You are on page 1of 5

Butterfly Theorem

Let M be the midpoint of a chord PQ of a circle, through which two

other chords AB and CD are drawn; AD cuts PQ at X and BC cuts PQ
at Y. Prove that M is also the midpoint of XY.

Proof 0 (William Wallace, 1805)

William Wallace's proof is based on the following diagram:

Proof 1 [Coxeter and Greitzer, p. 45]

Let's drop perpendiculars x1 and x2 from X and y1 and y2 from Y on AB and CD.
Let's also introduce a = MP = MQ and x = XM and y = YM. Observe several
pairs of similar triangles, which implies the following proportions:


and My
and My
and Cy
and By

x/y = x /y
x/y = x /y
x /y = AX/CY
x /y = XD/YB

from which
x/y = x1/y2 x2/y1 = x1/y1 x2/y2 = AXXD/ CYYB = PXXQ/ PYYQ.
So that
x/y = (a - x)(a + x)/ (a - y)(a + y) = (a - x)/(a - y) = a/a = 1.
And finally x = y.

Proof 2 [Shklyarsky, problem 104, solution 1]

Let O be the center of the given circle. Since OM XY, in order to show
thatXM = MY, we have to prove that XOM = YOM.
Drop perpendiculars OK and ON from O onto AD and BC, respectively.
Obviously, K is the midpoint of AD and N is the midpoint of BC. Further,
DAB = DCB and ADC = ABC,
as angles subtending equal arcs. Triangles ADM and CBM are therefore similar,
and AD/AM = BC/CM, or AK/AM = CN/CM. In other words, in triangles AKM
and CNM two pairs of sides are proportional. Also, the angles between the
corresponding sides are equal. We infer that the triangles AKM and CNM are
similar. Hence, AKM = CNM.
Now, have a look at the quadrilaterals OKXM and ONYM. Both have a pair of opposite straight angles, which implies that
both are inscribable in a circle. In OKXM, AKM = XOM. In ONYM, CNM = YOM. From which we get what we've been
looking for: XOM = YOM.

Proof 3 [Shklyarsky, problem 104, solution 2]

For convenience, denote the angles as on the diagram on the right. Let x =
XMand a = PM. As in Proof 1,


= a - x.

In DXM, by the Law of Sines, we have


= xsin()/sin(180 - ( + + ))
= xsin()/sin( + + ).

And in AXM
AX = xsin()/sin(),
which leads to
AXDX = xsin()sin()/sin()sin( + + ) = a - x.
From here we may find x:
x = asin()sin( + + ))/(sin()sin() + sin()sin( + + )).
The latter expression is symmetric in and . Therefore, if we repeat the derivation for the segment y = MY, we'll get exactly
same result. Hence, x = y.
According to Leon Bankoff (Mathematics Magazine, Volume 60, No. 4, October 1987, pp. 195-210) the appellationThe
Butterfly made its first appearance in the Solutions section of the American Mathematical Monthly in the February 1944
issue. Since the diagram of the theorem (with two wings on each side) established by Qiu Fawen and his students gives a
more realistic depiction of a butterfly, I suggest to call the statement A Better Butterfly Theoremwhich might be interpreted as
both A "Better Butterfly" Theorem and A Better "Butterfly Theorem", the former being my preference.

A Better Butterfly Theorem

Let there be two concentric circles with the common center O. A line crosses the two circles at points P, Q and P', Q', M
being the common midpoint of PQ and P'Q'. Through M, draw two lines AA'B'B and CC'D'D and connect AD', A'D, BC', and

B'C. (This is the Butterfly.) Let X, Y, Z, W be the points of intersection of PP'Q'Q with AD', B'C, A'D, and BC', respectively.


1/MX + 1/MZ = 1/MY + 1/MW.

(Since X coincides with Z and Y with W when the two circles coalesce into one, The Butterfly Theorem is an immediate
consequence of A Better Butterfly Theorem.)
The proof depends on the following


Let in RST, RU be a cevian through vertex R. Introduce angles = SRU and = URT. Then


sin( + )/RU = sin()/RT + sin()/RS.

The proof follows from the fact that Area( RST) = Area( RSU) + Area( RUT) by a consistent application of the sine
formula for the area of a triangle. (For example, Area( RST) = RSRTsin( + )/2.)

Proof of the theorem

We apply Lemma to triangles AMD', A'MD, B'MC, and BMC':


sin( + )/MX = sin()/MD' + sin()/MA,

sin( + )/MZ = sin()/MD + sin()/MA',
sin( + )/MY = sin()/MC + sin()/MB',
sin( + )/MW = sin()/MC' + sin()/MB.

In view of (3), (1) will follow from


sin()/MD' + sin()/MA + sin()/MD + sin()/MA' =

sin()/MC + sin()/MB' + sin()/MC' + sin()/MB,

or, which is the same, from


sin()(1/MA - 1/MB) + sin()(1/MA' - 1/MB') =

sin()(1/MC - 1/MD) + sin()(1/MC' - 1/MD').

Now, drop perpendiculars OM1 and OM2 from O onto AA'B'B and CC'D'D. M1 is the midpoint of both AB and A'B', whereas
M2 is the midpoint of CD and C'D'. Obviously,


MB - MA = MB' - MA' = 2OM = 2OMsin() and

MD - MC = MD' - MC' = 2OM = 2OMsin().

With (6) in mind, (5) is equivalent to


1/MAMB + 1/MA'MB' = 1/MCMD + 1/MC'MD'.

However, as is well known, MAMB = MCMD and MA'MB' = MC'MD'. Therefore, (7) is true, as is (5), which in turn, implies