How the Ju|’hoan Make Poison Arrows

A modified version of a map in a paper by Caroline S. Chaboo, Megan Biesele, Robert K. Hitchcock, Andrea Weeks

When I was a child, my grandmother taught me how to make arrows from the softwood of a poplar, a tree that was abundant in the woods that we could step into from my backyard. Later in adolescence, biased from the movies that I watched,  I perceived arrows to be “primitive”.

It is another case of truth differing from popular perception.

Human ancestors were hunter-gatherers millions of years ago, but archery appeared relatively late, around 65 000 years ago. To make arrows far more effective hunting weapons, humans have learned to extract poisons from plants, frogs and beetles. Unfortunately, indigenous communities across the planet who hunt with arrows are gradually becoming sedentary and are losing traditional knowledge. They include  some native North American Indians; Paraguay’s Ache; the Pumè in Venezuela; the Hadza in Tanzania and 113 000 people of various San communities (use “click” languages) from Angola to South Africa (see map). Some San groups are the only ones in the world who are known to use beetles to make their poison arrows. Based on their local conditions, these invariably isolated communities have developed their own specialized poison use and preparation.

Authors of a 2016 study, focused on two San groups from Namibia: the Ju|’hoan and Hai||om. The former who live near Nyae Nyae told researchers that, thanks to older hunters, they had learnt to locate plants of the genus Commiphora, which are host to the larvae of two beetle species shown below.


At margin of the shrub, using either a metal stick or a traditional wooden one, the hunter either digs a new 1 meter ditch encircling the plant or extends that of a preexisting one. ( (1) in picture set below)  The hunter uses his finger to sift the loosened sand with his fingers, straining out the beetles’ oval-shaped cocoons . When enough cocoons have been collected into an ostrich egg shell or plastic container (2), the hunter heads home.  The hunter makes sure that the concave surface of an animal knuckle bone faces upwards in the sand in front of him, and he places the beetle cocoons nearby. A small fire is lit. He breaks open a cocoon and taps out the single larva (3), discarding any pupae and adult insects.  Using a stick as a pestle, he rubs hard against the skin to loosen tissue, then extracts it to mix on the bone mortar (4), using about 10 larvae for every arrow. He chews the bark of Acacia mellifera to produce saliva which is mixed with the larval tissue and their “blood”, hemolymph. A bean of the toxic snake bean plant (Bobgunnia madagascariensis)  is heated over the fire, cooled and added to the mixture. Ju|’hoan informants at|Xai|Xai in Botswana told one of the researchers that they use the juice of Sansevieria plants to improve the poison. Without touching the poisonous mix, the hunter applies the ‘beetle paste’ of D. nigroornata larva with a twig to the dried sinew that fastens the arrowhead to the shaft.  The arrows are then finally dried. Their poison becomes less potent with time and expires within a year.


The poison is a  protein known as diamphotoxin. As far as proteins go, it is very small and simple structure. It’s a single chain polypeptide with a molecular weight of only 60000 grams per mole. The poison is a so-called ionophore, affecting the permeability of red blood cells. Many small ions, including those of calcium (Ca2+), enter indiscriminately. The wounded animal’s red bloods, about 75% of them, then burst.  This leads to a severe shortage of oxygen throughout the body. The release of ions from ruptured cells also wreaks havoc in kidneys. Convulsions, paralysis, and death ensue. As far as natural poisons go, diamphotoxin is highly potent. Whereas the lethal dose that kills 50% of animals (LD50) for chlorotoxin (scorpions), nicotine, amatoxin (amanita mushrooms) and frogs’ batrachotoxin (from poison dart) is 4.3, 3.3 , ~0.5 and ~0.002 mg/kg, respectively, diamphotoxin’s is only 0.000000025 mg/kg (25 picograms/kg) for mice. That implies that one gram of pure diamphotoxin would be enough to kill (50%)*(1012 pg/g)*(kg/25pg)*(mouse/0.020 kg) = one trillion twenty-gram mice. 

Although the San groups have no molecular explanations for their weapon of choice, in order to develop the method of preparation, they have performed a type of laboratory science, with all its trial and error. And just like with modern science, it took a leap of imagination to think of extracting the poison from an insect’s larvae in the first place. Unfortunately, unlike practitioners of  Western science, the San groups have not left any written records of how and what their previous generations have discovered.


Chaboo and Al. Beetle and plant arrow poisons of the Ju|’hoan and Hai||om San peoples of Namibia. Zookeys Feb 01, 2016 958: 154

Harpe and Al. Diamphotoxin. The arrow poison of the !Kung Bushmen. Journal of Biological Chemistry. October 10, 1983 258, 11924-11931.

Jacobsen and al. Effect of Diamphidia toxin, a bushman arrow poison, on ionic permeability in nucleated cells. Toxicon Volume 28, Issue 4, 1990, Pages 435-444


Other LD50 data:


Why Only Five Platonic Solids in Our Geo-Bio-Chemical World?

The DNA-surrounding capsid of the cold virus and the atomic arrangement of  a certain allotrope of  boron consists of  20 triangles arranged in a three dimensional shape known as a icosahedron.

If you form a three-dimensional structure with 8 triangles, you get an octahedron. An example is the molecular structure of the electrical insulator and potent greenhouse gas, sulfur hexafluoride, SF6. It can also appear in the mineral pyrite.

If you limit the number of triangles to four, you get a tetrahedron. The tetrahedral silicate unit,  SiO42- , is the basic component of most silicates in the Earth’s crust. On our planet’s surface, any time carbon makes four single bonds with other carbons or other elements in a wide variety of life’s hydrocarbons, we also get a tetrahedral arrangement. This allows the four bonding pairs of electrons to get as far from each other as possible.

Changing polygon, we can use 6 squares to make a cube. The similarly-sized ions of cesium and chloride can form an arrangement where each whole ion centers 8 vertices occupied by an ion of the opposite charge. But viewed within the lattice, the boundaries of the cube are such that only 1/8 of each ion is at a vertex. Given that there are 8 vertices, this maintains the ratio of one cesium for every one chloride ion. In the mineral halite, which is composed of NaCl, the chloride ion is considerably bigger than the sodium. They don’t pack into a cube in the same manner, but overall they still form the same shape.

Then there is the dodecahedron, consisting of a dozen pentagons. Cubes of pyritohedron form the macro-illusion of a dodecahedron, but otherwise this “perfect” solid is rare in nature.pyrite_really_cubes_molecular

Are there more than these 5 possible perfect or Platonic solids? No. But why?

There is a relationship that holds for all solids of this type. If you let V represent the number of vertices, E = number of edges and F = the number polygonal faces, slide1-l

you will observe that in a tetrahedron V= 4, E = 6 and F = 4.

For a cube, download

Notice that for both solids, the following simple formula(called Euler’s Formula)holds true:

V + F – E = 2.                                 Equation (1)

That doesn’t constitute a proof for why it should apply to all cases, but you can find one here. We can use this formula to prove that there are only a limited number of Platonic solids.

First let’s introduce two other variables, N = the number of sides in a polygonal face and R = number of edges that meet at a vertex.

Since each edge is shared by two vertices, if we multiply R by the number of vertices,V , and divide by two, we will get the total  number of  edges:

RV / 2  = E;   Solving for V we get

V = 2E / R                                        Equation (2)

The number of edges can also be obtained by the number of faces. Each face has one edge for each of the number of sides, N. But each edge is shared with another face, so again not to count things twice:

NF / 2 = E.   Solving for F we get

F = 2E / N                                        Equation (3)

Substituting equations (2) and (3) into equation 1:

2E / R  + 2E / N   – E = 2.     

Now divide each term by 2E:

1/R + 1/N1/2 = 1/E. 

We need at least three edges to get a 3D shape so R ≥ 3. Similarly to get a polygon, N ≥ 3. Interestingly N and R cannot simultaneously be greater than 3,  because as they create progressively smaller fractions, 1/R  and 1/N will add up to a maximum sum of 1/2 ( if R=N=4), which in the formula will yield 0 = 1/E.  

Letting N = 3,  if R = 3  then E = 6. Using this result and equation(3),  F = 2E/N = 2(6)/3= 4: the tetrahedron.

Having no choice due to the restriction we mentioned, we have to keep either N or R at 3 while increasing the other, so

letting N = 3,  if R = 4, E = 12 and F = 8, the octahedron. 

Reversing the values and letting N = 4 and R = 3, E = 12 and F = 2(12)/4= 6,  the cube.

We can then try the combinations of N= 3,  R= 5 and N= 5, R= 3, which will solve for the icosahedron ( E = 30; F = 20) and dodecahedron ( E = 30; F = 12), respectively.

But 5 is the limit because if we try values of 3 and 6:

1/3 + 1/6 – 1/2 = 0, which means the impossibility of no edges. A value larger than 6 yields a negative value for E.

Given that there are only these 5 solutions to Euler’s formula , then only five Platonic solids can exist in three dimensional space.

Zebra Finches At a STEM Conference

In the spring of 2016, I attended a STEM conference. Expanding the acronym as science, technology, engineering and math doesn’t shed too much light on the intentions and philosophy of STEM.  The premise is that math, science and technology subjects should not be taught in isolation; there should be more integration and emphasis on applications. All of this is largely inspired by the job market’s need for a larger number of better-trained people in these specialised fields. It all seems reasonable as long as the approach is not taken to an extreme.

A society, regardless of its bent, functions best when a wide range of talents are cultivated, even if they seem to serve no practical purpose. Similarly, we have a healthier situation in schools and colleges when educators don’t sail on the same ship. There was at least one organiser at the conference who shared my views because a particular lecture went against the grain of STEM.  95% of the auditorium featuring the lecture was empty and attended mostly by the speaker’s university students, a couple of bird-lovers and a little cluster of Canada Wide Science Fair attendees. Rudely, the latter group even walked out before it ended. But if you stick to the premise that attendance at public events is very often inversely proportional to its quality, you don’t worry about numbers.

Parentese or “baby talk“ is far from being just indulgence on the part of parents. In humans it helps draw attention from babies and promotes the learning of speech. Regardless of language, there are universal characteristics of parentese. Voice pitch is modulated; speech is slower, more repetitive and more attention-grabbing than adult-talk.

Image of baby zebra finch from

A similar situation arises in a small Australian bird known as the zebra finch. For those of you unfamiliar with the small bird, one of its distinguishing characteristics is its song, which is reminiscent of the squeaky sound of a rubber duckie. With their form of baby talk, adult zebra finches change their vocalisations when singing to young birds. They slow them down and use more repetition. The juvenile finches in return pay more attention to such songs than to those used between adults.In their young lives, they the simpler versions. With time, in the physical presence of adults, the chick’s song converges with that of their tutor. When zebra finches were isolated and exposed to mere recordings, they developed different songs.

The social interaction between tutor and chick stimulates communication between the midbrain’s ventra legmental area (VLA) and regions of the cerebrum. The VLA is partly a reward centre and uses dopamine.  When humans acquire language, neural bridges of that type are also made. In case of the finch, the evidence for such a pathway comes from the fact that in the absence of tutor’s physical presence, a marker for gene expression of catecholamines (which include dopamine) remained inactive.

Another revelation which made my morning was that the zebra finch researcher had originally majored in economics, reinforcing my notion that to get to an island you don’t have to board any specific boat.


Up ↑