Download as txt, pdf, or txt
Download as txt, pdf, or txt
You are on page 1of 32

OK.

We've skirted around


them long enough.

We've talked about


Thales' theorem

on constructing right
triangles in circles

from their diameters.

And as we've mentioned


in lecture 1,

this is the oldest recorded


result, with proof,

in geometry.

And we've talked


about circle-ometery.

But we've never


talked explicitly

about circles themselves.

It is time to study
circles in their own right

and prove facts about them.

Particular, we need to
prove Thales' discoveries.

So let's start with acute


activity, as follows.

Here on the whiteboard


are two nails.

Actually, that's not true.

I've just drawn two dots.

I didn't want to ruin the board.

But what I'm going to do is


take the 90-degree corner

of a piece of paper and bring


it up between the two nails,

and see where that corner lies.

And, in fact, I'll mark the


corner of that 90-degree angle.
There it is.

Not very exciting.

But I'm going to do


this lots of times,

coming in at
different positions.

Over here now, now the


top of the paper is there.

At this angle of
insertion, it's about here.

But I come in all


different directions.

In fact, if I do it this I
don't know, 20, 30, 500 times,

you can see that these


dots are starting

to trace out a curve,


[inaudible] very slightly

to the left, right there.

And [inaudible] very slightly


to the right, as well.

Why not?

There we go.

Right there.

And my question
is, yes, these dots

seem to be tracing a curve.

What curve could that be?

Well, since I seem to be


talking about circles,

I'm probably going to


guess that it is a circle.

In which case, all


these green dots

should be some equidistant


from a given special dot.
So my first question is, what
could be the possible center

in this picture?

What location seems to be


special in this diagram?

Well, if you think


about that for a moment,

I think when I'm


thinking equidistance,

maybe the midpoint of my two


nails is going to be special.

All right.

So let's give a name


to that distance.

I'll call it r
inches and r inches.

So the question is, do these


dots from the 90-degree corner

of my piece of paper
actually lie on a circle

with maybe this midpoint-- which


is a guess-- as its center?

That is, could I


prove, for example,

this point up here is also


r inches from that midpoint?

Oh, that seems tricky.

The only thing I


know about this point

up here is coming from the


corner of a piece of paper,

so it really is part
of a right triangle.

So let me draw on
that right triangle.

So it comes in from a corner


of a piece of paper like so,

and has a 90-degree angle.


So is there any
reason to believe

its distance from the


midpoint is also r inches?

Doesn't look like it right now.

Hmm.

But I do have a right triangle,


with a right angle in it,

for sure, which means the other


two angles-- so I'll call them,

say, x degrees and y


degrees-- we definitely have

that x plus y is
the remaining 180.

Already got 90, so x plus


y must be the other 90

degrees in that triangle.

Does that help at all?

Not sure.

All right.

So I'm going to stare


at this for a while

and wait for an epiphany.

And this could be a


very boring lecture,

if you're going to wait for


me to have an epiphany, so I'm

going to speed this


along a little bit.

I did have an epiphany one day


when I realized a clever thing

to do is to recopy
this right triangle,

but as a rotated version


underneath itself.

Here goes.

What I mean by that, literally


copy this right triangle,

put the right angle down here,


so I flipped it 180 degrees.

This angle x is now over here.

This angle y is now over here.

And I guess this line


I'm talking about now

looks like that, but down below.

Now why is that helpful?

Well, I see shape with two


90-degree angles in it,

and I see two angles, x plus y.

But what's x plus y?

It's 90 degrees.

So actually, these two


triangles, together,

make a perfect rectangle.

And look at the


lines I've drawn.

The original line segment is


one diagonal of the rectangle,

and it looks like the other line


segment is the other diagonal.

And if you remember,


we actually proved

that the diagonals of a


rectangle are congruent,

they're the same length, and


they cut each other perfectly

in half-- which is perfect,


because this diagonal has

the length 2 r, twice the


radius, or twice the length r,

and, which case, this other


diagonal also has length 2 r,

but it's cut in half,


must be r and r.
So I said radius.

I was actually right.

It really is the
radius of a circle.

This point is, indeed, r


inches from that midpoint.

In fact, the same argument


would work for any point

r inches, r inches, r inches,


coming from this construction

of paper.

So we've just proved


that actually bringing

in a 90-degree angle up
between two given points

is, in fact, going to trace


a perfect circle for us.

Absolutely brilliant.

So this has led us to


some theory of circles.

It's not quite Thales' theorem.

In fact, that's the


backwards version of Thales'.

So to get to Thales' proper,


let's talk about some jargon

and notation for circles.

So just to be absolutely
clear, a circle

is a set of points all the same


distance from a given point.

And that given


point's usually called

the center of the circle.

All right?

The notation for this, if I


give the center point a name,
say, maybe center
point p, geometrists

like to write circle


with a dot in it,

p, to be read as the circle


with center labeled p.

Grand.

So that's the
notation for circles.

And then we give special


names to features of circles.

For example, if I draw a line


segment from that center, p,

out to some point,


a, on the circle,

then that line segment's


called a radius of the circle.

For that line segment,


it's technically a picture.

So there's a picture of
a radius of a circle.

But we sometimes use the word


radius for just the number,

for the length of


that line segment.

So the word radius is


used in two senses--

as an actual picture
of a line segment,

or as a number being the


length of that line segment,

just the distance between


the points p and a.

So there is the
radius of a circle.

More jargon?

If I have a line
segment which is

connecting any two points on


the circle, it's called a chord.

And if that chord happens to


also go through the center,

then the chord is


given a special name.

It's called the


diameter of the circle.

Again, it's actually


two senses of the word.

It could actually be the


physical picture of a diameter,

or [inaudible] length of
that chord, in which case,

the diameter is a number.

And as numbers, it's pretty


clear from the picture,

that the diameter


equals twice the radius.

Grand.

Carrying on, a line


that goes infinitely

far in both directions, that


cuts the circle at two places,

is called a secant line.

Just a name.

And if that line happens


to just touch the curve,

then it's called a tangent line.

Actually, this is
kind of curious.

If you think of our


lecture on trigonometry,

Joachim Rhaeticus
gave the name tangent

to a ratio of sides
of a right triangle,

and he also gave the name


secant to another ratio of sides
in a right triangle.

It's very curious he chose


the words secant and tangent

for that work.

Hmm.

That's a little
bit of a mystery.

Anyhow, carrying on, if I just


name two points on the circle,

they will define a region,


some part of the circumference

of the circle.

Circumference is the whole


outer rim of the circle.

And recall the region


between two points, the arc.

So example, here's
a shaded region

of the circumference
of the circle,

and that would be called the


arc between the points A and B,

as denoted with a little


curved symbol above the names

of the two points.

Now this is actually a


little bit confusing,

because there technically


are two arcs in that picture.

And you see the small


one that I've shaded.

But there's also the big


one that goes all the way

outside that shaded region.

So there's a little
bit of ambiguity.

When I say arc AB, I could


mean one of two arcs.

It's become the


convention to assume

that the author doesn't specify,


he or she means the small arc.

In fact, the name of the


smaller arc is the minor arc,

and the name of the larger


one is the major arc.

So if the author wanted


me to think of the bigger

arc in this picture,


she would say, please

think about the major arc AB.

She doesn't specify otherwise,


assumes the smaller one.

All right.

Carries on with this jargon.

One usually measures


arcs, and, now this

is strange, by an angle,
because an arc represents,

actually, an amount of turning.

So if, for example, I say the


arc AB has measure 62 degrees,

I don't mean the


length of 62 degrees.

Doesn't make any sense.

I mean the amount of


turning it represents

is 62 degrees of turning.

So it's actually an angle.

So the measure of
an arc is an angle.

And to be very clear,


the actual angle

sits right at the


center of the circle.

If I were to draw the two radii


to the end points A and B,

that makes an angle,


and that angle

is 62 degrees in this picture.

So it really does represent


62 degrees of turning.

It's a little confusing.

People tend not to


draw the central angle.

That's the name


of an angle coming

to the center of the circle.

People tend not to


draw those radii,

and just write the


numbers on the arcs.

But if you need to, you can


always draw back in the radii

and see where that


actual angle lives.

Now, the reason why


this is confusing,

because on different
sized circles, arcs

can have exactly


the same measure

and clearly not be


the same length.

For example, here's a whole


series of concentric circles

with the same center.

And all these three arcs have


the same measure of 50 degrees.

In fact, you can see


they each represent

the same amount of turning


along there own circumferences.

However, clearly, the lengths,


the actual physical lengths,

of each arc is different.

So watch out.

Arc are measured by


angles representing

the mass of turning,


have nothing

to do with their lengths.

All right.

The jargon continues.

If I'm given an arc, I


might, for some reason,

need to draw an angle


coming from that arc.

And what I mean by that


is an inscribed angle,

draw an angle from


the two end points

to some point on the circle.

And that creates an angle


called the inscribed angle

for that arc.

In fact, here on the


picture on the left,

I have three inscribed


angles from the same arc.

Some people call them


peripheral angles,

but I think the preferred term


nowadays is inscribed angles.

If the arc is particularly


big, like the one on the right,

then the inscribed angle


can be a bit tricky to see.

But it is an angle, if
you look at the picture,

coming from the two


endpoints of the arc

to some other point on


the rest of the circle.

So the inscribed angle is


kind of a big, obtuse angle

down near the bottom there.

Bit tricky to
read, but there it.

Ah!

All right.

I feel like that's


enough jargon for now.

What I'd like to do is


go back to the whiteboard

and play with some ideas.

So let me give you


a practice question.

So here's a very typical


geometry practice question.

Suppose I've got two circles,


circle with center A,

and circle with center


B. And I'll tell you

these are congruent circles,


which I haven't defined yet,

so I have to think what


that means in a moment.

But here's two circles


with center A and center

B. I will tell you


they're congruent.

I'll give you some information.

The two points of


intersection are

going to be labeled C and D. And


in this question I'm making up,
I'll tell you that AC is 17
units, and AB is 30 units.

And my question for


the day is, what

is the length of the common


chord of these two circles?

So what chord do these two


circles have in common?

Well, if I look at CD, I


think this is the chord.

It's definitely a chord


of the left circle,

circle A. It's a chord of the


right circle, circle B. It's

a chord in common.

I want the length of


that common chord, CD.

And the only piece I


haven't put on the board

is that these are


congruent circles.

Oh, I shouldn't circle that.

It's a bit confusing.

All right.

What's going on here?

Well, starters, let me draw


the information we've got.

We're told that the


length AC is 17,

so at least I know this


length here is 17 units.

We're told that AB is 30 units.

OK?

Let me draw that.

30 units.

That's all I've got.


And somehow I have to
figure out length CD.

Very scary.

All right.

Let's take it slowly.

Congruent.

Two circles are


probably congruent

means they're identical


in some sense.

After all, congruent


line segments

were line segments


the same length.

Congruent angles were angles


that had the same measure.

So I'm guessing congruent


circles means two circles that

are basically the same.

They must have the same radius.

So I'm going to guess since


the radius of the circle A

is clearly 17-- there's a


radius-- then the circle B also

has radius 17.

In fact, if I draw this radius,


there must be 17, as well.

I bet that's what


congruent means.

I keep going.

There's another radius.

A to D is a radius, so
that must be 17, and B to D

is also a radius of
the right-hand circle.

All right?
Now I've got some more
information on the picture.

Can I get to the length CD yet?

Hmm.

It still looks hard.

All right.

Stare at this for a while.

Does anything come to me?

Ah-- I'm seeing four 17s.

This shape is
actually a rhombus.

And, in fact, if I remember


from my days of equidistance,

since I've got to a point


equidistant from A and B,

another point
equidistant from A and B,

they must lie on the


perpendicular bisector of AB.

In fact, we did prove that


the diagonals of a rhombus

bisect each other and


are perpendicular.

So, which case the line I


want, the dotted line, CD,

must be perpendicular
to AB, and they're

chopping each other in half.

Now my picture's
getting messy, but I

think I'm basically there.

Because 30, A to B is 30.

It gets chopped in half.

Let me erase that out


then and write 15 and 15.
And I have a right
triangle to play with,

and I can't help


but think Pythagoras

whenever I see a right triangle.

Look at this.

Here's a right triangle.

One side of 15, one side I


don't know, hypotenuse is 17.

However, I can now work


out at least this side

of the right triangle.

Let me do Pythagoras's theorem.

All right.

So 15 squared plus blank


squared equals 17 squared.

Rearrange that
formula in my mind.

This must be the square root


of 17 squared minus 15 squared.

All right, arithmetic.

That's always the hardest part.

17 squared.

For some reason, I have


that number in my head.

It's 289.

15 squared.

I actually have that


number in my head, too.

It's 225.

There's no shame in
getting out a calculator

to just find these things out.

289 and 225.

They differ by 64.


This is the square root of 64.

Square root of 64?

Ah.

The author of this


question was very

nice, chose very nice numbers.

It's 8.

Oh!

But that's the answer


to the question.

This part is 8 units


long, and it's half of CD.

So CD must actually be 8
units plus another 8 units.

We've got it.

It's 16 units long.

That is lovely.

That circle's brought back


the ideas of rhombuses,

and properties of
equidistance, and so forth.

It's not actually


surprising it does that.

After all, circles are


sets of equidistant points.

All of these things


will come back to us.

But let me bring back


some old ideas, as well,

and play with this some more


in a very beautiful way.

Let's just clean the board,


give yourself some space.

Do you recall that the shortest


way to get from a point

to a line?
We proved it's given by-- excuse
me-- the perpendicular line

segment.

Hold on to that thought.

Suppose I gave us a
circle with some center,

and I draw a tangent line.

Remember, a tangent
line's a line

that just comes and touches.

Maybe that's the point


of contact right there.

My question is,
what's the shortest

way to get from the center


of the circle to anywhere

on that tangent line?

Now, if I draw a length like


this, which is clearly wrong,

but just to get a feel


for what's happening,

you can see that this distance


is the radius of the circle

plus a bit more.

This length is the radius of the


circle plus still a bit more.

Maybe that's shorter.

This length is the radius of


the circle plus a bit more.

But this length to


the point of contact

is just the radius


of the circle.

So I can actually
say, that must be

the shortest way to get from


the center to the tangent line.
Go straight to the
point of contact.

But, by what we know


before, the shortest path

actually makes a 90-degree


angle to the line.

So we just proved a
pretty amazing fact.

Whenever you have a


radius to a tangent line,

it must make a 90-degree


angle, because it

has to be the shortest


path to the tangent line.

I call that the


Radius-Tangent theorem.

That's a very handy result.

The Radius-Tangent theorem,


radii and tangents always

make 90-degree angles


at the point of contact.

But let's keep going.

Here's a second charming


little result. Instead

of talking about
one tangent, what

if I talked about two tangents?

Here's a circle.

There's its center.

I'm going to choose a


point outside the circle.

And I want to draw the


whole tangent lines,

but there's two tangent


lines I can draw,

and I'll just draw the segments


to each point of contact.
There's one tangent
line segment.

Well, I mean, it's actually,


if it were a full line

it would be just touching once.

Here's a second
tangent line segment,

meaning if it were a
full line it would just

touch the circle just once.

My question is,
is there anything

interesting going on in
this picture of, what is it?

An ice cream cone, or a


megaphone, or something?

OK.

So I guess the first question


is, what's the question?

I can't help but wonder that


this may be something going on

with the lengths.

If this is a inches long,


and that's b inches long,

doesn't it feel suspiciously


like it could be that a equals

b?

Could we prove that?

Let's try.

So we've just established


the radius tangent theorem

and I feel very


compelled to draw

some radiuses to these tangents,


so I'm going to do that.

For, with the


Radius-Tangent theorem,
I know that's going to
make a 90-degree angle.

Ditto for the second tangent.

All right.

That feels helpful.

Does that prove that


a equals b in any way?

Oh, I don't to see it yet.

Well, these are both radiuses.

They're at the same length.

I can write that.

Well, as with many proofs


in geometry, sometimes

it's good to draw just


a single diagonal line

[inaudible] triangle.

So let me draw
this line, as well.

It's some length.

I don't know-- call it h.

Make up a name.

Does that prove


that a equals b yet?

Yes, it does.

We have a right triangle.

r squared plus a squared


equals h squared.

So if I do the algebra
on that, by Pythagoras,

this must be the square root


of h squared minus r squared.

Look at the second


triangle down below.

It's also a right triangle.

r squared plus a squared


equals b squared.

Do the algebra on this.

b squared must be
the square root

of h squares minus r squared.

It's identical.

Yes.

a does equal b.

So we've just proved what I


like to call the Two Tangents

theorem.

Whenever two tangents


from an outside point

meet a circle at the


two points of contacts,

then those two segments


of those tangents

are going to be the same length.

Grand.

Actually, that's a lot to go on.

We can do some more


practice questions.

Let's have a look


this sort of exercise.

Try this.

Here's a whole bunch of


circles, and a whole bunch

of line segments coming


from those circles.

Very confusing.

Hmm.

The question is, what is the


length of the line segment x?

Without knowing any


geometry, I could just
guess the answer is going
to be 6, because 6 is

the only number on the board.

But actually, can we


see why it has to be 6?

Well, yes.

Look at the first circle.

These are two tangent


line segments,

and we just proved


they're congruent.

So this length is 6.

These are a pair of


tangent line segments.

They must be congruent, so


this is 6 and 6, and 6 and 6,

and 6 and 6, in which


case, x is, indeed, 6.

Ah.

Grand.

How about this question?

I've drawn a hexagon


around a circle.

All right.

So far, so good.

Then, I'm going to


ask, in doing so,

I alternated colors-- blue,


red, blue, red-- with my pens.

Did I use more blue ink than


red ink when I drew this figure?

Or did I use more red ink


than blue when I drew this?

Or is it impossible to tell?

Hmm.

All right.
Well, let's have a look at this.

Ah-- I actually see lots


of tangent segments.

This section of blue is


a tangent line segment,

which is matched with


this tangent line segment.

So actually, they're
the same length.

That portion of the blue


matches that portion of the red.

In fact, this next


portion of the red

is also a tangent line segment.

Matches this
portion of the blue.

Another tangent side of


the segments must match.

So if I go all the
way around, there's

always a portion of blue that's


matched with a portion of red,

and a portion of red that's


matched with a portion of blue.

We have, in fact, matching


tangent line segments

all over the place.

In fact, I can now say the


amount of red I used here

in this diagram is exactly the


same as the amount of blue,

which is crazy.

Draw a random hexagon around


a circle, alternate colors,

the amount of red and blue ink


you'll use is exactly the same.

All right.
Let me now come to what I
think is my most favorite,

and what I think is the


most stunning, theorem

in all of circle geometry.

It goes as follows.

If I draw an arc, and I draw a


whole bunch of inscribed angles

from that arc, and if I


actually measure the angles

in these inscribed positions,


you will find-- well,

I will find.

If you try it, you


will find that they're

all the same measure.

Then it gets even


more surprising.

If this red arc


represented, say,

30 degrees of turning,
than each inscribed angle

equals 15 degrees,
actually half the measure

of the turning of the arc.

In fact, in general,
we can prove

that no matter which inscribed


angle you draw from an arc,

it is guaranteed to be half
the measure of the arc.

Seems completely unbelievable,


it's utterly surprising,

and it's truly beautiful.

So what I need to do is see why


that could possibly be true.

Here goes.
It's going to be a
bit involved, so it's

going to be a bit of a roller


coaster ride for our brains.

But we can do it.

So here's a picture.

I've just drawn one inscribed


angle, angle x, coming

from an arc of measure y.

So our goal is to prove


that x is half of y.

Just to be very clear,


y is actually an angle.

It's actually an angle from


the center of the circle.

So let me draw that


in, to be very clear.

There's the actual


angle y in the picture.

So we're going to prove that x


is half of y, or y is double x.

Either way would do.

I can't help but be tempted


to draw it in a radius.

That feels compelling,


so I'm going to do it.

But when I do that, I actually


get an isosceles triangle

on the left.

Radius and radius


makes isosceles.

And an isosceles
triangle on the right.

I've actually got two


isosceles triangles

in this picture, which means


I must have congruent base

angles going on.


Those two angles
a are base angles

of an isosceles triangle.

Ditto for the angles b.

Which means x, at the


very least, is a plus b.

All right.

But what does that do for me?

I mean, there's a formula


for x, in terms of a and b.

I wonder if I can get a formula


for y, in terms of a and b.

And hopefully, it will


be double that formula.

Well, there's 180


degrees in a triangle.

The triangle on the


left has two angles a,

so it means the remaining,


unlabelled angle

must be 180 minus 2 a.

Ditto for the


triangle on the right.

It's blank angle must


be 180 minus 2 b.

So is it getting
me to angle y yet?

Why, yes.

Because I happen to
know that there's

a full 360 degrees


around that center.

So y plus 180 minus 2


a, plus 180 minus 2 b,

must be a full 360 degrees.

That's a mouthful, but


I'm going to write it
out and see what it gives me.

If I do the algebra, 360


minus 2 a minus 2 b plus y

is the full 360.

It gets me that y is
2 a plus 2 b-- that

is, y is double a plus b.

Yes.

x was a plus b, y is double a


plus b, y is, indeed, double x,

or x is half of y.

Now, I was particularly


nice for us here, I

chose an inscribed angle


that was sort of positioned

above the central angle.

It is possible that
the inscribed angle

could be lopsided to one


side, in which case, we'll

have to do this proof


again for that case.

But I suggest you


draw in a radius.

And I bet you could chase


through some angles,

just like we did now,


using isosceles triangles,

because I happened to see


several isosceles triangles

in that picture, and I


bet you can prove again

that y is double x.

Wow.

So that gives us the what I


call the Inscribed/Central Angle
theorem, that a whole
bunch of inscribed angles

from a central angle that


is coming from the same arc,

and have the same


measure, and, in fact,

their measure is half


the measure of that arc.

In fact, when my students


see this picture,

they sometimes call this theorem


the Sydney Opera House theorem.

And you kind of see why.

It's absolutely beautiful.

All right.

So we're almost at
the point now where

we can finally, finally


prove Thales' theorem, which

I mentioned way
back in lecture 1,

and we've gone all this


length of this course so far

and have never


actually proved it.

Here it is.

Now, Thales' theorem


was about angles coming

from the diameter of a circle.

Here's a diameter.

How much turning does


the diameter represent?

Well, for the arc


underneath a diameter,

that's half the circle, so


that's 180 degrees of turning.

Now, we just proved


that any inscribed angle
coming from an arc must have
half the measure of that arc.

What's half of 180 degrees?

90.

So any angle coming


from this diameter

must have measure 90 degrees.

And there is Thales'


theorem, at long, long last.

Now, I should
mention, this is not

how Thales proved


this result first off.

There's actually a much


easier way to do it.

And I invite you to actually do


the isosceles triangle method.

I'm going to say,


draw this radius here,

and you will see two


isosceles triangles.

Using those isosceles


triangles, I

bet you can prove that


the angle at the top,

just like we did earlier,


has a nice property-- namely,

that the angles must add up


to 90 degrees at the corner.

So there we have it.

Paper pushing, if I'd


start with 90-degree angles

and bring them up,


and make a curve,

that curve must be a circle.

And reverse, if
we have a circle,
then any angle coming
from that diameter

must, in fact, be a right angle.

This is absolutely beautiful.

It's brought the whole


story full circle,

so it's a lovely place


to end for today.

I feel very satisfied.

Thanks very much.

You might also like