Saturday Online Class 4 (1st class on Polynomials)

May 30, 2020

What is a polynomial? A polynomial in the variable x is just an expression of the form

a0 + a1 x + a2 x2 + · · · , upto some an xn , where an 6= 0. For example, 2 + 3x, x4 − 101x + 2,
√
4x2 − 2πx + 3 etc. The coefficients can be anything, but the powers of x must be non-
√
negative integers. Thus 5/x, x + x are not polynomials. If you consider the expression
√
5xy + x2 y, then it will be a polynomial in x, but not a polynomial in y.
Consider the polynomial a0 +a1 x+a2 x2 +· · ·+an xn . Note that a0 , a1 , . . . , an are constants
(independent of the variable x), they are called coefficients, and the condition an 6= 0 imposed
to ensure that there is no ambiguity about what is the largest power of x in the polynomial.
The term an is called the leading coefficient and n is called the degree of this polynomial.
When n = 2, the polynomial a0 + a1 x + a2 x2 is called a quadratic polynomial. When n = 3,
it is called a cubic polynomial. Here a0 , a1 , a2 are just a notation for the coefficients, you can
also denote them as ax2 + bx + c or px2 + qx + r whatever.
We write b(x) = x2 + 3x + 4 to indicate that x2 + 3x + 4 is a polynomial, as a function
of x. Now what does that mean? It means that, when you plug in any value of x, b(x)
returns the value of the expression x2 + 3x + 4 for that particular value of x. For example,
b(1) = 12 + 3 × 1 + 4 = 8, b(1/2) = (1/2)2 + 3(1/2) + 4. Obviously, here b can be replaced
by any letter f, g etc.
Synthetic division: Suppose I give you two polynomials, say a(x) = 2x + 3, and b(x) =
x2 + 3x + 4. Suppose a, b are integers. When I say divide b by a, what do you do? You express
b as aq + r where q and r are integers, such that 0 ≤ r < |a|. Similarly, for polynomials, when
I say divide b(x) = x2 + 3x + 4 by a(x) = 2x + 3, I mean that express b(x) = a(x)q(x) + r(x)
where q(x) and r(x) are polynomials and 0 ≤ deg r(x) < deg a(x). Look up some examples
of synthetic division (we shall return to them in the next class).
Let us now learn some facts about the quadratic polynomial: p(x) = ax2 + bx + c where
a 6= 0. The roots of the quadratic equation ax2 + bx + c = 0 are called zeros of the polynomial
p(x) = ax2 + bx + c and we know that they are given by the quadratic formula
√
−b ± b2 − 4ac
.
2a

D = b2 − 4ac is called the discriminant of the quadratic equation. It determines the nature
of the roots of the quadratic equation, e.g. the roots are real and distinct when D > 0, real
and equal if D = 0, and non-real complex numbers√when D < 0. √
2 2
Let us call the roots as α and β. Say, α = −b+ 2ab −4ac and β = −b− 2ab −4ac . We observe
that α + β = −b/a and αβ = c/a. (Check!)
√ √
Example: The roots of x2 − 4x + 1 = 0 are 2 + 3 and 2 − 3. Note that the sum of
the roots is 4 and the product of the roots is 1.
This idea of relating sum and product of the roots to the coefficients is very important.

1
 A result to remember: If α and β are the zeros of the polynomial ax2 + bx + c then
α + β = −b/a and αβ = c/a.

Here I emphasize on the fact that a, b, c are just names of the coefficients, you should keep
in mind their roles (a is the leading coefficient, b is the coefficient of x and c is the constant
term) while applying the above fact.

Problem 1. If a, b, c are all odd integers, show that the equation ax2 + bx + c = 0
cannot have any rational root.

Solution: We will show that b2 − 4ac is not a perfect square. Recall that if n is any odd
integer, then n2 must be of the form 8k + 1. Here a, b, c are odd. So b2 = 8l + 1 for some l.
Also, let ac = 2y + 1. b2 − 4ac = 8l + 1 − 4(2y + 1) = 8 × something − 3. But if b2 − 4ac
were a square, it must have the form 8k + 1. Therefore, D = b2 − 4ac is not a perfect square.
√
not a perfect square, we can conclude that D is not rational,
Since D is an integer which is √
−b ± D
implying that the roots are not rational.
2a

Problem 2. Suppose that a, b are integers and b 6= −1. Show that if x2 + ax + b + 1 = 0

has an integer root then a2 + b2 must be composite.

Solution: Let the roots be α and β. Then, α + β = −a and αβ = b + 1. Since sum of the
roots is an integer, so if one of the roots is an integer then the other one must also be an
integer. It is given that (at least) one of them is integer. Hence we may assume that both α
and β are integers. Next, observe that

a2 + b2 = (α + β)2 + (αβ − 1)2 = α2 + β 2 + α2 β 2 + 1 = (α2 + 1)(β 2 + 1).

We are asked to show that a2 + b2 is composite and we derived a factorization of it. Are
we done? No, because we have to show that α2 + 1 and β 2 + 1 are both greater than 1. Since
α and β are integers, α2 + 1 and β 2 + 1 are already ≥ 1. So the only way in which one of
them can be equal to 1 is when α or β is 0. But αβ = b + 1 6= 0. Therefore, each of these
factors is > 1 and we are through.

Problem 3. Suppose that k is a real number such that the equations x2 − 4x + k = 0

and x2 + kx − 4 = 0 have exactly one root in common. Find the value of k.

Solution: Suppose that the common root is r. Since r satisfies both the equation, we have
r2 − 4r + k = 0 and r2 + kr − 4 = 0. Subtract the 2nd equation from the 1st equation (to
get rid of the term r2 ) and get −(4 + k)r + k + 4 = 0 =⇒ (k + 4)(r − 1) = 0. This implies
that either k = −4 or r = 1.
If k = −4 then the two equations are identical, which is not allowed. Therefore, we must
have r = 1. Putting r = 1 in any of the two equations, we get k = 3.

2
 Problem 4. Solve the equation 6x4 − 25x3 + 12x2 + 25x + 6 = 0.

Solution: Observing the symmetry in the equation, it would be a good idea to divide the
equation by x2 . Since x = 0 does not satisfy the equation, we may divide by x2 without losing
any root. Performing that, we get

25 6
6x2 − 25x + 12 + + 2 =0
   x x
1 1
or, 6 x2 + 2 − 25 x − + 12 = 0.
x x
1
Note that x2 + x2
= (x − x1 )2 + 2. Hence the above equation is equivalent to
 2  
1 1
6 x− + 12 − 25 x − + 12 = 0.
x x

Now if we put t = (x − x1 ) then the last equation will be a quadratic equation in t, which you
can solve (do it!). Indeed, it becomes 6t2 − 25t + 24 = 0, which has the solutions t = 23 , 83 .
Going back to x, we have two cases to solve: (i) x − x1 = 32 , and (ii) x − x1 = 38 .
Solving these, we get x = 2, − 12 , 3 and − 31 . (Answer)

A few problems as homework:

1. Prove that the equation (x − a)(x − b) + (x − b)(x − c) + (x − c)(x − a) = 0 has real

roots for any three real numbers a, b, c. Also show that the two roots are equal if and only
if a = b = c.
2. Solve the equation (for real number x): (x2 + x − 2)3 + (2x2 − x − 1)3 = 27(x2 − 1)3 .
3. Suppose that the roots of x2 + αx − a = 0 are β, γ, and the roots of x2 + βx − b = 0 are
γ, α. If 2(a + b + c) = α2 + β 2 + γ 2 , then show that the roots of x2 + γx − c = 0 are α, β.
4. Let f (x) = x3 − 3x + b and g(x) = x2 + bx − 3. Find all possible values of the real number
b such that the equations f (x) = 0 and g(x) = 0 have a common root.
5. If x = 2 is a root of the equation mx3 − 12x2 + 10x − m = 0 then find the other roots.
6. Let f (x) be a quadratic polynomial. Prove that there exists quadratic polynomials g(x), h(x)
such that f (x)f (x + 1) = g(h(x)).

Before the next class, please look up some examples of Synthetic division, and read about
Remainder theorem and Factor theorem from any book.

3
 Saturday Online Class 5 (2nd class on Polynomials)
June 06, 2020

Division algorithm

First we shall discuss the division algorithm for polynomials, then the remainder theorem and
the factor theorem. Suppose I give you two polynomials, say a(x) = x − 3, and b(x) =
2x3 + 4x2 + 5x − 1. What do we mean when we say divide b(x) by a(x)?
Suppose a, b are integers. When I say divide b by a, what do you do? You express b as
aq + r where q and r are integers, such that 0 ≤ r < |a|. Similarly, for polynomials, when I say
divide b(x) = 2x3 +4x2 +5x−1. by a(x) = x−3, I mean that express b(x) = a(x)q(x)+r(x)
where q(x) and r(x) are polynomials and 0 ≤ deg r(x) < deg a(x). Following figure shows
how to carry out this long division.

Division algorithm for integers: Given any two integers a, b where a 6= 0, there exist unique
integers q and r such that b = aq + r where 0 ≤ r < |a|.
Division algorithm for polynomials: Given any two polynomials a(x), b(x) where a(x)
is not the zero polynomial, there exist unique polynomials q(x) and r(x) such that b(x) =
a(x)q(x) + r(x) where 0 ≤ deg r(x) < deg a(x).
Remarks.
(i) In context of the above, we assume that degree of the zero polynomial is zero.
(ii) If b(x) = a(x)q(x) + r(x) where r(x) is the remainder, then deg b(x) = deg a(x) +
deg q(x).

Remainder theorem and factor theorem

What can you say about the remainder if a(x) is a linear polynomial (i.e. if deg a(x) = 1)?
The above theorem forces that deg r(x) must be zero, implying that r(x) must be a constant
polynomial.

4
 What will be the remainder if you divide a polynomial p(x) by x−a? The above discussion
tells us that the remainder will be a constant, say r. So we can write p(x) = (x−a)q(x)+r for
some polynomial q(x) and some constant r. Putting x = a, we get p(a) = 0 × q(a) + r =⇒
r = p(a).

Remainder Theorem. When we divide a polynomial p(x) by x − a, the remainder we

get is p(a).

Example: Find the remainder when the polynomial 2x3 + 4x2 + 5x − 1 is divided by x − 3.
Answer: Using the above theorem, the remainder should be the value of p(x) = 2x3 + 4x2 +
5x − 1 at x = 3, i.e. p(3) = 2 × 33 + 4 × 32 + 5 × 3 − 1 = 104.
Recall, we say that ‘x = a is a zero of the polynomial p(x)’ or ‘a root of the equation
p(x) = 0’ if the value of the polynomial at x = a is zero, i.e. if p(a) = 0. For example, the
zeros of the polynomial x2 − 4 are ±2.
According to the remainder theorem, we can write any polynomial p(x) as p(x) = (x −
a)q(x) + p(a). Therefore, if p(a) = 0 then (x − a) will be a factor of p(x), i.e. p(x) =
(x − a)q(x) for some polynomial q(x).

Factor Theorem. If p(a) = 0 then (x − a) will be a factor of p(x), i.e. p(x) =

(x − a)q(x) for some polynomial q(x).

Next let us solve a few problems involving the concepts above.

Problem 1. If x = 2 is a root of the equation mx3 − 12x2 + 10x − m = 0 then find

the other roots.

Since x = 2 is a root, we can put x = 2 in this equation to get the value of m.

(m × 23 − 12 × 22 + 10 × 2 − m = 0 =⇒ m = 4.) Therefore, our polynomial of interest is
p(x) = 4x3 − 12x2 + 10x − 4.
Applying the factor theorem, we know that x − 2 will be a factor of p(x), i.e. p(x) =
(x − 2)q(x) for some polynomial q(x). We have to find this q(x) using long division/vanishing
method. Performing that, we get 4x3 − 12x2 + 10x − 4 = (x − 2)(4x2 − 4x + 2).
Now the other roots of p(x) = 0 can be found from the equation 4x2 −4x+2 = 0. Using the
√
2 √
quadratic formula, x = −b± 2ab −4ac we find that the other roots are (1 ± −1)/2 = (1 ± i)/2.

Problem 2. Suppose that f (x) is a polynomial that leaves a remainder of 1 when

divided by x − 1 and leaves a remainder of 4 when divided by x + 2. What is the
remainder when f (x) is divided by (x − 1)(x + 2)?

In view of the remainder theorem, we can say that the remainder obtained when f (x) is
divided x − 1 is f (1) and the remainder obtained when f (x) is divided by x + 2 is f (−2).
Hence we must have f (1) = 1 and f (−2) = 4 (according to the question).
Now we can assume that the remainder obtained when f (x) is divided by (x − 1)(x + 2)
is ax + b for some constants a and b. Then, f (x) = (x − 1)(x + 2)q(x) + (ax + b), for every

5
x. Putting x = 1 and x = −2, we get a pair of simultaneous equations for a and b, which we
can solve! Details are provided below.
Putting x = 1, we get f (1) = a × 1 + b, i.e. a + b = 1. Putting x = −2, we get
f (−2) = a × (−2) + b, i.e. −2a + b = 4. Solving these two equations, viz. a + b = 1 and
−2a + b = 4, we get a = −1, b = 2. Therefore, the required remainder, ax + b, is −x + 2.
Wrong approach: We start with assuming that the remainder is r, so that f (x) = (x −
1)(x + 2)q(x) + r. Now putting x = 1 you get r = f (1) = 1 and if you put x = −2, you get
r = f (−2) = 4. This is absurd! The fallacy here lies in the assumption that the remainder
must be a constant, r.
Moral: If we are dividing by a quadratic, we must start with the assumption that the
remainder is of the form ax + b. Note that we are not ruling out the possibility that the
remainder can be a constant – we are allowing a to be 0 here.

Problem 3. Suppose that p(x) is a polynomial with integer coefficients such that
p(a) = p(b) = p(c) = p(d) = 3. Where a, b, c, d are distinct integers. Show that it is
not possible to have p(n) = 5, for any integer n.

Consider the polynomial f (x) = p(x) − 3. Since a, b, c, d are zeros of this polynomial,
we can say that (x − a), (x − b), (x − c), (x − d) are factors of this polynomial f (x). Since
a, b, c, d are distinct, we can say that (x − a)(x − b)(x − c)(x − d) is a factor of f (x). So we
can assume that f (x) = (x − a)(x − b)(x − c)(x − d)q(x) for some polynomial q(x).
Let, if possible, p(n) = 5 for some integer n. Then, f (n) = 5 − 3 = 2. On the other hand,
f (n) = (n − a)(n − b)(n − c)(n − d)q(n). Combining these, we get

(n − a)(n − b)(n − c)(n − d)q(n) = 2.

Does this give any contradiction? If we can show that q(n) is an integer, then we get a
contradiction, because 2 can be written as the product of at most 3 distinct integers, while
there are at least 4 distinct integers in the LHS above (since a, b, c, d are distinct).
Now we shall argue why q(x) must be a polynomial with integer coefficients. Recall that
q(x) is the quotient when f (x) is divided by its factor (x − a)(x − b)(x − c)(x − d). Since
the leading coefficient of this divisor polynomial is 1 and since f (x) = p(x) − 3 has integer
coefficients, so it follows from our algorithm of long division that the quotient must have
integer coefficients.
Moral: If we divide a polynomial b(x) with integer coefficients by another polynomial
a(x) whose leading coefficient is 1, then the quotient ought to be a polynomial with integer
coefficients. Polynomials whose leading coefficient is 1 are called monic polynomials.
Lets end today’s class by discussing two more problems from the last homework.

1. Prove that the equation (x − a)(x − b) + (x − b)(x − c) + (x − c)(x − a) = 0 has real

roots for any three real numbers a, b, c. Also show that the two roots are equal if and only
if a = b = c.

6
 Solution: First simplify the equation to 3x2 − 2(a + b + c)x + (ab + bc + ca) = 0. Now look
at its discriminant, which turns out to be D = B 2 −4AC = 4(a2 +b2 +c2 −ab−bc−ca) =
2((a − b)2 + (b − c)2 + (c − a)2 ). The conclusion follows from here.
2. Solve the equation (for real number x): (x2 + x − 2)3 + (2x2 − x − 1)3 = 27(x2 − 1)3 .
Solution: Factorize! x2 + x − 2 = (x − 1)(x + 2) and 2x2 − x − 1 = (x − 1)(2x + 1).
Thus, the given equation is equivalent to

(x − 1)3 (x + 2)3 + (x − 1)3 (2x + 1)3 = (x − 1)3 · 27(x + 1)3 .

Clearly, x = 1 is a root of the above equation (with multiplicity 3). The other roots can
be found from the following equation

(x + 2)3 + (2x + 1)3 = (3x + 3)3 .

Take a = x + 2, b = 2x + 1, then the above reads a3 + b3 = (a + b)3 , which implies

3ab(a + b) = 0, i.e. 3(x + 2)(2x + 1)(3x + 3) = 0 ⇐⇒ x = −2, −1/2, −1.
Therefore, the roots of the given equation are 1, 1, 1, −2, −1, −1/2.

Following problems shall be discussed in the next class.

1. Suppose that the roots of x2 + αx − a = 0 are β, γ, and the roots of x2 + βx − b = 0 are

γ, α. If 2(a + b + c) = α2 + β 2 + γ 2 , then show that the roots of x2 + γx − c = 0 are α, β.
2. Let f (x) = x3 − 3x + b and g(x) = x2 + bx − 3. Find all possible values of the real number
b such that the equations f (x) = 0 and g(x) = 0 have a common root.
3. Let f (x) be a quadratic polynomial. Prove that there exists quadratic polynomials g(x), h(x)
such that f (x)f (x + 1) = g(h(x)).
Hint. Instead of starting with f (x) = ax2 + bx + c, start with f (x) = a(x − r)(x − s)
where r, s are the two zeros of f (x).
What does g(h(x)) mean? Suppose g(x) = x2 − x + 2, and h(x) = 2x + 1. Then
g(h(x)) = h(x)2 − h(x) + 2 = (2x + 1)2 − (2x + 1) + 2. Another way of thinking:
g(h(x)) = g(2x + 1) = (2x + 1)2 − (2x + 1) + 2.
4. Consider the polynomial P (x) = x4 − 3x3 + mx2 + nx + 2 Suppose that when P (x) is
divided by x − 2, the remainder is 6 and when P (x) is divided by x − 1, the remainder is
4. Determine m and n.
5. In an attempt to discover a formula for the Fibonacci numbers, Alex finds a cubic poly-
nomial h(x) such that h(1) = 1, h(2) = 1, h(3) = 2 and h(4) = 3. What is the value of
h(5)?

7
 Saturday Online Class 6 (3rd class on Polynomials)
June 13, 2020

1. Suppose that the roots of x2 + αx − a = 0 are β, γ, and the roots of x2 + βx − b = 0 are

γ, α. If 2(a + b + c) = α2 + β 2 + γ 2 , then show that the roots of x2 + γx − c = 0 are α, β.
Solution: Since β and γ are the roots of the first equation, we have β + γ = −α and
βγ = −a. Similarly, from the second equation we have γ + α = −β and γα = −b. Thus,
we have α + β + γ = 0 and βγ = −a and γα = −b.
(α + β + γ)2 = 0 =⇒ (α2 + β 2 + γ 2 ) + 2(αβ + βγ + γα) = 0. From this we obtain
2(a + b + c) + 2(αβ − a − b) = 0 =⇒ αβ = −c.
Since we have αβ = −c and α + β = −γ, we can say that α and β are the roots of the
equation x2 + γx − c = 0.
2. Let f (x) = x3 − 3x + b and g(x) = x2 + bx − 3. Find all possible values of the real number
b such that the equations f (x) = 0 and g(x) = 0 have a common root.
Solution: Let the common root be r. Then r3 − 3r + b = 0 and r2 + br − 3 = 0.
Multiplying the second equation by r and subtracting it from the first equation, we get
(r3 − 3r + b) − r(r2 + br − 3) = 0 =⇒ b(1 − r2 ) = 0. Hence either b = 0 or r = ±1.
Putting r = ±1 in one of these polynomials, say g, and setting it to be 0, we get two more
possible values of b, namely b = ±2.
Therefore, the all possible values of b are 0, 2 and −2.
3. Let f (x) be a quadratic polynomial. Prove that there exists quadratic polynomials g(x), h(x)
such that f (x)f (x + 1) = g(h(x)).
What does g(h(x)) mean? Suppose that g(x) = x2 + x + 2 and h(x) = x2 + 10x + 99.
Then, g(h(x)) = g(x2 + 10x + 99) = (x2 + 10x + 99)2 + (x2 + 10x + 99) + 2. Another
way of thinking: g(h(x)) = h(x)2 + h(x) + 2 = (x2 + 10x + 99)2 + (x2 + 10x + 99) + 2.
Key idea for this problem: Instead of starting with f (x) = ax2 + bx + c, start with
f (x) = a(x − r)(x − s) where r, s are the two zeros of f (x).
Solution: Let us write f (x) in the form a(x − r)(x − s) where r, s are the zeros of f (x).
Then what is f (x)f (x + 1)? We observe that

f (x)f (x + 1) = a(x − r)(x − s) · a(x + 1 − r)(x + 1 − s)

= a(x2 − (r + s − 1)x + rs − r)a(x2 − (r + s − 1)x + rs − s)
= a2 (x2 − (r + s − 1)x + rs−r)(x2 − (r + s − 1)x + rs − s)
= a2 (h(x) − r)(h(x) − s) where h(x) = x2 − (r + s − 1)x + rs
= g(h(x)) where g(y) = a2 (y − r)(y − s).

8
4. Consider the polynomial P (x) = x4 − 3x3 + mx2 + nx + 2 Suppose that when P (x) is
divided by x − 2, the remainder is 6 and when P (x) is divided by x − 1, the remainder is
4. Determine m and n.
Since many of you could do this problem (also, it is similar to one of the problems discussed
in the last class) I am not including its solution here. Answer: m = n = 2.
5. In an attempt to discover a formula for the Fibonacci numbers, Alex finds a cubic poly-
nomial h(x) such that h(1) = 1, h(2) = 1, h(3) = 2 and h(4) = 3. What is the value of
h(5)?
Fibonacci numbers: 1, 1, 2, 3, 5, 8, · · · (any term = sum of last two terms)
Solution: Note that x = 2, 3, 4 satisfies the equation h(x) = x − 1. So, if we consider
the polynomial h(x) − (x − 1) then we can say that x = 2, 3, 4 are the roots of this new
polynomial h(x) − (x − 1). Since h(x) − (x − 1) is a cubic polynomial, we can say that
x = 2, 3, 4 are its only roots. Therefore

h(x) − (x − 1) = k(x − 2)(x − 3)(x − 4) for some constant k.

It only remains to find the value of k. We put x = 1 and use h(1) = 1 to find the value
of k. We find that k = −1/6. Now find the value of h(5) yourself.

Vieta’s Formulae

Earlier we saw how to relate the sum and product of the two zeros of a quadratic polynomial
with its coefficients. Next we shall see the same for cubic equations.
Suppose that r, s, t are the zeros of the cubic polynomial ax3 + bx2 + cx + d. In light of
the factor theorem, we can say that this polynomial must be same as a(x − r)(x − s)(x − t).
Note, this last polynomial can be expanded as a(x3 − (r + s + t)x2 + (rs + st + tr)x − rst.
Since this polynomial must be identical to the polynomial ax3 +bx2 +cx+d, we can compare
the coefficients in each of these polynomials to conclude that

r + s + t = −b/a, rs + st + tr = c/a, and rst = −d/a.

Similar calculation holds for polynomials of higher degree as well. If α, β, γ, δ are the roots of
the equation ax4 + bx3 + cx2 + dx + e = 0 then you can show that
b d
α + β + γ + δ = − , αβγ + βγδ + γδα + δαβ = − ,
a a
c e
αβ + αγ + · · · + γδ = , αβγδ = .
a a
These formulae, that connects the sum of roots, product of roots, sum of the roots taken two
at a time etc. with the coefficients, are known as Vieta’s formulae.

9
The general form of Vieta’s formulae, for an n-th degree polynomial, is given below.

coefficient of xn−k
The sum of the roots taken k at a time is given by (−1)k · , where
leading coefficient
n is the degree of the polynomial.

We shall end today’s class with the following example problem.

Example. Suppose that 1, 2, 3 are the roots of x4 + ax2 + bx + c = 0. Find the value c.
Solution. Let r be the fourth root. Since the coefficient of x3 is 0, so the sum of the roots
must be 0. Hence 1 + 2 + 3 + r = 0 =⇒ r = −6. Again, by Vieta’s theorem, the product
of the roots must be (−1)4 · 1c = c. Therefore c = 1 × 2 × 3 × (−6) = −36.

I am giving the following problems as homework. We shall discuss them in the next class.

b2 − 4ac
1. Given that α and β are the roots of the quadratic ax2 + bx + c = 0, express in
a2
terms of α and β.
2. Find the sum of the reciprocals of the roots of the equation 2x4 − 3x3 + 2x2 − 4x + 6 = 0.
3. Let p, q, r be the roots of the polynomial 5x3 − 11x2 + 7x + 35. Evaluate p3 + q 3 + r3 .
4. Suppose that the polynomial f (x) = x3 − 3x2 − ax + 15 has three roots, two of which
sum to 8. What is the value of a?
5. Suppose a ≤ b ≤ c and a + b + c = 9, ab + bc + ca = 26 and abc = 24. Find the value of
100a + 10b + c.

Answers to the homework problems: (α − β)2 , 2/3, −2449/125, 37, 234.

10
 Saturday Online Class 7 (4th class on Polynomials)
June 20, 2020
Let us begin by recalling Vieta’s Theorem:

coefficient of xn−k
The sum of the roots taken k at a time is given by (−1)k · , where
leading coefficient
n is the degree of the polynomial.

First let us solve the problems given as homework in the last class:
b2 − 4ac
1. Given that α and β are the roots of the quadratic ax2 + bx + c = 0, express in
a2
terms of α and β.
Solution: α + β = −b/a and αβ = c/a. Observe that,
 2
b2 − 4ac b c
= − −4 = (α + β)2 − 4αβ = (α − β)2 .
a2 a a

2. Find the sum of the reciprocals of the roots of the equation 2x4 − 3x3 + 2x2 − 4x + 6 = 0.
Solution: Suppose the roots of this equation are p, q, r, s. We want to find the value of

1 1 1 1 pqr + qrs + rsp + spq

+ + + = .
p q r s pqrs

Now, by Vieta’s theorem,

   
3 −4 4 6
pqr + qrs + rsp + spq = (−1) = 2 and pqrs = (−1) = 3.
2 2

Therefore, the value of the required quantity is 2/3.

3. Let p, q, r be the roots of the polynomial 5x3 − 11x2 + 7x + 35. Evaluate p3 + q 3 + r3 .
Solution: Since p, q, r are the roots, so by Vieta’s theorem we can say that

(−11) 7 35
p+q+r =− , pq + qr + rp = + , and pqr = − = −7.
5 5 5

We know that (p + q + r)3 = p3 + q 3 + r3 + 3(p + q + r)(pq + qr + rp) − 3pqr. Using it

we can find that p3 + q 3 + r3 = −2449/125.
4. Suppose that the polynomial f (x) = x3 − 3x2 − ax + 15 has three roots, two of which
sum to 8. What is the value of a?
Solution: Suppose that the roots are p, q, r, such that p + q = 8. By Vieta’s theorem,
p + q + r = 3 =⇒ r = 3 − 8 = −5. Also, pqr = −15 =⇒ pq = 3 and pq + qr + rp =
−a =⇒ −a = pq + r(p + q) = 3 + (−5) × 8 =⇒ a = 37.

11
5. Suppose a ≤ b ≤ c and a + b + c = 9, ab + bc + ca = 26 and abc = 24. Find the value of
100a + 10b + c.
Solution: The values of a + b + c, ab + bc + ca, and abc are given. So, in light of Vieta’s
theorem, a, b, c are the roots of the equation

x3 − (a + b + c)x2 + (ab + bc + ca)x − abc = 0

or, x3 − 9x2 + 26x − 24 = 0
or, (x − 2)(x − 3)(x − 4) = 0.

So the roots of the above equation are 2, 3, 4. Therefore, {a, b, c} = {2, 3, 4}. Since it is
given that a ≤ b ≤ c, hence a = 2, b = 3, c = 4. Ans: 234.

Problem. A polynomial P (x) with degree n satisfies P (k) = k/(k + 1) for k = 0, 1, . . . , n.

Find P (n + 1). (Funnier version: If 1 = 21 , 2 = 23 , 3 = 34 , then 4 =?)
Solution: We are given that P (0) = 01 , P (1) = 21 , P (2) = 23 , . . . , P (n) = n+1
n
. Will it be
x
advantageous to consider P (x) − x+1 ? Since it won’t be a polynomial, we won’t be able
to apply factor theorem on it. So instead of that, we consider (x + 1)P (x) − x. Since
k
P (k) = k+1 =⇒ (k + 1)P (k) = k for k = 0, 1, 2, . . . , n, we can say that x = 0, 1, . . . , n
are the roots of Q(x) = (x + 1)P (x) − x. Since Q(x) is a polynomial of degree n + 1, and
since x = 0, 1, . . . , n are its (n + 1) roots, we can say that they are all the roots of Q(x).
Hence we can apply factor theorem to say that Q(x) = c(x − 0)(x − 1) · · · (x − n) for some
constant c. Going back to P (x), we have shown that

(x + 1)P (x) − x = cx(x − 1)(x − 2) · · · (x − n) for some constant c. (∗)

How to find the value of c? Just put x = −1 in the above identity. We find that 1 =
c(−1)(−2) · · · (−1−n) = c(−1)n+1 ×1×2×· · ·×(n+1) =⇒ c = (−1)n+1 /(n+1)!. Now we
can put x = n+1 in the identity (∗) we get (n+2)P (n+1)−(n+1) = c×(n+1)! = (−1)n+1 .
This gives P (n + 1) = (n + 1 + (−1)n+1 )/(n + 2).
(Note, if n is odd, it gives P (n + 1) = 1 and if n is even then P (n + 1) = n/(n + 2).)

Fact. If P (x) is a polynomial with integer coefficients, then a − b | P (a) − P (b) for
any integers a and b.

Proof: Suppose that P (x) = c0 + c1 x + c2 x2 + · · · + cn xn . Observe that P (a) − P (b) =

c1 (a − b) + c2 (a2 − b2 ) + c3 (a3 − b3 ) + · · · + cn (an − bn ). We know that a − b | ak − bk for
every k ≥ 1. [Why? a = b (mod a − b) =⇒ ak = bk (mod a − b) =⇒ a − b | ak − bk .]
Since c1 , c2 , . . . , cn are integers, P (a) − P (b) = c1 (a − b) + c2 (a2 − b2 ) + · · · + cn (an − bn )
will be a multiple of a − b.

12
 The fact a − b | P (a) − P (b) is really important for problems involving polynomials with
integer coefficients. We shall end today’s class with the following problem.
Problem. Given a polynomial f (x) with integer coefficients whose value is divisible by 3 for
three integers k, k + 1, and k + 2, prove that f (m) is divisible by 3 for all integers m.
Solution: It is given that f (k), f (k + 1), f (k + 2) are multiples of 3. Fix any integer m. We
shall show that 3 | f (m). We know that m − b | f (m) − f (b) for any integer b. Now if we
pick b among k, k + 1, k + 2, then we would have 3 | f (b) and m − b | f (m) − f (b). Which
b should we pick? We shall pick that b among k, k + 1, k + 2, for which 3 | m − b. Because
then we would get 3 | f (m) − f (b), and 3 | f (b), implying that 3 | f (m).
Since m − k, m − (k + 1), m − (k + 2) are three consecutive integers, one of them must
be a multiple of 3. We shall use that particular m − b. (Thus, the choice of b depends on m.)
For that particular b, we have 3 | m − b and m − b | f (m) − f (b), hence 3 | f (m) − f (b).
And since b is one of the numbers k, k + 1, k + 2, we also have 3 | f (b). Hence 3 | f (m).

I am giving the following problems as homework. We shall discuss them in the next class.

1. Show that for all real numbers x, y, z satisfying x + y + z = 0 and xy + yz + zx = −3,

the value of the expression x3 y + y 3 z + z 3 x is a constant.
Explanation: The value of the three variables x, y, z cannot be uniquely determined from
two given conditions. You need to show that for any possible choice of x, y, z (satisfying
the two given conditions), the value of x3 y + y 3 z + z 3 x will remain a constant.
2. Suppose that P (x) is a polynomial with integer coefficients such that P (0) and P (1) are
both odd. Show that the equation P (x) = 0 cannot have an integer root.
3. Define the Fibonacci numbers as: F1 = F2 = 1 and Fn+2 = Fn+1 + Fn for all n ≥ 1.
(In simple words, any term = sum of previous two terms. Thus, the first few terms are:
1, 1, 2, 3, 5, 8, · · · .) Consider the polynomial P (x) = F1 x + F2 x2 + F3 x3 + · · · + F2020 x2020 .
Find the number of integer zeros of P (x).
4. Suppose P (x) is a polynomial with P (2) = 2017 and P (5) = 2002. If it is given that
P (x) = 0 has exactly one integer root, find that root.

13
 Saturday Online Class 8 (5th class on Polynomials)
June 27, 2020
1. Show that for all real numbers x, y, z satisfying x + y + z = 0 and xy + yz + zx = −3,
the value of the expression x3 y + y 3 z + z 3 x is a constant.
Explanation: The value of the three variables x, y, z cannot be uniquely determined from
two given conditions. You need to show that for any possible choice of x, y, z (satisfying
the two given conditions), the value of x3 y + y 3 z + z 3 x will remain a constant.
Solution: Consider a cubic equation whose roots are x, y, z. By vieta’s theorem, such a
cubic equation is
t3 − (x + y + z)t2 + (xy + yz + zx)t − xyz = 0.

According to the question, x + y + z = 0 and xy + yz + zx = −3. Therefore, the cubic

equation with roots x, y, z is given by t3 − 3t − xyz = 0. Since x, y, z are the roots of this
equation, we have
x3 − 3x − xyz = 0 =⇒ x3 y = 3xy + xyz · y,
y 3 − 3y − xyz = 0 =⇒ y 3 z = 3yz + xyz · z,
z 3 − 3z − xyz = 0 =⇒ z 3 x = 3zx + xyz · x.

Summing up these, we get x3 y + y 3 z + z 3 x = 3(xy + yz + zx) + xyz(x + y + z) =

3 × (−3) + 0 = −9, which is a constant.
2. Suppose that P (x) is a polynomial with integer coefficients such that P (0) and P (1) are
both odd. Show that the equation P (x) = 0 cannot have an integer root.
Solution: Let, if possible, m be an integer root of P (x) = 0, i.e., P (m) = 0. We know
that for any two integers a and b, a − b divides P (a) − P (b). Here, m − 0 | P (m) − P (0)
and m − 1 | P (m) − P (1). Since P (m) = 0, we get that m | P (0) and m − 1 | P (1).
Now one among m − 1 and m must be even and that cannot divide an odd number, so
we get a contradiction here.
Alternate way: m − 1 and m being coprime, we get from the above that m(m − 1) divides
P (0)P (1). But m(m − 1) is even and P (0)P (1) is odd, so we get a contradiction.
3. Define the Fibonacci numbers as: F1 = F2 = 1 and Fn+2 = Fn+1 + Fn for all n ≥ 1.
(In simple words, any term = sum of previous two terms. Thus, the first few terms are:
1, 1, 2, 3, 5, 8, · · · .) Consider the polynomial P (x) = F1 x + F2 x2 + F3 x3 + · · · + F2020 x2020 .
Find the number of integer zeros of P (x).
Solution: Note that P (x) is a polynomial with integer coefficients and P (0) = 0, so x = 0
is a root of P (x) = 0. To find the other roots (if any), we have focus on the rest part,
which is Q(x) = F1 + F2 x + F3 x2 + · · · + F2020 x2019 .
Note that Q(x) is again a polynomial with integer coefficients, and Q(0) = F1 = 1 (odd).
Now we shall find out whether Q(1) is odd. Because if Q(1) is also odd, then we can
apply the previous problem to conclude that Q(x) does not have any integer zero.

14
 It turns out that Q(1) = F1 + F2 + · · · + F2020 is indeed odd, we shall show that now.
Look at the Fibonacci sequence: 1, 1, 2, 3, 5, 8, 13, 21, 34, · · · . Their parity is seen to follow
this pattern: (odd, odd, even), (odd, odd, even), (odd, odd, even), · · · .
How to show that this pattern will continue? We can see this pattern to follow for the
first few blocks of size 3. Suppose this pattern holds to continue till the n-th block of size
3, which consists of F3n−2 , F3n−1 , F3n . What is the next block? F3n+1 , F3n+2 , F3n+3 .
Now F3n+1 = F3n + F3n−1 = even + odd = odd. Similarly, F3n+2 = F3n+1 + F3n =
odd + even = odd and F3n+3 = F3n+2 + F3n+1 = odd + odd = even. Therefore, the
pattern should hold for all the blocks of size 3.
Hence (F1 , F2 , F3 ), (F4 , F5 , F6 ), . . . , (F2017 , F2018 , F2019 ) are all (odd, odd, even), and F2020
is odd. So sum of each block is even. Hence Q(1) = F1 + F2 + · · · + F2019 + F2020 is odd.
Therefore Q(x) = 0 has no integer root and hence 0 is the only integer root of P (x) = 0.
4. Suppose P (x) is a polynomial with P (2) = 2017 and P (5) = 2002. If it is given that
P (x) = 0 has exactly one integer root, find that root.
Solution: Let m be the required root. Since P (x) has integer coefficients, we can use the
fact that a − b | P (a) − P (b). Using this, we get m − 2 | P (m) − P (2) and m − 5 |
P (m) − P (5). Since P (m) = 0, it gives m − 2 | 2017 and m − 5 | 2002. Since 2017 is
a prime, its only divisors are ±1, or ±2017. Therefore, all possible values of m − 2 are
−2017, −1, 1, 2017. This gives m = −2015, 1, 3, 2019. For each of these values of m, we
check whether m − 5 divides 2002. This holds only for m = 3. So our answer is 3.

Next we shall learn one more theorem, which more or less completes our bag of tools, for now.

Theorem. (Rational root theorem) Let A(x) = an xn + an−1 xn−1 + · · · + a1 x + a0 be a

polynomial with integer coefficients. Suppose that x = p/q is a rational root of the equation
A(x) = 0, where p and q are coprime integers. We shall show that p | a0 and q | an .
Proof: Since p/q is a root of the equation A(x) = 0, we have
 n  n−1  
p p p
an + an−1 + · · · + a1 + a0 = 0
q q q
or, an pn + an−1 pn−1 q + · · · + a1 pq n−1 + a0 q n = 0.

All the terms except the last one contains p as a factor. We write the above equation as
−a0 q n = an pn + an−1 pn−1 q + · · · + a1 pq n−1 , which will tell us that p | a0 q n . Since p and q
are coprime, this implies that p | a0 . Similarly, q | an pn , which gives q | an .

We shall remember this theorem as follows: If pq (in reduced form) is a rational root, then
q divides the leading coefficient and p divides the constant term.

15
Corollary. If P (x) is a monic polynomial with integer coefficients then any rational root of
P (x) must be an integer.
(Proof: If p/q (in reduced form) is a rational root, then the rational root theorem tells us
that q | 1, which implies that p/q is an integer.)

Problem. Find all rational roots of the equation 2x10 − 66x6 − 135 = 3x9 − 99x5 − 90x.
Solution: First we bring everything to the left hand side,

(2x10 − 3x9 ) − (66x6 − 99x5 ) + (90x − 135) = 0

and then factorize it as: (2x − 3)(x9 − 33x5 + 45) = 0. The first factor gives the rational root
x = 3/2. What about the second factor? Does it give any rational root?
The polynomial (x9 − 33x5 + 45) is a monic polynomial with integer coefficients. Hence
by rational root theorem, any rational root of it must be an integer.
Does this polynomial P (x) = x9 − 33x5 + 45 have any integer root? Since P (x) is a
polynomial with integer coefficients such that P (0) = 45 and P (1) = 1 − 33 + 45, both odd,
we conclude that P (x) does not have any integer root.
Therefore the given equation has only one rational root, which is 3/2.

I am giving the following problems as homework. We shall discuss them in the next class.

1. Find all rational roots of the equation x4 − 14x3 + 31x2 + 24x − 22 = 0. (Hint: Plug in
small values of x, guess using rational root theorem.)
2. Let f (x) = x3 + ax2 + bx + c and g(x) = x3 + bx2 + cx + a where a, b, c ∈ Z, c 6= 0.
Suppose f (1) = 0 and roots of g(x) = 0 are squares of the roots of f (x) = 0. Find the
value of a2020 + b2020 + c2020 .
3. The product of two of the four roots of the equation x4 − 18x3 + kx2 + 200x − 1984 = 0
is −32. Determine the value of k.
4. Let P (x) = a0 xn + · · · + an−1 x + an be a polynomial with integer coefficients. Suppose
the equation P (x) = 0 has n distinct integer roots which are pairwisely coprime. Then
prove that an−1 and an must be coprime.
5. Find all polynomials P (x) with real coefficients that satisfy xP (x − 1) = (x − 7)P (x) for
every x ∈ R. (Hint: Plug in small values of x; say x = 0, 1, 2, · · · etc.)
6. Let P (x) = x2 + 21 x + b and Q(x) = x2 + cx + d be two polynomials with real coefficients,
such that, P (x)Q(x) = Q(P (x)) for all real x. Find all real roots of P (Q(x)) = 0.

Also, practice problems from An Excursion in Mathematics and other books.

16
 Saturday Online Class 9 (6th class on Polynomials)
July 04, 2020
1. Find all rational roots of the equation x4 − 14x3 + 31x2 + 24x − 22 = 0. (Hint: Plug in
small values of x, guess using rational root theorem.)
Solution: First we try small values of x, e.g., x = 0, ±1, ±2. We find that among these
trial values, only x = −1 is a root of the equation.
Since we are interested only in rational roots, the next thing comes to our mind is the
rational root theorem. This theorem tells us that if x = p/q (in reduced form) is a rational
root of the equation then q | 1 and p | 22. Therefore p/q can be ±1, ±2, ±11, ±22. Note
that if we put x = 2k (an even integer) then the LHS is of the form 2 (mod 4), which
means any even integer can not be a root of the given equation. This rules out ±2 and
±22. So all we need to check now is whether x = ±11 is a root of the equation. Checking
this reveals that the x = 11 is the only rational root other than x = −1. (Answer)

2. Let f (x) = x3 + ax2 + bx + c and g(x) = x3 + bx2 + cx + a where a, b, c ∈ Z, c 6= 0.

Suppose f (1) = 0 and roots of g(x) = 0 are squares of the roots of f (x) = 0. Find the
value of a2020 + b2020 + c2020 .
Solution: Let 1, α, β be the roots of f (x) = 0. Then the roots of g(x) = 0 are 1, α2 , β 2 .
Now we can apply Vieta’s formulae to related 1, α, β to a, b, c, as follows:

(i) 1 + α + β = −a, (ii) α + β + αβ = b, (iii) αβ = −c,

(iv) 1 + α2 + β 2 = −b, (v) α2 + β 2 + α2 β 2 = c, (vi) α2 β 2 = −a.

Now we shall combine them. First, compare (i)2 with (iv) + 2 × (ii) to get a2 = b . Next,
combine (iii) and (vi) to get c2 = −a . Finally, combine (i), (ii) and (iii) or use f (1) = 0
to get 1 + a + b + c = 0 .
Let us write a and b in terms of c (using the first two relations above) and then put them
in the last relation to obtain the possible values of c. Doing this, we get 1 − c2 + c4 + c = 0.
This equation can be factorized as (1 + c)(c3 − c2 + 1) = 0. Note that the equation
c2 (c − 1) = −1 does not have any integer root because the LHS is even for any integer c.
Therefore, there is only one possible value of c (that comes from the first factor) is c = −1.
And hence a = −c2 = −1 and b = a2 = 1. Answer: a2020 + b2020 + c2020 = 1 + 1 + 1 = 3.

3. The product of two of the four roots of the equation x4 − 18x3 + kx2 + 200x − 1984 = 0
is −32. Determine the value of k.
Solution: Let a, b, c, d be the roots of this equation, where ab = −32. Using Vieta’s
theorem, we obtain abcd = −1984, abc + bcd + cda + dab = −200, a + b + c + d = 18,
and k = ab + ac + ad + bc + bd + cd.

17
 First, cd = abcd/ab = −1984/(−32) = 62. Now write the second equation as

ab(c + d) + cd(a + b) = −200 =⇒ −32(c + d) + 62(a + b) = −200.

We can combine this with (c + d) + (a + b) = 18 to obtain a + b = 4 and c + d = 14.

Now you may use ab = −32 and cd = 62 to get the values of a, b, c, d, and hence find k.
A cleverer way: write k = ab + cd + (a + b)(c + d) = (do the calculation) = 86. (Answer)

4. Let P (x) = xn + a1 xn−1 + · · · + an−1 x + an be a polynomial with integer coefficients.

Suppose the equation P (x) = 0 has n distinct integer roots which are pairwisely coprime.
Then prove that an−1 and an must be coprime.
Solution: Let r1 , r2 , . . . , rn be the roots. It is given that r1 , r2 , . . . , rn are distinct integers
that are pairwisely coprime. Applying Vieta’s theorem, we obtain r1 r2 · · · rn = (−1)n an
and r2 r3 · · · rn + r1 r3 r4 · · · rn + · · · + r1 r2 · · · rn−1 = (−1)n−1 an−1 .
Let, if possible, an and an−1 be not coprime. Then they must share some prime, say p, as
a common divisor. Now p | an =⇒ p | r1 r2 · · · rn . Since p is a prime, this implies that
p | ri for some 1 ≤ i ≤ n.
Now in the expression for an−1 , all but one of the summands contain ri and hence
are divisible by p. Hence p | an−1 implies that the term not containing ri (which is
r1 · · · ri−1 ri+1 · · · rn ) must also be divisible by p. This implies that p | rj for some j 6= i.
But p | ri and p | rj contradicts the fact that ri and rj are coprime. 

5. Find all polynomials P (x) with real coefficients that satisfy xP (x − 1) = (x − 7)P (x) for
every x ∈ R. (Hint: Plug in small values of x; say x = 0, 1, 2, · · · etc.)
Solution: It is given that xP (x − 1) = (x − 7)P (x) for every x ∈ R. Putting x = 0
we get P (0) = 0. Then putting x = 1 we get P (1) = 0. Continuing in this way, we get
P (0) = P (1) = P (2) = · · · = P (6) = 0. This means that x = 0, 1, 2, . . . , 6 are zeros
of P (x), so it can be factorized as P (x) = x(x − 1)(x − 2) · · · (x − 6)Q(x) for some
polynomial Q(x). Now put this into the given identity xP (x − 1) = (x − 7)P (x) to get

Q(x − 1) = Q(x) for every x ∈ R.

This implies that Q(x) must be a constant polynomial, say Q(x) ≡ c. Then our P (x) is
given by P (x) = cx(x − 1)(x − 2) · · · (x − 6), where c is any constant. (Answer)
Why should Q be a constant? Note that Q(x − 1) = Q(x) implies Q(8) = Q(9) =
Q(10) = Q(11) = · · · . Therefore the polynomial Q(x) − Q(8) has infinitely many roots,
implying that it must be the zero polynomial. Hence we can say that Q(x) = Q(8), a
constant, for every x.

18
6. Let P (x) = x2 + 12 x + b and Q(x) = x2 + cx + d be two polynomials with real coefficients,
such that, P (x)Q(x) = Q(P (x)) for all real x. Find all real roots of P (Q(x)) = 0.
Solution: If we put x = α in the given identity where α is a zero of P (x), then we get

P (α)Q(α) = Q(P (α)) =⇒ 0 = Q(0) =⇒ d = 0.

Hence Q(x) = x(x + c). Now we can rewrite the identity P (x)Q(x) = Q(P (x)) as

P (x)Q(x) = P (x)(P (x) + c).

Since this is true for all x, we can cancel out P (x) and hence obtain that Q(x) = P (x)+c,
i.e., x(x + c) = x2 + 21 x + b + c. Now we can compare coefficients to get c = 21 and
b + c = 0 =⇒ b = − 21 .
We are asked to find the real roots of P (Q(x)). We found that P (x) = x2 + 21 x − 1
2
=
(x + 1)(x − 21 ). So, P (x) = 0 ⇐⇒ x = −1 or 12 . Hence

1
P (Q(x)) = 0 ⇐⇒ Q(x) = −1 or .
2

Recall, Q(x) = x(x + c) = x(x + 12 ). Hence

1 1 1 1
Q(x) = −1 or ⇐⇒ x2 + x = −1 or x2 + x = .
2 2 2 2
Among these quadratic equations, the first one does not have any real roots, and the
second one has two real roots: −1 and 1/2. Therefore, the equation P (Q(x)) = 0 has
only two real roots, namely x = −1 and 1/2. (Answer)

19