Geometry of the Cardioid

Arseniy V Akopyan

Abstract
In this note, we discuss the cardiod. We give purely geometric proofs of its well-known properties.

Definitions and Basic Properties

The curve in Figure 1 is called a cardioid. The name is derived from the Greek word
“καρδια” meaning “heart”. A cardioid has many interesting properties and very often
appears in different fields of mathematics and physics. The study of geometric properties
of remarkable curves is a classical topic in analytic and differential geometry. In this note,
we focus mainly on the purely synthetic approach to the geometry of the cardioid.
In the polar coordinate system, the cardioid has the following equation:
Fig. 1
r = 1 − cos ϕ. (*)

In this article, we consider the geometric properties of a cardioid, so let us give a geometric definition.
Take a circle of diameter 1 and let another circle of the same size roll around the exterior of the first one.
Then the trace of a fixed point on the second circle will be a cardioid (Fig. 2).

O0
P
V
O A
A
A

Fig. 2 Fig. 3

This definition is not so useful for studying a cardioid (but actually it will help us later), so we restate
the same definition in purely geometrical terms. Let ω be a circle with center O, A a point on it, and P
a point moving along ω (Fig. 3). Suppose O0 is the point symmetric to O in P . Let B be the reflection of
A in the perpendicular bisector to OO0 . Notice that ∠BO0 P = ∠P OA and O0 B = OA. Then, the locus
of points B is a cardioid.
The point A is called the cusp of the cardioid, and the point V of the intersection of the cardioid and
the ray AO is called the vertex (see Fig. 3).
Since the triangle AOP is isosceles and AB is parallel to OP , an easy angle count shows the following.
Lemma 1. AP is the bisector of the angle OAB.

1
 B B

Q
O0 Q
P

O A
A A
A
ω ω

Fig. 4 Fig. 5

Denote by Q the second point of intersection of AB and ω (Fig. 4). Note that in the quadrilateral
BO0 OQ, two of pair of opposite sides are parallel and other two are equal. Therefore it is a parallelogram
or an isosceles trapezoid. But it is not the isosceles trapezoid, since in this case it will coincide with
AOO0 B. Therefore, BQ equals O0 O, which, in turn, is equal to diameter of the circle ω. If the angle AOP
is acute, then the point Q lies outside the segment AB and the construction looks a little different from
Figure 4. But in this case, the proof is analogues to the considered case.
Thus, we obtain a kinematic definition of cardioid. Let ω be a circle of the diameter 1 and A a point
on it. Let us take a point Q that is moving along ω. Let B be the point on the line AQ such that BQ
equals 1, the diameter of ω (Fig. 5). Then B moves along a cardioid path. The are two choices for point B
(on either side of the point Q). Both of these points lie on the cardioid. From this construction it is easy
to see that this curve is defined by the equation (*) in a polar coordinate system with center at A. Thus,
all given definitions of a cardioid are equivalent.

Tangents to a Cardioid and its Image under Inversion

How can tangent lines to cardioids be constructed? Let us return to the very first definition and look
at the situation informally. Suppose a circle ω 0 is rolling around ω and at some moment these circles touch
each other at a point P (Fig. 6). How can we find the direction of the velocity vector of the marked point
B? From mechanics, we know that the speed of the point P (on the circle ω 0 ) is equal to zero and the
point is at rest. The speed of the point B is perpendicular to the segment BP , since the distance between
the points B and P does not change. We obtain the following statement.

Lemma 2. A tangent line to the cardioid at point B is perpendicular to BP .

Thus, the cardioid touches all circles centered at points P on the circle ω 0 having radius P B = P A. In
other words, the cardioid is the envelope of this family of circles (Fig. 7).

2
 B

ω0
P
P
P

A
A
A
A

Fig. 6 Fig. 7

Consider the images of these circles under the inversion with center A and with radius equal to the
diameter of ω. The circle ω maps to the line `, which is the perpendicular bisector of the segment with
ends at the cusp and the vertex of the cardioid. Consider any circle ω t from our family and let P be its
center. The ray AP crosses ω t at some point Y . Then, the image of the circle ω t under the inversion is
the line passing through the image of the point Y and perpendicular to AP . Since P lies on the circle
ω, its image is the point X of intersection of lines AP and `. Since Y is twice as far from A as P is, the
distance between its image and A is half of the distance between A and X. Thus, we have proved that the
inversion maps all circles from our family to the lines which are perpendicular bisectors of the segments
with ends at A and at points X moving along the line ` (Fig. 8).
It is well-known that all the above lines touch the parabola with the focus A and directrix `. Thus,
our cardioid is its inversion image.

X
Y

X
P
A A
A

`
`
` Fig. 9 Fig. 10
Fig. 8

We formulate this statement as a theorem and provide a rigorous proof.

Theorem 3. Let κ be a cardioid with the cusp A and the vertex V . Let ` be the perpendicular bisector
AV
of the segment AV . Then, the inversion with the center at A and the radius maps κ to the parabola
2
with the focus at A and directrix `.
Proof. We show directly that a point on the cardioid is mapped to a point on the parabola. Here, our
proof follows from the construction described above.
Denote the midpoint of AV by M and the intersection of AP and ` by P 0 (Fig. 11). Let P be a point
on the circle with diameter M A and let B be any point of the cardioid κ. Suppose B 0 is the image of

3
the point B. Then, it follows from the properties of the inversion that the quadrilateral P 0 P B 0 B is cyclic,
since the points P and P 0 are inverse image of each other.
Note that the following equalities on the angles hold:

∠B 0 P 0 P = ∠B 0 BP = ∠P AB = ∠M AP = 90◦ − ∠M P 0 A.

Thus, the angle M P 0 B 0 is right angle and the triangle AB 0 P 0 is isosceles. As a consequence, the
point B 0 is equidistant to the line ` and to the point A, i.e., B 0 lies on the parabola from the statement of
the Theorem.
Since inversions map tangent curves to tangent curves, we obtain another proof of Lemma 2.

P0 B0
P
M
A

`
Fig. 11

Some Properties of a Cardioid

Consider again Figure 4. From Lemma 1, and symmetry we see that the angle between BP and BA is
equal to the half of ∠OAB. Since the tangent line to the cardioid at the point B is perpendicular to BP ,
simple calculation of angles give us the following Lemma.

Lemma 4. The oriented angle between the tangent line to the cardioid at a point B and the line AV equals
3
90◦ − ∠BAV , where A and V are the cusp and the vertex of the cardioid.
2
By the oriented angle we mean the angle needed to rotate the tangent line clockwise so that it becomes
parallel to AV (if the angle is negative, we rotate counterclockwise).
We see that the angle of slope of the tangent line changes one and a half times faster than the angle
of slope of BA. Here are two nice corollaries of this observation.

Corollary 5. Suppose X and Y are two points on a cardioid and the segment XY passes through the
cusp. Then, tangent lines at points X and Y are perpendicular (Fig. 12).

Corollary 6. Let A be the cusp of a cardioid and X, Y , and Z three points on it. If

∠XAY = ∠Y AZ = ∠ZAX = 120◦ ,

then the tangent lines at X, Y , and Z are parallel (Fig. 13).

4
 X

X 120◦◦◦◦◦ Y
120

A 120◦◦◦◦◦
120
120
Y
Z

Fig. 12 Fig. 13

We continue studying the construction from Figure 4. Denote by C the second point of intersection of
the line AB and the cardioid. Let Ω be the circle with center at O and the radius OV . Let M and N be
the points of intersection of the rays OP and OQ with Ω (Fig. 14).

N
B
B
B
M

Q
O0
P
A
A
V
O C
ω

Ω
Fig. 14 Fig. 15

From the second definition of a cardioid, we know that Q is the midpoint of the segment BC. Therefore,
QN equals half of BC and the angle BN C is right angle. Let us show that BN and N C are tangent
lines to the cardioid. Since ∠P BQ = 21 ∠QAO and the triangle BQN is isosceles, we have ∠QBN =
90◦ − 12 ∠BQN = 90◦ − 21 ∠QAO. Therefore, the angle P BN is right angle and BN is a tangent line
(analogously for CN ). We showed, that the tangent lines from Corollary 5 are not only perpendicular, but
also intersect on the circle Ω.
Since the angles P OV and P OQ are equal (and equal to ∠QAO), we get that
1
∠V M O = ∠OM N = 90◦ − ∠QAO.
2
So, we have shown that a ray from the vertex V after reflecting in the interior of the circle Ω travels along
a tangent to the cardioid (Fig. 15). In other words, the cardioid is a caustic of the circle if a light source
lies on it. So it is possible to meet it in everyday life (Fig. 16, photo by Gérard Janot). In fact, you can
try this experiment yourself. Take a regular metal kitchen pan and a source of light such as a lighter, and
hold the light source near the edge of the pan. The light reflecting on the walls of the pan will form a
cardioid-like shape on the bottom of the pan.
Another interesting fact about the cardioid is that it is the central part of the well-known Mandelbrot
set (Fig. 17, picture by Wolfgang Beyer). For more details we recommend the book [4].

5
 Fig. 16 Fig. 17

Acknowledgments
The author thanks Vadim Gorin, Brittany Terese Fasy, Sergey Dorichenko, and Eugene Epifanov for
the discussion and useful advice.

References
[1]

[2] R. A. Johnson. Advanced Euclidean Geometry. Dover Publications, Inc, Mineola, New York, 2007.

[3] X. Lee, Cardioid, http://xahlee.info/SpecialPlaneCurves dir/Cardioid dir/cardioid.html.

[4] H. Peitgen and P. Richter. The beauty of fractals: images of complex dynamical systems. Springer,
1986.