The Case of Boulder AP3-172, Majes, Peru

Print Friendly, PDF & Email

This paper describes a huge boulder at Alto de Pitis in the Majes Valley of southern Peru. It has some enigmatic petroglyphs that will be fully discussed. Especially one type of image is most idiosyncratic. It may depict or symbolise domestication of felines.

By Maarten van Hoek



The Case of Boulder AP3-172, Alto de Pitis

Majes Valley, Southern Peru


Maarten van Hoek



Alto de Pitis is a very important rock art site in the Majes Valley of southern Peru. Yet it is almost unknown, especially when compared with its “legendary” neighbour, Toro Muerto, which is located on the other side of the valley, about 7 km WNW of Alto de Pitis (Figure 1). Because “celebrated” Toro Muerto is signposted along the road from Arequipa and marketed online, hardly anybody visits Alto de Pitis, although all visitors to Toro Muerto unknowingly cross the site of Alto de Pitis. Boulder AP2-001 is even only 25 m north of the Arequipa – Toro Muerto highway (the construction of which destroyed part of the rock art site).

Figure 1: Location of Alto de Pitis in the Central Majes Valley, southern Peru. Map © by Maarten van Hoek, based on the map © by OpenStreetMap – Contributors.

Click on any illustration to see an enlargement.

The whole complex of Alto de Pitis measures about 4000 m from south to north by 700 m maximum from west to east. During several surveys my wife Elles and I documented over 400 boulders with petroglyphs at Alto de Pitis (all of rather soft volcanic rock). Many of the petroglyphs at this extensive site are unique, not only for Alto de Pitis, the Majes Valley and Arequipa, but also for the whole of the Desert Andes, the extremely dry desert area west of the High Andes. In a previous paper I discussed the images on Boulder AP1-001, which featured some exceptional petroglyphs (Van Hoek 2022a). In  this study I focus on another very large boulder at Alto de Pitis, Boulder AP3-172 in Sector AP3-B (Figure 2) (see Van Hoek 2013: Fig. 65 for the location of the four Sections; the boundaries between the Sections AP3-A to E are not shown and their locations are only arbitrarily indicated).

Figure 2. Map of Alto de Pitis. Map © by Maarten van Hoek, based on Google Earth 2021.

The subject of this paper is an enormous block of stone with numerous petroglyphs. Unfortunately it is impossible to describe all petroglyphs on Boulder AP3-172 in detail. Many of the petroglyphs cannot be distinguished anymore since they have severely been blurred because of the strong solar radiation and the destructive erosion of the often sand-particles-laden wind that especially eroded the lower parts of the boulder. Moreover, also the many petroglyphs, including superimpositions, blur the picture even more. Therefore only some specific images on Boulder AP3-172 will be discussed in more detail and will be put into a wider context.


The Location

Boulder AP3-172 is located at 461 m asl (in Google Earth 2021), 47 km inland and about 410 m due east of the edge of the east bank of the Majes Valley at “La Curva”; the main Arequipa-Corire highway that runs through the village of Mezana or Mesana (valley floor at 391 m; see Figure 2). Thus, Boulder AP3-172 sits about 70 m above the valley floor, but is invisible from the valley floor, also because a low ridge to the west of the boulder blocks the view. However, the boulder is so large that – despite the undulating landscape – it can easily be seen from several points further east. Figure 3 illustrates the view of Boulder AP3-172 from Boulder AP3-040 across – for instance – Boulders AP3-170 (and AP3-171; see Figure 2). This prospect also includes glimpses of the Majes Valley and the important Red Spot at Punta Colorada (Van Hoek 2013: 139), where more rock art was discovered recently (Van Hoek 2022a: 3). Apu Coropuna is not visible from Boulder AP3-172 (it must be mentioned that this Sacred Mountain is also visible from a second rock art site in Majes: Punta Colorada: Van Hoek forthcoming).

Figure 3. View from Boulder AP3-040 (at 505 m asl; note the wind erosion) looking towards AP3-172, which is visible about 468 m to the SW. Photograph © by Maarten van Hoek.


Boulder AP3-172

The boulder was probably first recorded in a publication by archaeologist Paúl Jofrey Álvarez Zeballos (2009), who described “seven” panels, but unfortunately only (partially) illustrated three panels. One of the reasons for not illustrating all panels is that he states that many panels have weathered considerably. However, he does not illustrate Panel A, which has very clear and important petroglyphs. Moreover, in my opinion his Panel 3 and Panel 5 actually form one large panel (Panel A in this study). Regrettably he also does not provide any scale of the boulder (and only of one – almost inaccessible – petroglyph), while it is obvious that Boulder AP3-172 is huge. This is evident when comparing its size with the “dwarfed” figure of my wife Elles in Figure 4. Not having measured anything of this boulder myself, I still estimate the boulder to be about or even over 300 cm in height at certain spots.

Figure 4. Boulder AP3-172, looking north. Panel A (facing north) is invisible in this photo. My wife Elles is standing in front of a petroglyph panel (which is not discussed in this study). Photograph © by Maarten van Hoek.

Actually, Boulder AP3-172 may be considered to represent to be one split boulder, or it may be viewed as two separate boulders (the smaller part partially quarried?). In Figure 3 my wife is standing in front of a decorated panel on the south side of the smaller part of the “split” boulder, which will not be discussed here (this panel may have been described by Álvarez Zeballos as “his” panel 2). Because of the weathering (and also because we did not venture to climb the rock to inspect the top surface [Panel D], as was obviously done by Álvarez Zeballos; see his photograph of Panel C) this study only focusses on Panels A, B and C (see Figure 4).


The Petroglyphs

Although (severe) weathering and erosion has blurred many details, it is safe to state that most petroglyphs on Boulder AP3-172 have been made by pecking, although there are notable differences. The majority was probably manufactured by rather superficially pecking out the lines to a certain depth (about 1 cm), often followed by lightly abrading those lines (although this effect may also have been caused and/or enhanced by natural sand-blasting). However, several petroglyphs were clearly made by rather crudely hacking out the images. Especially these hacked images are (relatively) hardly patinated and may therefore represent the last (?) layer of rock art production. In some cases there may be older images that have been enhanced (re-worked by hacking) at a later stage (see for instance Figure 7: Element h). What will go unnoticed, is that the possible first layer of anthropic marking on Boulder AP3-172 comprises most of the many very thinly incised lines, many of which are superimposed by subsequent layers of rock art. Those incised lines do not seem to build (recognisable) images.

Unfortunately, there is a final, most destructive layer of “rock art”. Ignorant visitors often intentionally scratch rock art images with a sharp (metal) object, while others use chalk or some kind of paint to “enhance” the art. Also at Alto de Pitis I know of several petroglyphs that have been chalked-in, while the academic archaeologist who thus enhanced the petroglyphs even justified his malpractice twice (see next paragraph). This malpractice is evident in the publication by Peruvian archaeologist Paul Álvarez Zeballos (2009), in which he included several photos of petroglyphs at Alto de Pitis that he recorded and that he himself chalked-in. Also many petroglyphs on Boulder AP3-172 have been chalked-in, remnants of which are still visible in my photos of 2009.

Álvarez Zeballos (2009: 29: my emphasis) justifies his most unwanted malpractices as follows: “Para poder apreciar mejor las figuras se las marco con tizas alcalinas”. Besides his unprofessional conduct, Álvarez Zeballos even included photos of petroglyphs that have been chalked-in by him in an incorrect way, for instance Boulder AP4-017 (2009: 18), thus also publishing an incorrect drawing of his Boulder P-Ps-1: “Todas estas imágenes se encuentran semi-borradas y para resaltar su imagen se las repaso con tiza alcalina o básica.” (2009: 27: my emphasis). Chalking-in is not only detrimental, it also forces a subjective, often incorrect image upon subsequent observers. Any kind of on-site enhancement is destructive and should always be avoided and discouraged. Some respected rock art journals even categorically reject the publication of photos of chalked-in rock art images (for instance: Rock art Research; see also Bednarik 1987 in this respect).


The Panels


Panel AP3-172A

North facing Panel A is huge. It may measure up to 3 meters in height. Its vertical surface can be divided into two parts. The lower part is crowded with numerous (often much weathered) petroglyphs, whereas the upper part has only a few images. Most petroglyphs are also deeply patinated (and weathered/eroded), although a number of images – most of them arranged in a horizontal strip across the centre – stands out rather clearly, being much less patinated, displaying a rather whitish hue. These are the petroglyphs that are hacked out (or re-worked by hacking?). Some specific petroglyphs on this panel will now be discussed in random order. Please keep in mind that not all images are being discussed. The numbers in Figure 5 indicate one petroglyph (referred to as an Element) or more petroglyphs (referred to as a Group).

Figure 5. Panel AP3-172A. Photograph © by Maarten van Hoek.

Near the bottom of the panel is a small, fully pecked anthropomorphic figure with two short, parallel lines between the legs; a female? (Element 5 in Figure 5). In the centre are two small, outlined figures that just possible may depict anthropomorphs (Group 6 in Figure 5). Nearby is an unfinished quadruped (Element 7 in Figure 5) and further east is a distinct solar or stellar motif (Element 8 in Figure 5). High up the panel are an outlined cross framed by a square and a simple cross framed by a square (Group 9 in Figure 5). At the very top are a faint rectangle with three parallel lines from the top (a “Trophy” Head?) and a small quadruped (Group 10 in Figure 5).

Near the west edge is a large outlined square with three parallel lines from the top line and six or seven parallel lines from the bottom-line. The square has in internal cross and several crudely pecked (hacked?) dots. It most likely depicts a “Trophy” Head (Element 3 in Figure 6). A much smaller square also has six lines from its bottom line and an inverted U-shape from its top line (Element 4 in Figure 5). It most likely is another “Trophy” Head, also as it seems to have facial features and at least one small ear.

Figure 6. Details of Panel AP3-172A. Photographs © by Maarten van Hoek.

Figure 7. Details of Group 2 on Panel AP3-172A. Photograph © by Maarten van Hoek.

Across the centre is a group of petroglyphs that has clearly been hacked out (and often partially superimposes earlier images). Group 1 (see Figure 5) includes a very large hacked quadruped with an exceptionally long, curled tail. It is preceded by a smaller example and above the larger one is another, simpler hacked quadruped. Almost between the legs of the largest animal is the more ancient (more patinated) petroglyph of a Rectangular Bird (to be discussed in another section). Importantly, it seems to have two short, vertically arranged, parallel tail-feathers.

The tail of this Rectangular Bird just possibly might have been the inspiration to manufacture the two large (hacked) petroglyphs of confronting zoomorphs (Group 2; Elements 2a and 2b in Figure 7). At first sight they seem to depict quadrupeds, but it is interesting to note that also Álvarez Zeballos stated that they have bird’s heads: “dos figuras zoomorfas de animales con cabeza de aves, cuerpo de animales y la cola compuesta como con plumas” (2009). Indeed, birds with circular heads do occur in MRAS (see for instance Figure 19: Element 6). Moreover, instead of a tail they seem to have feathers (which – like in Rectangular Birds – are more or less horizontally arranged). However, the zoomorphic figures of 2a and 2b do not have a wing (see also Van Hoek 2018: 117) and for that reason they may well represent an intentional conflation between birds and quadrupeds, called Bird-Quadrupeds in this study.

Very near Bird-Quadruped 2b are four profile bird petroglyphs (2c, d, e and f in Figure 6) that are clearly much more patinated and thus – in my opinion – (much!?) older, also because Birds 2d, e and f are partially superimposed by Bird 2b (and other hacked petroglyphs). Those Birds 2c to 2f belong to the Three-Digit-Claw Birds type (for more information on Majes bird-types see Van Hoek 2018 and 2021a).

Finally, the large animal of Group 1 partially superimposes with its head a rectangle enclosing facial features (of a feline?) involving an open circular mouth showing (incised) teeth and eyes that have long, slightly curved (incised) tears (Element 1c in Figure 6). The image also seems to have one, small fully pecked ear. Touching this head petroglyph is partially re-hacked Element 2h (h in Figure 7) enclosing a small biomorphic figure, while Bird-Quadruped 2a partially superimposes another square petroglyph (Element g in Figure 7).


Panel AP3-172B

South facing Panel B is equally large, but has mainly (only?) been decorated across the lower half, which, unfortunately, is severely eroded because of the prevailing southerly winds that are often very strong, because just south of Alto de Pitis the narrowest part of the Central Majes Valley acts as a funnel. For that reason the exact layout of most images is hard to make out, especially because of a chaos of superimpositions and close-by lines. For those reasons only a number of simple, match-stick quadrupeds can be seen (with difficulty). Element 3 seems to depict a snake and Element 2 clearly is an anthropomorphic figure with both arms raised, superimposed upon an outlined square (Figure 8).

Most interesting is the intriguing petroglyph of a large quadruped (Element 1 in Figure 8). Álvarez Zeballos (2009; my emphases and additions) confusingly describes this quadruped as: “… un camélido atado con una cuerda que sale desde su cuello y cabeza [however, only from the back of its head] hasta una figura humana que se ubica en la parte posterior del animal, dicha figura [the camelid] tiene la cabeza como de felino mostrando los dientes”. First of all, for reasons yet to be discussed I question whether the quadruped is a camelid. I also doubt whether – in this case – the line is indeed a rope (thus not even a leash), because it is attached to the back of the head (an inadequate spot for fixing a leash) and because this extra back-line joins three other, shorter lines from the rear of the animal.

Figure 8. Details of Panel AP3-172B. Photographs © by Maarten van Hoek.

I also disbelieve the suggested association between the quadruped and the anthropomorphic figure, especially as the purported “rope”-line is not at all linked to the anthropomorphic figure. Instead, the final part of the purported “rope”-line curves downwards and then runs parallel to three shorter, also downward curving lines that emerge from the bottom of the animal; two lines most likely representing the tail of the quadruped. In my opinion the purported “rope”-line represents a still enigmatic element of this unusual zoomorph, which will be discussed in another section. Finally, what Álvarez Zeballos seems not to have noticed in his own photograph (2009: 38), is that the front leg and (less clear) the hind leg both end in an arc of five cupules; elements most likely representing the toes of a feline in MRAS (see for instance Figure 19: inset). Camelids have only two toes at each foot.

Between the uppermost, long extra back-line and the actual back of the quadruped is a “rectangle” that just possibly could represent the head of a (much?) earlier anthropomorphic figure, the body and legs of which have been superimposed by the rear end and hind legs of the quadruped. Unfortunately, in the photos that I have available relevant details possibly confirming my suggestion cannot be distinguished because of the much blurred condition of the lower part of the panel.


Panel AP3-172C

First of all, Panel AP3-172C was incorrectly labelled by me as AP3-172B (Van Hoek 2018a: 50), yet correctly referred to as AP3-172C in the same book (Van Hoek 2018a: 52). This small panel is located at the very top of the boulder (Figure 9), above Panel B and is very hard to photograph (see Figure 4). The most conspicuous petroglyph is a long, straight zoomorphic figure, its interior mainly decorated with circles having an internal dot and some other markings. Its head is a most characteristic type of profile feline head, ascribed by Álvarez Zeballos (2009) to be of Wari origin, but in my opinion more likely belonging to the MSC-Style of the Formative Period – Páracas influenced – rock art of Arequipa (Van Hoek 2013: 80 – 81; 2018).

Figure 9. Panel AP3-172C. Photograph © by Maarten van Hoek.


The Images

Although the imagery on Boulder AP3-172 definitely belongs to the Majes Rock Art Style (MRAS), there are several iconic and even unique petroglyphs on this boulder. There are some  specific images – like “Trophy” Heads – that occur in every major valley of the MRAS (from Caravelí in the west to Vítor in the east; see Figure 1; for more details see Van Hoek 2018a). The image of the Rectangular Bird occurs (with only one exception) only in the Central Majes Valley, while other images are unique to Alto de Pitis. Most of the idiosyncratic images on Boulder AP3-172 concern petroglyphs of zoomorphs. But I will start with images of anthropomorphs and parts thereof.


“Trophy” Heads – Masks

On Panel AP3-172A are two petroglyphs of “Trophy” Heads (and a few more masks or heads). Both are squarish and seem to have indications of the hanging cord, with which they were carried. At Alto de Pitis are several other petroglyphs of “Trophy” Heads (for instance on Boulder AP2-051; AP3-076, 077 and 134; AP4-017, 018 and 043). Interestingly, several of the images on Boulder AP3-172 may well be related to specific petroglyphs on Boulder AP3-171, which is located only 205 m to the ENE of AP3-172 (see Figure 2 for location). Panel AP3-171A has six (partially unfinished) “Trophy” Heads (marked a to f in Figure 10; example c with a distinct T-shaped hanging cord), only one of which is circular (b in Figure 10). The square “Trophy” Heads also include the unique example of the “Trophy-Bird” (a in Figure 10), which was fully described and explained by me in an earlier publication (Van Hoek 2018b). In my opinion this “Trophy-Bird” conveys a special spiritual message.

Figure 10. Panel AP3-171A. Photographs and drawing © by Maarten van Hoek.


Rectangular Birds

There is only one petroglyph of a Rectangular Bird (rather hidden) on Panel AP3-172A. At Alto de Pitis there are only very few examples of the Rectangular Bird (one example occurs on distant Boulders AP1-001 (Van Hoek 2022a) and AP2-005, and three (or four; see next Section) on nearby Panel AP3-170A [Figure 11; see Figure 3 for location]). However, most examples (more than one hundred) occur at Toro Muerto on the west bank of the Río Majes (Van Hoek 2022b), while only three petroglyphs of the Rectangular Birds have been recorded by me at La Laja further north in Majes (see Figure 1 for locations; Van Hoek 2022c). One example was “recently” recorded at Querullpa Chico; yet another at Torán. Outside Majes the image of the Rectangular Bird has only been recorded once, in the Manga Valley, west of Majes.



Very special on Panel AP3-172A, but not unique, are the two petroglyphs of the so-called Bird-Quadrupeds (see also Van Hoek 2018a: 117). They have the body of a quadruped, but bird-like features. Especially the tail (composed of parallel, horizontally arranged lines) reminds us of the tails of the Rectangular Birds. I already mentioned that on Panel AP3-170A (284 m ENE of Boulder AP3-172) three petroglyphs of the Rectangular Bird have been recorded by me (Elements 1 to 3 in Figure 11; note the two “Carcanchas” – Skeleton-Anthropomorphs – on the same panel), but there is a fourth “bird” petroglyph (Element 4 in Figure 11; rough sketch presented in the inset) that could represent a “missing link” between the Rectangular Bird and the Bird-Quadruped. It has a tail composed of four parallel lines. It seems to have two legs. The hind leg is outlined and the front leg is depicted as a simple, crudely pecked line, which may have been added at a later stage. Emerging from the lower part of the rather long neck is an (indistinct) curved wing-line with downward pointing feathers. Unfortunately the (camelid- or bird-?) head is indistinct because it seems to merge with the wing of the largest (and older) Rectangular Bird petroglyph (Element 3 in Figure 11).

Figure 11. Panel AP3-170A. Photograph and drawing © by Maarten van Hoek.

As far as I know only one other decorated boulder at Alto de Pitis has similar petroglyphs. On Panel AP3-127A (tellingly only 30 m SSW of AP3-172; see Figure 2 for location) are two petroglyphs that definitely are related to the Bird-Quadrupeds on Panel AP3-172A. They both have the (fully pecked) bodies of a quadruped and the two legs of a quadruped (Figures 12 and 13; Elements 1 and 2). Yet they have a (bird?) tail composed of four (slightly splayed) lines. Only seemingly Element 1 has a large circular head, while Element 2 seems to have been re-worked some time in prehistory, since the petroglyph shows parts that are less patinated than other parts. Especially the head of Element 2 seems to be more recently added or remodelled. Below Element 2 sits a petroglyph of the so called “Venus Cross” (see detail photo of Figure 13), of which only two examples have been recorded so far in Majes (also the second one found at Alto de Pitis).

Figure 12. Panel AP3-127A; looking south. Detail of Element 1. Photographs © by Maarten van Hoek.

Figure 13. Panel AP3-127A; looking south. Detail of Element 2. Photographs © by Maarten van Hoek.



Also conspicuous on Panel AP3-172A is the large quadruped that is characterised by its rather long body ending in a distinctly curled tail (Group 1 in Figure 5). It is important to note that also this petroglyph is clearly hacked out. A comparable petroglyph – also hacked out – is found on nearby Panel AP3-127A (Element 4 in Figure 12) and there might be a second one on this panel still buried by the loose sands. Similar (hacked) petroglyphs are found on Boulder AP3-001 (at 446 m asl; 480 m SE of AP3-172; see Figure 2 for location) (Figure 14), on Boulder AP3-151 (630 m north of AP3-172) (Figure 15) and on nearby Boulder AP3-159. It proves that certain images – especially those that seem to be hacked out – are (only-mainly?) found very near the western edge of Sector AP3 at Alto de Pitis.

Figure 14. Panel AP3-001; looking south. Photograph and drawing © by Maarten van Hoek.

Figure 15. Panel AP3-152; looking SW. Photograph © by Maarten van Hoek. 



On Panel AP3-172C is a most characteristic petroglyph of a straight biomorph with internal decoration and a typical profile feline head. They may depict ceremonial staffs (like those held by the well-known Andean Staff God. Similar images have been recorded on Panel AP3-114A (Van Hoek 2013: Fig. 73) and (slightly differently) on Panel AP3-146A (Van Hoek 2018a: Fig. 82) and on Panel AP4-018. On Panel AP3-044A is a small example in an inverted position. Also the broken-off example on Boulder AP3-018 may belong to the same type of staff-like Snake-Felines. Also the complex, straight and long petroglyph on Panel AP2-052A may belong to this group of biomorphs.

Opposite Sector AP1 at Alto de Pitis and thus across the narrowest point of the Central Majes Valley, is a newly reported petroglyph site at Punta Colorada (see Figure 3 for an approximate location). Earlier I mentioned one of the newly discovered boulders that shows a “Trophy” Head petroglyph (Van Hoek 2022a: 3) and somewhat later I found out that there is a second boulder with a feline and a “Trophy” Head petroglyph at the same site on a small terrace just above the prominent Red Spot (Van Hoek forthcoming). Outside Majes examples have been recorded at a rock art site in the Caravelí Valley (Van Hoek 2018a: Fig. 28) and on Panel CHY-D-015A at Chillihuay in Ocoña (Van Hoek 2011: Fig. 70). Most (possibly all) of those examples of the Snake-Feline may be very old (belonging to the Andean Formative Period). Alternatively they may display Middle Horizon Wari influence.

Interestingly, another (recently?) reported boulder at Punta Colorada shows three groups of serpentine grooves, the centre one (Element 1 in Figure 16) composed of two undulating lines, which end in an inverted profile feline-like head, showing teeth. The two neighbouring sets have three lines each, but have no heads (anymore?). Yet the heads may have existed once, but a large part of the boulder has completely worn off (exfoliated and severely wind-eroded) because the site is found in the funnel of the Majes Valley, causing the prevailing winds to be very strong at times and especially destructive when laden with sand particles. The three sets most likely are in the original position, which seems to be confirmed by the “correct” vertical position of the anthropomorphic petroglyph (holding a straight object – a staff?) on the panel (Element 4 in Figure 16). Yet, the boulder may have been disturbed, the anthropomorph being added later still. Together, the nine or ten decorated boulders at Punta Colorada (only two or three recorded by Núñez Jiménez [1986: 527]) probably mark an ancient crossing place, connecting Alto de Pitis with Toro Muerto (Van Hoek forthcoming).

Figure 16. Petroglyphs of one (possibly three) Snake-Feline(s) at Punta Colorada, Majes Valley. Drawing © by Maarten van Hoek, based on a photograph by Jose Aroquipa Choque).


Multiple-Back Quadrupeds

In most cases petroglyphs of quadrupeds in MRAS are rather easily identifiable and prove to depict species like camelids, deer, foxes, dogs or felines. However, in some cases determining the species of the quadruped is complicated because of the anomalous and confusing layout of the animal. For instance, the purported feline petroglyph on Panel 2-113B at the Río Caravelí Site in the same drainage seems to depict a conflation between a feline and an (imaginary?) sea creature (Van Hoek 2022d: Fig. 47). And what to think of the “deformed” quadrupeds on Boulder PAJ-043 (Figure 17A) at the extensive site of Illomas in the Manga drainage (see Jennings, Van Hoek et al. 2019 for more info) and the very strange (subjective term) quadrupeds at Chillihuay in the Ocoña drainage (Figure 17B).

Figure 17. A: Petroglyphs on Boulder PAJ-043, Illomas, Manga. Photograph © by Grupo Andaray, Arequipa. B: Some of the petroglyphs on Boulder CHY B-011 at Chillihuay, Ocoña. Photograph © by Rainer Hostnig.


Boulder AP3-172

Also the rock art of the Majes drainage has its share of often incomprehensible (subjective term) images of quadrupeds. One of the – in my opinion – most enigmatic and interesting example at Alto de Pitis is Petroglyph 1 on Panel AP3-172B, the Multiple-Back Quadruped (see Figure 8). It concerns a large quadruped, which most likely depicts not a camelid (as suggested by Álvarez Zeballos [2009]), but more likely an imaginary zoomorph, possibly combining feline and camelid properties. But still, petroglyphs of the Multiple-Back Quadruped may equally represent some other species or even a completely imaginary animal. Petroglyphs depicting this type of zoomorphic figure have primarily been depicted fully laterally, often with a menacing open mouth showing teeth. Only two examples at Alto de Pitis have been laterally depicted, yet showing a frontally depicted head.

Most characteristic and critical in identifying this type of quadruped however, is the long line mainly emerging from the back of the head (thus not from the neck). In case of Petroglyph 1 on Panel AP3-172B this line joins three shorter, downward curved lines that seem to form the tail of the animal. Because of this enigmatic line (running more or less parallel to and hovering over the back-line of the animal), I have baptised this zoomorphic figure “The Multiple-Back Quadruped”. It must be taken into account however, that (some of) the extra back-line(s) of the Multiple-Back Quadruped may have been added at a later stage.

Surprisingly, despite the most unusual extra back-line, the rather idiosyncratic image of the Multiple-Back Quadruped on Panel AP3-172B is not unique. It is also found on at least six other boulders that are found unevenly distributed at Alto de Pitis (see Figure 2), while possibly three different species (or conflations of species) are involved.


Boulder AP2-053

Only one example is found south of Boulder AP3-172, about 700 m to the SSE (see Figure 2 for location). It concerns a rather small rock, Boulder AP2-053, which is found at the very north end of Sector AP2. On Panel AP2-053B is a rather clear, frontally depicted bird petroglyph and a confusion of often deep lines, among which – with difficulty – a large Multiple-Back Quadruped can be distinguished (Figure 18). It seems to have two extra back-lines (differently coloured in Figure 18), one not even attached to the animal (thus also arguing against a leash being depicted). Again the two extra back-lines join the two purported (shorter)  tail-lines that curve downwards. Its unusual short legs end in an arc of cupules (partially damaged at the front leg) and may thus be compared with Petroglyph 1 on Panel AP3-172B.

Figure 18. Panel AP2-053B. Drawing and photographs © by Maarten van Hoek.


Boulder AP3-171

North of Boulder AP3-172 are five boulders with altogether ten petroglyphs of the Multiple-Back Quadruped (thus bringing the total to a minimum of 12 examples at Alto de Pitis). On Panel AP3-171A (about 205 m ENE of Boulder AP3-172) are – besides the earlier mentioned “Trophy” Heads – two candidates, apparently depicting two different species. Petroglyph 1 (see Figure 10) has a large rectangular body (filled with squares) , short legs ending in circular feet with cupules as toes. Two parallel, curved grooves form the tail, but from the “strange” head runs a long line parallel to the back, which joins the two downwards curved tail-lines.

Petroglyph 2 (see Figure 10) has a long, segmented body and tail, two short legs, the front leg being bent forward and ending in a circle with possibly cupules attached to the circle as claws (the hind legs shows similar cupules much less distinct). Its frontally depicted head has two fully pecked ears, two circles depicting the eyes and a circular mouth showing two (?) teeth. From one of the ears (!) runs a long line, parallel to the back and the tail.

This animal (Petroglyph 2) is said by Álvarez Zeballos (2009) to represent a “chungungo” or sea otter, an animal which – in ancient times – abundantly occurred in the Majes Valley, but which is now extinct. Álvarez Zeballos (2009) further argues that the line represents a rope that is tied “to the animal’s neck” (which is incorrect; it is attached to an ear). For that reason he reasons that it is supposed to be an animal in captivity. Of course this may be a possibility. Earlier I argued that just possibly felines were once held as pets in captivity in the Central Majes Valley, based on petroglyphs featuring possible leashes from the neck (Van Hoek 2021b: 9). Therefore, I prefer to leave open the possibility that Petroglyph 2 on Panel AP3-171A in fact depicts a (sacred?) “pet-feline”. This also may be the case with two other petroglyphs, one on Boulder AP3-091 at Alto de Pitis and yet another boulder at Toro Muerto (both still to be discussed).


Boulder AP3-044

About 600 m NE of Boulder AP3-172 is the most important Boulder AP3-044 (first recorded by Hiram Bingham in 1911 [see Van Hoek 2013: 4; Fig. 16]; not by Álvarez Zeballos [2009]; see my comments on Álvarez Zeballos in Van Hoek 2013). It has two panels with altogether four examples of the Multiple-Back Quadruped (and – on Panel A – possibly two further yet incomplete [damaged?] examples, indicated with a “?” in Figure 19). On Panel A are at least three fully laterally depicted Multiple-Back Quadrupeds, all with multiple lines from the head area. They all show complex interior decoration and some have cupules for toes. The extra back-lines have been manufactured immediately on top of the real back-line.

Figure 19. Panel AP3-044A. Photographs © by Maarten van Hoek.

On the same panel sits a very large feline (measuring about 65 cm in length; F in Figure 19) and between its cupelled feet is a petroglyph of what possibly is one of the smallest feline petroglyphs of the Majes Valley. It measures only 5.9 cm in length (see Figure 19: inset). However, the probably smallest feline petroglyph in Andean rock art is found on Panel QPB-Ny-001 at the newly recorded rock art site of Quebrada Pampa Blanca (Van Hoek 2020), five kilometres north of Toro Muerto. This incised feline image is estimated by me to measure no more than 4 cm.

On Panel B of Boulder AP3-044 is a collection of quadrupeds, including a bird with circular head (6 in Figure 19), some felines (one marked “5” in Figure 19) and at least one petroglyph depicting the Multiple-Back Quadruped (marked 4 in Figure 19). It seems to be a conflation between a feline (front leg) and a camelid (hind leg), while the very long and narrow body ends in only one curved line for a tail. From its head (also featuring two diagonally arranged, outlined ears?) run three extra back-lines, two of them joining the factual tail.


Boulder AP3-091

A short distance to the NNE and 748 m NE of Boulder AP3-172 is an interesting petroglyph of a feline (superimposed by a zigzag line) on Panel AP3-091A (Figure 20). It has been depicted laterally, but with a frontally orientated head. From its neck runs a line (first short bit vertically arranged) parallel to the body and part of its tail. This extra back-line might have been added (somewhat) later to represent a leash used in domestication (an early ceramic – to be discussed further on – seems to confirm this possibility). Most interesting is the head which seems to have been framed by a U-shaped, segmented band. Another nearby panel shows a feline with an even more distinct U-shaped band (Van Hoek 2013: Fig. 76). This U-shaped feature may well have a Páracas origin (Van Hoek 2013: 83; please, disregard my earlier – incorrect – claim of a local Chuquibamba origin for this petroglyph).

Figure 20. Panel AP3-091A. Photograph © by Maarten van Hoek.


Boulder AP3-134

Located very near the western end of Alto de Pitis, directly overlooking the Majes Valley is Boulder AP3-134, located about 718 m north of Boulder AP3-172 (see Figure 2 for location). On Panel A are several petroglyphs, including a rather large petroglyph of a Multiple-Back Quadruped (with a small Mummy Bundle inside?). It has been depicted fully frontally, a position which enhances the open mouth with menacing teeth (Figure 21A). It has short legs, of which the front leg is bent forwards (like the example on Panel AP3-171A – see Figure 10). Unfortunately the toes have been superimposed by a later (hacked) petroglyph thus blurring the factual layout of any possible toes. The hind leg(s) seem(s) to be unfinished. The extra back-line seems to emerge  (again) from one of the ears. There is no neck, which also argues against a camelid being depicted. This is one of the (many) petroglyphs at Alto de Pitis that has been scandalously chalked-in by Álvarez Zeballos (2009), traces of which are still visible.

Figure 21. Boulder AP3-134; Panels A and B. Photographs © by Maarten van Hoek.

On Panel AP3-134B is a mixture of older and later petroglyphs, the latter recognisable by their light patina. Those later elements often superimpose a very large zoomorphic petroglyph with a massive, outlined rectangular body (Figure 21B). Although it is somewhat difficult to make out the layout of the animal, it is obvious that it concerns a Multiple-Back Quadruped. The extra back-line starts from the back of the head (actually there is no neck) and joins the short tail. But it is also the head that confirms that we are dealing with a Multiple-Back Quadruped here. It is almost identical to the head of the Multiple-Back Quadruped on the other panel of this boulder, especially the eye and the dented mouth. Its hind leg is short and featureless, while its front leg has been damaged and is now lost. Also this zoomorph most likely is not a camelid.


Boulder AP3-049

Even more enigmatic is the biomorphic petroglyph on Boulder AP3-049 (located about 732 m NE of Boulder AP3-172). It has a most complex layout, mainly consisting of five parallel, undulating lines that seem to build the body, while two long, extra back-lines and four very short lines (two ears and an open mouth?) emerge from the right-hand end of the biomorph (Figure 22). The other end seems to have three cupules enclosed by the real, backwards curving back-line (facial features?). There are several more cupules between the lines. Also enigmatic are its four (or two?) legs depicted as two groups of four, rather short, parallel lines. Each group of four, vertically arranged lines ends in two parallel, horizontally arranged rows of five small cupules (the digits or claws?). The other lines on this panels are more crudely pecked or hacked and possibly are later additions. Moreover, they do not seem to be associated with the purported Multiple-Back Quadruped in any way.

Figure 22. Boulder AP3-049. Photograph © by Maarten van Hoek.



Concluding, at Alto de Pitis the icon of the Multiple-Back Quadruped is found over a distance of 1460 m; from Boulder AP3-134 in the north of the site to Boulder AP2-053B in the south. It is remarkable that in all twelve cases the extra back-line is running parallel to the real back-line and the (often curved) tail. Moreover, in several cases there is more than one extra back-line. Especially for those reasons it is unlikely that those extra back-lines represents a kind of leash or in some way symbolises captivity.

However, when only one extra back-line has been depicted (as is the case with for instance the feline on Boulder AP3-091; see Figure 20), the presence of a leash cannot be ruled out, of course. The possibility of holding felines on leash seems to be confirmed by an exceptional piece of pottery stored in the Museo Chileno de Arte Precolombino – Santiago, Chile (Código de pieza: MCHAP-3098). It shows a purported female figure who is holding an unambiguous feline with one hand and a leash (clearly wrapped around the neck of the feline) with the other hand (Figure 23). Moreover, this piece of ceramic may also prove that the tradition to keeping felines as (sacred?) pets is very old, as the ceramic is said to be from the Andean Formative Period (900 to 200 B.C.).

Figure 23. Ceramic stored at the Museo Chileno de Arte Precolombino – Santiago, Chile. Drawing © by Maarten van Hoek, based on a photo of Código de pieza: MCHAP-3098.

Moreover, it is a fact that the feline (together with the Condor and Amaru-Snake) is one of the most important creatures in Andean worldview. It is also certain that felines also played a very important role in the worldview of the people living in the Central Majes Valley. In this respect it is striking to see that there exists an enormous statistical contrast between feline representations in the rock art of the Central Majes Valley and the rest of the Majes Valley. At the moment (updating November 2022) there has been recorded a minimum of 247 petroglyphs of felines in the Central Majes Valley, against only 11 in the remainder of the valley (Van Hoek 2021b: Fig. 23).

Therefore, it also cannot be ruled out that also certain members of the Majes elite kept felines as a (sacred?!) pet, also on leash. This tradition would also be illustrated in the rock art repertoire of Toro Muerto. Indeed, several petroglyphs of felines at Toro Muerto have a short line across the neck, possibly indicating a kind of collar. Moreover, the Multiple-Back Quadruped is not exclusive to Alto de Pitis. I also recorded one telling example in the extensive repertoire of Toro Muerto on the west side of the valley. It is found on one of the most interesting panels at Toro Muerto, Panel TM-Ba-014A. It features a very large and complex Mummy Bundle, joined by a long line (a leash?) to a feline petroglyph (Van Hoek 2021b: Fig. 27). More informative in view of the subject of this study is the fact that on the same panel there is also the petroglyph of a feline – a Multiple-Back Quadruped – with an extra back-line (not attached to the body, though) and a possible collar (Van Hoek 2021b: Fig. 27 – inset).

Finally, in view of the often menacing open mouth, showing teeth and the complete lack of a neck, it is unlikely that the Multiple-Back Quadruped depicts a camelid. Only in one case it is certain that a feline has been depicted (see Figure 20) and in one case an otter or feline (see Figure 10). For the rest it seems to depict an unknown animal, a conflation of two or more species or even an imaginary animal.

In view of the enormous importance of the domesticated camelid in Andean subsistence economy it is surprising to see that in MRAS only very rarely camelids have been  depicted domesticated or in captivity (like in corrals or on leash). Indeed I know of only very few scenes in MRAS where it is acceptable that a quadruped like a camelid is held on leash or kept in some kind of corral has been depicted (mainly at Toro Muerto). Importantly however, in all cases known to me of quadrupeds depicted on leash, the purported leash runs from the neck in front of the animal, not from the neck backwards across the back (and certainly not curving down, parallel to the tail).

The most convincing example of domestication occurs on Boulder TM-Nn-006 at Toro Muerto. It shows a camelid attached with a line from its neck (the leash) to a shorter line (probably a wooden pole) in front of the animal (Figure 24). At Quilcapampa in the Sihuas Valley are a few camelid petroglyphs that have a line from the neck (the leash?) attached to a small circle (a pole or a tethering stone?) in front of the animals, a configuration that may be compared with camelids depicted on a Wari textile (Van Hoek 2021a: 44; Fig. 20; Fig. 38: inset). At Cerro Jaguay (a rock art site 26 km due south of the city of Arequipa) is a petroglyph of a quadruped with a leash attached to what seems to be a wooden structure, a fence perhaps (Valle Alvarez and Gamboa Tomimaya 2018: Fig. 36). Again the purported leash runs from the neck to the purported fence in front of the animal.

Figure 24. Boulder TM-Nn-006 at Toro Muerto. Photograph © by Maarten van Hoek.

Regarding the distribution of the Multiple-Back Quadruped it is obvious that no specific spot was preferred to manufacture those petroglyphs. They are found randomly distributed across a very large area of the Alto de Pitis boulder field. Regarding the Bird-Quadrupeds and the Curled-Tail-Quadrupeds (and several other hacked petroglyphs) it seems possible that they all are contemporaneous (and later than the petroglyphs of the Multiple-Back Quadruped), showing a slight preference for the western edge of the boulder field, at spots not far from the steep cliff bordering the Majes Valley at its east edge.



I am – again – much indebted to Rainer Hostnig from Cusco, Peru, for his kind permission to use and publish his illustrative material (in this paper Figure 17B) that he collected at many rock art sites in Peru over the past decennia. I am also grateful to César Eliud Barriga Manrique of AndarayGrupo de Excursionismo – from  Arequipa, Peru for his kind permission to use and publish photographs of rock art sites in Arequipa, made by several group members (Figure 17A in this paper). Last but not least I am grateful to my wife Elles for her assistance during several surveys at Alto de Pitis and for her ongoing support at home.


Álvarez Zeballos, P. J. 2009. Petroglifos de Cantas, Pitis, La Mezana y La Laja; Valle de Majes. No longer accessible at Arqueología de Perú. PDF available via me.

Andaray – Grupo de Excursionismo. Websites: Blogspot and Flickr.

Bednarik, R. G. 1987. The chalking of petroglyphs: a response. La Pintura. Vol. 15 (2+3); pp. 12 – 13.

Jennings, J., M. van Hoek, W. Yépez Álvarez, S. Bautista, R. A. San Miguel Fernández and G. Spence-Morrow. 2019. Illomas: the three thousand year history of a rock art site in Southern Peru. Ñawpa Pacha, Journal of Andean Archaeology. Vol. 39-2; pp. 1 – 31.

Núñez Jiménez, A. 1986. Petroglifos del Perú. Panorama mundial del arte rupestre. 2da. Ed. PNUD-UNESCO – Proyecto Regional de Patrimonio Cultural y Desarrollo, La Habana.

Van Hoek, M. 2013. The Carcancha and the Apu. Rock Art in the Death Valley of the Andes. Oisterwijk, Holland. Book only available as PDF at ResearchGate.

Van Hoek, M. 2018a. Formative Period Rock Art in Arequipa, Peru. An up-dated analysis of the rock art from Caravelí to Vítor. Oisterwijk, Holland. Book only available as PDF at ResearchGate.

Van Hoek, M. 2018b. The Supernatural Flight of the ‘Trophy-Bird’ of Alto de Pitis, Majes Valley, Peru. In: TRACCE – On-Line Rock Art Bulletin, Italy. PDF available at ResearchGate.

Van Hoek M. 2020. New “Carcancha” Petroglyphs in Arequipa, Peru. Illustrating the “Road to Coropuna”. In: TRACCE – Online Rock Art Bulletin, Italy. PDF available at ResearchGate.

Van Hoek, M. 2021a. Accessing the Inaccessible. Rock Art of Quilcapampa, southern Peru. Oisterwijk, the Netherlands. Book only available as PDF at ResearchGate.

Van Hoek, M. 2021b. Contextualising the Unexpected Plethora of Feline Petroglyphs in the Majes Valley, Peru. In: TRACCE – Online Rock Art Bulletin, Italy. PDF available at ResearchGate.

Van Hoek, M. 2022a. The Case of Boulder AP1-001, Alto de Pitis, Majes Valley, Southern Peru. In TRACCE – Online Rock Art Bulletin, Italy. PDF available at ResearchGate.

Van Hoek, M. 2022b. The Status of Sector-X within the Rock Art Complex of Toro Muerto, Peru. In: TRACCE – Online Rock Art Bulletin, Italy. Updated: February 2022. Available in ResearchGate as PDF. 2022bb. Sector-X, Toro Muerto, Peru – Addendum September 2022. In: TRACCE – Online Rock Art Bulletin, Italy. PDF at ResearchGate.

Van Hoek, M. 2022c. The Rock Art Site of La Laja – Majes Valley, Peru. In: TRACCE – Online Rock Art Bulletin, Italy. PDF at ResearchGate.

Van Hoek, M. 2022d. The Book of the Río Caravelí Petroglyphs, Peru – Further Analyses. Oisterwijk, Holland. Book available as PDF only at ResearchGate.

Van Hoek, M. Forthcoming. Rock Art at Punta Colorada, Majes, Peru – An Update. In: TRACCE – Online Rock Art Bulletin, Italy. Check my web page to see my “New Publications”.

Valle Alvarez, L. and C. Gamboa Tomimaya (Eds). 2018. Arqueología en Cerro Verde. Arequipa, Perú. Lima, Perú.

Leave a Reply

Your email address will not be published. Required fields are marked *