Solutions to Exercises

In this chapter we present solutions to some exercises along with hints for
helping to solve others.

Solutions to Exercises in Chapter I

Exercise 1.8. We prove first the validity of the associative law of addition,
that is, the equality

n + (m + p) = (n + m) + p (1)

N
for all n, m, p ∈ . We do this by induction on p. For p = 0, the assertion is
clear. Suppose, then, that assertion (1) holds for an arbitrary but fixed p ∈ N
N
and for all n, m ∈ . Then from the definition of addition and the induction
hypothesis, we have the equality

(n + m) + p∗ = ((n + m) + p)∗ = (n + (m + p))∗

= n + ( m + p ) ∗ = n + ( m + p ∗ ),

as desired. This proves the associativity of addition. The first distributive

law, that is, the equality

(n + m) · p = n · p + m · p (2)

N
for all n, m, p ∈ , can be proved using the associativity and commutativity
of addition along with induction on p. For p = 0, the assertion is clear. Sup-
N
pose now that (2) holds for an arbitrary but fixed p ∈ and for all n, m ∈ . N
Then we have

(n + m) · p∗ = (n + m) · p + (n + m) = (n · p + m · p) + (n + m)
= (n · p + n) + (m · p + m) = n · p∗ + m · p∗ ,

as desired. This proves the asserted distributivity. Using the first distributive
law, one can then prove the commutativity of multiplication by induction.
This then also yields the validity of the second distributive law. Finally, one
can prove the associativity of multiplication by induction.

N
Exercise 1.10. Let m, n ∈ . If m = 0 or n = 0, then one immediately has
the equality m · n = 0 by Definition 1.5 (2) and the commutativity of mul-

© Springer International Publishing AG 2017

J. Kramer and A.-M. von Pippich, From Natural Numbers to Quaternions,
Springer Undergraduate Mathematics Series, https://doi.org/10.1007/978-3-319-69429-0
248 Solutions to Exercises

tiplication. To prove the converse, we assume m 6= 0 and n 6= 0 and prove

the inequality m · n 6= 0. Since m 6= 0 and n 6= 0, there exist a, b ∈ with N
m = a∗ = a + 1 and n = b∗ = b + 1. Therefore,

m · n = m · b ∗ = ( m · b ) + m = ( m · b ) + ( a + 1) = ( m · b + a ) + 1 = ( m · b + a ) ∗ ,

that is, the natural number m · n is the successor of the natural number
m · b + a. By the third Peano axiom, we must then have m · n 6= 0.

Exercise 1.14. The power law from Lemma 1.13 can by proved by induction.

Exercise 1.17. The proof of properties (i), (ii), and (iii) of Remark 1.16 are left
to the reader.

Exercise 1.20. Properties (i) and (ii) of Remark 1.19 can be proved by induc-
tion.

Exercise 1.23. We leave it to the reader to come up with suitable examples.

Exercise 1.25. Let m, n ∈ N

with m ≥ n. We first prove the existence of a
N
natural number x ∈ with n + x = m. If m = n, then for x = 0, we have the
equality n + x = n + 0 = m. If m > n, then there exists a ∈ , a > 0, such N
that m = n∗···∗ ( a times). Then for x = 0∗···∗ (a times), we have the equality
n + x = n + 0∗···∗ = n∗···∗ = m. To prove uniqueness, let y ∈ be another N
natural number with n + y = m. If x < y, then by Remark 1.19 (i) and the
commutativity of addition, we have the inequality

m = n + x = x + n < y + n = n + y = m,

that is, we conclude that m < m, a contradiction. If x > y, then there fol-
lows analogously the contradiction m > m. Therefore, we must have x = y,
proving the asserted uniqueness.

Exercise 2.5. (a) By assumption we have 3 | ( a1 · · · ak + 1), that is, there exists
n∈ N with a1 · · · ak + 1 = 3 · n. If now 3 | a j for some j ∈ {1, . . . , k }, then
by Lemma 2.4 (ix), we have 3 | ( a1 · · · ak ), that is, there exists m ∈ N with
a1 · · · ak = 3 · m. We therefore have the equality

1 = 3 · n − a1 · · · a k = 3 · n − 3 · m = 3 · ( n − m ),

a contradiction. Thus none of the numbers a1 , . . . , ak is divisible by 3.

(b) We assume that none of the numbers a1 + 1, . . . , ak + 1 is divisible by 3.
One first shows that then one must have a j + 1 = 3 · n j + r j for certain n j ∈ N
and r j ∈ {1, 2} (j = 1, . . . , k), from which follows a j = 3 · n j + (r j − 1), which
for j = 1, . . . , k implies the equality r j = 2, since by part (a), no a j is divisible
by 3. By multiplying out, one shows that the number
Solutions to Exercises in Chapter I 249

k
a1 · · · a k − 1 = ∏ (3 · n j + 1) − 1
j =1

is divisible by 3. But this contradicts the assumption that a1 · · · ak + 1 is di-

visible by 3. Therefore, at least one of the numbers a1 + 1, . . . , ak + 1 must be
divisible by 3.

Exercise 2.12. To make the proceedings clear, we calculate first, with a1 = 2

N
and an+1 = ( an − 1) · an + 1 for n ∈ , n ≥ 1, the numbers a2 = 3, a3 = 7,
a4 = 43, a5 = 1 807, a6 = (1807 − 1) · 1806 + 1 = 3 263 443. With Mn = { p ∈ P |
p | an }, we obtain the first six sets,

M1 = {2}, M2 = {3}, M3 = {7}, M4 = {43},

M5 = {13, 139}, M6 = {3 263 443}.

N
We claim that we have the equality an = 5 · bn + rn with bn ∈ and rn ∈ {2, 3}
N
(n ∈ , n ≥ 1). This can be proved by induction on n. If n = 1, the assertion is
clear. We now assume that the assertion holds for arbitrary but fixed n ∈ , N
n ≥ 1. Then we have

a n + 1 = ( a n − 1 ) · a n + 1 = ( 5 · bn + r n − 1 ) · ( 5 · bn + r n ) + 1
= 5 · (5bn2 + 2bn rn − bn ) + rn2 − rn + 1.

If rn = 2, then rn2 − rn + 1 = 3; if rn = 3, then rn2 − rn + 1 = 5 + 2. Therefore, in

N
both cases, an+1 is of the form 5 · bn+1 + rn+1 with bn+1 ∈ and rn+1 ∈ {2, 3},
as desired. We have thus shown that 5 - an , that is, 5 ∈ / Mn , for all n ∈ , N
n ≥ 1.

Exercise 2.13. Let us assume, contrary to the assertion, that there are only
N
finitely many prime numbers p1 , . . . , pn in 2 + 3 · . We then consider the
natural number
a : = 3 · p1 · · · p n − 1 .
We have that a > 1, and by Lemma 2.9, a has a prime divisor p. Since 3 - a,
N
it follows that p 6= 3. We now show that p ∈ 2 + 3 · ; that is, we must show
that 3 | ( p + 1). If p = a, we are done. If p < a, then there exists q ∈ , q > 1, N
with a = p · q. Since 3 | ( p · q + 1), it follows by Exercise 2.5 (b) that 3 | ( p + 1)
or 3 | (q + 1). In the first case, we are done. In the second case, we repeat the
process for a prime divisor of q. Finally, after finitely many steps, we obtain
N
a prime divisor p of a with p ∈ 2 + 3 · . We now proceed as in Euclid’s
proof, for on the assumption that there are only finitely many prime num-
N
bers in the set 2 + 3 · , we must have p ∈ { p1 , . . . , pn }. In particular, we have
p | ( p1 · · · pn ). However, since we have the divisibility relation p | a, we must
have p | 1 from the divisibility rules, which is a contradiction.

Exercise 2.15. We prove the contrapositive of the asserted implication.

250 Solutions to Exercises

(i) Suppose that n is not a prime. Then there exist natural numbers a, b ∈ N
with n = a · b and 1 < a, b < n. We thus obtain

2n − 1 = 2a·b − 1 = (2a − 1) · (2a·(b−1) + 2a·(b−2) + · · · + 2a + 1).

Since 1 < 2a − 1 < 2n − 1, it follows that 2a − 1 is a nontrivial divisor of

2n − 1, which proves that 2n − 1 is not prime.
N
(ii) Let n ∈ , n > 0, with n not a power of 2. Then n > 2, and there exist
N
natural numbers a, b ∈ , b odd, with n = a · b and 1 ≤ a < n, 1 < b ≤ n.
Since b is odd, we obtain

2n + 1 = 2a·b + 1 = (2a + 1) · (2a·(b−1) ∓ · · · − 2a + 1).

Since 1 < 2a + 1 < 2n + 1, it follows that 2a + 1 is a nontrivial divisor of

2n + 1, which proves that 2n + 1 is not prime.

Exercise 2.18. Let { a1 , . . . , an } and {b1 , . . . , bm } denote the sets of all divisors
of a and all divisors of b. Since a and b are relatively prime, the set of all
divisors of a · b is equal to { a j · bk | j = 1, . . . , n; k = 1, . . . , m}. It follows that
n m
S( a) · S(b) = ( a1 + · · · + an ) · (b1 + · · · + bm ) = ∑ ∑ a j · bk = S ( a · b ) .
j =1 k =1

This completes the proof of the assertion.

Exercise 2.19. The assertion can be proved by induction on m.

Exercise 2.20. (a) We have the equalities

S(220) − 220 = 1 + 2 + 4 + 5 + 10 + 11 + 20 + 22 + 44 + 55 + 110 = 284,

S(284) − 284 = 1 + 2 + 3 + 71 + 142 = 220.

Therefore, we have S(220) = 220 + 284 = S(284), which proves that the
numbers 220 and 284 are amicable.
(b) We must show that S( a) = a + b = S(b). Since x, y, z are distinct odd
primes, it follows that

S( a) = S(2n · x · y) = S(2n ) · S( x ) · S(y) = (2n+1 − 1)( x + 1)(y + 1),

S(b) = S(2n ) · S(z) = (2n+1 − 1)(z + 1),

where we have used the well-known equality S(2n ) = 2n+1 − 1. A direct

calculation shows that x · y = 9 · 22n−1 − 9 · 2n−1 + 1, and therefore, x · y +
x + y = z. This implies S( a) = (2n+1 − 1)(z + 1) = S(b). Finally, we calculate
Solutions to Exercises in Chapter I 251

a + b = 2n · ( x · y + z) = 2n · (9 · 22n − 9 · 2n−1 ) = 22n−1 · 9 · (2n+1 − 1)

= ( z + 1 ) · (2n +1 − 1 ) = S ( a ) = S ( b ),

which proves that the numbers a and b are amicable.

Exercise 3.2. We obtain the prime factorizations 720 = 24 · 32 · 5, 9 797 = 97 ·

101 and 360360 = (23 · 32 · 5)360 = 21 080 · 3720 · 5360 . Finally, on using the third
binomial formula four times, we obtain

232 − 1 = (22 − 1) · 22 + 1 · 24 + 1 · 28 + 1 · 216 + 1

= 3 · 5 · 17 · 257 · 65 537 .

Exercise 3.7. Let a = 232 − 1 and b = 255 with prime factorizations (see Ex-
ercise 3.2)

a= ∏ pa p = 3 · 5 · 17 · 257 · 65 537, b= ∏ pbp = 3 · 5 · 17;

p ∈P p ∈P

here a3 = 1, a5 = 1, a17 = 1, a257 = 1, a65 537 = 1, a p = 0 for all p ∈ P \

{3, 5, 17, 257, 65 537} and b3 = 1, b5 = 1, b17 = 1, b p = 0 for all p ∈ P \ {3, 5, 17}.
Therefore, b p ≤ a p for all p ∈ P, which by the criterion of Lemma 3.5, proves
that b | a.

Exercise 4.6. With the help of Theorem 4.3, we obtain (3 600, 3 240) = 360,
(360360 , 540180 ) = ((23 · 32 · 5)360 , (22 · 33 · 5)180 ) = 2360 · 3540 · 5180 , (232 −
1, 38 − 28 ) = 5, where for the last equality, we used the prime factorization
from Exercise 3.2 and the prime factorization 38 − 28 = (32 − 22 ) · (32 + 22 ) ·
(34 + 24 ) = 5 · 13 · 97.

Exercise 4.13. We have (2 880, 3 000, 3 240) = (120, 3 240) = 120 and [36, 42, 49]
= [252, 49] = 1 764.

Exercise 4.15. For example, the numbers a1 = 6, a2 = 10, a3 = 15 are rela-

tively prime, since ( a1 , a2 , a3 ) = (6, 10, 15) = (2, 15) = 1. The numbers a1 , a2 ,
a3 , however, are not pairwise relatively prime, since we have ( a1 , a2 ) = 2.

Exercise 4.17. Let a1 , . . . , an ∈ N. We leave to the reader the proof of the fol-
lowing equivalence:

( a1 , . . . , an ) · [ a1 , . . . , an ] = a1 · · · an ⇐⇒ a1 , . . . , an pairwise relatively prime.

This proves the desired criterion.

Exercise 5.2. We obtain 773 = 2 · 337 + 99. Further, we calculate 25 · 34 · 52 =

(22 · 32 ) · (23 · 32 · 52 ) = (5 · 7 + 1) · (23 · 32 · 52 ) = 7 · (23 · 32 · 53 ) + (23 · 32 · 52 ).
Since 216 + 1 = 48 + 1, it follows that 232 − 1 = (216 − 1)(48 + 1) + 0.
252 Solutions to Exercises

Exercise 5.4. This process can be carried out for arbitrary natural numbers
g > 1. One obtains the unique representation

n = q` · g` + q`−1 · g`−1 + · · · + q1 · g1 + q0 · g0

with natural numbers 0 ≤ q j ≤ g − 1 (j = 0, . . . , `) and q` 6= 0, called the g-adic

representation of the natural number n.

Solutions to Exercises in Chapter II

Exercise 1.3. For a, b, c ∈ Rn , one has the equalities

( a ⊕ b) ⊕ c = Rn ( a + b) ⊕ c = Rn (Rn ( a + b) + c),
a ⊕ (b ⊕ c) = a ⊕ Rn (b + c) = Rn ( a + Rn (b + c)).

Division with remainder of a + b and b + c by n yields the uniquely deter-

N
mined numbers q1 , q2 ∈ such that

a + b = q1 · n + R n ( a + b ) and b + c = q2 · n + R n ( b + c ),

whence follows

Rn (Rn ( a + b) + c) = Rn ( a + b + c − q1 · n) = Rn ( a + b + c)
= Rn ( a + b + c − q2 · n) = Rn ( a + Rn (b + c)).

We have thereby shown that the operation ⊕ is associative. Analogously,

one can prove that the operation  is associative.

N N
Exercise 1.4. (a) The set 2 · = {2 · n | n ∈ } of even natural numbers is a
N
nonempty subset of . If 2 · m, 2 · n ∈ 2 · , then N
2 · m + 2 · n = 2 · (m + n) ∈ 2 · N, (2 · m) · (2 · n) = 2 · (m · 2 · n) ∈ 2 · N.

N
Thus both + and · are operations on 2 · . Since the operations + on and N
N
· on are associative, it follows that in particular, the operations + on 2 · N
N N
and · on 2 · are associative. Therefore, both (2 · , +) and (2 · , ·) are N
semigroups.
N N
The set 2 · + 1 = {2 · n + 1 | n ∈ } of odd natural numbers is a
N
nonempty subset of . If 2 · m + 1 and 2 · n + 1 are in 2 · + 1, then N
(2 · m + 1) + (2 · n + 1) = 2 · (m + n + 1) ∈ 2 · N,
(2 · m + 1) · (2 · n + 1) = 2 · (m · 2 · n + m + n) + 1 ∈ 2 · N + 1.
Solutions to Exercises in Chapter II 253

N
Therefore, while · is an operation on 2 · + 1, we see that + is not an opera-
N N
tion on 2 · + 1. Therefore, (2 · + 1, +) is not a semigroup. The operation ·
N
on is associative, and so in particular, the operation · on 2 · + 1 is asso- N
N
ciative. Therefore, (2 · + 1, ·) is a semigroup.
N N N
(b) Let k ∈ , k > 1. The set k · = {k · n | n ∈ } is a nonempty subset of
N N N
. One shows, as in (a), that both (k · , +) and (k · , ·) are semigroups.

Exercise 1.5. Because of the inequality

(2 ◦ 3) ◦ 2 = (23 ) ◦ 2 = (23 )2 = 23·2 = 26 6= 29 = 2 ◦ (32 ) = 2 ◦ (3 ◦ 2),

the operation ◦ on N is not associative, and therefore (N, ◦) is not a semi-

group.

Exercise 1.8. If A1 = { a1 } is a one-element set, then

map( A1 ) = {id},

where the mapping id : A1 −→ A1 is given by the assignment a1 7→ a1 . The

semigroup (map( A1 ), ◦) is abelian. If A2 = { a1 , a2 , . . . } is an arbitrary set
that contains at least two elements a1 6= a2 , then

map( A2 ) = {id, f , g, . . . },

where the mapping id : A2 −→ A2 is given by a j 7→ a j (a j ∈ A2 ), the mapping

f : A2 −→ A2 by a1 7→ a2 , a j 7→ a j (a j ∈ A2 \ { a1 }), and the mapping g : A2 −→
A2 by a2 7→ a1 , a j 7→ a j (a j ∈ A2 \ { a2 }). But then we have

( f ◦ g)( a1 ) = f ( g( a1 )) = f ( a1 ) = a2 6= a1 = g( a2 ) = g( f ( a1 )) = ( g ◦ f )( a1 ),

whence (map( A2 ), ◦) is a nonabelian semigroup.

Exercise 1.12. Let e` be a left identity element and er a right identity element
of H. Then
e ` = e ` ◦ er = er ,
where the first equality follows from the fact that er is a right identity el-
ement of H, and the second from the fact that e` is a left identity element
of H.

N N
Exercise 1.14. (a) By Exercise 1.4, (2 · , +) and (2 · , ·) are semigroups.
It remains to show that there exists an additive identity element in 2 · . N
By the definition of addition, 0 is this element. Since 1 6∈ 2 · , there is noN
N
multiplicative identity element, so that (2 · , ·) is only a semigroup.
(b) We leave it to the reader to find other examples of semigroups that are
not monoids.
254 Solutions to Exercises

Exercise 2.3. (a) Suppose that g0 and g00 are two inverse elements to an ele-
ment g ∈ G. Then

g0 = g0 ◦ e = g0 ◦ ( g ◦ g00 ) = ( g0 ◦ g) ◦ g00 = e ◦ g00 = g00 ,

where the second equality follows from the fact that g00 is in particular a
right inverse to g, and the fourth equality follows from the fact that g0 is in
particular a left inverse to g.
(b) Let g`0 be a left inverse and gr0 a right inverse to an element g ∈ G. Then it
follows that

g`0 = g`0 ◦ e = g`0 ◦ ( g ◦ gr0 ) = ( g`0 ◦ g) ◦ gr0 = e ◦ gr0 = gr0 ,

analogously to part (a).

Exercise 2.6. (a) Let g−1 ∈ G be the inverse element to g ∈ G. Then

g ◦ g−1 = e = g−1 ◦ g.

Thus g is the inverse element to g−1 , that is, ( g−1 )−1 = g.

(b) Let g−1 ∈ G be the inverse element to g ∈ G and h−1 ∈ G the inverse
element to h ∈ G. Then

(h−1 ◦ g−1 ) ◦ ( g ◦ h) = h−1 ◦ ( g−1 ◦ g) ◦ h = h−1 ◦ e ◦ h = h−1 ◦ h = e,

( g ◦ h) ◦ (h−1 ◦ g−1 ) = g ◦ (h ◦ h−1 ) ◦ g−1 = g ◦ e ◦ g−1 = g ◦ g−1 = e.

Thus h−1 ◦ g−1 is the inverse element to g ◦ h, that is, ( g ◦ h)−1 = h−1 ◦ g−1 .
The calculational rules (c) and (d) follow directly from the definition.

Exercise 2.9. We compare only the groups that have the same numbers of el-
ements. The Cayley tables of the groups (R4 , ⊕), (R5 \ {0}, ), and ( D4 , ◦)
have, reading from left to right, the following form:

⊕ 0 1 2 3  1 2 3 4 ◦ d0 d1 s0 s1
0 0 1 2 3 1 1 2 3 4 d0 d0 d1 s0 s1
1 1 2 3 0 2 2 4 1 3 d1 d1 d0 s1 s0
2 2 3 0 1 3 3 1 4 2 s0 s0 s1 d0 d1
3 3 0 1 2 4 4 3 2 1 s1 s1 s0 d1 d0
One may conclude from the Cayley tables that all three groups under con-
sideration are abelian. We now determine the smallest nonzero natural num-
ber n such that gn = e for g 6= e. In the group (R4 , ⊕), we have e = 0 and

12 = 2, 13 = 3, 14 = 0; 22 = 0; 32 = 2, 33 = 1, 34 = 0.

In the group (R5 \ {0}, ), we have e = 1 and

Solutions to Exercises in Chapter II 255

22 = 4, 23 = 3, 24 = 1; 32 = 4, 33 = 2, 34 = 1; 42 = 1.

Thus in each group there are two elements with n = 4 and one element with
n = 2. In the group ( D4 , ◦), however, d21 = s20 = s21 = e with e = d0 , that is,
there is no element with n = 4.
The Cayley tables for (R6 , ⊕) and ( D6 , ◦), reading from left to right, have
the following form:

⊕ 0 1 2 3 4 5 ◦ d0 d1 d2 s0 s1 s2
0 0 1 2 3 4 5 d0 d0 d1 d2 s0 s1 s2
1 1 2 3 4 5 0 d1 d1 d2 d0 s2 s0 s1
2 2 3 4 5 0 1 d2 d2 d0 d1 s1 s2 s0
3 3 4 5 0 1 2 s0 s0 s1 s2 d0 d1 d2
4 4 5 0 1 2 3 s1 s1 s2 s0 d2 d0 d1
5 5 0 1 2 3 4 s2 s2 s0 s1 d1 d2 d0
From these tables, one can see that the group (R6 , ⊕) is abelian. The group
( D6 , ◦) is nonabelian, since s0 ◦ s1 = d2 6= d1 = s1 ◦ s0 . We again determine
the smallest nonzero natural number n such that gn = e for g 6= e. In (R6 , ⊕),
we have e = 0 and

12 = 2, 13 = 3, 14 = 4, 15 = 5, 16 = 0; 22 = 4, 23 = 0; 32 = 0;
42 = 2, 43 = 0; 52 = 4, 53 = 3, 54 = 2, 55 = 1, 56 = 0.

There exist, therefore, in (R6 , ⊕) two elements with n = 6, two elements

with n = 3, and one element with n = 2. In ( D6 , ◦), we have e = d0 and

d21 = d2 , d31 = d0 ; d22 = d1 , d32 = d0 ; s20 = d0 , s21 = d0 , s22 = d0 .

There exist, therefore, in ( D6 , ◦) no element with n = 6, two elements with

n = 3, and three elements with n = 2.

Exercise 2.10. (a) One shows using the Cayley table

12
1 12
2 21
for (R3 \ {0}, ) and the Cayley table from Exercise 2.9 for (R5 \ {0}, )
that (R3 \ {0}, ) and (R5 \ {0}, ) are groups.
(b) We leave to the reader the task of verifying the assertions of Exam-
ple 2.8 (iii) regarding the dihedral group ( D2n , ◦).
(c) Let n ≥ 3. We consider the elements

1 2 3 ··· n 1 2 3 ··· n
   
π1 = , π2 =
2 1 3 ··· n 3 1 2 ··· n
256 Solutions to Exercises

of Sn , where for n > 3, each of the elements 4, . . . , n is mapped to itself. Then

1 2 3 ··· n 1 2 3 ··· n
   
π1 ◦ π2 = 6= = π2 ◦ π1 ,
3 2 1 ··· n 1 3 2 ··· n

where for n > 3, each of the elements 4, . . . , n is mapped to itself. This proves
that (Sn , ◦) for n ≥ 3 is nonabelian.

Exercise 2.13. This is proved by induction on n.

Exercise 2.19. We have S3 = {π1 , π2 , π3 , π4 , π5 , π6 } with

123 123 12 3
     
π1 = , π2 = , π3 = ,
123 231 31 2
123 123 12 3
     
π4 = , π5 = , π6 = .
132 321 21 3

We calculate ord(π1 ) = 1, ord(π2 ) = ord(π3 ) = 3, and ord(π4 ) = ord(π5 ) =

ord(π6 ) = 2.

Exercise 2.23. Since d1k = dk (k = 0, . . . , n − 1) and d1n = d0 , we have hd1 i =

{d0 , . . . , dn−1 }. Using the subgroup criterion, one can show that the nonempty
subset h gi = {. . . , ( g−1 )2 , g−1 , g0 = e, g1 = g, g2 , . . . } ⊆ G is a subgroup of G
for every group G. In particular, hd1 i = {d0 , . . . , dn−1 } is a subgroup of D2n .

Exercise 2.26. We have S3 = {π1 , π2 , π3 , π4 , π5 , π6 } with π j (j = 1, . . . , n) as

in Exercise 2.19. We have the cyclic subgroups

h π1 i = { π1 }, h π2 i = { π1 , π2 , π3 } = h π3 i ,
h π4 i = { π1 , π4 }, h π5 i = { π1 , π5 }, h π6 i = { π1 , π6 } ,

and the subgroup S3 itself, which is not cyclic. One can see that S3 has no
other subgroups.

Exercise 3.3. We have S3 = {π1 , π2 , π3 , π4 , π5 , π6 } with π j (j = 1, . . . , n) as

in Exercise 2.19 and D6 = {d0 , d1 , d2 , d0 ◦ s0 , d1 ◦ s0 , d2 ◦ s0 }, where s0 is re-
flection in the median of the side joining vertices 1 and 2. By the definition
of the group homomorphism f : D6 −→ S3 , we have

f ( d0 ) = π1 , f ( d1 ) = π3 , f ( d2 ) = π2 ,
f ( d0 ◦ s0 ) = π6 , f ( d1 ◦ s0 ) = π5 , f ( d2 ◦ s0 ) = π4 ,

which proves that f is bijective and therefore in fact a group isomorphism.

k
Exercise 3.5. Let d j1 ◦ s01 and d j2 ◦ s0k2 with j1 , j2 ∈ {0, . . . , n − 1} and k1 , k2 ∈
1
{0, 1} be two elements of D2n . Since d j ◦ s0 = s0 ◦ d− j , we have
Solutions to Exercises in Chapter II 257




d j1 ◦ d j2 , if k1 = 0, k2 = 0;
d ◦ d j2 ◦ s0 , if k1 = 0, k2 = 1;

j1
(d j1 ◦ s0k1 ) ◦ (d j2 ◦ s0k2 ) = 1
d
 j1
 ◦ d−j2 , if k1 = 1, k2 = 1;
1
◦ d−

j2 ◦ s0 , if k1 = 1, k2 = 0.
d

j1

It follows that
k
k

sgn d j1 ◦ s01 ◦ d j2 ◦ s0k2 = k1 ⊕ k2 = sgn d j1 ◦ s01 ⊕ sgn d j2 ◦ s0k2 .

Therefore, sgn is a group homomorphism, and we have im(sgn) = R2 and

ker(sgn) = {d j | j = 0, . . . , n − 1}.

Exercise 3.7. By Lemma 3.6, we have f injective ⇐⇒ ker( f ) = {eG }. It there-

fore suffices to prove, under the assumption | G | < ∞, the equivalence f in-
jective ⇐⇒ f surjective. We have

f injective ⇐⇒ g 6= h for g, h ∈ G implies f ( g) 6= f (h)

⇐⇒ | f ( G )| = | G |
| G |<∞
⇐⇒ for every g ∈ G there exists h ∈ G with f (h) = g
⇐⇒ f surjective.

This proves the assertion.

Exercise 3.8. If g
∈ G and ord( g) = n, then e = f (e) = f ( gn ) = f ( g)n , which

implies ord f ( g) ≤ n = ord( g).

Exercise 3.9. Suppose that f : D24 −→ S4 is a group isomorphism. Then for

each g ∈ D24 , we must have the equality

ord( g) = ord( f ( g)).

Since ord(d2 ) = 12, we must have that f (d2 ) ∈ S4 is an element of order

12. We first determine the orders of the elements of S4 . We obtain the nine
elements of order 2,

1234 1234 1234

     
, , ,
2134 3214 4231
1234 1234 1234
     
, , ,
1324 1432 1243
1234 1234 1234
     
, , ,
2143 3412 4321

the eight elements of order 3,

258 Solutions to Exercises

1234 1234 1234 1234

       
, , , ,
2314 3124 2431 4132
1234 1234 1234 1234
       
, , , ,
3241 4213 1342 1423

and the six elements of order 4,

1234 1234 1234

     
, , ,
2341 2413 3421
1234 1234 1234
     
, , .
3142 4312 4123

The group S4 has, therefore, only elements of orders 1, 2, 3, and 4. There can
therefore be no group isomorphism between D24 and S4 .

Exercise 3.11. (a) If f : (R4 , ⊕) −→ (R4 , ⊕) is a group homomorphism, then

f (0) = 0. Since R4 = h1i, it follows that

f (2) = f (1 ⊕ 1) = f (1) ⊕ f (1), f (3) = f (1 ⊕ 1 ⊕ 1) = f (1) ⊕ f (1) ⊕ f (1),

that is, f is uniquely determined by specifying the image of f (1). Therefore,

there are precisely four distinct group homomorphisms, f 1 , f 2 , f 3 , f 4 , given
by the assignments

f 1 (0) = 0, f 1 (1) = 0, f 1 (2) = 0, f 1 (3) = 0,

f 2 (0) = 0, f 2 (1) = 1, f 2 (2) = 2, f 2 (3) = 3,
f 3 (0) = 0, f 3 (1) = 2, f 3 (2) = 0, f 3 (3) = 2,
f 4 (0) = 0, f 4 (1) = 3, f 4 (2) = 2, f 4 (3) = 1,

whence follows ker( f 1 ) = R4 , ker( f 2 ) = {0}, ker( f 3 ) = {0, 2}, ker( f 4 ) = {0},
im( f 1 ) = {0}, im( f 2 ) = R4 , im( f 3 ) = {0, 2}, im( f 4 ) = R4 . This shows in par-
ticular that f 2 and f 4 are bijective.
(b) Since R p = h1i, every group homomorphism g : (R p , ⊕) −→ (Rn , ⊕)
is uniquely determined by specifying the image g(1). One now shows that
only g(1) = 0 is possible, since n and p are relatively prime. There is, there-
fore, only one group homomorphism g, which is given by the assignment
g(m) = 0 (m ∈ R p ). We have ker( g) = R p and im( g) = {0}.

Exercise 4.3. (a) The verification of the statement of Example 4.2 is left to the
reader.
(b) The order relation ≤ is not an equivalence relation on , since ≤ is not N
symmetric.
N
(c) The relation ∼ is not an equivalence relation on , since ∼ is not transi-
tive.
Solutions to Exercises in Chapter II 259

Exercise 4.6. Let M be a set with an element m ∈ M. The equivalence class

of m with respect to equality “=” is Mm = {m0 ∈ M | m0 = m}, that is, the
set of all elements of M that are equal to m. We leave it to the reader to find
additional equivalence relations and determine the associated equivalence
classes.

Exercise 4.11. We leave the solution of this exercise to the reader.

Exercise 4.12. Let U = hπ4 i = {π1 , π4 }. The left coset of an element π ∈

S3 with respect to U is given by π ◦ U = {π ◦ π1 , π ◦ π4 }. Therefore, the
following are all left cosets of S3 with respect to U:

π1 ◦ U = U = π4 ◦ U,
π2 ◦ U = {π2 , π6 } = π6 ◦ U,
π3 ◦ U = {π3 , π5 } = π5 ◦ U.

Exercise 4.15. (a) If g ∈ G, then ord( g) = |U | for U = h gi ≤ G, which implies

by Lagrange’s theorem that ord( g) | | G |.
(b) Suppose | G | = p for a prime number p. If g ∈ G, g 6= e, then ord( g) > 1,
and therefore, by part (a), we must have the equality ord( g) = p, which im-
plies G = h gi.
(c) Let | G | = 4, and write G = {e, a, b, c}. If G has an element g ∈ G with
ord( g) = 4, then G = h gi, that is, G is cyclic, and therefore isomorphic to the
group (R4 , ⊕). If G has no element of order 4, then every element g ∈ G,
g 6= e, has order 2, that is, a2 = b2 = c2 = e. Since G is a group, it follows
that a ◦ b = c = b ◦ a, a ◦ c = b = c ◦ a, and b ◦ c = a = c ◦ b. Therefore, G is
isomorphic to the group ( D4 , ◦). Therefore, up to group isomorphism, there
are precisely two groups of order 4, given by the following Cayley tables:

◦ e a b c ◦ e a b c
e e a b c e e a b c
a a b c e a a e c b
b b c e a b b c e a
c c e a b c c b a e
Both groups are abelian.
One shows further that up to group isomorphism, the only groups of order
6 are (R6 , ⊕) and ( D6 , ◦). Thus every abelian group of order 6 is isomorphic
to (R6 , ⊕), and every nonabelian group of order 6 is isomorphic to ( D6 , ◦).

Exercise 4.19. Exercises 4.11 and 4.12 can be solved analogously for right
cosets.

Exercise 4.21. Let U = hπ4 i = {π1 , π4 }. Then U ◦ π2 = {π2 , π5 } 6= {π2 , π6 } =

π2 ◦ U, which implies that U is not a normal subgroup of S3 . Analogously,
one can show that hπ5 i and hπ6 i are not normal subgroups of S3 .
260 Solutions to Exercises

Exercise 4.24. (a) For every element h ∈ H, one has h ◦ H = H = H ◦ h. Now

let g ∈ G \ H. Then g ◦ H 6= H and H ◦ g 6= H. Because of [ G : H ] = 2, we
obtain the disjoint decomposition H ∪˙ ( g ◦ H ) = G = H ∪˙ ( H ◦ g) of G. It
then follows that g ◦ H = H ◦ g must hold. Altogether, one has the equality
g ◦ H = H ◦ g for all g ∈ G. This proves that H is a normal subgroup of G.
(b) The mapping f : G −→ R2 , given by
(
0, if g ∈ H;
f ( g) =
1, if g ∈
/ H,

is a surjective group homomorphism.

Exercise 4.26. Since ker( f ) is a subgroup, indeed a normal subgroup, of S3 ,

it must be the case that ker( f ) is one of the groups {π1 }, A3 , S3 . If ker( f ) =
{π1 }, then f is injective, which is impossible because of 6 = |S3 | > |R3 | = 3.
If ker( f ) = { A3 }, then ord(π4 ) < ord( f (π4 )), which is impossible because
of Exercise 3.8. The only possibility is then ker( f ) = S3 , which implies that
f (π ) = 0 for all π ∈ S3 .

Exercise 5.10. Let f : G −→ R2 be the surjective group homomorphism of

Exercise 4.24. Then ker( f ) = H, and by Corollary 5.8, we have an isomor-
phism G/H ∼ = R2 . This isomorphism can also be read off from the Cayley
table
• H g◦H
H H g◦H
g◦H g◦H H

for the group G/H = { H, g ◦ H }, where g ∈ G \ H is arbitrary, and H =

eG ◦ H. We note here that ( g ◦ H ) • ( g ◦ H ) = H must hold, since otherwise,
we would have from ( g ◦ H ) • ( g ◦ H ) = ( g ◦ g) ◦ H = g ◦ H the equalities
g ◦ g ◦ h1 = g ◦ h2 ⇐⇒ g ◦ h1 = h2 ⇐⇒ g = h2 ◦ h1−1 for certain h1 , h2 ∈ H, and
thus the contradiction g ∈ H.

Exercise 5.11. It is clear that from H E G, K E G, and K ⊆ H, one has K E H.

That H/K is a normal subgroup of G/K comes immediately from the proof
of isomorphism. To this end, we define the mapping f : G/K −→ G/H by
the assignment
g ◦ K 7→ g ◦ H.
Since K ⊆ H, it follows that f is well defined. On account of K E G and
H E G, it is clear that f is a homomorphism. Furthermore, we have

ker( f ) = { g ◦ K | f ( g ◦ K ) = H } = { g ◦ K | g ◦ H = H }
= { g ◦ K | g ∈ H } = H/K.
Solutions to Exercises in Chapter III 261

This proves in particular that H/K E G/K. Since f is surjective, it follows

from the homomorphism theorem that

( G/K )/( H/K ) = ( G/K )/ker( f ) ∼

= G/K,

as claimed.

Exercise 6.4. (a) Let A = { a1 , a2 , . . . } be a set with a1 6= a2 . Then map( A) =

{id, f , g, . . . }, whereby the mapping id : A −→ A is given by the assignment
a j 7→ a j (a j ∈ A), the mapping f : A −→ A by a1 7→ a2 , a j 7→ a j (a j ∈ A \ { a1 }),
and the mapping g : A −→ A by a2 7→ a1 , a j 7→ a j (a j ∈ A \ { a2 }). It then
follows that

( f ◦ g)( a1 ) = f ( a1 ) = a2 = ( f ◦ id)( a1 ), ( f ◦ g)( a2 ) = f ( a1 ) = a2 = ( f ◦ id)( a2 ),

which proves f ◦ g = f ◦ id; however, we have g 6= id, which proves that in

the semigroup (map( A), ◦), the first cancellation law is invalid.
One may show analogously that the second cancellation law is also in-
valid.
(b) We leave it to the reader to find further examples of semigroups that are
not regular.

Exercise 6.6. (a) The solution to this exercise is left to the reader.
N N N
(b) On (2 · + 1) × (2 · + 1) = {( a, b) | a, b ∈ 2 · + 1} we define the
relation

( a, b) ∼ (c, d) ⇐⇒ a · d = b · c ( a, b, c, d ∈ 2 · N + 1).

N
If we write ba for the equivalence class [ a, b] of ( a, b) ∈ (2 · + 1) × (2 · + 1), N
N N
then the group G := ((2 · + 1) × (2 · + 1))/ ∼ can be identified with the
N
set of all fractions of the form ba , where a, b ∈ 2 · + 1 and a, b are relatively
prime. We leave the detailed construction from Theorem 6.5 to the reader.

Z
Exercise 7.6. The generalization to of the addition and multiplication rules
in Remark 1.19 of Chapter I is left to the reader.

Exercise 7.9. The verification of the assertions of this example are left to the
reader.

Solutions to Exercises in Chapter III

Exercise 1.2. The proof of the calculational laws from Lemma 1.1 are left to
the reader.
262 Solutions to Exercises

Exercise 1.5. The proof of Theorem 1.4 is left to the reader.

Exercise 1.9. We give an idea of the proof. We assume that

a = e · p1 · · · pr = e · q1 · · · q s

for e ∈ {±1} and for prime numbers p1 , . . . , pr , q1 , . . . , qs (r ∈ , s ∈ ), not N N

necessarily distinct. Since now we have p1 | a and therefore p1 | e · q1 · · · qs , it
follows with the help of Euclid’s lemma, Lemma 1.7, that p1 | q j for some j =
1, . . . , s. Since p1 is prime, we must have p1 = q j . Without loss of generality,
we may assume (by renumbering if necessary) that p1 = q1 . Application of
the cancellation law implies the equality

p2 · · · pr = q2 · · · q s . (3)

Since p2 divides the left-hand side of (3), p2 must also divide the right-hand
side. As in the first step, we conclude that p2 = q2 . Proceeding in this way,
we obtain the equalities r = s and p j = q j for j = 1, . . . , r, which proves the
asserted uniqueness.

Exercise 2.5. We leave it to the reader to prove that the polynomial ring
( R[ X ], +, ·) is a commutative ring if and only if ( R, +, ·) is commutative.

Exercise 2.6. Let A be a nonempty set and ( R, + R , · R ) a ring. Then, 0R ∈ R,

and therefore, the mapping 0 : A −→ R, a 7→ 0R , is an element of map( A, R).
Hence, the set map( A, R) is not empty. We now show that + on map( A, R)
is associative. To this end, let f , g, h ∈ map( A, R). Then for all a ∈ A, we have

(( f + g) + h)( a) = ( f + g)( a) + R h( a) = ( f ( a) + R g( a)) + R h( a)

def of + def of +
= f ( a) + R ( g( a) + R h( a))
+ R associative,
since R is a ring

= f ( a) + R ( g + h)( a) = ( f + ( g + h))( a),

def of + def of +

which proves that + is associative. The mapping 0 : A −→ R is the identity

element with respect to +, since for all f ∈ map( A, R), we have the equality

(0 + f )( a) = 0( a ) + R f ( a ) = 0R + R f ( a) = f ( a)
def. of + def. of 0 0R id. el.
w.r.t. + R

and analogously the equality ( f + 0)( a) = f ( a) for all a ∈ A. The other ring
properties of map( A, R) are proved similarly using the ring properties of R.

Exercise 2.7. The solution of this exercise is left to the reader.

Solutions to Exercises in Chapter III 263

Exercise 2.11. If n > 1 is not prime, then there exist natural numbers a, b ∈
Rn , a > 1, b > 1, with a · b = n. But then we have a  b = 0. Therefore, a and
b are zero divisors of Rn .

Exercise 2.12. We show here only that the lack of zero divisors in ( R, +, ·)
implies the lack of zero divisors in ( R[ X ], +, ·). We assume that the ring R[ X ]
has zero divisors and show that then the ring R must have zero divisors. Let,
then, f ( X ) = an · X n + · · · + a1 · X + a0 (an 6= 0) and g( X ) = bm · X m + · · · +
b1 · X + b0 (bm 6= 0) be zero divisors in R[ X ], so that

f ( X ) · g( X ) = ( an · bm ) · X n+m + · · · + ( a1 · b0 + a0 · b1 ) · X + a0 · b0 = 0,

where 0 denotes the zero element of R[ X ], i.e., the zero polynomial. In par-
ticular, we must have an · bm = 0, which proves that R has zero divisors.

Exercise 2.13. The ring (map( A, R), +, ·) from Exercise 2.6 has as its zero
element 0 : A −→ R, a 7→ 0R . If now, for example, we have R = (R6 , ⊕, ),
then for f : A −→ R, a 7→ 2 (a ∈ A) and g : A −→ R, a 7→ 3 (a ∈ A), we have
the equality

( f · g)( a) = f ( a)  g( a) = 2  3 = 0 = 0( a ),
def of · def of f , g

that is, f and g are zero divisors of (map( A, R), +, ·). Therefore, the ring
(map( A, R), +, ·) also has zero divisors if R has zero divisors. We note that
the ring (map( A, R), +, ·) can possess zero divisors even if R has no zero
divisors.

Z
Exercise 2.19. In the polynomial ring ( [ X ], +, ·), the unit element is given
by the unit polynomial 1. Let f ( X ) = an · X n + · · · + a1 · X + a0 (an 6= 0) be a
Z
unit in [ X ]. Then there exists a polynomial g( X ) = bm · X m + · · · + b1 · X +
Z
b0 (bm 6= 0) in [ X ] such that

f ( X ) · g( X ) = ( an · bm ) · X n+m + · · · + ( a1 · b0 + a0 · b1 ) · X + a0 · b0 = 1.

If n > 0, then we must have in particular that an · bm = 0, in contradiction

Z
to the fact that has no zero divisors. Therefore, f and hence g must be
of the form f ( X ) = a0 and g( X ) = b0 . The equality a0 · b0 = 1 shows that f
and g are units if and only if a0 ∈ {1, −1} and a0 = b0 . The polynomial ring
Z
( [ X ], +, ·) has therefore only the units {1, −1}.

Exercise 2.20. That the units of a ring ( R, +, ·) with unit element 1 form a
group under multiplication with multiplicative identity element 1 follows
directly from how units are defined.

Exercise 2.21. The group of units of R5 is (R5 \ {0}, ) and is therefore

isomorphic to the group (R4 , ⊕). The group of units of R8 is ({1, 3, 5, 7}, ).
264 Solutions to Exercises

We have 3  3 = 1, 5  5 = 1, 7  7 = 1, which shows that this group is

isomorphic to ( D4 , ◦). The group of units of R10 is ({1, 3, 7, 9}, ). We have
3  7 = 1, 9  9 = 1, which shows that this group is isomorphic to (R4 , ⊕)
and therefore to (R5 \ {0}, ). The group of units of R12 is ({1, 5, 7, 11}, ).
We have 5  5 = 1, 7  7 = 1, 11  11 = 1, which shows that this group is
isomorphic to ( D4 , ◦) and therefore to ({1, 3, 5, 7}, ).

Exercise 2.25. For n ∈ N, we have that (nZ, +, ·) is a subring of (Z, +, ·).

Exercise 2.26. We leave it to the reader to show that ( R, +, ·) is a subring of
the polynomial ring ( R[ X ], +, ·).

Exercise 3.2. (a) The mapping f 1 is a ring homomorphism.

(b) The mapping f 2 is not a ring homomorphism, since for g1 ( X ) = X and
g2 ( X ) = 1, we have f 2 ( g1 ( X ) · g2 ( X )) = f 2 (1 · X ) = 1 6= 0 = 1 · 0 = f 2 ( X ) ·
f 2 (1) = f 2 ( g1 ( X )) · f 2 ( g2 ( X )).
(c) The mapping f 3 is a ring homomorphism if and only if r = 0.
(d) The mapping f 4 is a ring homomorphism.
(e) The mapping f 5 is a ring homomorphism.

Exercise 3.5. The proof of Lemma 3.4 is left to the reader.

Exercise 3.6. We obtain

n o
ker( f 1 ) = ∑ a j · X j  a j ∈ R, a0 = 0 , im( f 1 ) = R .

j∈ N
The mapping f 2 is not a ring homomorphism. Furthermore, we have

ker( f 3 ) = map( A, R), im( f 3 ) = {0} (if r = 0) ;

ker( f 4 ) = { g ∈ map( A, R) | g( a) = 0}, im( f 4 ) = { g( a) | g ∈ map( A, R)} ;
ker( f 5 ) = { f ( X ) ∈ R[ X ] | f (r ) = 0}, im( f 5 ) = { f (r ) | f ( X ) ∈ R[ X ]} = R,

for the ring homomorphisms under consideration.

Exercise 3.13. To prove the equality a = R, we must show that a ⊇ R. To this

end, let r ∈ R. Since 1 ∈ a and a is an ideal, it follows that r · 1 = r ∈ a. This
proves that R ⊆ a.

Z
Exercise 3.14. No, since for every subring U of ( , +, ·), we have in particu-
Z
lar that (U, +) is a subgroup of ( , +). Therefore, we must have U = n for Z
N
some n ∈ , which proves that (U, +) is an ideal of . Z
Exercise 3.15. For example, Z is a subring of the polynomial ring (Z[X], +, ·)
Z
that is not an ideal of [ X ].
Solutions to Exercises in Chapter III 265

Z Z
Exercise 3.16. The principal ideals of [ X ] are of the form { h · f | f ∈ [ X ]}
Z
for some h ∈ [ X ]. We leave it to the reader to show that the ideal

Z
a := {2 · f + X · g | f , g ∈ [ X ]}

Z
is not a principal ideal of ( [ X ], +, ·).

Exercise 3.18. We have ker( f 1 ) = ( X ) and ker( f 5 ) = ( X − r ). If the ring R

possesses a unit element, then ker( f 3 ) = (1).

Z Z Z
Exercise 3.27. Let a ∈ . Then the mapping f : [ X ] −→ given by the as-
signment f ( X ) 7→ f ( a) is a surjective ring homomorphism with ker( f ) =
( X − a) by Exercises 3.2, 3.6, and 3.18. Invoking Corollary 3.25, we see that
Z [ X ]/( X − a) ∼ Z
= is a ring isomorphism, as desired.

Exercise 3.28. An analogy to the group isomorphism from Exercise 5.11 of

Chapter II can also be formulated and proved for rings by replacing “sub-
group” and “normal subgroup” throughout with “ideal” and observing that
the group homomorphisms that arise are also ring homomorphisms. We
leave this task to the reader.

Exercise 4.4. This is impossible, since one can show that every skew field
with finitely many elements is in fact a field.

Exercise 5.3. The proofs of associativity of ⊕, commutativity of , and the

second distributive law are left to the reader.

Exercise 5.5. We consider the ring homomorphism f : K −→ Quot(K ) given

by a 7→ [ a, 1]. The ring homomorphism f is well defined, since K is a field,
and therefore, we have 1 ∈ K. Moreover, f is injective, which can be shown
as follows: Assume that [ a, 1] = [b, 1] for a, b ∈ K. Then ( a, 1) ∼ (b, 1), which
implies a · 1 = 1 · b and hence a = b. We now show that f is also surjective.
To this end, let [ a, b] ∈ Quot(K ) with a, b ∈ K, b 6= 0, be an arbitrary element.
Since K is a field, there exists an inverse b−1 ∈ K to b. Therefore, we have
a · b−1 ∈ K and
f ( a · b−1 ) = [ a · b−1 , 1] = [ a, b],
where the last equality follows from ( a · b−1 , 1) ∼ ( a, b) ⇐⇒ a · b−1 · b =
1 · a ⇐⇒ a = a. Therefore, f is a ring isomorphism, and we have the ring
isomorphism K ∼ = Quot(K ).

Z N
Exercise 6.3. Let r = s/t for s ∈ and t ∈ \ {0}. Then r corresponds to the
Z Z
element [s, t] ∈ × ( \ {0}). If now d = (s, t) > 0 is the greatest common
divisor of s and t, then we may write s = d · a and t = d · b with a ∈ , Z
N
b ∈ \ {0}, and a, b relatively prime. From the equality s · b = (d · a) · b =
(d · b) · a = t · a we infer the equality [s, t] = [ a, b]. This proves the existence
266 Solutions to Exercises

of the claimed representative. To prove the uniqueness of the representative,

Z N
we assume that [c, d] with c ∈ , d ∈ \ {0}, and c, d relatively prime is also
an element such that [s, t] = [c, d]. Then in particular, we have [ a, b] = [c, d],
and therefore
a · d = b · c.
If p is a prime number such that p | a, then we must have p | b · c and conse-
quently, since p is prime, p | b or p | c. But since a and b are relatively prime,
we must have p | c. Conversely, one can show that for an arbitrary prime p
with p | c, we must also have p | a. From this it follows that a | c and also c | a,
which, since a and c have the same sign, proves the equality a = c. We con-
clude by an analogous argument the equality b = d, completing the proof of
uniqueness.

Exercise 6.4. Since one usually learns this proof in a first course in real anal-
ysis, we leave this exercise to the reader.

Exercise 6.7. The generalizations of the addition and multiplication rules

from Remark 1.19 of Chapter I are left to the reader.

Z
Exercise 7.4. We obtain in [ X ] the decompositions 20X = 22 · 5 · X and
10X 2 + 4X − 6 = 2 · ( X + 1) · (5X − 3). Therefore, 2 is the greatest common
divisor, and 22 · 5 · X · ( X + 1) · (5X − 3) = 100X 3 + 40X 2 − 60X the least
common multiple, of the polynomials 20X and 10X 2 + 4X − 6 in [ X ]. Z
Exercise 7.9. We leave it to the reader to come up with relevant examples.

Exercise 7.15. Part (ii) of the proof of Lemma 7.14 is left to the reader.

Exercise 7.25. For example, the polynomial ring K [ X, Y ] in two variables

over a field K is not a principal ideal domain, since the ideal

a := { X · f + Y · g | f , g ∈ K [ X, Y ]}

is not principal.

Exercise 7.38. (a) We obtain (123 456 789, 555 555 555) = 9.
(b) We calculate

X 4 + 2X 3 + 2X 2 + 2X + 1 = ( X + 1) · ( X 3 + X 2 − X − 1) + (2X 2 + 4X + 2) ,
1 1
X3 + X2 − X − 1 = X− · (2X 2 + 4X + 2) + 0 .
2 2
We thereby obtain ( X 4 + 2X 3 + 2X 2 + 2X + 1, X 3 + X 2 − X − 1) = 2X 2 +
4X + 2.
Solutions to Exercises in Chapter IV 267

Solutions to Exercises in Chapter IV

Exercise 1.6. (a) We obtain the decimal expansions 15 = 0.2, 13 = 0.3, 16

1
=
1 1
0.0625, 11 = 0.09, and 7 = 0.142857.
Z
(b) One shows that a reduced fraction ba (a, b ∈ ; b 6= 0) has a terminating
decimal expansion if and only if b = 2k · 5l with k, l ∈ . N
N
(c) We consider the fraction m1 for m ∈ , m 6= 0, with 2 - m and 5 - m. One
shows that m − 1 is the maximal period length of the decimal expansion
of the fraction m1 , considering that there can be at most m − 1 remainders
on division by m. The fraction 17 , for example, has a period length that is
maximal.
(d) Without loss of generality, we may assume that the periodic decimal
fraction has the form

0. q−1 . . . q−v q−(v+1) . . . q−(v+ p)

with natural numbers v ≥ 0, p > 0. Then

p p− j
a ∑vj=1 q− j 10v− j 1 ∑ j=1 q−(v+ j) 10
= + v· .
b 10v 10 10 p − 1

For example, for 0.123, one obtains the fraction

123 123 41
= = .
103 − 1 999 333

Exercise 2.2. (a) Without loss of generality, we may assume that e ∈ Q, 0 <
N
e < 1. For n ∈ , we then have
 1  1−e
 
 n + 1  < e ⇐⇒ e < n.
 

If we set N (e) := [(1 − e)/e], where [ x ] is the greatest integer less than or
N
equal to x, then for all n ∈ with n > N (e), we have the inequality
 1 
 
 n + 1  < e.
 

 
1
This proves that the sequence n+ 1 n≥0 is a rational null sequence. Using
the inequality
n 1
<
2n n+1
N
for n ∈ , n ≥ 5, one shows analogously that 2nn n≥0 is also a rational null

sequence.
268 Solutions to Exercises
 
1
(b) Other examples of rational null sequences are the sequences
N
with k ∈ , k ≥ 2.
( n +1) k n ≥0

Exercise 2.18. Let α = ( an ) + n, β = (bn ) + n be two real numbers with α <

Q
β, that is, there exist q ∈ , q > 0, N (q) ∈ Nwith bn − an > q for all n ∈
N with n > N (q). To prove the asserted independence of the choice of the
representing rational Cauchy sequences, we assume that ( an ) + n = ( a0n ) + n
and (bn ) + n = (bn0 ) + n for rational Cauchy sequences ( a0n ) and (bn0 ). Then
in particular, we must have ( a0n ) = ( an ) + (cn ) = ( an + cn ) and (bn0 ) = (bn ) +
(dn ) = (bn + dn ) for rational null sequences (cn ) and (dn ). This yields

bn0 − a0n = (bn − an ) + (dn − cn ) ≥ (bn − an ) − |dn − cn |. (4)

Since, moreover, (dn − cn ) is a rational null sequence, there exists for e :=

Q
q/2 ∈ an N N N
e (e) ∈ such that for all n ∈ with n > N e (e), we have the
Q
inequality |dn − cn | < e. So if we set q0 := q − e = q/2 ∈ , we have q0 > 0,
and with (4) we obtain the inequality

bn0 − a0n > q − e = q0

N
for all n ∈ with n > N (q0 ) := max( N (q), N
e (e)). This proves the claimed
independence of representative.

Exercise 2.22. We leave the proof of Lemma 2.21 to the reader.

Exercise 2.25. Real null sequences whose

 √ elements are irrational numbers
N

2
include, for example, the sequences (n+1)k with k ∈ , k ≥ 1.
n ≥0

Exercise 2.30. We begin by considering that for a rational number a0 ∈ Q,

a0 > 0, we have the following equivalences with an error of δ0 :
√
a0 + δ0 = 2 ⇐⇒ ( a0 + δ0 )2 = 2 ⇐⇒ 2a0 δ0 + δ02 = 2 − a20
2 − a20 δ2
⇐⇒ δ0 = − 0 .
2a0 2a0

So if we set
2 − a20 2 + a20
a1 : = a0 + = ,
2a0 2a0
√
then because a0 > 0, we have the inequality a21 > 2, that is, a1 > 2. This
implies 2 − a21 /(2a1 ) < 0, and we therefore have for

2 − a21 2 + a21
a2 : = a1 + =
2a1 2a1
Solutions to Exercises in Chapter IV 269
√
both a1 > a2 and a22 > 2, that is, a2 > 2. We consider now the rational se-
Q
quence ( an )n≥0 with a0 ∈ , a0 > 0, arbitrary and

2 + a2n
a n +1 : = (n ∈ N, n ≥ 1). (5)
2an
One first shows by induction
√ that both an > an+1 for all n ∈ , n ≥ 1, and N
2 N
an > 2, that is, an > 2, for all n ∈ , n ≥ 1. Using these inequalities, one
then shows in a second step that ( an )n≥0 is a rational Cauchy sequence. This
R
rational Cauchy sequence has limit α ∈ , α > 0. Because of the recurrence
formula (5), we see that α satisfies the equation

2 + α2 √
α= ⇐⇒ α = 2,
2α
as desired.

Exercise 3.12. (a) The decimal representation 0.101 001 000 100 001 . . . is nei-
ther terminating nor periodic. Therefore, this number cannot be rational.
One can find analogous examples, such as 0.121 331 222 133 331 . . . .
(b) Using the rational
√ Cauchy √series ( an )n≥0 constructed in Exercise 2.30
for calculating 2, one obtains 2 ≈ 1.4142135623, accurate to ten decimal
places, by choosing, for example, a0 = 1 and iterating four times.

Exercise 4.2. We consider, for example, the sequence

n2 + 2
 
( a n ) n ≥0 : = .
2n n ≥0

We have a0 = 2 and a1 = a2 = 3/2. For n ∈ N, n ≥ 3, one shows the inequality

2(n2 + 2) > (n + 1)2 + 2. Since

n2 + 2 ( n + 1)2 + 2
2(n2 + 2) > (n + 1)2 + 2 ⇐⇒ n
> ⇐⇒ an > an+1
2 2n +1
N
for n ∈ , n ≥ 3, it follows that the sequence ( an )n≥0 is monotonically de-
creasing, but not strictly. One shows analogously that the sequences

n3 + 3
 
1
 
12 n+1 ,
n ≥0 3n n ≥0

are strictly monotonically decreasing and that the sequence

n3 − 2
 

n2 − 2 n ≥0
270 Solutions to Exercises

is monotonically increasing. The sequence

 1 
n n +1
n ≥0

is neither monotonically increasing nor monotonically decreasing. How-

ever, the sequence
 1 
n n +1
n ≥4

is strictly monotonically decreasing.

Q
Exercise 4.6. The subset M ⊆ consisting of all sequence terms an (n > 0)
in the rational Cauchy sequence ( an )n≥0 constructed
√ in Exercise 2.30, which
is bounded below, has greatest lower bound 2 ∈ / . Q
√
Exercise 4.7. The greatest lower bound of the set { x x | x ∈ , x √
≥ 0} is at- Q
tained when x = 0 and is equal to zero. The least upper bound is e e.

Exercise 5.3. The solution of this exercise is left to the reader.

Exercise 6.7. We leave the completion of the details in the sketch of the proof
of Theorem 6.5 to the reader.

Solutions to Exercises in Chapter V

Exercise 1.1. The solution of this exercise is left to the reader.

Exercise 1.8. The completion of the proof of Theorem 1.7 is left to the reader.

C √
Exercise 1.10. Let α = α1 + α2 i ∈ , α 6= 0. If α2 = 0 and α1 > 0, then ± α1 are
the solutions to the equation x2 = α. If α2 = 0 and α1 < 0, then ± |α1 |i are
p

the solutions to the equation x2 = α. It remains to consider the case α2 6= 0.

C
Let β = β 1 + β 2 i ∈ with β 1 6= 0. We then have the equivalence

β2 = α ⇐⇒ ( β21 − β22 ) + (2β 1 β 2 )i = α1 + α2 i ⇐⇒ β21 − β22 = α1 , 2β 1 β 2 = α2 .

If we substitute the second equation, β 2 = α2 /(2β 1 ), in the first equation,

we obtain the equation 4β41 − 4α1 β21 − α22 = 0. Setting y := β21 , we obtain the
quadratic equation 4y2 − 4α1 y − α22 = 0, which has the solutions
q
α1 ± α21 + α22 α1 ± | α |
y1,2 = = .
2 2
Solutions to Exercises in Chapter V 271

√
R
Since now β 1 ∈ , we need consider only the nonnegative solution y1 . With
√
β 1 = ± y1 and β 2 = ±α2 /(2 y1 ), we obtain the solution formula

α2 i
p
α1 + | α |
β=± √ ±q
.
2 2 α1 + | α |

Altogether, we obtain for the solutions to the equation x2 = α the following

solution formula:
 √

 ± α1 , if α1 > 0, α2 = 0;

 ± | α1 | i , if α1 < 0, α2 = 0;

 p

x1,2 = q
|α|+α1
q 
±

 2 + |α|−
2
α1
i , if α2 > 0;
  q q 
|α|+α1
− |α|− α1

± i , if α2 < 0.

2 2

From this it follows that the solutions to the equation x2 = i are

1+i
x1,2 = ± √ ,
2

those to the equation x2 = 2 + i are

 p√ p√
5+2 5−2

x1,2 = ± √ + √ i ,
2 2

and those to the equation x2 = 3 − 2i are

 p√ p√
13 + 3 13 − 3

x1,2 = ± √ − √ i .
2 2

Exercise 1.11. Since by Exercise 1.10 we have the equality ((1 + i )/2)2 = i/2,
we obtain, on completing the square,
 1 + i 2 i
x2 + (1 + i ) · x + i = 0 ⇐⇒ x + + = 0.
2 2
If we now substitute y := x + (1 + i )/2, we obtain the quadratic equation
y2 = −i/2. With the solution formula from Exercise 1.10, we obtain the so-
lution
1+i 1 i 1+i
x1,2 = y1,2 − =± ∓ − ,
2 2 2 2
that is, x1 = −i and x2 = −1 are the solutions of x2 + (1 + i ) · x + i = 0, which
can be easily checked by substitution. One can prove analogously that the
equation x2 + (2 − i ) · x − 2i = 0 has the solutions x1 = i and x2 = −2.
272 Solutions to Exercises

Exercise 1.14. (a) First, one verifies the equality α · β = α · β for all α, β ∈ C.
One then has

| α · β |2 = ( α · β ) · α · β = α · α · β · β = | α |2 · | β |2 ,

which proves the assertion.

N
(b) Let α1 , α2 , β 1 , β 2 ∈ . The product rule from part (a) with α = α1 + α2 i
and β = β 1 + β 2 i yields

(α21 + α22 ) · ( β21 + β22 ) = (α1 β 1 − α2 β 2 )2 + (α1 β 2 + α2 β 1 )2 .

This implies the assertion.

Exercise 2.3. The statement of Remark 2.2 results immediately from use of
the product rule from Exercise 1.14.

R
Exercise 2.5. Let A ∈ M2 ( ) be an invertible matrix. One may convince one-
C R
self that the mapping f A : ( , +, ·) −→ (M2 ( ), +, ·) given by
 
α1 α2
α = α1 + α2 i 7 → A · · A −1
− α2 α1

is an injective ring homomorphism. In particular, f induces an isomorphism

C ∼ C
= im( f ), that is, is isomorphic to the subring im( f ) of M2 ( ). R
Exercise 2.8. The solution of this exercise is left to the reader.

Exercise 2.12. We first observe that for two complex numbers α, β ∈

C \ {0} with polar coordinate representations α = |α| · (cos( ϕ) + i sin( ϕ))
and β = | β| · (cos(ψ) + i sin(ψ)), one has the following multiplication for-
mula:

α · β = |α|| β| · cos( ϕ) cos(ψ) − sin( ϕ) sin(ψ)

+ i sin( ϕ) cos(ψ) + cos( ϕ) sin(ψ)

= |α|| β| · cos( ϕ + ψ) + i sin( ϕ + ψ) ,

where for the second equality we have invoked the addition theorems for
sine and cosine. From this follows for m ∈ the equalityN
αm = |α|m · cos(mϕ) + i sin(mϕ) .

Likewise, for n ∈ N with n 6= 0, we obtain the equality

1 1
  ϕ  ϕ 
α n = |α| n · cos + i sin .
n n
This completes the proof of the general formula.
Solutions to Exercises in Chapter VI 273

Exercise 4.4. Let p be a prime number. Then f ( X ) = X 2 − p is a quadratic

√
polynomial with integer coefficients, and we have f ( p) = 0. We now as-
sume that there exists a linear polynomial g( X ) = aX + b (a, b ∈ , a 6= 0) Z
√ √
with g( p) = 0. But then we must have p = −b/a ∈ , a contradiction.
√
Q
We have therefore shown that p is algebraic of degree 2.

Exercise 4.12. We leave it to the reader to find additional transcendental

numbers following the pattern of the Liouville number.

Exercise 5.8. The calculation of better approximations to e is left to the

reader.

Solutions to Exercises in Chapter VI

Exercise 1.6. For example, the quadratic polynomial X 2 + 1 ∈ [ X ] has zeros H

±i, ± j, ±k and every purely imaginary quaternion α2 i + α3 j + α4 k ∈ Im( ), H
that satisfies the condition α22 + α23 + α24 = 1.

Exercise 1.7. It is clear that R H

⊆ Z ( ). We therefore have to show that
H R H
Z ( ) ⊆ . To this end, let α = α1 + α2 i + α3 j + α4 k ∈ Z ( ). For each β =
H
β 1 + β 2 i + β 3 j + β 4 k ∈ , we then have α · β = β · α. Since

α · β = ( α1 β 1 − α2 β 2 − α3 β 3 − α4 β 4 ) + ( α1 β 2 + α2 β 1 + α3 β 4 − α4 β 3 ) i
+ ( α1 β 3 − α2 β 4 + α3 β 1 + α4 β 2 ) j + ( α1 β 4 + α2 β 3 − α3 β 2 + α4 β 1 ) k

and

β · α = ( β 1 α1 − β 2 α2 − β 3 α3 − β 4 α4 ) + ( β 1 α2 + β 2 α1 + β 3 α4 − β 4 α3 ) i
+ ( β 1 α3 − β 2 α4 + β 3 α1 + β 4 α2 ) j + ( β 1 α4 + β 2 α3 − β 3 α2 + β 4 α1 )k,

however, we have

α · β = β · α ⇐⇒ 2(α3 β 4 − α4 β 3 )i + 2(−α2 β 4 + α4 β 2 ) j + 2(α2 β 3 − α3 β 2 )k = 0

⇐⇒ α3 β 4 = α4 β 3 ∧ α4 β 2 = α2 β 4 ∧ α2 β 3 = α3 β 2 .

If α2 6= 0, then from the third equality, it follows that β 3 = (α3 α2−1 ) · β 2 for
H
every β ∈ . This contradiction implies that we must have α2 = 0. Similarly,
one shows that we must also have α3 = 0 and α4 = 0. Altogether, therefore,
R
we have that α = α1 ∈ . This proves the inclusion Z ( ) ⊆ . H R
Exercise 1.14. (a) Let α = α1 + α2 i + α3 j + α4 k. We then calculate
274 Solutions to Exercises

α2 = (α21 − α22 − α23 − α24 ) + 2α1 α2 i + 2α1 α3 j + 2α1 α4 k

= −(α21 + α22 + α23 + α24 ) + 2α1 α,

which yields the desired result.

(b) This can be established by a direct calculation.

Exercise 1.15. Let α = Im(α) · i with Im(α) = (α2 , α3 , α4 ) and β = Im( β) · i

with Im( β) = ( β 2 , β 3 , β 4 ). We calculate

α · β = ( α2 i + α3 j + α4 k ) · ( β 2 i + β 3 j + β 4 k )
= (−α2 β 2 − α3 β 3 − α4 β 4 ) + (α3 β 4 − α4 β 3 )i
+ (−α2 β 4 + α4 β 2 ) j + (α2 β 3 − α3 β 2 )k
= − Im(α)t , Im( β)t + Im(α)t × Im( β)t · i,



which proves the assertion.

Exercise 1.18. Let α, β ∈ H. We calculate

2 · hα, βi = 2 · Re(α · β) = α · β + α · β = α · β + β · α.

Multiplication of this equality on the right by β yields

2 · hα, βi · β = α · β · β + β · α · β,

which on account of β · β = h β, βi ∈ R proves the result.

Exercise 1.19. (a) The equality α · β = β · α can be verified by a direct calcu-
lation.
(b) Using part (a), one obtains

| α · β |2 = ( α · β ) · α · β = α · ( β · β ) · α = | β |2 · ( α · α ) = | β |2 · | α |2 = | α |2 · | β |2 ,

which proves the assertion.

(c) Let α1 , α2 , α3 , α4 , β 1 , β 2 , β 3 , β 4 ∈ N. Using the product rule from part (a),
we obtain

(α21 + α22 + α23 + α24 ) · ( β21 + β22 + β23 + β24 )

= ( α1 β 1 − α2 β 2 − α3 β 3 − α4 β 4 )2 + ( α1 β 2 + α2 β 1 + α3 β 4 − α4 β 3 )2
+ ( α1 β 3 + α3 β 1 + α4 β 2 − α2 β 4 )2 + ( α1 β 4 + α4 β 1 + α2 β 3 − α3 β 2 )2 .

This implies the assertion.

Exercise 1.21. The completion of the proof of Theorem 1.20 is left to the
reader.
Solutions to Exercises in Chapter VI 275

Exercise 1.25. Verification of the assertions of this problem is left to the

reader.

Exercise 1.27. The solution to this problem is left to the reader.

Exercise 2.3. The statement of Remark 2.2 is an immediate result of the prod-
uct rule from Exercise 1.19.

C
Exercise 2.5. Let A ∈ M2 ( ) be an arbitrary invertible matrix. One may con-
H C
vince oneself that the mapping f : ( , +, ·) −→ (M2 ( ), +, ·) given by

α1 + α2 i α3 + α4 i
 
α = α1 + α2 i + α3 j + α4 k 7 → A · · A −1
− α3 + α4 i α1 − α2 i

is an injective ring homomorphism. In particular, f induces an isomorphism

H ∼ H
= im( f ), that is, is isomorphic to the subring im( f ) of M2 ( ). C
R
Exercise 2.6. We show that f is an -linear mapping. Let α = α1 + α2 i +
H H
α3 j + α4 k ∈ and β = β 1 + β 2 i + β 3 j + β 4 k ∈ . For arbitrary µ, ν ∈ , one R
has

f (µα + νβ) = f ((µα1 + νβ 1 ) + (µα2 + νβ 2 )i + (µα3 + νβ 3 ) j + (µα4 + νβ 4 )k)

(µα1 + νβ 1 ) + (µα2 + νβ 2 )i (µα3 + νβ 3 ) + (µα4 + νβ 4 )i
 
=
−(µα3 + νβ 3 ) + (µα4 + νβ 4 )i (µα1 + νβ 1 ) − (µα2 + νβ 2 )i
α1 + α2 i α3 + α4 i β1 + β2 i β3 + β4 i
   
=µ +ν
− α3 + α4 i α1 − α2 i − β3 + β4 i β1 − β2 i
= µ f ( α ) + ν f ( β ),

that is, f is R-linear. The verification of the remaining assertions are left to
the reader.

Exercise 2.9. The solution to this problem is left to the reader.

Exercise 3.4. The solution to this problem is left to the reader.

R
Exercise 3.6. We first prove that every -linear mapping v 7→ A · v (v ∈ 3 ) R
R R
with A ∈ SO3 ( ) is an orientation-preserving rotation of 3 about an axis
passing through the origin. To this end, we begin by noting that for A ∈
R
SO3 ( ), on account of

det( A − E) = 1 · det( A − E) = det( At ) · det( A − E)

= det( At · ( A − E)) = det( E − At ) = det( E − A)t
= det( E − A) = (−1) · det( A − E),
276 Solutions to Exercises

we have the equality det( A − E) = 0, which proves that 1 is an eigenvalue

of A. Now let a1 denote a normalized (to have length 1) eigenvector of A
associated with the eigenvalue 1. We shall see that the mapping v 7→ A · v
R
(v ∈ 3 ) describes a rotation about an axis passing through the origin that
is determined by a1 . For this, we extend a1 to an orthonormal basis of 3 R
R
by choosing an additional vector a2 ∈ 3 , normed to have length 1, that is
R
perpendicular to a1 , and then setting a3 := a1 × a2 . If now S ∈ M2 ( ) denotes
a matrix with S · (1, 0, 0)t = a1 , S · (0, 1, 0)t = a2 , and S · (0, 0, 1)t = a3 , then we
R
must have S ∈ SO3 ( ). We further obtain

S−1 · A · S · (1, 0, 0)t = S−1 · A · a1 = S−1 · a1 = (1, 0, 0)t ,

which implies the equality

10 0
 

S −1 · A · S = 0 α β 
0γδ

R R
for certain α, β, γ, δ ∈ . But since because of A, S ∈ SO3 ( ), we have also
R
the inclusion S−1 · A · S ∈ SO3 ( ) and thereby

R
 
αβ
∈ SO2 ( ),
γδ

it follows that there exists a uniquely determined ϕ ∈ [0, 2π ) with

1 0 0
 

S−1 · A · S = 0 cos( ϕ) − sin( ϕ) =: D ϕ .

0 sin( ϕ) cos( ϕ)

R
The mapping v 7→ D ϕ · v (v ∈ 3 ) is an orientation-preserving rotation in the
x2 , x3 -plane through the angle ϕ about the x1 -axis (counterclockwise if a1
points toward the observer). Altogether, we have shown that the mapping
R
v 7→ A · v (v ∈ 3 ) is an orientation-preserving rotation in the a2 , a3 -plane
through the angle ϕ about the a1 -axis (counterclockwise if a1 points toward
the observer).
We now describe, conversely, an arbitrary orientation-preserving rotation
R
of 3 through the angle ϕ ∈ [0, 2π ) about an axis passing through the origin
that is determined by the vector (normed to have length 1) a1 := (ν1 , ν2 , ν3 )t ∈
R 3 . To this end, we first consider the matrices

√ ν21 2 0 √ ν23 2
  q 
ν + ν ν + ν ν12 + ν32 ν2 0
R
1 3 1 3
D1 :=  0 1 0  , D2 :=  ∈ SO3 ( ).
  q 
− ν ν 2 + ν2 0
2 1 3 
− √ ν23 2 0 √ ν21 2
  
ν1 +ν3 ν1 +ν3 0 0 1
Solutions to Exercises in Chapter VI 277

If in the first step we apply D1 to a1 , we rotate a1 about the x2 -axis in such a

way that D1 · a1 lies in the x1 , x2 -plane. If in the second step we apply D2 to
D1 · a1 , then we rotate D1 · a1 about the x3 -axis, so that finally, D2 · D1 · a1 is
parallel to the x1 -axis. Altogether, the orientation-preserving rotation under
R
discussion is given by the mapping v 7→ A · v (v ∈ 3 ) with A := D1−1 · D2−1 ·
D ϕ · D2 · D1 . Multiplying out yields

ν12 µ + cos( ϕ) ν1 ν2 µ − ν3 sin( ϕ) ν1 ν3 µ + ν2 sin( ϕ)

 

A = ν1 ν2 µ + ν3 sin( ϕ) ν22 µ + cos( ϕ) ν2 ν3 µ − ν1 sin( ϕ) ,

ν1 ν3 µ − ν2 sin( ϕ) ν2 ν3 µ + ν1 sin( ϕ) ν32 µ + cos( ϕ)

where we have set µ := 1 − cos( ϕ); this can now be easily decomposed as

A = E + sin( ϕ) · N + (1 − cos( ϕ)) · N 2

with
0 −ν3 ν2
 

N :=  ν3 0 −ν1  ,
−ν2 ν1 0

as asserted.
Selected Literature

The following list of books on (elementary) number theory and algebra can
serve to fill in some of the gaps in this book’s presentation. Some of these
books will take the reader much deeper into various topics. The literature
on the concept of number and the representation of numbers is of cultural-
historical significance, while the works of a historical nature provide insight
into the historical development of algebra and number theory. Finally, we
offer the interested reader two books on approaches to the teaching of alge-
bra and number theory.
Selected literature for the appendices is listed at the end of the respective
appendix.

Literature on Number Theory

[1] D. Burton: Elementary number theory. McGraw-Hill Education, 7th edi-

tion, 2010.
[2] W. A. Coppel: Number theory. An introduction to mathematics. Springer,
Berlin Heidelberg New York, 2nd edition, 2009.
[3] G. H. Hardy, E. M. Wright: An introduction to the theory of numbers. Ox-
ford University Press, 6th edition, 2008.
[4] H. Hasse: Number theory. Translated from the 3rd German edition. Sprin-
ger, Berlin Heidelberg New York, 1980.
[5] L. -K. Hua: Intoduction to number theory. Translated from the Chinese
original by P. Shiu. Springer, Berlin Heidelberg New York, 1982.
[6] K. Ireland, M. Rosen: A classical introduction to modern number theory.
Springer, Berlin Heidelberg New York, 2nd edition, 1990.
[7] F. Jarvis: Algebraic number theory. Springer, Cham Heidelberg New York
Dordrecht London, 2014.
[8] G. A. Jones, J. M. Jones: Elementary number theory. Springer, London,
1998.
[9] M. B. Nathanson: Elementary methods in number theory. Springer, Berlin
Heidelberg New York, 2000.
[10] I. Niven, H. S. Zuckerman, H. L. Montgomery: An introduction to the the-
ory of numbers. John Wiley & Sons, Hoboken, NJ, 5th edition, 2008.
[11] D. Redmond: Number theory. Marcel Dekker, New York, 1996.
[12] K. H. Rosen: Elementary number theory and its applications. Pearson, Bos-
ton, 6th edition, 2010.

© Springer International Publishing AG 2017

J. Kramer and A.-M. von Pippich, From Natural Numbers to Quaternions,
Springer Undergraduate Mathematics Series, https://doi.org/10.1007/978-3-319-69429-0
280 Selected Literature

[13] W. Sierpinski: Elementary theory of numbers. Elsevier, Amsterdam, PWN,

Warsaw, 2nd edition, 1988.
[14] A. Weil: Basic number theory. Springer, Berlin Heidelberg New York,
3rd edition, 1995.

Literature on Abstract Algebra

[15] M. Artin: Algebra. Pearson, Boston, 2nd edition, 2017.

[16] R. Cooke: Classical algebra: its nature, origins, and uses. John Wiley &
Sons, Hoboken, NJ, 2008.
[17] D. S. Dummit, R. M. Foote: Abstract algebra. John Wiley & Sons, Hobo-
ken, NJ, 3rd edition, 2003.
[18] B. Fine, A. M. Gaglione, G. Rosenberger: Introduction to abstract algebra.
Johns Hopkins University Press, Baltimore, MD, 2014.
[19] J. Gallian: Contemporary abstract algebra. Brooks Cole, 9th edition, 2016.
[20] R. S. Irving: Integers, polynomials, and rings. Springer, Berlin Heidelberg
New York, 2004.
[21] S. Lang: Algebra. Springer, Berlin Heidelberg New York, 3rd edition,
2002.
[22] F. Lorenz: Algebra. Volume I. Fields and Galois theory. Translated from the
1987 German edition by S. Levy. Springer, Berlin Heidelberg New York,
2006.
[23] W. K. Nicholson: Introduction to abstract algebra. John Wiley & Sons,
Hoboken, NJ, 4th edition, 2012.
[24] J. J. Rotman: A first course in abstract algebra. Pearson, Boston, 3rd edi-
tion, 2005.
[25] L. H. Rowen: Algebra: groups, rings and fields. A K Peters, Wellesley, MA,
1994.
[26] J. Stillwell: Elements of algebra: geometry, numbers, equations. Springer,
Berlin Heidelberg New York, 1994.
[27] B. L. van der Waerden: Algebra. Volume I. Springer, Berlin Heidelberg
New York, 9th edition, 1993.

Literature on the Concept of Number

[28] J. H. Conway, R. K. Guy: The book of numbers. Springer Copernicus, New

York, 1996.
[29] L. Corry: A brief history of numbers. Oxford University Press, Oxford,
2015.
[30] H. Ebbinghaus et al.: Numbers. Translated from the 2nd German 1988
edition by H. L. S. Orde. Springer, Berlin Heidelberg New York, 1991.
[31] G. Ifrah: From one to zero: a universal history of numbers. Translated from
the French original by L. Bair. Penguin Books, New York, 1987.
 281

[32] K. Menninger: Number words and number symbols: a cultural history of

numbers. Translated from the German revised edition by P. Broneer.
Dover, New York, 1992.
[33] R. Taschner: Numbers at work: a cultural perspective. Translated from the
2005 German original by O. Binder and D. Sinclair-Jones. A K Peters,
Wellesley, MA, 2007.

Literature on the History of Algebra and Number Theory

[34] I. G. Bashmakova, G. S. Smirnova: The beginnings and evolution of algebra.

Translated from the Russian original by A. Shenitzer. Mathematical As-
sociation of America, Washington, DC, 2000.
[35] V. J. Katz, K. H. Parshall: Taming the unknown: a history of algebra from an-
tiquity to early twentieth century. Princeton University Press, Princeton,
NJ, 2014.
[36] A. Weil: Number theory. Birkhäuser Boston, Boston, MA, 1984.

Literature on the Teaching of Algebra and Number Theory

[37] A. Arcavi, P. Drijvers, K. Stacey: The learning and teaching of algebra. IM-
PACT Series, Routledge, 2016.
[38] J. D. Sally, P. J. Sally: Integers, fractions, and arithmetic: a guide for teachers.
American Mathematical Society, Providence, RI, 2012.
Index

associative operation, 45 element

identity, 47
bounded set, 160 inverse, 48, 101
irreducible, 120
left identity, 47
Cauchy sequence
left inverse, 48, 101
of an ordered field, 168
prime, 121
rational, 145
right identity, 47
real, 153
right inverse, 48, 101
Cayley table, 51
unit, 98
Cayley’s octonions, 231
zero, 98
characteristic, 101
equivalence class, 57
completeness
equivalence relation, 57
axiom of geometric, 166
Euclid’s lemma, 23, 96
of an ordered field, 168
Euclidean algorithm, 126
of real numbers, 155
extended, 127
completeness principle, 162
Euclidean domain, 125
complex numbers, 184 exponential function, 197
absolute value, 187
complex conjugate, 186 field, 110
imaginary part, 184 absolute value, 168
modulus, 187 algebraically closed, 193
purely imaginary, 184 archimedean, 169
real part, 184 order, 168
complex plane, 184 field of fractions, 117
coset, 61 fraction, 76, 118
left, 59 fundamental theorem
right, 61 of algebra, 191
of arithmetic, 22, 96
decimal, 155
genuine, 159 group, 48
infinite, 155 abelian/commutative, 49
terminating, 155 alternating, 68
decimal expansion cyclic, 52
period, 143 dihedral, 50
periodic, 143 symmetric, 51
purely periodic, 143 group homomorphism, 54
Dedekind cuts, 167 group isomorphism, 54
division with remainder, 30, 95, 125
divisor, 15, 94, 120, 121 Hamilton’s quaternions, 219
common, 15, 95, 120 homomorphism theorem
greatest common, 25, 97, 120, 122 for groups, 66
proper, 17 for rings, 108
trivial, 17, 95
domain, 100 ideal, 104
© Springer International Publishing AG 2017
J. Kramer and A.-M. von Pippich, From Natural Numbers to Quaternions,
Springer Undergraduate Mathematics Series, https://doi.org/10.1007/978-3-319-69429-0
284 Index

divisibility, 121 octonions, 231

greatest common divisor, 122 order
intersection, 122 of a group, 52
least common multiple, 122 of an element, 53
principal, 105 orthogonal group
sum, 122 R
O2 ( ), 189
unit, 105 R
O3 ( ), 228
zero, 105
image Peano axioms, 9
of a group homomorphism, 55 predecessor, 9
of a ring homomorphism, 104 prime number, 17, 95
index of a subgroup, 61 Fermat, 19
induction, 9 Mersenne, 19
infimum, 161 principal ideal domain, 124
infimum principle, 162
integers, 74 quaternions, 219
absolute value, 76 conjugate quaternion, 221
decimal representation, 141 imaginary part, 220
difference, 75 imaginary space, 221
order, 75 modulus, 221
product, 93 purely imaginary, 220
quotient, 118 real part, 220
integral domain, 100 quotient group, 65
quotient ring, 108
kernel
of a group homomorphism, 55 R-algebra, 222
of a ring homomorphism, 104 associative, 222
commutative, 222
dimension, 222
least common multiple, 26, 97, 120, 122
division algebra, 222
lower bound, 160
R -subalgebra, 223
monoid, 47
R -algebra homomorphism, 223
rational numbers, 118
multiplicative group of a ring, 111 absolute value, 119
decimal representation, 143
natural numbers, 9 order, 119
decimal representation, 31 real number line, 164
difference, 14, 75 real numbers, 151
order, 13 absolute value, 153
product, 11 decimal expansion, 159
sum, 10 decimal representation, 159
nested intervals, 162 order, 152
nested intervals principle, 162 real sequence
normal subgroup, 61 (strictly) monotonically decreas-
null sequence ing, 160
rational, 145 (strictly) monotonically increasing,
number 159
algebraic, 193 convergence, 153
amicable, 21 limit, 153
Euler, 197 relatively prime, 28
irrational, 159 pairwise, 28
Liouville, 196 ring, 97
perfect, 20 commutative, 98
transcendental, 194 factorial, 123
Index 285

polynomial, 99 supremum principle, 162

zero, 98
ring homomorphism, 103 theorem
ring isomorphism, 103 Gauss’s, 123
Lagrange’s, 60
semigroup, 45 Liouville’s, 194
abelian/commutative, 46 of Euclid, 18
regular, 69
skew field, 110 unique factorization domain, 123
special orthogonal group unit, 101
R
SO2 ( ), 189 imaginary, 184
R
SO3 ( ), 228 unitary group
special unitary group C
U2 ( ), 225
C
SU2 ( ), 225 upper bound, 160
subgroup, 53
subgroup criterion, 53 well-ordering principle, 14
subring, 102
subring criterion, 102 zero divisor, 100
successor, 9 left, 100
supremum, 161 right, 100