Topic 5. Harder Unit Topics.

Level 3.
Problem HAR3_01.
1 1
If a > 0 , show that a 2 + 2
≥a+ ≥ 2.
a a

2 2 2
 1  1  1 1 1
Solution:  a +  =  a −  + 4 ⇒  a +  ≥ 4 ⇒ a + ≥ 2 (equality iff a = , i.e., a = 1 );
 a  a  a a a
1  1  1 1  1
a a
(
a
)
a 2 + 2 −  a +  = a 2 − a +  2 −  = a(a − 1) − 2 ( a − 1) = ( a − 1)  a − 2  =
a  a
1
 a 
( a − 1)
3

( a − 1) a = ( a − 1)( a − 1)
(a 2
+ a +1 ) = (a − 1) ( a
2
2
+ a +1 ) ≥0⇒ a2 +
1 1
≥ a + with
2 2 2 2
a a a a
equality iff a = 1.

Problem HAR3_02.
If a ≥ 0 , b ≥ 0 and c ≥ 0 , show that ab + bc + ca ≥ a bc + b ca + c ab .

x+y
Solution: (x + y ) 2
= (x − y ) + 4 xy ≥ 4 xy ⇒ if x ≥ 0 and y ≥ 0 we get
2
xy ≤ . Hence
2
ab + ca
a bc = ab ca ≤ .
2
ab + cb
Similarly b ac = ab cb ≤ ,
2
ac + cb
and c ab = ac cb ≤ .
2
By addition a bc + b ca + c ab ≤ ab + bc + ca (equality iff a = b = c ).

Problem HAR3_03.
If a 1 ≥ 0 , a 2 ≥ 0 , b1 ≥ 0 and b 2 ≥ 0 , show that (a 1 + b1 )(a 2 + b 2 ) ≥ a 1 a 2 + b1 b 2 .

Solution: By squaring the sides of the inequality to be proved we get:

(a 1 + b1 )(a 2 + b 2 ) ≥ a 1 a 2 + b1 b 2 + 2 a 1 a 2 b1 b 2 .
x+y
Use the inequality xy ≤ with x = a 1 b 2 and y = a 2 b1 :
2
a 1 b 2 + a 2 b1
a 1 a 2 b1 b 2 = a 1b 2 a 2 b1 ≤ .
2

Problem HAR3_04.
Show that if a , b and c are the sides of a triangle, then (a + b − c )(b + c − a )(c + a − b ) ≤ abc .
Hint: Show that if x ≥ 0 , y ≥ 0 and z ≥ 0 , then ( x + y )( y + z )( z + x ) ≥ 8 xyz .

Solution: (x + y ) = (x − y )
2 2
+ 4 xy ≥ 4 xy .
Hence ( x + y ) ≥ 2 xy ,
similarly ( z + y ) ≥ 2 zy ,
and ( x + z ) ≥ 2 xz .
By multiplication ( x + z )( y + z )( z + x ) ≥ 8 xyz .
The substitution x = a + b − c , y = b + c − a , z = c + a − b gives
2 b 2 c 2 a ≥ 8 (a + b − c )(b + c − a )(c + a − b ).
Hence abc ≥ (a + b − c )(b + c − a )(c + a − b ).
Note that since a , b and c are sides of a triangle, x , y and z are non-negative.

Problem HAR3_05.
If a 1 , a 2 , ⋅ ⋅⋅, a n are positive numbers such that a 1 a 2 ⋅ ... ⋅ a n = 1, show that
(1 + a )(1 + a ) ⋅ ... ⋅ (1 + a ) ≥ 2
1 2 n
n
.

Solution: (1 + x ) = (1 − x ) + 4 x ≥ 4 x ⇒ 1 + x ≥ 2 x for x ≥ 0 .
2 2

Hence (1 + a 1 )(1 + a 2 ) ⋅ ... ⋅ (1 + a n ) ≥ 2 a 1 2 a 2 ⋅ ... ⋅ 2 a n = 2 n a 1 a 2 ⋅ ... ⋅ a n = 2 n .

Problem HAR3_06.
1 1 1
If a > 0 , b > 0 and c > 0 , show that (a + b + c ) + +  ≥ 9 .
a b c

Solution: Deduce that (a + b + c )(ab + bc + ca ) ≥ 9 abc and hence deduce that

a b + b c + c a + ab + bc + ca ≥ 6 abc .
2 2 2 2 2 2

Problem HAR3_07.
If a > 0 , b > 0 , c > 0 and d > 0 , show that
16 3 3 3 3 1 1 1 1
≤ + + + ≤ + + +
a +b+c+ d b+c+ d a+b+c a+ b+ d a+c+ d a b c d
1 1 1
Hint: Use the inequality ( x + y + z ) + +  ≥ 9 with positive x , y and z .
x y z

9 1 1 1
Solution: It is clear that ≤ + + . Hence
x+y+z x y z
9 1 1 1 9 1 1 1
≤ + + , ≤ + + ,
a+b+c a b c a+c+d a c d
9 1 1 1 9 1 1 1
≤ + + , ≤ + + .
a+b+d a b d b+c+d b c d
By addition
3 3 3 3  1 1 1 1
+ + + ≤ + + + 
b+c + d a+b+c a+b+ d a+c + d  a b c d
( equality iff a = b = c = d ).
1
Furthermore, if we employ the inequality t + ≥ 2 , t > 0 , We have
t
 1 1 1 1  a b  a c   b d 
(a + b + c + d )  + + +  = 1 + 1 + 1 + 1 +  +  +  +  +  +  +
 a b c d  b a  c a  d b 
 b c  c d   d a
 +  +  +  +  +  ≥ 4 + 2 + 2 + 2 + 2 + 2 + 2 = 16 .
 c b  d c   a d 
 1 1 1 1
Hence ( a + b + c + d )  + + +  ≥ 16 ( equality iff a = b = c = d ).
 a b c d
The substitution a → a + b + c, b → b + c + d , c → c + d + a, d → d + a + b in this
1 1 1 1
inequality permits to obtain the desired factor + + + ,
b+c+d a+b+c a+b+d a+c+d
and we get
 1 1 1 1 
3( a + b + c + d ) + + +  ≥ 16 ,
 a+b+c a+b+ d b+c + d a+c + d
( equality iff a = b = c = d ).
Then
16 3 3 3 3
≤ + + + .
a +b+c+ d b+c+ d a+b+c a+ b+ d a+c+ d
Finally, using the inequalities given above
3 11 1 1 3 1  1 1 1
≤  + + , ≤  + + ,
b + c + d 3 b c d  a + b + c 3  a b c
3 1 1 1 1 3 1 1 1 1
≤  + + , ≤  + + ,
a + b + d 3 a b d  a + c + d 3 a c d 
we come to the desired result
16 3 3 3 3 1 1 1 1
≤ + + + ≤ + + +
a +b+c+ d b+c+ d a+b+c a+ b+ d a+c+ d a b c d
( equality iff a = b = c = d ).

Problem HAR3_08.
2
a +b +c a +b +c
2 2 2

Show that ≥  .
3  3 

Solution: a 2 + b 2 = (a − b ) + 2 ab ≥ 2 ab
2

Hence 2 ab ≤ a 2 + b 2 (equality iff a = b ),

Similarly 2 bc ≤ c 2 + b 2 and 2 ac ≤ a 2 + c 2 .
By addition 2 ab + 2 bc + 2 ac ≤ 2 a 2 + b 2 + c 2 .( )
But (a + b + c ) = a + b + c + 2 ab + 2 bc + 2 ca
2 2 2 2

(
So (a + b + c ) ≤ 3 a 2 + b 2 + c 2 .
2
)
a +b +c 
Hence   ≤
2
a +b +c
2 2 2
(
(equality iff a = b = c ).
)
 3  3

Problem HAR3_09.
6 6
x y
If x ≠ 0 and y ≠ 0 , show that x 4 + y 4 ≤ 2
+ 2
.
y x
Solution: Let a = x 2 and b = y 2 . Then we can rewrite the inequality to be proved in the
3 3
a b
equivalent form a + b ≤ 2 2
+ , a > 0 and b > 0 .
b a
Multiplication by ab > 0 gives ab a 2 + b 2 ≤ a 4 + b 4 . ( )
a +b
2 2

Clearly a 2 + b 2 = (a − b ) + 2 ab ⇒ ab ≤
2
(1)
2

) ≤ (a ) 2
+b2 a + b + 2a b
(
2 4 4 2 2

Hence ab a + b 2 2
= .
2 2
a +b
4 4

The substitution a → a 2 , b → b 2 in (1) gives a 2 b 2 ≤ ⇒ 2a 2b 2 ≤ a 4 + b 4 .

2
a + b + 2a b a +b +a +b
( )≤
4 4 2 2 4 4 4 4

Hence ab a + b 2 2
≤ .
2 2

( )≤ a
3 3 6 6
a b x y
So ab a + b 2 2 4
+ b and a + b ≤
4 2 2
+ ⇒x +y ≤ 4 4
2
+ 2
.
b a y x

Problem HAR3_10.
a c lg a lg c
If a > 1, b > 1 and ñ > 1 such that ≥ , show that ≥ .
b a lg b lg a

a c lg b + lg c
Solution: ≥ ⇒ a 2 ≥ bc ⇒ lg a 2 ≥ lg (bc ) ⇒ 2 lg a ≥ lg b + lg c ⇒ lg a ≥ .
b a 2
Clearly ( x + y ) = ( x − y ) + 4 xy ≥ 4 xy ⇒ ( x + y ) ≥ 4 xy . Hence for x ≥ 0 and y ≥ 0
2 2 2

x+y lg b + lg c
≥ xy . The substitution x = lg b and y = lg c gives ≥ lg b lg c .
2 2
lg a lg c
Hence lg a ≥ lg b lg c . Squaring, (lg a ) ≥ lg b lg c ⇒ ≥
2
.
lg b lg a

Problem HAR3_11.
x +y
2 2

If x > y and xy = 1, show that ≥2 2 .

x−y

Solution: The multiplication of the inequality to be proved by x − y > 0 gives

x + y ≥ 2 2 (x − y ) .
2 2

Squaring, x 2 + y 2 ( ) 2
(
≥ 8 x 2 + y 2 − 2 xy = 8 x 2 + y 2 − 2 . ) ( )
( )
2
(
Hence x 2 + y 2 − 8 x 2 + y 2 + 16 ≥ 0 . Let t = x 2 + y 2 then it suffices to show that )
t − 8 t + 16 ≥ 0 . This is the case as the discriminant of this quadratic ∆ = 8 − 4 ⋅ 16 = 0 .
2 2

Problem HAR3_12.
Show that the geometric mean of n positive numbers can not exceed their arithmetic mean, i.e.,
x + x 2 + ... + x n
if x 1 , x 2 ,..., x n are positive numbers, show that ( x 1 x 2 ... x n ) ≤ 1
1/ n
.
n
x −1
Hint: Show that x ≤ e for all x .
Solution: Let f (x ) = e x −1 − x . Then f ' ( x ) = e x −1 − 1 . Clearly f ' ( x ) = 0 when x = 1,
f ' ( x ) > 0 when x > 1 and f ' ( x ) < 0 when x < 1 . Hence f ( x ) has an absolute minimum of 0
x 1 + x 2 + ... + x n
xk
x −1 xk xk −1
when x ≤ e . Let A = . The substitution x → gives ≤e A
, k = 1,..., n .
n A A
x1 x 2 xn x x x 
By multiplication ⋅⋅⋅ ≤ exp  1 − 1 + 2 − 1 ... n − 1 .
A A A A A A 
x1 x2 xn 1
But −1 + − 1 ... − 1 = ( x 1 + x 2 + ... + x n ) ⋅ −n =0.
A A A A
x1 x 2 xn
⋅⋅⋅ ≤ 1 ⇒ x 1 x 2 ... x n ≤ A n . So ( x1 x 2 ... x n ) ≤ A.
1/ n
Hence
A A A

Problem HAR3_13.
1 1 1 
If a 1 , a 2 ,..., a n are positive numbers, show that (a 1 + a 2 + ... + a n ) ⋅  + + ... +  ≥ n 2 .
 a1 a 2 an 

a 1 + a 2 + ... + a n 1 1
Solution: We have (a 1 a 2 ... a n ) ≤ . The substitution a 1 → ,..., a n →
1/n

n a1 an
1 1 1
1/ n + + ... +
1 1 1  a1 a2 an
gives  ...  ≤ .
 a1 a 2 a n  n
1 1 1  1
By multiplication (a 1 + a 2 + ... + a n ) + + ... +  ⋅ 2 ≥ 1 .
 a1 a 2 an  n
1 1 1 
Hence (a 1 + a 2 + ... + a n ) + + ... +  ≥ n 2 .
 a1 a 2 an 

Problem HAR3_14.
1 1 1
If a > 0 , b > 0 , ñ > 0 and a + b + c = 1, show that 2
+ 2
+ 2
≥ 27 .
a b c

1 2 1 1
1 1  1
Solution: Clearly = (a + b + c )  2 + 2 + 2 ; a + b + c = 1 .
+ +
a c 
2 2 2
a b c b
Use the inequality between the arithmetic mean and the geometric mean for a , b , c and then for

a
1
2
1 1
, 2 , 2 . We have
b c
a +b +c 3
≥ abc ⇒ (a + b + c ) ≥ 9 3 abc ,
2

3
2
( )
1 1 1
+ +
a
2
b
2
 1 1 1 
c
2
1 1 1 1
≥3  2 ⋅ 2 ⋅ 2  ⇒ 2 + 2 + 2 ≥3 .
3 a b c  a b c 3 2
a b c
2 2

2 1 1 1 
By multiplication, (a + b + c )  2 + 2 + 2  ≥ 9 3 abc 3
2 1
= 27 . ( )
a b c  3 2
a b c
2 2

Problem HAR3_15.
If a 1 , a 2 ,..., a n are positive numbers, show that
 a 1 + a 2 + ... + a n ≥ na 1 a 2 ... a n
n n n
(i)
n
(ii) n
a 1 a 2 ... a n ≥ .
1 1 1
+ + ... +
a1 a2 an

Solution: Use the inequality between the arithmetic mean and the geometric mean
x + x + ... + x n
(x1 x 2 ... x n )1 / n ≤ 1 2 , x 1 > 0 ,..., x n > 0 .
n
(i) The substitution x 1 = a 1n ,..., x n = a nn gives
a 1 + a 2 + ... + a n
(a )
n n n
n 1/ n
n
1
n
a ... a
2 n ≤ ⇒ a 1 + a 2 + ... + a n ≥ na 1 a 2 ... a n . n n n

n
(ii) The substitution x 1 = a 1−1 ,..., x n = a n−1 gives
1 1 1
1/ n + + ... +
1 1 1  a1 a2 an n
  ≤ ⇒ (a 1 a 2 ... a n ) ≥
1/n
... .
 a1 a 2 a n  n 1 1 1
+ + ... +
a1 a2 an

Problem HAR3_16.
1 1 1
Show that x − x 3 < tan −1 x < x − x 3 + x 5 for x > 0 .
3 3 5

1
Solution: Let us prove that for t > 0 < 1− t2 + t4 . 1− t2 <
1+ t 2

First, it is easily seen that (1 − t )(1 + t ) = 1 − t < 1 for t > 0 , and hence
2 2 4

1
1− t2 < for t > 0 . Further, it is clear that (1 − t 2 + t 4 )(1 + t 2 ) − 1 = t 6 > 0 for t > 0 .
1+ t 2

1
Thus, we have < 1 − t 2 + t 4 for t > 0 .
1+ t 2

1
Hence, we arrive to the desired result 1 − t 2 < < 1 − t 2 + t 4 for t > 0 .
1+ t2
By integrating the last inequality between 0 and x, we derive
x x x
dt 1 3 1 3 1 5
∫ (1 − t )dt < ∫ 1 + t 2 < ∫ (1 − t + t )dt , x − 3 x < tan x < x − 3 x + 5 x for x > 0 .
2 2 4 −1

0 0 0

Problem HAR3_17.
Show that xy ≤ e x −1 + y ln y for all real x and all positive y . When does equality hold?

Solution: Let f ( x ) = e x −1 + y ln y − xy be the function with parameter y > 0 . It is easily to get

f ′( x ) = e x −1 − y , f ′( x ) = 0 ⇔ e x−1 = y or x = 1 + ln y .
Furthermore, f ′( x ) > 0 for x > 1 + ln y and f ′( x ) < 0 for x < 1 + ln y , as we can see that, if
x = ∆ x + 1 + ln y , then f ′( x ) = y ( e ∆ x − 1) , and f ′( x ) > 0 if ∆ x > 0, and f ′( x ) < 0 if ∆ x < 0 .
Hence f ( x ) has an absolute minimum of 0 when x = 1 + ln y . As a result, we get for all x
f ( x ) ≥ 0 , and e x−1 + y ln y ≥ xy with equality iff x = 1 + ln y .
Problem HAR3_18.
11 x (1 − x )4
1 4 1
Show that ∫ x 4 (1 − x )4 dx < ∫ dx < ∫ x 4 (1 − x )4 dx and hence deduce that
0 1+ x
20 2
0
22 1 22 1
− <π < − .
7 630 7 1260

Solution:
1 1 1 1 1
1 1  x5   x6   x7   x8   x 9 
∫ x (1 − x ) dx = ∫ ( x − 4 x + 6 x − 4 x + x )dx =  5  − 4 6  + 6 7  − 4 8  +  9 
4 4 4 5 6 7 8

0 0  0   0   0   0   0
1 2 6 1 1 1
= − + − + = .
5 3 7 2 9 630
x (1 − x )4
1 4
x − 4 x 5 + 6 x 6 − 4x 7 + x8
1 4
∫ dx = ∫ dx
0 1 + x 2
0 1 + x 2

By using the representation x 8 − 4 x 7 + 6 x 6 − 4 x 5 + x 4 = (1 + x 2 )( x 6 − 4 x 5 + 5 x 4 − 4 x 2 + 4) − 4 ,

we get
x (1 − x )4
1 4 1 1
dx
∫ 1 + x2 dx = ∫ ( x 6
− 4 x 5
+ 5 x 4
− 4 x 2
+ 4 ) dx − 4 ∫ 1 + x2
0 0 0
1 1 1 1
 x7   x6   x5   x3 
=   − 4  + 5  − 4  + 4[ x ]0 − 4 tan −1 x = − + 1 − + 4 − π = [ ]
1 1 1 2 4 22
− π.
 7  0  6  0  5  0  3  0 0 7 3 3 7
It is easily seen that for 0 < x < 1
1 1
< <1,
2 1 + x2
1 1 1 − x2 1 x2
because of − = > 0, 1 − = > 0.
1 + x 2 2 2(1 + x 2 ) 1 + x 2 1 + x2
1 4 x 4 (1 − x )4
Since x (1 − x ) > 0 , we get x (1 − x ) <
4 4 4
< x 4 (1 − x )4 .
2 1+ x 2

By integrating this inequality with respect to x between 0 and 1 , we deduce that

11 4 x (1 − x )4
1 4 1
∫ x (1 − x ) 4
dx < ∫ dx < ∫ x 4 (1 − x )4 dx .
0 1+ x
20 2
0
x (1 − x )
1 1 4 4
1 22
∫ x (1 − x ) dx = ∫ dx = − π , we obtain
4 4
In view of and
1+ x
2
0
630 0
7
1 22 1 1 22 1 22 1 22 1
< −π< ⇒− <π− <− ⇒ − <π< − .
1260 7 630 630 7 1260 7 630 7 1260

Problem HAR3_19.
π/ 2
π π 2
Show that 1 − e − π / 2 < ∫e
− sin x
dx < ( e − 1) . Hint: if > x > 0 , show that x > sin x > x .
0 2e 2 π
 π
Solution: Let us show that sin x < x for 0 < x < . It is easily to deduce that for
2
f ( x ) = x − sin x , we get f ′( x ) = 1 − cos x ≥ 0 . Hence for x ≥ 0 f ( x ) is a non-decreasing
function with absolute minimum 0 when x = 0 .
π
Thus f ( x ) > 0 for x > 0 , and sin x < x for 0 < x < .
2
2 π
Let us show that sin x > x for 0 < x < . It is not difficult to establish for
π 2
2 2 2
g( x ) = sin x − x that g ′( x ) = cos x − and g ′( x ) = 0 when x = arccos .
π π π
π
Furthermore, for 0 < x < function g( x ) has the only absolute maximum of
2
sin(arccos 2 / π ) − 2 / π arccos 2 / π > 0 when x = arccos 2 / π , since
2 2
g ′( x ) = cos x − > 0 for x < arccos ,
π π
2 2
g ′( x ) = cos x − < 0 for x > arccos .
π π
π
Function g( x ) reaches absolute minimum of 0 when x = 0, .
2
π 2
Thus g( x ) ≥ 0 for 0 < x < , that is, x < sin x , and, finally,
2 π
2 2 − x ⋅2 / π
x < sin x < x , − x > − sin x > − x , e > e −sin x > e − x .
π π
π
By integrating the last inequality with respect to x between 0 and , we come to
2
π/ 2 π/ 2 π/ 2 π/ 2
[ ] − x π/ 2
[
π − x ⋅2/ π
]
−x − sin x − x ⋅2/ π − sin x π/ 2
∫ e dx < ∫e dx < ∫e dx , − e
0
< ∫e dx < −
2
e
0
,
0 0 0 0
π/ 2
π
1 − e− π/ 2 < ∫e
− sin x
dx < ( e − 1) .
0 2e
 y

y = 1 /x

A
B

D C
0 a b x

Problem HAR3_20.
ln b − ln a 11 1
If b > a > 0 , show that <  + .
b−a 2a b

Solution: Shaded area < area ABCD ,

y = 1 /x

A
B
D C
0 a b x

ln b − ln a
b
1 1 1 1 11 1
∫ x dx < 2  a + b (b − a ),
a
b−a
<  + .
2a b

Problem HAR3_21.
b −a a +b
If b > a > 0 , show that < .
ln b − ln a 2

Solution: Rewrite the inequality to be proved in the equivalent form

b 
2 ( b − a ) < (ln b − ln a )( a − b ) or ln   ⋅ ( b + a ) − 2 ( b − a ) > 0 .
a
t  t  a
Let f ( t ) = ln  ( t + a ) − 2 ( t − a ) ⇒ f ( a ) = 0 , f ′( t ) = ln   + − 1 ⇒ f ′( a ) = 0 ,
a  a  t
1 a 1 a
f ′′( t ) = − 2 = 1 −  > 0 for t > a .
t t t t 
So f ′( a ) = 0 and f ′( t ) is an increasing function for t > a ⇒ f ′( t ) > 0 for all t > a .
But f ( a ) = 0 ⇒ f ( t ) > 0 for all t > a . Consequently,
b 
f ( b ) > 0 as b > a ⇒ ln  ( b + a ) − 2 ( b − a ) > 0 .
a 

Problem HAR3_22.
n ( n + 1)
2 2

Show that for n ≥ 1, 1 + 2 + 3 + ... + n = 3 3 3 3

.
4

Solution: Let u n = 1 3 + 2 3 + .. + n 3 ⇒ u n +1 = u n + ( n + 1) 3 , u 1 = 1, n = 1, 2 ,... Hence it suffices to

n ( n + 1) n ( n + 1)
2 2 2

show that u n = . Define the statement S ( n ) : u n = , n = 1, 2 ,...

4 4
1 (1 + 1)
2 2

Consider S (1) : u 1 = 1 = ⇒ S (1) is true. Let k be a positive integer. If S (k ) is true, then

4
k ( k + 1)
2 2

uk = .
4
k ( k + 1)
2 2

Consider S ( k + 1) : u k +1 = u k + ( k + 1) = 3
+ ( k + 1) 3 , if S ( k ) is true
4
( k + 1) ( k + 4k + 4) ( k + 1) ( k + 2 )
2 2 2 2

= = ⇒ S ( k + 1) is true.
4 4
Hence for all k ≥ 1, S ( k ) true implies S ( k + 1) true. But S (1) is true, hence by induction S (n ) is
n ( n + 1)
2 2

true for all n ≥ 1, 1 3 + 2 3 + 3 3 + ... + n 3 = u n = .

4

Problem HAR3_23.
1 2 3 n 1
Show that for n ≥ 1, + + + ... + =1− .
2! 3! 4! ( n + 1)! ( n + 1)!

1 2 3 n 1
Solution: Define the statement S ( n ) : + + + ... + =1− , n = 1, 2 ,...
2! 3! 4! ( n + 1)! ( n + 1)!
1 1 1
Consider S(1) : n = 1, = 1 − = . Hence S(1) is true.
2! 2! 2
1 2 3 k 1
Let k be a positive integer. If S ( k ) is true, then + + + ... + =1− .
2! 3! 4! ( k + 1)! ( k + 1)!
Consider S ( k +1) . As S ( k ) is true, we get
1 2 3 k k +1 1 k +1
+ + + ... + + =1− + =
2! 3! 4! ( k + 1)! ( k + 2 )! ( k + 1)! ( k + 2 )!
k + 2 − ( k + 1) 1
= 1− = 1− .
( k + 2)! ( k + 2)!
Hence for all positive integers k , S ( k ) true implies S ( k +1) true. But S(1) is true, therefore by
induction, S ( n ) is true for all positive integers n :
1 2 3 n 1
+ + + ... + =1− for n ≥ 1.
2! 3! 4! ( n + 1)! ( n + 1)!
Problem HAR3_24.
x x ( x + 1) x ( x + 1)...( x + n − 1) ( x + 1)( x + 2 )...( x + n )
Show that for n ≥ 1 , 1 + + + ... + = .
1! 2! n! n!

Solution: Define the statement S ( n ) :

x x ( x + 1) x ( x + 1)...( x + n − 1) ( x + 1)( x + 2 )...( x + n )
1+ + + ... + = , n = 1, 2 ,... .
1! 2! n! n!
x x +1
Consider S(1) : n = 1, 1 + = 1 + x = . Hence S(1) is true.
1! 1
Let k be a positive integer. If S ( k ) is true, then
x x ( x + 1) x ( x + 1)...( x + k − 1) ( x + 1)( x + 2 )...( x + k )
1+ + + ... + = .
1! 2! k! k!
Consider S ( k +1) . As S ( k ) is true, we get
x x ( x + 1) x ( x + 1)...( x + k − 1) x ( x + 1)...( x + k )
1+ + + ... + +
1! 2! k! ( k + 1)!
( x + 1)( x + 2 )...( x + k ) x ( x + 1)...( x + k ) ( x + 1)( x + 2 )...( x + k )( k + 1 + x )
= + = .
k! ( k + 1)! ( k + 1)!

Hence for all positive integers k , S ( k ) true implies S ( k +1) true. But S(1) is true, therefore by
induction, S ( n ) is true for all positive integers n :
x x ( x + 1) x ( x + 1)...( x + n − 1) ( x + 1)( x + 2 )...( x + n )
1+ + + ... + = .
1! 2! n! n!

Problem HAR3_25.
Show that for n ≥ 1, 2 ⋅ 1!+5 ⋅ 2!+10 ⋅ 3!+... + ( n 2 + 1) n! = n ( n + 1)! .

Solution: Define the statement S ( n ) : 2 ⋅ 1!+5 ⋅ 2!+10 ⋅ 3!+... + ( n 2 + 1) n! = n ⋅ ( n + 1)! , n = 1, 2 ,... .

Consider S(1) : n = 1, 2 ⋅ 1! = 1 ⋅ 2 !. Hence S(1) is true.
Let k be a positive integer. If S ( k ) is true, then
2 ⋅ 1!+5 ⋅ 2!+10 ⋅ 3!+... + ( k + 1) k ! = k ⋅ ( k + 1)! .
2

Consider S ( k +1) . As S ( k ) is true, we get

2 ⋅ 1!+5 ⋅ 2!+10 ⋅ 3!+... + ( k + 1) k !+ (( k + 1) + 1)( k + 1)! = k ( k + 1)!+ (( k + 1) + 1)( k + 1)! =
2 2 2

= ( k + 1)!( k + ( k + 1)2 + 1) = ( k + 1)!( k + 1)( k + 2 ) = ( k + 2)!( k + 1) .

Hence for all positive integers k , S ( k ) true implies S ( k +1) true. But S(1) is true, therefore by
induction, S ( n ) is true for all positive integers n :
2 ⋅ 1!+5 ⋅ 2!+10 ⋅ 3!+... + ( n + 1) n! = n ⋅ ( n + 1)! for n ≥ 1.
2

Problem HAR3_26.
n ( n − 3)
Show that a convex polygon with n ≥ 4 sides has diagonals.
2

Solution: Let the function f ( n ) define the quantity of diagonals for a convex polygon with
n ≥ 4 sides. It is easily seen that f ( n + 1) = f ( n ) + n − 1 (see figure 15), since including an
additional point An+1 for a polygon with n sides leads to new n − 2 diagonals with respect to
the points A1 , A 3 ,..., A n −1 , besides the side A1 An becomes a new diagonal.

n ( n − 3)
Define the statement S ( n ) : f ( n ) = , n = 4 ,5 ,... . An+1
2
4 ⋅1
Consider S ( 4 ): f (4) = = 2 ⇒ S (1) true.
2
Let k be a positive integer, k ≥ 4 . If S ( k ) is true, then A1 An
k ( k − 3)
f (k ) = . A2
2 An-1
A3
Consider S ( k +1) . As S ( k ) is true, we get
k ( k − 3) k 2 − k − 2 ( k + 1)( k − 2)
f ( k + 1) = f ( k ) + k − 1 = + k −1= =
2 2 2
.
We see that for all k ≥ 4 , S ( k ) true implies S ( k +1) true. But S( 4) is true. Hence by induction,
S ( n ) is true for all integers n ≥ 4 .

Problem HAR3_27.
Show that (1 + x ) n − nx − 1 is divisible by x 2 for n ≥ 2 .

Solution: Define the statement S ( n ) : (1 + x ) n − nx − 1 is divisible by x 2 , n = 2 ,3,...

Consider S ( 2 ) : (1 + x ) 2 − 2 x − 1 = x 2 ⇒ S ( 2 ) is true.
Let k be a positive integer more that 1. If S (k ) is true, then (1 + x ) k − kx − 1 = x 2 M , where M is
a polynomial in x ⇒ (1 + x ) k = 1 + kx + x 2 M .
Consider S ( k + 1) : (1 + x ) k +1 − ( k + 1) k +1 − ( k + 1) x − 1 = (1 + x )(1 + x ) k − ( k + 1) x − 1
= (1 + x )(1 + lx + x 2 M ) − ( k + 1) x − 1, if S ( k ) is true
= x 2 M + kx 2 + x 3 M = x 2 ( m + k + xm ) ⇒ (1 + x ) k +1 − ( k + 1) x − 1 is divisible by x 2 .
Hence for all positive integers k ≥ 2 , S ( k ) true implies S ( k + 1) true. But S ( 2 ) is true, hence by
induction S (n ) is true for all positive integers n ≥ 2 : (1 + x ) n − nx − 1 is divisible by x 2 .

Problem HAR3_28.
Show that for n ≥ 10 , 2 n > n 3 .

Solution: Define the statement S ( n ) : 2 n > n 3 , n = 1, 2 ,...

Consider S (10 ) : 2 10 = 1024 > 10 3 ⇒ S (10 ) is true.
Let k be a positive integer, k ≥ 10 . If S (k ) is true, then 2 k > k 3 .
Consider S ( k + 1) : 2 k +1 = 2 ⋅ 2 k > 2 ⋅ k 3 , if S ( k ) is true. But
2 k − ( k + 1) = k − 3 k − 3 k − 1 > k − 3 k − 3 k − k = k ( k − 7 ) > 0 ⇒ 2 k > ( k + 1) .
3 3 3 2 3 2 2 2 2 3 3

Hence 2 k +1 > ( k + 1) 3 ⇒ S ( k + 1) is true.

Hence for all positive integers k ≥ 10 , S ( k ) true implies S ( k + 1) true. But S (10 ) is true, hence
S (n ) is true for all n ≥ 10 , 2 > n .
n 3

Problem HAR3_29.
 Show that for n ≥ 2 , n n > ( n + 1) n −1 .

Solution: Define the statement S ( n ) : n n > ( n + 1) n −1 , n = 2 ,3,...

Consider S ( 2 ) : 2 2 > ( 2 + 1) ⇒ S ( 2 ) is true. Let k be a positive integer, k ≥ 2 . If S (k ) is true,
then k k > ( k + 1) k −1 .
k +1 k +1
( k + 1) ( k + 1)
Consider S ( k + 1) : ( k + 1) k +1 = k
⋅kk > k
⋅ ( k + 1) k −1 , if S ( k ) is true
k k
( k + 1)
2k

= k
.
k
( k + 1)
2k

> ( k + 2 ) k ⇔ ( k + 1) 2 k > (k ( k + 2 ) ) ⇔ ( k + 1) 2 > k ( k + 2 ) ⇔

k
Let us show that k
k
k
2
+ 2k + 1 > k 2
+ 2k .
So we have ( k + 1) k +1 > ( k + 2 ) k ⇒ S ( k + 1) is true. Hence for all k ≥ 2 , S ( k ) true implies
n −1
S ( k + 1) true. But S ( 2 ) is true, hence by induction S (n ) is true for all n ≥ 2 , n > ( n + 1) .
n

Problem HAR3_30.
Show that for n ≥ 1, 1 +
1
+
1
+ ... +
1
(
> 2 n +1 −1 . )
2 3 n

1 1 1 1
Solution: Let u n = 1 + + + ... + ⇒ u n +1 = u n + , u 1 = 1, n = 1, 2 ,...
2 3 n n +1
Hence it suffices to show that u n > 2 ( n + 1 − 1) .
Define the statement S ( n ) : u n > 2 ( n + 1 − 1), n = 1, 2 ,...
Consider S (1) : 1 > 2 ( 2 − 1) ⇔ S (1) is true.
Let k be a positive integer. If S (k ) is true, then u k > 2 ( k + 1 − 1) .
1 1
Consider S ( k + 1) : u k +1 = u k + > 2 ( k + 1 − 1) + , if S ( k ) is true.
k +1 k +1
1
But 2 ( k + 1 − 1) + > 2 ( k + 2 − 1) ⇔ 2 ( k + 1) + 1 > 2 ( k + 1)( k + 2 ) ⇔
k +1
( 2 k + 3 ) > 4 ( k + 1)( k + 2 ) ⇔ 4 k + 12 k + 9 > 4 k 2 + 12 k + 8 . So we have that
2 2

u k +1 > 2 ( k + 2 − 1) ⇒ S ( k + 1) is true. Hence for all k ≥ 1, S ( k ) true implies S ( k + 1) true. But

S (1) is true, hence S (n ) is true for all n ≥ 1, 1 +

1
+
1
+ ... +
1
(
> 2 n +1 −1 . )
2 3 n

Problem HAR3_31.
1 1 1 4 1 
Show that for n ≥ 1, + + ... + ≤ 1 − n  .
(1! )
2
( 2! )
2
( n! )
2
3 4 

Solution: Show that for n ≥ 1, n ! ≥ 2 n −1 .

Define the statement S ( n ) : n! ≥ 2 n −1 , n = 1, 2 ,... .
Consider S (1) : 1! ≥ 2 0 = 1 . Hence S(1) is true.
Let k be a positive integer. If S ( k ) is true, then k ! ≥ 2 k −1 .
Consider S ( k +1) . As S ( k ) is true, we get
k +1 ( k + 1)
( k + 1)! = k !( k + 1) ≥ 2k −1( k + 1) = 2k ⋅ ≥ 2k , since ≥ 1 for k ≥ 1 .
2 2
Hence ( k + 1)! ≥ 2k , k ≥ 1.
We see that for all positive integers k , S ( k ) true implies S ( k +1) true. But S(1) is true. Hence
n −1
by induction, S ( n ) is true for all positive integers n ≥ 1 : n! ≥ 2 .
1 1 1 4 1 
Define the statement S ( n ) : + + ... + ≤ 1 − n , n = 1, 2 ,... .
(1! )
2
( 2! )
2
( n! ) 3
2
4 
4 1 1
Consider S (1) : ≤ 1 −  = 1 . Hence S(1) is true.
(1! ) 3 4 2

Let k be a positive integer. If S ( k ) is true, then

1 1 4 1  1
+ + ... +
≤ 1 − k  .
(1! )
2
( 2! ) (k!)
2
3 4 
2

Consider S ( k +1) . As S ( k ) is true, we get

1 1 41 1  1 1
+ + ... + ≤ 1 − k  + + .
(( k + 1)! ) 3 4  (( k + 1)! )
2 2 2 2 2
(1! ) ( 2! ) (k!)
1 1
Since k ! ≥ 2 k −1 for k ≥ 1 we get ≤ k . By using this inequality, we come to
(( k + 1)!) 2
4
1 1 1 1 4 1  1 4 1 
+ + ... + + ≤ 1 − k  + k = 1 − k +1  .
(( k + 1)! ) 3 4  4 3 
2 2 2 2
(1! ) ( 2! ) (k!) 4
Hence for all positive integers k , S ( k ) true implies S ( k +1) true. But S(1) is true. Hence by
induction, S ( n ) is true for all positive integers n :
1 1 1 4 1 
+ + ... + ≤ 1 − n  .
3 4 
2 2 2
(1! ) ( 2! ) ( n! )

Problem HAR3_32.
Show that 35 n + 3 ⋅ 7 n + 2 ⋅ 5 n + 6 is divisible by 12 for n ≥ 0 .

Solution: Let u n = 35 n + 3 ⋅ 7 n + 2 ⋅ 5 n + 6 , then

u n +1 = 35 ⋅ 35 + 21 ⋅ 7 + 10 ⋅ 5 + 6 = ( 36 − 1) ⋅ 35 + ( 24 − 3 ) ⋅ 7 + (12 − 2 ) ⋅ 5 + (12 − 6 )
n n n n n n

= 36 ⋅ 35 n + 24 ⋅ 7 n + 12 ⋅ 5 n + 12 − u n = 12 ( 3 ⋅ 35 n + 2 ⋅ 7 n + 5 n + 1) − u n
Define the statement S ( n ) : u n is divisible by 12, n = 1, 2 ,...
Consider S ( 0 ) : u 0 = 12 is divisible by 12 ⇒ S ( 0 ) is true.
Let k be an integer, k ≥ 0 . If S (k ) is true, then u k = 12 ⋅ M for some integer M .
S ( k + 1) : u k +1 = 12 ( 3 ⋅ 35 + 2 ⋅ 7 + 5 + 1) − u k =
k k k

Consider
12 ( 3 ⋅ 35 + 2 ⋅ 7 + 5 + 1) − 12 M , if S ( k ) is true.
k k k

u k +1 is divisible by 12 ⇒ S ( k + 1) is true. Hence for all integer k ≥ 0 , S ( k ) true implies

S ( k + 1) true. But S ( 0 ) is true, hence by induction S (n ) is true for all integers n ≥ 0 :
35 + 3 ⋅ 7 + 2 ⋅ 5 + 6 is divisible by 12.
n n n

Problem HAR3_33.
Show that 7n + 15n is divisible by 11 for odd n ≥ 1.
Solution: Let us introduce f ( n ) = 7n + 15n . It is easily seen that
f ( n + 2) = 7n+ 2 + 15n+ 2 = 72 ( 7n + 15n ) − 49 ⋅15n + 225 ⋅15n = 49 f ( n ) + 176 ⋅15n
Define the statement S ( n ) : f ( n ) is divisible by 11, n = 1,3,5 ,... .
Consider S(1) : f (1) = 7 + 15 = 22 = 11 ⋅ 2 ⇒ S (1) is true, since f (1) is divisible by 11.
Let k be a positive odd integer. If S ( k ) is true, then f ( k ) = 11 ⋅ M for some integer M .
Consider S ( k + 2 ) ( k + 2 is the next odd integer). As S ( k ) is true , we get
f ( k + 2) = 49 f ( k ) + 15k ⋅ 176 = 49 ⋅ 11 M + 15k ⋅ 16 ⋅ 11 = 11( 49 ⋅ M + 15k ⋅ 16) .
Since 49 M + 16 ⋅ 15k is integer, we see that f ( k + 2 ) is divisible by 11.
Hence for all odd positive integers k , S ( k ) true implies S ( k + 2 ) is true. But S(1) is true.
Therefore by induction, S ( n ) is true for all odd positive integers n : 7n + 15n is divisible by 11
for odd n ≥ 1.

Problem HAR3_34.
The equation x 2 − x + 1 = 0 has roots α and β , and A n = α n + β n for n ≥ 1 .
(i) Without solving the equation, show that A1 = 1, A 2 − 1 and A n = A n −1 − A n − 2 for n ≥ 3 .
nπ
(ii) Hence show by induction that for n ≥ 1, A n = 2 cos .
3

Solution: (i) If α, β are roots of x 2 − x + 1 = 0 , then α + β = 1 and αβ = 1 by Vieta’s theorem.

Hence (α + β) = 1 ⇒ α 2 + β 2 = 1 − 2αβ ⇒ α 2 + β2 = −1 , An = α n + β n = α n −1 ⋅ α + β n −1 ⋅ β ,
2

α + β = 1 ⇒ α = 1 − β and β = 1 − α . Hence An = α n −1(1 − β) + β n −1(1 − α ) ⇒

( )
An = α n −1 + β n −1 − α n −1β − β n−1α ⇒ An = An −1 − αβ α n −2 + β n −2 ; αβ = 1 ⇒
An = An−1 − An −2 for n ≥ 3 .
nπ
(ii) Define the statement S ( n) : A n = 2 cos , n = 1, 2 ,... .
3
π
Consider S (1) : n = 1, A1 = 2 cos = 1 ⇒ S (1) is true.
3
2π
Consider S (2) : n = 2, A2 = 2 cos = −1 ⇒ S (2) is true.
3
Let k be a positive integer, k ≥ 2 . If S ( n) is true for all integer n ≤ k , then
nπ
An = 2 cos , n = 1, 2, 3,... k .
3
Consider S ( k + 1) . As S ( n) is true for n = 1, 2 ,3,..., k , we get
kπ ( k − 1)π  kπ ( k − 1) π 
Ak +1 = Ak − Ak −1 = 2 cos − 2 cos = 2  cos − cos .
3 3  3 3 
 a + b   b − a 
But cos a − cos b = 2 sin   sin   , a, b ∈ R.
 2   2 
 (2 k − 1) π   1 π  π  1 π 
Hence Ak +1 = 4 sin  ⋅  sin  − ⋅  ⇒ Ak +1 = −4 sin sin   k −   ⇒
 2 3  2 3 6  2 3 
 π π π
Ak +1 = −2 sin  ( k + 1) −  ⇒ Ak +1 = 2 cos( k + 1) .
 3 2 3
Hence for k ≥ 2 , S ( n) true for all positive integers n ≤ k implies S ( k + 1) is true. But S (1) ,
S (2) are true. Therefore by induction, S ( n) is true for all positive integers n :
nπ
An = 2 cos for n ≥ 1 .
3

Problem HAR3_35.
If u 1 = 1, u 2 = 1 and u n = u n −1 + u n − 2 for n ≥ 3 , show that for
 n
 1 − 5  
n
1  1+ 5
n ≥1 un =   −  .
5  2   2  


1   
n n
 1 + 5 
 −  1 − 5   , n = 1, 2 ,... .
Solution: Define the statement S ( n ) :u n =    2 
5  2    

1 + 5 1 − 5 
1
Consider S (1): u1 =  −  = 1 ⇒ S (1) is true.
 2 5 2 
 2 2
1  1 + 5   1 − 5   1 4 5 
Consider S ( 2 ): u2 =   −   =   = 1 ⇒ S(2) is true.
5  2   2   5 4 
 
Let k be a positive integer, k ≥ 2 . If S ( n ) is true for all integers n ≤ k , then
1   1 − 5  
n n
 1 + 5
un =   − 
 2   for n = 1, 2 ,..., k .
5  2    

Consider S ( k + 1):

1   1 − 5   1  1 + 5   1 − 5  
k k k −1 k −1
 1 + 5
u k +1 = u k + u k −1 =   −   
  2   + 5  2  −  2  
 
5  2         

as S ( k − 1), S ( k ) are true

1  
k −1 k −1
 1 + 5 1 + 5  1 − 5  1 − 5
=   ⋅  + 1  −  ⋅ + 1 
5  2 
  2   2 
  2 

1   1 − 5   1  1 + 5   1 − 5  
k −1 2 k −1 2 k +1 k +1
 1 + 5  1 + 5  1 − 5 
=  ⋅   −   
  2  ⋅  2   = 5  2  −  2  ,
   
5  2   2           

Hence for all integers k ≥ 1 S ( n ) true for all positive integers n ≤ k For k = 2, 3, K, S ( n ) is
true for all positive integers n ≤ k implies S ( k +1) true. But S (1), S ( 2) are true. Hence by
induction, S ( n ) is true for all positive integers n:
 n
 1 − 5  
n
1  1+ 5
un =   −  .
5  2   2  


Problem HAR3_36.
 2un3 −1 + 27
If u1 = 1 and un = for n ≥ 2 , show that
3un2−1
(a) u n > 3 for n ≥ 2 ;
(b) u n +1 < u n for n ≥ 2 .

Solution: (a) Define the statement S ( n ) : u n > 3, n = 2 ,3,... .

2 + 27 29
Consider S ( 2 ): u2 = = > 3 ⇒ S ( 2) is true.
3 ⋅1 3
Let k be a positive integer k ≥ 2 . If S ( k ) is true then u k > 3 .
Consider S ( k +1) . Show that uk +1 > 3. To this end, let the function f ( x ) be given by
2 9
f ( x ) = x + 2 , x ≥ 3.
3 x
It is easily seen that
2 18
f ′( x ) = − 3 .
3 x
We obtain that f ′( x ) = 0 when x = 3 and f ′( x ) > 0 for x > 3 . Thus the function f ( x ) has an
absolute minimum of f (3) = 3 on the set x ≥ 3 . Hence
2 9
f ( x) = x+ 2
>3
3 x
Thus, as S ( k ) is true ( uk > 3, k ≥ 2 ), using this inequality, we get
2uk3 + 27 2 9
S ( k + 1): uk +1 = 2
= uk + 2 > 3 .
3uk 3 uk
Hence for all positive integers k ≥ 2 , S ( k ) true implies S ( k +1) true. But S( 2) is true, therefore
by induction, S ( n ) is true for all positive integers n ≥ 2: un > 3 .
(b) Show that un +1 < un for n ≥ 2 .
1 9 2 9
One can deduce that un − un +1 = un − 2 , since un +1 = un + 2 .
3 un 3 un
1 9
Let the function g( x ) be given by g( x ) = x − 2 , x ≥ 3 .
3 x
1 18
It is easy to see that g ′( x ) = + 3 > 0 for x ≥ 3 . Hence g( x ) is a monotonically increasing
3 x
1 9
function for x ≥ 3 and g(3) = 0 . Thus g( x ) > 0 for x > 3 , and x − 2 > 0, x > 3 .
3 x
1 9
By using this inequality and the fact proved in (a) that un > 3, n ≥ 2 , we get un − 2 > 0 .
3 un
Hence un +1 < un , n ≥ 2 .

Problem HAR3_37.
1 3 
If u1 = 1 and un =  un −1 +  for n ≥ 2 , show that
2 un −1 
(a) u n > 3 for n ≥ 2 ;
2

(b) u n +1 < u n for n ≥ 2 .

 1 9 
Solution: (a) Define the statement S ( n ) :u n2 =  u n2 −1 + 6 + 2  > 3, n = 2 ,3,... .
4 u n −1 

Consider S ( 2 ): u22 = (1 + 6 + 9) = 4 > 3 ⇒ S ( 2) is true.

1
4
Let k be a positive integer k ≥ 2 . If S ( k ) is true, then u k > 3 i.e.,
2

1 9 
uk =
2
 u k2 −1 + 6 + 2  > 3 .

4 u k −1 

Consider S ( k +1) . Show that uk +12 > 3 , that is,

1 9 
u k +1 =
2
 u k2 + 6 + 2  > 3 .
4  u k 
1 9
Show that S ( k +1) is true. Let the function f ( x ) be given by f ( x ) =  x + 6 +  , x ≥ 3.
4 x
1 9
It is easy to see that f ′( x ) =
1 − 2 
4 x 
⇒ f ′( x ) = 0 when x = 3 and f ′( x ) > 0 for x > 3 .
Hence the function f ( x ) has an absolute minimum of 3 when x = 3 , and thus
1 9
 x + 6 +  > 3 for x > 3 .
4 x
1 2 9
By using this inequality, in view of uk2 > 3 , we get uk2+1 =  uk + 6 + 2  > 3.
4 uk 
Hence for all positive integers true implies S ( k +1) true. But S( 2) is true, therefore by
induction, S ( n ) is true for all positive integers n ≥ 2: un 2 > 3 .

(b) Show that un +1 < un for n ≥ 2 .

1 3  1 3
As u n +1 =  u n +  , we have that un − un +1 = un − and un > 0 for n ≥ 2 .
2 un  2 2un
1 3
Let the function g( x ) be given by g( x ) = x − , x > 0 .
2 2x
1 3
It is easy to see that g ′( x ) = + 2
> 0 for all x > 0 .
2 2x
Hence g( x ) is a monotonically increasing function and g( x ) = 0 when x = 3 . Thus g( x ) > 0
1 3
for x > 3 , and x − > 0 for x > 3 .
2 2x
By using this inequality and the fact that un2 > 3 or un > 3 , proved in (a), we get
1 3
un − > 0 for n ≥ 2 . Hence un +1 < un for n ≥ 2 .
2 2un

Problem HAR3_38.
 ABC is a triangle inscribed in the a circle. P is a point on the minor arc AB . The points L, M,
and N are the feet of the perpendiculars from P to CA produced, AB, and BC respectively. Show
that L, M and N are collinear. (The line NL is called the Simpson line.)

L
P
A

B C
N

Solution:

L
P S
A

M
K
B C
N

In order to prove that L, M and N are collinear, it is sufficient to show that ∠ LMA = ∠ NMB .
For this purpose we show, that ∠ NMB = ∠ BPN = ∠ SPA = ∠ LMA .
The first step: ∠ NMB = ∠ BPN . The triangles PKM and BKN are rectangular and
BK NK
∠ PKM = ∠ BKN ⇒ ∆ PKM are similar ∆ BKN ⇒ = . But
PK MK
∠ PKB = ∠ MKN ⇒ ∆ PKB are similar ∆ MKN ⇒ ∠ NMB = ∠ BPN .
The second step: ∠ BPN = ∠SPA . The point P lies on the circle ⇒ PACB is a cyclic
quadrilateral ⇒ ∠ PAC + ∠ PBC = 180 ° . But ∠ PAC + ∠ PAL = 180 ° . Hence ∠ PBC = ∠ PAL .
From here, as the triangles PNB and PLA are rectangular, we have ∆ PNB are similar ∆ PLA ⇒
∠ BPN = ∠ APL .
The third step: ∠SPA = ∠ LMA . It is obvious that ∆ ALS is similar ∆ PMS, as these rectangular
PS MS
triangles have the common angle ∠ PSM . Hence = ⇒ ∆ MLS is similar ∆ PAS ⇒
AS LS
∠SPA = ∠ LMA .

Problem HAR3_39.
∆ABC is inscribed in a circle. P is a point on a minor arc AB . The points M, L and N are the
feet of the perpendiculars from P to AB produced, AC and BC respectively.
Show that ∆PNL ||| ∆PBA . Hint: use the fact that the points N, M, and L are collinear.

L
P
A

B C
N

Solution:
 L
P
A

M
K
B C
N

We use the known fact that the points N, M and L are collinear (the line NL is the Simpson line,
see problem 39). It suffices to show that ∠PNL = ∠PBA and ∠NPL = ∠BPA .
The first step: ∠PNL = ∠PBA . The rectangular triangles KMP and KNB are similar as
KN KB
∠PKM = ∠BKN ⇒ = . But ∠NKM = ∠BKP ⇒ ∆NKM ∆BKP ⇒ ∠PNL = ∠PBA .
KM KP
The second step: ∠NPL = ∠BPA . The sum of any quadrilateral is 360 0 . But the quadrilateral
PNCA has two right angles ⇒ ∠NPL + ∠BCA = 180 0 . The quadrilateral PBCA is a cyclic one
⇒ ∠BPA + ∠BCA = 180 0 (as opposite angles). Hence ∠NPL + ∠BCA = ∠BPA + ∠BCA ⇒
∠NPL = ∠BPA .

Problem HAR3_40.
ABCD is a cyclic quadrilateral. P is a point on the circle through ABC and D. PH, PX, PK and
PY are the perpendiculars from P to AB produced, BC, DC produced and DA, respectively.
(i) Show that ∆XPK ||| ∆HPY ;
PX ⋅ PK
2
( XK )
(ii) Hence show that PX ⋅ PY = PH ⋅ PK and = .
PH ⋅ PY ( HY )
2

Hint: Use the result of the problem 39.

 A
Y
D

X C
B
H
P K

Solution:

A
Y
D

X C
B
H
K
P

(i) The triangle ABD and CBD are inscribed in a circle. Hence according to the result of the
problem 39 ∆HPY is similar to ∆BPD and ∆BPD is similar to ∆XPK .
Therefore ∆XPK ||| ∆HPY .

PX PK PX XK
(ii) ∆ XPK is similar to ∆ HPY, hence = ⇒ PX ⋅ PY = PH ⋅ PK . Also = and
PH PY PH HY
PK
=
XK
, multiplying these equalities,
PX
⋅
PK
=
( XK ) 2

.
PY HY PH PY (HY ) 2