67P/Churyumov-Gerasimenko. A Single Body That’s Been Stretched- Part 41



Orange- delaminated layers that rode up the head lobe due to crustal tensile stress after shearing from the body lobe.

Light blue- the four ridges that correspond to the delaminated ‘gull wings’ on the body.

Photo 2- a different view showing the four cuboids that match to the light blue ridges. The cuboid slide was described in Part 40. 


Light blue- as for photo 1.

Orange- as for photo 1 on the head lobe but orange shows the Aten perimeter on the body. 

Fuchsia- four quadrilaterals that are the faces of the four cuboids. 

Red- cuboid slide lines. The red kinked sections directly behind the quadrilaterals were yellow in Part 40, depicting the sloping backs of the cuboids. Although these kinked sections retain lines that are a signature of the slide they are strictly speaking the signature of the delaminated ‘Venetian’ bands that bunched up behind the quadrilaterals to give them their bulk. But their direction is very nearly along the slide vector anyway. The other sections of the red lines leading to the green gull wings are along the true slide vector. The kink is largely due to the sloping backs of the cuboids. 

Green- the gull wing delaminations.

Dark blue- north pole.

Brown- the paleo pole preliminary adjustment (Parts 37 and 40).

Photo 3 -another view. The isolated orange dot is a rogue dot. 



The upper, slid/torn layer corresponds exactly to this line of prolific sunset jets as described by the OSIRIS team on the Rosetta blog:


These jets are partially if not fully explained by the tearing of the layer at the shear line.


Apologies for being a bit repetitive in this post. It’s because we’re dealing with two delamination processes that happened at roughly 90° to each other and then looking at the signatures of both in one area on the head. One delamination process was the long axis, core-directed delamination and the other was the radial sliding of crust up the head lobe once the crust had been loosened enough to slide radially. Both involve the same chunk of crust obeying the long axis delamination first and then sliding (riding up the head) into two delaminated layers. This is actually no different from the ‘Venetian blind’ layers delaminating across Babi along the long axis and then sliding radially to form the cuboids (Part 40). But the signature is harder to see on the head lobe. And it’s especially difficult to visualise what is effectively a delamination of a delamination. 

All references to up, down, left and right are in the ‘upright duck’ mode with the head lobe directly above the body lobe.

Long, narrative keys to photos are concluded with ‘/////’.

Here’s the ESA regions map for navigation purposes.

Photo 4- ESA regions 



In Part 39 we looked at three onion layer levels at Babi and designated them as levels 1,2 and 3. Level 1 was the top layer that slid radially from the north pole. Level 2 was the present-day Babi surface containing the shear line at its rim, overlooking Hapi. Level 2 looks almost like riven rock as if it’s an exposed fracture plane and if level 1 slid from above it that must indeed be what we see today. Level three was the next onion layer down, in Hapi, peeping out from under level 2.

Photo 5- This is the Part 39 header, reproduced and showing the three levels. 


Terracotta- three lines of dots show the front lip of each of the onion layers, levels 1,2 and 3. The two short lines of terracotta at right are the back edge of the radially recoiled level 1, where visible. It’s largely visible around the back of Site A. It’s also just visible in the bottom-right corner where it traces the bottom edge of the cuboids at Aten. The level 2 line continues on into the distance past Site A because it’s the classic shear line. The other two lines for levels 1 and 3 could be extended too because there are level 1 and 3 onion layers beyond Site A as well but it’s not relevant to this post. Those layers have been dealt with in the past: level 1 is the red triangle recoil; and the fracture plane below the four coloured anchors is level 3 (both in Part 26). They just weren’t called levels 1 and 3 at that time.

The two sets of numbers in photo 3 denote the same three layers, but the farther set traces them across Site A and the nearer set traces them across Babi. 

Light blue- delaminated ridges embedded in the surface of level 2. These delaminated predominantly along the long axis but succumbed to a small amount of radial tugging without breaking free and sliding like the level 1 crust above. The radial tugging is evident in the fan shapes. Their widest extent betrays the width of the properly delaminated ‘Venetian’ bands above level 2, i.e. the bands in level 1. And those four bands became the four cuboids perched over Aten by sliding radially and ‘scooping themselves up’ along their length. 

Dark green- the five sets of gull wings.

Mauve, yellow, orange, green- the four anchors (Part 24) in the distance. These aren’t relevant to this post but are included for context. 


The wiggly nature of the classic shear line is probably due to all the radial tugging of the crust above it, which was the level 1 onion layer. The level 1 crust slid away radially from the north pole point and bunched up further down the body at the back of Site A and Babi (Parts 32 and 40). One telling signature in Photo 3 is that the terracotta line defining the shear line (rim of level 2) is remarkably parallel to the level 1 line (rim of level 1). That’s despite the fact that both lines are quite wiggly. They stay parallel across both site A and Babi and even across the central divide between the two. That central divide is the short, dusty section of Ash and in fact the level 1 rim line traces that short section via the dividing line between dust and no dust. As for Babi, that’s a signature of the division between level 1 recoiled material and the bare (riven) level 2 surface it slid over.

It should be borne in mind that today’s head rim matches prove that the head was still attached to the level 2, classic shear line while all this radial tugging was happening on the layer above it. Both head and body had level 1 crust unzipping and riding around radially above the level 2 layer. Level one unzipped and so level 1 crust rode up the head in mirror fashion to the level 1 crust riding down the body. That exposed the head rim and shear line which we see today and tend to think of as the true surface of the comet. But the head rim and shear line used to be neatly covered by the level one layer that’s now split and shared between head and body- way up the head and way down the body. So even the head rim and a short section above it is level 2 as well. That’s because it was attached at the shear line. The short expanse between the head rim and first section of ridden-up layer 1 material on the head would be the short section of level 2 that was attached to today’s Babi just before shearing. 

Indeed, the level 1 crust on the body needed something to tear away from in the first place and we know it couldn’t have torn from today’s exact head rim line because that rim matches to today’s body shear line. Today’s shear line was originally buried under level 1 and hadn’t yet sheared when the level 1 crust was unzipping and sliding away radially above it. The head rim certainly doesn’t match to all that crumpled-up level 1 crust now concertinaed up at the back of Site A and Babi; it definitely matches the shear line directly below it. 

That concertinaed crust at the back of Site A and Babi actually tore from its counterpart crust on the head lobe when the head was still seated on the body and just before the head sheared away for good. The level 1 crust on the body simply continued on over the herniating head lobe when it was still seated. The level 1 crust tore along a line that was, yes, directly over today’s shear line but since the head hadn’t departed yet, today’s classic shear line underneath hadn’t quite opened up yet. After tearing, the two components of the level 1 crust on head and body both recoiled, one down the body to the back of Site A and Babi (plus the slab A extension) and the other up the head i.e. up Ma’at and towards Hatmehit. 

The preliminary tearing of the level 1 crust along the same line as the shear line could be considered as the precursor to the classic shear line at level 2 directly below it. It suggests that the classic shear line was already beginning to weaken and on the point of sudden failure followed immediately by head lobe release.

Since the head was herniating from the body, it could be visualised as being like an expanding balloon: as core material migrated from the future body lobe into to the interior of the head, it became more and more of a bell shape sitting on top of the body. If you coated a half-blown-up balloon with papier mâché layers and let them dry before blowing it up more, the layers of papier mâché would crack and move apart as it expanded. If the glue between the layers were still slightly wet, the layers would tear, possibly along one line, and slide away from each other in two halves under the released tension, rather than crack into many islands that move apart. If they did tear along one line, the layers would recoil in two directions at 90° to the tear. That’s what happened along the shear line. 

The recoil under tension in the case of the comet was due to the pent-up radial tension in the level 1 crust due to spin-up force. 

There are clear signatures of this sliding happening at Imhotep and the south pole as well (see the Imhotep slide page in the menu bar and Part 30 for the south pole slide). Both have crustal slides of the same order of magnitude as for Babi, Seth and Ma’at, around 800 metres).

So pent-up tensile stress in the crust and possibly lubrication could have facilitated the crust recoiling process. There certainly was a dramatic recoil of layers on the comet and there are slurry signatures all over Seth and Babi so perhaps there was some lubrication that explains the degree of sliding. But you still need the tensile force stretch vector (achieved via spin-up) in the first place so slurry would only be helping along the process of crustal sliding, not causing it. 


The ridden-up (recoiled) head layer component that matches to the cuboids was shown to be constrained between the light blue, vertical ridges in Part 40. However, if you recall, there are four light blue ridges enclosing three interstices. This means that it is the left hand three cuboids that match to the three interstices. The fourth cuboid was attached to a point somewhere in the cove. Since the left side of the cove is defined by the right hand member of the four light blue ridges, the right hand cuboid is outside the array of four ridges. So it doesn’t match to the three interstices, only its three companions to the the left do so. 

The four light blue ridges correspond to the gull wing delaminations directly below them on the body. The level 1 recoil component on the head corresponds to two thin layers that rose above the head rim and parallel to the rim. Those two layers constitute what was the head component of our supposed single level 1 onion layer. It might have been discernibly two layers or simply susceptible to delamination (as were the cuboid components). Those two ridden-up layers are dotted orange in all the photos in this post. 

The two ridden-up layers are at 90° to the blue ridges and that’s because they are the recoiled layers, recoiling only after they had delaminated along the long axis stretch vector. And the recoil vector, being radial, meant that it was at about 90° to the delamination vector. That’s why the orange lines in the photos cross the light blue lines at 90°. The light blue lines represent the vestiges of the delaminated layer fronts that spread out in bands along the long axis stretch vector. They were likened to the slats of a Venetian blind in part 40. That was for the body but in truth, they spread on up the soon-to-be head lobe from Babi and into the future Ma’at area. And the reason they succumbed to the long axis vector before sliding radially is that the core stretch was dominant, stealing core material from the comet’s short axis and donating it to the long axis. The crust had to keep up by delaminating and only once it was loosened by the delamination could the more feeble radial forces induce the radial slide at the surface. And the radial forces were able to act only after the delaminated level 1 had torn in half along a line directly above today’s classic shear line. That tear or upper shear line was across the delaminated layers and at 90° to them. That’s why the recoil, at 90° to the tear, bunched up the delaminated layers along their length as they rose up the head (as happened for the cuboids). That left the signature we see, the vestiges of the layer fronts. Those are the light blue ridges and they look slightly crumpled, probably due to the old layer fronts bunching up under recoil, becoming more flaccid and zig zagging a bit. 

The vestige of the long tear, unzipping just above the actual shear line, is the two orange layers running roughly parallel to the head rim. And they run parallel because of course they tore from a line that ran along the same line as to soon-to-tear head rim. The two layers then recoiled up Ma’at, above the rim line, just before the head rim itself was released (and once the rim was released, it allowed the head lobe to rise on the growing neck). 

The two ridden-up layers are actually extensions of the second and fourth ribs of the cove next to them to their right. That’s because the cove was subjected to exactly the same process of tear and radial recoil at or around same time. The two lines of the layers are even traceable right through the second and fourth cove ribs and on down to a point where they almost meet on the Ma’at perimeter to the right of the cove. This fact will be seen to be significant for the Site A slide in a future post. 

Photo 6- the two lines that comprise the two delaminated layers in this post as well as the second and fourth cove ribs and and two extensions to the head rim perimeter.  


Orange- the two layers in this post.

Yellow- the second and fourth cove ribs. 

Red- the two extensions along and down to the Ma’at perimeter. The Ma’at perimeter is the head rim. The bottom line extension actually reaches the head rim immediately after traversing the second cove rib. However its path between there and the bend in the head rim is about parallel to the upper extension and in that sense they are structurally linked. 


The classic shear line along the level 2 onion layer at Babi and Seth is the catastrophic shearing of the entire head lobe. That shearing happened only after the level 1 layer components had torn along the same line as the shear line and recoiled from each other, up the head and down the body. Only then did the classic shear line actually shear, allowing the head lobe to rise on the growing neck with the matches perfectly preserved at the moment of shearing. The matches were preserved on the head rim and shear line. The level 1 crust, which tore into two halves, slid up the head and down the body under the influence of radial forces. One half ended up as the four cuboids in Babi, overhanging Aten and the other half ended up as two parallel layers that rode up Ma’at. Since both the head lobe shear and the level 1 tearing above it happened only after the long axis delamination into ‘Venetian’ bands, the slid components retain a crumpled signature of those bands. The signature is in the form of wiggly lines tracing the crumpled layer fronts. They are crumpled due to being flaccid after all tension was released after the tear. 

The same principle applies to Seth. The level 1 crust tore into two halves. The body half slid to the back of Site A and the slab A extension. The head lobe half is hiding in plain view on Ma’at. It will be weeks before it’s characterised in detail on this blog. Perhaps someone else will identify its location and slide vector (including a sideways split) before it gets posted.


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 40



4 Fuchsia dots- these denote each of four cuboid lumps that slid across Babi from the gull wings.

Small fuchsia- perimeter of the front faces of the cuboids.

Red- slide direction which is consistent with the centrifugal force vector. It emanates from near the north pole. 

Yellow- many bunched-up mini onion layers are to be found between these yellow lines. They look like shards of thin slate that have been stacked in a sloping pile, each layer peeping out from the one above. They were once spread out over Babi towards the gull wings at the shear line and formed what was ostensibly one main onion layer, the top layer of cometary crust (‘level 1′ in Part 39). They contribute to the bunched-up bulk behind the two middle cuboids’ faces. Each yellow line rises up quite a steep slope of material, which is part of the reason for their being at an angle to the red lines. These are real lines (see unannotated version). They appear to be linked to each cuboid vertex and each set of gull wings (via the red lines which are also real lines in the unannotated version).

Green- the five sets of delaminated gull wings.

Dark blue- the north pole.

Brown- the paleo pole adjustment (Part 37).

Orange- the perimeter of the Aten depression. The bottom edges of the four fuchsia cuboids join in one remarkably straight line and this line completes the perimeter of Aten. 


The ESA regions map is at the end of this post (photo 10). All references to up, down, left, right are in the ‘upright duck’ view with the head lobe above the body lobe, unless otherwise stated. Longer photo keys are concluded with ‘/////’.


This part builds on the material presented in Parts 37 and 38. It may be difficult to see how this sliding of the Babi crust layer happened without reading those parts. However, the essence of the arguments laid out in those posts is as follows. 

The gull wings delaminated along the long axis stretch vector, which is in line with the Hapi border and the Hapi border is the shear line where the head lobe sheared from the body lobe. The delamination was therefore towards Aker at the comet’s long axis extremity. This is the main stretch vector as dominated by the comet’s core and evidenced by the fact that the body is nearly a perfect diamond shape. You can click the second video in this Rosetta blog link to see the diamond shape as viewed from below Imhotep:


The stretched long axis of the diamond had to involve the core. Core material was stolen from the short axis and given to the long axis during stretch. The crust at Babi had to yield and succumb to that long axis direction of stretch. It was too brittle to stretch so it delaminated in that direction. You could call that a preliminary landslide but it’s in the long axis direction along the Hapi shear line which is itself directed along the long axis. And it was probably more of a prising apart of layers than a slide. The landslide in this post involves the same section of Babi crust but now that it was much more loosened by the initial long axis stretch, it was able to slide radially. That radial slide was in sympathy with the centrifugal force of the fast (circa 2-hour) comet rotation. 

The centrifugal force wasn’t enough for this crust near the north pole to escape but the next best thing was for it to slide to a higher radius. That radius was at a point where it could conserve angular momentum and the slide would be in a line directly away from the north pole. It would explain why all four of the cuboid lumps ended up with their bottom edges in a straight line overhanging Aten. That was the desired radius from the rotation axis for AM conservation.

The initial delamination along the long axis was in banded layers that were delaminating like the blades of a Venetian blind. Those delaminated layers ended up as four bands across Babi running lengthwise from Hapi to Aten. They delaminated in the direction of Aker at the comet’s long axis extremity. Those four bands correspond roughly to the five red lines in the header (five lines needed to enclose four bands). However the red lines are really the slide vector and the actual bands were at the angle which is now preserved in the yellow lines. The gull wings were simply slightly more recognisable features at the ends of each delaminated band, spread along the shear line. They are more recognisable due to being pushed into a double lump (gull wing) by slurry emerging at the shear line prior to delaminating into 5 sets. The slurry emergence was described in Part 7.

Notice that the left hand red line doesn’t emanate from a gull wing set and, by the same token, the right hand red line emanates from the second set not the first. This means that instead of it being a perfectly tidy scenario of five gull wing sets enclosing four delaminated layers and their eventual cuboids, we have four sets (2,3,4 and 5) enclosing three of the delaminated layers/cuboids and a fourth layer/cuboid on the outside of the fifth gull wing set. So that left hand layer/cuboid was pulled or ‘delaminated out’ between the fifth gull wing set and Aker. There’s some evidence that there’s a sixth set of vestigial gull wings at Aker anyway but that’s not certain. It’s besides the point though because you don’t need gull wings to prove a delamination. The left hand cuboid, being of the same character as the first three shows that this last section between set 5 and Aker did indeed delaminate. Also the surface of Babi has the same characteristic look all the way to Aker, suggesting the same delaminating process was at play across the whole swathe of Babi. 

As for the gap between the first two gull wing sets which don’t have an accompanying cuboid, that’s because it appears not to have been a delamination (layers sliding like the card deck) but a straight tear and rifting apart. 

Once these delaminated crustal layers were loosened by that long axis slide towards Aker, they slid radially along the direction of the red lines. 

The cuboid faces slid from their former positions between the sets of gull wings. Each section slid along and between the red lines, preserving something of its original form. And each section scooped up its own delaminated band as it slid across Babi. Or rather than scooping, it slid across the top of the mini layers (‘shards of slate’) within that band and by the time they all reached Aten they were bunched together in a cuboid that was about three times the thickness of the original layer. That’s because the cuboid areas are roughly a third of the area of the delaminated ‘Venetian’ bands. What are now the faces of the cuboids overhanging Aten used to be between the gull wings and they stayed on top all the way along the slide. In fact the faces eventually overtook the layers below them which are today spread back across Babi in a slope behind the faces. The yellow lines run down that slope and stop at the bottom. 

Each yellow line preserves a vestigial signature of the original gull wing delamination (see unannotated versions of the header and of photos 7 and 8). That signature is in the form of larger chunks stepping up slightly more markedly than the thinner shards and doing so in a line from top to bottom of the slope. That causes a visible line running up the slope of shards to each cuboid vertex. The line is probably not just a signature the original gull wing signature but a signature of the edge of the entire ‘Venetian’ band. The bands delaminated along the long axis towards Aker. Only then did they slide out to Aten and bunch up along their length in the process. They bunched up into the sloping shards which were now a third of the length of the original band. Therefore the Venetian layer edges also bunched up to a third of their original length but still managed to remain as a step (the yellow lines) because they were of course stepped when originally delaminated as bands across Babi towards Aker. 

The slope of shards that’s bounded by the yellow lines consists of very fine mini delaminations of what was ostensibly a single top onion layer. That would be the ‘level 1’ onion layer as described in Part 39. The mini delaminations, being so numerous, might have facilitated the slide just like the analogy of the card deck sliding across the croupier’s table. They’re more evident in photo 8.

Photo 2- side view 


In photo 2, the left hand red line is omitted. Also, the dark blue and brown dots for the north pole are hovering just above their actual positions in Hapi behind the rim of the shear line. All other colours are as for the header photo.

Photo 3- underside view. 


Photo 3 is a four-frame composite NAVCAM image. The top left frame shows the left hand three cuboid faces and the beginning of the fourth, right hand one (you can see its dot chopped in half). The top right frame shows the whole of that right hand cuboid face and a bit of the next one to its left. Aten is outlined in orange on all four frames so as to build up an impression of its extent. 

You can see from this photo that the four cuboid lumps are very distinct, protruding much further out over the Aten depression than the rest of the Aten perimeter. This is because they all slid to a radius above the rotation axis that allowed them to conserve angular momentum. That radius was irrespective of the Aten perimeter. They overshot the perimeter and that is why they look alien to the rest of Aten. 

An OSIRIS photo shows the base of the cuboids inside the Aten depression to have an obvious crack running along the bottom. This lends weight to the idea that they were dumped there and aren’t really integral or native to the Aten perimeter. This side of the four cuboids that drops down into Aten is known as the ‘cliffs of Aten’ which is a recognition of their spectacularly tall and blocky morphology. 


OSIRIS scientists have puzzled over how the Aten depression came about, specifically how it could have lost its 0.1km cubed of material via sublimation alone. Explosive ejection of material has even been suggested (via pressurised pockets of gas). In other words, they wonder why it’s so deep. One reason it appears so deep is that one side has been built up by the four cuboid lumps that slid from Babi to above the original Aten rim. They should really be called the cliffs of Babi. The base of the cliffs is where the genuine, original Aten depression starts. 

And the original, shallower Aten ‘depression’ is of course very likely the crater left by a missing slab. That’s because it extends along exactly one side of the stretched diamond and is diametrically opposite the Anubis slab that also takes up exactly one side of the diamond. 

Why else would two diamond sides exhibit elongated craters that take up the exact length of a side? It’s consistent with the tangential, sliding departure of the slabs and not consistent with random sublimation- even when, rotation, obliquity and solar irradiance is invoked. Slabs sliding away from any one of the four sides will simply make that side more distinct. But no daily or seasonal irradiance pattern can explain the huge loss of material from exactly two diamond sides (Aten and Anubis) while leaving the diamond side between them completely intact and free of material loss (Ash).

The loss of the Aten slab may well have instigated the Babi landslide by leaving the delaminated layers unrestrained along their Aten edges. The Aten border is likely a signature of a threshold. It’s the cusp between a radius of rotation that induced a tangential escape velocity and a radius that would only allow radial slide. Anything above that radius would escape (tensile adhesion failure permitting). Anything that was sliding radially outwards from below that radius of rotation couldn’t slide beyond that cusp. The four cuboids therefore stopped at the Aten border and beyond the border, the Aten slab escaped. This is simplistic though because the radius of rotation varies along the Aten border. However, it becomes workable when bulk sections of crust, or massifs, such as the four cuboids are sliding about while clamped together. They would be sharing the above average and below average components of the tangential velocities. The same applies to the Aten and Anubis slabs.

The escape/slide cusp would also explain why all the mini delaminations in the four bands bunched up in such a well-behaved manner at the Aten border after their radial slide: those that were already at the Aten border didn’t need to slide further out because they were in equilibrium. Those layers a little way towards the gull wings (lower radius of rotation) needed just a small adjustment so they slid out and over the outermost layers to sit directly on top of them at the border too. Those layers that were at a still lower radius slid to the same Aten border cusp and ended up on top of the growing pile. Finally, the faces of the cuboids, originally sited between the gull wings slid the furthest but to the same Aten cusp. Adding in some discrepancies in friction between layers due to the lessening weight of upper layers and you have the overshooting of the faces and sloping yellow lines of deeper layers trailing behind them. There was probably some yanking too of the material at a lower radius between the gull wings by the material at a higher radius (see the ‘pie-shaped crust effect’ sub-heading in Part 34).


Photo 4- this is photo 3 with extra head annotations.  


Extra annotations on the head lobe in photo 4:

Yellow- the cove (Part 35) which constitutes part of the head rim shear line.

Terracotta- the rest of the head rim shear line that’s visible.

Dark green- the sites on the head rim underside where the third, fourth and fifth sets of gull wings on the body kissed. The third set is on the right, fifth on the left. The third set is the classic set that matches head cove to body. That is, the third set on the head was matched to the identically shaped third set on the body. This was done both in plan view and side view (Part 5).

Light blue- these are four subtle ridges on the head that correspond to the left hand three cuboids. In other words, those three cuboids used to be joined to these four lines (four lines enclose three interstices). So each cuboid was attached to each head interstice when it was still a delaminated ‘Venetian’ band spread across Babi. And of course, that would have been when the head was seated on the body. 

Since the cuboid faces are said to be the topmost layer that stayed on top all the way through the slide from the gull wings to Aten, it follows that the faces we see today, teetering over Aten, used to be attached to the head interstices. That is consistent with the fact that both the cuboid faces and the head lobe surface between the interstices are the same lighter colour.


The right hand cuboid appears to be related to a now-eradicated signature in the cove. That’s because its slide track is to the right of the third or classic gull wing set which defines the left tip of the cove. 

The four subtle head ridges are more pronounced under certain lighting. Their extent up the head corresponds to the cove delamination height (the fourth yellow curve in photo 6, below). There’s a sloping line that’s common to the top of the third and topmost cove scallop and to the top ends of these subtle, light blue ridges. That line is a head lobe crust recoil behaving in mirror fashion to all the radial body crust recoils we’ve seen from Parts 32 to 40. The recoil line is responsible for the cove delamination itself (Part 35). It will get an airing in due course. 

Photo 5- another view of the slid cuboids. It has the same key as for photos above.  


Photo 5 shows that the middle two cuboids don’t have uniform straight faces as they appear to have in other photos. They are kinked with the upper portion facing somewhat upwards in this view. Their two respective dots have been placed on these two slightly angled sub-facets. The lower facet counterparts on each of these two middle cuboids seem to be the same ‘height’ as the two smaller, flanking cuboids. That’s because all four seem to share a common ridge that runs across the tops of the flanking ones but halfway through the middle two. But this isn’t at all apparent in other photos. 

Photo 6- this is the same as photo 5 but with extra head lobe annotations.  


The head lobe dot colours are as for photo 4 but with a small difference. The cove shows all three cove scallops (four ribs enclosing three scallops). 

Photo 7- the overshoots (light pinky orange). 


Red- just the outside two slide vectors are shown.

Pinky orange- these are overhangs. In Part 39 (and above) it was suggested that if the delaminated layers that had spread across Babi then slid radially, a signature of that radial slide would be that the topmost layers might overshoot the rest of the sliding layers beneath them. This can be seen in photo 7 with each overhang pointing out along the slide vector from the north pole. The description of this phenomenon in Part 39 used the stack of cards analogy. The stack is first spread across the ‘table’ along the comet’s long axis but since there are a lot of cards (mini layers) they are still quite bunched up despite the long spread. That leaves the spread stack open to further delamination in another direction. The whole stack then slides out radially and the topmost cards slide that much further (delamination in the radial direction resulting in overshoot for the very top cards). The pinky orange protrusions didn’t just carry on and on sliding because they were travelling upwards in the gravity field to their preferred equilibrium point. They were behaving just like the cuboids they’re sitting on, stopping at the Aten ‘escape cusp’ but overshooting the rim slightly due to less weight and commensurately less friction. 


Photo 8- a similar view to photo 7 (this is the previous NAVCAM photo taken on that orbit). This is presented simply as an overview of all we’ve covered here but with a slightly different viewing angle and different lighting. 


Photo 8 shows the mini delaminations between the yellow lines in more detail. It also shows how the two outside cuboids slid further and symmetrically from their initial positions and yet still managed to stop along the same straight line as the middle two cuboids. This is potentially the signature of all four cuboids reaching AM equilibrium at this radius from the rotation axis- or rather, these radii because they all started and finished along a line of slightly different radii. 


This part concludes the major radial crust-sliding vectors across Seth and Babi around the north pole. They encompass an angle of about 120°.

Photo 9- the radial vectors at Babi and Seth.  


Large brown- the paleo pole adjustment.

Brown dotted line- the paleo y axis (sitting below the paleo long axis- see Part 39).

There will be other posts on radial vectors, for example, looking at the Site A delaminations in the same detail as in this post. But the general direction of the Site A radial slide vector has already been established with the two vectors at either end (sink hole delamination and monolithic slide) and the Ash recoil behind it. 

Photo 10- ESA regions map. 


UPDATE 13th February 2016

Photo 11 shows the yellow lines sloping up the delaminated pile that constitutes the bulk of material sitting behind the two middle cuboids.  


The cuboids’ faces are out of sight, over the horizon at the far end of the yellow lines. However, the first cuboid (fuchsia dot) that slid from between sets 2 and 3 of the gull wings is very visible as is the red ridge that denotes its slide vector. 

The visible gull wings (green) are the classic set (set 3) in the foreground and the left ‘wing’ of set 4 behind that. Set four is really just a double hump and not very visible even as the single, left hand half of the hump in this view. But it’s obvious as a hump in photo 2.

The yellow lines show the distinctive change in the delaminated layers. Each line denotes a step-down as we move across the delaminated debris field (or step-up if moving the opposite way). Each step-down corresponds to a cuboid vertex at the far end of the yellow lines and a gull wing set in the foreground (three step-downs enclosing the middle two cuboids). 

The gull wing sets don’t attach directly to the yellow lines but do so via the red slide vector lines in the header photo. Those red lines aren’t shown here because they’re too foreshortened to be ascertained. 

Even the exact path of the yellow lines is open to some interpretation due to foreshortening but they are undeniably there as evidenced by the unannotated version and the header photo (unannotated version as well). 

The fact that the yellow lines are related to the cuboids at one end and gull wings at the other shows that they retain the stepped down signature of the delaminated bands from which they are derived. That’s because each band once extended lengthwise from Hapi to Aten and the four bands delaminated like Venetian slats towards Aten. On sliding radially, the bands were scooped up along their length into the bulky pile we see here behind the cuboid faces. Each individual band (Venetian slat) was originally stepped up from the delaminated band below it and that’s why today’s bulky pile retains the stepped-up signature along its length, as depicted by the yellow lines. 

Photos 12/13 are very similar. Photo 13 is actually photo 11, reproduced. Photo 12 is a very subtly different view and is slightly clearer. It’s a shame it’s cut off on the right hand side but if you toggle between the two you can see that the stacked up pile behind the two middle cuboids is all there. Photo 12 shows up both the mini delaminations and the step-ups on the slope better than photo 11/13. However, photo 11/13 shows the step-ups along the top of the slope in better relief. 

photos 12/13


UDATE 8th April 2016


Yellow- visible slide tracks

Red- Part 1 matches

UPDATE 20th May 2016

This update concerns the 4th cuboid matches and slide.

Matches with slide track (red). Note the track is different from the April update above:

Without slide track:



(A further update was cut from here due to realisation that the dots were badly placed and the photo wasn’t therefore very useful).


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 39



Wavy terracotta line- this is the classic shear line along Hapi. It also runs off into the distance round to the back of Anuket and in doing so, it runs through the four coloured anchors which are also matches to the head lobe (Part 24). The mauve match has both of its ‘ears’ marked. The dark green match is practically out of sight behind the neck.

Straight terracotta line- a suggested initial shear in a lower onion layer, which is along the stretch vector line and from which the layer above (Site A and Babi) were shunted slightly.


The ESA regions map is photo 5 at the bottom of this post.

It should be noted that the shear line now includes the front rim of site A, that is, the lip of the flat area. This hasn’t been included before because the missing slab above it was what matched to the head rim but that was no longer in place. That meant the Site A rim couldn’t logically match to the head rim. However, the Site A rim shows sufficient matching to a line just below the head rim. That evidence won’t be presented for a while due to a backlog of stretch vector posts. But we have to move ahead with related phenomena and this added section across Site A helps to inform the arguments for those phenomena. It’s the last 1-kilometre section in an already well-matched 7 to 8-kilometre shear line.

The straight terracotta line is a line that links the outer lips of the two obvious, lower sections of onion layer in Hapi. The line is then extended towards us to what appears to be part of the same lower, sheared layer running in line with the other two sections. That section stops at the marked curve in the classic shear line in the foreground. The curve is the exact site of the fifth and last gull wing delamination, with one wing kissing the curve (although the gull wings are more like a hump with two pimples at the fifth delamination). As we shall see in an upcoming post, the other end of this straight line stops dead at the point of the furthest delamination in the opposite direction. This delamination and the fifth gull wing set are almost equidistant from the north pole and exactly equidistant from the paleo pole adjustment point, insofar as the paleo adjustment can be said to be accurate (see Part 37). This is strong evidence that the straight line was an inner shear line brought about by the long axis stretch vector. The line is parallel to the paleo y axis (adjusted from the current y axis) and the paleo y axis runs through the paleo rotation plane. Thus, the straight line is exactly along the long axis stretch vector when looking from above (but not quite in line from the side view in upright duck mode- see below). This alignment should be no surprise because the straight line is clearly parallel to the line along which all the crust delaminations occurred at the Hapi shear line. The lower layer preserves the straightness of that line better than the classic shear line presumably because it was somewhat pinned under the onion layer above it (Site A and Babi). It would have been pinned even though Site A and Babi were themselves sliding back, as one layer and and parallel, on that pinned layer.

The idea of Site A and Babi sliding albeit just a short distance, is a new proposition in this blog just in case you thought you’d missed it. The slide was short and translational for the full length of the straight line, not radial like the top layer crust delaminations. You can see that from the fact that the classic shear line and the newly proposed straight tear are close and parallel. 

That short slide was a shunt back of a middle onion layer, which is actually site A and Babi. Although that ‘middle’ layer is today’s top layer of crust, we know from previous parts that there was another layer on top in the form of the two missing slabs. Those slabs are missing slabs A and B (Part 9) and were the former top layer. They were said to have escaped in Part 9. However, given the amount of radial sliding signatures that we’ve been discussing in recent posts, it appears that slabs A and B didn’t escape but slid and bunched up dramatically. It was the delamination of the main sink hole on the body into three (Part 32) that made me realise there might be slide signatures for the whole site A area and even Babi. And indeed those signatures are there (see the “implications for slabs A and B” heading further down). So we now have three levels of crust or onion layers when considering the shear line and crust sliding at Hapi. 


The three levels are described separately below. Levels 1 and 2 are quite long. Level 2 has a photo too.

Level 1:

This is the classic top layer sliding radially as described in Parts 32 to 38. It includes the delamination of the three sink holes on the body at site A (Part 32); the Ash slide (Part 32); the monolithic slide out of the horseshoe (Part 33); the Babi slide (to be presented in Part 40). Also, the red triangle from Part 26. The result of this sliding back of the level one onion layer was that it revealed the level 2 onion layer beneath.

Although level 1 slid back radially it also delaminated along with the gull wings prior to the radial slide. It had to do this in order to be loosened enough to slide radially. The level 2 gull wing delamination was in the long axis stretch vector direction which is between 30° and 90° to the various level 1 radial slide vectors around the pole. Level 1 behaved exactly like level 2 below it as they delaminated together along the long axis. Only then did level 1 slide radially. 

So it would seem the only thing that distinguishes level 1 from level 2 is the very fact that level 1 was loosened enough to slide radially. The deeper layers were less loosened, being deeper into the comet and those, by definition became level 2. The level 2 surface is the present-day surface at Babi (including gull wings) and the flat part of Site A. This is why Babi looks riven, as if scalped, and it’s probably why the fourth and fifth gull wing sets are smooth bumps with pairs of pimples, which gives the impression of material having been flayed from above them.

That said, it seems likely that the dividing line between being loosened enough to slide, or not, was governed by the onion layers. Babi is quite smooth and Site A very smooth. That suggests a decoupling along the fracture line between two onion layers served as the weak spot between loosened enough and not loosened enough crust. It would also explain the ease with which the loosened top crust slid if it slid on its own smooth onion layer fracture plane. 

The idea of delaminating one way and then sliding off in a completely different direction isn’t pure speculation. There are signatures of this all round the rims of Site A and Babi. The card pack sliding analogy fits perfectly. The stack was slid one way (long axis delamination) with a slight bunching of cards at each gull wing. Then the whole delaminated stack slid radially and up to 90° from the initial card slide. It would be nominally a line of cards moving outwards but arcing under the radial influence. It would still have cards that were very much overlapping but nevertheless clearly delaminated. When the line of cards had slid and arced out to a new, higher radius from the rotation axis (further from the pole) it stopped again, but not before the topmost cards had slid a little further radially, jutting out further over the rims of both Site A and Babi. That signature is actually there at the rims and, once you see it, it’s very obvious indeed. The configuration even retains the initial long axis delamination signature in the form of radial lines across the two arced lines of stacked card. That would be one radial line across the arc for each gull wing in the case of Babi. The lines are there too at the back of Site A but not currently traceable to the shear line at Site A where there are no gull wing type delaminations. Those may be on the head lobe. Close up photos of the two arced rims of cards, one for Babi, one for Site A, will be shown in Part 40 and beyond. 

Seeing as levels 1 and 2 may have been separate onion layers we shall assume that to be the case for ease of explanation and they will be referred to as such as well as levels 1 and 2. It doesn’t make any difference to the sliding mechanisms if they were actually one onion layer that sheared into levels 1 and 2. In both cases it involves the layer or layers delaminating into many finer layers (the ‘cards’) in one direction then sliding radially in another direction. Level 3 seems even more distinct as a separate onion layer. It appears to be about the same thickness as the other two, which is an indication that they are indeed separate.

Level 2:

This is the middle layer that was predominantly delaminating along the long axis. That was the gull wing delamination along with their light blue ridges in Part 39. But this onion layer also shunted back a little way in sympathy with the radial slide of the top layer, level 1. The short shunt of level 2 is depicted in photo 2, a stunningly symmetrical illustration of the radial vectors around the pole and their close relationship with the straight section of shear line.

Photo 2- radial stretch vectors around the pole. This is a top-down view with the head lobe in the foreground. 


(the red arrow should be ignored. It’s the current x axis which is irrelevant here). 

Wavy terracotta- the shear line, which is the rim of level 2. Level 2 was originally the middle onion layer at Site A and Babi but is today’s top layer after level 1 slid back radially.

Straight terracotta- these smaller dots depict the straight line tear of the lower level, level 3.

Red dots- radial stretch vectors. Progressing clockwise: the bright green ridge (Parts 22-29); the three delaminated sink holes on Site A; The Ash slide matches (slightly angled to radial); the monolithic slide; two of four Babi radial stretch vectors, due in Part 40. The sink hole line and monolithic slide line are extended to the comet’s horizon in this view so as to make the radial stretch vector pattern more evident. The other four actually do extend to the horizon. The sink hole line may extend that far anyway if the large, isolated crater on Ash is related to the sink hole. That’s somewhat speculative but it did slide almost that far, a fact that’s betrayed by being entwined with the Ash slide matches (see Part 32). 

Brown lines- the paleo y axis and paleo z rotation axis. They’re adjusted from the green and blue lines respectively because the rotation plane precessed by this amount. Notice how the paleo y axis is exactly parallel to the straight line tear. The paleo y axis is closely related to the paleo long axis since they are both in the same plane, at right angles to the paleo rotation axis and stacked almost one above the other when looking straight down from the top of the head.

Large brown- this is the paleo pole preliminary adjustment (see Part 37). The paleo rotation axis runs through it when viewed from above, as in this view. But when the head was on the body, it would have in fact run in line with it but below it. See part 37 for the paleo pole adjustments. You can see that the brown dot is at the exact mid point along the straight terracotta tear line and central to the whole symmetrical picture of radial vectors.


Photo 2 shows how (a) the radial force vectors of level 1 are focussed on the north pole or just beyond it and even more accurately so on the paleo adjustment point (b) their average direction tugged level 2 (with the classic shear line at its rim) at 90° away from the straight line tear on level 3, the lowest level. The result was that the shear line ended up parallel to the straight tear even though it became wavy at a finer scale due to vagaries of the local force vectors in the tug. It’s also wavy due to the local matrix weaknesses dictating the shear line tear path around stronger ridges like the horseshoe.  

Although the shear line shunted in sympathy with the radial sliding of the top layer it didn’t exhibit any radial behaviour or at least, almost none. Its shunt was a short translational slide at 90° to the straight line tear of the layer below. That left it parallel to the straight tear. This probably happened because it was presumably tugged by the average radial vector fanning symmetrically around the pole (about 120° for the two radial stretch vectors at either end of the straight line tear). That average force vector shunted the level 2 shear line away at 90° to the original straight tear, leaving it parallel to it. The average force vector was also at 90° to the paleo y axis and therefore it acted along a line parallel to the paleo rotation axis. This scenario is especially likely for two reasons. Firstly, the length of the shunted section extended along a path that straddled the pole symmetrically and so this layer was tugged symmetrically by the 120° of radial vectors. Secondly, the extent of the line along which the shunt happened is exactly the same line as the straight tear and that line in turn comprises the delaminating layers and radial vectors emanating from it. With the exception of the red triangle, a separate entity as we’ll eventually see, there are no radial vectors, delaminations or straight tearing beyond the two ends of the straight terracotta line (the other end is Aker which is virtually unstretched crust riding on the tip of the stretching diamond shape). The parallel shunt of level 2 sits wholly inside this highly constrained geometric setup and extends along its full length, no more, no less.

Level 2 includes the gull wings, which delaminated along their long axis stretch vector line. It was along the long axis because it was directed by the comet’s core stretching underneath. That vector line is at 90° to the outward layer shunt away from Hapi. It also includes the entire flat area of Site A. As for the Site A portion of this second level, it doesn’t look as though it delaminated like the gull wings but that flat area appears to be under a lot of dust, possibly disguising signatures. However, the front rim of Site A doesn’t betray delaminations either whereas its mirror counterpart on the other side of the pole (and along the opposite half of the shunted section of shear line) exhibits the gull wing delaminations on its rim. That said, Site A may have simply stretched and that would be a neat explanation for its uncanny flatness. There are reasons for thinking it performed such a stretch and those reasons will be presented in a future post.  

Level 3:

The third layer is the deepest of the three onion layers, peeping out from under the shunted middle layer, level 2. It exhibits the straight line tear along its rim that faces Hapi. The straight line is along the tensile stretch force vector. Although the three sections that make it up have gaps between them, making this level look like a sporadic assortment of jutting out chunks, at least two match to each other. They are the quasi rectangular section below the Site A rim and the section below the horseshoe crater. Their matching faces are the fuchsia V shapes in Parts 35-37. When joined, they would have made one long strip on one side of the pole. Their tearing and sliding apart along the straight line shouldn’t be a surprise because much or all of the second layer delaminated in the same direction along Hapi. After all, if the comet stretched along that straight line in Hapi before actually causing this massive shear stress fracture, level 3 would be expected to have either delaminated or torn into sections. It seems that these two sections tore. They are the furthest two sections along the straight terracotta tear line in the header photo. That tear left the matchable V-shaped gap between these two sections. 

The other half of level 3 along the straight terracotta line seems to have delaminated instead of tearing in two. That’s the two fuchsia India shapes (Parts 35-37) and the stepped feature extending along to the fifth gull wing set. 

It’s interesting that the pole was the dividing point between tearing on one half of the straight line tear and delamination on the other half. I have no explanation for that yet, but it will probably inform us in due course regarding the flatness of Site A. 

That concludes the description of the three onion layers, levels 1-3.


Looked at in the round, it looks as if the cometary matrix along the straight tear line of the third level onion layer used to take up something approaching half the length of the straight line. And that’s why it now consists of sparsely spaced chunks. It had to develop gaps in order to stretch along Hapi and along the long axis vector. If it originally took up half the length, it would mean the comet was originally shorter along its long axis to the tune of about 1km. That mass would have to go somewhere if we reversed the stretch of the comet. The short axis widening to accommodate it is the obvious candidate. This is especially likely seeing as the existing shape of the body is an elegant, elongated diamond that got that way due to stretch. The converse is that it had to have stretched from a stubbier diamond, even a former square. And seeing as stretch tensile vectors likely caused the appearance of the straight sides of the diamond (formerly a stubby diamond) it probably used to be more rounded. It was probably just a stubby potato shape with a faintly protruding proto head lobe lump that was ripe for future herniation on spin-up.


The direction of the shear line along Hapi is predominantly along the long axis direction. The line also happens to be running away radially from the pole in both directions. It’s actually straddling the pole diametrically along Hapi, which is why it can be both straight and radial from the pole at the same time. It’s no coincidence that the shear line along Hapi is parallel, or nearly parallel, to the long axis. It’s also at 90° to the rotation axis. The main stretching of the comet’s core under the surface crust was along the long axis due to that axis being the most susceptible axis to stretch under spin-up. That’s why the delamination of the gull wings happened along the that line. The crust couldn’t stretch by over a kilometre in the long axis direction to accommodate the core stretch, so it delaminated (or performed straightforward tears) in that direction instead. It was the core’s long axis stretching that was causing the gull wings’ direction of delamination along the soon-to-be shear line. That presumably caused shear stresses in the long axis direction as well i.e. causing stress fractures in that direction and not in some other random direction. That would eventually become the shear line. Or perhaps it was just one gigantic stress fracture, the shear line itself succumbing as the weakest point at a very early stage, and without any collateral stress fractures nearby. That’s why the Hapi shear line runs in the long axis direction. It’s also why the neck’s cross section is comparatively long and thin. It’s somewhat longer in the long axis direction. 

Obviously, to have a head shearing away, the shear line had to curve round at both ends which would be across the long axis direction. That might seem to contradict what’s laid out above. However, the point being made here is that the shear line on the north pole side and, it seems, on the south pole side, was longer and straighter than the two curved ends and that the length and straightness were due to the stretch vector. The south pole pictures released so far seem to support this. 

The fact that the shear line crosses the long axis where it does at either end of the neck would be due to random factors to do with head herniation from the body. Those factors would be the initial potato shape and the relative tensile strengths of different areas of core and crust. These would dictate the line along which the head started herniating, front and back. They would also dictate up to a point where it started herniating at the sides. But those side lines would also be strongly affected by the stretch vector. They would be moderated into longer, straighter lines as the comet stretched whereas the two curving end tears would remain the same width, or even narrow slightly under stretch. They would also remain curved or become slightly more curved under stretch. This all follows from the fact that any initial, quasi circular shear line will be stretched into an oval or, as it appears to be on 67P, a stubby rectangle with very curved ends. That would be the cross sectional profile of the neck if we sliced through it to take a look. And the straightening of the oval’s sides would be due to the overwhelming influence of the stretch vector along the long axis. 

Incidentally, the influence of the same force that straightened the sides of the shear line may have been so strong as to slice the vertical wall onion layer on either side to create the vertical wall itself (Part 27). That would explain why it symmetrically straddles the paleo rotation plane and is comparable to the width of the neck. However, this aside is a work in progress because the width of the neck and vertical wall isn’t the same thing as the width of the shear line oval due to the fact that the neck was extruded and narrowed as the lobes separated after shear.

Only after the long axis, core-directed delamination of all three levels, 1 to 3, did the level 1 pieces of crust start sliding away from the north pole. Level 1 was the top layer and it was now loosened enough to slide radially in all directions down Seth, Ash and (as we shall see, Babi). Those were mostly very different directions from the initial long axis delamination stretch vector. In fact, the definition of level 1 is that it was the layer which was loosened enough to slide. At some depth it wouldn’t be loose enough to slide and that depth defines the top of level 2 which is today’s Site A and Babi. 

The following photos, 3 and 4, serve to illustrate the geometry of the intimate relationship between the straight shear line at Hapi, the long axis and the rotation axis. Photo 3 is included to show that, from the side, the Hapi stretch vector isn’t actually quite in line with the long axis. However, it’s nearly as close as it can be because it’s fairly close to parallel when viewed from the side and exactly parallel when viewed from above. And although the stretch vector isn’t quite parallel overall, the straight tear in Hapi straddles the rotation axis (and pole) exactly. That’s strong evidence that it was centrifugal stretch forces along the long axis that brought about the straight shear. And that would be why it’s also almost in line with the long axis. 

Photo 3- the long axis viewed from the side and back. 


Dark orange- the current long axis.

Light orange- the paleo long axis, estimated for when the head sat on the body. It passes below the current rotation axis (blue) because the head was 1000 metres lower when seated. In the view from behind, this fact makes the light orange line appear to be ‘to far’ to the left at first glance. But it’s deeper into the comet than it appears to be. That’s because our eye is drawn to the blue rotation axis line and assumes it runs through the axis and not below it. 

Terracotta- the shear line. 

Photo 4- the long axis direction of the gull wing delamination. This also includes the two fuchsia V shapes from Part 37. 

Brown line- the paleo rotation plane (see the relevant page in the menu bar and Part 26 onwards)

Dark blue line- (very small dots to the right of the brown line). The current rotation plane or equator of the comet. 

Single dark blue dot- this larger blue dot is today’s north pole 

Two brown dots- the right hand one of these two larger brown dots (next to the blue dot) is the preliminary adjustment of the paleo pole from the north pole (see Part 37). The other one is the estimated position of the actual paleo pole which has to be somewhere down the body along a line that is at 90° to the paleo rotation plane. It’s positioned very slightly differently from its position in Part 37 due to a better view of the paleo rotation plane here.

Dark green- the delaminated gull wing sets. Firstly, there are only three sets shown instead of five. That’s because the first set, nearest to us, is in fact the first three sets attached together at the horseshoe where they started. Their current position is skewed towards Ash from their original delamination vector due to the monolithic slide (Part 33). Secondly, due to the scale, there aren’t pairs of green dots because they’d be on top of each other. The three we see are kissing the shear line so they nominally denote the wing on the shear line side because the sets of wings themselves kiss the shear line. 

Fuchsia- these two dots correspond to the fuchsia v shapes from Part 37. They aren’t the India shapes but the clean tear of the lower onion layer, level 3. The dots aren’t on the V shapes which are only important as matches. They’re still sited either side of the same gap but are nestled to the end of it, against the shear line because we’re looking at the direction of the delamination vector along the shear line. Seeing as the two fuchsia dots are the same distance from their matched V’s while still kissing the shear line we’re not fudging the force vector along the shear line here. This has all been ascertained from a multitude of close up NAVCAM photos from different angles, not from this grainy shape model.

Red- this is an old photo so the red dots aren’t significant for the stretch vectors. They’re interesting in that they show the well-defined diamond shape of the body, although only one end of the diamond was annotated in this case. That’s because it was used to show how the V- shapes on the head match to the body diamond shape which is itself, of course, a V at each end. You can even see the oft-mentioned 15° anticlockwise head lobe rotation clearly displayed here. It’s apparent in the difference between the V-shape orientation angles on head and body. It’s also very obvious in the path of the paleo rotation plane (brown). There’s a distinct angle between its path on the head and on the body. The paleo plane is the centreline of the V’s anyway but its path is defined by eleven stretch signatures running round the whole comet of which these V’s are only one. 


The main point of photo 4 is to show that the line of the three green and two fuchsia dots and their associated straight terracotta line are all parallel to the brown rotation plane line. The rotation plane line has the long axis running between its two furthest extremities (at Apis and the Hatmehit/Bastet border) so it’s closely related to our stretch vector studies. But it doesn’t look quite as convincing as the top-down photo above. However, it has to be remembered that the brown line on the body is going ‘uphill’ all the time and never gets a chance to arc over to direct itself along the level of Hapi. Instead, it starts to arc over a bit and then gets flipped upwards at the neck base. So it looks more off than it really might be. If it could be arced right over and directed through the neck, staying within its plane and following a line that was the same ‘level’ as the straight terracotta line, it would indeed be parallel to that terracotta line. That’s the assumption because the top-down photo is showing it to be so.

Since the long axis also runs through the rotation plane it means that the five delaminations depicted are nearly parallel to the long axis stretch vector. And this jumps to seven delaminations when the nested triad at the horseshoe rim is included. All seven were in a roughly straight line along the long axis stretch vector. Two of them (gull wing sets 1 and 2) subsequently got yanked back by the monolithic slide which was radial from the north pole and towards Ash rather than instigated by the long axis stretch vector. See the next section for an explanation of this difference between the gull wing delamination vector and subsequent radial crust-sliding vectors.


The surface crust-sliding was radial, away from the pole, as we know. The long axis stretching of the core was also radial because the long axis runs through the z-axis of rotation. Except there’s a difference. Really, the core stretch was in the two long axis directions that are opposite to each other and away from the z axis of rotation, the axle of the comet. That’s because the core is inside the comet, not sitting around on the surface near the pole like the three levels of crust we’re looking at here. So core stretch was, strictly speaking, not away from the pole itself- each chunk of buried core was stretching away from whichever part of the rotation axis it was rotating around. The whole core was stretching away from the whole rotation axis. 

Since the core was wholly dominant in the stretching process, the crust had to do whatever it could to keep up. It was too brittle to stretch but it was clearly composed of onion layers so those layers delaminated in line with the predominant stretch force vector, which was along the long axis and *not* radially away from the north pole. Hence, any actual layer delamination was in the long axis direction. Only after that did the now-loosened crust (level 1) respond to the radial vectors on the surface of the comet. Only that layer (and not the core or lower layers) was sitting loosened on the surface and so it was the only layer that could slide radially. Before delamination, the integrity of the onion layers was perfectly adequate to resist these radial forces on the surface. But after delamination, the top level slid away radially from the pole. And crucially, it wasn’t just the actual gull wings that delaminated, it was the blue ridges they were a part of (Part 38) that slid across the Babi long axis direction as far as Aker. The whole area of Babi was delaminated in the long axis direction. The width of the delamination was from Aten to the shear line. The gull wings were just the most obvious features, perched on the ends of the delaminated layers at the shear line. The layers delaminated in the long axis direction but as we shall see, the loosened crust then immediately slid radially. 

You may be wondering why we’re talking about crust sliding radially when it’s clear from the Part 38 annotations that the light blue ridges prove the delaminated layers are still spread out along the long axis direction. However, as stated above, the current Babi surface is in fact level 2. Level 1 was above this layer, was loosened, and is now missing. Babi is known to be a gravitationally low region in comparison to Ash and Seth. That was stated in one of the OSIRIS morphology papers and it’s fairly obvious just looking at it. That’s why it was always cited on this blog as being the crater left by missing slab B. Described in Part 9, slab B was said to have departed the comet at escape velocity. And it had to have had a kick from the detaching head lobe to do so because its radial velocity close to the pole wasn’t enough to eject it from the comet at escape velocity (0.8 metres per second). If there’s a missing slab it means by definition that there was more crust sitting on top of the present day Babi region. It follows that the annotated blue ridges in Part 38 were originally sitting well below the surface and did delaminate but weren’t loosened enough to start sliding radially from the pole. There was a depth limit at which the delaminating crust was loosened enough to slide radially and that depth limit was the present-day Babi surface. In fact, you’d expect some sympathetic attempt by this layer to slide radially as well, along with its companion layers above and we do see that. It’s apparent in the fanning of three of the delaminated layers, which means by definition that two at least were arcing round to the radial vector towards Aten. And the first two sets of gull wings drifted back in sympathy with the monolithic slide towards Ash. These current surface layers on Babi, as annotated in Part 38, were the first long axis delaminated layers that were deep enough to survive the subsequent radial slide. But they still eased their way in that direction a tad. Their direction of movement betrays the radial direction of the full-blown level 1 slide above them. 

As for that level 1 slide which involved the rest of the crust sitting loosened above the gull wings and their associated delaminated ridges, it slid in the most spectacular fashion and in its entirety across Babi and towards Aten. It never did escape, after all. It became the cliffs of Aten: four cuboid lumps corresponding to the gull wing delaminations


Technically speaking, the loosened sections of crust were also trying to move in a line that was away from the actual axis of rotation and not the pole. And if there was zero gravity and cohesion they would have left in such a radial direction (notwithstanding an initial tangential component if detaching and escaping). But there was significant enough gravity and cohesion, which meant that the net effect of these sections of crust trying to fly away from the rotation axis actually meant sliding away radially from the north pole. That was due to the gravitational and cohesive forces keeping the pieces of crust stuck to the comet but nevertheless sliding up to a larger radius of rotation. This was especially the case for all this crust we’re looking at around the north pole. This crust, near the axis of spin, was less susceptible to being flung away outright into free space, from centrifugal force alone (like the Imhotep and Hatmehit slabs were). So the next best option was to slide to a larger radius. And for a piece of crust consigned to staying stuck to the comet, the quickest path to a larger radius is radially, away from the pole. That’s why all the crust sliding and delamination vectors across Seth, Ash and Babi are radial, away from the pole. Beyond this area of sliding but not escaping is Imhotep and the Imhotep slab escaped because it’s at the longest radius from the axis. 


This clearly has implications for missing slabs A and B (Part 9). Those are the famous (in this blog) missing slabs on Site A and Babi. Site A and Babi are close to the north pole and it was always known that the centrifugal force of a two-hour spin-up was nowhere near enough to eject them from the comet. So it was suggested that they were flipped up by the head shearing, “like yanking a bollard out of a slab pavement and watching the slabs getting flipped upwards around it”. That wasn’t all that satisfactory though, because the head itself probably never quite reached escape velocity. It obviously didn’t escape but could conceivably have sheared at just above escape velocity before being attenuated by the neck. However, the tensile strength of the neck is paltry compared with the force per square metre exerted by the head lobe on the neck even when departing at just below escape velocity (270 pascals exerted vs 50 pascals of tensile attenuation- a calculation but with no more than informed guesses for head lobe mass and neck cross-sectional area). That 50 pascals leaves a very small margin for being just above escape velocity on shear but getting attenuated enough not to escape. It’s also an absolute maximum as ascertained by various scientific papers on 67P and other comets. Moreover, it’s tricky to get the head to flip the slabs to escape velocity from one side only.

And yet, there are clearly two missing slabs on this side of the comet. Well, they’ve now been found. Slab A didn’t escape- it’s all that bunched-up material at the back of Site A. It slid radially to the back of the crater.

And the missing Babi slab is the crust described above that was delaminated on the long axis vector but by doing so, was loosened enough to slide radially. The whole lot slid the best part of a kilometre across Babi to a satisfactorily higher radius and is now sits as four cuboid lumps teetering over the Aten depression. 

The evidence for the radial Babi slide, with photos, will be presented in Part 40.

So Part 9 gets full marks for noticing the slabs were missing but zero for thinking they’d escaped. It’s taken over a year to garner enough information about the head shear, sliding layers, head stretch before shearing, stretch vectors, and much more, to be able to realise where the missing material went. And it’s been right under our noses for all that time, just like every other discovery and doubtless, more to come. Still, 39 steps forward, one step back, can’t be bad. 


Marco Parigi has a stretch theory blog as well. It describes many of the aspects of stretch in more concise terms than here and with annotated photos. It then links to the relevant posts in this blog for those readers who want the full, in-depth analysis:


Marco thought of stretch theory, as it would be applied to comets in general, and did so long before Rosetta arrived at 67P-CG. My first realisation regarding the possibility of stretch was on seeing the first published close-up of 67P on August 6th 2014. 

Photo 5- The ESA regions. 



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

To view a copy of this licence please visit:


All dotted annotations by Scute1133.