Appendix A

Exercises A 1. 1, 2, 3, 4 5. none 9. 1, 2 15. 5 1, 3, 5, 7, p 6 23. 5 5n : n H 6 3. 5 1, 2 6, 5 3, 4 6 11. none 17. 5p , 5, 1, 3, 7, p 6 21. 5 3n 2 : n H 6 25. A B C, D E

7. 2, 1, 0, 1, 2, 3

13. 3, 2, 1, 0, 1, 2, 3, 4, 6 19. 5n H : n 36

27. A B 5 2, 4 6, A B 5 1, 2, 3, 4, 5, 6, 8 6
25

29. A B , A B 5 1, 2, 3, 5, 7, 10, 15, 17, 26, 316 31. a 2k

k1 11

33. a 1> 2k1

8 k1

35. a 1 3k 2 2
k1 2 2 i0 j0

37. a r k 1
k1 2

n1

39. a a i j a 1 i i 1 i 2 2 1 0 1 2 2 1 1 0 1 2 1 2 1 0 2 8,
2 2 2 i0

j0 i0

a a i j a 1 j 1 j 2 j 2 10 1 2 2 11 0 1 2 12 1 0 2 8
j0

45. (By contrapositive) If n is not odd, then n is even. Thus, n 2k for some integer k and so 3n 1 3 1 2k 2 1 2 1 3k 2 1, which is odd. Hence 3n 1 is not even. 47. (By contradiction) Suppose that m n is odd but it is not the case that one of m or n is even and the other is odd. Then either m and n are both even or they are both odd. In either case, m n is even, a contradiction. Conclude that one of m or n is even and the other is odd. 49. 3 1 4 (By contrapositive) Assume that it is not the case that both m and n are even. If m and n are both odd, then mn is odd; if only one of m or n is even, then m n is odd. In either case, it is not the case that mn and m n are both even. 3 B 4 Assume that both m and n are even. Then clearly mn and m n are both even. 51. If 12 is rational, then it can be written in the form 12 a> b, where a and b are integers with no common factors. Then 2 a2> b2, so a2 2b2 and, hence, a2 is even. Hence, a is even (by Exercise 49 with m n a) and so a 2k for some integer k. But then 4k2 2b2, so b2 2k2 and, hence b2 is even. Hence, b is even and a and b have a common factor of 2, a contradiction. Conclude that 12 must be irrational.

n3 n 1 2k 1 2 3 1 2k 1 2 8k3 12k2 4k 2 1 4k3 6k2 2k 2 , which is even.

43. If n is even, then n 2k for some integer k and so n3 n 1 2k 2 3 2k 2 1 4k3 k 2 , which is even. If n is odd, then n 2k 1 for some integer k and so

41. If n is even, then n 2k for some integer k. Then 3n 5 3 1 2k 2 5 2 1 3k 3 2 1, which is odd.

Appendix B
Exercises B 1. For n 1, we have 1 2 # 12 1. Assume that 1 5 9 p 1 4k 3 2 2k2 k. Then
1

Answers to Selected Odd-Numbered Exercises

3. For n 1, we have 1 Assume that 12 22 32 p k2 k 1 k 1 2 1 2k 1 2> 6. Then 12 22 p k2 1 k 1 2 2 k 1 k 1 2 1 2k 1 2> 6 1 k 1 2 2 1 k 1 k 1 2 1 2k 1 2 6 1 k 1 2 2 2> 6 1 k 1 2 1 k 1 2k 1 2 6 1 k 1 2 2> 6 1 k 1 2 1 2k2 7k 6 2 > 6 1 k 1 2 1 k 2 2 1 2 1 k 1 2 1 2> 6.
2

1 5 p 141k 1 2 3 2 1 5 p 1 4k 3 2 1 4k 1 2 2k2 k 4k 1 2 1 k2 2k 1 2 1 k 1 2 2 1 k 1 2 2 1k 1 2 .

17. For n 0, we have 1 ab 2 0 1 1 # 1 a0b0. Assume that 1 ab 2 k akbk. Then 1 ab 2 k 1 1 ab 2 k 1 ab 2 akbkab akabkb ak 1bk 1. 19. For n 1, we have x11 x1, which is certainly divisible by x1. Assume that xk1 is divisible byx1, so that xk1 1 x1 2 f 1 x 2 , for some polynomial f 1 x 2 . Then xk 11 xk 1xk xk1 1 x1 2 xk 1 x1 2 f 1 x 2 1 x1 2 1 xk f 1 x 2 2 , which is divisible by x1. 21. For n 0, we have a set with no elements: the empty set . The only subset of is itself, so has 1 20 subsets. Assume that any set with k elements has 2k subsets. Now let S be a set with 2k 1 elements, say S 5 x1, x2, p , xk, xk1 6. If A S, then either xk 1 H A or xk 1 x A. If xk 1 H A, then A 5 xk1 6 A , where A is a subset of 5 x1, x2, p , xk 6 ; by the induction hypothesis, there are 2k such subsets. If xk 1 x A, then A is a subset of 5 x1, x2, p , xk 6 ; by the induction hypothesis, there are 2k such subsets. It follows that the total number of subsets of S is 2 # 2k 1 2k 2, as was required to be proved. 23. Hint: The basis step is for n 3, in which case we have a triangle and the sum of its interior angles is 180 1 32 2 180 . Assuming that a convex k-gon has an interior angle sum of 1 k2 2 180 , consider a convex 1 k 1 2 -gon P. Subdivide P into a triangle and a k-gon. 25. n>1 n 1 2 .

1 1 1 1 1 2 1 2 # 1 1 2 > 6.

11. For n 0, 501 11 0, which is divisible by 4. Assume that 5k1 is divisible by 4, so that 5k1 4m for some integer m. Then 5k 11 5k 15k 5k1 1 51 2 5k 1 5k1 2 4 # 5k 4m 4 1 5k m 2 , which is divisible by 4. 13. For n 5, we have 25 32 7 25 52. Assume that 2k 7 k2. Then 2k 1 2 # 2k 7 2k2. But, since k 5, k 1 k2 2 1 so k2 2k 1 and hence 2k2 k2 2k 1 1 k 1 2 2. It follows that 2k 1 7 1 k 1 2 2.
1 k 1 2 2 k k 1 k1 2 2 b .

5. For n 0, we have 1 20 11. Assume that 1 2 4 8 p 2k 2k 11. Then 1 2 4 p 2k 2k 1 1 2k 1 1 2 2k 1 2 # 2k 1 1 2k 2 1. 7. For n 1, we have 1 # 1! 1 2 1 1 1 1 2 ! 1. Assume that 1 # 1! 2 # 2! p k # k! 1 k 1 2 ! 1. Then 1 # 1! 2 # 2! p k # k! 1 k 1 2 1 k 1 2 ! 11 k 1 2 ! 1 2 1 k 1 2 1 k 1 2 ! 1 k 1 2 ! 11 1k 1 2 2 1 1k 1 2 !1k 2 2 1 1 k 2 2 ! 1. 9. For n 0, 02 0 0 is even. Assume that k2 k is even, so that k2 k 2m for some integer m. Then 1 k 1 2 2 1 k 1 2 k2 2k 1 k 1 1 k2 k 2 2k 2 2m 2 1 k 1 2 2 1 m k 1 2 , which is even.

1 15. For n 1, we have 1 2 1 1 . Assume that 1 4 1 1 1 1 1 p p 9 k 2 2k . Then 1 4 k2

27. For n 1, we have 1 20 # 1. Assume that, for all integers n such that 1 n k , n can be factored as n 2im for some integer i 0 and some odd integer m. Consider k 1. If k 1 is odd, then k 1 20 1 k 1 2 is the required factorization. If k 1 is even, then k 1 2a for some integer a. Since 1 k , k 1 2k and so a 1 k 1 2> 2 k. By the induction hypothesis, a 2im for some integer i 0 and some odd integer m. Then k 1 2a 2i 1m is the desired factorization. 29. For n 8, we have 8 3 # 1 5 # 1. Assume that, for all integers n such that 8 n k , n can be written as n 3a 5b for some nonnegative integers a and b. Consider k 1. Since 9 3 # 3 5 # 0 and 10 3 # 0 5 # 2, we may assume that k 1 11. Hence 8 1 k 1 2 3 k2 k and so k2 3a 5b for some nonnegative integers a and b, by the induction hypothesis. Then k 1 1 k2 2 3 1 3a 5b 2 3 3 1 a 1 2 5b, as required.

Now 1 k 1 2 2k k2 k 1 k2 k k1k 1 2 . Therefore,

1 4 1 k 1 2 2 k k 1 k1 2 2

1 1 k1 2 2

21 k

1 1 k1 2 2

2 a

k 1 k1 2 2 2

1 k1 2

1 1 k1 2 .

It follows that

1 p

1 1 k1 2 2

1 1 k1 2 .

Answers to Selected Odd-Numbered Exercises

31.For n 0, we have f0 0 11 f21. Assume

k k1 i0 k i0 i0

that a fi fk 21. Then a fi a a fi b fk1

f1k 22 1 fk 1 1 fk 1 fk 2 2 1 fk 3 1 f1k 12 2 1.
k k1 i0 k

disks to peg C; this takes 2k1 moves, by the induction hypothesis. Now move the largest disk to peg B (1 move) and then transfer the tower of k disks from peg C to peg B (2k1 moves). The total number of moves is thus 2 1 2k 1 2 1 2k 1 1, as required. 41. The basis step is not true.

2 # 33. For n 0, we have f 2 0 0 0 0 1 f0 f1. Assume 2 2 2 that a f 2 i fk fk 1. Then a f i a a f i b f k 1 i0 i0

fk fk 1 f k2 1 fk 1 1 fk fk 1 2 fk 1fk 2.

Appendix C
Exercises C 1. 84i 5. 74i
1 13 9. 10 10 i

35. For n 0, we have fm1f0 fmf1 fm1 # 0 fm # 1

fm fm 0. Assume that fm1fn fm fn 1 fm n for all 0 n k . Then fm 1 fk 1 fm fk 2 fm 1 1 fk 1 fk 2 fm 1 fk fk 1 2 1 fm 1 fk 1 fm fk 2 1 fm 1 fk fm fk 1 2 fm k 1 fm k fm k 1.

3. 13 11i
1 7. 1 2 2i

11. i 15. 5 17. 3

13. 10i

37. For n 1, a 2 2 board with a square removed is just a single L-tile. Assume that a 2k 2k with a square removed can be tiled with L-tiles. Consider a 2k 1 2k 1 board with a square removed. Subdivide the board into four 2k 2k quadrants. One of the quadrants contains the missing square, so it can be tiled with L-tiles, by the induction hypothesis. Now place a single L-tile at the center of the board so that it covers one square in each of the remaining three quadrants. By the induction hypothesis, the remaining squares in each quadrant can be tiled with L-tiles, and we are done.
2k II 2k 2k 1 2k 2k 1 2k I

19. 2 12 1 cos 1 p> 4 2 i sin 1 p> 4 2 2 21. 2 1 cos 1 p> 6 2 i sin 1 p> 6 2 2 23. zw 2 12 1 cos 1 11p> 12 2 i sin 1 11p> 12 2 2 , z> w 1 2> 12 2 1 cos 1 7p> 12 2 i sin 1 7p> 12 2 2 , 1> z 1 12> 2 2 1 cos 1 3p> 4 2 i sin 1 3p> 4 2 2 25. zw 8 12 1 cos 1 3p> 4 2 i sin 1 3p> 4 2 2 , z> w 2> 12 1 cos 1 p> 4 2 i sin 1 p> 4 2 2 , 1> z 1 12> 8 2 1 cos 1 p> 4 2 i sin 1 p> 4 2 2 27. 16 29. 1616 13 i 31. ; 1, ; i, ; 1 12> 2 2 ; 1 12> 2 2 i
Im i 2 2 i 2 2 2 2 i 2 2

Re

III

IV

39. For n 1, clearly a single disk can be transferred to a different peg in 1 211 move. Assume that a tower of k disks can be transferred to a different peg in 2k1 moves. Consider a tower of k 1 disks on peg A, say. In order to move it to peg B, we need to move the largest disk to peg B. To do this, we rst must transfer the top k

2 2 i 2 2 i

2 2 i 2 2

Answers to Selected Odd-Numbered Exercises

33. ; 13> 2 i> 2, i

Im

17. f 1 x 2 g 1 x 2 x 3 x 2 2x, f 1 x 2 g 1 x 2 x 3 x2 2, f 1 x 2 g 1 x 2 x 4 1

3 1 i 2 2

3 1 i 2 2 Re

19. f 1 x 2 g 1 x 2 x 4 1 1 12 2 x2 12x 2, f 1 x 2 g 1 x 2 x4 1 12 1 2 x2 12x, f 1 x 2 g 1 x 2 x6 12x5 1 1 12 2 x4 2x3 1 1 12 2 x2 12x 1 21. x 2 1 x 1 2 1 x1 2 1 27. 2> 5, 2 23. 2x 3 x 2 1 x 2 2 1 2x2 3x 6 2 12 29. no rational roots

25. x 4 x 3 3x 2 2x 2 1 x 2 x 1 2 1 x 2 2 2 31. 1> 2 ; 15> 2 33. 1, ; 12 35. 1, 2, 4 ; i 37. There is one sign change, so p has at most one positive zero. But p 1 0 2 1 and p 1 1 2 1, so there is a zero in the interval 1 0, 1 2 . Therefore p has exactly one positive zero. 38. There are no sign changes, so p has no positive zeros. Since p 1 x 2 2x3 3x2 4 has one sign change, p has at most one negative zero. We nd that p 12 2 0 and so, since p 1 0 2 0, p has exactly one real zero. Since p has degree 3, p has three zeros altogether. Hence, p has exactly two complex (nonreal) zeros. 41. There is one sign change so p has at most one positive zero. Since p 1 0 2 1 and p 1 1 2 2, there is a zero in the interval (0, 1). Since p 1 x 2 x4 5x2 3x1 also has one sign change, p has at most one negative zero. From p 1 1 2 8 and p 1 0 2 1, we see that there is a zero in the interval 1 1, 0 2 . Since p10 2 0, p has exactly two real zeros and so, because p has degree 4, it must have two complex (nonreal) zeros as well. 43. If p 1 x 2 a0 a1x p an1xn1 anxn, then p* 1 x 2 an an1x p a1xn1 a0xn.

35. i

37. e

39. (a) a bi abi a bi (c) 1 a bi 2 1 c di 2 1 acbd 2 1 ad bc 2 i 1 acbd 2 1 ad bc 2 i 1 a bi 2 1 c di 2 1 a bi 2 1 c di 2 (e) Let z a bi. If z is real, then b 0 and hence z a a z. Conversely, if z z , then a bi abi. Thus, 2bi 0 and so b 0. Hence z a is real.

41. (a) From cos 2u i sin 2u 1 cos u i sin u 2 2 1 cos 2 u sin2 u 2 i 1 2 cos u sin u 2 we see that cos 2u cos2 u sin2 u and sin 2u 2cosu sinu. (c) From cos 4u i sin 4u 1 cos u i sin u 2 4 1 cos 4 u 6 sin2 u cos 2 u sin4 u 2 i 1 4 cos3u sin u 4 cos u sin3 u 2 , we see that cos 4u cos4u 6 sin2u cos 2u sin4u and sin 4u 4cos3usinu 4cosu sin3u.

Appendix D
Exercises D 1. polynomial 3. polynomial 5. polynomial for x 0 7. not a polynomial 9. not a polynomial 11. not a polynomial 13. polynomial 15. f 1 x 2 g 1 x 2 3x2 x 1, f 1 x 2 g 1 x 2 3x2 x 3, f 1 x 2 g 1 x 2 3x3 6x2 x 2

45. First note that x1 makes sense since x 0 is not a solution of p 1 x 2 0, by Exercise 44(a). Since p is palindromic of degree 2n, we have p 1 x 2 a0x2n a1x2n1 p anxn p a1x a0, so, multiplying by xn, p 1 x 2 0 can be rewritten as a0xn a1xn1 p an p a1x1n12 a0xn 0, or a0 1 xn xn 2 a1 1 xn1 x1n12 2 p an 0. It is now enough to prove that xn xn is a polynomial of degree n in t x x1 for all n 1. For n 1, it is clear. Assume that xn xn is a polynomial of degree n in t, for all 1 n k. Then xk xk f 1 t 2 , where f has degree k, and xk1 x1k12 g 1 t 2 ,

Answers to Selected Odd-Numbered Exercises

47. (a) From a e2pi>5 cos 1 2p> 5 2 i sin 1 2p> 5 2 and a1 e2pi>5 cos 1 2p> 5 2 i sin 1 2p> 5 2

where g has degree k1. Therefore, xk 1 x1k 12 1 x x1 2 1 xk xk 2 1 xk1 x1k12 2 tf 1 t 2 g 1 t 2 , which is a polynomial in t of degree k 1.

(c) 1 15 1 2> 4

cos 1 2p> 5 2 i sin 1 2p> 5 2 , we obtain a a1 2 cos 1 2p> 5 2 .