1/23

Successive Differentiation
(For B.Sc./B.A. Part-I, Hons. And Subsidiary Courses of Mathematics)

Poonam Kumari
Department of Mathematics, Magadh Mahila College
Patna University
 2/23

Contents
1. Definitions and Notations

2. Derivation of nth differential coefficient of Some Standard Functions

3. Method of finding nth derivative of an Algebraic Rational Function

4. Leibnitz’s Theorem

1. Definitions and Notations

Successive differentiation is the differentiation of a function successively to derive its higher order
derivatives.

If y  f ( x ) be a function of x , then the derivative (or differential coefficient) of y w. r. t. x is

denoted by dy or Dy or f ' ( x ) or y 1 and this is called the first derivative of y w. r. t. x, where
dx
d
D .
dx
dy
If can be differentiated once again, i.e., if y  f ( x ) is derivable twice w. r. t. x , then the
dx
dy d2y 2
derivative of w. r. t. x is denoted by 2
or D y or f ' ' ( x ) or y 2 and this is called the
dx dx
second derivative of y w. r. t. x .

d2y
Similarly, if can be differentiated once again, i.e., if y  f ( x ) is derivable thrice w. r. t. x ,
dx 2
d2y d3y 3
then the derivative of 2
w. r. t. x is denoted by 3
or D y or f ' ' ' ( x ) or y 3 and this is called
dx dx
the third derivative of y w. r. t. x .
In a similar manner, we can find the fourth derivative, fifth derivative and, in general, the nth
derivative of y w. r. t. x by differentiating successively the given function y w. r. t. x four times,
five times and n times.
Following notations are generally used for the successive derivatives of y w. r. t. x :

First derivative Second derivative Third derivative …. nth derivative

or y1 y2 y3 …. yn
or f '( x) f ''( x) f '' ' ( x) …. f n(x)
or dy d2y d3y …. dny
dx dx 2 dx3 dxn
or Dy D2 y D3 y …. Dn y
 3/23

2
Example 1 : If x  y  y  x  c, prove that y 2  .
c2
Solution : Given that x y  yx c
Squaring both sides, we get

x  y  y  x  2 y2  x2  c2

This  2 y 2  x 2  c 2  2 y

Again, squaring both sides, we get

 
4 y 2  x 2  c 4  4c 2 y  4 y 2
2 2 4
This  4 x  4 c y  c  0

Differentiating both sides w. r. t. x , we get

8 x  4c 2 y1  0
2
This  2 x  c y1  0

Again, differentiating both sides w. r. t. x , we get

2  c 2 y2  0
2
This  y 2  .
c2

Example 2 : If y  a cos ( log x )  b sin(log x ), prove that x 2 y 2  xy1  y  0.

Solution : Given that y  a cos( log x )  b sin(log x ).

Differentiating both sides w. r. t. x , we get

1 1
y 1   a sin( log x ) .  b cos(log x ) .
x x

This  xy 1   a sin(log x )  b cos(log x )

Again, differentiating both sides w. r. t. x , we get

1 1
xy 2  y 1   a cos(log x ) .  b sin(log x ) .
x x
This  x 2 y 2  xy 1   a cos( log x )  b sin(log x )

 a cos( log x )  b sin(log x ) 

  y  y  a cos ( log x )  b sin(log x ) given 

2
This  x y 2  xy1  y  0.
 4/23

1 x
Example 3 : If y  e a sin , prove that (1  x 2 ) y2  xy1  a 2 y.
1 x
Solution :Given that y  e a sin .
Differentiating both sides w. r. t. x , we get

1 x  1 
y1  e a sin a
 2 
 1 x 

This  y1 
ay
1 x2
 y  e a sin 1 x
given 
 y1 1  x 2  ay
Squaring both sides, we get
2
y1 (1  x 2 )  a 2 y 2
Again, differentiating both sides w. r. t. x , we get
2
2 y1 y 2 (1  x 2 )  y1 ( 2 x )  a 2 ( 2 yy1 )

 
This  2 y1 (1  x 2 ) y 2  xy 1  2 y 1 ( a 2 y )

 (1  x 2 ) y 2  xy 1  a 2 y .

 log y 
Example 4 : If x  cosh  , prove that ( x 2  1) y 2  xy1  m 2 y  0.
 m 
 log y 
Solution : Given that x  cosh  .
 m 
log y
This  cosh 1 x 
m
Differentiating both sides w. r. t. x , we get

1 1
 y1
x2 1 my

This  y1 x 2  1  my

Squaring both sides, we get

2
y1 ( x 2  1)  m 2 y 2
Again, differentiating both sides w. r. t. x , we get
2
2 y1 y 2 ( x 2  1)  y1 ( 2 x )  m 2 ( 2 yy1 )

 
This  2 y 1 ( x 2  1) y 2  xy 1  2 y1 ( m 2 y )

 ( x 2  1) y 2  xy 1  m 2 y .
 5/23

2. Derivation of nth differential coefficient of Some Standard Functions

1. nth differential coefficient of x m , where m is a positive integer  n

Let y  x
m

Differentiating both sides w. r. t. x successively, we get

y1  m x m1
y2  m (m  1) x m2
y3  m (m  1)(m  2)x m3
………………………………..
………………………………..
Now we suppose that

y n  m (m  1)(m  2)(m  3)...............(m  (n  1)) x mn ........(1)

Differentiating both sides of equation (1) w. r. t. x , we get

y n 1  m (m  1)(m  2)(m  3)...............( m  ( n  1))(m  n) x m( n 1)

Clearly, y n 1 is of the same form as y n , given by equation (1). This  If equation (1) is true for a
particular value of n, then equation (1) is true for next higher value of n also.
But we have already proved that equation (1) is true for n = 1, 2, 3.
Since equation (1) is true for n = 3, it is true for n = 4 also. Similarly, since equation (1) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (1) is true for every positive integer n.
From equation (1), we can write y n as follows:

m (m  1)(m  2)(m  3)...............(m  (n  1))(m  n)(m  (n  1))...........3.2.1 m n

yn  x
(m  n)(m  (n  1))...........3.2.1

m!
i.e., yn  x mn
(m  n) !

Corollary : If y  x n , then y n  n !

2. nth differential coefficient of (ax  b)m , where m is a positive integer  n

Let y  ( ax  b)
m

Differentiating both sides w. r. t. x successively, we get

y1  m a(ax  b) m1
y2  m (m  1)a 2 (ax  b) m2
y3  m (m  1)(m  2)a 3 ( ax  b)m3
………………………………..
………………………………..
Now we suppose that
y n  m (m  1)(m  2)(m  2)...............(m  (n  1) )a n (ax  b) mn ........(2)
 6/23

Differentiating both sides of equation (2) w. r. t. x , we get

y n 1  m (m  1)(m  2)(m  3)...............(m  (n  1))(m  n)a n 1 (ax  b) m ( n 1)

Clearly, y n 1 is of the same form as y n , given by equation (2). This  If equation (2) is true for a
particular value of n, then equation (2) is true for next higher value of n also.
But we have already proved that equation (2) is true for n = 1, 2, 3.
Since equation (2) is true for n = 3, it is true for n = 4 also. Similarly, since equation (2) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (2) is true for every positive integer n.
From equation (2), we can write y n as follows:

m (m  1)(m  2)(m  2)...........( m  (n  1) )(m  n)(m  n  1).........3.2.1 n

yn  a (ax  b) m  n
(m  n)(m  n  1)...........3.2.1

m!
i.e., yn  a n (ax  b) m  n
(m  n)!

1
3. nth differential coefficient of
ax  b
1
Let y   ( ax  b ) 1
ax  b
Differentiating y w. r. t. x successively, we get
y1  ( 1)a( ax  b) 2
y 2  (1)(2) a 2 (ax  b) 3
y3  (1)(2)(-3)a 3 ( ax  b) 4
………………………………..
………………………………..
Now we suppose that

y n  (1)(2)(3)...............(n)a n ( ax  b) ( n 1) ........(3)

Differentiating both sides of equation (3) w. r. t. x , we get
y n 1  (1)(2)(3)...............(n)((n  1))a n 1 (ax  b)  ( n  2)
Clearly, y n 1 is of the same form as y n , given by equation (3). This  If equation (3) is true for a
particular value of n, then equation (3) is true for next higher value of n also.
But we have already proved that equation (3) is true for n = 1, 2, 3.
Since equation (3) is true for n = 3, it is true for n = 4 also. Similarly, since equation (3) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (3) is true for every positive integer n.
From equation (3), we can write y n as follows:

yn  (1) n n ! a n (ax  b) ( n1)

i.e., (1) n a n n !
yn 
(ax  b) ( n 1)
 7/23

1
4. nth differential coefficient of
(ax  b) 2
1
Let y   (ax  b) 2
(ax  b) 2

Differentiating both sides w. r. t. x successively, we get

y1  (2)a(ax  b) 3
y 2  (2)(3) a 2 (ax  b) 4
y3  (2)(3)( 4)a 3 (ax  b) 5
………………………………..
………………………………..
Now we suppose that

y n  (2)(3)(4)...............((n  1))a n (ax  b)  ( n 2) ........(4)

Differentiating both sides of equation (4) w. r. t. x , we get
)
yn1  (2)(3)(4)...............((n  1)) ((n  2))a n1 (ax  b)  ( n3)
Clearly, y n 1 is of the same form as y n , given by equation (4). This  If equation (4) is true for a
particular value of n, then equation (4) is true for next higher value of n also.
But we have already proved that equation (4) is true for n = 1, 2, 3.
Since equation (4) is true for n = 3, it is true for n = 4 also. Similarly, since equation (4) is true for n =
4, it is true for n = 5 also and so on.

Therefore, by mathematical induction, equation (4) is true for every positive integer n.
From equation (4), we can write y n as follows:

y n  (1) n n  1! a n (ax  b)  ( n 2)

i.e., (1) n a n n  1!

yn 
(ax  b) ( n  2 )

5. nth differential coefficient of log (ax  b)

Let y  log (ax  b )
Differentiating both sides w. r. t. x successively, we get
a
y1   a ( ax  b ) 1
ax  b
y 2  ( 1) a 2 ( ax  b) 2
y 3  (1)(2)a 3 ( ax  b) 3
………………………………..
………………………………..
Now we suppose that
 8/23

y n  (1)(2)(3)...............((n  1))a n (ax  b)  n ...........(5)

Differentiating both sides of equation (5) w. r. t. x , we get

y n 1  (1)(2)(3)...............(( n  1))(n)a n 1 ( ax  b) ( n 1)

Clearly, y n 1 is of the same form as y n , given by equation (5). This  If equation (5) is true for a
particular value of n, then equation (5) is true for next higher value of n also.
But we have already proved that equation (5) is true for n = 1, 2, 3.
Since equation (5) is true for n = 3, it is true for n = 4 also. Similarly, since equation (5) is true for n =
4, it is true for n = 5 also and so on.

Therefore, by mathematical induction, equation (5) is true for every positive integer n.
From equation (5), we can write y n as follows:

y n  ( 1) n1 ( n  1)! a n (ax  b)  n

i.e., (1) n1 a n (n  1) !

yn 
(ax  b) n

6. nth differential coefficient of e mx

mx
Let y  e
Differentiating both sides w. r. t. x successively, we get

y1  memx
y 2  m 2 e mx
y 3  m 3 e mx
………………………………..
………………………………..
Now we suppose that

y n  m n e mx .........(6)
Differentiating both sides of equation (6) w. r. t. x , we get

y n1  m n (memx ) i.e., y n1  m n1e mx

Clearly, y n 1 is of the same form as y n , given by equation (6). This  If equation (6) is true for a

particular value of n, then equation (6) is true for next higher value of n also.

But we have already proved that equation (6) is true for n = 1, 2, 3.

Since equation (6) is true for n = 3, it is true for n = 4 also. Similarly, since equation (6) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (6) is true for every positive integer n,

i.e., y n  m n e mx
 9/23

7. nth differential coefficient of a mx

mx
Let y  a

Then y  e mx log a  a mx
 e mx log a 
Differentiating both sides w. r. t. x successively, we get


y1  m log a e mx log a
y2  m log a m log a e   (m log a) e 
mx log a 2 mx log a

y3  ( m log a ) m log a e
2
  (m log a) e 
mx log a 3 mx log a

………………………………..
………………………………..
Now we suppose that


y n  (m log a) n e mx log a  .........(6)
Differentiating both sides of equation (6) w. r. t. x , we get


y n 1  (m log a) n m log a e mx log a     (m log a) e
n1 mx log a

Clearly, y n 1 is of the same form as y n , given by equation (6). This  If equation (6) is true for a

particular value of n, then equation (6) is true for next higher value of n also.

But we have already proved that equation (6) is true for n = 1, 2, 3.

Since equation (6) is true for n = 3, it is true for n = 4 also. Similarly, since equation (6) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (6) is true for every positive integer n,


y n  (m log a ) n e mx log a 
i.e., y n  (m log a) n a mx  a mx
 e mx log a 
i.e., y n  m n (log a ) n a mx

8. nth differential coefficient of sin (ax  b)

Let y  sin(ax  b)
Differentiating both sides w. r. t. x successively, we get
 
y1  a cos( ax  b )  a sin  ax  b  
 2
      
y 2  a 2 cos  ax  b    a 2 sin  ax  b     a 2 sin  ax  b  2 
 2  2 2  2
      
y 3  a 3 cos  ax  b  2   a 3 sin  ax  b  2    a 3 sin  ax  b  3 
 2   2 2   2
………………………………..
………………………………..
Now we suppose that
 10/23

 
y n  a n sin  ax  b  n  .......... (7)
 2
Differentiating both sides of equation (7) w. r. t. x , we get
    
y n 1  a n 1 cos  ax  b  n   a n 1 sin  ax  b  n  
 2  2 2
 
i.e., y n1  a n1 sin  ax  b  (n  1) 
 2
Clearly, y n 1 is of the same form as y n , given by equation (7). This  If equation (7) is true for a
particular value of n, then equation (7) is true for next higher value of n also.
But we have already proved that equation (7) is true for n = 1, 2, 3.
Since equation (7) is true for n = 3, it is true for n = 4 also. Similarly, since equation (7) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (7) is true for every positive integer n,

i.e.,  n
y n  a n sin  ax  b  
 2 

9. nth differential coefficient of cos (ax  b)

Let y  cos( ax  b)
Differentiating both sides w. r. t. x successively, we get
 
y1   a sin ( ax  b )  a cos  ax  b  
 2
      
y 2   a 2 sin  ax  b    a 2 cos  ax  b     a 2 cos  ax  b  2 
 2  2 2  2
      
y 3   a 3 sin  ax  b  2   a 3 cos  ax  b  2    a 3 cos  ax  b  3 
 2  2 2  2
………………………………..
………………………………..
Now we suppose that
 
y n  a n cos  ax  b  n  ..........(8)
 2
Differentiating both sides of equation (8) w. r. t. x , we get
    
y n 1   a n 1 sin  ax  b  n   a n 1 cos  ax  b  n  
 2  2 2
 
i.e., y n1  a n1 cos  ax  b  (n  1) 
 2
Clearly, y n 1 is of the same form as y n , given by equation (8). This  If equation (8) is true for a
particular value of n, then equation (8) is true for next higher value of n also.
But we have already proved that equation (8) is true for n = 1, 2, 3.
Since equation (8) is true for n = 3, it is true for n = 4 also. Similarly, since equation (8) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (8) is true for every positive integer n,

i.e.,  n
y n  a n cos  ax  b  
 2 
 11/23

nth differential coefficient of e sin(bx  c)

ax
10.
Let y  e sin(bx  c)
ax

Differentiating both sides w. r. t. x , we get

y1  a e ax sin(bx  c)  be ax cosbx  c 
b
Put a  r cos  and b  r sin  so that r  a 2  b 2 and   tan 1
a
Then we have
y1  r cos  e ax sin(bx  c)  r sin  e ax cosbx  c 
 r e ax sin(bx  c) cos   cosbx  c sin 
 r e ax sin(bx  c  )
Similarly, differentiating y w. r. t. x twice, thrice, ………, we get

y 2  r 2 e ax sin(bx  c  2)
y 3  r 3 e ax sin(bx  c  3)
………………………………..
………………………………..
Now we suppose that
y n  r n e ax sin(bx  c  n) .......(9)
Differentiating both sides of equation (9) w. r. t. x , we get


y n1  r n ae ax sin(bx  c  n)  be ax cos(bx  c  n) 
 r n e ax a sin(bx  c  n)  b cos(bx  c  n)

 r n e ax r cos  sin(bx  c  n)  r sin θ cos(bx  c  n)

(using a  r cos , b  sin  )

 r n 1e ax sin(bx  c  n) cos   cos(bx  c  n) sin 

 r n 1e ax sin(bx  c  n  )

 r n 1e ax sin bx  c  (n  1)

Clearly, y n 1 is of the same form as y n , given by equation (9). This  If equation (9) is true for a
particular value of n, then equation (9) is true for next higher value of n also.
But we have already proved that equation (9) is true for n = 1, 2, 3.
Since equation (9) is true for n = 3, it is true for n = 4 also. Similarly, since equation (9) is true for n =
4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (9) is true for every positive integer n,

i.e., b
y n  r n e ax sin(bx  c  n ), where r  a 2  b 2 and   tan 1
a
 12/23

11. nth differential coefficient of e cos(bx  c)

ax

Let y  e cos( bx  c)
ax

Differentiating both sides w. r. t. x , we get

y1  a e ax cos(bx  c)  be ax sin bx  c 

b
Put a  r cos  and b  r sin  so that r  a 2  b 2 and   tan 1
a
Then we have
y1  r cos  e ax cos(bx  c)  r sin  e ax sinbx  c 
 r e ax cos(bx  c) cos   sin bx  c sin 
 r e ax cos(bx  c  )
Similarly, differentiating y w. r. t. x twice, thrice, ………, we get

y 2  r 2 e ax cos(bx  c  2)
y 3  r 3 e ax cos(bx  c  3)
………………………………..
………………………………..
Now we suppose that

y n  r n e ax cos(bx  c  n) .......(10)

Differentiating both sides of equation (10) w. r. t. x , we get


y n 1  r n ae ax cos(bx  c  n)  be ax sin(bx  c  n) 
 r n e ax a cos(bx  c  n)  b sin(bx  c  n)
 r n e ax r cos  cos(bx  c  n)  r sin θ sin(bx  c  n)
(using a  r cos  , b  sin  )

 r n 1e ax cos(bx  c  n) cos   sin(bx  c  n) sin 

 r n 1e ax cos(bx  c  n  )
 r n 1 e ax cosbx  c  (n  1)
Clearly, y n 1 is of the same form as y n , given by equation (10). This  If equation (10) is true for a
particular value of n, then equation (10) is true for next higher value of n also.
But we have already proved that equation (10) is true for n = 1, 2, 3.
Since equation (10) is true for n = 3, it is true for n = 4 also. Similarly, since equation (10) is true for n
= 4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, equation (10)) is true for every positive integer n,

i.e., b
y n  r n e ax cos(bx  c  n ) , where r  a 2  b 2 and   tan 1
a
 13/23

Example 5 : Find nth differential coefficient of 1

.
(1  x ) 2
1
Solution : Let y 
(1  x) 2
Differentiating both sides w. r. t. x successively, we get

y1  (2)(1  x) 3
y 2  (2)(3) (1  x) 4
y3  (2)(3)(-4)(1  x) 5
………………………………..
………………………………..
Proceeding in this way, we get

y n  (2)(3)(4)............... (n  1)(1  x)  ( n 2)
(1) n ( n  1) !
i.e., yn  .
(1  x) ( n  2)

Example 6 : Find y n if y  sin 2 x cos 2 x.

Solution : We have
1 1 1 1  1  cos 4 x 
y  sin 2 x cos 2 x  ( 2 sin x cos x ) 2  (sin 2 x ) 2  sin 2 2 x   
4 4 4 4 2 
1 1
i.e., y  cos 4 x
8 8
n
This  y n  0  4 n cos  4 x 
1

8  2 
n
(using the formula : yn  a n cos  ax  b   if y  cos(ax  b) )
 2 
 n
i.e., y n  2 n2 cos  4x  .
 2 
Example 7 : Find y n if y  e x sin 2 x.
Solution : We have
 1  cos 2 x 
y  e x sin 2 x  e x  
 2 
1 x 1 x
i.e., y  e  e cos 2 x
2 2
n
This  y n 
1 x 1 2
2

e  (1  2 2 ) 2 e x cos 2 x  n tan 1 2
2

 using the formulae : y n  m n e mx if y  e mx 
 
 n
 b 
 and y n  (a 2  b 2 ) 2 e ax cos  bx  c  n tan 1  if y  e ax cos(bx  c) 
  a 
n
1 x 
e 1  (5) 2 cos  2 x  n tan  1 2   .

i.e., yn 
2   
 14/23

Example 8 : If y  sin mx  cos mx, prove that yn  m 1  (1) sin 2mx n n 2

.

Solution : We have
y  sin mx  cos mx

This  y n  m n sin  mx  n    m n cos  mx  n  

 2   2 
  n 
 using the formulae : y n  a n sin  ax  b   if y  sin( ax  b) 
  2  
  n 
 and y n  a cos  ax  b 
n
 if y  cos( ax  b) 
  2  
  n    n   
i.e., y n  m n sin  mx  
  cos  mx   .
  2   2  
1
    2
2
This  y n  m n sin  mx  n    cos  mx  n    
 

  2  
 2   
 
1
  n    n    n    n    2
 m sin 2  mx 
n 2 
  cos  mx 

  2 sin  mx 

 cos  mx 
2  

  2   2   2  
1
  n    n    2
 m 1  2 sin  mx 
n 
 cos  mx 
2  
  
  2  

1
n
 n    2 (using the formulae 2 sin  cos   sin 2)
 m 1  sin 2 mx 
2  

 

 m 1  sin 2 mx  nπ 
n 2

 m 1  sin 2mx cos n   cos 2mx sin n 

n 2

 m 1  ( 1) sin 2mx 

2
( since cos n   (1) and sin n   0 )
n n n

3. Method of finding nth derivative of an Algebraic Rational Function

In order to find the nth derivative of an algebraic rational function, it is first required to check
whether the denominator of the given algebraic rational function is resolvable into real linear factors
or not.

If the denominator of the given algebraic rational function is resolvable into real linear factors, then
the given algebraic rational function is resolved into partial fractions. In this case, the denominator of
each partial fraction thus obtained consists of real linear factors, repeated or non-repeated,
therefore the nth derivative of the given algebraic rational function can be found out by directly using
the formulae derived earlier for nth derivative of some standard functions.

If the denominator of the given algebraic rational function is not resolvable into real linear factors,
then its nth derivative is found out by the application of De Moivre’s theorem.
 15/23

x4
Example 9 : Find y n if y  .
( x  1)( x  2)
Solution : We have
x4 x4 15x  14
y  2  x2  3x  7  2
( x  1) ( x  2) x  3x  2 x  3x  2

2 15x  14
i.e., y  x  3x  7  ....(1)
( x  1) ( x  2)
15 x  14
Now let us resolve into partial fractions.
( x  1) ( x  2)

15x  14 A B
Let   ....(2)
( x  1) ( x  2) x  1 x  2

15x  14 A( x  2)  B( x  1)
This  
( x  1) ( x  2) ( x  1) ( x  2)
 15 x  14  A ( x  2)  B ( x  1) ....(3)

Putting x  1 on both sides both sides of equation (3), we get A  1 .

Putting x  2 on both sides both sides of equation (3), we get B  16.

Substituting the values of A and B in equation (2) and using equation (1), we get
16 1
y  x 2  3x  7  
x  2 x 1
 ( 1) n n !   ( 1) n n ! 
This  y n  0  0  0  16  n 1 
 n 1 
 ( x  2)   ( x  1) 
 (-1) n a n n ! 1 
 usig the formula : y n  if y  
 (ax  b) n 1
ax  b 
 16 1 
i.e., y n  ( 1) n n !  n 1
 n 1 
.
 ( x  2) ( x  1) 

x2
Example 10 : Find y n if y  .
( x  1)2 ( x  2)
Solution : We have
x2
y ....(1)
( x  1)2 ( x  2)

x2
Let us resolve into partial fractions.
( x  1) 2 ( x  2)

x2 A B C
Let    ....(2)
( x  1) ( x  2) x  1  x  1
2 2
x2
 16/23

x2 A( x  1) ( x  2)  B( x  2)  C ( x  1) 2
This  
( x  1) 2 ( x  2) ( x  1) 2 ( x  2)

 x 2  A ( x  1) ( x  2 )  B ( x  2 )  C ( x  1) 2 ....(3)

Putting x  1 on both sides of equation (3), we get B  1 / 3 .

Putting x   2 on both sides of equation (3), we get C  4 / 9.

Equating the coefficients of x 2 from both sides of equation (3), we get

4 5
1  A  C , which  A  1  C  1   .
9 9
Substituting the values of A, B and C in equation (2) and using equation (1), we get

5  1  1 4  1 
y     
9  x  1  3  x  1 2
9  x  2

 ( 1) n n !  1  ( 1) n n  1! 4  ( 1) n n ! 

This  y n  5  n 1 
  n2 
  n 1 
.
9  ( x  1)  3  ( x  1)  9  ( x  2) 
 (-1) n a n n ! 1 
 using the formulae : yn  n 1
if y  
 (ax  b) ax  b 
 
 (-1) a n  1!
n n
1 
 and yn  if y  
 (ax  b) n  2 ( ax  b) 2 

1
Example 11 : Find y n if y  .
x  a2
2

Solution : We have

1 1 1  ( x  ia)  ( x  ia) 
y  
2
x a 2
( x  ia) ( x  ia) 2ia  ( x  ia) ( x  ia) 

1  1 1 
i.e., y 
2ia  ( x  ia) ( x  ia) 


 ( 1) n n ! ( 1) n n !   (-1) n a n n ! 1 

This  y n  1   
n 1  
usig the formula : y  if y  
ax  b 
n 1 n n 1
2ia  ( x  ia ) ( x  ia )   (ax  b)

(1) n n !  1 1 
i.e., yn   ( x  ia) n1  ( x  ia) n1  .
2ia  
a
Now put x  r cos  and a  r sin  so that r  x 2  a 2 and   tan 1
x
Then we have

(1)n n !  1 1 
yn  
2ia  (r cos   i r sin ) n1 (r cos   i r sin )n1 
 17/23

(1) n n !  1 1 
This  yn   (cos   i sin ) n1  (cos   i sin )n1 
2ia r n1  

(1) n n !  1 1   usig the formula : e i   cos   i sin  

 n1  i  n1  i  n1 
2ia r  (e ) (e )   

( 1) n n !  1 1 
 n1  i ( n1)   i ( n1)  
2ia r  e e 


2ia r n1
e 
( 1) n n ! i ( n1) 
 e i ( n1)  
( 1) n n !
 cos( n  1)   i sin( n  1)   cos( n  1)   i sin( n  1) 
2ia r n1

(1) n n !
 2i sin( n  1) 
2ia r n1

( 1) n n !  a 
 n1 sin( n  1) 
 since a  r sin , i.e., r  
 a   sin  
a 
 sin  

(1) n n !
i.e., yn  sin n1  sin( n  1)  .
a n 2

Example 12 : Find y n if y  tan 1 x.

Solution : We have
y  tan 1 x
Differentiating both sides w. r. t. x , we get
1
y1 
1  x2
1 1  ( x  i)  ( x  i) 
This  y1  
( x  i ) ( x  i ) 2 i  ( x  i ) ( x  i ) 

1  1 1 
y1  
2 i  ( x  i) ( x  i) 

i.e.,

Differentiating y 1 w. r. t. x (n  1) times, we get

1  d n 1  1  d n 1  1 
yn      
2 i  dx n 1  ( x  i )  dx n 1  ( x  i ) 
 18/23

This  y n  1 
 ( 1) n 1
n  1!  ( 1) n 1 n  1!

2i  ( x  i) n ( x  i) n 
 (-1) n a n n ! 1 
 using the formula : y n  if y  
 (ax  b) n 1
ax  b 

(1) n  1 n  1!  1 1 
i.e., yn   ( x  i)n  ( x  i)n  .
2i  
1
Now put x  r cos  and 1  r sin  so that r  x 2  1 and   tan 1
x
Then we have
(1) n  1 n  1!  1 1 
yn   (r cos   i r sin ) n  (r cos   i r sin ) n 
2i  
(1) n  1 n  1!  1 1 
This  yn   (cos   i sin ) n  (cos   i sin )n 
2 i rn  

(1) n  1 n  1!  1 1  i
  (e i  ) n  (ei  ) n  (using e  cos   i sin  )
2i rn  

( 1) n  1 n  1!  1 1 
  e i n   ei n  
2 i rn

( 1) n  1 n  1! i n  i n 

2 i rn
e e 
( 1) n  1 n  1!
 cos n   i sin n  cos n   i sin n 
2i rn
(1) n  1 n  1!
 2i sin n 
2 i rn

(1) n  1 n  1!  1 
 sin n   since 1  r sin , i.e., r  
 1 
n
 sin  
 
 sin  

i.e., yn  ( 1) n  1 n  1!sin n  sin n  .

4. Leibnitz’s Theorem
Leibnitz’s Theorem is used to find the nth derivative of the product of two functions. Its statement
and proof are as follows :
Statement : If u and v are two functions of x , possessing derivatives of the nth order, then
uv n  n
C 0 u n v  n C1 u n1v1  n C 2 u n 2 v 2  .......... n C r u n r v r  ......... nC n u vn .

Proof : Let y  u v , where u and v are functions of x .

Differentiating both sides w. r. t. x , we get
 19/23

y1  u1 v  u v1
Clearly, y1 can be written as

y1  1C0 u1 v  1C1 u v1  C0  1, 1C1  1

1

This  Leibnitz’s Theorem is true for n  1.

Differentiating y1 w. r. t. x , we get

y 2  u 2 v  u1 v1   u1 v1  u v 2   u 2 v  2 u1 v1  u v 2
Clearly, y 2 can be written as

y 2  2C0 u 2 v  2C1 u1 v1  2C2 u v2  2

C0  1, 2C1  2, 2C2  1 
This  Leibnitz’s Theorem is true for n  2.

Differentiating y 2 w. r. t. x , we get

y 3  u 3 v  u 2 v1   2 u 2 v1  u1 v 2   u1 v 2  u v 3   u 3 v  3 u 2 v1  3 u1 v 2  u v 3
Clearly, y 3 can be written as

y3  3C0 u3 v  3C1 u 2 v1  3C2 u 1 v2  3C3 u v3  C0  1, 3C1  3, 3C2  3, 3C3  1

3

This  Leibnitz’s Theorem is true for n  3.

Now we suppose that Leibnitz’s Theorem is true for a particular value of n, say, m, i.e., we suppose
that

y m  m C 0 u m v  m C1 u m 1v1  m C 2 u m  2 v 2  ............ m C r 1 u m  r 1 v r 1  m C r u m  r v r

 ............  m C m 1 u1 v m 1  m C m u v m .....(1)
Differentiating both sides of equation (1) w. r. t. x , we get

y m 1  mC 0 u m 1 v  u m v1   m C1 u m v1  u m 1v 2  m C 2 u m 1v 2  u m  2 v 3 

 .......... m C r 1 u m  r  2 v r 1  u m  r 1 v r  m C r u m  r 1v r  u m  r v r 1 

 .......... .......... .....  mC m 1 u 2 v m 1  u1 v m   m C m u1 v m  uv m 1 

This  y m 1 
m
C 0 u m1v   m

C 0  mC1 u m v1   m

C1  mC 2 u m 1v2  ...................... 

 m

C r 1  mC r u m r 1vr  ........   m

C m1  mC m u1vm  mC m uv m1 .....(2)

m1
But we know that
m
Cr 1  mCr  Cr
Putting r  1, 2, 3,......., m successively in this well-known result, we get
m
C0  mC1 m1C1 , mC1  mC2 m1C2 , ............, mCm1  mCm m1Cm

m1 m1
Also, we know that
m
C0  C0  1 and m
Cm  Cm1  1.
 20/23

Substituting these values in equation (2), we get

m 1
y m 1  C 0 u m 1 v  m 1C1 u m v1  m 1
C 2 u m 1v 2  ................
m 1 m 1 m 1
 C r u m 1 r v r  ......  C m u1 v m  C m 1 uv m 1 .....(3)

Equation (3)  Leibnitz’s theorem is true for n  m  1.

Thus we have proved that if Leibnitz’s Theorem is true for a particular value of n, say, m, then it is
true for next higher value of n, i.e., for m+1 also.
But we have already proved that Leibnitz’s Theorem is true for n = 1, 2, 3.
Since Leibnitz’s Theorem is true for n = 3, it is true for n = 4 also. Similarly, since Leibnitz’s Theorem is
true for n = 4, it is true for n = 5 also and so on.
Therefore, by mathematical induction, Leibnitz’s Theorem is true for every positive integer n, i.e., we
have

uv n  n
C 0 u n v  n C1 u n 1v1  nC 2 u n  2 v 2  .......... n C r u n r v r  ......... nC n u v n

Example 13 : Find y n if y  x 2e a x .
Solution : By Leibnitz’s Theorem, if y  u v , then

yn  unv  nC1 un1v1  nC2 un2v2  .............. nCn u vn ...(1)

We are given that y  x 2e a x .

Therefore, we can take u  e a x , v  x 2 .

Then we have u1  ae ax , u 2  a 2 e ax , u 3  a 3 e ax ,......... .., u n  a n e ax

and v1  2 x, v2  2, v3  0, v2  0,......... .., vn  0.

Substituting these values in equation (1), we get

yn  a neax x 2  nC1 a n1eax (2 x) nC2 a n2eax (2)  0  0  .....

n ( n  1) n  2 ax
 a n e ax x 2  n a n 1e ax ( 2 x )  a e ( 2)
2

 a n2 e ax [a 2 x 2  2n ax  n (n  1)].

Example 14 : Find y n if y  x 2 log x, where n  3 .

Solution : By Leibnitz’s Theorem, if y  u v , then

yn  unv  nC1 un1v1  nC2 un2v2  ............ nCn u vn ...(1)

We are given that y  x 2 log x.

Therefore, we can take u  log x, v  x 2 .

 21/23

Then we have
1 1 ( 1)( 2) ( 1)( 2).........  ( n  1)  ( 1) n1 ( n  1) !
u1  , u 2   2 , u3  ,......... .., u n  
x x x3 xn xn
and v1  2 x, v 2  2, v3  0, v2  0,......... ........., v n  0.
Substituting these values in equation (1), we get

(1) n1 (n  1) ! 2 n (1) n2 (n  2) ! n (1) n3 (n  3) !

yn  x  C1 ( 2 x )  C2 (2)  0  0  .......
xn x n1 x n 2
(1) n3
 n 2
[(1) 2 (n  1) !  nC1 (1)(n  2) !(2) nC2 (n  3) !(2)]
x
(1) n3 n (n  1)
 n 2
[(n  1) !  n(n  2) !(2)  (n  3) !(2)]
x 2
(1)n3
 [(n  1)(n  2)(n  3) !  2n (n  2)(n  3) !n (n  1) (n  3) !]
x n2
(1) n3 (n  3) !
 [(n  1)(n  2)  2n (n  2)  n (n  1) ]
x n 2
(1) n1 (n  3) ! 2
 2 n 2
[(n  3n  2)  (2 n 2  2n)  (n2  n) ]
(1) x
(1) n1 (n  3) !
 (2)
x n 2
2 (1) n1 (n  3) !

x n2
Example 15 : If y  (sin 1 x) 2 , prove that (1  x 2 ) y n 2  ( 2 n  1) xy n1  n 2 y n  0.
Solution : Given that y  (sin 1 x ) 2 .
Differentiating both sides w. r. t. x, we get
 1 

y1  2 sin 1 x   

 1 x
2

This  y1 1  x 2  2 sin 1 x

  2
 y1 1  x 2  4 (sin 1 x ) 2 (on squaring both sides)

 y 1  x 2   4 y  y  (sin 
2
1
1
x ) 2 given

Differentiating both sides w. r. t. x, we get

2 y1 y2 (1  x 2 )  y1  2 x   4 y1
2

This  2 y1[ y 2 (1  x 2 )  xy1 ]  4 y1

 y 2 (1  x 2 )  xy1  2

Differentiating both sides n times w. r. t. x and using Leibnitz’s Theorem, we get

[ y n  2 (1  x 2 )  nC1 y n 1 (  2 x )  nC 2 y n ( 2)]  [ y n 1 x  nC1 y n (1)]  0
 22/23

n n n
(by Leibnitz’s Theorem, (uv) n  u n v  C1 u n1v1  C 2 u n2 v2  .......... Cn u vn )

 n (n  1) 
This   yn2 (1  x 2 )  n yn1 (2 x)  yn (2)   yn1 x  n yn   0
 2 
 y n  2 (1  x 2 )  ( 2 n  1) xy n 1  [ n ( n  1)  n ] y n  0

 (1  x 2 ) yn  2  ( 2 n  1) xy n 1  n 2 y n  0 .

1 1

Example 16 : If y m  y m  2 x, prove that ( x  1) y n  2  ( 2 n  1) xy n 1  ( n  m ) y n  0.
2 2 2

1 1

Solution: Given that y  y m m
 2 x.
Differentiating both sides w. r. t. x, we get

1 y m1  1 y    1  y  m1  1 y  2
1 1
m  m
1  1
ym y m
This  1 y1  1 y1  2
m y m y

 y1  y m  y 
1 1
m
 2my
 
2

 y1
2  y m1 
 1
y m   4m 2 y 2 (on squaring both sides)
 

 1 2
1  1 
 2 1
 m 
 y1  y  y   4 y m y m   4m y
m 2 2

   
 
 
1 1

 y1 2 x   4  4m 2 y 2  y m  y m  2 x given
2 2
 
 
2
 4 y1 ( x  1)  4m y
2 2 2

2
 y1 ( x 2  1)  m 2 y 2
Differentiating both sides w. r. t. x, we get
2 y1 y2 ( x  1)  y1 2 x   m 2 ( 2 yy1 )
2 2

This  2 y1[ y 2 ( x 2  1)  xy1 ]  2 y1 (m 2 y )

 y 2 ( x 2  1)  xy1  m 2 y.

Differentiating both sides n times w. r. t. x and using Leibnitz’s Theorem, we get

[ y n  2 ( x 2  1)  nC1 y n 1 ( 2 x )  nC 2 y n ( 2)]  [ y n 1 x  nC1 y n (1)]  m 2 y n
n n n
(by Leibnitz’s Theorem, (uv) n  u n v  C1 u n1v1  C 2 u n2 v2  .......... Cn u vn )

 n (n  1) 
2
This   yn 2 ( x  1)  n yn1 (2 x)  yn (2)   yn1 x  n yn   m 2 yn
 2 
 23/23

 y n 2 ( x 2  1)  ( 2 n  1) xy n1  [ n ( n  1)  n ] y n  m 2 y n

 ( x 2  1) y n 2  ( 2 n  1) xy n1  ( n 2  m 2 ) y n  0 .

Exercises
1. If y  A sin mx  B cos mx, prove that y 2   m 2 y.

2. If y  sin(sin x), prove that y 2  y1 tan x  y cos 2 x  0 .

x
If y  log 
x  3 2
3.  , prove that x y 2  ( y  x y1 ) .
 a  bx 
d2y h 2  ab
4. If y  ax 2  2hxy  by 2  1, prove that  .
d x 2 (h x  b y ) 3
1
5. Find y n if y  3
.
( x  1) ( x  2)
xn
6. Find y n if y  .
(1  x)

7. Find y n if y  x 2 e 2 x sin x.

2
8. If y  e x , prove that y n 1  2 xy n  2 n y n 1  0.

y n2
9. If y  sin m (sin 1 x), prove that lim  n2  m2.
x 0 yn
1 x
10. If y  e a tan  a 0  a1 x  a 2 x 2  ......... , then prove that (n  2)an2  nan  an1.

Answers
(1) n n ! (1) n n ! (1) n (n  1) ! (1) n (n  2) !
5. y n    
( x  2) n1 ( x  1) n1 ( x  1) n2 2 ( x  1) n3
n!
6. yn 
( x  1) n1
n n1
 1  1
7. yn  5 2 x 2e 2 x sin  x  n tan 1   2 n x 5 2 e 2 x sin  x  (n  1) tan 1 
 2  2
n2
 1 
 n (n  1) 5 2 e 2 x sin  x  (n  2) tan 1 
 2