Gilf Kebir, southwest Egypt, Egypt
Wadi Sura II
About 6500 to 4400 BC by context (no date on the paint; a hunter-gatherer window in the later Green Sahara)
Under a sandstone lip in the Gilf Kebir, about 900 hands were sprayed onto the rock, and thirteen of them are far too small for the babies they were first called. The shelter is the Cave of Beasts. Its age is the Green Sahara's, borrowed from camps and pots, not from the paint.
Why this is extraordinary
Wadi Sura II is a rock shelter, not a closed cave, at the southwestern foot of the Gilf Kebir, in the Western Desert of Egypt. The other Egyptian rooms are on the Egypt page. The overhang is about 20 metres across and about 8 metres deep, open to the day, and the painted panel rises about 4 metres on nearly 100 square metres of wall. It was found in 2002. About 10 kilometres to the east, the shelter called the Cave of Swimmers had been found in 1933 by László Almásy. The name Wadi Sura is the valley of pictures. The 2002 wall holds more than 8,000 figures, people and animals and hybrid creatures, laid over one another until the order of painting can be read where one body covers another. The catalogue called it the Cave of Beasts for a few dozen creatures drawn without heads.
Among the earliest paintings is a crowd of stencils, hands and also arms, feet, disks and sticks. Counts of the hands have moved. One said about 120, another about 400, and the count made for a 2016 study says about 900. Thirteen of them are tiny. Eleven could be measured. From the base of the palm to the end of the middle finger they average 45.3 millimetres, and one table in the same study prints 45.33. The fingers are longer than the palm and they thin to a point. The tiny ones sit about 1.8 metres above the floor as it was when the shelter was found, five of them in a line like a frieze, and one pair is nested inside a pair of larger human hands. They are under 1.5 percent of the stencils.
They were first read as babies' hands. The 2016 study set the eleven measurable stencils against 36 hands of babies born at term and 25 hands of babies born early. Term newborns averaged 62.01 millimetres. On four of eight measurements, the chance that the tiny stencils and the term babies share an average was under 0.01 percent. The early babies failed in the same way on four criteria, even where a couple of single measurements did not rule a shared population out. Proportions differ as well. A human thumb sits off to the side, where it can meet the fingers, and on these stencils the thumb lines up with the other digits. To see whether a stencil simply warps a hand, the study compared 30 large stencils from this wall with 30 living adult hands. The average deformation stayed within about 5.5 percent, too small to turn a baby hand into these shapes. A young monkey's hand matches the length and the long fingers, and it was set aside because the thumb still opposes and the tips are not this pointed. A wooden or clay cut-out is a weaker fit. The fingers shift a little from stencil to stencil, and no retouch shows, which reads as a jointed foot rather than one carved stamp. The closest match in the paper is the forefoot of a desert monitor, Varanus griseus. One adult male specimen matched six of the seven measurements. A young crocodile is not ruled out, and that comparison was still unfinished when the paper was published. The catch is local. The same paper says neither animal had yet been identified in the Gilf Kebir, in the pictures or in the bones. A foot can be carried in. A species missing from the rest of the evidence is a comparison, not a capture.
The small horizontal bodies are the swimmers. Almásy gave the 1933 shelter that name because the figures look as if they are in water, and a vanished lake is part of the ground in front of the later shelter too. Wadi Sura II stands over a playa, a dry lake floor of stacked sediments. The 2002 wall has its own row of the same horizontal figures. A body painted level with the ground can be a swimmer, a person floating, a spirit, or a standing person the wall has laid on its side. That menu is already open for the round-head figures at Tassili n'Ajjer and in the Tadrart Acacus. No source used here names a star, a solstice, or an azimuth for either Wadi Sura shelter. The headless beasts are the other strange body. A few dozen hybrid creatures are drawn without heads, and some of them are surrounded by small humans who touch them. A missing head can be a being, a carcass, or a head painted in a pigment that has since gone. In the beasts photograph on this page the large dark animals still have heads, so that frame is the crowd and the animals, not a labelled headless one.
No pigment from this wall has been dated. The hand stencils belong to the earlier layers in the pile, and a relative chronology puts that phase, tentatively, in the second half of the seventh millennium BC and in the sixth, around 6000 BC. The wider project places the prevailing pictures with hunter-gatherers of the Gilf B phase, about 6500 to 4400 cal BC, and most of the neighbourhood pottery is that phase. Ostrich, giraffe and gazelle on the wall match a hunting menu. Cattle pictures belong with a later pastoral phase, after about 4400 BC, and they are a different subject. The climate is the Green Sahara, and it is not one date for the whole desert. The Sphinx water room lines the gauges up. Off Mauritania a dust core runs from about 14,800 to 5,500 years before 1950. In the eastern Sahara, between about 16 and 24 degrees north, lakes begin near 8500 BC and the stable humid run lasts until about 5300 BC, with refuge country into about 3500 BC. The Gilf Kebir is near 23.65 degrees north, at the north edge of that band. A window of 6500 to 4400 BC sits in the later wet centuries and runs past 5300, which fits a plateau that kept water after the open desert had started to dry. Those years belong to the camps and the pots. The paintings inherit them.
Hands elsewhere on this map: Cueva de las Manos, El Castillo, Cosquer, Gargas, Pech Merle, Maros, Lubang Jeriji Saléh, and Monte Alegre. What this changes: if the tiny stencils are a monitor's forefoot, this shelter has a pentadactyl animal stencil, thirteen times, sometimes set inside a human hand, and the baby story was a trick of scale. If they are infants after all, the hospital sample is what failed, and people here did stencil babies. The swimmers need the wet phase to be a lake, or they need a meaning that does not use water. The paint still has no clock of its own.
Why it matters
The tiny stencils are a measurement argument, infant hands against a reptile's forefoot, and the swimmers are a climate argument, a real lake against a floating body. Neither is settled by a date on the pigment.
The wide wall, the hand field, and the beasts are this shelter. The single horizontal figure and the cliff with two mouths are the Cave of Swimmers, about 10 kilometres east, so they are the neighbour and not this pin. The diagram is the 45-millimetre stencil beside a newborn hand, and the art window inside the eastern wet phase.
Photo tour
6 photographs, unedited. On a phone, swipe the row. Choose one to look closely.
- The thirteen tiny stencils are drawn in the 2016 paper. None of these photographs isolates one and names it. A small mark in the hand field is not identified here as one of the thirteen.
- A headless beast is a catalogue motif. The large animals in the beasts photograph still read as having heads.
See the evidence at the source
These pictures are not openly licensed, so they are not copied onto this page. Each link opens in a new tab.
- Non-human stencil hands at Wadi Sūra IIThe measurements, the newborn comparison, and the monitor forefoot. The paper's figures are not reprinted here.
- Wadi Sura, the Cave of BeastsThe catalogue of about 8,000 figures. The plates are not openly licensed for this gallery.
What is still unexplained
These sit beside the consensus account, with the same visual weight. None of them is marked solved.
Location and dating
- Location
- Wadi Sura II, also called the Cave of Beasts, at the south-western foot of the Gilf Kebir, New Valley, Egypt. The pin is the published shelter coordinate, 23°39′12″ N, 25°09′35″ E. The Cave of Swimmers, Wadi Sura I, is about 10 kilometres east. Both sit with the other rooms on the Egypt page.
- Period label
- Current consensus — open to revision Hand stencils tentatively in the later seventh millennium BC and the sixth, around 6000 BC. Prevailing art about 6500 to 4400 cal BC, Gilf B, hunter-gatherers. Pastoral cattle pictures after about 4400 BC are a later subject. No pigment date. Current consensus — open to revision.
- How this dating was done
- Relative order, which colour and which body lies on top, plus the settlement around the shelter. Pottery and camps of the Gilf B phase carry radiocarbon ages, and the pictures are tied to that phase by the animals and by the pots, which is also typology. The eastern Sahara lake window used beside them is the one already set out in the Sphinx water room. The pigment itself was not dated.
- What could change it
- A date on binder from one hand stencil and from one swimmer, or a crust over a beast, would show whether the Gilf B window is the painting or only the century people camped here.
What is actually there
- A shelter about 20 metres wide and 8 metres deep, with more than 8,000 figures.
- About 900 hand stencils. Thirteen are tiny. Eleven average 45.3 millimetres.
- Term newborns in the comparison average 62.01 millimetres.
- One desert-monitor forefoot matched six of seven measurements. The crocodile comparison was unfinished.
- Horizontal "swimmers" and headless hybrid beasts.
- No published star, solstice, or azimuth.
Open questions
| Question | Status |
|---|---|
| Are the thirteen tiny stencils infant hands, a monitor's forefoot, or a crocodile's? | Open. The newborn comparison is a poor match. The animal match is one specimen, and that animal is not yet known from Gilf Kebir bones. |
| Are the horizontal figures swimming in a lake, or floating in some other sense? | Open. The playa is real. A level body is not, by itself, a stroke in the water. |
| Do the paintings belong inside 6500 to 4400 BC? | Open until something in the pigment, or a crust on it, is dated. |
The current narrative
Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.
The current narrative — open to revision
Tiny stencils that miss a baby hand, in a wet-phase shelter
The thirteen tiny stencils do not match measured human newborn hands, and a reptile forefoot is the closest comparison published. The pictures sit with hunter-gatherers of about 6500 to 4400 BC by the camps and the pots, not by a date on the paint.
What supports this
- Average length 45.3 millimetres against 62.01 for term newborns, with four criteria under 0.01 percent.
- Thumb position and proportions sit outside the human cloud, and adult stencil distortion is only about 5.5 percent.
- One monitor forefoot matched six of seven measurements.
- Gilf B pottery and the hunting animals frame the same centuries as the eastern Sahara's later wet phase.
What pushes back
- The animal match is one specimen, and the crocodile comparison was unfinished.
- Neither candidate species is identified in Gilf Kebir art or bone in that paper.
- The babies measured were a modern European hospital sample. The adult check bounds that bias. It does not erase it.
- No pigment was dated, so the century can still move.
What would settle it
The mismatch with newborn hands is the firm result. The species of the foot, and the year of the paint, stay open inside that result.
Emerging narratives
Other published readings. Each may be right. None is adopted here unless the evidence carries it.
An emerging narrative
They are the hands of human infants
The tiny stencils are babies after all. The first readers of the wall, including the pair of small hands set inside a pair of large ones, had the human picture right, and the hospital sample is the wrong comparison for a Holocene Saharan child.
What supports this
- The nested pair looks like a child held against an adult, which is a human scene if both are hands.
- Hand size does vary with population, and a stencil is not a caliper on skin.
- A couple of the early-baby measurements did not, by themselves, exclude a shared group.
What pushes back
- Four criteria on term babies, and four on early babies, fall under 0.01 percent.
- The thumb lies in line with the fingers, which is a shape, not only a length.
- The adult stencils on the same wall do match living adult hands within about 5.5 percent, so the method is not wildly distorting human hands.
What would settle it
The reading the wall first suggested. The measurements are the weight against it, and they are why the room does not treat "baby" as the name of the mark.
An emerging narrative
The horizontal figures are people in a lake
The swimmers are swimming. The Green Sahara put seasonal water in front of these shelters, and the painters recorded bodies in that water.
What supports this
- The 1933 shelter was named for the pose, and the 2002 wall repeats it.
- A playa with lake sediments lies in front of Wadi Sura II.
- The art window falls inside the later eastern Sahara wet phase, when surface water in this band is published.
What pushes back
- A horizontal body is a pose. It does not show a waterline.
- The same pose is read as floating or spiritual at other Saharan sites, without a lake in the picture.
- The wet phase dates the climate. It does not date a particular stroke.
What would settle it
A fair use of the landscape. It needs the pose to mean water, which the pose alone does not say.
An emerging narrative
Floaters and headless beings
The horizontal figures are weightless bodies, spirits or dreamers or a convention of the wall, and the headless beasts are beings rather than portraits of animals that lived in the wadi.
What supports this
- The figures do not stand on a ground line, and the beasts lack the head a portrait usually keeps.
- Small humans gather around and touch some of the headless creatures, which is a scene, not a hunt tally.
- Round-head paintings at Tassili and in the Acacus already carry this kind of reading for floating bodies.
What pushes back
- The playa and the wet-phase window give the swimming reading a real lake to point at.
- A head can be lost to fading pigment, and the beasts photograph shows animals that still have heads.
- No myth text from these painters survives to name a spirit.
What would settle it
As open as the lake reading, and no more closed. Nothing in the sources turns either into a star or a solstice.
Should the current view be reevaluated?
Open to revision
The newborn mismatch is a real measurement, and the monitor is the best animal comparison in print, with the species still missing from the local bones. The swimmers have a lake and a pose, and the pose has other readings. The years are the Gilf B context. The paint is undated. No sky alignment is published.
What drives this
- 45.3 millimetres against 62.01, four criteria under 0.01 percent
- One monitor forefoot, six of seven measurements, crocodile unfinished
- Gilf B about 6500 to 4400 cal BC, no pigment date
- A playa in front of the shelter, and headless hybrids in the catalogue
This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.
Weigh it yourself
Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.
Visitor lean so far — not evidence. On this browser only; a shared tally is not collected.
If the contested reading is right
What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.
The tiny stencils are a monitor's forefoot
People at this shelter stencilled a reptile among human hands, the first pentadactyl animal stencil identified this way, and the infant story would be a scale mistake. The monitor would still need a local bone or a picture of the whole animal.
Why it is still worth a test
The unfinished crocodile comparison, and a search of Gilf Kebir bone for the monitor.
They are human infants the sample missed
The Sahara would have the infant hand stencils it was thought to lack, nested in adult hands, and the morphometric cutoff would be a population problem rather than a species one.
Why it is still worth a test
A comparison with hands of Holocene North African children, if any measurable sample exists, run with the same seven lengths.
The swimmers record a real lake, and the art is Gilf B
The shelter would be a picture of the later Green Sahara made while the water was there, a hunter-gatherer century before the cattle painters, still waiting on a pigment date to lock it.
Why it is still worth a test
A date on one swimmer and one hand, and a sediment age from the playa directly in front of the wall.
Sources
- First identification of non-human stencil hands at Wadi Sūra II Journal of Archaeological Science: Reports, 2016
- Open copy of the stencil study HAL
- Wadi Sura — The Cave of Beasts Heinrich-Barth-Institut, 2013
- Review of Wadi Sura — The Cave of Beasts Pastoralism, 2014
- The Cave of Beasts and its chronological affiliation Royal Academy for Overseas Sciences, 2012
Image credits
- The painted wall of Wadi Sura II, the Cave of Beasts, with sand heaped along the floor. Rows of hand stencils run across the lower rock. Smaller figures sit above them. The thirteen tiny stencils are not labelled in this frame.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- A field of hand stencils. The larger ones show spread fingers. Smaller marks sit among them. This frame does not identify which, if any, are the thirteen tiny stencils.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- Large dark animals and a crowd of small humans, some of them horizontal. The big animals in this frame still have heads, so they are not a labelled headless beast.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- The overhang and the dune in front of it. The painted panel is not in this view. The shelter is open to the sky, which is why 'cave' is the popular word and 'shelter' is the closer one.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- One horizontal figure from the Cave of Swimmers, Wadi Sura I, about 10 kilometres from this pin. The body is the dark mass and the limbs stick out. Swimming and floating are both readings of that pose.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- Shelter mouths at the foot of the Wadi Sura cliff, the valley of the 1933 swimmers. The paintings are inside. This is the neighbour, not the Cave of Beasts wall.. Roland Unger, CC BY-SA 3.0. Shown unedited.
- The tiny stencils, the newborn average, and the context window inside the eastern Saharan lakes. A bar is not a date on paint.. Diagram: Our Deep History.