Ideals, Varieties, and Algorithms.

An introduction to Computational Algebraic Geometry and

Commutative Algebra. Third Edition (2007). Springer.

David Cox, John Little, Donal O’Shea

Solved exercises by Gerardo Manuel García

Managua-Nicaragua

May, 2021

Exercises for §𝟏 (𝐺𝑒𝑟𝑎𝑟𝑑𝑜)

2. Let 𝔽𝟐 be the field from exercise 1.
a. Consider the polynomial 𝒈(𝒙, 𝒚) = 𝒙𝟐 𝒚 + 𝒚𝟐 𝒙 ∈ 𝔽𝟐 [𝒙, 𝒚]. Show that
𝒈(𝒙, 𝒚) = 𝟎 for every (𝒙, 𝒚) ∈ 𝔽𝟐𝟐 , and explain why this does not
contradict Proposition 5.

Solution:
Since, the affine space 𝔽22 has only as its elements
(0,0), (0,1), (1,0)𝑎𝑛𝑑 (1,1), we can see, at a glance, that the first three
listed elements yield 0 as the value of 𝑔, and with the latter we get 1 + 1 =
0. So, for every (𝑥, 𝑦) ∈ 𝔽22 we get 𝑔(𝑥, 𝑦) = 0.

This result does not contradict Proposition 5, because this proposition

refers to infinite fields, but 𝔽2 isn’t infinite.

b. Find a nonzero polynomial in 𝔽𝟐 [𝒙, 𝒚, 𝒛] which vanishes at every point of

𝔽𝟑𝟐 . Try to find one involving all three variables.

Solution:
We have to consider that 𝔽32 = {(0,0,0), (0,0,1), (0,1,0), (0,1,1), (1,0,0),
(1,0,1), (1,1,0), (1,1,1)}. So, we can check that the following nonzero
polynomials vanish at every point of 𝔽32 .
𝑔(𝑥, 𝑦, 𝑧) = 𝑥 2 𝑦 + 𝑦 2 𝑥

𝑓(𝑥, 𝑦, 𝑧) = 𝑥𝑦𝑧 + 𝑥𝑦 2 𝑧

We can see that the nonzero polynomial ℎ(𝑥, 𝑦, 𝑧) = 𝑥 𝑚 𝑦 𝑛 𝑧 𝑝 + 𝑥 𝑟 𝑦 𝑠 𝑧 𝑡 ,

where 𝑚, 𝑛, 𝑝, 𝑟, 𝑠, 𝑡 are positive integers, vanishes at every point of 𝔽32 .

Even more, if we have a polynomial with an even number of terms and, in

which term, we have the three variables, then this will vanish at every point
of 𝔽32 .
 c. Find a nonzero polynomial in 𝔽𝟐 [𝒙𝟏 , … , 𝒙𝒏 ] which vanishes at every point
of 𝔽𝒏𝟐 . Can you find one in which all of 𝒙𝟏 , … , 𝒙𝒏 appear?

Solution:
If the polynomial
𝛼 𝛼 𝛼
𝑓 = ∑ 𝑥 𝛼 𝑤ℎ𝑒𝑟𝑒 𝛼 = (𝛼1 , … , 𝛼𝑛 ) 𝑎𝑛𝑑 𝑥 𝛼 = 𝑥1 1 ∙ 𝑥2 2 , … , 𝑥𝑛 𝑛
𝛼
has an even number of terms and each term has all the variables 𝑥1 , … , 𝑥𝑛 ,
then it vanishes at every point of 𝔽𝑛2 . We can see this by thinking that all the
elements of the affine space 𝔽𝑛2 , but one, has at least one zero as entry, so
with all of this 𝑛 − 1 points, 𝑓 vanishes. Then, we have one element that

has no zeros, and only ones, (1,

⏟… ,1), then, since the polynomial has an
𝑛 𝑡𝑖𝑚𝑒𝑠

even number of terms, we will have 𝑓 (1,

⏟… ,1) = 0.
𝑛 𝑡𝑖𝑚𝑒𝑠

3. (Requires abstract algebra.) Let 𝒑 be a prime number. The ring of integers

module 𝒑 is a field with 𝒑 elements, which we will denote 𝔽𝒑 .

a. Explain why 𝔽𝒑 − {𝟎} is a group under multiplication.

Solution:
For 𝔽𝑝 − {0} to be a group, we have to prove associativity, identity and
inverse.
Since the numbers 1,2, … , 𝑝 − 1 are integers, the associativity holds with
this set of numbers too.

The identity element is 1.

We have to prove that if 𝑥 ∈ 𝔽𝑝 − {0}, then there exists 𝑥 −1 ∈ 𝔽𝑝 − {0}

such that 𝑥 ∙ 𝑥 −1 = 1 = 𝑥 −1 ∙ 𝑥. To do this, we will use the Fermat’s Little
Theorem.

Fermat’s Little Theorem

Let 𝑝 be a prime number and 𝑎 any integer not divisible by 𝑝. Then,
𝑎𝑝−1 ≡ 1 (𝑚𝑜𝑑 𝑝).

So, if 𝑥 ∈ 𝔽𝑝 − {0}, then 𝑝 ∤ 𝑥 because 0 < 𝑥 < 𝑝. So, by the Femat’s Little
Theorem we have that 𝑥 𝑝−1 ≡ 1 (𝑚𝑜𝑑 𝑝), if we take 𝑥 𝑝−2 ∈ 𝔽𝑝 − {0},
then we can see that 𝑥 𝑝−2 is the inverse element of 𝑥:
𝑥 𝑝−2 ∙ 𝑥 ≡ 𝑥 ∙ 𝑥 𝑝−2 ≡ 𝑥 𝑝−1 ≡ 1 (𝑚𝑜𝑑 𝑝).

b) Use Lagrange’s Theorem to show that 𝒂𝒑−𝟏 = 𝟏 for all 𝒂 ∈ 𝔽𝒑 − {𝟎}.

Solution:
Lagrange’s Theorem states that if H is a subgroup of a finite group 𝑮, then
the order of H is a divisor of the order of 𝑮. As a collorary, we have that if
𝑮 is a finite group and 𝒈 ∈ 𝑮, then the order of 𝒈 divides the order of 𝑮.

Let us suppose that 𝑥 ∈ 𝔽𝑝 − {0} = 𝐺. Since 𝐺 is a finite group, then the

order of 𝑥 cannot be infinite. So, let 𝑘 be the order of 𝑥. We have the
following three cases:

Case 1: 0 < 𝑘 < 𝑝 − 1

In this case, the set of numbers {𝑥, 𝑥 2 , 𝑥 3 , … , 𝑥 𝑘 } is a subgroup of 𝐺 under
multiplication. So, by Lagrange’s Theorem, 𝑘 divides 𝑝 − 1. This means that
𝑝 − 1 = 𝑘𝑚, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑚 ∈ ℤ, then 𝑥 𝑝−1 = 𝑥 𝑘𝑚 = (𝑥 𝑘 )𝑚 = 1𝑚 = 1.

Case 2: 𝑘 = 𝑝 − 1
We use a property of the residual classes module a prime p:
If 𝒑 is a prime number, then 𝒂𝒑 ≡ 𝒂 (𝒎𝒐𝒅 𝒑), ∀𝒂 ∈ ℤ. Since the elements of 𝐺
are integers, this property holds in 𝐺, then we have that 𝑖𝑓 𝑥 ∈ 𝐺, 𝑡ℎ𝑒𝑛 𝑥 𝑝 ≡
𝑥 (𝑚𝑜𝑑 𝑝), and multiplying by the inverse of 𝑥, we have: 𝑥 𝑝 ∙ 𝑥 −1 ≡ 𝑥 ∙
𝑥 −1 (𝑚𝑜𝑑 𝑝) from which we get 𝑥 𝑝−1 ≡ 1 (𝑚𝑜𝑑 𝑝) for all 𝑥 ∈ 𝔽𝑝 − {0}.

Case 3: 𝑘 > 𝑝 − 1
This is impossible, because, if so, we have to have a subgroup of an order
greater than the order of the group.

So, cases 1 and 2, show us the desired result.

c) Prove that 𝒂𝒑 = 𝒂 for all 𝒂 ∈ 𝔽𝒑 . Hint: treat the cases 𝒂 = 𝟎 and 𝒂 ≠ 𝟎

separately.

Solution
Case 1: 𝒂 = 𝟎
In this case, we have that 𝑎𝑝 = 0𝑝 = 0 = 𝑎.

Case 2: 𝒂 ≠ 𝟎
Recalling that 𝐺 = 𝔽𝑝 − {0} is a group under multiplication, let us suppose
that 𝑎𝑝 ≠ 𝑎 (𝑚𝑜𝑑 𝑝). Then, 𝑎𝑝−1 ∙ 𝑎 ≠ 𝑎, so the order of 𝑎 is a positive
integer 𝑘, such that 𝑘 < 𝑝 − 1. This means that there exists a subgroup of
𝐺 = 𝔽𝑝 − {0}, with order 𝑘, but this implies, by Langrange’s Theorem,
that 𝑘 | 𝑝 − 1. Hence 𝑝 − 1 = 𝑘𝑚, 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑚 ∈ ℤ and 𝑎𝑝−1 = 𝑎𝑚𝑘 = 1,
which leads us to a contradiction. So, 𝑎𝑝 = 𝑎 (𝑚𝑜𝑑 𝑝) for all 𝑎 ∈ 𝔽𝑝 − {0}.

By cases 1 and 2, we can assert that 𝑎𝑝 = 𝑎 for all 𝑎 ∈ 𝔽𝑝 .

 d) Find a nonzero polynomial in 𝔽𝒑 [𝒙] which vanishes at every point of 𝔽𝒑 .
Hint: use part c.

Solution
We can use part c in this way. Since 𝑎𝑝 = 𝑎 (𝑚𝑜𝑑 𝑝) for all 𝒂 ∈ 𝔽𝒑 , then, if
𝑥 ∈ 𝔽𝑝 we have that 𝑎𝑝 = 𝑎 (𝑚𝑜𝑑 𝑝) → 𝑎𝑝 − 𝑎 = 0 (𝑚𝑜𝑑 𝑝), so we
can write the polynomial 𝑓(𝑥) = 𝑥 𝑝 − 𝑥 which is not the zero
polynomial and vanishes at every point of 𝔽𝒑 . Even more, the
polynomials of the form 𝑔(𝑥) = 𝑠𝑥𝑝 − 𝑠𝑥, 𝑤𝑖𝑡ℎ 𝑠 ∈ 𝔽𝑝 are nonzero
polynomials that vanish at every point of 𝔽𝑝 .

4. (Requires abstract algebra.) Let 𝑭 be a finite field with q elements. Adapt the
argument of Exercise 3 to prove that 𝒙𝒒 − 𝒙 is a nonzero polynomial in 𝑭[𝒙]
which vanishes at every point of 𝑭. This shows that proposition 5 fails for all
finite fields.

By Lagrange’s Theorem we know that if 𝑮 is a finite group and 𝒈 ∈ 𝑮, then the order
of 𝒈 divides the order of 𝑮. This means that, if 𝒎 is the order of 𝑥 ∈ 𝐺 = 𝑭 − {𝟎}, then
= (𝒙 ) = 𝒆 = 𝒆.
𝒒−𝟏 𝒎𝒌 𝒎 𝒌 𝒌
𝒒 − 𝟏 = 𝒎𝒌 for some 𝒌 ∈ ℤ+ . So we have that 𝒙 =𝒙
Where 𝑒 is the multiplicative identity of 𝐺. Hence, 𝑓(𝑥) = 𝑥 𝑞−1 𝑥 − 𝑥 = 𝑥 𝑞 −
𝑥, vanishes for all 𝑥 ∈ 𝐹, this assertion is for all 𝑥 ∈ 𝐹, because the polynomial
vanishes when 𝑥 = 0.

5. In the proof of Proposition 5, we took 𝒇 ∈ 𝒌[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] and wrote it as a

polynomial in 𝒙𝒏 with coefficients in 𝒌[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 ]. To see what this look
like in a specific case, consider the polynomial

𝒇(𝒙, 𝒚, 𝒛) = 𝒙𝟓 𝒚𝟐 𝒛 − 𝒙𝟒 𝒚𝟑 + 𝒚𝟓 + 𝒙𝟐 𝒛 − 𝒚𝟑 𝒛 + 𝒙𝒚 + 𝟐𝒙 − 𝟓𝒛 + 𝟑

a. Write 𝒇 as a polynomial in 𝒙 with coefficients in 𝒌[𝒚, 𝒛 ].

b. Write 𝒇 as a polynomial in 𝒚 with coefficients in 𝒌[𝒙, 𝒛 ].
c. Write 𝒇 as a polynomial in 𝒛 with coefficients in 𝒌[𝒙, 𝒚 ].

Let us recall what was done in the proof of Proposition 5:

Let 𝒇 ∈ 𝒌[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ]. By collecting the various powers of 𝑥𝑛 , we can
write 𝑓 in the form

𝒇 = ∑ 𝒈𝒊 (𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 )𝒙𝒊𝒏

𝒊=𝟎

where 𝒈𝒊 ∈ 𝒌[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 ].

For the case a, we have:

𝒇(𝒙, 𝒚, 𝒛) = 𝒙𝟓 𝒚𝟐 𝒛 − 𝒙𝟒 𝒚𝟑 + 𝒚𝟓 + 𝒙𝟐 𝒛 − 𝒚𝟑 𝒛 + 𝒙𝒚 + 𝟐𝒙 − 𝟓𝒛 + 𝟑
 where 𝑛 = 5, so 𝑔5 (𝑦, 𝑧) = 𝑦 2 𝑧, 𝑔4 (𝑦, 𝑧) = −𝑦 3 , 𝑔3 (𝑦, 𝑧) = 0, 𝑔2 (𝑦, 𝑧) = 𝑧,
𝑔1 (𝑦, 𝑧) = 𝑦 + 2 and 𝑔0 (𝑦, 𝑧) = 𝑦 5 − 𝑦 3 𝑧 − 5𝑧 + 3. Substituting the 𝑔𝑖 in 𝑓, we
have the required form of the polynomial:

𝒇(𝒙) = 𝑔5 (𝑦, 𝑧)𝒙𝟓 + 𝑔4 (𝑦, 𝑧)𝒙𝟒 + 𝑔3 (𝑦, 𝑧)𝑥 3 + 𝑔2 (𝑦, 𝑧)𝒙𝟐 + 𝑔1 (𝑦, 𝑧)𝒙 + 𝑔0 (𝑦, 𝑧)

Thinking in the same way, we have the following form for 𝒇(𝒚) and 𝒇(𝒛):

Case b
With ℎ5 (𝑥, 𝑧) = 1, ℎ4 (𝑥, 𝑧) = 0, ℎ3 (𝑥, 𝑧) = −𝑥 4 − 𝑧, ℎ2 (𝑥, 𝑧) = 𝒙𝟓 𝒛,
ℎ1 (𝑥, 𝑧) = 𝑥 and ℎ0 (𝑥, 𝑧) = 𝑥 2 𝑧 + 2𝑥 − 5𝑧 + 3, then

𝟒
𝒇(𝒚) = ℎ5 (𝑥, 𝑧)𝒚𝟓 + ℎ4 (𝑥, 𝑧)𝑦 + ℎ3 (𝑥, 𝑧)𝑦 3 + ℎ2 (𝑥, 𝑧)𝒚𝟐 + ℎ1 (𝑥, 𝑧)𝒚 + ℎ0 (𝑥, 𝑧)

Case c
With 𝑞1 (𝑥, 𝑦) = 𝑥 5 𝑦 2 + +𝑥 2 − 𝑦 3 − 5 and 𝑞0 (𝑥, 𝑦) = −𝑥 4 𝑦 3 + 𝑦 5 + 𝑥𝑦 +
2𝑥 + 3, then

𝒇(𝒛) = 𝑞1 (𝑥, 𝑦)𝒛 + 𝑞0 (𝑥, 𝑦)

6. Inside of ℂ𝒏 , we have the subset ℤ𝒏 , which consists of all points with integer
coordinates.

a. Prove that if 𝒇 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] vanishes at every point of ℤ𝒏 , then 𝒇 is

the zero polynomial. Hint: Adapt the proof of Proposition 5.

To prove this proposition, let us analyze the proof given by the textbook for
Proposition 5. I will give some remarks, as footnote, in order to get a better
understanding of the proof.

Proposition 5: Let 𝑘 be an infinite field and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛 ]. Then 𝑓 = 0

in 𝑘[𝑥1 , … , 𝑥𝑛 ] if and only if 𝑓: 𝑘 𝑛 → 𝑘 is the zero function.

The proof given by the textbook includes two parts:

Part one: →
Let 𝑘 be an infinite field and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛 ]. If 𝑓 = 0 in 𝑘[𝑥1 , … , 𝑥𝑛 ] then
𝑓: 𝑘 𝑛 → 𝑘 is the zero function.

One direction of the proof is obvious1 since the zero polynomial clearly gives
the zero function.

1
Let 𝑘 be an infinite field and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛 ]. If 𝑓 = 0 in 𝑘[𝑥1 , … , 𝑥𝑛 ]; it means that 𝑓 =
∑𝛼 𝑎𝛼 𝑥 𝛼 = 0 𝑖𝑛 𝑤ℎ𝑖𝑐ℎ 𝑎𝛼 = 0. This implies that 𝑓(𝑥1 , … , 𝑥𝑛 ) = 0, 𝑓𝑜𝑟 𝑎𝑙𝑙 (𝑥1 , … , 𝑥𝑛 ) ∈ 𝑘 𝑛 , so 𝑓: 𝑘 𝑛 →
𝑘 is the zero function.
 Part two: ←
Let 𝑘 be an infinite field and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛 ]. If 𝑓: 𝑘 𝑛 → 𝑘 is the zero
function, then 𝑓 = 0 in 𝑘[𝑥1 , … , 𝑥𝑛 ].

To prove the converse, we need to show that if 𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛 ) = 0 for all

(𝑎1 , 𝑎2 , … , 𝑎𝑛 ) ∈ 𝑘 𝑛 then 𝑓 is the zero polynomial. We will use induction on
the number of variables 𝑛.

When 𝑛 = 1, it is well known that a nonzero polynomial in 𝑘[𝑥] of degree 𝑚

has at most m distinct roots. (This is by the Corollary 3, §5, Chapter 1.) For
our particular 𝑓 ∈ 𝑘[𝑥], we are assuming 𝑓(𝑎) = 0 for all 𝑎 ∈ 𝑘 (this is the
hypothesis in the direction ←). Since 𝑘 is infinite, this means that the
polynomial 𝑓 has infinitely many roots, and, hence 𝑓 must be the zero
polynomial (because, if 𝑓 is not the zero polynomial, this would contradict
the Corollary 3).

Now assume that the converse is true2 for 𝑛 − 1, and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛 ] be
a polynomial that vanishes at all points of 𝑘 𝑛 . By collecting the various
powers of 𝑥𝑛 , we can write 𝑓 in the form3
𝑁

𝑓 = ∑ 𝑔𝑖 (𝑥1 , … , 𝑥𝑛−1 )𝑥𝑛𝑖 ,

𝑖=0
where 𝑔𝑖 ∈ 𝑘[𝑥1 , … , 𝑥𝑛−1 ]. We will show that each 𝑔𝑖 is the zero polynomial
in 𝑛 − 1 variables, which will force 𝑓 to be the zero polynomial in
𝑘[𝑥1 , … , 𝑥𝑛 ].

If we fix (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) ∈ 𝑘 𝑛−1 , we get the polynomial

𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 , 𝑥𝑛 ) ∈ 𝑘[𝑥𝑛 ]. By our hypothesis4 on 𝑓, this vanishes for
every 𝑎𝑛 ∈ 𝑘. It follows from the case 𝑛 = 1 that 𝑓(𝑎1 , … , 𝑎𝑛−1 , 𝑥𝑛 ) is the
zero polynomial in 𝑘[𝑥𝑛 ]. Using the above formula for 𝑓, we see that the
coefficients of f 𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 , 𝑥𝑛 ) are 𝑔𝑖 (𝑎1 , … , 𝑎𝑛−1 ), and thus,
𝑔𝑖 (𝑎1 , … , 𝑎𝑛−1 ) = 0 for all 𝑖. Since (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) was arbitrarily chosen in
𝑘 𝑛−1 , it follows that each 𝑔𝑖 ∈ 𝑘[𝑥1 , … , 𝑥𝑛−1 ] gives the zero function on
𝑘 𝑛−1 . Our inductive assumption then implies that each 𝑔𝑖 is the zero
polynomial in 𝑘[𝑥1 , … , 𝑥𝑛−1 ]. This forces f to be the zero polynomial in
𝑘[𝑥1 , … , 𝑥𝑛 ] and completes the proof of the proposition.

2
Here the assumption is: Let 𝑘 be an infinite field and let 𝑓 ∈ 𝑘[𝑥1 , … , 𝑥𝑛−1 ]. If 𝑓: 𝑘 𝑘−1 → 𝑘 is the zero
function, then 𝑓 = 0 in 𝑘[𝑥1 , … , 𝑥𝑛−1 ].
3
In this form of the polynomial 𝑓 the summation goes from 0 to 𝑁, where 𝑁 is the highest power of 𝑥𝑛 .
3
So, for example, let 𝑓 = 2𝑥 3 𝑦 2 𝑧 + 𝑦 3 𝑧 3 − 3𝑥𝑦𝑧 + 𝑦 2 . In this polynomial, the highest power of 𝑥 is
2
𝑁 = 3, so, we can write 𝑓, collecting the various powers of 𝑥, in the form 𝑓 = ∑3𝑖=0 𝑔𝑖 (𝑦, 𝑧)𝑥 𝑖 ,
3
where 𝑔0 (𝑦, 𝑧) = 𝑦 3 𝑧 3 + 𝑦 2 , 𝑔1 (𝑦, 𝑧) = −3𝑦𝑧, 𝑔2 (𝑦, 𝑧) = 0 and 𝑔3 (𝑦, 𝑧) = 2𝑦 2 𝑧.
2
4
This is the hypothesis of the converse proposition (←), i.e., If 𝒇: 𝒌𝒌 → 𝒌 is the zero function, then ...so
this is stated in this part of the proof: Now assume that the converse is true4 for 𝑛 − 1, and let 𝑓 ∈
𝑘[𝑥1 , … , 𝑥𝑛 ] be a polynomial that vanishes at all points of 𝑘 𝑛 .
 Now, we are going to prove that if 𝒇 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] vanishes at every
point of ℤ𝒏 , then 𝒇 is the zero polynomial, by adapting the proof of
Proposition 5.

We need to show that if 𝒇 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] and 𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛 ) = 0 for

all (𝑎1 , 𝑎2 , … , 𝑎𝑛 ) ∈ ℤ𝑛 , then 𝑓 is the zero polynomial. We will use induction
on the number of variables.

If 𝑛 = 1, we know that a nonzero polynomial in ℂ[𝑥 ] of degree 𝑚 has at

most 𝑚 distinct roots (This is considering what is stated by the Corollary 3,
§5, Chapter 1). For our particular polynomial 𝑓 ∈ ℂ[𝒙 ], we are assuming
𝑓(𝑎) = 0, for all 𝑎 ∈ ℤ ⊂ ℂ. Since ℤ is infinite this means that 𝑓 has infinitely
many roots, and hence, 𝑓 must be the zero polynomial.

Now assume that the converse is true for 𝑛 − 1, and let 𝒇 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ]
be a polynomial that vanishes at all points of ℤ𝑛 . By collecting the various
powers of 𝑥𝑛 we can write 𝑓 in the form

𝑓 = ∑ 𝑔𝑖 (𝑥1 , … , 𝑥𝑛−1 )𝑥𝑛𝑖 (∗)

𝑖=0
where 𝑔𝑖 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 ]. We have to show that each 𝑔𝑖 is the zero
polynomial in 𝑛 − 1 variables, which will force 𝑓 to be the zero polynomial in
ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ].

If we fix (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) ∈ ℤ𝑛−1 , we get the polynomial

𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 , 𝑥𝑛 ) ∈ ℂ[𝒙𝒏 ]. By our hypothesis on 𝑓, this vanishes for
every 𝑎𝑛 ∈ ℤ. It follows from the case 𝑛 = 1 that 𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 , 𝑥𝑛 ) is the
zero polynomial in ℂ[𝒙𝒏 ]. Using the formula (∗) for 𝑓, we can see that the
coefficients of 𝑓(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 , 𝑥𝑛 ) are 𝑔𝑖 (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ), and thus,
𝑔𝑖 (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) = 0 for all 𝑖. Since (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) was arbitrarily
chosen in ℤ𝑛−1 , it follows that 𝑔𝑖 (𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) = 0 for all
(𝑎1 , 𝑎2 , … , 𝑎𝑛−1 ) ∈ ℤ𝑛−1 𝑔𝑖 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 ] gives the zero function on
ℤ𝑛−1 . Our inductive assumption then implies that each that each 𝑔𝑖 is the
zero polynomial in ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏−𝟏 ]. This forces 𝑓 to be the zero polynomial
in ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] and completes the proof of the proposition.

b. Let 𝒇 ∈ ℂ[𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 ] and let 𝑴 be the largest power of any variable that
appears in 𝒇. Let ℤ𝒏𝑴+𝟏 be the set of points of ℤ𝒏 , all coordinates of which
lie between 1 and 𝑴 + 𝟏. Prove that if f vanishes at all points of ℤ𝒏𝑴+𝟏 , then
𝒇 is the zero polynomial.