67P/Churyumov-Gerasimenko. A Single Body That’s Been Stretched- Part 20.


Red- gouges on head and body lobes.

Blue- comet’s rotation plane. The blue line is plotted onto the relevant longitude line using a still from the video on the ESA blog post that’s linked below. The longitude line was traced round the body from where it runs through the x and y axis lines. These two axes are like virtual spokes on the rotational z axis so they define the rotation plane as they rotate. 



In Parts 10 and 11 the dynamics of the head lobe shearing and lifting from the body were looked at in the light of the rotation plane effects. Prior to lift-off, the head lobe’s position wasn’t at the extremity of the long axis of rotation. This means that there would have been a tendency for it to slide forward on the weakest fracture plane that happened to be sheared. The top face of that fracture plane would therefore be the Hathor cliff face we see today (which is the underside of Ma’at in the regional map in photo 3, below). In practice, the two planes couldn’t have slid directly on each other, i.e. grinding together across their entire faces. Most probably, the head tipped slightly before sliding as will be described in detail below.

If the top face of the fracture plane is the present day Hathor cliff, the bottom face is a little harder to locate because there’s a drop-down from the shear line on the body to the dusty Hapi region. Hapi can’t be said to resemble Hathor in its large scale structure at all. However, there are lines in the dust that continue the fracture planes in the neighbouring Babi region (Part 6 header photo). And the abundant dust on Hapi is largely due to the gravity field so it’s a bit of a red herring when comparing features. 

But it does seem as though one whole stratum is missing from the body as evidenced by the drop-down or small cliff edge all along the shear line at Hapi. It’s assumed to have been lost via sublimation and collapse but why wasn’t Hathor eroded in the same way? That’s a general problem, whatever theory is propounded for the comet’s morphology. 

The above two paragraphs are included because it would be a bit dishonest to talk of sliding strata layers without observing that somehow there’s a stratum missing on the body. We can only assume that it was there when the head sheared, otherwise there would have been a slab-shaped void through the middle of the comet prior to shearing- not very likely to say the least. Also, despite the dusty part of Hapi not resembling Hathor, there are matches running a long way down the underside of the Hathor cliff which match to features in the apparently eroded slopes of the smaller cliff along the shear line (described mostly in Part 7; also three stepped features that have been labelled in terracotta in various photos). So it’s a reasonable assumption that the stratum layer was there and has now disappeared. We’ll continue with the strata-sliding scenario based on that assumption.


If there was any resistance to the head sliding forward, in the form of protrusions between the strata layers or because the shearing hadn’t yet ‘unzipped’ its way all around the perimeter, there would be a tendency for the head to tip forward. It may even have tipped a tad sideways, initially, if the unzipping didn’t start exactly at the back of the head due to structural anomalies- this is expanded on later.

The tipping forward would have happened in line with the rotation plane in the forward direction. It would have been a hinging process with the chin of the ‘duck’ head (Bastet region) acting as a fulcrum as it ground against the body (Aker region).  And it would have occurred prior to the head being lifted in the translational sense from the body. A translational lift means a straightforward ‘vertical’ pulling away of the head with no further tipping or rotation. Here’s an annotated version of the ESA regional map of the comet showing the axis of rotation, rotation plane and the expected location of any hinge evidence on the separated lobes (please ignore bottom-left image which is not relevant to this post):

Photo 3  

Green- rotation axis (z axis).

Blue- rotation plane (swept by the x and y axes that are arranged like virtual spokes on the z axis).

Red- two large dots showing expected positions of the hinge on the now-separated lobes. “Expected” means the hinge locations if the tip-up behaved perfectly according to theory. 

Fuchsia- these superimpose blue dots on the rotation plane that are common to both views. One is at the top of the head lobe, the other is at the bottom i.e. the head rim. This shows that the smaller picture is displaying the opposite side of the comet from the hinge so it’s out of view.

N- north pole of comet rotation z axis.

S- south pole of comet rotation z axis.

Large mauve dot- location of suspected initial tear of the shear line (discussed below).

So the hinging involved an arcing up of the back of the head only. The translational lift followed, with the head lifting in its entirety, pulling the neck up with it (the main stretch) and maintaining the tipped attitude during that stretch. This behaviour is recorded in the present day configuration of head, body and neck. Furthermore, the tip is indeed in line with the rotation plane with the transverse cracks in the back of the neck neatly traversing the rotation plane line at 90° as you would expect (Part 10).

If the above scenario did occur, one wouldn’t be surprised to find a gouge in both head and body where the initial hinging occurred. There certainly would have been a lot of grinding of the surfaces at this point. These gouges would be opposite each other on the now-separated lobes and the rotation plane should ideally run through the middle of both of them. This is what we see in the header photo, which is reproduced below, except that the gouges are just to one side of the rotation plane:

Photo 4  

There are also matches in the two gouges which is surprising, given the apparent mayhem in this area. However, seeing as matter appears to have been gouged out, they probably aren’t matches in the sense of perimeter rim structures that were married tight before the stretch. They appear to be matching strata configurations that were submerged in the interior but are now exposed. There are four in total, three on the body and one on the head:

Photo 5  

Red- head lobe stratum feature.

Yellow- three matching body lobe strata features.

Blue- a ridge common to both lobes and noted in Part 3.

Green- (requires zooming) denotes matching striations in the curved extensions of the top and bottom strata. These are small scale matches within a large scale match. 

The left hand, curved extension to the head stratum configuration matches the left hand, curved extension to the lowermost body configuration. So, even though there are two more strata between these two, they match faithfully despite saying that there can’t be any properly married matches in the conventional sense. This is a puzzle but what is certain is that there are gouges in both lobes where we would expect them to be and well-defined matches between the lobes at this point whatever their true character may be. 


There’s another neat correlation that supports the hinging scenario. The 15° anticlockwise rotation of the head lobe has been referred to several times in the last 19 parts. It’s been revealed when the matches around the perimeter appear to be near-perfect but for this rotational misalignment. 

It must be emphasised that this is a different concept from the main comet rotation axis. It’s a small, angular displacement of the head lobe with respect to the body that has presumably stopped (for now) and has nothing to do with the continuous rotation of the comet in a 12.4-hour period. The axis of this displacement therefore has nothing to do with the comet’s z rotation axis. It’s actually at almost 90° to the comet’s z axis, causing the head to swing directly across the rotation plane of the comet. This probably isn’t mere coincidence as we’ll see below (ref: rebalancing of mass across the rotation plane).

The axis of this head lobe displacement was never referred to because it didn’t seem very important. But like everything else on the comet it’s evidence of a process, staring out at us, and we’ll eventually twig if we stare back at it for long enough. This axis isn’t centered in the middle of the head; it’s displaced to a point on the perimeter so that the head had a tendency to swing out from the body like sliding one end of the lid of a sarcophagus to one side. This would reveal a long V-shaped aperture along one side of the casket and the lid would overhang on the opposite side. A similar effect can be seen on 67P:

Photo 6  

Yellow- angle of eccentric head swing.

Red- delineates the head perimeter against the body in the background.

We therefore see head features that stay in line circumferentially with their shear line matches but get displaced in an almost translational straight line from those matches due to swinging on an eccentric axis. 

There’s a ridge that matches from head to body in the above photo and it’s annotated in blue below, as it was in Part 3. The head portion of the ridge can be seen to be translated (upwards towards top of frame in this view), while staying in line with its body match. It doesn’t get rotated along its head ‘circumference’ on a head-centered axis which would result in a noticeable sideways disparity across the frame. Instead, it’s a pure upward, eccentric displacement:

Photo 7  

To be exact, the displacement is almost a translational straight line but it’s actually a very gentle arc subtended by 15° from the eccentric axis located on the perimeter. This eccentric swing is also noticeable in the Part 16 header photo:

Photo 8  

The reason for describing the exact nature and location of this eccentric axis of head rotation is that it’s located on the Aker/Bastet perimeter, almost exactly where the hinge is located:

Photo 9  

Red oval- body gouge.

Red line- top perimeter of head gouge that’s out of view under the head rim but immediately below this point.

Green dot- axis of eccentric head rotation which is centred at the very edge of the body gouge or just to the side of it in open space. If we moved our view to a point that was exactly above the head perimeter (the view is currently very slightly offset), this axis may move in to a location that is comfortably inside the gouge.

Blue- comet rotation plane. Note that the anticlockwise head lobe rotation (yellow lines) is at almost exactly 90° to the comet rotation plane. This is as one would expect if the head was both hinging and swivelling from the same point- a point that’s on the perimeter and also straddling the rotation plane (but in actuality just to one side of it).

Large mauve dot- location of suspected initial tear of the shear line (discussed below).

‘N’ and ‘S’ are the poles of the comet’s z rotation axis.

Photo 10 may show evidence of the head rotation axis actually being in the gouge. Although it too is an offset view, the offset is at about 90° to that small offset in photo 9. That means we’re looking down the length of what were the angled, yellow lines. In this case it’s a red line showing the body shear line and a yellow line showing the head rim and we’re looking down their length towards their crossing point which is the axis. The angle appears to have closed up a bit, nearer to 10°, as suspected. However, everywhere else it’s referred to as being 15° so we’ll stick with that for now. The axis would again be out in free space (not shown) from this view but with a little bit of mental swivelling, we can see that it may slide in to a point that is within the gouge if we are looking from directly above. This would be a view with the two large, red dots aligned across the frame but with the head dot just below the body dot due to being swung eccentrically, just along from the axis. The two red dots are the same ones as in photo 3 i.e. the “ideal” hinge points sited on the rotation plane. The body gouge can be seen just to the right of the body’s red dot (not annotated). The head gouge is out of sight and just below the head’s red dot.

Photo 10  

Red- shear line.

Yellow- head rim.

2 large red dots- ideal hinge points.

Incidentally, it can be seen how faithfully the head rim follows the shear line in the above photo, both in terms of angle of each turn and the length of each portion. Measuring with a ruler shows them to be exact for all intents and purposes, bearing in mind the 1-2% foreshortening due to the head being closer and the tiny small scale anomalies that you’d expect. The only exception is the marked dog-leg on the body with no corresponding dog-leg on the head. However, the true head rim is just out of sight after the 90° turn. There is such a dog-leg along the head rim in photo 9, but it’s slightly less kinked. 

The location of the axis of head rotation in or near the gouge means that when the head tipped on its hinge and ground out the gouge, it also swivelled 15° sideways on that same fulcrum. The fact that the swivel axis is also where we expect the hinge to be is, of itself, extra evidence for the hinge.

More circumstantial evidence for this 15° head rotation is that the Anuket region of the neck shows signs of anticlockwise twist. There appear to be torsion lines running almost vertically up the neck but twisting to the right at the top (see Part 17 header photo). They have a scalloped cross-section which is reminiscent of when anything straight is subjected to torsional forces, such as box steel.

Some sideways swing correction would be entirely expected. This would happen as random mass-balance anomalies corrected themselves so as to balance equally either side of the rotation plane during the tip. This is an important phenomenon, the subtleties of which are analysed later on in this post.


“Random” suggests the head could have swivelled either way. However, it’s interesting that it went in the direction that’s directly opposite the catastrophic outgassing catalogued in Parts 7 and 8. This outgassing was never cited as a force capable of lifting the head or even tipping it. On the contrary, it was cited as a symptom of the very initial stages of the head trying to shear and lift. This would generate colossal heat and runaway sublimation in the core, leading to explosive outgassing through a very narrow aperture at the soon-to-be head rim. The evidence for this is in the form of conduits, vast slurry piles and, possibly, a thin line of precipitate draped along the top of a 500-metre, elongated slurry pile (which can be inferred from Part 1). All these features are located in a concentrated area and aligned exactly along the shear line. This point of initial shear is marked with a mauve dot in photo 9, above, and on the regional map, photo 3. The mauve dot is right next to the north pole and therefore at almost 90° to the rotation plane.

So if this evidence is a symptom of the initial head lobe lift, it follows that this was the unique point where the shear line started to unzip. That’s because once the rim had lifted even ten metres, the gases would escape without leaving the concentrated slurry piles and ridges that betray an escape through a thin aperture under pressure. And there’s no such evidence anywhere else around the shear line. 

Although the head should have tipped from the back first, structural anomalies may therefore have allowed this side section to be the start point of the unzipping. This would have been just before the rotation plane dominated proceedings. That would mean that although the rotation plane would dictate the direction of head tip, there was initially a lack of tensile resistance at this point, 90° to the hinge area. Since this unzipping point is right next to the north pole of rotation it implies that the head succumbed to an initial sideways kick-up at its north pole perimeter just before it started trying to tip forward in earnest. That in turn would mean that the tip was directed towards the opposite side of the comet (the south pole) at the very beginning of the process as the shear line unzipped fully on the outgassing side of the main hinge but had yet to unzip along the south pole perimeter. 

So the unzipped south pole acted as a temporary hinge before the main hinge took over. A tip towards the south pole would be at 90° to the expected rotation plane tip. On the face of it, this would be highly speculative but there’s more evidence for this scenario later in this post.

If there was an initial tip towards the south pole, it would explain how the mass-balance correction involved a sliding of the head on its incipient neck in a way that sent it on a 15° eccentric swing in that direction. The eccentric location of the head rotation axis at the gouge, combined with the swing direction, meant that the net displacement of the head lobe was southward. This should mean that the head was swung out to overhang the body on the south pole side like the overhanging sarcophagus lid. That would happen once the south pole perimeter line had unzipped along its shear line. This head swing would therefore have happened after the initial head tip in that direction but before or during the period where the main head tip from the back dominated the process due to rotation plane influences. Even today, the Hathor cliff seems to be tipped slightly towards the south pole as well as exhibiting the main tip-up along the rotation plane. It’s as if the whole tipping process was very slightly biased to the south pole side of the rotation axis resulting in a slight smudging of the ideal head tip-up that would be in line with the rotation plane. It would also explain the displacement of the hinge to one side of the rotation plane and its slight smudging along that end of the south pole perimeter.

Another piece of circumstantial evidence for the south pole tip is that there are no missing slab signatures on this side of the body. Yes, there appears to have been a lot of crushing and crumbling away but no clear crater signature like those of missing slabs A and B on the opposite side (see Part 9). The north pole runs almost exactly through the point where slab A was lifted (and almost through the initial zip opening) so it’s directly opposite that short-lived south pole fulcrum. Also, the residual slab A crater exhibits the hallmarks of a slab that’s been lifted cleanly and vertically by an upward-tipping head rim: it has a flat base, and just a few rocks displaced from its rim (rocks A-E; Parts 14,15,18). There’s no such clean slab-uplift signature on the south pole side, just the evidence of crushing and grinding. Thus, missing slabs A and B lend some weight to the notion of a southward head-tip with the fulcrum at the south pole and the head rim lift at the north pole.

Further evidence for the south pole tip is to be found in the fact that the long ‘vertical’ striations on the Hathor cliff show no sign of having been scoured at 90° to their alignment. The striations would have been directed north-to-south when the head lobe was seated on the body. If the head had undergone a slide along the rotation plane before locking into its hinge at Bastet/Aker, these striations would have been scoured across with new lines at 90°, or even obliterated. The fact that they remain intact and highly visible strongly suggests there was an initial southward tip that lifted the Hathor stratum cleanly away from the stratum below with no sideways movement along the rotation plane at all. This would have occurred if this north pole side of the head had indeed unzipped first. It would only need to lift a few metres from the stratum below before being at liberty to slide in any way it wished without its striations being damaged. That slide would have been forward along the rotation plane, if only for a few metres, before hinging at the gouge area. Because it had been tipped southwards and then forwards, it would slide on its chin only. That may be the specific reason for the southward offset and smudge of the hinge from the rotation plane: i.e. there was a southward bias to what was a predominantly forward slide and hinging. 

The evidence for the unzipping at the north pole side is detailed in Part 7 (major outgassing). Those long Hathor striations are conjectured to be the dykes that channelled the abundant, heat-generated gases from the core (Part 8). 


Whatever the exact process may have been, the south pole region of the head lobe appeared to have some tendency towards being the section that was destined to slide out and become the expected overhang. 

And yet, having described the whole process in detail above, there’s no overhang apparent on today’s comet as can be seen in this photo:

Photo 11  

Instead, the head lobe exhibits a remarkably straight facet on its south pole side and the head rim itself at this point is therefore remarkably straight as well. It’s also in line with the body perimeter. It’s as if it’s making a real effort to make a perimeter match and prove stretch theory once and for all. But it should be overhanging the body by 15°.

Looking more closely, we can see transverse lines running at an angle across that very straight facet. Those lines can be traced to the ‘corner’ of the head lobe where they meet up with the strata lines that curve right around the head. The transverse lines are therefore strata lines and that remarkably straight facet is where those strata layers have been sliced clean through. It should really be a continuation of the curved head lobe running right round to meet the curved strata lines at the opposite ‘corner’ on the right. This would leave the circular Hatmehit crater neatly placed in the middle of a circular head lobe without this strange slice taken off the side:

Photo 12  

Blue- rotation plane. Each of the six dots on the left of the head lobe is located on a fracture plane that curves round the head.

Red- gouges betraying the hinge. The head gouge is again mostly hidden under the head rim.

Yellow- dots at each end of the fracture planes that run across the flat facet. These can be traced to the sharp ‘corner’ and round to the blue-dotted strata lines.

Green- extensions of three of the fracture planes into the more crumbled region.

Photo 13 shows the fracture lines extending around the other side of the head lobe:

Photo 13  

So how did this slice get taken off the side of the head lobe? When it started to tip, that curved section of the head was still intact and it would indeed have slid into the overhang position if it hadn’t broken off first. It was suggested above that the head lobe was tipped towards the south pole initially (even if it was a rim tip-up of just a few metres on the north pole side) before embarking on its main rotation plane tip. The weight of the head was therefore trying to pivot up along this one edge on the south pole side. This caused the head rim to collapse along the line we see today at the bottom of the sliced facet. This in turn prompted all the strata above the rim to crack along the same fracture line, each one above the other in quick succession, leaving that strangely flat facet. So it wasn’t exactly sliced; it was cracked under the stress of tipping on its flimsy rim and the rest of the strata cracked in line, somewhat like the tiles in a karate chop display.

This proposition is supported by the fact that whilst the middle part of the chopped facet is very smooth, it is flanked by two crumbly ends. This would be expected when giving the rim a sudden shove upwards from its underside: the central part of the fracture was deeper into the head where it was held in, constrained by surrounding material so it broke cleanly. The two ends were not so constrained on their inside flanks and were completely exposed on their outside flanks, even tapering to nothing at the very corners at the head rim. This allowed them to crumble away upwards and away rather than crack in a short, sharp snap from a tightly constrained seating.

What’s possibly even more significant is that the strata in the middle are seen to be sandwiched tightly on top of each other but at their exposed ends they are feathered rather like concentric tiers of tiles (as evidenced by their continuation round the head). This means that at their very extremities, each layer could be lifted directly upwards from the one below. This process is seen at its most extreme in the left hand bottom corner where the tops of the strata layers are seen exposed after the layers above have been simply lifted off by the upward force from below the rim, instead of being snapped off leaving the vertical face. This lifting rather than snapping off would probably have involved some lengthwise help from their attached central portions doing some of the lifting once they had cracked cleanly in the flat facet area.

This crumbly area is denoted by the three lines of green dots. So that almost vertical, clean cut that characterises most of this south pole facet, faltered at this flaky end, becoming less vertical as larger and larger pieces of horizontal sections of strata were lifted off. This is because they were constrained by less and less material above. The gradation from perfect vertical slice to exposed horizontal stratum at the very tip is gradual and so the strata were sliced or lifted to varying degrees, causing the crumbly morphology we see today.

Indeed, the entire morphology of the chopped facet is explained by invoking a sudden force from below due to the head tipping up on this side of its rim. 

The right hand end of the slice is somewhat crumbly too and although the flaking process appears to be less evident, it appears nevertheless to be at play.

There might seem to be a problem with the idea of the head rim undergoing this torqueing force due to the head tip on its outer edge. If the comet was spinning fast enough to eject the head, there wouldn’t be any effective weight of the head bearing down on its southward rim. However, the rotation axis (z axis) ran close to the centre of the line that gave way. There would be less ‘centrifugal’ force at this point than at the centre of gravity of the head lobe, above. The only reason the bottom of the head lifted away at all is because it was attached to the rest of the head lobe whose centre of gravity, some 500-700 metres above, was indeed subjected to the required force. So, until the south pole ‘unzipped’ (or rather collapsed) the head was trying to tip and rotate southward, causing a torque on the underside of the rim on that side.

Coupled to this, we have to remember that the head wasn’t quite at the extremity of the long axis so it was still attempting to slide sideways towards the main hinge (the gouge) and couldn’t therefore lift directly off its fracture plane. As soon as it met any resistance it would tip. Again, that would involve rotation about the head’s centre of gravity, causing torque on the rim.

As well as the head lobe rim cracking, the body crumbled somewhat too as it succumbed to the grinding process, which is why the body gouge extends some way down the south pole side. That’s probably why we see that strange, curved protrusion on the south pole side of the body. It matches the two curved extension matches in the head and body described above in photo 5- the curves with the matching green-dotted striations. This suggests these curved structures extended further out and were sandwiched on top of the protrusion. The protrusion was evidently far enough down the body to avoid being broken off.

The head lobe strata-snapping scenario is made more plausible if we consider the alternative which is that the head tipped, neatly pivoting on impossibly small and resilient sections of head and body rim. The comet has the density of dog food and, given the proliferation of papers speculating on its fine structure being the size of marbles, it probably behaved with all the resilience of dog food as the head attempted to tip, initially southward and then along the rotation plane. What would have been a comparatively neat hinge in a rocky structure, simply collapsed until an area the size of quite a few football pitches was acting as the hinge point. Only at that point could the head start tipping in earnest and before that point all manner of upheaval was occurring on both lobes to accommodate the tip. This ultimately established a hinging area that could take the torquing force exerted by the head lobe without compressing or shearing any further. During this process, the body lobe had no choice but to grind away and compress until something akin to the socket in a ball and socket joint was established. The head lobe had no choice but to snap its flimsy strata, allowing it to collapse down, slide forward to the hinge and offer a larger area against the body as the ball in the ball and socket. 

This scenario was dominated by rotation plane forces but modified by the sideways tip of the head lobe as it initially tipped southward. It would explain the subsequent anticlockwise rotation as the head adjusted its position, sliding southwards (not just tipping). If the strata hadn’t been sliced off the head lobe, the head would indeed be overhanging today. Instead, the fracture line that was by definition set back when the head rim cracked, was rotated the 15° and that brought it back exactly parallel to the body lobe perimeter line. This is probably not the amazing coincidence it may appear to be, but a requirement of rebalancing across the rotation plane (see below).

As mentioned above, the southward tip and consequent rotation might explain why the hinge and the gouges that betray its existence are set just to one side of the rotation plane: the sideways slide smeared the gouge to that side. Furthermore, the fact that there was extra mass on the head and body that’s now gone means that the gouges were originally nearer to the centre of mass, or rather, the y axis line to be precise. By extension, this would mean that the hinge was closer to what was the former rotation plane. This is because the y-axis defines the rotation plane along with the x axis. Since the loss of that south pole mass, the rotation plane may have adjusted itself that small amount to its current alignment in order to compensate. Hence the hinge/gouges are offset with respect to today’s rotation plane. 

Conversely, it could viewed as being the mass that rebalanced across the rotation plane rather than the rotation plane adjusting to the mass loss. This would mean that the head was being flung round 15° to a point that rebalanced the comet mass either side of the rotation plane.

Viewed this way, the 15° head swing had more to do with compensating for all that head lobe mass loss than being triggered by the small, initial southward tip. Indeed, it’s probably quite significant that the 15° swing was exactly enough to bring the newly sliced head perimeter into line with the body perimeter again- it was simply rebalancing the head mass across the rotation plane by exactly filling the void that had been left by the sliced-off portion. That’s why the head swing just happening to stop with the new facet perfectly aligned with the body perimeter isn’t the coincidence it may appear to be- the old head rim was naturally in line with the body perimeter and balanced across the rotation plane; so the new head perimeter was destined to find its balanced niche in the same configuration by swinging obediently into place. 

If the snapped-off head strata were retrieved from their current deep space locations and stuck back on, we’d find that the continuation of the curve round the head lobe rim would indeed overhang the body like the displaced sarcophagus lid. It wouldn’t stick out by any old amount- the overhang would be commensurate with the 15° displacement of matches all round the rim. And the axis of the swing would of course be offset to the perimeter where the hinge was located and present day gouges can be seen:

Photo 14  

Yellow- former head rim line of section that was sliced off.

Orange- current rim of the sliced head and the rough line of the body perimeter that it sits on which is directly below it. This seating is what we usually call the shear line, the line along the body from where the head sheared. However, seeing as it didn’t shear from this side but hinged and crumbled, it’s more of a crumble line. The newly sliced head rim then swung into place above this line to replace the missing volume of lost material. The outermost body perimeter line is further out but that’s further down towards the base because the body slopes outwards further down. 

Red- rest of the head perimeter delineated to distinguish it from the body below.


It would only be after this complex hinging procedure played out that the translational stretch took place with the head lobe dragging the neck up with it. The translational stretch would have begun because the rotation plane head tip would have propelled the head far enough forward to become the axis extremity. It would also have overcome the sliding/shear resistance that was causing it to hinge- which is the same thing as saying that it eventually tipped enough to pop out of its rather sorry looking hinge socket. The point at which that pop-out happened may be written into the present day head tip attitude of about 30°-40° although some subsequent movement due to outgassing and mass depletion is presumably quite likely.

Although the scenario described above fits the features on the present-day comet, it probably didn’t play out exactly as described. As usual, it’s described as if fact so as to avoid burdening the prose with the obvious caveats. However, the features on the comet do suggest that something close to this scenario did indeed occur.


There are two flies in the ointment that were left out for the sake of fluency but will be mentioned here for transparency:

1) although the eccentric hinging should produce a gentle translational arc as described for the blue-dotted ridge, Part 16 header photo, and photo 10 in this post, it should actually produce a perimeter swing to the matches opposite the eccentric axis. This would involve the Anuket to Anubis matches of Part 17. Although those matches are very clear, it’s difficult to discern any such swing. However, none of the photos are very conducive to seeing the swing because they’re at unfavourable angles. But the best candidate, the Part 17 header, doesn’t seem to show it. It can only be assumed that it is there because the other three examples are so clear and identical in their swing.

2) there are matches between the present day south pole head rim and body (PART 19 header photo). But they are both supposed to have been crushed and rotated with respect to each other. However, seeing as the nearby gouges and the large south pole protrusion show resilient matches between the displaced, gouged out lobes, it can only be assumed that this is also the case for the apparent matches along the main south pole stretch. In other words, the strata features extended into the body of the comet and so still matched roughly when material was removed from both lobes to expose the interior. The matches in Part 19 along that south pole perimeter look fairly strong but the single-dot matches along the dark, shadowed area are indeed less strong. They can still be discerned in the gouge photo above but the lighting shows them to be rougher than the part 19 photo would suggest. The matches either side of the single-dotted area remain strong, that is, the line running up from the orange-dotted corner (already recorded in Part 17) and the green lines to the right (although those are slightly misplaced in Part 19 if compared to their correct placing in the gouge photo above).

Neither of these potential obstacles is a deal breaker. They both throw a little bit of doubt on the swing-out of the head towards the overhang position and suggest that the bottom of the flat facet is the correct match to the body perimeter below it after all. That would imply that (a) the body lost exactly the same amount of perimeter as the head and (b) that there was no swing at all. (a) is unlikely, given the mechanism for the head lobe slicing was unique to the head lobe itself. (b) flouts the evidence for a swing axis and rotationally misaligned features (blue ridge; Part 16 photo, photo 10) and the fact that the head used to have extra mass which would have to be overhanging if stuck back on.

These potential objections do therefore appear to be surmountable. Even if they were valid, they’d have no bearing on the initial southward head tip and ‘slicing’. It certainly doesn’t affect the rotation plane tip and subsequent stretch. And of course, it doesn’t affect the all-important matches, which are so well documented that nowadays any new matches are being introduced almost as being predictable and incidental to the more advanced implications of stretch theory.


There isn’t much to say in conclusion except to note that, once again, the morphology of 67P that’s described by the scientists as “tough” to explain, is explicable via stretch theory. 


The Anubis missing slab has gone to the back of the queue again. That’s because while writing this post, I realised that there’s distinct evidence for the head lobe stretching even before shearing. This was always a suspicion but certain features seemed to rule this out so it was dropped. It now appears that those features were a red herring. This is of course a much more significant finding than yet another missing slab.


Copyright ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0

To view a copy of this licence please visit:

All dotted annotations by scute1133.


67P/Churyumov-Gerasimenko. A Single Body That’s Been Stretched- Part 19.



The first photo to show the south pole of 67P with both head rim and body rim illuminated shows matches between the two. As usual, this comes as no surprise at all in stretch theory. No need for the usual long explanations. The header photo says it all.

This means that matches have now been made all the way around the perimeter of both lobes except at missing slab A, where they are expected not to be present. Doubtless, more detailed ones will be found between the others but any new matches are purely academic. 

This brief post has kicked Part 18 off the top, which was there for only a couple of days. Part 18 is a more interesting post with another displaced rock, a structure running from head to body and a shear line jump between strata.

Copyright ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0

67P/Churyumov-Gerasimenko. A Single Body That’s Been Stretched- Part 18.



Blue dots- rock D perimeter.

Red dots (very small, zooming required)- probable line of the rock’s bottom perimeter, judging by its shadow. The dots are small so as not to compromise the reader’s own judgment in estimating the line.

Photo 2- rock D in context with its rotation plane and rock C.  

Pair of bright blue dots- these show each end of rock D (top pair) and each end of its suggested seating position (bottom pair).

Dark blue- rock D’s  rotation plane corridor.

Yellow- rock C (middle-right of frame) and its seating position next to rock D’s seating.

Red- square feature on rock C’s seating that will serve for comparison and orientation with another photo later on.

Mauve- companion rocks to rock D, with similar rectangular features. They’re in line with the rock D rotation plane corridor and so are likely to have come from the same place. The two nearer companions are fuzzy in all photos so haven’t been given designations just in case they are proven to be fixed features and not rocks. The farthest one was discovered, just before posting, as being clearly detached so it should really be called rock E. Its position is significant, suggesting it should be twinned with rock D.


In part 14 we analysed how Rock A with a mass of around 30 million kilograms was detached from its original seating position and drifted 170 metres across the crater known as Landing Site A. The features on the detached rock matched those on the seating position. The mini crater that was actually sliced in two and shared between rock and seating made that match definitive. Rock B was dealt with in the same post, its seating position being slightly less definitive but its uplift probably collateral to rock A’s uprooting. And of course, the uprooting happened when the head lobe rim sheared away from this exact point and rose to its current position. Rock C was covered in Part 15. It was the biggest of the three rocks and whilst it didn’t have a sliced mini crater, two of its matches made its seating position and consequent 500-metre excursion virtually unequivocal. 

They’re all dubbed ‘rocks’ for the sake of convenience. We know they look like rocks but have a very different composition. As usual, scenarios involving rock D are described as if they are fact so as to save on peppering the prose with endless conditionals and qualifications. There are already plenty enough qualified statements in this post, just in order to avoid possible ambiguities as we twist and turn in three dimensions between strata.

So, unlike rocks A and C, rock D’s seating position and subsequent excursion is a hypothesis that can’t be proven definitively. However, in the light of the evidence presented here, the seating, detachment and rotation plane travel scenario is quite plausible. When looked at in the light of the accumulated mass of corroborating evidence for stretch presented in Parts 1-17 it’s quite probable. 


It would be easier to understand the references to the ‘rotation plane corridor’ and ‘missing slab A’ if you’ve already read Parts 14 and 15 but here’s a recap. Missing slab A was about 1 km across (see header photo, Part 9, for its shape and position). It was pulled up by the head lobe too, like yanking a bollard (the head lobe) violently out of a slabbed walkway and watching the surrounding slabs lift with it, each one lifting at one end and hinging at the other. Being near the rotation axis, slab A probably wouldn’t have escaped due to ‘centrifugal’ force alone. But the sudden shearing would have released pent-up tensile forces to do work, lifting the head and slab at an acceleration that was greater than slab A could otherwise muster. This pent-up tensile force would be commensurate with that needed to eject the head lobe at or near 1m/sec which is the comet’s escape velocity, implying spin-up to a circa 3.5 hour rotation period. The head lobe’s escape was prevented by having to drag the neck up with it but not before slab A had detached from the head (and body) at 1m/sec or more and escaped. 

The rotation plane corridor is the path that a detached rock would follow, backwards along the comet’s rotation plane, if it were detached from the surface. It would travel backwards along this line before being caught up by the body lobe rotating into it from behind. This would happen due to the rock being in a suborbital trajectory because the rocks in this vicinity, being near the rotation axis, were not subject to the same ‘centrifugal’ force that sheared the head lobe from the body. So, like slab A, they needed a helping hand from the uplifting head but they broke away from both head and slab early on in the upheaval process. They were therefore floating free at suborbital speeds.


Rock D is a more tentative match than the definitive matches of rocks A and C in terms of an exact mirror image of features between the rock and its seating position. However, the evidence for its former position being at that point or within a very small distance of it is compelling. 

Rock D is the last large rock on Site A that can be safely discerned as being a detached entity like rocks A to C and not a fixed stratum feature that has the appearance of being a detached rock. Doubtless, other true rocks will become discernible as the OSIRIS images are released. There are probably two more just behind rock D but they’re a bit fuzzy.

[UPDATE: after this post was largely completed a third companion was discovered. Like rock D, it’s discernible as a detached entity. It’s strewn in-line with rock D and the other two, along the rotation plane corridor, but landed much further down. It also exhibits similar surface features. So it should be dubbed rock E and borne in mind as being an equally valid twin to rock D for the rest of this post. Being in-line constitutes a signature- it would have come from the same vicinity and been subject to the same detachment and displacement process as rock D. It was added to the photo annotations at the last minute. It’s the isolated rock, dotted mauve and sitting close to the Site A crater rim].

Rock D is smaller than rocks A and C, and a little bigger than rock B. Its mass is still tens of millions of kilograms. Like the others, it’s sitting on Landing Site A, known informally as the amphitheatre. Whilst rocks A,B and C are towards one end of the crater, rock D is at the other end. It’s much further from its suggested seating position than the other three are from theirs. Readers of Parts 14 and 15 may discern a potential problem with this but it’s addressed lower down.

It can be seen from photo 2 that the suggested seating position for rock D is right next to that of rock C but more on the head lobe side of the shear line i.e. extending down the narrow, sloping ridge that falls away from the ‘horizontal’ seating area of all the other rocks. This ridge resembles a hollow-block wall, exhibiting square or rectangular dips along its top and two vertical ‘skins’ either side. The ‘wall’ is very straight and has the appearance of stunted remains, having been demolished almost to its sloping base. As we’ll see, the quasi rectangular hollows can be discerned to a greater or lesser degree in different photos and the two outside skins stick up slightly higher than the main body that contains the hollows. Rock D exhibits the same characteristic hollows at one end and has an S-shape that vaguely resembles the line of the sloping top of the wall, whichever way up it’s swiveled to make the fit.

The following photo, photo 3, is a close-up from the one used for the Part 15 header. It shows rock C’s seating position, including the red-dotted square that was mentioned in that part. The sloping ‘hollow-block wall’ remains are shown with two rows of bright blue dots delineating each of its top edges where rock D is thought to have been seated. The upper three pairs of dots enclose two square hollows. These look like one shadowed rectangle in most other photos. Two more square dimples extend below the two rows of dots as the hollow-block wall extends further down from the rock D position. These are out of focus in this photo but visible in photo 4. Being S-shaped it’s possible that rock D peeped over the top edge of the ridge and actually kissed rock C as well as rock B (not shown).

Photo 3  
Bright blue- suggested rock D seating position.

Yellow- rock C seating position.

Red square- this is retained here for comparison with photo 2 to allow for orientation. This is because the hollow-block ridge doesn’t look the same in both photos. It’s overexposed in photo 2 (witness rock A, 30 million kg, apparently turned to dust in the middle of Site A). The rock C seating position and its red square proves we’re looking at the same hollow-block ridge.

Orange dot- the pointed end of rock A’s seating position.

Photo 4 is taken from almost the same angle as photo 3 but with subtly different lighting conditions. It shows the other two square dimples along with the pair on the suggested seating position (very small yellow dots). So there are at least two pairs of squares, totaling four. Seeing as two squares weren’t apparent in photo 3 but are apparent here under different lighting, there may be more sitting before us but out of focus or overexposed. Four more rectangles that are further down in the same ridge are annotated as well, to show that rectangles are a common feature along this line.

Photo 4  
Red- each end of the hollow-block ridge as depicted in photo 3.

Yellow- two pairs of squares on the straight section and four more where the ridge curves round towards us slightly (below the bottom red dot).

Photo 5 shows this ridge in plan view in order to demonstrate how straight it is. This observation will be important for later discussion in this post. The photo is the only plan view of the ridge and shows its fairly long slope as being a bit stubby because it’s foreshortened.

Photo 5  
Red- one edge of the hollow-block ridge (zooming required). Its two pairs of squares aren’t shown here.

Yellow- three of the four extra rectangles mentioned in photo 4. At this point, the ridge starts curving away from the main, straight section that we’ll be focussing on but the existence of more rectangles proves that these are a common feature on the ridge.

Photo 6 is a close-up of photo 2. Again, it shows how straight the hollow-block ridge is, albeit from a profile viewpoint.   
Red- either end of the straight section of the hollow-block ridge.

Yellow dots on ridge- denote two shaded areas each of which accommodates a pair of squares, as mentioned in photos 3 and 4.

Whatever relationship rock D may have had with the hollow-block ridge, the main piece of evidence that suggests it came from this seating position is that the rock and its seating are sitting at either end of the rock D rotation plane corridor. The corridor can only be aligned at one angle from the rock where it sits today and that line goes through the seating position. Here’s a plan view of the corridor.

Photo 7  
Bright blue- rock D. Its seating is at the other end of the corridor, specifically, between the red and yellow dots.

Dark blue- rock D rotation plane corridor. Note that the bottom (outer) perimeter line almost touches rock A and its seating. This will be further discussed below in photo 8.

Terracotta- shear line.

Yellow- continuation of shear line around the crater that mirrors the ‘cove’ in the head above. This has often been dotted yellow by convention in former posts.

Mauve- rock D’s 2 blurry companion rocks and rock E (isolated to the right).

The rock D corridor line is all the more accurate because we can piggyback on the rock A corridor which relates to an unequivocal 170-metres of rock travel from seating to present day position (photo 8, below).

The suggested seating position also happens to be where the corridor intersects the shear line. In fact, this intersection is plumb on the three-way join of head, body and the perimeter of missing slab A. This is where the most upheaval was likely to have happened. 

There’s much evidence for this. As well as kissing the shear line and rock C’s seating position, rock D is likely to have kissed rock B. Rock B was almost certainly lifted at the same time as rock A because their configuration as they are arranged today on the flat crater floor of Site A is almost identical to their seating configuration but for the fact they tipped on their sides when they landed. Rock B’s seating is contiguous with rock C’s for its entire length, completing a suite of connections and actual movements that furnish strong evidence that all four rocks were detached in the same event. That event can only be explained by something sudden and catastrophic- the head lobe shearing and lifting. The join between rocks B and C also constitutes part of the perimeter of missing slab A, i.e. the rim of the residual crater we see. So rock D came from the exact point of the three-way join of head, body and missing slab. 

Rock A’s line of travel back across the crater floor is mappable to a very accurate degree by drawing a line between its pointed tip and the corresponding pointed end of its seating position. This line can then be used for the rock D rotation plane corridor line which would have been parallel. In fact, in order for it to be plotted parallel to rock A’s corridor, its outside line has to be adjusted so that it’s virtually contiguous with rock A’s inside line. Any other line isn’t parallel. 

Photo 8 (culled from Part 14 and reannotated- apologies for having three different types of blue).  
Bright blue- rock D seating at the top-left end of its corridor (needs zoom). Rock D itself is off-frame.

Dark blue- shows both the rock A corridor and the rock D corridor. Rock A’s corridor is solidly derived from drawing straight lines from the two end points of the rock to the end points of its seating. Rock D’s corridor is then constrained by having to be parallel to this line while also embracing rock D where it sits today. This constrains rock D’s seating to being somewhere along this line. 

Pastel blue- shows both rocks A and B plus their seating positions. Note how they retained their seating configuration even after the 170-metre drift. The dots depicting their seating positions show only the right hand edge for each rock in this plan view. Rock B’s is a little more debatable but it uses the shadow of the low ridge (mentioned below) for guidance. If this is correct, its other long edge was contiguous with rock C. So if it kissed rock D as well, it might have represented a continuation of the hollow-block ridge formation.

Yellow- rock C seating position (rock C is off frame).

Terracotta- shear line.

It can be seen from photo 8 that it’s likely that rock D kissed rock B when originally seated. This is because rock B was right next to rock A and protruded a fraction closer to the shear line. In fact, rock B’s shape and seating suggest it may have been a continuation of the hollow-block wall. It left behind a low ridge that appears to be a continuation of one of the skins of the wall that accommodated rock D. However there are no hollows visible along that line and the relevant facet on rock B is always either in shadow or too fuzzy to make out. 

So there’s some strong evidence that rock D came from this seating position but its actual seating position, as proposed, doesn’t have the sort of definitive matches that rocks A and C had. However, there’s more.

Photo 9 shows the head lobe seating position for rock D. The long line of rectangles between the two yellow dots shows what is in effect the seating position for the entire straight section of hollow-block ridge sitting below on the body. But with one proviso- they wouldn’t quite meet because mass went missing from between them in the form of rock D and other ejected rocks. The top facet of rock D would have been seated against this matching head lobe feature. It probably would have been a little way down the line from the ‘top’ rectangle, as viewed in this photo, and would have taken up around half of its total length. 

This very straight line of rectangles faithfully hugs a straight section of the oft-mentioned ‘cove’ on the head rim. The rectangles are just a fraction away from the perimeter itself, running along the underside of the head lobe. The whole line constitutes one side of the cove so the ‘top’ rectangle is at one end of the cove.

This straight section has been documented before as being the match to the sloping hollow-block ridge below it by virtue of the cove on the head matching the crater on the body. So it’s already irrevocably linked to rock D’s suggested seating position (whether rock D sat between them or not). Such a line of rectangles is to be found only here, positioned directly above the line of rectangles on the body ridge, and nowhere else on the head lobe rim. 

You can even see the two parallel skins of the ‘wall’ that were noted on the body’s hollow-block ridge. Notice also how the cove perimeter turns at a sharp 90° when it reaches the base of the line of rectangles. This will be of some significance in the discussions below and in future posts.

Photo 10 is a profile view of the line of rectangles. It shows the cove in context within a larger view of the head lobe.

Photo 10  
So the character and shape of the hollow-block ridge on the body isn’t the only evidence for rock D’s seating position. The seating position on the head lobe, now 1 km ‘vertically’ above, exhibits the same feature in great detail. The head lobe seating position would have been attached to the upper facet of rock D or possibly to a larger section of ‘wall’ that fragmented into rock D, rock E and their fuzzy companions. In other words, it would have sandwiched Rock D between head and body. It would follow that rock D was part of a seamless continuation of the hollow-block wall formation from head to body. Rock E and the two fuzzy companions do appear to be dimpled with squares or rectangles and are about the same width as rock D and the ‘wall’. They’re also perfectly positioned to have been part of the collateral damage.

The fact that the head lobe ridge hugs the perimeter of the cove means that, in structural terms, the perimeter is actually defined by the hollow-block wall feature, not vice versa. This means that there’s what might be considered to be a retaining wall running up this side of the cove, with mass on one side and the void of the cove on the other. This behaviour is mirrored on the body by the sloping hollow-block wall feature. It too acts, or rather, acted as a demarcation line with mass on one side and a void on the other. The mass has largely been ripped away but was once sitting there in the form of slab A. The void is in the form of the crater that mirrors the cove. The detailed matches around the curve of crater and cove were documented in Part 8. So the hollow-block ridge acts as a demarcation line between mass on one side and void on the other, in exactly the same way on both head and body.

This matching line of rectangles on the head lobe wasn’t found as an afterthought. In other words, there was no trawling around under the head lobe rim after a sudden realisation there might be something up there to match the rock D body seating. The head lobe line was already constrained to match the body seating by all the other matches along the shear line that were presented in Parts 1 to 8. Each match dictated the exact point of the next match, head-to-body, in lock-step all along the shear line. This was the penultimate match in the line before the matches disappear at the edge of missing slab A. These two hollow-block ridges were paired up 3 months and 15 parts ago, in Part 3. That’s why the match is said to be previously documented and irrevocable. It was just the hollow-block aspect that wasn’t mentioned before.

Photo 4 from Part 3 is reproduced below (photo 11 for this post). The two pairs of red dots next to the two yellow curves match from head to body. Rock D’s seating positions i.e. top-to-head lobe and bottom-to-body lobe are represented by the left hand dot of each pair, directly adjoining the beginning of each yellow curve. The bright blue dot on site A is rock D’s present position. 

Photo 11  
More specifically, the body seating suggested above would extend between the left hand red dot and first yellow dot; the head seating would extend downwards from the corresponding head dot or even starting a little below it, so it’s just out of sight. The right hand dot of each pair corresponds to the very last match before missing slab A took the rest of the matches with it along its top edge. The large scale cove-to-crater match is there too, denoted by the two yellow curves. 

So, as always, there’s simply no wiggle room in this current post to go casting around on the head rim for random ‘hollow-block’ matches in the vague vicinity of where they’re thought to be. The two points, body seating and head seating, are straitjacketed by the other matches along the shear line, which precludes any woolly match-making. We’re constrained to focus our gaze on one point and one point only on the head lobe rim whether there are hollow-block features there or not. And yet at that very point of focus we see the same hollow-block formation, set in the same very straight line, aligned at the same angle, holding mass back on side, open to the void on the other and found nowhere else on the head rim.

At present, the exact lengthwise matching of the two hollow-block lines is hampered by several factors, not least of all the several tens of metres of missing mass thickness. Lengthwise alignment refers to which rectangle on the head ridge sits above which rectangle on the body ridge to initiate the match, rather like sliding a set of two ladders a little way along their length and deciding whether to lock them on the second or third rung. The ‘straitjacket’ reference was to the ridges’ sideways alignment which is very exact but even the lengthwise alignment is straitjacketed by the other matches- it’s just harder to judge it exactly until someone somewhere cuts the neck out of their newly minted shape model and sits the head directly on the body.

This lengthwise alignment could be pinned down much more accurately using the more detailed ESA computer shape model and cutting the neck out virtually. The head to body match would also have to align features at 90° to the two ridges discussed in this post. Such a feature is the slurry pile at the back of the rectangle mentioned in parts 1-7. This would automatically lock the two ‘ladders’ on their correct rungs. The resolution would have to be at the current 5 metres at the very least and preferably 2 or 3 metres to get a really good picture of what’s going on. Clamping the head rim onto the shear line around the rest of the body would automatically show the two hollow-block ridges trying to clamp together but ending up parallel, one directly above the other and with a large gap between them. The distance between the tip of the head ridge and the scoured slab A crater rim below it would be the thickness of the slab A perimeter. It would be the same or thicker than rock D.

This exercise may produce the unsurprising conclusion that the the top tip of the head ridge protruded a little beyond rock D and up the side of rock B to its top surface. This would be an outcome that satisfies the fact that the head rim broke away from the true surface as it existed before slab A and the rocks were dispersed. After all, it was the head shearing from the actual top surface of slab A that lifted it and caused all this upheaval. That’s why I said that the seating for rock D might start just a little way down the head lobe line of rectangles from the exact tip. The actual tip would in that case reach to the top of rock B, or a bit higher, when it was seated. This obscure point wouldn’t have been mentioned but for the fact that it will crop up in future posts when grappling with the three-dimensional character of the shear line at this point.

That 3D nature is actually caused by the sloping ridge itself, coupled with the slab A thickness. And the underlying cause of all this weird sloping up of the shear line is due to it jumping from one stratum layer to another like tearing a piece of pleated paper at an angle to the pleats. The tear line jumps from pleat to pleat, running straight along each pleat for some distance before the next jump. If you put sticky tape across one of the pleats on both sides, the tear will very likely do a 90° turn at the tape and jump to the next pleat as a consequence. The shear line did exactly this because it couldn’t tear through the very ridge we’re discussing in this post. So it turned an abrupt 90° to go round it, jumping up to the next stratum layer in the process. That was indelibly recorded in that emphatic 90° turn at the base of the head lobe ridge of rectangles. There will be a post on this stratum jump but the evidence is all there in Part 6, either stated or implied. 

The reason this head-to-body match wasn’t elaborated on before by highlighting its matching hollow-block ridges, was because it was clear that there was some missing mass between them. It wouldn’t have been correct to say that they matched in the sense of actually marrying together although it was clear they shared the same morphological structure that ran through both points in both lobes. Rock D and rock E are possibly part of that missing mass that was sandwiched between the two, perhaps along with their two fuzzy companions.

The missing mass between rock D’s head seating and body seating comes as no surprise in stretch theory because the thickness of the missing mass is also the thickness of missing slab A as its edge tapered up to its perimeter line at this point (it was thicker across the low, flat crater floor).

All four rocks, A to D, were at that 3-way join of head, body and missing slab perimeter, giving rise to the focus of mayhem at this point. Rocks A,B and C were more attached to missing slab A (but kissing the shear line) while rock D was right on the shear line or even on the head lobe side. They all represent part of that somewhat tapered thickness of the 50-70-metre stratum that’s missing. Rock D’s shape, current position and integral hollows would simply suggest it’s part of the missing piece of hollow-block wall that originally continued seamlessly from head to body. 


There are several apparent objections to rock D coming from this seating position:

1) It’s four times further down the rotation plane than rock A so that suggests it must have been dropped from a great height, possibly above 750 metres (see calculations for rock C in Part 15). But slab A was only about 750 metres high even when hinged to vertical.

2) It couldn’t be dropped from the underside of slab A at vertical anyway because it’s in line with its own rotation plane corridor. 

3) It couldn’t be dropped from slab A low down because it would’ve landed alongside rock A.

The answer is that it didn’t drop from slab A. It’s on shear line or even on the head side of the shear line. It must have risen with the head lobe and been dropped from a good height, yes, but from directly above its seating position. This height probably wouldn’t need to be 750 metres anyway because on release it would be under a significant influence from the head lobe gravity but still experiencing a net draw to the body. This would allow it to drop much more slowly than rock A did and so travel four times further but from a lower release height than if it had been released from slab A. Remember that 4 times further down the rotation plane means a 16 times higher drop height if slab A is involved (see Part 15). That’s higher than the dimensions of slab A so it must have been released from the head  lobe.

Dropping from the head, and from directly above its seating position would also have allowed rock D to stay within its rotation plane corridor. It would still be fairly close to the rotation axis of the comet, meaning the body lobe gravity would still be dominant but compromised by rock D’s proximity to the head lobe gravity field at the beginning of its fall. This would mean a slower, longer descent resulting in capture further down the rotation plane.

The head would continue to rise despite rock D dropping because its centre of gravity was rotating at around twice the radius of the release point on its underside so it would have twice the ‘centrifugal’ force. Its original ~1m/sec rise would have to have been attenuated before release of rock D, though. Otherwise rock D would have escaped too. However, after several hundred metres of stretching, there would have been considerable shearing and heat generation in the emerging neck, leading to slowing of the stretch and retention of the head lobe.


Despite the circumstantial evidence for the rough original siting of rock D, its exact seating position is less definitive than the unequivocal sitings for rocks A and C. However, whether it’s rock D, rock E, the fuzzy companions or even escaped rocks that sat in the void between head and body lobes, this post has finally firmed up the significant match along the shear line that was made in Part 3: the hollow-block wall ridge that ran as a single structure through both head and body. 

Photo credits:

Copyright ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0


New evidence for stretch theory is emerging all the time and there will be more posts to come. Recent photos of the newly illuminated south pole of the comet confirm the prediction made in Part 11 that the head lobe perimeter profile would match the body lobe perimeter on that side. They also show evidence of the initial tipping forward of the head as it lifted off. So the next post might focus on that, but the Anubis slab needs a post too as well as matches from Bastet to Aker.