabove, V 20-174, A 179-4, and RC 10- tention drawn to dius; 49 show no lithic evidence of breaks in approximately 2°, sedimentation and appear to represent choir. He will see a further deviation continuous of 2° to the south between the choir deposition in a relatively stable environment. Since the dates asthe misalignment, between and the Trinity chapel nave (Fig. and 1). The (v) the arcs connecting major and minor ends have a common center at y. | Figure 2c shows the triangle «By on one side of the axis only. Circular arcs | signed to the boundaries of the X zone chapel is apse-shaped with “horseshoe” from a and £ differing in radius by 1 are based on radiometric measurements rather than parallel sides. At the far megalithic yard are connected by an arc from y to form the oval. Arcs from of consider end, a small circular chapel, the princores RC 10-49 and A 179-4 the most Caribbean cipal apse of the cathedral, named the a corresponding point on the other side reliable and Corona, or Becket’s Crown, lies on the of the axis duration of the magnetic event. On the axis of the Trinity chapel. St. Andrew’s plete the oval. The point y must be 1 basis of the rate of sedimentation within the X zone of these for boundaries placed at sediments, estimating of the we the age would comchapel on the north and St. Anselm’s megalithic yard two cores, the on the south are symmetrically spaced from a in order to connect the two Blake event are with respect to the Trinity chapel but ends. If the triangle is a right triangle, not with respect to the choir. the hypotenuse will exceed one side by 108,000 and 114,000 years ago + 10 percent. Woodhenge | (not shown) is a megalithic monufarther from ß than one unit. JERRY D. SMITH ment 3 km from Stonehenge. It is as- The marvel of Woodhenge is that the JOHN H. FOSTER sumed to be of the same age as the triangle «By has the proportions 12, 35, Lamont Geological Observatory, early Stonehenge phase of construction and 37. This makes a perfect Pythag- Columbia (1800 B.C.). It consists of 160 postorean triangle, that is, a right triangle holes arranged in six concentric symhaving integral sides so that the sum University, Palisades, New York 10964 References and Notes pur . N. Bon Hommet and J. Babkine, Comptes Rend. 264, 93 (1967). . A. Cox, J. Geophys. Res. 73, 3247 (1968). X(©AW0NIa . J. H. Foster, Earth Planet. Sci, Lett. 6, 463 (1966). . N. D. Opdyke, B. Glass, J. D. Hays, J. H. metrical figures, and is surrounded by of the square of the two sides equals ditch and outer bank (7). Thom (2) the square of the hypotenuse, 122 + made a precise survey of the site (Fig. 35? 2c) and finds that (i) the arcs at the yards, the major and minor radii would large (major) end have a common cendiffer by 2 megalithic yards. If, how- . = 372. If laid out in megalithic Foster, Science 154, 349 (1966). . D. B. Ericson, M. Ewing, G. Wollin, B. C. Heezen, Geol. Soc. Amer. Bull. 72, 193 (1961). A. Cox, R. R. Doell, G. Dalrymple, Nature 198, 1049 (1963). B. C. Heezen, C. D. Hollister, W. F. Ruddiman, Science 152, 502 (1966). F. L. Parker, Bull. Amer. Paleontol. 52 (235), 182 (1967). W. S. Broecker, D. L. Thurber, J. Goddard, .T. Ku, R. K. Matthews, K. J. Mesolella, Science 159, 297 (1968). fn = . Lamont Geological Observatory Contribution No. 1295. Based on material provided by the . Lamont Geological Observatory core library. Supported by NSF grant GA 824, NSF 1193 and ONR grant TO-4. J.D.S. held a Duke University traineeship under the Cooperative Research and Training Program in . Biological Oceanography. We thank A. Kaneps, who carried out the paleontological analysis of the cores, helped describe the lithological stratigraphy of the cores, and contributed to the preparation of the manuscript; J. D. Hays, B. C. Heezen, and N. D. Opdyke for advice and criticism; and D. R. Horn, M. Delach, W. F. Ruddiman, J. V. Gardner, W. L. Prell, and W. Ryan for their support. 30 August 1968; revised 7 FEBRUARY 1969 15 October

ever, the half megalithic yard was used, of William the Englishman (Fig. 2, a small ends, and the major radii are 6 making a triangle 6, 17.5, and 18.5 megand b). The distance between centers a and 12 megalithic yards. The circular alithic yards, the relation of the sides and ß is 2 rods but, within the precision cavity below the Corona (Fig. would continue to be 12, 35, and 37, but of the available plans, this is 12 megais 1, D) 5 megalithic yards in radius and is the major radii would exceed the minor lithic yards. The Woodhenge-type ovals tangent to the larger oval. The smaller by 1 megalithic yard, and the distance of Fig. 2b are derived from triangles oval between centers af would be 6 megawith sides of 12, 72, and 73 megalithic centers lithic yards, as found by Thom. There yards. Although these are not mathe- 2a, E), beneath St. Anselm’s, and the is no matically symmetrical chapel of the Holy Innodoubt that Pythagoras’ theorem true Pythagorean triangles is tangent of St. to the line Gabriel’s between chapel (Fig. was being applied, although not proven (122 + 72? = 5328, while 732 = 5329), cents (Fig. 2a, F) beneath St. Andrew’s. or the deviation is 1 in 5000 and would The centers of the two chapels and the escape detection. The large ends of the center of curvature of the large ends Woodhenge megaa form three corners of a square. The lithic yard larger in radius than the double columns of the crypt, and those understood, 1000 years before Pythagoras. Now let us analyze the crypt below the Trinity chapel, known as the crypt. ovals are again 1 of Trinity chapel above, are oriented perpendicularly to the smaller Woodhenge oval and are therefore not perpendicular to the center line. The two - columns p,p of Fig. 2, a and b, fall on the larger Woodhenge oval. Although the circle of the Corona is tangent to the circular wall of the Trinity chapel, in the crypt the horseshoe is slightly deformed from a true circle for structural reasons. The centers of the three small apsidal chapels form an isosceles triangle of base 34 megalithic yards and altitude 34 megalithic yards. This triangle can be divided into two right triangles with a med Ä WILLLAM THE ENGLISHMANS common side, each approximately Pythagorean: 172 + 342 = 1445; 382 = 1444. The elegance of the geometrical design shows a carefully conceived and executed plan. Little is known of the Anglo-Saxon church which occupied the site before the Norman Conquest. Bede (3) informed that it had been was built by Roman Christians. However, it burned in A.D. 1067 and was entirely removed by Lanfranc before the construction of the Norman cathedral. The original Norman nave and transepts were finished in A.D. 1077 (4). The undercroft, choir, St. Anselm’s and St. Andrew’s chapels, initiated Ernulf about A.D. by Prior 1100, were dedicated by Prior Conrad in 1126. After the assassination of Archbishop Thomas Becket in 1170, the body of the martyr was enshrined “behind the altar of Our Lady Undercroft” (5). The choir was gutted by fire in 1174, but was immediately rebuilt. Between 1180-81, the cathedral was extended into the monk’s cemetery by William the Englishman to form the crypt, Trinity chapel, and the ment FE 9 10 20 s0,,,, Corona. At the same time, the two columns p,p of Fig. 2a were added to Ernulf’s crypt. Fig. 2. Comparison of eastern end of Canterbury Cathedral with Woodhenge. (a) The crypt. (4) Altar of Our Lady Undercroft; (B) Ernulf’s crypt; (C) William the Englishman’s crypt; (D) small chapel under Corona; (E) chapel of St. Gabriel; (F) chapel of the Holy Innocents; p,p columns added by William the Englishman. (b) Geometry the horizon at which the sun rises at of the crypt of William the Englishman. (c) Geometry of Woodhenge. the summer solstice. This rising point 568 The alignment at Woodhenge at Stonehenge is toward the point on SCIENCE, VOL. 163

has been constant within 0.5° during of the the past 4000 years. Other solar system then have been: (1) a Woodhenge type plans under Canterbury must In alignments such as the equinoxes and structure of wood or stone with circles was built upon the old crypt, and the the winter solstice are equally fixed. In at the positions of the three chapels; present Trinity chapel and Corona reother megalithic monuments, however, (ii) placed a small rectangular crypt and 400 years later a structure with the walls but did not damage the crypt. the reconstruction, the new choir the axis may be aligned upon a point apsidal end added west of the existing chapel. Nothing was altered in Ernulf’s on be structure but aligned with the contem- | crypt except for the addition of the two assigned to sun or moon. Many circles porary rising point of Betelgeuse; (iii) columns p,p (Fig. 2a). These columns, are aligned upon the rising and setting after still another 400 years, a second the positions of the two side chapels points of the star Capella, a smaller addition added west of existing strucand the asymmetry of the columns are number upon precession the horizon too far north Deneb. The to of the equinoxes causes the rising and tures, again aligned with the contemporary rising point of Betelgeuse. Of strong evidence Ernulf’s crypt that had construction been adapted of in setting points to migrate so that such the second addition, nothing remains, 1100 B.C. to accommodate venerated monuments can be dated. Ten monufor the Anglo-Saxon plans, positions or structures to the east ments a church were removed before erecting which were part of the monks’ ceme- At Canterbury, the Norman nave. Only the alignment tery. of the Trinity chapel made use of these analigned date near on Capella 1800 B.C. indicate the axes are oriented slightly south of east for nave, choir, all vestiges of Anglo-Saxon church was preand the Trinity served. When the choir was added, it chapel. The declination of the star may was placed upon the earlier plan (ii). be calculated from the direction of the In this addition, the chapels of The later construction of the cient positions and relations. © The available surveys of Ernulf’s St. crypt show no evidence of a horseshoe incorplan; columns and walls are shown as horizon point, the height of the horizon Anselm and point above the true horizontal, and the porated even though they latitude of the geographical location. symmetrically located. After When so determined, the axis of the assassination, when Trinity chapel points toward the rising planned for the display of his sacrophlithic point of a star of declination —6°; the agus, eastmegalithic yards near the transept. If axis of the choir, —4°; and the nave, ward into the monks’ cemetery. Again, one attributes the lack of parallel walls —2°. The change of old again to a Woodhenge oval, the disstars near the the [perhaps of declination equinoxes, where rate is greatest, is 42° per century. The interval between parts of the years, so that planning cathedral 800 adjacent must 400 Andrew were church foundations a was or were Becket’s chapel extended an old not was geometry preserved by postholes, (i)] were used. | Corroborative evidence for this proparallel. However, in the choir above, the walls are curved and not quite parallel, being separated by 32 megayards near the altar, tance between the centers and 30 of the circular ends would be about 50 megalithic yards. The new Pythagorean posal is obtained from an analysis of triangle required to generate a Woodhenge oval in which major radii exceed must have the side chapels, the two columns deselapsed between nave and the Trinity ignated p,p, and the arrangement of minor radii by 1 megalithic yard would chapel. The known construction dates columns in Ernulfs crypt (Fig. 2a). be 50, (1067 to do not provide suffi- Literary evidence leaves no doubt that would certainly represent a more ad- 1180) years be the St. 1250, and 1251. Such a plan cient time to account for the observed the chapels of St. Anselm and St. Anvanced culture than that responsible for deviations of axes. | Stars near the vernal equinox drift drew were part of the choir construction form Woodhenge or the Trinity chapel, and hence a later date. The best measurenorth; those near the autumnal equinox Pythagorean triangles with the Corona, ments which can be made upon availdrift south. and If the sequence of planof A.D. so were 1100. Yet related to they the Corona. able plans indicate a center of curvature ning were nave, choir, Trinity chapel, Their incorporation into the choir was 700 megalithic yards the axes south and the then a matter of convenience. The colmeasurements at Canterbury would be star would be near the autumnal equinox. No first magnitude star provides umns p,p Of Fig. 2a serve no structural required to confirm this interpretation. plausible sethe Trinity chapel construction. Their chapel, choir, positions on the larger Woodhenge oval fore appear to conform to the axis of would move are evidence that their locations were the north and the star would be near the part Ernulf's vernal equinox. Among first magnitude Trinity quence nave, would move dates. were the If, however, Trinity rising point the purpose. They were added as part of of the henge plan 7th-century crypt Anglo-Saxon when added of therechurch. in A.D. stars, Betelgeuse (in Orion) is the obtherefore preserve the location of pretion, yet its structure was asymmetric vious vious its declination standing columns 1077 nave at Canterbury would 1100 followed an old plan or foundabecause The construction in A.D. Further must choice chapel. used for the The the distant. stones or postholes. and provides evidence of a yet older columns forming plan east of the cathedral in the monks’ was —6° in 2300 B.C.; —4° in 1900 Examination of the B.C.; and —2° in 1500 B.C.—quite in the apsidal end of Ernulf’s crypt discemetery. When the Trinity chapel and accord with dates at Arminghall (6) Closes that they are not symmetrically the Corona were constructed in A.D. (carbon dated 2300 B.C.); Woodhenge, located with respect to the center line. 1180, 1800 B.C.; and Stonehenge 1800-1500 Those to the south have smaller spac- The result was the horseshoe shaped this older plan was followed. B.C. The dates for Canterbury are aping than those to the north, so that the Trinity chapel with a geometry similar proximate and are subject to correction center line grazes one of the columns. to based upon a more precise survey of the site. The axis of the Trinity chapel, Corona, aligned with the rising point of Beteland geuse in 2300 B.C. Anglo-Saxon built one after churches were another, often end-to-end along an axis (7). Perhaps this was a pre-Christian tradition. 7 FEBRUARY 1969 The sequence the two side chapels, however, Woodhenge, passes symmetrically between the two columns. The fire which was originally LYLE B. Borst Department of Physics and Astronomy, of A.D. 1170, which destroyed the roof of the choir, weakened State University of New York at Buffalo Buffalo 14214

References 1. M. E. Cunnington, Woodhenge (Simpson, Devizes, England, 1929), 2. A. Thom, Megalithic Sites in Britain (Clarendon Press, 3. Bede, Oxford, 1967), Ecclesiastical quoted in R. Canterbury Willis, p. 73. History, vol. 1, p. 33, Architectural History of Cathedral (Longman, London, 4. R. Willis, Architectural History of Canterbury Cathedral (Longman, London, 1845). 5. Gervase, Decem Scriptores, quoted in R. Willis, Architectural History of Canterbury Cathedral (Longman, London, 1845). 6. J. G. D. Clark, Proc. Prehist. Soc, 2, 1 (1936). 7. H. M. Taylor and J. Taylor, Anglo-Saxon Architecture (University Press, Cambridge, England, 1965). 26 September 1968; revised 12 December 1968 1845). y about 2500 to 3000 million years; that is, the half-mass age of all sedimentary rocks is about 600 million years. However, the half-mass ages of the various components of the sedimentary lithosphere appear to be different; that of carbonate rocks is about 300 to 400 million years and that of evaporites perhaps 200 to 300 million years (Fig. 1) (2). | Sedimentary Rock Types: Relative Proportions as a The near absence of evaporite deposits in Precambrian rocks may then be largely attributable to a rapid turn- Function of Geological Time Abstract. Proportions of sedimentary rock types remaining today differ from period to period. These differences may be chiefly the result of differential rates of deposition and erosion of the various components of the rocks. Lower percentages of limestones and evaporites in Precambrian rocks than in post-Precambrian rocks probably represent selective loss of these more easily removable components from the original deposits. The proportions of sedimentary rock that time. We propose that age differtypes in the geologic column vary as a ences in ratios of rock types may be function of age; for example, evaporites largely due to differential rates of overamount to several percent of post-Preturn of sedimentary materials in which cambrian the they sedimentary are far less rocks than 1 whereas percent various components of the sediof mentary rock mass circulate at mark- Precambrian rocks of sedimentary oriedly different rates controlled by their gin. It is often assumed that the ratios erodibility. of sedimentary rock types of a particu- Sediments have been continuously destroyed throughout lar age represent the relative propor-| deposited tions geologic time. The rates of deposition of sediment types deposited at and and destruction certainly have not been constant, as evidenced today by the irregular distribution of sediment mass as a function of age; however, one can construct highly simplified models in an = . +160 Half-mass age Ame 1.4 x 10°years B=—-0,6x 10° years attempt to simulate the gross aspects of today’s mass distribution. I mass that is constant with time, unyRieotmasc/srk1s08 constant and equal rates of deposition and destruction, and an equal probability of destruction of equal masses (independent of age) tary mass distributions of sedimenas a function of The present mass-age relations indicate that a cycling rate two to three times that of the shales would be sufficient to account for present day distributions. Also, there has long been an apparent discrepancy between the 20 to 30 the post-Precambrian percent age are ality; the 10x model predicts too small calculations cycle (3). If approximately 4,0 3.0 2.0 1.0 Units of 10° years Fig. 1. Distribution of mass of sedimentary rocks as a function of age. Curves based on models that assume the total mass of sediments existing today has remained constant throughout geologic time. Curves A, B, and C represent total deposition of a mass of sediments equal to two, five, and ten times the existing mass, respectively. The corresponding half-mass ages are 1.4, 0.6, and 0.3 billion years. Hachured histogram is an estimate of the actual mass distribution based on observed occurrence (1) and our interpretation of the Precambrian distribution. | 570 far too much (Fig. 1). in and the carbonate rocks 1.5 times to 2 faster than shale, they would be predicted to make up decreasing percentages of existing sedimentary rocks a function though mass of their may increasing percentage always as age, even of the total been very have nearly the same. The predicted present distribution of shale, carbonate, and evaporite as a function of age (Fig. 2) based on half-mass ages of 600, 300, and 200 agrees million years, favorably with (2). Finally, respectively, the actual the possibility disemerges that the very large percentages of cherty rocks, particularly of middle Precambrian age, may reflect the slow cycling of chert because of its high resistance to erosion and its characteristic protected position at the bottom of sedimentary basins. From these suggest that tarianism” | qualitative relations we geochemical should be a mass of older sediments, and the 2x model record 5 to 10 percent predicted for all sedipredicted (Fig. 1). The 5x model is of the right order of magnitude to fit actuof carbonate rocks mentary rocks by geochemical balance tribution Based on the assumptions of a total G—:— 0.3 x 10%years over of these relatively soluble materials. “uniformistrongly consid- The choice of the 5x model, as opposed to A n either of the others, is compelling. This conclusion would not change even if there were large errors in estimates of 80 he Percent Shale > 60H the actual age distribution of existing sediments. The mass distribution required to fit the 10x and 2x models is contrary to the experience of many geologists. Approximately half the total mass of existing sediments is younger than 600 million years, whereas the rest is distributed irregularly over an interval of Present Units of 10? years Fig. 2. Calculated present distribution of shale, carbonate, and evaporite as a function of time. The calculation is based on half-mass ages of shale, carbonate, and evaporite of 600, 300, and 200 million years, respectively. SCIENCE, VOL. 163