II. Four Vectors: Vertical Sections

Running in parallel with the fortnightly observations which continued to be made at the 'Eight' data points, observations on a Daily basis across other parts of the site were initiated in late December 2014. These were noted down on specially printed record sheets to permit ready transfer to computer spreadsheets.

Having reached a basic understanding of how the Grid & Lattice was arranged in the horizontal plane, the next priority for investigation was the 'Annual Wave Movement'. The perspex & wood 3-D wave model produced in 2014 (see photo) seemed to explain most of the 'static' features defined to date; it captured both the vertical orientation of the wave and also the resultant S-shaped form in the horizontal plane.

However, there was a need to explain the vertical plane movement in 'opposite' directions seen over time between Dn3-Ground floor and that of Dn3-1st floor. The first hypothesis to be tested was that of a 'rigid wave' moving in an 'upward direction' i.e., a wave possessing little or no change in amplitude or wavelength but simply flowing from Summer to Winter in an upwards direction relative to the Earth's surface. Such behaviour seemed to fit the 'back-and-forth' movement in the horizontal plane when seen at two points - as illustrated in Fig. 2.31 below; and measurements taken at the two locations over time seemed to fit the hypothesis reasonably well (see also the monthly plots already shown in figs. 2.19/20/21 in the previous section).

Fig. 2.31: Initial Hypothesis - a 'Rigid Wave' rising. 

For the remainder of 2015, vertical observations of Vector Dn3 were carried out daily at section CC (Ground Floor & 1st floor). However, to test whether the 'upward flow' Wave movement theory was supported or unsupported required observations of another vector's vertical wave movement. A similarly related pair of locations at the opposite end of the house were chosen: the expectation being that a second pair, with similar height separation, would behave in a similar way if the hypothesis was supported.

Unfortunately, the layout of the house made it impossible to find a second pair aligned exactly vertically above each other. The best option that could be found were two points on vector De6, separated vertically from each other by about 3 metres. They ran through the house from SW to NE, and were perpendicular to, and about 7 metres from the two Dn3 locations. It was assumed that such a relatively short horizontal distance from true vertical, between the two De6 points, would not impact too significantly on the validity of any comparisons with those from Dn3, bearing in mind the apparent 30 metre elongation in the horizontal plane already seen in the site plans to date. However, this assumption did not take into account the dramatic change in the underlying bedrock angle at this second location, which later proved to be a factor influencing the comparison.

When a graph of the daily observations collected over the course of 2015 from Dn3 with De6 was produced,(see Fig. 2.32 below) it showed that behaviour of the wave was significantly different at the two locations thus:

• Firstly, at ground floor level, the range of movement at De6-Study(~2.32 m) was less than half that of Dn3-Hall (~4.80 m); whereas, at 1st floor level, the De6-BedRm-1 range (~2.55 m) was a little more similar to that of Dn3-BedRm.3 (~2.84 m).
• Secondly, the wave movement at De6-BedRm.1 appeared to follow very closely the ground level De6-Study movement. In contrast, the movement seen between the two points on Dn3 were mirror images of each other - shifting in opposite directions.
• Thirdly, although the oscillating horizontal movement of De6 shifted 'negatively' towards the NW, from Winter (+1.0 m) to Spring (-1.5 m), this did not continue. After the Spring Equinox it reversed, moving back towards 'positive' SE, and reached a summer peak (+1.0 m) before shifting NW (-1.38 m) again as the Autumn Equinox approached. The graph showed that it finally returned SE, arriving at the previous Winter peak position and thus, completed what appeared to be an annual cycle; but one completely different to Dn3, the other vector with which it was being compared. 

Fig. 2.32: Graphs of the Year cycle - comparing Dn3 (top) with De6 (bottom).

These differences between Dn3 & De6 were puzzling at the time and required further investigation - a reasonable explanation for these movements could only be made once substantially more evidence was at hand. It should be born in mind that the divergence between Dn3 and De6 was a gradually emerging picture over the course of the whole of 2015. Each new corresponding exploration of the form threw up more data in need of elucidation. With the benefit of hindsight, the lack of a mirror image movement at De6 did not appear to support the large 'up/down' shift proposed by the initial 'wavelength expansion/contraction' model, particularly when considered together with the 'variations in range' in the horizontal plane at various locations around the site. A relatively 'rigid' wave, with unvarying amplification, like that being hypothesised, would be expected to produce a fairly uniform annual range of movement across the whole site, no matter what the vertical starting point. However, the steadily accumulating evidence showed this was not the case, but these facts were not entirely obvious at the time. Trying to visualise an alternative to this prototype rigid 'up/down' wave movement theory proved to be initially elusive.

A closer examination of the shape in the vertical plane was an essential next step to creating a better image of the form and unravelling the complexity of this moving Three-Dimensional shape. The two vertically related vector pairs Dn3 & De6, within the house, were too widely spaced to deduce the full curve of the wave, and gave little scope to be extended (limited by the 50 cm head-room beneath the ceiling).

The Workshop building, on the other hand, offered considerably more ceiling height (1.75 m of head-room) and additionally had sufficient floor space to make use of 'two ladders & plank' method to follow reactions upwards in stages. There were also a few exterior areas around the site where it was sufficiently level to do something similar, but they were quite limited in number due to the steep granite outcrop and its associated slopes.

Within the workshop, Vector Pe3/4 ran on an almost North/South axis - roughly parallel with workshop's East wall; it offered a little more potential to explore the vertical form of the wave not only indoors but also outdoors. There was level access to Pe3/4 in the Upper Courtyard to the South of the workshop, and also on the Path and the Drive to the North. Thus. on 18th Jan. 2015 (two days before New Moon), vertical measurements were taken along a 30 m length of the vector at six points in the horizontal plane, designated P 0,1,2...6 (see Fig.2.33[a] below). The varying vertical angles subtended at the six locations were recorded on a separate sheet (see Fig. 2.33[b] below) and then each fragmentary profile was transferred to a single compressed North Elevation drawing of the wave (see Fig. 2.33[c] below).

Fig. 2.33[c]: Pe3/4 North Elevation Composite section of six verticals.

The wave's transition through to a vertical angle lay between P5 and P4, and also at P1. Knowing that the bedrock was less than 30 cm below the surface at these locations suggested yet again, that vector Pe3/4 possessed a tilt - radiating upwards, perpendicular to the bedrock slope, at an angle of roughly 3° to 7° (See Fig. 2.34 below). The curve plotted at P1, Upper Courtyard (see top of fig. 2.33[b]), indicated a peak to trough distance a little greater than 2.6 m thus, suggesting a possible wavelength ≥ 5.2 m with an amplitude of ±1.6 m. This was a dimension not dissimilar to that found earlier (5.5 m) on Vector Dn3.

Fig. 2.34: Slope Angles along Pe3/4

Almost a fortnight later (30th Jan.'15) and four days before Full Moon, a similar set of observations was undertaken using Vector Pn2. This vector ran East to West i.e., at right angles to the previous vector Pe3/4. Positions were noted at 11 locations along a 38-metre section across the site (see Figure 2.35 below). They comprised the 1st floor of the house and several level areas outdoors including the foundry roof. The data offered further insights into the form of the wave as it crossed over the centre of the granite outcrop at the heart of the site.

Fig. 2.35: Geology & Plan view of Vector Pn2 across the site.

The individual measurements were combined in West and East Elevation drawings which again suggested a vector wavelength in the region of 6.0 m. However, the amplitude between the buildings was notably smaller - oscillating as little as ±0.7 m. It was difficult to know whether such a measurement related to the timing of the observations or to their situation - sandwiched between masses (see Elevations in Figs 2.36[a] & 2.36[b] below).

The problem of why there was an opposing movement between Dn3-Ground floor and Dn3-1st floor, while parallel movements occurred between the two points on De6, despite the similarity of their locations, still remained unresolved at the Summer Solstice 2015. To shed a little more light on this problem and see if a reason for the difference would manifest itself, a 3-day, detailed examination of vertical sections along the length of Vector De6 was undertaken at around the time of Aphelion (4th & 6 -7th July 2015). Wave angles were measured at 5 exterior and 7 interior site locations at a variety of heights along a 22 metre section. The SE face of the Chimney together with the corresponding Interior Wall were used as a single co-ordinate reference for locating each De6 position; they were then amalgamated on a NE elevation drawing (see fig. 2.37[a] below) and also on a section BB drawing (fig. 2.37[b] - top). The results again appeared to show a wave emanating perpendicular to the underlying slope of the bedrock but also possessing a slight tilt (10° to the SE from point No.1 to point Nos. 3-A,B, & C). Further, when viewed in section, the NW elevation drawing (Section BB [top] in Fig. 2.37[b]) appeared to show the wave draped over the Granite Outcrop and fanning out at a variety of angles in apparent response to the underlying mass - emanating at angles ranging from in excess of 15°, as the outcrop emerged from deep underground, to vertical as it crossed the summit. Notably the curves all had quite small arcs, suggesting a Midsummer wavelength of possibly as little as 3.0 m.

Close examination of the De6 results from the ground floor & 1st floor key locations (points 3.4-Study & 3.A-Bedroom 1), showed that the wave sections appeared to be at their peak NW amplitude at both locations (see fig. 2.37[a]). They did not appear to be in a 'Peak and Trough' arrangement, as was the case with the two Dn3 points. This seemed to partially explain the difference in recorded movement between Dn3 and De6 seen in the 2015 graphs shown earlier (fig. 2.37). The horizontal dislocation of 3.0 metres between the two points on De6, when considered alongside the estimated tilt angle, suggested the broad upward curve of the wave movement at De6-1st floor would logically follow the De6-Ground floor movement in a 'Peak to Peak' layout at shorter wavelengths.

At this point in time (Summer 2015), the 'moving wave; expansion/contraction' model did not appear to have much evidence to support it; however, there was still insufficient data to allow consideration of an alternative hypothesis. More detailed information was needed on the 'day-to-day' curve of the wave over short vertical distances, but this involved prioritising the amount of time that a single researcher could devote to any one aspect of the total form under investigation; and, in the short term, a fuller understanding of the ground level horizontal movement continued to be the first priority.

However, in late June 2015 a little more time was set aside for this second aspect. An additional daily vertical observation was made 37 cm above the Dn3-1st Floor level; and a similar second point of observation at Dn3-Hall, Ground Floor level, followed in Nov'15 . This allowed the production of a sectional graph (see Fig. 2.38 below) which did not appear to support the idea of a large change in upward wave movement to account for what had been observed. However, an alternative proposition could not be put forward at the time due to a lack of intervening vertical data points. 

* In fact, (but only with the benefit of hindsight) the sectional graph can be seen to suggest a wave 'oscillating in the vertical plane', peaking in one direction towards mid-winter, and then peaking in the opposite direction at mid-summer; with a 'Point of Inflection' hovering at around 46.0 m a.s.l. i.e., somewhere in the region of 3.0 metres above bedrock. Unfortunately, lack of familiarity with such a large scale harmonic, oscillating wave-form (together with a scarcity of confirmatory data from other parts of the site) resulted in a failure to recognize this pattern until another year or so had elapsed. 

Fig. 2.38: Dn3 Vertical Sections over Time - Midsummer 2015 to February 2016.