0% found this document useful (0 votes)
528 views84 pages

HW

1) The document contains proofs of various properties of groups and permutations. 2) It proves that every permutation can be written as a product of adjacent transpositions, and that a transposition of elements i and j is an odd number of adjacent transpositions. 3) It also proves properties of the sign of a permutation, including that the sign of a product of permutations is the product of the signs.

Uploaded by

Tg Dg
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
528 views84 pages

HW

1) The document contains proofs of various properties of groups and permutations. 2) It proves that every permutation can be written as a product of adjacent transpositions, and that a transposition of elements i and j is an odd number of adjacent transpositions. 3) It also proves properties of the sign of a permutation, including that the sign of a product of permutations is the product of the signs.

Uploaded by

Tg Dg
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 84

Debra Griffin Math 210A HW #1, 2.2 p. 50 #2, 7, 9, 11, 15 2.2 If Sn , prove that sgn (1) = sgn().

). Proof: If = 1 2 k (a complete factorization into disjoint cycles), then 1 = k1 21 11 (Proposition 2.4). By definition, sgn (1) = (1)nk = sgn(). 2.7 Given X ={1, 2, , n}, let us call a permutation of X an adjacency if it is a transposition of the form (i, i + 1) for i < n. (i) Prove that every permutation in Sn , for n 2, is a product of adjacencies. Proof: Every permutation Sn , n 2, is a product of transpositions (Prop. 2.10), so we only need to show that any transposition can be written as a product of adjacencies. Let = 1 2 t . Let i = (i j), i < j. (Note that i = (j i) as well.) Then i = (i j) = (i + 1 i)(i + 2 i + 1)(j 1 j 2)(j 1 j)(i + 1 i + 2)(i i + 1). For example, (1 5) = (2 1)(3 2)(4 3)(4 5)(3 4)(2 3)(1 2). is a product of adjacencies. (ii) If i < j, prove that (i j) is a product of an odd number of adjacencies. Hint. Use induction on j i. Proof: If j i = 1, then (i j) is an adjacency. Assume (i j) is a product of an odd number of adjacencies for some j i > 1. Then, by part (i), (i j) = (i + 1 i)(i + 2 i + 1)(j 1 j 2)(j 1 j)(i + 1 i + 2)(i i + 1) is an odd number of adjacencies and (i j + 1) = (i + 1 i)(i + 2 i + 1)(j 1 j 2) (j j 1) (j j + 1) (j 1 j) (i + 1 i + 2)(i i + 1). This is 2 more adjacencies than the odd number of adjacencies whose product is the transposition (i j). Since the sum of an odd number and 2 is odd, then the number of adjacencies whose product is the transposition (i j +1) is odd.

Page 1

2.9 If is an r-cycle and 1 < k < r, is k an r-cycle? No. is an r-cycle if and only if gcd(r, k) = 1. Proof: Let i1, i2, , ir be distinct integers in {1, 2, n}. : We will show this direction by contradiction. Suppose gcd(r, k) = d > 1. Then r = dm and k = dn for some positive integers m and n. Note that m > n. Thus k(i1) = dn(i1) = idn+1, dn(idn+1) = i2dn+1, , dn(i(m1)dn+1) = idmn+1 = i1. k produces a cycle that does not include ir, and, hence, cannot be an r-cycle. : Now, suppose gcd(r, k) = 1. Then m, n {1, 2, ..., r}, such that m n, (mk + 1)(mod r) (nk + 1)(mod r). Otherwise, mk + 1 rp = nk + 1 rq for some integers p, q. Thus, mk nk = rp rq = r(p q), forcing r | (m n)k. Since r /| k, then r | (m n). But 1 m r, 1 n r, and m n, so r < 1 n m n r n < r. This gives us that |m n| < r, hence r /| (m n). {(jk + 1)(mod r)| j {1, 2, ..., r}} = {1, 2, ..., r}. So then j {0, 1, 2, ..., r}, k(ijk+1) = i(j+1)k+1. Thus { (ijk+1) | j {0, 1, 2, ..., r}} = {i1, i2, , ir}. k is an r-cycle.
k

Page 2

2.11 (i) Prove, for all i, that Sn moves i if and only if 1 moves i. Proof: : Assume moves i. Suppose = 1 2 (i c1 c2 ck) t is the complete factorization of into disjoint cycles. Then 1 = t 1 (ck c1 i) 2 1 1 1 by Proposition 2.4. By definition of (ck c1 i), 1 moves i. : Proof is similar to as we can simply interchange and 1 in the proof above to conclude that if 1 moves i then moves i. For all i, Sn moves i if and only if 1 moves i. (ii) Prove that if , Sn are disjoint and if = (1), then = (1) and = (1). Proof: Let , Sn be disjoint permutations such that = (1), We will show by contradiction that = (1) and = (1). Suppose that (1). Then i such that (i) = j i. Since , are disjoint, then (i) = (i) = j. (1), a contradiction. = (1). So now we have = (1) = = (1). = (1) and = (1). 2.15 If n 3, show that if Sn commutes with every Sn, then = (1). Proof: Assume such that = for every Sn. We shall show = (1) by assuming (1) and arriving at a contradiction. Suppose (1), then i , j {1, 2, ..., n} such that (i) = j and i j. Let Sn such that (i) = i, (j) = k, and (k) = j where i k j. Since n 3, then we know such a j exists. So (i) = (i) = j k =(j) = (i), and we have our contradiction. = (1).

Page 3

Thm 2.12 For all , Sn, sgn () = sgn()sgn(). Proof: Let a, b {1, 2, , n} such that a b. Let = (a b). We will show * sgn() = sgn() for any Sn. Case 1: Suppose = (a c1 c2 ck )(b d1 d2 dL)1 2 t , a complete factorization into disjoint cycles where K, L 0, Sn. Since (a b) (a c1 c2 cK)(b d1 d2 dL) = (a c1 c2 ck b d1 d2 dL), then sgn() = sgn((a c1 c2 cK b d1 d2 dL)1 2 t) = (1)n (t + 1) = (1)n (t + 2) = sgn((a c1 c2 cK )(b d1 d2 dL)1 2 t ) = sgn(). Case 2: Suppose = (a c1 c2 ck b d1 d2 dL)1 2 t , a complete factorization into disjoint cycles where K, L 0, Sn. Since (a b) (a c1 c2 cK)(b d1 d2 dL) = (a c1 c2 ck b d1 d2 dL), then (a b) (a b) (a c1 c2 cK)(b d1 d2 dL) = (a b) (a c1 c2 ck b d1 d2 dL), hence (a c1 c2 cK)(b d1 d2 dL) = (a b) (a c1 c2 ck b d1 d2 dL). sgn() = sgn((a c1 c2 cK )(b d1 d2 dL)1 2 t) = (1)n (t + 2) = (1)n (t + 1) = sgn((a c1 c2 cK b d1 d2 dL)1 2 t ) = sgn(). Now suppose = 1 2 m where each i is a transposition (apply Prop. 2.10) and = 1 2 t a complete factorization into disjoint cycles. We will show sgn () = sgn()sgn() by induction on m. Let m = 1. Then = 1 (c1) (c2) (cn2), hence sgn() = (1)n (n 1) = 1. By definition, sgn() = (1)n t. And by *, sgn() = sgn() = (1)n t. sgn () = sgn()sgn(). Let m > 1 and assume sgn () = sgn() sgn () for any that is a product of m transpositions. Define = m+1 1 2 m, where m+1 is a transposition in Sn. Let = 1 2 m. By *, sgn() = sgn(m+1) = sgn(). So then, by our induction hypothesis, sgn() = sgn (m+1) = sgn () = sgn()sgn() = sgn()sgn(). sgn () = sgn()sgn() for all , Sn.

Page 4

Debra Griffin Math 210A HW #2, 2.3 p. 61 #18, 21 28 2.18 Assume that G is a set with an associative binary operation. Prove that (ab)(cd) = a[(bc)d] without using generalized associativity. Proof: If G is a set with an associative binary operation and a, b, c, d G, then (ab)(cd) = a[b(cd)] = a[(bc)d], as desired. 2.21 If G is a group, prove that the only element g G with g2 = g is 1. Proof: Assume g G with g2 = g. Then g1 G and multiplying both sides by g1 gives us g = g2 g1 = g g1 = e. 2.22 This exercise gives a shorter list of axioms defining a group. Let H be a set containing an element e, and assume that there is an associative binary operation on H satisfying the following properties: 1. e x = x for all x H; 2. for every x H, there is x H with x x = e. (i) Prove that if h H satisfies h h = h, then h = e. Hint: If h h = e, evaluate h h h in two ways. Proof: If h H satisfies h h = h, then h = e h = h h h = h h = e. h = e. (ii) For all x H, prove that x x = e. Hint: Consider (x x)2. Proof: For all x H, (x x)2 = (x x)(x x) = x (x x ) x= x e x= x x. By result in (i), (x x)2 = x x x x = e. (iii) For all x H prove that x e = x. Hint: Evaluate x x x in two ways. Proof: Since is associative, then x e = x (x x)= (x x) x = e x = x.

Page 1

(iv) Prove that if e H satisfies e x = x for all x H, then e = e. Hint: Show that (e)2 = e. Proof: If e H satisfies e x = x for all x H, then (e)2 = e e = e. Again by (i), e = e. (v) Let x H. Prove that if x H satisfies x x = e, then x = x. Hint: Evaluate x x x in two ways. Proof: If x H satisfies x x = e, then by (ii), x x = e. x = x e = x x x = e x = x. (vi) Prove H is a group. Proof: By (i), if h H satisfies h h = h, then h = e. This gives us cancellation for h h = h. By (ii), for all x H, x x = e. This gives us an element x such that x x = e = x x. By (iii), for all x H, x e = x. This gives us that H has an element, e, such that x H, e x = x = x e. By (iv), if e H satisfies e x = x for all x H, then e = e. This give us uniqueness of the identity, e. By (v), if x H satisfies x x = e, then x = x. This gives us uniqueness of each inverse. We have a set H equipped with a binary operation such that (i) the associative law holds; (ii) there is an element e H, with e x = x = x e for all x H; (iii) every x H has an inverse; there is x H with x x = e = x x. H is a group.

Page 2

2.23 Let y be a group element of order m; if m = pt for some prime p, prove that yt has order p. Hint: Clearly, (yt)p = 1. Use Theorem 2.24 to show that no smaller power of yt is equal to 1. Proof: Assume (y) = m = pt for some prime p, then 1 = ym = ypt = (yt)p. Suppose d < p such that (yt)d = 1. Then td | m by Theorem 2.24 (If a
order n, then am = 1 if and only if n | m.) G is an element of

However, d /| p as p is prime, hence td /| pt = m.

(yt) = p.

2.24 Let G be a group and let a G have order k. If p is a prime divisor of k, and if
there is x G with xp = a, prove that x has order pk. Proof (My newer elegant proof): Since (a) = k and xp = a, then xpk = (xp)k = ak = 1. x has finite order, w, such that w | pk. This gives us that k | w. Since p is a prime divisor of k, then p | k | w. So then we have that w/p is an integer. Now, ( x
p w p

= x w = 1, hence k | (w/p) or kp | w.

w = pk, the order of x.

Earlier Proof: Since (a) = k and xp = a, then xpk = (xp)k = ak = 1. x has finite order, w, such that w | pk, hence pk = wm. So, aw = (xp)w = (xw)p = 1p = 1. This gives us that k | w, hence w = km. So pk = wm = kmm, hence p = mm. Since p is prime, then m = 1 and m = p or m = p and m = 1. Suppose m = 1 and m = p, then pk = wp, hence k = w. Since p is a prime divisor of k, then for some t Z+ such that t < k, k = pt. So 1 = xw = xk = xpt = at, a contradiction (as (a) = k). m = p and m = 1, which gives us pk = w 1 = w. (x) = pk.

Page 3

0 1 0 1 2.25 Let G = GL(2, Q), and let A = and B = . 1 0 1 1 1 0 Show that A4 = I = B6, but that (AB)n I for all n > 0, where I = is the 2 2 identity 0 1 matrix. Conclude that AB can have infinite order even though both factors A and B have finite order (this cannot happen in a finite group). Proof: 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 B = 4 A = 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 = 1 0 1 1 1 1 1 1 1 1 1 0 0 1 0 1 1 0 0 1 0 1 0 1 = = 0 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 0 1 0 1 0 1 0 1 = = 1 1 1 1 1 1 1 0 1 0 1 1 0 1 = 1 0 1 1 1 0 = 1 0 = 0 1 0 1
6

0 1 0 1 1 1 AB = = 1 0 1 1 0 1

1 1 1 1 1 1 ( AB) 2 = = 0 1 0 1 0 1

1 1 1 0 Clearly, n > 1, ( AB) n = . AB has infinite order. 0 1 0 1


2.26 If G is a group in which x2 = 1 for every x G, prove that G must be Abelian. [The Boolean groups B(X) of Example 2.18 are such groups.] Proof: Let a, b G. Then ab G. We are given that x2 = 1 for every x G, and so 1 = (ab)2 = a2 b2. Thus, abab = aabb. Left multiplying by a1 and right multiplying by b1 gives us ba = ab. G is Abelian.

Page 4

2.27 If G is a group with an even number of elements, prove that the number of elements in G of order 2 is odd. In particular, G must contain an element of order 2. Hint: Pair each element with its inverse. Proof: Let S = {a G: (a) 3}. Then a S, a1 S such that a a1. This gives us a pairing of each element in S with its inverse, hence |S| is even. Let M = {b G: (b) < 3}. Then |M| = |G| |S|, hence |M| is even. Since 1 has order 1, then 1 M and all other elements in M have order 2. The number of elements of order 2 in G is odd. In fact, since 1 G and |M| is even, there must be at least one element of order 2. 2.28 What is the largest order of an element in Sn, where n = 1, 2, , 10? (We remark that no general formula is known for arbitrary n, although, in 1903, E. Landau found the asymptotic behavior.) Proof: n 1 2 3 4 5 6 Sn Cycle structures/Orders (1) = 1 (1) = 1, (12) = 2 (1) = 1, (12) = 2, (123) = 3 (1) = 1, (12) = 2 , (123) = 3, (1234) = 4, (12)(34) = 2 (1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123)(45) = 6 (1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123456) = 6, (12)(45)(56) = 2, (12)(3456) = 4, (123)(456) = 3 (1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123456) = 6, (1234567) = 7, (12)(45)(56) = 2, (12)(34567) = 10, (123)(4567) = 12 (1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123456) = 6, (1234567) = 7, (12345678) = 8, (12)(45)(56)(78) = 2, (123)(45678) = 15, (1234)(5678) = 4 Largest order element 1 2 3 4 6 6

12

15

Page 5

10

(1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123456) = 6, (1234567) = 7, (12345678) = 8, (123456789) = 9, (12)(45)(56)(78) = 2, (12)(34)(56)(789) = 6, (123)(456)(789) = 3, (123)(456789) = 6, (1234)(5678) = 4, (1234)(56789) = 20 (1) = 1, (12) = 2, (123) = 3, (1234) = 4, (12345) = 5, (123456) = 6, (1234567) = 7, (12345678) = 8, (123456789) = 9, (123456789 10) = 10, (12)(45)(56)(78)(9 10) = 2, (12)(34)(56)(789 10) = 4, (12)(34)(567)(89 10) = 6, (12)(345)(6789 10) = 30, (123)(456)(789) = 3, (123)(456789 10) = 21, (1234)(56789) = 20, (1234)(56789 10) = 12, (12345)(12345) = 25,

20

30

Page 6

Debra Griffin Math 210A HW #3, 2.4 p. 72 #29, 31-35, 37 Hand in #35 2.29 Let H be a subgroup of a group G. (i) Prove that right cosets Ha and Hb are equal if and only if ab1 H. Proof: : Assume Ha and Hb are equal. Then h H, h H such that ha = hb. hab1 = hbb1. 1 h hab1 = h1h. ab1 = h1h H. : Assume ab1 H. We will first show Ha Hb. Let ha Ha. Then ab1 H hab1 H hab1b Hb ha Hb. Ha Hb. To show Hb Ha, we let hb Hb. Then ab1 H (ab1)1 = ba1 H hba1 H hba1a = hb Ha. Hb Ha. Ha = Hb. (ii) Prove that the relation a b if ab1 H is an equivalence relation on G whose equivalence classes are the right cosets of H. Proof: Let a, b G. We first show reflexivity. Since 1 = aa1 H, then a a. is reflexive. To show symmetry, we note that ab1 H gives us that (ab1)1 = ba1 H also. a b b a. is symmetric. And to show transitivity, we assume a b and b c and note that ab1 H and bc1 H gives us ab1 bc1 = ac1 H by closure. is transitive. Let the equivalence class of a be X. Then X = {g G | g b} = {g G | gb1 H} = {g G | g Hb} = Hb. By part (i) Hb = Ha, and these are the right cosets of H.

Page 1

2.31 (i) Give an example of two subgroups H and K of a group G whose union, H K is not a subgroup of G. Hint: Let G be the four-group V. Proof: Let G = {1, i, j, k} where i2, j2, k2 = 1, ij = k, jk = i, and ki = j . Let H = {1, i} and K = {1, j}. It is clear that H and K are subgroups of G. But H K = {1, i, j} does not have closure as ij = k H K. (ii) Prove that the union, H K, of two subgroups is itself a subgroup iff either H is a subset of K or K is a subset of H. Proof: Let G be a group, H G, and K G. : Assume H K G. Let h H and k K. Then hk H or hk K. Suppose hk K. Then hkk1 = h K. H K. Similarly, if hk H, then h1hk = k H. K H. H K or K H, as desired. : If H K, then H K = K G. If K H, then H K = H G. H K G. 2.32 Let G be a finite group with subgroups H and K. If H K, prove that [G:H] = [G:K][K:H]. Proof: Let n = [G:H], m = [G:K], and t = [K:H]. Then n|H| = |G| = m|K|. Since |K| = t|H|, then n|H| = mt|H|, hence n = mt. [G:H] = [G:K][K:H].

Page 2

2.33 If H and K are subgroups of a group G and if |H| and |K| are relatively prime, prove that H K = {1}. Hint: If x H K, then x|H| = x|K|. Proof: Let a H K with (a) = d. Then ad = 1. Since a H and a K, then d | |H| and d | |K|. But |H| and |K| are relatively prime, so d must be 1. (a) = 1 and the only element with order 1 is the identity. Hence a = 1. 2.34 Prove that every subgroup S of a cyclic group G = a is itself cyclic. Hint: If S 1, choose k to be the smallest positive integer with ak S. Proof: Let S be a subgroup of G. If S = {e}, then S = e and we are done. Assume S {e}, then S contains a nonidentity element of G, say ai with i 0. Thus ai S also. One of i or i is positive, so S contains positive powers of a. Let k be the smallest positive integer such that ak S. If x is an arbitrary element in S, then x = am for some m. By the Division Algorithm, m = kq + r with 0 r < k. Thus r = m kq and ar = am kq = amakq = am(ak)q. Both am and ak are in S. ar S by closure. Since ak is the smallest positive power of a in S and since r < k, we must have r = 0. m = kq and x = am = akq = (ak)q ak. Hence S = ak.

Page 3

2.35 Prove that a cyclic group G of order n has a unique subgroup of order d for every d dividing n. Hint: If G = a and n = dk, consider ak. Proof: Let G be a cyclic group a of order n and suppose d | n. Then n = dk for some integer k 1. Since G contains every integral power of a, then ak G. And since (a) = |a| = n, then we can apply the gcd theorem to find the order of ak.
( a k ) = n dk = = d . d = (ak) = |ak|. gcd( n, k ) gcd( dk, k )

And we know ak a as t Z , (ak)t = akt where kt Z .

To show ak a we first note that ak as ak ak. Let aki, akj ak. Since aki (akj)1 = ak(i j), then ak(i j) ak, hence ak a. We now will show the uniqueness of ak. Suppose x is a subgroup of a and |x| = d (dk = n, as defined above). Then for some integer m, x = am. Let xt x. Then xt = amt. And amtd = 1 as (am) = d. Thus, as (a) = n we have n | md md = nr for some r m = kr. We can write xt = amt = akrt = (ak)rt. xt ak, hence x ak. And since |x| = |ak|, then x = ak.

Page 4

2.37 If H is a subgroup of a group G, prove that the number of left cosets of H in G is equal to the number of right cosets of H in G. Hint: The function : aH Ha1 is a bijection from the family of all left cosets of H to the family of all right cosets of H. Proof: Define : {aH }a G {Hb}b

by (aH) = Ha1. We will show this is a bijective function.

Let x, y G where x y. We have that (xH) = (yH) Hx1 = Hy1 y1x = y1(x1)1 H xH = yH. is well-defined and injective. Let Hb be a right coset in G. Since G is a group, then there is a unique inverse of b in G. And since ((b1)H) = H(b1)1 = Hb, then is surjective. is a bijection from {aH }a G to {Hb}b G.

The number of left cosets of H in G is equal to the number of right cosets of H in G.

Page 5

Debra Griffin Math 210A HW #4, 2.5 p. 80 #42, 47, 51, 52, 53, 64 2.42 This exercise gives some invariants of a group. Let f : G H be an isomorphism. (i) Prove that if a G has infinite order, then so does f (a), and if a has finite order n, then so does f (a). Conclude that if G has an element of some order n and H does not, then ~(G H). Proof: Let a G and suppose that the order of a is infinite. Then n, m Z , where n m, an am (otherwise 1 = anan = aman = amn where m n is finite). Then by injectivity and homomorphism properties of f, f (a)n = f (an) f (am) = f (a)m. The order of f (a) is infinite. Now suppose the order of a is n < . Then, 1 = an f(an) = f (a)n = 1. The order of f (a) divides n. The order of f (a) is finite. Additionally, we have that, if [ f (a)] = m < , 1 = f (a)m = f(am) am = 1, hence n | m. And, as previously noted, [ f (a)] | n. m = n. In conclusion, this gives us that if G has an element of some order n and H does not, then ~(G H). (ii) Prove that if G H, then, for every divisor d of |G|, both G and H have the same number of elements of order d. Proof: Since G H, then f : G H that is a bijection. Let d be a divisor of |G|. Let Gd = {g G | (g) = d} and Hd = {h H | (h) = d}. We shall show there is a bijection : Gd Hd, hence |Gd| = |Hd|. Define : Gd Hd by (g) = f (g), and let a, b Gd such that (a) = (b). Then f (a) = f (b). Since f is an isomorphism, then a = b. is an injection. Let y Hd. Then (y) = d. Since f is a surjection, then x G such that f (x) = y. And by (i), (x) = (f (x)) = (y). (x) = d, hence x Gd. is a surjection and we have our desired bijection. For every divisor d of |G|, both G and H have the same number of elements of order d.

Page 1

2.47

(i) If f : G H is a homomorphism and x G has order k, prove that f (x) H has order m, where m | k.

Proof: Let x G with (x) = k. By homomorphism properties, if m = (f (x)), then 1 = f (1) = f (xk ) = f (x)k. Hence, m | k. (ii) If f : G H is a homomorphism and if (|G|, |H|) = 1, prove that f (x) = 1 for all x G. Proof: Let |G| = n and |H| = q. Let x G, (x) = k, and (f (x)) = m. By (i), m | k, and by corollary to Lagranges Theorem, k | n and m | q. So now we have m | k | n which implies m | n and m | q. But since (n, q) = 1, then m = 1. So then x G, (f (x)) = 1, but the only element with order one is the identity. f (x) = 1 for all x G. 2.51 (i) Prove that if Sn, then and 1 are conjugate.

Proof: Let Sn. Then = (c1 c2 ...cr) for some r n and 1 = (cr cr1 ...c2 c1). Let Sn such that (c1) = cr, (c2) = cr1, ..., (cr) = c1. Then 1 = 1. and 1 are conjugate. (ii) Give an example of a group G containing an element x for which x and x1 are not conjugate. Proof: Let G = Z 3 = {[0], [1], [2]}. Let x = [1], then x1 = [2]. [0]([1]) = [0] + [1] + [0] = [1] [2]. [1]([1]) = [1] + [1] + [2] = [0] [2]. [2]([1]) = [2] + [1] + [1] = [0] [2]. g G, gxg1 x1.

Page 2

2.52 Prove that the intersection of any family of normal subgroups of a group G is itself a normal subgroup of G. Proof: Let {K}G be a family of normal subgroups of G. Since e K for each , then e K . Let j, k

K . Then j, k K for each .

Since K is a group for each , then j 1 K . 1 And by closure, kj K for each , hence kj 1 K . Let g G. Then, since K G for each , gkg1 K for each . gkg1 K . K G .

2.53 Define W = (12)(34), the cyclic subgroup of S4 generated by (12)(34). Show that subgroup of V, but that W is not a normal subgroup of S . Conclude that W is a normal 4 normality is not transitive: W V and V G do not imply W G. Proof: V = {(1), (12)(34), (23)(41), (13)(42)}. W = (12)(34) = {(1), (12)(34)}. If W V, then w W, v V, vwv1 W. So, we verify W V by computation: Since [(12)(34)]2 = [(23)(41)]2 = [(13)(42)]2 = (1), then (1)(1)(1) = (12)(34)(1)(12)(34) = (23)(41)(1)(23)(41) = (13)(42)(1)(13)(42) = (1) W. and (1)(12)(34)(1) = (12)(34)(12)(34)(12)(34) = (23)(41)(12)(34)(23)(41) = (13)(42)(12)(34)(13)(42) = (12)(34) W. W V. To verify V S4, we will simply note that S4, (1)1 = (1) and where V and (1), 1 is of the same cycle structure as (i.e. (12)(34)). Since there are 3 permutations of the form (12)(34) in V and the same 3, and only 3, occur in S4, then 1 is guaranteed to live in V. V S4. However, (123)(12)(34)(321) = (32)(14) W. ~(W S4). Thus, we can conclude that normality is not transitive.

Page 3

2.64

(i) For every group G, show that the function : G Aut (G), given by g g (where x is conjugation by g), is a homomorphism.

Proof: Let a, b, x G. Then ab(x) = (ab)x(ab)1 = abxb1a1 = aba1(x) = ab(x). Thus = (ab) = ab = (a) (b). is a homomorphism. (ii) Prove that ker = Z(G) and im = Inn(G); conclude that Inn(G) is a subgroup of Aut(G). Proof: We first show ker = Z(G). Note that the identity of Aut (G) is the identity map i: G G where g G, i(g) = g. Then ker = {g G | (g) = i}. To show ker Z(G) we let a ker and note that x G, a(x) = axa1 = x. x G, xa = ax. a Z(G). To show ker Z(G), let b Z(G), let g G, and note that bgb1 = g. Thus, b(g) = bgb1 = g, hence b ker . ker = Z(G). And now we will show im = Inn(G). We have im Inn(G) by definition of . To show im Inn(G), let b Inn(G). Then b = x = (x) for some x G. im = Inn(G). We now show, in conclusion, that Inn(G) Aut(G). We know Inn(G) Aut(G) by definition. Since e Inn (G), Inn (G) . Let a, b, x G and let a, b Inn (G). Since b1 G, then b1xb G and b(b1xb) = bb1xbb1 = x = bb (x). So x G, a b (x) = a (b1xb) = a(b1xb)a1 = (ab1) x(ba1) = (ab1) x(ab1)1 = ab 1 (x). And ab 1 (x) Inn (G) (since ab1 G). Inn(G) Aut(G).
1 1

(iii) Prove that Inn(G) Aut(G).

Proof: By (ii) we know Inn(G) Aut(G). Let Aut (G), g Inn (G), where g G. Then by homomorphic and surjective properties of , x G, g1(x) = (g1(x)g1) = (g)(1(x))(g1) = (g) x (g1) = (g)(x). Since (g) G, then (g) Inn (G), as desired. Inn(G) Aut(G).

Page 4

Debra Griffin Math 210A HW #5, 2.6 p. 95 #67, 68, 69, 71, 76 2.67 (i) Prove that Aut(V) S3 and that Aut (S3) S3. Conclude that nonisomorphic groups can have isomorphic automorphism groups. Proof: S3 = {1, a, b, c, ab, bc} = {1, 2, ..., 6}. We first show that Aut(V) S3. Let : Aut(V) S3 where Aut(V) , () = (V). We know V = {1, a', b', c'} and Aut(V), (1) = 1. Since V consists of 3 nonidentity elements, a', b', c' of the same order, then every Aut(V) permutes these 3 elements and there are 6 distinct permutations of 3 elements. Then we can let Aut(V) = {1, 2, ..., 6}. One permutation, 1, is the identity permutation. Three permutations, 2, 3, 4, permute only 2 elements and fix the other, hence have order 2. And two permutations, 5, 6, permute all 3 elements, and thus have order 3. Define : Aut(V) S3 by (i) = i. Clearly, is an isomorphism. We now show that Aut (S3) S3. Let S3 = {1, a, b, c, ab, bc} where a, b, c have order 2; and ab and bc have order 3. Since S3 consists of 3 elements of order 2, 2 elements of order 3 and 1, then every Aut(S3) permutes the 3 elements of order 2 and there are 6 distinct permutations of the 3 elements. By homormorphism properties, (ab) = (a)(b), thus each permutation of the order 2 elements determines the assignment of the order 3 elements. Thus Aut(S3) has exactly 6 distinct elements. Let Aut(S3) = {1, 2, ..., 6}. One permutation, 1, is the identity permutation. Three permutations, 2, 3, 4, permute only 2 of the order 2 elements, a, b, and c, and fix the other, hence have order 2. And two permutations, 5, 6, permute all 3 of a, b, and c, and thus have order 3. Define : Aut(S3) S3 by (i) = i; hence is an isomorphism from Aut(S3) S3. (ii) Prove that Aut(Z ) Z 2. Conclude that an infinite group can have a finite automorphism group. Proof: Since Z = 1 = 1, then if Aut (Z ), (1) and (1) must be generators of the group. Since 1 and 1 are the only generators of Z , then there are only 2 automorphisms of Z . Hence Aut(Z ) Z 2.

Page 1

2.68 If G is a group for which Aut (G) = {1}, prove that |G| 2. Proof: From an online source Assume that the only automorphism of G is the identity mapping. Then all inner automorphisms of G are trivial, so G must be abelian. Next, the function : G G defined by (x) = x 1, for all x in G, is an automorphism, so by hypothesis it must be trivial. This forces x = x 1 for all x in G. If G is written additively, then G has a vector space structure over the field Z2. (Since every element of G has order 2, it works to define 0 x = 0 and 1 x = x, for all x in G.) With this vector space structure, any group homomorphism is a linear transformation (and vice versa), so the automorphism group of G is a group of invertible matrices. Therefore Aut (G) is nontrivial, unless G is zero or one-dimensional. For |G| = 1, the result holds trivially. For |G| = 2, G = {e, a}. Since Aut(G), (e) = e, then the bijection forces (a) = a. Thus we have only the identity map. (*) Suppose |G| 3 and Aut(G) = {1}. We will show that over the field 2, G satisfies the axioms of a vector space. We start with the commutativity axiom. g G, the inner automorphism : G G defined by (g) = gxg1 is the identity map. So then g, x G, gx = xg, hence G is abelian. The automorphism : G G defined by (x) = x1 must be the identity map, hence x G, x = x1. This gives us that every nonidentity element of G has order 2. So then we can define 02 x = 0G and 12 x = x (If x has order > 2, the distributive law fails; for example, if x has order 3, then (1 + 1) x = 0 x = 0 x + x = 1 x + 1 x.) So now we can assume G satisfies the axioms of a vector space. Consider G*, the set of all linear transformations from G to itself. The set of all automorphisms of G is a subgroup of G*. Since automorphisms are bijective linear transformations, then Aut(G) is a group of invertible matrices (that is matrix representations of the change of bases). Since |G| 3, then Aut(G) is a group of n n matrices with entries from 2 where n 2. Since more than 1 invertible matrix exists for any n 2, then |Aut(G)| 1, a contradiction to our assumption (*). |G| 2. 2.69 Prove that if G is a group for which G/Z(G) is cyclic, where Z(G) denotes the center of G, then G is Abelian. Hint: If G/Z(G) is cyclic, prove that a generator gives an element that is possibly outside of Z(G) which commutes with each element of G. Proof: Assume G/Z(G) is cyclic. Then for some a G, G/Z(G) = aZ(G) = {Z(G), aZ(G), (aZ(G))2, }. And since Z(G) G, then (aZ(G))n = an(Z(G))n = anZ(G). Let x, y G.

Page 2

Then for some integers i and j, xZ(G) =(aZ(G))i =aiZ(G) and yZ(G)=(aZ(G))j =ajZ(G). Since 1 Z(G) and x = x 1, then x xZ(G). Thus, x = aic and y = ajd where c, d are elements in Z(G). As c, d Z(G), then xy = (aic)(ajd) = ai(caj)d = ai(ajc)d = (aiaj)(cd) = (ai+j)(cd) = (aj+i)(dc) = (ajai)(dc) = aj(aid)c = aj(dai)c = (ajd)(aic) = yx. G is Abelian.

Page 3

2.71 Let G be a finite group with K G. If (|K|, [G:K]) = 1, prove that K is the unique subgroup of G having order |K|. Proof: Let |K| = m, then by Lagranges Theorem, |G| = mn for some positive integer n and [G:K] = n. Assume (m, n) = 1. Suppose G has another subgroup H that has order m. Let h H, h 1. Then (hK) | n since hK G/K and |G/K| = n. Let d = (h). Then K G K = hdK = (hK)d, hence (hK) | m as (h) | |H|. By assumption, (m, n) = 1. So (hK) = 1. hK K. H K and |H| = |K|. H = K. K is the unique subgroup of G having order |K|.

Page 4

2.76 If H and K are normal subgroups of a group G with HK = G, prove that G/(HK) (G/H) (G/K). Hint: If : G (G/H) (G/K) is defined by x (xH, xK), then ker = HK; moreover, we have G = HK so that aH = HK = bK .
a b

Proof: (xH, xK). We will show (1) is a homomorphism, Define : G (G/H) (G/K) by x (2) (G) = (G/H) (G/K), and (3) ker = H K. (1) Let x, y G, and note that normalcy gives us that xHyH = xyH and xKyK = xyK. Then (x)(y) = (xH, xK)(yH, yK) = (xHyH, xKyK) = (xyH, xyK) = (xy). is a homomorphism. (2) We will show containment in both directions. Let (X, Y) (G). Then X = xH and Y = xK for some x G. (x) = (xH, xK) = (X, Y) (G/H) (G/K). (G) (G/H) (G/K). Let (X, Y) (G/H) (G/K). Then X = aH and Y = bK for some a, b G. Note that since HK = G, then a = hk and b = hk for some h H, h H, k K, and k K. Also note that normalcy gives us that hkH = h(Hk) = (hH)k = Hk = kH. (aH, bK) = (hkH, hkK) = (hkH, hK) = (kH, hK). Let x = hk, then (xH, xK) = (hkH, hkK) = (kH, hK) = (aH, bK). (X, Y) (G). (G/H) (G/K) (G), as desired. (G) = (G/H) (G/K). (3) Again, we will show containment in both directions. Let x ker . Then (x) = (xH, xK) = (H, K). xH = H and xK = K. x H and x K, hence x H K. Let x H K. Then x H and x K. (x) = (H, K). ker = H K. And by the 1st Isomorphism Theorem, G/(HK) (G/H) (G/K).

Page 5

Debra Griffin Math 210A HW #6, Direct Products Homework Handout #1 4 Let G be a group with H G and K G. Define HK = {hk| h H, k K}. 1. Prove that the following are equivalent. (a) HK G. (b) HK is closed under the binary operation of G restricted to HK. (c) HK = KH. Proof (a) (b): Assume HK G. Then HK is closed under the binary operation of G restricted to HK by definition of subgroup. Proof (b) (c): Assume HK is closed under the binary operation of G restricted to HK. We will show HK KH and HK KH. Let x KH, then x = kh where h H, k K. Since H G and K G, then e H and e K. he = h HK and ek = k HK. x = kh HK by assumption of closure. KH HK. For the reverse inclusion, let y HK. Then y = hk for some h H and some k K. We know k1h1 KH, and since k1h1 = (hk)1 = x1, then x1 KH. But also k1 = ek1 HK and h1 = h1e HK, hence k1h1 HK by assumption of closure. So x1 = hk for some h H and k K. Since x = (x1)1 = (hk)1 = k1h1 KH, then HK KH. HK = KH. Proof (c) (a): Assume HK = KH. We will show HK is nonempty and then apply the subgroup test to show HK G. Since e H and e K, then e = e e HK. HK is nonempty. Let x, y HK. Then x = h1k1 for some h1 H and some k1 K. And y = h2k2 for some h2 H and some k2 K. So xy1 = h1k1(h2k2)1 = h1k1k21h21. Let k3 = k1k21. Clearly k3 K. Since KH = HK, then k3h21 = h3k4 for some h3 H and some k4 K. Let h4 = h1h3. Clearly h4 is in H. And now, we shall put it all together: xy1 = h1k1(h2k2)1 = h1k1k21h21 = h1k3h21 = h1h3k4 = h4k4 HK. HK G.

2. Prove that if H G or K G, then HK G. Proof: We will show HK and then apply the subgroup test. Since e H and e K, then e = e e HK. HK is nonempty. Let x, y HK. Then x = h1k1 for some h1 H and some k1 K. And y = h2k2 for some h2 H and some k2 K. So xy1 = h1k1(h2k2)1 = h1k1k21h21. Let k3 = k1k21. Clearly k3 K. So xy1 = h1k3h21. If H G then we have ghg1 H, g G, h H. If K G then we have gkg1 K, g G, k K. Either way, we have k3h21 = h3k4 for some h3 H and some k4 K. Let h4 = h1h3, also clearly in H. And now we put it all together. xy1 = h1k1(h2k2)1 = h1k1k21h21 = h1k3h21 = h1h3k4 = h4k4 HK. HK G.

3. Prove that if H G and K G, then HK G. Proof: Assume H G and K G. We have show in Exercise #2 that HK G, so we only need to establish normalcy. Let g G, h H, and k K. Since g G, (ghg1 H, h H, and gkg1 K, k K), then ghg1gkg1 = ghkg1 HK. HK G.

4. Suppose H and K are finite subgroups of G. (a) For each x = hk HK, define : H K {(h, k) H K | hk = x} by (d) = (hd, d1k). Prove that is a bijection. Proof: Let x = hk HK. We first show is injective. Let m, n H K such that (m) = (n). Then (hm, m1k) = (hn, n1k). hm = hn and m1k = n1k. Note that m, n H K m, n H, and m1, n1 K. We have cancellation for hm = hn which gives us m = n. And we have cancellation for m1k = n1k, which also gives us m = n. is injective. And now we show is surjective. Let (a, b) {(h, k) H K | hk = x}. Then ab = hk. Since h G, k G, and G is a group, then b = a1hk and a = hkb1. Since a H, then hkb1 H. And since h H, then kb1 H. But k K and b K, which implies b1 K, hence kb1 K. kb1 H K. Similarly a1h H K. Let d = kb1. We will show that d1 = a1h. Consider a1hkb1 = a1abb1 = 1 = kk1h1h = k(hk)1h = k(ab)1h = kb1a1h. a1h = (kb1)1 and (kb1) = (hkb1, a1hk) = (a, b), hence, is surjective.

(b) Prove that if H K = {e}, then each element of HK can be expressed uniquely as a product hk where h H and k K. Proof: Assume H K = {e}. Let x = hk HK. Suppose hk = hk for some h H and k K. Then h1h = k1k. Since h1h H and k1k K, then by assumption, e = h1h = k1k. h = h and k = k. x can be expressed uniquely as a product hk where h H and k K.

(c) Prove |HK| =

HK . H K

Proof: Define f : H K HK by f (h, k) = hk = x, clearly a surjection. each x HK, we have f 1 (x) = {(h, k) H K | hk = x}. Then for By part (a), for each x = hk HK, there is a bijection : H K {(h, k) H K | hk = x} defined by (d) = (hd, d1k). For each x, | f 1(x)| = |H K|. And since H K =

x HK

f 1 ( x ) where each f 1 (x) is disjoint, then

|H K| = | x HK f 1 ( x ) | = |HK| | f 1(x)|. |H| |K| = |H K| = |HK| |H K|.


Hence, |HK| =

HK . H K

Debra Griffin Math 210A HW #7, 2.7 p. 112 #81, 83, 84, 85, 89, 90, 94, 98, 99 2.81 Prove that if p is prime and G is a finite group in which every element has order a power of p, then G is a p-group (i.e |G| = pn for some n Z +). (A possibly infinite group G is called a p-group if every element in G has order a power of p.) Elegant proof (mostly by Mark Roberts): We will prove the result by contradiction. Let p be a prime. Assume every element has order a power of p, and suppose G is not a p-group. If |G| = 1, then were done, so well assume |G| = pnm, (p, m) = 1, n 0, m > 1. Then, q, a prime, such that q p and q | |G|. And by Cauchys theorem, G contains an element of order q, a contradiction to our assumption that every element has order a power of p. If p is prime and G is a finite group in which every element has order a power of p, then G is a p-group. Alternate proof (less elegant): Let g G. Then (g) = pn for some n Z +. Then |G| = pnm for some m Z +, by corollary to Lagranges theorem. By the Fundamental Theorem of Arithmetic,
r2 m = q1r1 q2 qtrt where each qiri are prime.

By Cauchys Theorem, G contains an element of order qi for each integer i {1 i t}.

But each element of G has order a power of p. m = pw for some w Z +, hence |G| = pnpw = pn+w, a power of p. And G is a p-group.

Page 1

2.83 Show that S4 has a subgroup isomorphic to D8. Proof: D8 = {r0, r1, r2, r3, h, v, d, t} where r0, r1, r2, r3 are rotations of a square, (r0 has order 1, the others, order 4) and h, v, d, t are, respectively, horizontal, vertical, diagonal (1), and diagonal (2) flips, each of order 2. Let H = h. Then H D8, and [D8:H] = 8/2 = 4. By Representation on Cosets, there is a homomorphism : D8 S4 with ker H and by the First Isomorphism Theorem, D8/ker ( D8) S4. We know |H| = 2 and H is not normal to D8, while ker H and ker D8. ker = {e}. D8/ker = D8/{e}. It is easily shown that D8/{e} D8. So we have D8 D8/ker ( D8) S4, as desired.

2.84 Prove that S4/V S3. Hint. Use Proposition 2.90 (If G is a group of order 6, then G Z 6 or G S3.) Proof: Since S4 is a group and V S4, then S4/V is a group. And |S4/V| = | S4| / | V| = 24 / 4 = 6. We know the following about the 24 elements of S4, the 6 elements of S3, and the 6 elements of Z 6: S4 No. Cycle Structure 1 (1) 6 (12) 8 (123) 6 (1234) 3 (12)(34) S3 No. Cycle Structure 1 (1) 3 (12) 2 (123) Z6 Equiv. Class [0] [1] [2] [3] [4] [5]

Order 1 2 3 4 2

Order 1 2 3

Order 1 6 3 2 3 6

If is the natural homomorphism from S4 to S4/V (i.e. () = V), then S4, (()) | (). So, each element V of S4/V ( S4) has order 1, 2, 3, or 4. Since [1] Z 6, [1] has order 6, and S4/V has no element of order 6, then ~( S4/V Z 6). By Proposition 2.90, S4/V S3.

Page 2

2.85 (i) Prove that ~(A4 D12). Hint. Recall that A4 has no element of order 6. Proof: We know the following about the 12 elements of A4 and D12: A4 No. Cycle Structure 1 (1) 8 (123) 3 (12)(34) D12 (r = rotation, f = flip) Elmt Order Elmt Order 1 1 f 2 r1 6 rf 2 2 2 r 3 rf 2 r3 2 r3 f 2 4 4 r 3 rf 2 r5 6 r5 f 2

Order 1 3 2

Since D12 contains an element of order 6, but A4 does not, then ~(A4 D12).

(ii) Prove that D12 S3 Z2. Hint. Each element x D12 has a unique factorization of the form x = bia, where b6 = 1 and a2 = 1. Proof: We know the following about the elements of D12 and the elements of S3 Z2: D12 (r = rotation, f = flip) S3 Z2. Elmt Order Elmt Order Elmt Order Elmt Order 1 1 f 2 ((1),[0]) 1 ((23),[0]) 2 r1 6 rf 2 ((1),[1]) 2 ((23),[1]) 2 2 2 r 3 rf 2 ((12),[0]) 2 ((123),[0]) 3 r3 2 r3 f 2 ((12),[1]) 2 ((123),[1]) 6 r4 3 r4 f 2 ((13),[0]) 2 ((132),[0]) 3 5 5 r 6 rf 2 ((13),[1]) 2 ((132),[1]) 6 Since D12 and S3 Z 2 have the same number of elements of each order d | 6, then D12 S3 Z 2. 2.89 (i) If H is a subgroup of G and if x H, prove that CH(x) = H CG(x). Proof: If H acts on itself by conjugation then CH(x) = {h H| hxh1 = x}. If G acts on itself by conjugation then CG(x) = {g G| gxg1 = x}. If x H, then x G, hence both CH(x) and CG(x) are defined. We will prove the result by showing containment in both directions. To show CH(x) H CG(x), we let h CH(x). Then hxh1 = x. Since h H G, then h CG(x). So CH(x) H CG(x). For the reverse containment, let g H CG(x).

Page 3

Then g H and gxg1 = x. g CH(x). Hence CH(x) = H CG(x). 2.89 (ii) If H is a subgroup of index 2 in a finite group G and if x H, prove that |xH| = |xG| or |xH| = (1/2)|xG|, where xH is the conjugacy class of x in H. Hint. Use the second isomorphism theorem (If H, K G, H G, then HK G, HK K and K/(HK) HK/H.) Proof: Let H G such that [G:H] = 2. If H acts on itself by conjugation then xH = {hxh1 | h H}. If G acts on itself by conjugation then xG = {gxg1 | g G}. If x H, then x G, hence both xH and xG are defined. By corollary to Theorem 2.98 (If G acts on a set X and x X, then |O (x)| = [G:Gx]), |xH| = [H:CH(x)] = |H|/|CH(x)|, hence |H| = |xH| |CH(x)|, and |xG| = [G:CG(x)] = |G|/|CG(x)|, hence |G| = |xG| |CG(x)|. [G:H] = |G|/|H| =

x G CG ( x ) x
H

CH ( x)

= 2. Hence |xH| =

1 G CG ( x ) . x 2 CH ( x)

Since [G:H] = 2, then H is normal. And since CG(x) G and CH(x) G, then we can apply the Second Isomorphism theorem as follows: HCG(x) G. H CG(x) G, and CG(x)/( H CG(x)) (HCG(x))/H. By part (i) we know CH(x) = H CG(x). So CG(x)/CH(x) (HCG(x))/H. This gives us that |CG(x)/CH(x)| = |(HCG(x))/H|. And so we have H HCG(x) G. Let |H| = p, |HCG(x)| = q, and |G| = r. We know r = 2p by assumption, and by Lagranges theorem, q = pm, and r = pmn (m, n Z ), hence 2p = pmn. 2 = mn, the only possibilities for which are m = 2, n = 1 or m = 1, n = 2. |(HCG(x))/H| = 2 or |(HCG(x))/H| = 1. |xH| = or |xH| =

1 G CG ( x ) 1 G HCG ( x ) 1 G x = x = x 2 = xG , 2 CH ( x) 2 H 2 1 G CG ( x ) 1 G 1 x = x 1 = xG . 2 CH ( x) 2 2

Page 4

2.89 (iii) Prove that there are two conjugacy classes of 5-cycles in A5, each of which has 12 elements. Hint. If = (12345), 120 then CS5 ( ) = 5 because 24 = ; hence CS5 ( ) = . What is CA 5 ( ) ? CS5 ( ) Proof: Let = (12345). By Thm 2.9 (All permutations and in Sn have the same cycle structure iff Sn with = 1.) we know all 5-cycles are conjugate to (12345) in S5. S5 = (5 4 3 2 1)/5 = 24. And |S5| = 5 4 3 2 1 = 120. Since S5 =

S5 CS5 ( )

, then | CS5 ( ) | = 5.

This gives us that CS5 ( ) = as 5 is prime.


2 And we know , , 3, and 4 are also 5-cycles, hence, even. Since CA 5 ( ) = A5 CS5 ( ) , then CA 5 ( ) = , which gives us | CA 5 ( ) | = 5. Note that |A5| = 1/2 |S5| = (1/2) 5!.

A A A Thus, knowing |A5| = (1/2)5! = 60, we can apply 5 = 5 , to get that 5 = 12. CA 5 ( )

Since S5 = 24, then |S5 A 5 | = 12. Let S5 A 5 where is a 5-cycle. By similar proof, | CS5 ( ) | = 5, hence A 5 = 12.
A A 5 = S5 and A 5 A 5= , then A 5 and A 5 are the two conjugacy And since 5

classes of 5-cycles in A5, each of which has 12 elements.

Page 5

(iv) Prove that the conjugacy classes in A5 have sizes 1, 12, 12, 15, and 20. Proof: A5 No. Cycle Structure 1 (1) 20 (123) 24 (12345) 15 (12)(34) Order 1 3 5 2 We will show the conjugacy classes of A5 are as follows: | (1) A 5 | = 1 If = (12345), then | A 5 | = 12. If is a 5-cycle in S5 A 5 , then | A 5 | = 12. If = (12)(34), then A 5 = 15. And if = (123), then A 5 = 20.

We have | (1) A 5 | = 1 as A5, (1) 1 = 1 = (1) , and from part (iii), we have | A 5 | = 12 and | A 5 | = 12. We will show A5 has exactly 2 more conjugacy classes and their sizes are 15 and 20. Let = (12)(34). By Thm 2.9 (stated above) we know all permutations of structure (12)(34) are conjugate in S5. | S5 | = (5 4)(3 2)/8 = 15. to (12)(34) S5 120 And by Thm 2.98 (|O (x)| = [G:Gx]), | S5 | = = . | CS5 ( ) | = 8. CS5 ( ) CS5 ( ) We can easily find these 8 elements. Its a subgroup, so it must contain (1). Since 1 = , then CS5 ( ) . Since (12) and (34) are factors of , then it is easily checked that (12), (34) CS5 ( ) . Since CS5 ( ) is a group, then (12) and (34) CS5 ( ) . Since (13)(24) and (14)(23) fix 5, then it is quickly checked that they are also in C S 5 ( ) . So CS5 ( ) = {(1), (14)(23), (12), (34), (12)(12)(34), (34)(12)(34)}. (12)(34), (13)(24), even even even even odd odd odd odd Since CA 5 ( ) = A5 CS5 ( ) , then CA 5 ( ) = {(1), (12)(34), (13)(24), (14)(23)}, hence A5 | CA 5 ( ) | = 4. Since | A 5 | = , and |A5| = (1/2)5! = 60, then | A 5 | = 15. Since there CA 5 ( )

are exactly of in S5, then A 5 is the only conjugacy class of in A5, and it 15 conjugates has order 15. By Thm 2.9 (stated above) we know all Let = (123). 3-cycles are conjugate to (123) in S5 S5. | | = (5 4 3)/3 = 20. S5 120 And by Thm 2.98 (|O (x)| = [G:Gx]), | S5 | = = . | CS5 ( ) | = 6. CS5 ( ) CS5 ( ) We can easily find these 6 elements. Its a subgroup, so it must contain (1). Since 1 = , then CS5 ( ) . So 1 = (132) CS5 ( ) . Any cycle , disjoint to will then (123)(45) and (132)(45) give us 1 = , so (45) CS5 ( ) . Since CS5 ( ) is a group, are in CS5 ( ) . So CS5 ( ) = {(1), (123), (132), (45), (123)(45), (132)(45)}
Page 6

even even even odd

odd

odd

CA 5 ( ) = A5 CS5 ( ) , so CA 5 ( ) = {(1), (123), (132)}, hence | CA 5 ( ) | = 3. A5 Since A 5 = , and |A5| = (1/2)5! = 60, then A 5 = 20. Since there are exactly 20 CA 5 ( )

in S5, then A 5 is the only conjugacy class of 3-cycles in A5, and it has order 20. 3-cycles

2.90 (i) Prove that every normal subgroup H of a group G is a union of conjugacy classes of G, one of which is {1}. Proof: Let G be a group. Let H G.
G Let K = x G i where xi H for each i and each x i is a distinct conjugacy class of xi in G.
i

We will show H = K.

Let h H. Then g G, ghg1 H as H G. So ghg1 = xi for some xi H. Hence h = g1xig = kxik1 where k = g1 (which we know is in G).
1 h xG i = {gxig : g G}. Thus, h K, which gives us H K.

For the reverse containment, let k K. Then k x G i for some xi H.

So k = gxig1 for some g G. But gxig1 H since H G. k H, hence K H. Thus H = K.

Note that 1 H and 1G = {g 1 g1 : g G} = {1}, so 1G = {1} K, as desired.


(ii) Use part (i) and Exercise 2.89 to give a second proof of the simplicity of A5. Proof: Let H A5. Then, by part (i) above, H =

x
i

A5 i

{1} where xi H for each i and each

x iA 5 is a distinct conjugacy class of xi in A5.


By Exercise 2.89 part (iv), conjugacy classes in A5 have sizes 1, 12, 12, 15, and 20. So |H| {1, 13, 16, 21, 25, 28, 33, 36, 40, 45, 48, 60}. Since H A5, then |H| | |A5| = 60. Thus |H| = 1 or |H| = 60. H = {(1)} or H = A5, hence A5 is simple.

Page 7

2.94 (i) Prove, for all n 5, that the only normal subgroups of Sn are {(1)}, An and Sn. Proof: Let H Sn. Case 1: If H = {(1)}, then we are done. Case 2: Assume H {(1)} and suppose H An. Then H = H An {(1)}. Since H Sn and An Sn, then by the 2nd Isomorphism theorem (If H G, K G, then HK G, H K K, and K/(H K) HK/H.), H An An. And since An is simple n 5, and H An {(1)}, then H An = An. Hence H = An. Case 3: Assume H {(1)} and suppose ~( H An). Then H such is odd. Since H Sn, then Sn, 1 H. Also, 1 H. Hence 11 H and 11 is even, hence 11 H An. If 11 = (1), then = , Sn. And by Exercise 2.14, = (1), a contradiction to our assumption that is odd. H An {(1)}. And since H An = An, we have that An H. Hence |An| < |H|. But [Sn:An] = 2. So if we apply Lagranges theorem, [Sn:H] = 1. H = Sn. Each case gives us that H is either {(1)}, An, or Sn. (ii) Prove that if n 3, then An is the only subgroup of Sn of order (1/2)n!. Hint. If H is a second such subgroup, then H is normal in Sn and hence H An is normal in An. Proof: Suppose H Sn and |H| = (1/2)n!. Let n = 3. Then |H| = (1/2) 6 = 3. Since H S3, then H must contain all conjugates of its elements. H = A3, (otherwise {(1), (12), (13), (23)} H, which exceeds the order of 3). Let n = 4. Then |H| = (1/2) 24 = 12. Recall the following about the elements of S4: S4 No. Cycle Structure 1 (1) 6 (12) 8 (123) 6 (1234) 3 (12)(34) Order 1 2 3 4 2 Since H must contain all conjugates of its elements, then the only combination of 1, 6, 8, 6, 3 that includes 1 and produces a sum of 12 is 1, 8, and 3. Hence H = A4.

Let n 5. By part (i) above, H = An.

Page 8

2.94 (iii) Prove that S5 has no subgroup of order 30. Proof: Suppose H S5 and |H| = 30. Case 1: Suppose H A5. Then [A5:H] = 2, hence H A5. Since A5 is simple, H = {(1)} or H = A5, hence |H| = 1 or |H| = 60. This is a contradiction to our assumption that |H| = 30. Case 2: Suppose ~( H An). H such that is odd. We first will prove that |H/H An| = 2. Define f : H Z 2 by f () = 0 if is even and f () = 1 if is odd. Then it is easily checked that f is a surjective homomorphism. Since ker f = { H| is even} = H An, then, by the 1st Isomorphism theorem, H/H An Z 2. Thus |H/H An| = |Z 2| = 2. So |H/H An| = 30 / |H An| = 2, hence |H An| = 15. This gives us [A5 :H An] = 60/15 = 4. But |A5| = 60 | 24 = 4!. By the index factorial theorem, this implies A5 is not simple, which we have proven is not true. Hence we have a contradiction. S5 has no subgroup of order 30. (iv) Prove that S5 contains no subgroup of order 40. Proof: Suppose H S5 and |H| = 40. Since 40 | 60, then |H| | |A5|, hence ~(H A5). H such that is odd. We first will prove that |H/H An| = 2. Define f : H Z 2 by f () = 0 if is even and f () = 1 if is odd. Then it is easily checked that f is a surjective homomorphism. Since ker f = { H| is even} = H An, then, by the 1st Isomorphism theorem, H/H An Z 2. Thus |H/H An| = |Z 2| = 2. So we can use this to determine that |H/H An| = 30 / |H An| = 2, hence |H An| = 15. This gives us [A5 :H An] = 60/15 = 4. But |A5| = 60 | 24 = 4!. By the index factorial theorem, this implies A5 is not simple, which we have proven is not true, multiple times! Hence we have a contradiction. S5 has no subgroup of order 30.

Page 9

2.98 Let G be a group with |G| = mp, where p is a prime and 1 < m < p. Prove that G is not simple. Hint. Show that G has a subgroup H of order p, and use the representation of G on the cosets of H. Proof: By Cauchys theorem, a G such that (a) = p. Thus a G and |a| = p. By Lagranges theorem, [G:a] = pm/p = m. Since m < p and p is prime, then |G| = pm | m!. So, by the index factorial theorem, G is not simple. 2.99 (i) Let a group G act on a set X, and suppose that x, y X lie in the same orbit: y = g x for some g G. Prove that Gy = gGxg1. Proof: Let a Gy, then a y = y. Since y = g x for some g G, then (ag) x = a g x = g x. Since g1 G, then g1 (ag) x = g1 g x . which gives us (g1ag) x = x. g1ag Gx. Thus gg1agg1 = a gGxg1. So Gy gGxg1. For the reverse containment, let b gGxg1. Then b = ggg1 for some g Gx. Thus g x = x. Note that since y = g x for some g G, then g1 y = g1 g x = x. g g1 y = g1 y. Thus (ggg1) y = g (gg1) y = g g1 y = y. ggg1 Gy. b Gy. gGxg1 Gy. And thus we have Gy = gGxg1. (ii) Let G be a finite group acting on a set X; prove that if x, y X lie in the same orbit, then |Gx| = |Gy|. Proof: Since x, y X lie in the same orbit and orbits partition X, then O (x) = O (y), hence |O (x)| = |O (y)|. By Theorem 2.98 (If G acts on a set X and x X, then |O (x)| = [G:Gx]), we have |Gx| = |G|/|O (x)| = |G|/|O (y)| = |Gx|

Page 10

Debra Griffin Math 210A HW #8 Finite Abelian Groups #1 16 1. Prove that Z n Z m Z nm if and only if (n, m) = 1. Proof: Assume gcd (n, m) = 1. Let ([1]n, [1]m) = t. Then n | t and m | t. But (n, m) = 1 nm | t. Since ([1]n, [1]m)nm = ([0]n, [0]m), then t | mn, hence t = nm. Z n Z m is a cyclic group of order nm and thereby isomorphic to Z nm. Conversely, assume Z n Z m Z nm. Since Z nm contains an element of order nm, then Z n Z m also contains an element of order nm by Exercise 2.42 (If G has an element of some order n and H does not, then G is not isomorphic to H.). Let Z n = [x]n and Z m = [y]m. Assume (n, m) = d > 1. Then n = dr and m = ds for some integers r, s, and drs < mn. Then for any integers a, b, ([xa ]n, [yb ]m)drs = ([xadrs ]n, [ybdrs ]m) = ([xasn ]n, [ybrm ]m) = ([0]n, [0]m). The order of ([xa ]n, [yb ]m) is a divisor of drs, and, hence, strictly less than mn. And so every element in Z n Z m has order at most drs, a contradiction to our assumption that Z n Z m Z nm. (n, m) = 1. 2. Prove that Z n Z m is cyclic if and only if (n, m) = 1. Proof: Define Z n = [x]n and Z m = [y]m and suppose for some integers a, b, [xa ]n generates [x]n and [yb ]m. Let (n, m) = d. Then n = dr and m = ds for some integers r, s. Thus the order of ([xa ]n, [yb ]m) is a divisor of drs. Note that drs < mn if d 1. So then ([xa ]n, [yb ]m) generates Z n Z m , a group of order mn if and only if (n, m) = 1.

Page 1

3. Let H1, , Hk be subgroups of a group G such that (|Hi|, |Hj|) = 1 for all i j in {1, , k}. Prove that |H1H2Hk| = |H1| |H2| |Hk|. Proof: Let G be a finite Abelian group. We will induct on k. Let k = 2. Note that by Exercise 2.33 (If HG, KG, and (|H|, |K|) = 1, then H K = {1}.), we have Hi Hj, for all i j in {1, , k}. And by Direct Product Handout, Exercise #4(c), (If HG, KG, then |HK| = |H||K|/|HK|), we have |H1H2| = |H1||H2|/|H1 H2| = |H1||H2|. Let k > 2 and assume that if H1, , Hk are subgroups of G, then |H1H2Hn| = |H1| |H2| |Hn| n < k. Then if H1, , Hk1 are subgroups of G, we have |H1H2Hk1| = |H1| |H2| |Hk1|. Note that H1H2Hk1 G as G is Abelian. And since (|Hi |, |Hj |) = 1 for all i j, then (|H1H2Hk1|, |Hk|) = 1, hence H1H2Hk1 Hk = {1}, Thus |H1H2Hk1Hk| = |H1H2Hk1||Hk| = |H1| |H2| |Hk1||Hk|. |H1H2Hk| = |H1| |H2| |Hk| for all i j in {1, , k}. 4. Let G be a finite Abelian group of order pka where (p, a) = 1 and let k G(p) = {x G| x p = e}. Prove that G(p) G. Proof: k Clearly, G(p) G. And, since e p = e, then e G(p), hence G(p) . pk k k Let x, y G(p). Then x p = e, and y p = e. Since G is Abelian, then ( xy 1 ) =

( x)

pk

( y 1)

pk

= ( x)

pk

[ ]
y) (
pk

= e e1 = e. xy1 G(p), hence G(p) G.

5. Let G be a finite Abelian p-group and a an element of maximal order in G. If (a) = pk, pk x prove that = e for all x G.
Proof: Assume |G| = pm for some integer m. Let x G. Then (x) | pm. By Cauchys theorem, (x) {1, p, p2, , pm}, hence (a) = pk {1, p, p2, , pm}. Since (a) is maximal, then (x) (a). (x) {1, p, p2, , pk}. So (x) = pki where 0 i k. e = xp = xp
ki

( )
ki

pi

= xp

ki

pi

= x p , as desired.

Page 2

k2 l 6. Let G1 and G2 be Abelian groups of order p1k1 p2 pk . Prove that G1 G2 l if and only if G1(pi) G2(pi) for all i {1, 2, , l}.

Proof: Assume G1 G2. Then : G1 G2 where is an isomorphism.


k2 l iki p k Since G2(pi) G2 and [G2: G2(pi)] = p1k1 p2 p , then, by the Correspondence l

theorem, H G1 such that H/ker G2(pi). So |H| = piki . By Lemma 4 (If G is an Abelian group such that |G| = pka where p is prime and (p, a) = 1, then there exists a unique subgroup of order pk.) H = G1(pi). Thus G1(pi)/ker G2(pi) and since is injective, ker = {e}. Hence G1(pi) G2(pi) for all i {1, 2, , l}, as desired. Assume G1(pi) G2(pi) for all i {1, 2, , l}. We will proceed by induction on l. Suppose l = 2. Then there are isomorphisms, : G1(p1) G2(p1) and : G1(p2) G2(p2). Define f : G1(p1) G1(p2) G2(p1) G2(p2) by f (a, b) = ((a), (b)). Since and are bijective then f is bijective as well. To show f is a homomorphism, let (a, b) and (c, d) G1(p1) G1(p2). Then f (ac, bd) = ((ac), (bd)) = ((a)(c), (b) (d)) =((a), (b))((c), (d)) = f ((a, b))f ((c, d)). Assume l > 2 and k < l, G1(p1) G1(pk) G2(p1) G2(pk). Then, by similar proof we have G1(p1) G1(pl) G2(p1) G2(pl). Hence G1 G2 as desired.

Page 3

7. Given n1 n2 nr and m1 m2 ms where n1 + n2 + + nr = m1 + m2 + + ms, p is a prime, and Ct is a cyclic subgroup of order t, then we have that Cp n 1 Cp n 2 Cp nr Cp m 1 Cp m 2 Cp m s if and only if r = s and ni = mi for all i). Proof: Cp n 2 Cp nr Cp m 1 Cp m 2 Cp m s . Assume C p n 1 We know C p n 1 and C p m 1 are cyclic, n1 is maximal in {n1, n2, , nr }, m1 is maximal in {m1, m2, , ms}, and C p n 1 = a and C p m 1 = b where (a) = p n1 and (b) = p m1 .

So p n is the maximal order of all elements in C p n 1 C p n 2 C p n r and maximal order of all elements in C p m 1 C p m 2 C p m s . p m is the C C C C C . Hence p n = p m as C
1 1 1 1

pn1

pn 2

pnr

pm1

pm 2

pm s

Cp n 1 Cp m 1 .

/Cp n 1 Cp n 2 Cp n r And since C p n 1 C p n 2 C p n r C p m 2 C p m s , then Cp m 2 Cp m s /Cp m 1 and C p m 1


n 2 C nr C m 2 C m s . C p p p p can repeat this process until We we have exhausted all pairs. There must be the same number of C p n i ' s as C p m i ' s . Hence r = s.
And since each C p n i C p m i , then ni = mi for every i. To prove the converse, assume r = s and ni = mi for all i. Then since C p n i and C p m i are each i, and | C p n i | = p n i = p m i = | C p m i |, then cyclic for And since r = s, then C p n 1 C p n 2 C p n r C p m 1 C p m 2 C p m s .

Cp ni C p m i .

Page 4

8. Find all, up to isomorphism, Abelian groups of order (a) 105 Let |G| = 105. 105 = 3 5 7. By the Fundamental Theorem for Finite Abelian Groups, G Z 3 Z 5 Z 7. (b) 270 Let |G| = 270. 270 = 2 33 5. By the Fundamental Theorem for Finite Abelian Groups, G Z 2 Z 3 Z 3 Z 3 Z 5 or G Z 2 Z 9 Z 3 Z 5 or G Z 2 Z 27 Z 5. (c) 9801 Let |G| = 9801. 9801 = 34 112. By the Fundamental Theorem for Finite Abelian Groups, G Z 3 Z 3 Z 3 Z 3 Z 11 Z 11 or G Z 9 Z 3 Z 3 Z 11 Z 11 or G Z 27 Z 3 Z 11 Z 11 or G Z 81 Z 11 Z 11 or G Z 9 Z 9 Z 11 Z 11 or G Z 3 Z 3 Z 3 Z 3 Z 121 or G Z 9 Z 3 Z 3 Z 121 or G Z 27 Z 3 Z 121 or G Z 81 Z 121 or G Z 9 Z 9 Z 121. (e) 44100 Let |G| = 44100. 44100 = 2 32 5 72. By the Fundamental Theorem for Finite Abelian Groups, G Z 2 Z 3 Z 3 Z 5 Z 7 Z 7 or G Z 2 Z 3 Z 3 Z 5 Z 49 or G Z 2 Z 9 Z 5 Z 7 Z 7 or G Z 2 Z 9 Z 5 Z 49. (d) 320 Let |G| = 320. 320 = 26 5. By the Fundamental Theorem for Finite Abelian Groups, G Z 2 Z 2 Z 2 Z 2 Z 2 Z 2 Z 5 or G Z 4 Z 2 Z 2 Z 2 Z 2 Z 5 or G Z 8 Z 2 Z 2 Z 2 Z 5 or G Z 16 Z 2 Z 2 Z 5 or G Z 32 Z 2 Z 5 or G Z 64 Z 5 or G Z 4 Z 4 Z 2 Z 2 Z 5 or G Z 4 Z 4 Z 4 Z 5 or G Z 4 Z 4 Z 4 Z 5 or G Z 8 Z 4 Z 2 Z 5 or G Z 16 Z 4 Z 5.

Page 5

9. What is the smallest positive integer n such that (a) there are two nonisomorphic groups of order n? n = 4, V and Z 4. (b) there are three nonisomorphic Abelian groups of order n? n = 8, Z 8, Z 4 Z 2, and Z 2 Z 2 Z 2, (c) there are exactly four nonisomorphic Abelian groups of order n? n = 36. Z 36, Z 4 Z 3 Z 3, Z 2 Z 2 Z 9, Z 2 Z 2 Z 3 Z 3. 10. Calculate the number of elements of order 2 and or order 4 in each of the following groups: (a) Z 16,
Elmt Order [0] 1 [1] 16 [2] 8 [3] 16 [4] 4 [5] 16 [6] 8 [7] 16 [8] 2 [9] 16 [10] 8 [11] 16 [12] 4 [13] 16 [14] 8 [15] 16

So Z 16 has 1 element of order 2 and 2 elements of order 4. (b) Z 8 Z 2,


Elmt (0, 0) (1, 0) (2, 0) (3, 0) (4, 0) (5, 0) (6, 0) (7, 0) (0, 1) (1, 1) (2, 1) (3, 1) (4, 1) (5, 1) (6, 1) (7, 1) Order 1 8 4 8 2 8 4 8 2 8 4 8 2 8 4 8

So Z 8 Z 2 has 3 elements of order 2 and 8 elements of order 4. (c) Z 4 Z 4 ,


Elmt (0, 0) (1, 0) (2, 0) (3, 0) (0, 1) (1, 1) (2, 1) (3, 1) (0, 2) (1, 2) (2, 2) (3, 2) (0, 3) (1, 3) (2, 3) (3, 3) Order 1 4 2 4 4 4 4 4 2 4 2 4 4 4 4 4

So Z 4 Z 4 has 3 elements of order 2 and 12 elements of order 4. (d) Z 4 Z 2 Z 2,


Elmt Order
(0, 0, 0) (1, 0, 0) (2, 0, 0) (3, 0, 0) (0, 1, 0) (1, 1, 0) (2, 1, 0) (3, 1, 0) (0, 0, 1) (1, 0, 1) (2, 0, 1) (3, 0, 1) (0, 1, 1) (1, 1, 1) (2, 1, 1) (3, 1, 1)

So Z 4 Z 2 Z 2 has 7 elements of order 2 and 8 elements of order 4. (e) Z 2 Z 2 Z 2 Z 2.


Elmt Order
(0, 0, 0, 0) (1, 0, 0, 0) (0, 0, 0, 1) (1, 0, 0, 1) (0, 0, 1, 0) (1, 0, 1, 0) (0, 1, 1, 1) (1, 1, 1, 1) (0, 1, 0, 0) (1, 1, 0, 0) (0, 1, 0, 1) (1, 1, 0, 1) (0, 1, 1, 0) (1, 1, 1, 0) (0, 1, 1, 1) (1, 1, 1, 1)

So Z 2 Z 2 Z 2 Z 2 has 15 elements of order 2 and no elements of order 4.

Page 6

11. There are six groups, up to isomorphism, of order 108. Let G be a group of order 108 = 22 33. So G Z 2 Z 2 Z 3 Z 3 Z 3, or G Z 4 Z 3 Z 3 Z 3, or G Z 2 Z 2 Z 9 Z 3, or G Z 4 Z 9 Z 3, or G Z 2 Z 2 Z 27, or G Z 4 Z 27. Prove that (a) two of them have exactly one subgroup of order 3. Proof: G Z 2 Z 2 Z 27 and G Z 4 Z 27 each have exactly one subgroup of order 3. The subgroup of order 3 for Z 2 Z 2 Z 27 = {(0, 0, 0), (0, 0, x), (0, 0, x2)} where (x) = 3 in Z 27. Since Z 27 is cyclic, there is only one such subgroup. The subgroup of order 3 for Z 4 Z 27 = {(0, 0), (0, x), (0, x2)} where (x) = 3 in Z 27. Since Z 27 is cyclic, there is only one such subgroup.

(b) two of them have exactly four subgroups of order 3. Proof: G Z 2 Z 2 Z 9 Z 3 and G Z 4 Z 9 Z 3 have exactly 4 subgroups of order 3. The subgroups of order 3 for The subgroups of order 3 for Z2 Z2 Z9 Z3 = Z4 Z9 Z3 = {(0, 0, 0, 0), (0, 0, x, y), (0, 0, x2, y2)} {(0, 0, 0), (0, x, y), (0, 0, x2, y2)} where (x) = 3 in Z 9 and (y) = 3 in Z 3. where (x) = 3 in Z 9 and (y) = 3 in Z 3. There are 1 1 2 2 = 4 such subgroups. There are 1 1 2 2 = 4 such subgroups. (c) two of them have exactly 13 subgroups of order 3. Proof: G Z 2 Z 2 Z 3 Z 3 Z 3, G Z 4 Z 3 Z 3 Z 3 have exactly 13 subgroups of order 3. The subgroups of order 3 for The subgroups of order 3 for Z 2 Z 2 Z 3 Z 3 Z 3 ={(0, 0, 0, 0, 0), Z4 Z3 Z3 Z3 = 2 2 2 (0, 0, x, y, z), (0, 0, x , y , z )} {(0, 0, 0, 0), (0, x, y, z), (0, x2, y2, z2)} where (x) = (y) = (z) = 3 in Z 3, or 2 of where (x) = (y) = (z) = 3 in Z 3, or 2 of them have order 3 while one has order 1. them have order 3 while one has order 1. There are (1 1 2 2 2 1) + There are 1 1 2 2 = 4 such subgroups. 3(1 1 1 1 2) = 13 such subgroups. There are (1 1 2 2 2 1) + 3(1 1 1 1 2) = 13 such subgroups.

Page 7

12. How many Abelian groups, up to isomorphism, are there (a) of order pq, where p and q are distinct primes? There is one Abelian group of order pq. Proof: Let G be a group such that |G| = pq, where p and q are distinct primes. By Theorem 6 (If G is a finite Abelian group such that |G| = p1k1 plkl where each of the pi are distinct primes, then G G(p1) G(pi).) we have G G(p) G(q). Since p and q are distinct primes, then G(p) and G(q) are cyclic, hence G(p) Z p and G(q) Z q. By Exercise 1, above (Z n Z m Z nm if and only if (n, m) = 1.), G Z pq. Hence there is only one Abelian group of order pq. (b) of order pqr, where p, q, and r are distinct primes? There is one Abelian group of order pqr. Proof: Similar to part (a). 13. Suppose that G is an Abelian group of order 120 and that G has exactly three elements of order 2. Determine the isomorphism class of G. Proof: 120 = 23 3 5. By the Fundamental Theorem of Finite Abelian Groups, G Z 2 Z 2 Z 2 Z 3 Z 5, or G Z 4 Z 2 Z 3 Z 5, or G Z 8 Z 3 Z 5. An element of order 2 in Z 4 Z 2 Z 3 Z 5 looks like (x, y, 0, 0) where (x) = 2 in Z 4 and (y) = 2 in Z 2. So there are 1 1 + 1 2 = 3 elements of order 2 in G Z 4 Z 2 Z 3 Z 5. However, there is only one element of order 2 in G Z 8 Z 3 Z 5 and there are 7 elements of order 2 in G Z 2 Z 2 Z 2 Z 3 Z 5. The isomorphism class of G is Z 4 Z 2 Z 3 Z 5. 14. Suppose that G is an Abelian group of order 16 with at least one element of order 8 and at least two elements of order 2. Determine the isomorphism class of G. Proof: 16 = 24. By the Fundamental Theorem of Finite Abelian Groups, G Z 2 Z 2 Z 2 Z 2, or G Z 4 Z 2 Z 2, or G Z 4 Z 4, or G Z 8 Z 2, or G Z 16. Since G has at least one element of order 8, then G is not isomorphic to Z 2 Z 2 Z 2 Z 2, Z 4 Z 4, or or Z 4 Z 2 Z 2. And since G has at least 2 elements of order 2, then G cannot be cyclic. Hence the isomorphism class of G is Z 16.

Page 8

15. Suppose that G is an Abelian group of order 16 and that a, b G with (a) = (b) = 4 and a2 b2. Determine the isomorphism class of G. Proof: G Z 2 Z 2 Z 2 Z 2, or G Z 4 Z 2 Z 2, or G Z 4 Z 4, or G Z 8 Z 2, or G Z 16. We can rule out Z 2 Z 2 Z 2 Z 2 as it contains no elements of order 4. We can rule out Z 4 Z 2 Z 2, Z 8 Z 2, and Z 16, as Z 4, Z 8, and Z 16, each have 2 elements of order 4, but a2 = b2 in each case. But in Z 4 Z 4, (1, 0) = 4, (1, 0)2 = (2, 0), (0, 1) = 4 and (0, 1)2 = (0, 2) (2, 0), The isomorphism class of G is Z 4 Z 4. 16. Let G = U(Z 16) = {[1]. [3]. [5], [7], [9], [11], [13], [15]}. Recall that this is a group with respect to multiplication. (a) Determine the isomorphism class of G. The isomorphism class of G is Z 4 Z 2. Proof: By the Fundamental Theorem for Finite Abelian Groups, G Z 2 Z 2 Z 2, or G Z 4 Z 2, or G Z 8. By calculation, we have the following orders of elements of U(Z 16): Elmt [1] [3] [5] [7] [9] [11] [13] [15] Order 1 4 4 2 2 4 4 2 Thus, we can eliminate the isomorphism class of Z 2 Z 2 Z 2 as it has no elements of order 4. And, we can eliminate the isomorphism class of Z 8 as it is cyclic and has only one element of order 2. But Z 4 Z 2 has 2 2 = 4 elements of order 4, and 1 1 + 1 2 = 3 elements of order 2. The isomorphism class of G is Z 4 Z 2. (b) Write G as an internal direct product of cyclic groups. Consider H = [3] = {[1], [3], [9], [11]} and K = [7] = {[1], [7]}. Since G is Abelian, then H G and K G. And we can observe that H K = {[1]}. And by calculation, we have the following multiplication table: [1] [3] [9] [11] [1] [1] [3] [9] [11] [7] [7] [5] [15] [13] G = HK where H and K are cyclic.

Page 9

Debra Griffin Math 210A HW #9 The Sylow Theorems #1 16 Textbook Problems: 5.23 (i) Prove that if d is a positive divisor of 24, then S4 has a subgroup of order d. Proof: We have the following subgroups of S4: 1|24 {(1)} S4 2|24 {(1), (12)} S4 3|24 {(1), (123), (132)} S4 4|24 {(1), (12)(34), (13)(24), (14)(23)} S4 6|24 {(1), (12), (13), (23), (123), (132)} S4 8|24 {(1), (13), (24), (12)(34), (13)(24), (14)(23), (1234), (1432)} S4 12|24 A4 S4 24|24 S4 S4 (ii) If d 4, prove that any two subgroups of S4 having order d are isomorphic. Proof: Note that S4 No. Cycle Order Structure 1 (1) 1 6 (12) 2 8 (123) 3 6 (1234) 4 3 (12)(34) 2 Let Hd denote any subgroup of S4 having order d. 1|24 H1 {(1)} (there is only 1 isomorphism class of order 1) 2|24 H2 Z 2 {(1), (12)} (there is only 1 isomorphism class of order 2) 3|24 H3 Z 3 {(1), (123), (132)} (there is only 1 isomorphism class of order 3) 4|24 H4 V4 {(1), (12)(34), (13)(24), (14)(23)} or H4 Z 4 {(1), (13)(42), (1234), (1432)} 6|24 H6 {(1), (12), (13), (23), (123), (132)} (A group of order 6 is only isomorphic to S3 or Z 6. Since S4 has no element of order 6, then H6 S3.) 8|24 H8 {(1), (13), (24), (12)(34), (13)(24), (14)(23), (1234), (1432)} (Recall Prop 2.58(If G is a group and g G, then conjugation g :G G is an isomorphism.) Since H8 is a Sylow 2-group of S4 and all Sylow 2-subgroups are conjugates, then all Sylow 2-subgroups are isomorphic.) 12|24 H12 A4 S4 (there is only one subgroup of order 12) 24|24 H24 S4 (the orders are the same and H24 S4)

Page 1

5.27 Prove that a Sylow 2-subgroup of A5 has exactly five conjugates. Proof: |A5| = 60 = 22 3 5. n2 = 1, 3, 5, 15 n2|15 n2 1 (mod 2) n2 = 1, 3, 5, 15 n2 = 1, 3, 5, 7, 9, 11, 13, 15 n3 = 1, 4, 10, 20 n3|20 n3 1 (mod 3) n3 = 1, 2, 4, 5, 10, 20 n3 = 1, 4, 7, 10, 13, 17, 20 n5 = 1, 6 n5|12 n5 1 (mod 5) n5 = 1, 2, 3, 4, 6, 12 n5 = 1, 6, 11 n2 = the number of conjugates of a Sylow 2-subgroup of A5. Since A5 is simple, then n2 1, n3 1, and n5 1, hence n5 = 6. If n3 = 4, then [G:H3] = 4 and |G| = 60 | 4!, so n3 4. If n2 = 3, then [G:H2 ] = 3 and |G| = 60 | 3!, so n2 3. If n5 = 6 and n3 = 20, then A5 has 6 subgroups of order 5, and 20 subgroups of order 3. This would require 24 distinct elements of order 5 and 40 distinct elements of order 3. This would exceed the order of A5, 60. So n3 = 10. Now assume n5 = 6, n3 = 10, and n2 = 15. Then A5 has 6 subgroups of order 5, 10 subgroups of order 3 and 15 subgroups of order 4. Since A5 has no elements of order 4, then this would require 24 distinct elements of order 5, 20 distinct elements of order 3, A5 No. Cycle Structure 1 (1) 20 (123) 24 (12345) 15 (12)(34) Order 1 3 5 2 Notice that elements of order 2 in A5 are of the cycle structure, (12)(34).. So, for any Sylow 2-subgroups of A5, only 4 letters of {1, 2, , 5} can be used as using all 5 in one subgroup would create a permutation of cycle structure other that (12)(34). And we can only make 5 such subgroups. n2 15.

Debra Griffin 11/9/09 5:42 AM


Deleted: and 45 distinct elements of order 2. Again, we have exceeded order of A5, 60.

This leaves only one possibility. Assume n5 = 6, n3 = 10, and n2 = 5. Thus A5 has 6 subgroups of order 5, 10 subgroups of order 3 and 5 subgroups of order 4. This would require 24 distinct elements of order 5, 20 distinct elements of order 3, and 15 elements of order 2. These elements with the identity add up to 60, as desired. n5 = 10, which implies that A5 has exactly five conjugates.

Page 2

5.28 Prove that there are no simple groups of order 96, 300, 312, or 1000. Hint. Some of these are not tricky. Proof: Let G be group such that |G| = 96 = 25 3. n2|3 n2 1 (mod 2) n2 = 1, 3 n2 = 1, 3 n2 = 1, 3 n5|32 n5 1 (mod 5) n5 = 1, 16 n5 = 1, 2, 4, 8, 16 n5 = 1, 6, 11, 16 If G is simple, then n2 1, hence n2 = 3. But |G| = 96 | 3!. So by the Index Factorial theorem, G is not simple, a contradiction to our assumption. We have that n2 = 1, hence P2 G where P2 is a Sylow 2-subgroup of G. G is not simple. 5.28 (cont.) Let G be group such that |G| = 300 = 22 3 52. n2 = 1, 3, 5, 15, 25, 75 n2|75 n2 1 (mod 2)
n2 = 1, 3, 5, 15, 25, 75 n2 = 1, 3, 5, , 15, , 25, , 75

n3|100
n3 = 1, 2, 4, 5, 10, 20, 25, 50, 100

n3 1 (mod 3)
n3 = 1, 4, 10, , 25, , 100

n3 = 1, 4, 10, 25, 100 n5 = 1, 6

n5|12
n5 = 1, 2, 3, 4, 6, 12

n5 1 (mod 5)
n5 = 1, 6, 11

If G is simple, then n5 1, hence n5 = 6. But |G| = 300 | 6!. So by the Index Factorial theorem, G is not simple, a contradiction to our assumption. We have that n5 = 1, hence P5 G where P5 is a Sylow 5-subgroup of G. G is not simple. Let G be group such that |G| = 312 = 23 3 13. n2 = 1, 3, 13, 39 n2|39 n2 1 (mod 2) n2 = 1, 3, 13, 39 n2 = 1, 3, , 13, , 39 n3 = 1, 4, 13, 52 n3|104 n3 1 (mod 3) n3 = 1, 2, 4, 8, 13, 26, 52, 104 n3 = 1, 4, , 13, , 52 n13 = 1 n13|24 n13 1 (mod 13) n13 = 1, 2, 4, 8, 12, 24 n13 = 1, 14, 25 Since n13 = 1, then P13 G where P13 is a Sylow 13-subgroup of G. Let G be group such that |G| = 1000 = 23 53. n2 = 1, 5, 25, 125 n2|125 n2 1 (mod 2) n2 = 1, 5, 25, 125 n2 = 1, , 5, , 25, , 125 n5 = 1 n5|8 n5 1 (mod 5) n5 = 1, 2, 3, 4, 8 n5 = 1, 6, 11 Since n5 = 1, then P5 G where P5 is a Sylow 5-subgroup of G.

Page 3

5.29 Let G be a group of order 90. (i) If a Sylow 5-subgroup P of G is not normal, prove that it has six conjugates. Hint. If P has 18 conjugates, there are 72 elements in G of order 5. Show that G has more than 18 other elements. Proof: Let G be group such that |G| = 90 = 2 32 5. n2|45 n2 = 1, 3, 5, 15, 45 n3|10 n3 = 1, 2, 5, 10 n5|18 n5 = 1, 2, 3, 6, 9, 18 n2 = 1, 3, 5, 15, 45 n2 1 (mod 2) n2 = 1, 3, 5, , 15, , 45 n3 = 1, 10 n3 1 (mod 3) n3 = 1, 4, 7, 10, n5 = 1, 6 n5 1 (mod 5) n5 = 1, 6, 11, 16,

Let P5 be a Sylow 5-subgroup of G. If P5 is not normal, then n5 1, hence n5 = 6. P5 has 6 conjugates in G, (by Sylow theorem, part (2) (All Sylow p-supbgroups are conjugates.)) (ii) Prove that G is not simple. Hint. Use Exercises 2.95(ii) and 2.96(ii) on page 114. Proof: Assume G is simple, then n5 1, hence n5 = 6. Thus, there are 6 4 = 24 elements of order 5. If n3 = 10, there are 10 subgroups of order 9. If these 10 subgroups intersect trivially, then we have 10 8 = 80 non-identity elements (not of order 5). This is too many elements, 104 for our group of order 90. So we have Sylow 3-subgroups, P3 and P3 such that |P3 P3| = 3. Let Q = P3 P3. We know Q P3 and Q P3 as P3 and P3 are Abelian by Corollary 2.104 (If p is prime, then every group of order p2 is Abelian.). And we know P3 NG(Q) and P3 NG(Q) as the normalizer is the largest subgroup of G in which Q is normal. Also, |P3| = |P3| |NG(Q)| as P3 P3 NG(Q) and P3 P3. Let m = |NG(Q)|. And now we have, |P3| = |P3| |NG(Q)| |G|, hence 9 m90. So, m = 18, 45, or 90. If m = 18, then [G: NG(Q)] = 90/18 = 5. But we have assumed G is simple, yet 905!, a contradiction, by the Index Factorial theorem. m 18. If m = 45, then [G: NG(Q)] = 90/45 = 2, hence NG(Q) G, another contradiction. If m = 90, then [G: NG(Q)] = 90/90 = 1, hence G = NG(Q) which means Q is normal in G. So there is no escaping a contradiction to our assumption that G is simple, thus G is not simple.

Page 4

5.30 Prove that there is no simple group of order 120. Proof: Let G be group such that |G| = 120 = 23 3 5. n2|15 n2 = 1, 3, 5, 15 n3|40 n3 = 1, 2, 4, 5, 8, 10, 20, 40 n5|24 n5 = 1, 2, 3, 6, 12, 24 n2 = 1, 3, 5, 15 n2 1 (mod 2) n2 = 1, 3, 5, 15, n3 = 1, 4, 7, 10, 40 n3 1 (mod 3) n3 = 1, 4, 7, 10, , 40 n5 = 1, 6 n5 1 (mod 5) n5 = 1, 6, 11, 16,

Assume G is simple. Then n2 1, n3 1, and n5 1, hence n5 = 6. Then, by Representation on Cosets, :G S6 where ker NG(P5). Since G is simple by assumption, then ker = {e}. So G (G) S6. Notice that (G) A6 (G) (by 2nd Isomorphism theorem). And by Exam 1, # 7(c), |(G) A6| = |(G)| = 120 or |(G) A6| = (1/2)|(G)| = 60. If ~((G) A6), then (G) A6 (G), hence |(G) A6| = (1/2)|(G)| = 60. But since (G) A6 (G), (G) is simple by assumption, and |(G) A6| = 60, then we have a contradiction. If (G) A6, then (G) A6 = (G), hence |(G) A6| = |(G)| = 120. So [A6: (G) A6] = 360/120 = 3. Then by the Representation on Cosets theorem, : A6 S3 where ker A6/(G) A6. And since A6 is simple, then ker = {(1)}. Thus, by the 1st Isomorphism theorem, A6 (A6). But (A6) S3 and |A6| = 120 > 6 = |S3|. So ker {e}, hence G is not simple. 5.31 Prove that there is no simple group of order 150. Proof: Let G be group such that |G| = 150 = 2 3 52. n2 = 1, 3, 5, 9, 15, 45 n2|45 n2 1 (mod 2) n2 = 1, 3, 5, 9, 15, 45 n2 = 1, 3, 5, 7, ... n3 = 1, 10 n3|30 n3 1 (mod 3) n3 = 1, 2, 3, 5, 6, 10, 15, 30 n3 = 1, 4, , 10, n5 = 1, 6 n5|6 n5 1 (mod 5) n5 = 1, 2, 3, 6 n5 = 1, 6, 11, 16, If G is simple, then n5 1, hence n5 = 6. But |G| = 150 | 6!. So by the Index Factorial theorem, G is not simple, a contradiction to our assumption. We have that n5 = 1, hence P5 G where P5 is a Sylow 5-subgroup of G. G is not simple.

Page 5

Worksheet Problems: 2. Prove the following: (a) Every group of order 15 is Abelian. Proof: Let G be group such that |G| = 15 = 3 5. n3|5 n3 = 1, 5 n5|3 n5 = 1, 3 n3 1 (mod 3) n3 = 1 n3 = 1, 4, n5 1 (mod 5) n5 = 1 n5 = 1, 6,

Since n3 = 1, and n5 = 1, then P3 G and P5 G where P3 is a Sylow 3-subgroup of G and P5 is a Sylow 5-subgroup of G. Since |P3| = 3 and |P5| = 5, then by Corollary 2.45 (Every group of prime order is cyclic.) a, b G such that (a) = 3, (b) = 5, a = P3, and b = P5. Consider ab. We know ab G. And (|a|, |b|) = 1, hence |ab| = 15 = |G|. ab = G. Also, a b = {e}. And so G is the internal direct product of a and b. And by Direct Products Theorem 4 (If G is the internal direct product of H and K, then HK H K.), G a b. But a Z 3 and b Z 5 so G Z 3 Z 5. And by Finite Abelian Groups, Exercise 1, (Z n Z m Z nm if and only if (n, m) = 1.), we have G Z 15. Since Z 15 is cyclic, then G is cyclic, hence Abelian.

Page 6

2. (b) There are no more than 4 non-isomorphic groups of order 30. Proof: Let G be group such that |G| = 30 = 2 3 5. n2|15 n2 = 1, 3, 5, 15 n3|10 n3 = 1, 2, 5, 10 n5|6 n5 = 1, 2, 3, 6 n2 = 1, 3, 5, 15 n2 1 (mod 2) n2 = 1, 3, 5, 15, n3 = 1, 10 n3 1 (mod 3) n3 = 1, 4, 7, 10, n5 = 1, 6 n5 1 (mod 5) n5 = 1, 6, 11, 16,

If n3 = 10 and n5 = 9, then we need 24 distinct elements of order 5 and 20 distinct elements of order 3, a contradiction to |G| = 30. Either n3 = 1 or n5 = 1. If n3 = 1, then a G such that (a) = 3, a = P3 and a G. And since and n5 = 1, 6, we also have that b G such that (b) = 5, and b = P5. By Direct Products, Exercise 2 (H G, K G, H G or K G HK G), we have that ab G and ab Z 15.by part (a). Thus, c G such that (c) = 15 and c G. By similar argument, if n5 = 1 and n3 = 1, 10 we have the same result. Since [G: c] = 2, then c G. So by Cauchys theorem, d G such that (d) = 2. So c G gives us that dcd1 c. Hence dcd1 = cn where 0 n 14. 2 2 Thus c = d1cnd = dcnd1 = (dcd1)n = (cn)n = c n . e = c n 1. By corollary to Lagranges theorem, 15n2 1. So n2 1 = 1, 4, 11, or 14. This gives us the following possible groups isomorphic to G: c, d: c15 = 1 = d2; cdc1 = d, c, d : c15 = 1 = d2 ; cdc1 = d4, c, d: c15 = 1 = d2; cdc1 = d11, c, d: c15 = 1 = d2; cdc1 = d14. Thus, there are no more than 4 non-isomorphic groups of order 30. (c) There are at least 4 non-isomorphic groups of order 30. (Describe them in terms of groups that we know and explain how you know that the four youve described are nonisomorphic. Proof: |Z 30| = |D30| = |D10 Z 3| = |D6 Z 5| = 30. Z 30 has an element of order 30. D30 is not cyclic and has 15 elements of order 2. D10 Z 3 has only 1 element of order 2 and 5 elements of order 6. D6 Z 5 has only 1 element of order 6 and only 1 element of order 2. None of the 4 groups have the same number of elements of the same order, hence none of them are isomorphic to each other. There are at least 4 non-isomorphic groups of order 30.

Page 7

Page 8

3. Let G be a group of order 48. Show that the intersection of any two distinct Sylow 2-subgroups of G has order 8. Proof: 48 = 24 3. n2|3 n2 = 1, 3 n3|16 n3 = 1, 2, 4, 8, 16 n2 = 1, 3 n2 1 (mod 2) n2 = 1, 3, n3 = 1, 4, 16 n3 1 (mod 3) n3 = 1, 4, 7, , 16,

By hypothesis, we assume n2 = 3 (i.e. P2, P2, and P2 , distinct Sylow 2-subgroups of G, each of order 16.). Note that |P2 P2| 16 as the subgroups are distinct, hence |P2 P2| = 1, 2, 4, or 8. Since n2 = [G:NG(P2)] = 3 = [G:P2], then NG(P2) = P2. By the Representation of Cosets theorem, :G S3 such that ker P2. Since |G| = 48 and |S3| = 6, then the map is an 8 to 1 map, hence |ker | 8. We know ~( P2 G) but ker G, so ker P2. We also know P2 is conjugate to P2, so g G such that P2 = gP2g1. Since ker G, then g(ker )g1 = ker . And, since ker P2, ker = g(ker )g1 gP2g1 = P2. ker P2 P2. Since |ker | 8, ker P2 P2, and |P2 P2| 8, then |P2 P2| = 8. The intersection of any two distinct Sylow 2-subgroups of G has order 8. 4. Let G be a group with |G| = 56. Prove that G is not simple. Proof: 56 = 23 7. n2|7 n2 = 1, 7 n7|8 n7 = 1, 2, 4, 8 n2 1 (mod 2) n2 = 1, 7 n2 = 1, 3, 5, 7, n7 1 (mod 7) n7 = 1, 8 n7 = 1, 4, 7,

Assume G is simple, then n2 1 and n7 1. Thus, n2 = 7 and n7 = 8, which gives us 7 Sylow 2-subgroups each of order 8 and 8 Sylow 7-subgroups each of order 7. Since 7 is prime, then 7 6 = 42 elements of our group of order 56 have order 7. The 7 subgroups of order 8 have intersections of 1, 2, or 4 elements, hence we have a minimum of 5 4 = 20 elements of order 7, thus exceeding the size of our group. n2 = 1 or n7 = 1, hence G is not simple.

Page 9

5. What is the smallest composite integer n such that there is a unique group of order n? 15. Proof: Any group with order an even composite integer is isomorphic to a dihedral group as well as a cyclic group. So n must be odd. The smallest odd composite integer is 15. By Exercise 1 (a) above, any group of order 15 is cyclic and Abelian, hence unique. 6. Let G be a noncyclic group of order 21. (a) How many 3-Sylow subgroups does G have? 7. Proof: n3|7 n3 = 1, 7 n7|3 n7 = 1, 3 n3 1 (mod 3) n3 = 1, 7 n3 = 1, 4, 7, n7 1 (mod 7) n7 = 1 n7 = 1, 4,

If n3 = 1, and n7 = 1, then P3 G and P7 G where P3 is a Sylow 3-subgroup of G and P7 is a Sylow 7-subgroup of G. Since |P3| = 3 and |P7| = 7, then a, b G such that (a) = 3, (b) = 7, a = P3, and b = P7. We know ab G. And (|a|, |b|) = 1, hence |ab| = 21 = |G|. ab = G. Also, a b = {e}. And so G is the internal direct product of a and b. Thus, G a b. But a Z 3 and b Z 7 so G Z 3 Z 7. And so we have G Z 21, which is cyclic. n3 = 7, hence G has 7 Sylow 3-subgroups. (b) Prove that G has 14 elements of order 3. Proof: Since G has 7 distinct Sylow 3-subgroups, then G has 7 2 distinct elements of order 3.

Page 10

7. Let G be a group of order 60. Show that G has exactly four elements of order 5 or exactly 24 elements of order 5. Which of these cases holds for A5? Proof: 60 = 22 3 5 n2 = 1, 3, 5, 15 n2|15 n2 1 (mod 2) n2 = 1, 3, 5, 15 n2 = 1, 3, 5, 15, n3 = 1, 4, 10 n3|20 n3 1 (mod 3) n3 = 1, 2, 4, 5, 10, 20 n3 = 1, 4, 7, 10, n5 = 1, 6 n5|12 n5 1 (mod 5) n5 = 1, 2, 3, 4, 6, 12 n5 = 1, 6, 11, 16, If n5 = 1, then there is a unique Sylow 5-subgroup of G such that |P5| = 5, hence G has exactly four elements of order 5. If n5 = 6, then there are 6 unique Sylow 5-subgroups of G, each of order 5. Since the subgroups are unique and of prime order, the intersection of any 2 of them has order 1. In this case, G has exactly 6 4 = 24 elements of order 5. Since A5 is simple, then n5 1, hence A5 has exactly 24 elements of order 5. 8. Let G be a group of order 60 and let H G with |H| = 2. Show (a) G has normal subgroups of order 6, 10, and 30, Proof: Since H G with |H| = 2, then G/H is a group and |G/H| = 30. And by Exercise 2 (b) S* G/H such that S* is cyclic, S* G/H, and |S*| = 15. By Exercise 2 (a) T* S*/H such that T* is cyclic, T* G/H, and |T*| = 5, and U* S*/H such that U* is cyclic, U* G/H, and |U*| = 3, So, by the Correspondence theorem, S G, T G, and U G such that S* = S/H, S G, |S| = 30, T* = T/H, T G, |T| = 10, and S* = U/H, U G, |U| = 6 (b) G has subgroups of order 12 and 20, and Proof: Since |G/H| = 30, then P3*, P5* both normal to G/H, and P2* G/H. So P2*P3* G/H, and P2*P5* G/H, where |P2*P3*| = 6 and |P2*P5*| = 10. By the Correspondence theorem H12 and H20, subgroups of G, such that H12 = P2*P3*/H, H20 = P2*P5*/H, |H12| = 12, and |H20| = 20. (c) G has a cyclic subgroup of order 30. Proof: (stuck)

Page 11

9. Let G be a group of order 60. If the Sylow 3-subgroup is normal, show that the Sylow 5-subgroup is also normal. Proof: (stuck) 10. Let |G| = 72 13. Prove G is Abelian. Proof: n7 = 1 n7|13 n7 1 (mod 7) n7 = 1, 13 n7 = 1, 4, 7, n13|49 n13 1 (mod 13) n13 = 1 n13 = 1, 7, 49 n13 = 1, 12, Since n7 = 1 and n13 = 1, then by Proposition 5.39 (A finite group G all of whose Sylow subgroups are normal is the direct product of its Sylow subgroups.), G = P7 P13. Since |P7 |= 49, then P7 is Abelian. P7 Z 49 or P7 Z 7 Z 7. And since 13 is prime, P13 Z 13. Thus, G Z 49 Z 13, or G Z 7 Z 7 Z 13, hence G is Abelian. A group is said to be solvable if there exist subgroups G0, G1, , Gk such that {e} = Gk Gk1 G2 G1 G0 = G and such that Gi/Gi+1 is Abelian for all i. This sequence of subgroups is called a solvable series for G. 11. Prove that S4 is solvable. Proof: {(1)} V4 A4 S4. [S4: A4] = 2. S4/ A4 is Abelian since 2 is prime. [A4: V4] = 3. A4: V4 is Abelian since 3 is prime. V4/{(1)} = V4 and we know V4 is Abelian. S4 is solvable. 12. Prove that if G is solvable and H G, then H is solvable. Proof: G is solvable exist subgroups G0, G1, , Gk such that {e} = Gk Gk1 G2 G1 G0 = G. Let H0 = H and Hi = H Gi. Then by the 2nd Isomorphism theorem, H Gi H. (stuck)

Page 12

13. Suppose G is solvable and : G G is a homomorphism from G to G . Prove (G) is solvable. Proof: G is solvable exist subgroups G0, G1, , Gk such that {e} = Gk Gk1 G2 G1 G0 = G. Suppose ker = {e}, then by the Correspondence theorem, (Gi) (Gi1) for every i {0, 1, , k} and (G0) = ( G ). Suppose Gj = ker for some j {0, 1, , k}, then by the Correspondence theorem, (Gi) (Gi1) for every i {0, 1, , j}, (G0) = ( G ), and (Gm) { eG } for every m {j +1, , k}, hence (Gm) (Gm1). Need to show (Gi1)/(Gi) is Abelian for every i. (stuck) G is solvable.

14. Let G be a group with H G. Suppose H and G/H are both solvable. Prove G is solvable. Proof: H is solvable exist subgroups H0, H1, , Hk of H such that {e} = Hk Hk1 H2 H1 H0 = H. G/H is solvable exist subgroups N0, N1, , Nk of G/H such that {e} = Nk Nk1 N2 N1 N0 = N. By the Correspondence theorem, there are subgroups Pi of G with H Pi such that Pi / H = Ni and Pi Pi1. So H Pk/H Pk1/H P1/H P0/H = G/H. By the 3rd Isomorphism theorem, Pi/Pi1 (Pi/H)/(Pi1/H). And since (Pi/H)/(Pi1/H) is Abelian, then Pi/Pi1 is Abelian. {e} = Hk Hk1 H2 H1 H0 = H Pk Pk1 P1 P0 = G. G is solvable. 15. Let G be a group with H G and K G. Prove that if H and K are both solvable, then HK is solvable. Proof: 16. Let G be a group of order 495 = 32 5 11. (a) What are the possible numbers of Sylow subgroups? (ie. what are the possibilities for n3, n5, and n11?) Proof: n3|55 n3 = 1, 5, 11, 55 n5|99 n5 = 1, 3, 9, 11, 33, 99 n11|45 n11 = 1, 3, 5, 9, 15, 45 n3 = 1, 55 n3 1 (mod 3) n3 = 1, 4, , 55, n5 = 1, 11 n5 1 (mod 5) n5 = 1, 6, 11, 16, n11 = 1, 45 n11 1 (mod 11) n11 = 1, 10, , 45, Page 13

(b) Prove that a 5-Sylow subgroup or an 11-Sylow subgroup is normal. Proof: If n5 = 11 and n11 = 45, then there are 11 distinct subgroups of order 5 and 45 distinct subgroups of order 11. Thus, G contains 11 4 = 44 elements of order 5 and 45 10 = 450 elements of order 11. We also have at least one subgroup of order 9 which contains 8 elements of order 3 or 9. So G contain 44 + 450 + 8 = 502 elements, a contradiction to |G| = 495. n5 = 11 or n11 = 45, thus a 5-Sylow subgroup or an 11-Sylow subgroup is normal in G. (c) Let K be the normal subgroup from part (b). Prove G/K is isomorphic to Z m Z 3 Z 3 or Z m Z 9 where m {5, 11}. Proof: Let M be the other subgroup from part (b) and let P3 be a Sylow 3-subgroup of G. Suppose m = 5, then |G/K| = 32 5 11 = 5 = 32 11. So n3 = 1 and n11 = 1. Thus by Proposition 5.39 (A finite group G all of whose Sylow subgroups are normal is the direct product of its Sylow subgroups.), G/K P3P11. Since |P3 | = 9, then P3 is Abelian. P3 Z 9 or P3 Z 3 Z 3. And since 11 is prime, P11 Z 11. Thus, G Z 11 Z 9, or G Z 11 Z 3 Z 3. If m = 11, then by similar proof we have G Z 5 Z 9, or G Z 5 Z 3 Z 3. (d) Let H5 be a 5-Sylow subgroup of G and let H11 be an 11-Sylow subgroup of G. Prove H5H11 G. (part (c) might be helpful one of these two subgroups is the K in part (c)). Proof: By the Second Isomorphism theorem, H5H11/H5 H11/H5 H11. Since H5 H11 = {e}, then H5H11/H5 H11. And since | H11| = 11, then H11 Z 11. Since, by part (c) G/H5 is isomorphic to Z 11 Z 3 Z 3 or Z 11 Z 9, and Z 11 Z 11 Z 3 Z 3, and Z 11 Z 11 Z 9, then by the Correspondence theorem, H5H11 G. (e) Find a solvable series for G. Proof: {e} H5 H5H11 G.

Page 14

Debra Griffin Math 210A HW #10 3.2 #3, 5, 6, 7, 8, 10, 12, 13 + Worksheet #3 8 Hand in #3.12 Textbook Problems: 1. Read and study section 3.2 on your own. 3.3 (i) If R is a domain and a R satisfies a2 = a, prove that either a = 0 or a = 1. Proof: Assume R is an integral domain a R, a2 = a. Then a2 a = a(a 1) = 0. Since R is an integral domain, then either a = 0 or a = 1. (ii) Show that the commutative ring F(R ) in Example 3.7 (Let F(R ) be the set of all the functions R R equipped with the operations of pointwise addition and pointwise multiplication: Given f, g F(R ), define functions f + g and fg by f + g:a f (a) + g(a) and fg:a f (a)g(a).) contains infinitely many elements f 0, 1 with f 2 = f. Proof: 1: a nZ Let f n ( a) = . 0 : otherwise Then n N , fn2 = f, {fn}n

F(R ) and{fn}n

contains infinitely many elements.

3.5 Show that U(Z m) = {[k] Z m : (k, m) = 1}. Proof: Let m Z . Let T = {[k] Z m : (k, m) = 1}. Let [x] U(Z m). Then [y] U(Z m) such that [x] [y] = [xy] = [1]. xy 1 (mod m), hence xy 1 = rm for some integer r. xy rm = 1. Suppose (x, m) = d. Then x = sd and m = td for some integers s, t. So then xy rm = sdy rtd = d(sy rt) = 1. d = 1, hence [x] T. And this gives us U(Z m) T. Let [w] T. Then (w, m) = 1. s, t Z such that ws + tm = 1, ws 1 = tm, ws 1 (mod m), [ws] = [1], [w] [s] = [1]. [w] U(Z m). Hence T U(Z m). So, finally, we have U(Z m) = T.

HW #10, Ring Theory, 3.2 + Worksheet 3.6 Find all the units in the commutative ring F(R ) defined in Example 3.7. UF(R ) = {f F(R ): a R , f (a) 0}. Proof: Let f U(F(R )). Then a R , (f (a))1 0, such that f (a)(f (a))1 = 1. a R , f (a) 0, hence f {f F(R ): a R , f (a) 0}. Thus UF(R ) {f F(R ): a R , f (a) 0}. Let g {f F(R ): a R , f (a) 0}. Then since g: R R , then a R , (g(a))1 0, such that g(a)(g(a))1 = 1. Thus, g U(F(R )), hence {f F(R ): a R , f (a) 0} UF(R ). UF(R ) = {f F(R ): a R , f (a) 0}.

3.7 Generalize the construction of F(R ) to arbitrary commutative rings R: Let F(R) be the set of all functions from R to R, with pointwise addition, f + g : r f (r) + g(r), and pointwise multiplication, f g: r f (r)g(r) for r R. (i) Show that F(R) is a commutative ring. Proof: (1) The function f = 0F(R) where f (r) = 0R r R, is the additive identity for F(R). (2) Let f, g F(R). Since f (r), g(r) R r R, then f (r) + g(r) R. Thus f + g F(R). And we have closure under addition.

(3) Let f, g, h F(R). Since R is associative under addition, then f (r) + (g(r) + h(r)) = (f (r) + g(r)) + h(r) r R, then f + (g + h) = (f + g) + h. Thus, F(R) is associative under addition. (4) And since R is commutative under addition, then f (r) + g(r) = g(r) + f (r) r R, then f + g = g + f. Thus, F(R) is commutative under addition. (5) Since r R, r such that r + r = 0R, then f F(R), f (r) + f (r) = 0R r R, hence f + f = 0F(R). Thus, each element of F(R) has an additive inverse. (6) Since f (r), g(r) R r R, then f (r)g(r) R. Thus fg F(R). And we have closure under multiplication. (7) Since R is associative under multiplication, then f (r)(g(r)h(r)) = (f (r)g(r))h(r) r R, then f(gh) = (fg)h. Thus, F(R) is associative under multiplication. (8) Since R is commutative under multiplication, then f (r)g(r) = g(r)f (r) r R, then fg = gf. Thus, F(R) is commutative under multiplication.

Page 2

HW #10, Ring Theory, 3.2 + Worksheet (9) Since the distributive law holds in R, then f (r)(g(r) + h(r)) = f (r)g(r) + f (r)h(r) r R, then f (g + h) = f g + f h. Thus the distributive law holds in F(R). F(R) satisfies all the axioms of a commutative ring. (ii) Show that F(R) is not a domain. Proof:

r : r 0 r : r = 0 Let f ( r) = and g( r) = . Then f 0F(R) and g 0F(R), but fg = 0F(R). 0 : r = 0 0 : r 0 F(R) is not a domain.
(iii) Show that F(Z 2) has exactly four elements, and that f + f = 0 for every f F(Z 2). Proof: Z 2 = {[0], [1]}. By the fundamental counting principle, there are 2 2 possible maps from Z 2 to Z 2. Specifically, F(Z 2) = {f1, f2, f3, f4} where f1 = {([0], [0]), ([1], [1])} f2 = {([0], [1]), ([1], [0])} f3 = {([0], [0]), ([1], [0])} f4 = {([0], [1]), ([1], [1])} 3.8 (i) If R is a domain and S is a subring of R, then S is a domain. Proof: Let R be a domain and S be a subring of R. Let a, b S such that a 0S and b 0S. Since a, b R then ab R and ab 0R. Since S is a subring of R, then 0S = 0R. ab 0S. S is a domain. (ii) Prove that C is a domain, and conclude that the ring of Gaussian integers is a domain. Proof: We know C is a ring with the usual addition and multiplication (easily verified), so we only need to verify C is a domain. Let a + bi, c + di C where a, b, c, d R and (a + bi)(c + di) = 0. Then ac bd + (ad + bc)i = 0. Thus, ac bd = 0 and ad + bc = 0, which gives us ac = bd and ad = bc. So, by substitution, we have d2 = c2 or a2 = b2. Since a, d R , then this can only be true if d = c = 0 or a = b = 0.

Page 3

HW #10, Ring Theory, 3.2 + Worksheet 3.10 (i) Prove that R = {a + b 2 : a, b Z } is a domain. Proof: It is easily verified that R is a ring under the usual addition and multiplication, so we will only show that R is a domain. Let a + b 2 , c + d 2 R such that (a + b 2 )(c + d 2 ) = 0. Then ac + 2bd + (ad + bc) 2 = 0. Thus, ac + 2bd = 0 and ad + bc = 0, which gives us ac = 2bd and ad = bc. So, by substitution, we have 2 c = 2d2 or a2 = 2b2. Since a, b, c, d Z , then this can only be true if d = c = 0 or a = b = 0. a + b 2 = 0 or c + d 2 = 0., hence R is a domain. (ii) Prove that R = { (a + b 2 ): a, b Z } is not a domain. 2 Proof: Let a, b, c, d be odd integers. Then 1 ) 1 (c + d 2 ) = 1 (ac + 2bd) + 1 (ad + bc) 2 . (a + b 2
2 2 4 4 1

Since a, c are odd, then by simple number theory, ac + 2bd is odd.

Thus (ac + 2bd) cannot be written as k where k is an integer. 4 2 R is not closed under multiplication, hence R is not a ring, and ergo R is not a domain. (iii) Using the fact that = (1 +
2 1

19 ) is a root of x2 x + 5, prove that

R = {a + b: a, b Z } is a domain. Proof: The ring axioms for addition, multiplicative identity, closure under multiplication, and distributive law can be easily shown. So, we will only show R is a domain. Let a + b, c + d R such that (a + b)(c + d) = 0. (a + b)(c + d) = bd2 + (ad +bc) + ac = 0. Since = (1 +
2 1

19 ) is a root of x2 x + 5, then bd2 + (ad +bc) + ac = 0

bd = 1, ad +bc = 1, and ac = 5 or bd = 0, ad +bc = 0, and ac = 0. Since a, b, c, d are integers, then if bd = 1 we have b = d = 1. Hence ad +bc = a + c = 1. So then, a(a 1) = 5, a contradiction to a Z . bd = 0, ad +bc = 0, and ac = 0. If b 0, then d = 0, hence ad +bc = bc = 0, which implies that c = 0. If d 0, then b = 0, hence ad +bc = ad = 0, which implies that a = 0. a + b = 0 or c + d = 0. R = {a + b: a, b Z } is a domain.

Page 4

HW #10, Ring Theory, 3.2 + Worksheet 3.12 (i) If R is a commutative ring, define the circle operation a b by a b = a + b ab. Prove that the circle operation is associative and that 0 a = a for all a R. Proof: Let a, b, c R. Then, since R is a commutative ring where (1)a = a a R, we have (a b) c = a b + c (a b)c = a + b ab + c (a + b ab)c = a + b + (1)ab + c + (1)(ac + bc + (1)abc) = a + b + (1)ab + c + (1)(ac) + (1)(bc) + (1)(1)(abc) = a + b + c +(1)(bc) + (1)ab + (1)(ac) + (1)(1)(abc) = a + b + c bc a(b + c bc) = a + (b c) a(b c) = a (b c) The circle operation is associative. And since 0 a = a 0 = 0 for each a R, we have 0 a = 0 + a 0 a = a + (1) 0 = a + 0 = a. (ii) Prove that a commutative ring R is a field if and only if {r R: r 1} is an Abelian group under the circle operation. Hint. If a 0, then a + 1 1. Proof: Let T = {r R: r 1}. : Assume R is a field. We will show T is an Abelian group. (1) To show closure under , let a, b T, then a 1 and b 1. Suppose a b = a + b ab = 1. Then b ab = 1 a b(1 a) = 1 a b =

But this is clearly a contradiction as b 1. a b 1. a b T. (2) To show associativity under , note that since R is a field and T R, we have by part (i), T is associative under . (3) To show commutativity under , let a, b T. Since R is commutative under + and , then a b = a + b ab = b + a ba = b a, hence T is commutative under . (4) To show existence of an identity, 0 under , note that by commutativity of T and by result of part (i), we have 0 a = a 0 = a, a T. (5) To show existence of an inverse for each element of T under , let a T. If a = 0, then a is its own inverse. So, assume a 0. Since a T, then a 1, hence 1 a 0. We have (1 a)1 R (as R is a field) and (a)(1 a)1 T (as (a)(1 a)1 1). a a a a + a 2 a( a 1) a( a 1) Since a = a+ a = a+ = a+ = a+ = 0 , then 1 a 1 a 1 a 1 a 1 a a 1 each element a T has an inverse in T under .

1 a = 1. 1 a

Page 5

HW #10, Ring Theory, 3.2 + Worksheet 3.12 (ii) (continued) : Assume R is a commutative ring and T = {r R: r 1} is an Abelian group under the circle operation. Let x R such that x 0. Then x + 1 1, hence x + 1 T. Since T is an Abelian group under , then y T such that 0 = (x + 1) y. 1 = (x +1) y + 1 = x +1 + y (xy + y) = x(y 1). Since y T, then y 1, hence y 1 0. x is a unit in R. Every nonzero element of R is a unit, hence R is a field. 3.13 Find the inverses of the nonzero elements of Z 11. Z 11 = {[0]. [1], [2], [3], [4], [5], [6], [7], [8], [9], [10]}. [1] [1] = [1], hence [1]1 = [1]. [2] [6] = [12] = [1], hence [2]1 = [6] and [6]1 = [2]. [3] [4] = [12] = [1], hence [3]1 = [4] and [4]1 = [3]. [5] [9] = [45] = [1], hence [5]1 = [9] and [9]1 = [5]. [7] [8] = [56] = [1], hence [7]1 = [8] and [8]1 = [7]. [10] [10] = [100] = [1], hence [10]1 = [10]. Worksheet Problems: 3. Prove that in a ring with unity, U(R) ZD(R) = . Proof: We will prove the result by contradiction. Suppose U(R) ZD(R) . Then x U(R) such that x ZD(R). Thus, y R such that xy = yx = 1 and z R such that z 0 but xz = 0. From this we have 1 = xy = xy + 0 = xy + xz = x(y + z). Therefore, y = y 1 = y(x(y + z)) = (yx)(y + z) = 1 (y + z) = y + z. And since z 0, then we have a contradiction. U(R) ZD(R) = . 4. Let R be a ring and let a R. Prove that if a 0 and a ZD(R), then ab = ac implies b = c and ba = ca for all b, c R. Proof: Let a R such that a 0 and a ZD(R). Let b, c R such that ab = ac. Then 0 = ab ac = a(b c). Since a 0 and a ZD(R), then b c = 0. This gives us that b = c, hence ba = ca.

Page 6

HW #10, Ring Theory, 3.2 + Worksheet 5. Prove that every finite integral domain is a field. Proof: Let R be a finite integral domain. Since R is a commutative ring with identity, we need only show that for each nonzero element, a, in R, a is a unit. Since R is finite, then R = {a1, a2, , an} for some natural number n. Suppose at 0 for some t {1, 2, , n}. Consider the products ata1, ata2, , atan. If ai aj, then atai ataj. Thus, ata1, ata2, , atan are distinct elements of R. And since there are n of them, then they are all of the elements of R. Since 1 {ata1, ata2, , atan}, then for some j {1, 2, , n}, ataj = 1. 6. Let R be a ring in which x2 = x for all x R. (Such a ring is called a Boolean ring.) Prove that R is a commutative ring. Proof: Let x, y R. Then x + x = (x + x)2 = x2 + x2 + x2 + x2 = x + x + x + x 0=x+x x = x. And x + y = (x + y)2 = x2 + xy + yx + y2 = x + xy + yx + y 0 = xy + yx = xy + yx xy = yx. R is commutative. 7. Find all subrings of Z . (Compare to problem # 9 in book.) The subrings of Z are {0}, {nZ : n N }. Proof: We know {0} is a subring of any ring, R, so we will show (1) nZ is a subring of Z , n N and (2) T Z such that T nZ n 0, T is not a subring of Z . (1) Let n N . Let a, b Z. Then na nb = n(a b) nZ (as a b Z ). And na nb = n(anb) nZ (as anb Z ). And since 0 nZ then nZ . nZ is a subring of Z . (2) Let T Z such that T nZ n 0. Consider Y = {|a b| : a, b T, such that a b}. Then by the well-ordering principle, Y has a least element h. 8. Find an example of a ring with elements a and b such that a and b are zero divisors, but a + b 0 and a + b is not a zero divisor. In Z 6, [2] and [3] are zero divisors, but [2] + [3] = [5] is not a zero divisor and [5] [0].

Page 7

Debra Griffin Math 210A HW #11 Ch 3.4 #41, 42, 44, 45, 49, 50 (i), (ii), 51 Textbook Problems: 3.40 Let R be a commutative ring and let F(R) be the subring consisting of all functions f : R R with pointwise operations. (i) Show that R is isomorphic to the subring of F(R) consisting of all the constant functions. Proof: Let T be the subring of F(R) consisting of all the constant functions. Let r R and let fr :R R be defined by fr(x) = r x R. Let : R T be defined by (r) = fr. Since x R, (a + b) = fa+b(x) = a + b = fa(x) + fb(x), and (ab) = fb(x) = a + b = fa(x) + fb(x), then is a homomorphism under + and . To show is injective, let a, b R such that (a) = (b) and note that x R, a = fa(x) = fb(x) = b. To show is surjective, let y T and note that y = fy = (y) where y in R. (ii) If f (x) R[x], let f: R R be defined by r f (r) [thus, f is the polynomial function associated to f (x)]. Show that the function : R[x] F(R), defined by (f (x)) = f, is a ring homomorphism. Proof: Let f, g R[x]. Then (f + g) = f+g = {(r, (f + g)(r))| r R} = {(r, (f (r) + g(r))| r R} = {(r, f (r)) + (r, g(r))| r R} = f + g = (f) + (g). And, similarly, (fg) = fg = f g = (f )(g). is a ring homomorphism. (iii) Show that is injective if R is an infinite field. Proof: Assume R is an infinite field. Let f, g R[x] such that (f ) = (g). Then f = g. Thus r R, f (r) = g(r). Assume, without loss of generality, deg f (x) deg g(x) and let h(x) = f (x) g(x). If h(x) 0, then deg h(x) = n for some integer n 0. By Theorem 3.25 (If k is a field, f (x) k[x], deg f (x) = n, then f (x) has at most n roots in k.), h(x) has at most n roots. However, r R, h(r) = f (r) g(r) = 0. Thus every element of R is a root of h and R is an infinite field. But this contradicts that h(x) has at most n roots. h(x) = 0. f (x) = g(x), hence is injective.

HW #11, Ring Theory, 3.5 + Worksheet

3.41 Let I and J be nonzero ideals in a commutative ring R. If R is a domain, prove that I J {0}. Proof: Since I and J are nonzero ideals, then a I, a 0, and b J, b 0. Thus, ab 0 as a, b r and R is a domain. Since b R ab I and a R ab J, then I J {0}. 3.42 Let R be a commutative ring. Show that the function : R[x] R, defined by : a0 + a1x + a2x2 + + anxn a0, is a homomorphism. Describe ker in terms of roots of polynomials. Proof: We first show that is a ring homomorphism. Let f, g R[x]. Then f = f0 + f1x + f2x2 + + fnxn for some f0, f1, , fn R, and g = g0 + g1x + g2x2 + + gnxn for some g0, g1, , gn R. Since R is commutative, we have (f + g) = (f0 + g0 +( f1 + g1)x + +(fn + gn) xn) = f0 + g0 = (f ) + (g). n i And (fg) = ( f g j k x ) = f0g0 = (f )(g). i= 0 j + k = i is a ring homomorphism. To describe ker in terms of roots of polynomials, we note that ker = {f R[x]: (f ) = f (0) = 0}. f R[x] such that 0 is a root of f, f ker . 3.44 (i) Prove that F, the field with 4 elements (see Exercise 3.14 on page 125: a b F 4 = { | a, b Z 2}), and Z 4 are not isomorphic commutative rings. b a + b Proof: Suppose : F Z 4 is an isomorphism. 0 0 1 0 1 1 0 1 Let F = {0F , 1F, a, b}. where 0F = , 1F = , a = , b = . 0 0 0 1 1 0 1 1 By simple calculations we have ab = 1F. (ab) = (1F) = 1Z 4. But only 1Z 4 1Z 4 = 1Z 4 and 3 3 = 1Z 4, contradicting that a b. is not an isomorphism.

Page 2

HW #11, Ring Theory, 3.5 + Worksheet (ii) Prove that any two fields having exactly four elements are isomorphic. Hint. First prove that 1 + 1 = 0, and then show that the nonzero elements form a cyclic group of order 3 under multiplication. Proof: Let F be a field of order 4. Then F = {0, 1, a, b} where 1, a, and b are units. We have ab = 1 (otherwise ab = a or ab = b, causing b = 1 or a = 1, a contradiction). Thus, a2 = b and b2 = a, as ab = 1 rules out a2 = 1 or b2 = 1 (inverses are unique). So we can rewrite F = {0, 1, a, b} as F = {0, 1, a, a2}. This gives us that, under multiplication, U(F) = a where a is a cyclic group of order 3. Since F is a group under addition, then there is a group isomorphism, F Z 4 or F V4 by Proposition 2.89, (Every group G of order 4 is isomorphic to either Z 4 or V4. Moreover, Z 4 and V4 are not isomorphic.) However, by Proposition 3.12 (The commutative ring Z m is a field if and only if m is prime.) Z 4 is not a field. So, we shall assume that there is a group isomorphism, F V4 = {e, a, b, c} where e is the additive identity and a + a = b + b = c + c = e. And, by simple calculations, we can see that the multiplication table for a b F 4 = { | a, b Z 2}) matches the multiplication for F as described above, and the b a + b addition table for F 4 matches that of V4. Thus, we can establish an isomorphism 0 0 1 0 1 1 0 1 2 : F F 4 defined by (0F) = , (1F) = , (a) = , and (a ) = . 0 0 0 1 1 0 1 1 Any two fields having exactly four elements are isomorphic to F 4with multiplication defined as above.

3.45 (i) Show that every element a Z p has a pth root (i.e. there is b Z p with a = bp).
Proof: (I think we must assume p is prime. A quick check shows 2 a4 for any a Z 4.) We first note that 0p = 0 and 1p = 1. Assume p is prime. Then U(Z p) = Z p {0} = {a1, a2, , ap1}, a multiplicative group of order p 1. By corollary to Lagranges Theorem, ai U(Z p), (ai)p1 = 1. Thus, (ai)p1ai = (ai)p = ai. Every element a Z p has a pth root.

Page 3

HW #11, Ring Theory, 3.5 + Worksheet 3.45 (ii) Let k be a field that contains Z p as a subfield [e.g. k = Z p(x)]. For every positive integer n, show that the function n: k k, given by (a) = a p , is a ring homomorphism. Proof: Let p be a prime. We first note that the unity 1 Z p is the same as the unity of k (as ab = 1Z p for some a, b Z p, and ac = 1k for some c k ac = ( 1Z p a)c = 1Z p (ac) = abac = 1Z p 1k = 1Z p ). Secondly, note that a Z p, pa = 0 (by corollary to Lagranges theorem). Since p is prime, then p is the smallest positive integer such that p 1 = 0. Thirdly, note that b k, pb = p 1 b = 0 b = 0. Thus the characteristic of k is p. n Also, note that for n > 0, j 0 and j pn, p pn! = p j!(pn j)!. This gives us that
j
n

a, b k, (a + b) = (a + b ) =

pn

p n p n j j a b = j = 0

pn

(a + b ) p = a p + 0 + + 0 + b p = (a) + (b).

n n

as the only terms of divisible by pn are a p and b p . which are not n n n Also, a, b k, (ab) = (ab ) p =( a ) p (b ) p = (a)(b). n is a ring. 3.49 (i) If R and S are commutative rings, show their direct product R S is also a that commutative ring, where addition and multiplication in R S are defined coordinatewise: (r, s) + (r, s ) = (r + r, s + s) and (r, s)(r, s) = (rr, ss). Proof: Let (r, s), (r, s), and (r, s) R S. (1) (r, s) + (r, s) = (r + r, s + s) R S (since R and S are closed under addition). (2) (r, s) + [(r, s) + (r, s)] = (r, s) + (r + r, s + s) = (r + r + r, s + s + s) and [(r, s) + (r, s)] + (r, s) = (r + r, s + s) + (r, s) = (r + r + r, s + s + s) (since R and S are associative under addition). (3) (r, s) + (r, s) = (r + r, s + s) = (r + r, s + s) = (r, s) + (r, s) (since R and S are commutative under addition). (4) (r, s) + (0R, 0S) = (r + 0R, s + 0S) = (r, s). = (0R + r, 0S + s) = (0R, 0S) + (r, s). (5) (r, s) + (r, s) = (r + (r), s + (s)) = (0R, 0S) = ((r) + r, (s) + s) = (r, s) + (r, s) (since r and s have additive inverses, r and s in R and S respectively). (6) (r, s) (r, s) = (rr, ss) R S (since R and S are closed under multiplication). (7) (r, s) [(r, s) (r, s)] = (r, s) (rr, ss) = (rrr, sss) and [(r, s) (r, s)] (r, s) = (rr, ss) (r, s) = (rrr, sss) (since R and S are associative under multiplication). (8) (r, s) [(r, s) + (r, s)] = (r, s) (r + r, s + s) = (rr + rr, ss + ss) and (r, s) (r, s) + (r, s) (r, s) = (rr, ss) + (rr, ss) = (rr + rr, ss + ss) (since the distributive law holds in both R and S). (9) (r, s) (r, s) = (rr, ss) = (rr, ss) = (r, s) (r, s) (since R and S are commutative under multiplication). R S is a commutative ring.

Page 4

HW #11, Ring Theory, 3.5 + Worksheet (ii) Show that if m and n are relatively prime, then Z mn Z m Z n as rings. Hint. See Theorem 2.81. Proof: By Theorem 2.81, we have that Z mn Z m Z n as groups. Thus, we only need to show a, b Z , f (ab) = f (a)f (b) where f : Z Z m Z n by f (x) = ([x]m, [x]n). f (ab) = ([ab]m, [ab]n) = ([a]m, [a]n)([b]m, [b]n) = f (a)f (b). Z mn Z m Z n as rings. (iii) Show that if neither R nor S is the zero ring, then R S is not a domain. Proof: Since R {0} and S {0}, then r R and s S such that r 0 and s 0. Thus (r, 0) R S, (0, s) R S, (r, 0) (0, 0) and (0, s) (0, 0). However, (r, 0)(0, s) = (0, 0). Thus R S is not a domain. (iv) Show that R {0} is an ideal in R S. Proof: Let (r, s) R S, and (a, 0) R {0}, then (r, s)(a, 0) = (ra, 0) R {0} (as ra R). (v) Show that R {0} is a ring isomorphic to R, but it is not a subring of R S. Proof: Define : R R {0} by (r) = (r, 0). Then is a bijection and we have that a, b R, (a + b) = (a + b, 0) = (a, 0) + (b, 0) = (a) = (b) and (ab) = (ab, 0) = (a, 0)(b, 0) = (a)(b). R {0} is a ring isomorphic to R, However, if 0 0S, then R {0} is not even a subset of R S, hence cannot be a subring of R S. 3.50 (i) If R and S are nonzero commutative rings, prove that U(R S) = U(R) U(S), where U(R) is the group of units of R. Proof: Let (x, y) U(R S), then (w, z) R S such that (x, y)(w, z) = (xw, yz) = (1R, 1S). Thus, xw = 1 and yz = 1 where w R and z S. So x is a unit in R and y is a unit in S, hence (x, y) U(R) U(S). U(R S) U(R) U(S). Let (a, b) U(R) U(S), then c R and d S such that ac = 1R and bd = 1S. So then (a, b)(c, d) = (1R, 1S). Thus, (a, b) is a unit in R S, hence U(R) U(S) U(R S). U(R S) = U(R) U(S).

Page 5

HW #11, Ring Theory, 3.5 + Worksheet 3.50 (ii) Redo Exercise 2.65 on page 94 using part (i) (Prove U(Z 15) Z 4 Z 2). Proof: By Theorem 2.81 (If m and n are relatively prime, then Z mn Z m Z n as groups), thus Z 15 Z 3 Z 5. By part (i), U(Z 3 Z 5) = U(Z 3) U(Z 5). Since U(Z 3) = {1, 2} is a multiplicative group of order 2, then U(Z 3) Z 2. And since U(Z 5) = {1, 2, 3, 4} is a multiplicative group of order 4 in which (3) = 4 and V 4 has no element of order 4, then U(Z 5) Z 4. Thus, U(Z 15) U(Z 3 Z 5) = U(Z 3) U(Z 5) Z 4 Z 2.

a b 3.51 Let F be the set of all 2 2 real matrices of the form A = . Prove that F is a b a field (with operations matrix addition and matrix multiplication), and prove that there is an isomorphism : F C with det(A) = (A) ( A) . Hint. Define : F C by (A) = a + ib. Proof: a b c d Let A, C F such that A = , C = . b a d c a + c b + d ac bd ad + bc Then A + C = = C + A F. And AC = = CA F. (b + d ) a + c ( ad + bc ) ac bd This gives us that F is closed under addition and multiplication and commutative. 0 0 1 0 The zero matrix, F and the identity matrix, F. 0 0 0 1 a /( a 2 + b 2 ) b /( a 2 + b 2 ) If A 0F and X = , then AX = 1F. 2 2 2 2 b /( a + b ) a /( a + b ) F is a field. Define : F C by (A) = a + ib. To show is injective, let A, C F (as defined above) such that (A) = (C). Then a + ib = c + id. a = c and b = d A = C is injective. To show is surjective, let y C . Then y = w + iz for some real numbers w, z. x y W = F such that (W) = x + iy = y. y x To show is a homomorphism, define arbitrary elements A and C as above and note that (A + C) = a + c + i(b + d) = a + ib + c + id = (A) + (C) and (AC) = ac bd + i(ad + bc) = (a + ib)(c + id) = (A)(C). is an isomorphism. Lastly, we note that det(A) = a2 + b2 = (a + ib)(a ib) = (A) ( A) , as desired.

Page 6

HW #11, Ring Theory, 3.5 + Worksheet

Worksheet Problems 2. Recall that an element in a ring is called idempotent if a2 = a. Prove that in a commutative ring of characteristic 2, the set of idempotents forms a subring. Proof: Let R be a commutative ring of characteristic 2. Let T = {a R | a2 = a}. First, note that 02 = 0, so 0 T, hence T . Let a, b T. Then (a b)2 = a2 2ab + b2 = a2 + b2, since char (R) = 2. And a2 + b2 = a + b = a b + b + b = a b + 2b = a b T. Also (ab)2 = abab = a2b2 = ab T. T is a subring of R. 3. Let F be a finite field with n elements. Prove that xn1 = 1 for all nonzero x in F. Proof: First, note that F {0} is a multiplicative group with identity 1F and |F {0}| = n 1. By corollary to Lagranges theorem, x F {0}, xn1 = 1, as desired. 4. Let R be a ring with unity and let I be an ideal in R. (a) Prove I = R if and only if I contains an element of U(R). Proof: Assume I = R. Let r U(R). Since U(R) R = I, then r I. Conversely, assume I contains an element, x, of U(R). Then y R such that xy = 1. Since I is an ideal and y R, then xy = 1 I. r R, 1r = r R. Thus, R I. By definition, I R, thus I = R. (b) Prove that if R is a field, then R has exactly two ideals. Proof: Let R be a field. We know {0} is an ideal of R, so suppose I {0} is an ideal of R. Let a I where a 0. Since R is a field, then a is a unit in R. By part (a), I = R. {0} and R are the 2 and only 2 ideals of R. 4. (c) Show that a homomorphism from a field onto a ring with more than one element must be an isomorphism. Proof: Let F be a field. Let T be a ring with more than one element. Let : F T be an epimomorphism. By Proposition 3.50 (If f :A R is a ring homomorphism, then ker f is an ideal in A) we know ker is an ideal in F. Since F is a field, then by part (b), ker = {0} or ker = F. If ker = {0}, then is an injection and we are done. If ker = F, then T = {0}, a contradiction to our assumption that T is a ring with more than one element. ker = {0}, hence is a bijection.

Page 7

Debra Griffin Math 210A HW #12 Ch 3.8 #81, 82, 83 + Worksheet #2, 3, 4 Textbook Problems: 3.81 Prove that if I = {0}, then R/I R. Proof: Let : R/I R be defined by (r + {0}) = r. Since (r + {0} + s + {0}) = (r + s + {0}) = r + s = (r + {0}) + (s+ {0}) and ((r + {0})(s + {0}) = (rs + {0}) = rs = (r + {0})(s+ {0}), then is a homomorphism. is surjective since r R, (r + {0}) = r. is injective as (r + {0}) = (s + {0}) r=s rs=0 r s {0} r + {0} = s + {0}. : R/I R is an isomorphism. 2nd proof: Let i:R R be the identity map. Since i(r + s) = r + s = i(r) + i(s) and i(rs) = rs = i(r)i(s), then i is a homomorphism. We know i is surjective as r R, i(r) = r. Since {0} = ker f, then by the first isomorphism theorem, R/I R. 3.82 (Third Isomorphism Theorem for Rings) If R is a commutative ring having ideals I J, then J/I is an ideal in R/I and there is a ring isomorphism (R/I)/(J/I) R/J. Proof: We first show J/I is an ideal in R/I. Let a J/I and b R/I. Then a = j + I for some j J and b = r + I for some r R. Since j J R, then jr R, hence ab = (j + I)(r + I) = jr + I R/I. Thus, J/I is an ideal in R/I. We next will show (R/I)/(J/I) R/J. Define f : R/I R/J by f (r + I) = r + J. To verify f is well-defined, let c, d R/I such that c = d. Then c = r + I and d = s + I for some r, s R. Thus, r s I J, so r s J. Hence r + J = f (r + I) = f (s + I) = s + J, and we have f is well-defined. We verify surjectivity of f by letting r + J R/J and noting that f (r + J) = r + I. The kernel of f = {r + I R/I | f (r + I) = r + J = J}. Since r + J = J r J, then ker f = J/I. So, by the first isomorphism theorem, (R/I)/(J/I) R/J.

HW #12, Ring Theory, 3.5/3.8 + Worksheet 3.83 For every commutative ring R, prove that R[x]/(x) R. Proof: Let R be a commutative ring. Let f (x) = a0 + a1x + + anxn, where a0, a1, , an R. Define : R[x] R by (f (x)) = a0. We know is surjective as r R, (r) = r, where r is a polynomial of degree 0. We will now check that is a ring homomorphism. Let f, g R[x]. Then f = a0 + a1x + a2x2 + + anxn for some a0, a1, , an R, and g = b0 + b1x + b2x2 + + bnxn for some b0, b1, , bn R. Since R is commutative, we have (f + g) = (a0 + b0 +( a1 + b1)x + +(an + bn) xn) = a0 + b0 = (f ) + (g). n i And (fg) = ( a b j k x ) = a0b0 = (f )(g). i= 0 j + k = i Note that ker = {f R[x]| (f (x)) = (a0 + a1x + + anxn) = a0 = 0R}. Let g (x), then g = xf for some f R[x], and xf = x(a0 + a1x + + anxn). Relabeling the coefficients, we have x(a0 + a1x + + anxn) = b0 + b1x + + bnxn where b0 = 0. g ker . Let f ker , then (f (x)) = (a0 + a1x + + anxn) = a0 = 0R. Since a0 = 0R, then f (x) = 0 + x(a1 + a2x + + anxn) = x(g(x)) for some g R[x], hence f (x). ker = (x). By the first isomorphism theorem, R[x]/(x) R.

Page 2

HW #12, Ring Theory, 3.5/3.8 + Worksheet

Worksheet Problems 2. Let F be a field of order 8. (a) Prove that char (F) = 2. Proof: Let a F with order k. Note that (1 + 1)(1 + 1)(1 + 1) = 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 = 0. 1 + 1 = 0, as F has no zero divisors. Hence (1) = 2. Thus a(1 + 1) = 2a = 0. And this gives us that char (F) = 2. (b) Suppose a, b F with a2 + ab + b2 = 0. (i) Prove a3 = b3. Proof: Since a3 b3 = (a b)(a2 + ab + b2) = (a b) 0 = 0, then a3 = b3. (ii) Prove a2 = b2. Proof: If a = 0 or b = 0 , then a2 = b2. If a 0 and b 0, then a, b U(F). And since U(F) is a multiplicative group of order 7, then a7 = b7 = 1. Note that a3 = b3 a9 = a3 a3 a3 = b3 b3 b3 = b9. Thus, a2 = a2 1 = a2 a7 = a9 = b9 = b7 b2 = 1 b2= b2, as desired. (iii) Prove a = b = 0. Proof: Since a2 = b2, then (a + b)(a b) = (a b)(a b) = 0. Thus a = b. And 0 = a2 + ab + b2 = 3b2 = 2b2 + b2 = b2, thus 0 = b = a. 3. Let R be a commutative ring. Let S be a commutative ring without zero-divisors. Let : R S be a ring homomorphism with K = ker . Prove that for any a, b R, if ab K, then a K or b K. Proof: Let a, b R, such that ab K, then (ab) = (a)(b) = 0S. Since S has no zero-divisors, then (a) = 0S or (b) = 0S. Thus, a K or b K.

Page 3

HW #12, Ring Theory, 3.5/3.8 + Worksheet 4. Let A = 2Z and B = 8Z. (a) Show that the group A/B is isomorphic to the group Z 4. Proof: First, note that a A, a = 2h for some h Z. Define : A Z 4 by (a) = [h] (where h Z and a = 2h). We first will show is a homomorphism. Let a, b A. Then a = 2j, b = 2k for some j, k Z. (a + b) = [j + k] = [j] + [k] = (a) = (b). is a group homomorphism. Next, we show B = ker . Let b B, then b = 8h for some h Z. (b) = [4h] = [0]. So b ker . Let k ker , then k = 2h for some h Z and (k) = [h] = [0]. Thus, 4h, hence 8k. This gives us that k B. B = ker . So, by the first isomorphism theorem, A/B is isomorphic to the group Z 4. (b) Show that the ring A/B is not isomorphic to the ring Z 4. Proof: We know 2Z does not have a unity. Since 2Z/8Z = {2i + {8Z}: 0 i < 4}, and j 0, we have that (2j + {8Z}) (2i + {8Z}) 2i + {8Z}, then 2Z/8Z does not have a unity. However, Z 4 does have a unity, [1], hence, by basic ring homomorphism properties, A/B is not ring homomorphic to the ring Z 4. Thus, A/B is not isomorphic to the ring Z 4.

Page 4

Debra Griffin Math 210A HW #13 Ch 3.2 #17, Ch 3.4 #29, Ch 3.5 #46, 47, 48(i) 3.17 (i) Show that F = {a + bi: a, b Q} is a field. Proof: Since F C, and C is a field, then we know + and are well-defined, associative and commutative and 1, 0 F. We only need to show closure under + and and the existence of inverses for each element in F. Let w, z F. Then w = a + bi and z = c + di for some a, b, c, d Q. Since w + z = a + bi + c + di = a + c + (b + d)i where a + c, b + d Q, and wz = (a + bi)(c + di) = ac bd + (ad + bc)i where ac bd, ad + bc Q, then w + z, wz Q. Thus, F is closed under + and . a b Since (a + bi)( 2 2 i ) = 1 + 0i, and a2 + b2 0, then each element of F has a 2 a + b a + b2 multiplicative inverse. F is a field. (ii) Show that F is the fraction field of the Gaussian integers. Proof: Let x F. Then for some a, b Q where a = p/q and b = r/s where p, q, r, s Z such p r ps + rqi that q 0 and s 0, we have x = a + bi = + i = . Since ps, rq, qs Z, then q s qs + 0i ps + rqi, qs + 0i Frac(Z[i]). So F Frac(Z[i]). Let y Frac(Z[i]). Then for some a, b, c, d Z, where c and d are not both 0, we have a + bi ac bd + ( ad + bc )i = ac bd + ad + bc i . that y = = 2 2 c + di c +d c2 + d2 c2 + d2 Since c2 + d2 0, then y F. So Frac(Z[i]) F. F = Frac(Z[i]), as desired.

HW #13, 3.2/3.4/3.5 (Frac(R)) 3.29 Let R be a domain. If f (x) R[x] has degree n, prove that f (x) has at most n roots in R. Hint. Use Frac (R). Proof:
n

Let f (x) R[x] with degree n, then f (x) =


n

a x , a R.
i i
i

i= 0

Since Frac (R) is a field, then g(x) =

[a ,1]x
i i= 0

, ai R has at most n roots in Frac (R).


n

Suppose f (x) has n + 1 roots in R , say {1, 2, , n+1}. Assume [n+1, 1] is not a root of g(x). Since f (n+1) =
n n

a
i i= 0 n i= 0

i n +1

= 0,

then g([n+1, 1]) =

[a ,1][
i i= 0

n +1

,1] =

[a ,1][
i i= 0

i n +1

,1] =

[ a
i

i n +1

,1] = [0, 1], a

contradiction to our assumption that [n+1, 1] is not a root of g(x). If f (x) R[x] has degree n, prove that f (x) has at most n roots in R. 3.46 If R is a field, show that R Frac (R). More precisely, show that the homomorphism f : R Frac (R) in Example 3.45 (i), namely, r [r, 1], is an isomorphism. Proof: Define f : R Frac (R) by r [r, 1]. To show f is injective, we note that a ker f, f (a) = [a, 1] = [0, 1]. ker f = {0}, hence f is injective. To show f is surjective, we let y Frac (R) and note that y = [a, b] for some a, b R, b 0. Since R is a field, then ab1 R, thus [a, b] = [ab1, 1]. So f (ab1) = y. f is surjective. To show f is a homomorphism, we note that f (a + b) = [a + b, 1] = [a, 1] + [b, 1] = f (a) + f (b) and f (ab) = [ab, 1] = [a, 1][b, 1] = f (a) f (b). R Frac (R).

Page 2

HW #13, 3.2/3.4/3.5 (Frac(R)) 3.47 (i) If A and R are domains and : A R is a ring isomorphism, prove that [a, b] [(a), (b)] is a ring isomorphism Frac (A) Frac (R). Proof: Define f : Frac (A) Frac (R) by [a, b] [(a), (b)]. To show f is injective, we note that [a, b] ker f, f ([a, b]) = [(a), (b)] = [0, (b)] where (b) 0A. Since is an isomorphism, then [a, b] = [0, b] ker f = {[0, b] }, hence f is injective. To show f is surjective, we let y Frac (R) and note that y = [u, v] for some u, v R, v 0. Since is surjective, then a, b A such that (a) = u and (b) = v. So f ([a, b]) = [(a), (b)] = [u, v]. f is surjective. To show f is a homomorphism, we note that since is an isomorphism, we have f ([a, b] + [c, d]) = f ([ad + bc, bd]) = [(ad + bc), (bd)] = [(ad) + (bc), (bd)] = [(a)(d) + (b)(c), (b)(d)] = [(a), (b)] + [(c), (d)] = f ([a, b]) + f ([c, d]). And f ([a, b][c, d]) = f ([ac, bd]) = [(ac), (bd)] = [(a)(c), (b)(d)] = = [(a), (b)][(c), (d)] = f ([a, b]) f ([c, d]). f is a homomorphism. In conclusion, we have that f is an isomorphism. 3.47 (ii) Prove that if a field k contains an isomorphic copy of Z as a subring, then k must contain an isomorphic copy of Q. Proof: We have : Z (Z) where (Z) k is an isomorphism. Define f : Q k by f (a/b) = (a)(b)1 where a, b Z and b 0. To show f is injective, we note that if x ker f, then x = p/q where q 0 and f (p/q) = (p)(q)1 = 0. Since is an isomorphism and k is a field, we have (q)1 0, thus (p) = 0. And so p = 0 = x, hence ker f = {0}. f is injective. To show f is a homomorphism, we note that since is an isomorphism and k is a field, then for a/b, c/d Q (a, b, c, d Z, b 0, d 0), we have f (a/b + c/d) = f ((ad + bc)/(bd)) = (ad + bc)(bd)1 = [(ad) + (bc)](bd)1 = (a)(d)(b1)(d1) + (b)(c)(b1)(d1) = (a)(b1) + (c)(d1) = f (a/b) + f (c/d). And f (a/b c/d) = f ((ac)/(bd)) = (ac)(bd)1 = (a)(c)(b1)(d1) = f (a/b) f (c/d). f is a homomorphism. So, since f (Q) k and f : Q f (Q) is surjective, then f : Q f (Q) is an isomorphism. k contains an isomorphic copy of Q.

Page 3

HW #13, 3.2/3.4/3.5 (Frac(R)) 3.47 (iii) Let R be a domain and let : R k be an injective ring homomorphism, where k is a field. Prove that there exists a unique ring homomorphism : Frac (R) k extending ; that is, |R = . Proof: We have that : R k is an injective ring homomorphism. Define : Frac (R) k by ([a, b]) = (a)(b)1. Then ([a, b] + [c, d]) = ([ad + bc, bd]) = [(ad + bc)](bd)1 = [(ad) + (bc)](bd)1 = [(a)(d) + (b)(c)](b)1(d)1 = (a), (b)1 + (c), (d)1 = ([a, b]) + ([c, d]). And ([a, b][c, d]) = ([ac, bd]) = (ac), (bd)1 = (a)(c), (b)1(d)1 = = [(a)(b)1][(c)(d)1] = ([a, b]) ([c, d]). is a homormophism. Since ([r, 1]) = (r)(1)1 = (r), then |R = . To show uniqueness, we suppose 2: Frac (R) k such that 2|R = . Then |R = = 2|R. Since ([a, b]) = (a)(b)1 = ([a, 1])([b, 1])1 = ([a, 1])([1, b]) = 2([a, 1])2([1, b]) = 2([a, 1][1, b]) = 2([a, b]), then = 2. 3.48 R is a domain with fraction field F = Frac (R). (i) Prove that Frac (R[x]) F(x). Proof: Note that
n

f R[x], f (x) = h F[x], h(x) =

a x , a R, n N, and
i i
i

i= 0 n

[a ,b ]x , a , b R, b 0.
i i i
i i i

i= 0

Define : Frac (R[x]) F(x) by


n1 n2

n1 i i i

n2 i i

([f (x), g(x)]) = ([ ai x , i= 0


i

b x ]) = [ [a , 1]x , [b , 1]x ].
i i= 0 i= 0 i= 0

n i

To show is injective, note that if [f (x), g(x)] = [ ai x ,


i= 0

b x ] ker , then
i i i= 0 n

[ [ ai , 1] x i , [bi , 1] x i ] = 0F(x). Thus,


i= 0 i= 0

ai , 1] x i = 0F[x]. So [
i= 0

a x
i i= 0

= 0R[x].

ker = {0R[x]}, hence is injective.

Page 4

HW #13, 3.2/3.4/3.5 (Frac(R)) 3.48 (i) (continued)


n n

To show is surjective, let [h (x), j(x)] = [ [ ai , bi ] x i , [c i , di ] x i ], and let


i= 0
n1 n2 i

i= 0

r=

b
i= 0

where bi = 1 if ai = 0, and s =
n n

d
i= 0

where di = 1 if ci = 0.
n

Then [rs h (x), rs j(x)] Frac ( R[x]) and ([ rs h (x), rs j(x)]) =


n

[rs [ ai , bi ] x , rs [c i , di ] x ] = [ [ ai , bi ] x , [c i , di ] x i ]. i= 0 i= 0 i= 0 i= 0 To show is a homomorphism, we will simplify our notation to


i i i n n

([f , g]) = [f , g] where ([f , g]) = ([f (x), g(x)]) = [ [ ai , 1] x i , [bi , 1] x i ] = [f , g]. i= 0 i= 0
n1 n2
np np

Note that (fg) = ( ai x i


i= 0
np

bi x i ) = (
i= 0 n2 i= 0

a j bk x i ) =

[a b , 1]x
j k i= 0 i= j + k

i= 0 i= j + k

[a j , 1][bk , 1]x i = [ai , 1]x i [bi, 1]x i = fg (where np = max(n1, n2).


n1 i= 0 n2
np i bi x ) = ( (ai + bi ) x i ) = np

i= 0 i= j + k

n 1 And (f + g) = ( ai x i +
i= 0

([a + b , 1]) x
i i i= 0

i ([ai, 1] + [bi, 1]) x = = [ai , 1]x i + [bi, 1]x i = f + g (where np = max(n1, n2).
np i= 0

i= 0 n1

i= 0 n2

i= 0

i= 0

(gj)] = [f j + hg, gj] = So ([ f, g] + [h, j]) = [fj + hg, gj] = [(fj + hg), [f , g] + [h, j] = ([f, g]) + ([h, j]). And ([f, g][h, j]) = [fh, gj] = [(fh), (gj)] = [f h, gj] = [f , g][h, j] = ([f, g])([h, j]). is a homomorphism. In conclusion, Frac (R[x]) F(x).

Page 5

You might also like