Читать книгу Stonehenge: Neolithic Man and the Cosmos - John North - Страница 13
ОглавлениеThe date is of course very sensitive to the chosen directions, and in all of them there are uncertainties attendant on the relatively small dimensions of the barrow. The measured azimuths quoted earlier produce a year of 3740 BC and a viewing altitude of 11.6°, with various qualifications that are hardly worth explaining. It would not be unreasonable to quote the date 3700 BC and to set the range of uncertainty as a century and a half either way. A reason will be given shortly for preferring a slightly later date.
The broad scheme explained here, with two pairs of perpendicular sight-lines to Deneb and Bellatrix, seems to be the only viable one. (Betelgeuse with Deneb offers the only other significant solution with any plausible lines of sight, but it is in the fifth millennium.) Lines along the axis, for observers in the ditches north and south, seem to offer no alternatives, and viewing at right-angles to ditches is in any case something we have found used often enough elsewhere. But why, under these circumstances, did they not simply build a square barrow, or at least a rectangular one? Why did they produce a skewed parallelogram?
The answer is that the axis had an importance of its own, for it was aligned on the setting of Rigel over the natural horizon. The barrow mound, following the altitudes derived here, would have been no more than 1.36 m above the eye of the ditch observers, and so no obstacle to an observer standing at ground level. As for the natural horizon, the town of Abingdon has long stood in the line of sight, but it seems likely that in Neolithic times the passage to the river in this direction would have been clear of trees, and that it is reasonable to take the extinction angle for Rigel. But what should that be? The site was not on high ground with good viewing conditions, but looked across the Thames valley. Using the norm for favourable viewing, the extinction angle would have been 1.53°. The derived declination in this case fits a year around 3600 BC. The extinction angle here could well be half a degree higher, however, which would bring the date a century later.
The Abingdon ditched (‘causewayed’) enclosure produced radiocarbon dates in the first half of the third millennium bc, quite consistent with all estimates made here, but why are these spread over a century and a half? The most probable error is in the azimuths, which are probably all a degree too high. On this assumption, all dates fall within three decades of 3700 BC. Whether or not this is so, it seems likely that the Radley barrow is half a century or more later than Skendleby 2. This is not certain, but what does seem certain is that both made use of the same clever device for introducing the rising and setting of the same star.
What of parallels with Skendleby’s alignments on the Sun? At Radley, there is no sign whatsoever of a solar alignment. The southern extreme of the Moon’s setting might have been observed in the same direction as Rigel’s, but this idea presupposes a level of sophistication for which we have no other evidence, and it must be abandoned. An isolated direction is no evidence at all. What is needed is evidence that other extremes were also observed, whether here or in the same general context, and this is lacking.
The last earthen long barrow to be considered here in detail was intermediate in date and place between the solar barrows at Skendleby and South Street. It was revealed in another rescue excavation, this time in the Nene Valley near Peterborough in the seasons 1974 and 1975. This ‘short long barrow’ at Grendon, Northamptonshire, has an importance out of all proportion to the quality of our knowledge of it, for it seems to signal a transition between two different astronomical traditions of barrow construction. At first it was thought to be a round barrow, like others on the same site (which also included an Iron Age farm), for it was enclosed within a double ring ditch. It was eventually found to enclose a square structure with a façade strongly reminiscent of Skendleby 1 (Skendleby 2 was excavated only in 1975–6), and the ring ditch was seen to be a later addition. In advance of gravel quarrying on the site—which had been discovered from aerial photography—the barrow was excavated, but not completely, and an ‘overzealous removal of topsoil’ meant the loss of much information as to the sequence of construction. But in this case the very geometry of the barrow comes to the rescue, and it is in some ways reminiscent of Radley’s.
The ‘long barrow’ components of Grendon (strictly area C, barrow V) are shown in Fig. 47, while the surrounding ring ditches are added to Fig. 48. All signs of the mound have been lost, but there is one good and one passable line from the façade, and one excellent straight inner edge to the ditch. What is more, the curved ditch opposite the façade fits closely to two arcs of circles of different radii, and the resulting ensemble of construction lines has some remarkable geometrical properties, quite independently of any assumption we might choose to make about potential astronomical lines of sight. Briefly, the angles marked x in Fig. 47 radiate from the centre of the circle bounding the inner edge of the ditch and are all equal to 22.5° (a sixteenth of a circle), while the angles marked y are at the centre for the outer edge, and are within half a degree of 60°. The quadrant made up of the four x-angles is not set to the cardinal points of the compass, but in any case these are not sight lines.
Looking for lines of sight one may begin by taking parallels to the façade. (Perpendiculars to edges, which differ only slightly, will be considered shortly.) The azimuths are 314.5° and 125.4° (each of course might be reversed) for crosswise viewing. These at first make for uncertainty as to which stars were involved. Looking northwest, there can be little doubt that the setting of Deneb was observed, but in the other direction one might choose between Aldebaran (beginning of the fourth millennium), Bellatrix (mid-third millennium) and Betelgeuse. Following the standard procedure, the rising of Betelgeuse is indicated, with the setting of Deneb, both around 3450 BC and at altitude 12.7°. This fits well with stylistic considerations, but it is ultimately the radiocarbon dates obtained from the site that confirm this preference.
Mature oak from the façade trench gave a date equivalent to only 2900 ± 120 BC (2330 ± 70 bc), but there is now good reason to expect the renewal of posts, so this is no embarrassment, when taken in conjunction with a date of 3490 ± 140 BC (2750 ± 130 bc) for a sample of mixed oak and blackthorn from the ditch. One of the post holes—that nearest the bottom corner of the figure—contained ‘fairly mature oak’ datable to 3690 ± 140 BC (3000 ± 80 bc), which it was thought might have come from old charcoal or be from a first use of the site. Whilst we can bring our date forward by a century, by assuming half a degree error in the Deneb direction, it falls within the stated range even without doing so, if old timber is assumed. The pair of post holes at the lower corner is so similar in position to the split posts at Skendleby 2—lying across and to one side of the sight-lines though the gap in the façade—and makes for such a consistent pattern of star viewing, that the mid thirty-fifth century seems highly probable.
FIG. 47. The Grendon square barrow. Potential lines of sight are added as continuous lines, and possible construction lines are shown as broken lines, some of them being placed—to avoid confusion—on the right-hand figure.
FIG. 48. The two Grendon ring ditches surrounding the earlier square barrow. The broken circles are the suggested aims of those responsible for the ring ditches. Their centres are marked (near the upper crossing of lines of sight), and they are shown in the inset figure in more detail.
The similarities do not end here, for despite their utterly different orientations—one is almost at right-angles to the other—the two barrows were clearly designed for observing Deneb twice over. The rising of Deneb would have been seen at Grendon by an observer in the ditch at D, looking over the mound along a line parallel to the inner edge of the southeast ditch. The azimuth of this is taken to be 45.5°, so that it would have fitted perfectly with the angle assumed for cross-viewing. They are not perfectly at right angles, but only a degree removed, and if the error is ours, rather than theirs, then the date—as alluded to earlier—moves to about 3550 BC. Like Skendleby 2, therefore, Grendon was aligned twice over on Deneb and once on Betelgeuse; but now, four or five centuries on, the last alignment is not on the setting but the rising of Betelgeuse.
The other important similarity of function involves solar observation. To decide this question it is first necessary to make an estimate of the size of the mound. One cannot say exactly where were the preferred positions for crosswise viewing, but those marked A and D (for Deneb’s setting and rising, respectively, parallel to the lines a and d on Fig. 47) and B (for Betelgeuse’s rising, parallel to b) are the most likely, in view of the large post positions, and the lines crossing over the centre of the square seem the most significant of all. The proximity of the crossing point to the centres of the ring ditches adds credence to the idea that the mound was at its highest there. Even though there might originally have been a ridge, it would in any case have passed through this point. The height needed here, to set equal angles of 12.7° to observers in the three viewing positions, is about 1.9 m above eye-level. Perhaps the ditches were in fact 1.9 m deep at the appropriate places, so that by standing at ground level an observer could have seen the natural horizon over the mound from any position around it. If the eye of a normal male observer was at ground level—our previous findings fitted perfectly an assumption that the height of a man’s eye was critical—then the mound would have been about 23 cm above his eye were he to have stood at ground level. It is not even necessary to make this assumption, however, to make a claim for solar viewing at the site: the southernmost pair of posts would have been usable in connection with the natural horizon, using the line (at azimuth 132.4°/312.4°) marked on the figure, in order to observe the rising midwinter Sun in one direction and the setting midsummer Sun in the other. This is so because at a conservative estimate the height of the mound here would have been no higher than about a metre, and it might have been less. (The placement of such a sight line with respect to the posts in the manner shown was to become standard, as will be discovered in Chapter 7.)
What is surprising here is the fact that, at a mere glance at the Grendon plan, midsummer sunrise or midwinter sunset would have been thought the likeliest candidates, either of them viewed along the axis of the barrow, just as at Skendleby 2 we found midwinter sunrise and midsummer sunset along its axis. At Grendon, this method is impossible; but by taking a line across the barrow, it is clear that what was observed could have been exactly the same as at Skendleby 2, namely midwinter sunrise or midsummer sunset or both. (It is also possible that these phenomena were observed by looking along the façade in some way.)
The Grendon barrow is of greatest interest, not because it mirrors a much earlier style but because it seems to signal the approaching demise of the truly long barrow. This was not to occur for several centuries, but neither was the transition a sudden one. Beckhampton Road was built at least a century after Grendon, and was converted to a round barrow presumably more than a century after that. Long chambered tombs in stone were probably still being built in the third millennium. Grendon shows that it is unnecessary to postulate an invading culture, such as the ‘Beaker People’, to see how circular barrow-forms might have arisen quite naturally. Of course it is not being suggested that it was Grendon in particular that supplied Neolithic peoples with the idea that barrows might just as well be square, or round, as long—Radley, after all, had even greater economy of line than Grendon. It is rather that this barrow is symptomatic of changing Neolithic attitudes to barrow construction. It symbolizes too the great naturalness of three prominent features in those circular mounds that qualify as early ancestors of Stonehenge: (1) the circular ditch, which is no more than a rationalization of the encircling ditch of a shortened long barrow; (2) a low mound of height roughly equal to that of the human eye, within that ditch, and corresponding in some way to it; and (3) an assortment of posts within the mound area, at first arranged for crosswise viewing, and later supplemented by others in ever more perfect circles.
Hazleton North and Burn Ground—Cotswold–Severn Long Cairns
Whilst the chambered long barrows at Wayland’s Smithy and West Kennet fit easily into the Wessex tradition, they also have affinities with the Neolithic stone long barrows (long cairns) of the Cotswolds, which might just as reasonably be considered their natural neighbours. (West Kennet is little more than 20 km from the string of known examples, and the isolated Wayland’s Smithy is hardly further from several at the northern end of the string than it is from West Kennet.) The differences are chiefly in the material used, although—threatened by endless classifications of type—it is wise to add that there are many differences of design too. Where easily split stone was plentiful it was used for the mound, and if larger upright stones (orthostats) were not locally available, then they were imported, from appreciable distances, if need be. Most barrows in the group were essentially roughly trapezoidal in plan, often with a Cupid’s-Bow wall at the high end (usually referred to as ‘horns’). Almost all seem originally to have been surrounded by a dry-stone revetment, a containing wall that even when excavated is usually found to be remarkable for its straightness along the sides. Changes of direction are much rarer than with earthen barrows, no doubt because good viewing positions are much harder to create along the length of the barrow, in a stony district. Like so many earthen barrows, there is a clearly identifiable spine, now usually of slabs of stone pitched together. Some of the stone barrows—perhaps all—have internal walls running across the width of the mound, dividing them into cells.
These internal walls are exactly analogous to the much flimsier lines of stakes found in so many earthen barrows, and clearly mark the stages of construction. The conjecture made earlier in this chapter, that the stakes were deliberately set in the directions of lines of sight, and at right angles to the neighbouring or far barrow edge, is harder to test with a dry-stone wall, but all the evidence suggests that the same principles were indeed applied, and nowhere is that evidence clearer than in the report—by Alan Saville and many collaborators—on excavations carried out in the period 1979–82 at Hazleton North, in Gloucestershire. The pattern of internal cells at Hazleton North, taken from that report, is shown in Fig. 49.
The long cairn in question is one of a pair, in a field known as Barrow Ground, 13 km east of Cheltenham and about 45 km west of Oxford. Many other long mounds are nearby, including Notgrove, Belas Knap, and Burn Ground (Hampnett). Hazleton North is on a bedrock of limestone, rising gently to the northwest. It is well placed for viewing over reasonably distant horizons in the southern half of the sky. It lies east–west and is fairly unusual in one respect, for its highest part was to the west—although it was once thought to be otherwise. In common with so many others in this, the finest English group of stone chambered tombs, it was being steadily destroyed by ploughing. The rescue excavation, for which several groups had been pressing hard, resulted in one of the most meticulous Neolithic surveys ever carried out. It revealed traces of human activity on the site before the mound (cairn) was erected. Tentatively grouped with this phase in the estimation of the excavators were a number of post holes and stake holes beneath the mound. Some of these follow the line of one of the transverse internal walls so closely (Figs. 50 and 51 below) that they were almost certainly aligned on the same astronomical event. Some of the posts involved were too substantial for merely staking out the mound, however, and their arrangement roughly resembles that of post holes bounding burial areas on the sites of earthen mounds. For want of more surface detail, they are not discussed further here, beyond the remark that if used for viewing over the natural horizon, the heavier posts might possibly have been aligned on the rising of alpha Crucis a little before 4000 BC. Early or late, the fact that they all come so close to the line separating cells U and S suggests that the mound was deliberately related to them.
FIG. 49. The general arrangement of the dry-stone walls that gave the Hazleton North cairn its form (after the much larger plan in the excavation report by Alan Saville). For the quarries, a scale, and a key to the lettering used in the text, see Fig. 52.
The arguments offered for the West Kennet chamber’s having been set up before the surrounding mound was begun might have been applied here, had the upright stones been more precisely tooled. As it is, one can only say that there are certain significant directions that they seem to define—for example, the longest side of the north chamber probably aligned on the rising Pleiades, and the spine on the setting Pleiades. If this is not an illusion, then once more it can be said that the alignments would not have been possible had a significant part of the mound preceded the uprights. One would naturally expect any viewing of stars across the finished mound to have taken place opposite them, if nowhere else; and in fact the quarry-ditches to north and south are best suited to such activity opposite the middle of the mound. Those quarries have never been extensively excavated, but their estimated outlines are as shown in Fig. 52. The floors of quarries, by the nature of things, are mostly very uneven, but since in the act of quarrying the builders reached down to marl at some points—especially in the central areas—they had at their disposal two flat areas from which to view. The inner area marked on the figure, namely the south quarry, the better excavated of the two, was extremely flat and level. As in the case of chalk ditches, levels are virtually identical to both sides of the barrow. To the south, small ‘islands’ of stone were left in place that could have been used as steps. One of these, opposite the south chamber, would have brought the adult male observer’s eye to the level of the base of the mound, but there is no way of knowing whether such a step was used—an unfortunate fact, for it would have allowed us to estimate the height of the finished mound. The excavator’s discussion of heights was inconclusive. The favoured reconstruction suggested a maximum height of around 2.5 m, while analogies with other cairns seemed to hint at a somewhat higher figure. While our earthen barrows have shown us that analogy is an unreliable guide, the angle of view to be derived shortly (11.85°), taken together with the symmetrical viewing positions marked Y and Z on the figure, imply a height of about 2.9 m—or about 25 cm more than this if the eye had been precisely at the level of the mound base.
FIG. 50. A detail of Fig. 52, showing the relation of the posts to the division between cells U and S.
FIG. 51. The stake holes and post holes under the Hazleton North cairn.
The astronomical argument proceeds exactly as before. The possibilities are considered of viewing at right angles to near and far edges and—as justified by the perfect level of the quarry floors—at equal altitudes. In the half millennium centred on the many radiocarbon dates known from the site, the only stars that offer themselves are the setting Vega and the rising Sirius. Viewing perpendicular to near edges, the date derived from the azimuths assumed (340.3° and 171.8°) is 3870 BC, and the viewing altitude 11.85°. (If the principles on which it is found are accepted, then the date is unlikely to be in error by more than half a century.) Surprisingly enough, taking far edges instead, the date is only put back to 3950 BC, but the viewing angle becomes 9.6°, and the known facts of the structure seem to rule this out conclusively. Once again, therefore, the orientation of cell divisions seems to be a misleading guide to viewing direction. In fact the pattern of these divisions is extremely interesting: at the head, perpendiculars are to the axis in the middle, the business area, they are to far sides; and at the tail of the mound one of each sort, near and far, crossing the entire mound. This seems to suggest that if there was no perfect pairing of stars across, and perpendicular to, the line of the axis, then the mound was used as a ‘precise’ artificial horizon only across its middle region. In fact as far as can be seen there was no other reasonable option.
There remains the possibility that stars were observed along the line of the barrow, as found elsewhere. The barrow sloped upwards to the west, and on our estimate of heights (2.9 m over the chamber and 0.5 m at the tail, 32 m distant from it) it did so at an angle of about 4.3°. Although this working is very approximate, the figure is so close to the extinction angle of the Pleiades that it comes as no surprise to find that indeed the Pleiades cluster could have been seen setting over the mound by an observer at the lower end, looking along the spine (estimated azimuth 256°) in the thirty-ninth century BC. (Taking the extinction angle, the year comes out at 3870 BC, but the uncertainty is of the order of two centuries.) This finding fits the other evidence so well that it comes hard to modify it. The ideal arrangement would have been with the observer’s eye at or below (depending on distance) the tail edge of the barrow. Our estimate of the mound’s overall shape is shown in Fig. 53, for which the line of sight is the right-hand half of the white line along the ridge. In short, we should expect some sort of pit or quarry beyond the tail edge of the mound. None was found, although the excavation did not extend very far in this direction. Of course the Pleiades could still have been observed setting in this direction, but they would have vanished from view somewhat above the ridge, rather than descending into the mound. The Pleiades might have been observed ritually by a large number of people, in the course of setting over the natural horizon along the same direction.
FIG. 52. The proposed cell structure of the Hazleton North mound, with construction lines drawn at right angles to the south or north mound edges or to the axis.
FIG. 53. The probable overall shape of Hazleton North.
This example of a barrow of Cotswold–Severn construction is taken chiefly to show that a shared set of principles governed the design of mounds, whether they were of chalk or stone. One last brief visit to another long cairn of the same group may be made to show that Hazleton North is not an isolated case. Only 5 km to the southwest the remains of a much-damaged long cairn of similar overall shape were excavated as part of a rescue operation in 1940–41. The barrow was at Burn Ground, to the northeast of the village of Hampnett in Gloucestershire. About 30 m of its length remained at its head, and perhaps as much again had been lost from its tail. Its general orientation, and the shape of its tomb chamber, strongly resembled that of the West Kennet barrow, except that at Burn Ground the chamber faced to the west, and was entered from either side of the mound via a passage running right across the mound. The internal walls were ill-defined, but the fragmentary remains of the stone edging of the cairn were in excellent straight lengths, and provide an opportunity to work from an absolute minimum of data.
The orientations of the north and south sides—at the head of the mound, at least—are taken as 84.5° and 91.0° respectively. Our usual method then provides us initially with three solutions worth examining, the first-named star rising in the south and the second setting in the north:
(1) Sirius and Vega in 4270 BC at altitude 10.3°
(2) Sirius and Arcturus in 3530 BC at altitude 13.4°
(3) Rigel and Arcturus in 3020 BC at altitude 10.6°
There are no available radiocarbon dates now, but (1) is surely too early and (3) too late. Recommending (2) is the fact that it is just a century after our preferred date for West Kennet, and that it introduces the two brightest stars in the sky, as at West Kennet. The altitude is high, but not impossibly so. In his report of the excavation, W. F. Grimes showed that no great attention was being given to possible ditches or quarries. (Two strips at right angles to the mound, north and south, were probably not dug very deep.) If the original overall height was the same as at Hazleton North, and the eye was at ground level, then the observer was at Burn Ground between roughly 12 m and 16 m away from the axis—depending on the mound shape—and 7 m less, measuring from the edging stones. The higher the mound, the larger these figures.
Granted that the building of long barrows was guided by the behaviour of the stars, presumably held in esteem for religious reasons, many questions still remain unanswered. Some of these are technical. How, if at all, did their builders achieve such excellent levels at corresponding points in opposite ditches? Did they use water channels for this? Or long beams fitted with plumb lines, as in a type of builders’ level used from ancient times to the present day? And what standard of perfection did they achieve? The question of quality of level seems never to have been asked by excavators, although the excellence of verticals has occasionally drawn comment. And what of the levelling of mound sites? Are the Neolithic plough marks under the South Street barrow not an indication of how seriously this task was taken, rather than a sign of earlier agriculture?
Other questions are essentially social. Where did it all begin, and did any of the manifest advances in technique stem from overseas? Without doubt, chambered tombs entirely of stone from fifth-millennium Brittany served similarly as artificial horizons, across which risings and settings of stars were observed. Their similarity to the Cotswold–Severn group of long barrows (cairns) is too great for the two to be independent, although both might simply have shared a third common source of inspiration. The problem of finding where observers stood in relation to stone cairns is more difficult than in the case of long barrows on chalk, for suitable quarries are not as easily manipulated as ditches in chalk; but the example of Hazleton North strongly suggests that future excavation will produce suitable viewing points, possibly confined to places opposite the side chambers. Once again, while the general principles are likely to prove to have been imported from distant places, there is every expectation that many specific local variations will emerge.
