If you've been following the nerdy end of the news you will have seen headlines this week along the lines of
"Scientists discover new shape!" Now th—w—hang on
who were these scientists discovering a new shape? What d'you mean by new shape?
Are any mathematicians involved and obviously I wasn't the only person wondering this because
people start to email me and send me messages saying "Matt, can you explain
what this new shape the scu-toid or SQ-toid is?" and I was like
You know what? I want to find out. The problem is I'm currently travelling through
I don't know, generic foreign cities somewhere in the world because I'm doing shows for high school students. Details below.
So, to be able to explore this shape and see what it is and how it behaves and if it is indeed
new, it's time for another installment in everyone's favorite stand-up math segment
Matt makes a shape out of things he found around the place he's staying
while on holidays. Let's do it
Welcome to the rather
echoey dining room of my Airbnb.
It's furnished with all the most generic decor that IKEA has to offer and a lot of their walls.
So anyway, I've got some cardboard I found around the house. I've got all the Australian staples
There's BBQ Shapes and Little Creatures Pale Ale.
Thankfully, I've got tape with me and I found a pair of
scissors and I was able to locate some pipe cleaners.
Now the scutoid is a member of the wider family of
prism or prismatoid
shapes. So, before we get into the scutoid itself, we're gonna start with shape number 0: the prism.
I've now got a collection of pentagons that I've cut out of shapes and I've got hexagons from Little Creatures. Now, a
prism is just when you've got two identical
parallel faces and you join up all of the vertices. So I'm going to use the pipe cleaners to do that.
Okay, and there we go. There's my
pentagonal prism and my hexagonal prism. Now, of course, these faces should be filled in
rectangles. I've just done the edges and left these empty,
but they should be filled in the same way that the ends are filled in and these your standard-issue
prisms. Nothing that exciting yet.
that's not the only option for making a prism if we actually take our pentagonal prism and give it a twist,
I can sneak a few more edges in there.
So our original prism here had hexagonal faces and then all these faces joining them are
Because of the twists, I've now got in twice as many edges going around and two
edges meet at each vertex. I mean, plus the, you know, edges on the top.
And so what I've got here is an antiprism. If your two parallel identical faces
line up and you're using rectangles as a prism, hexagonal prism. If things are skewed and you use triangles,
antiprism. So that's now shape number one in the prism family: the antiprism.
Slight detour off the prismatoid path to a pyramid. It's a, you know, hexagon based
pyramid. This is not a prismatoid because it hasn't got a pair of parallel faces anywhere. But you could imagine,
like if you sliced off the top and put a new face in there, it would be parallel to the bottom one.
That's exactly what I've made here. So this is a frustum. That's shape
number two: the frustum. This is what happens if you get a
pyramid-like shape and you lop the top off. It works with cones as well. You get a frustum shape
It is a prismatoid. It's yet another
variation on the thing. So now we have the prism, the antiprism, and the frustum. One more to go.
Okay, so what I've done now is I've put a Pentagon on the top and a hexagon on the bottom.
But to join them together,
well there's more vertices at the bottom than the top
so two of the edges,
which I've done in yellow here, have to meet.
And so this bit here, to patch the six vertices at the bottom to the five on the top
I've used an antiprism triangle here whereas the rest of it is prism rectangles.
Now in maths, there's not a special name for this. This is just back to the generic name prismatoid, right?
It's got the parallel faces all the ones in between have four or three edges around them.
It's a prismatoid. But it's some combination of, it's a bit prism-y me in some bits,
it's a bit antiprism-y in others. It's a mix of the two. Doesn't get a special name.
So I guess our final category is just
prismatoid, which is kind of everything else. But now,
there's one more.
So this is it! This is our
scu-toid or es-cu-toid. It is the same with our generic prismatoid before, but instead of the antiprism
triangle face all the way out, you've got like a Y. So you've got a little triangle face and then over here
you've just got your standard single edge coming into the top. So you could say
a scutoid is a
prismatoid at one end and then it's a prism at the other.
prism and a
single triangle prismatoid
Though, there you are. I mean, that's-- that's it.
I guess the most important bit is this bit here where it merges. A little like Y or V shapes. So in short,
pentagons to the left of V,
hexagons to the right. Here, Y am
stuck in the middle
Hey, that's pretty good.
That's so good, I'm taking five minutes off.
Actually, while I am taking a break, earlier on today
I had a chat with my friend Laura Taalman who was working out how to 3D print one of these.
Which is not easy, because these faces no longer all flat
you've got this concave bit here. This weird kind of 5 edged
face is no longer well behaved and so 3D printing that is not trivial. And
we'll get to in a moment, these things are meant to kind of stack together.
So I asked, basically, I asked Laura how she went about designing a 3D model for these.
So when I first tried to make the scutoid, I did sort of the most boneheaded thing you can do which is just to put
vertices in the shape of a hexagon,
and then below it in the shape of a pentagon, and then have an extra vertex, and take the convex hull of all those vertices.
That didn't work. Some of the sides have to be curvy.
They're not actually planar faces. The scutoid isn't a polyhedron. The next thing I did after that is I
divided the object vertically into slices and kind of lofted a slice at a time
so each slice was not curved but then it approximated
this nice curved surface going up. And then, I did get an object that was nice and curvy and looked like a scutoid that
looked pretty much just like the scutoid that's in the picture in the article. However, it didn't pack with itself.
It turns out that the locations of the vertices and, in particular, the location of the extra vertex
are really key for determining whether a scutoid will pack with itself.
But there's ways that you can constrain it to make that happen
and the final model does have this property that it packs with itself. I was doing all this in OpenSCAD.
You can delete that if that's not interesting, but I was. I did this an OpenSCAD.
So what does a scutoid have to do with packing and biology? Well, there are cells in biology called
epithelium cells and they form an epithelium layer and biologists just always assumed that they would be prisms
that have f-- well, obviously distorted prisms. The cells would grow and fill the space and just, you know
have a face at each end of be joined up normally in the middle and whenever there was some kind of curvature or
the surface was distorted then it's a frustum. It's a truncated pyramid arrangement.
But then, biologists spotted patterns and curvature in these layers of cells which couldn't be explained by truncated pyramid
shaped cells. And so they got some mathematicians involved. Yeah, thank goodness
There were some mathematicians involved
once the biologists saw these strange shapes, some mathematicians came in and they thought well
how would we model
what shape these cells would form if they were packed into these various arrangements? And they used, well, a
variation on a thing called the Voronoi Technique
which is where you put a bunch of points in space
and then you expand them until they fill all the space and each region around each point are all the other points which are
closest to that
original starting seed point. And then they looked at, in the 3D
arrangement, how these cells would form if they were trying to like minimize the energy, but find a stable
arrangement where they pack together as efficiently as possible? And that is where they came up with the scutoid.
The mathematicians predicted that these
epithelium cells would have a different number of edges on one face compared to the other face and the edges would merge
somewhere in the middle giving this prismatoid arrangement.
I was actually lucky enough to talk to one of the mathematicians involved in the project, because I knew they were in there.
So, I had a chat with Clara Grima
and she explained a bit more about how they found this shape based on what the biologist told them.
Hey Matt, this has been the first time for me
that we work together with biologists, physicists, and computer scientists
and I think that the role of everyone was quite important.
In short, we the
mathematicians had to formalise how to define the new shape that
biologists had observed in the microscope. The group leaded by
Luis M. Escudero,
biologists had discovered that the usual representation of cells or the epithelial tissues
truncated pyramids could not explain some properties that they observed in the microscope.
So they ask us to try to find the correct model.
With our first model we
made some predictions,
but they will not still
accurate. So, we have to change our model and finally with the scutoid
everything worked fine.
Finally, we have the seal of approval of physicists and he proved that the
scutoids are stable when
they pack together.
The mathematics didn't predict that all epithelium cells would be scutoids just enough to give you the right curvature.
So for the case of a cylinder, which happens in animals apparently, or some kind of spheroid, you would have a certain
percentage of these and it's quite nice that it's built out of pipe cleaners because it's not some fixed rigid shape, right?
But it's not, it's not like a cube or something like that but nor is it quite topology because there is some geometry.
It's somewhere between topology and geometry. They worked out the structure
that would be required in some cells to give these sorts of shapes and then they found it. They looked in fruit flies.
They took what the mathematicians told them and the biologists were actually able to find these shapes inside
little creatures. It's ap--
a phenomenal collaboration between biology and mathematics. There may be applications as well.
I was chatting to Clara and she said that one day
if they look at tissue actually in, let's say, humans and they know what structure to
expect, they can tell if the cells are growing normally or not. Or if we ever make artificial cells,
we need to know what kind of structure they form. So at the moment, it's all very abstract. It's all very exciting.
You can read the full nature paper if you want. I'll put a link to it below.
It's publicly available. One last bombshell that Clara told me when we were talking is that this shape here
which gives it the name scutoid in the media. Officially, it's because of a beetle, right?
It's named after a very similar shape you will see in beetles.
Biologists! But when I was chatting to her, it turns out that was not the
original reason for the name scutoid.
This is funny and cute for me
Officially, it's because part of the body of a beetle is very similar to our shape the scutoid.
okay, but there is a hidden reason and I will tell you secret.
The biologist leader of the project is called Escudero
Translated to Latin is "escudo". So from "escudo" as a joke, we start to call it
Later, we cannot think another name for our shape, but scutoid. Bye Matt,
May the scutoid be with you. Thanks so much for watching the video.
There are more videos of me building things that are stuff
I find while traveling. You can subscribe to my channel for all your mathematical needs and thank you so much to all my patreon supporters
who make these videos possible. If you come across more maths in the news or something interesting, drop me an email.
I'll see if I can make a video about it
and finally the shows for high school students I was doing in Sydney are over
over but I do them all the time right across the UK and we'll be in New York for the first time in October.