Partial Solutions for Questions in

Appendix K of
A Companion to Analysis
T. W. K
orner

Introduction
Here is a miscellaneous collection of hints, answers, partial answers
and remarks on some of the exercises in the book. I expect that there
are many errors both large and small and would appreciate the opportunity to correct them. Please tell me of any errors, unbridgable gaps,
misnumberings etc. I welcome suggestions for additions. ALL COMMENTS GRATEFULLY RECEIVED. (If you can, please use LATEX 2
or its relatives for mathematics. If not, please use plain text. My e-mail
is twk@dpmms.cam.ac.uk. You may safely assume that I am both
lazy and stupid so that a message saying Presumably you have already
realised the mistake in question 33 is less useful than one which says
I think you have made a mistake in question 33 because not all left
objects are right objects. One way round this problem is to quote Xs
theorem.)
To avoid disappointment note that a number like K15* means that
there is no comment. A number marked K15? means that I still need
to work on the remarks. Note also that what is given is at most a
sketch and often very much less.
Please treat the answers as a last resort. You will benefit more from
thinking about a problem than from reading a solution. I am inveterate
peeker at answers but I strongly advise you to do as I say and not as
I do.
It may be easiest to navigate this document by using the table of
contents which follow on the next few pages.

Contents
Introduction
K1
K2
K3
K4
K5
K6
K7
K8
K9
K10
K11
K12
K13*
K14*
K15
K16
K17
K18
K19
K20
K21
K22
K23*
K24
K25
K26
K27
K28
K29
K30
K31
K32
K33
K34*
K35*
K36
K37*
K38
K39
K40
K41
K42
K43

2
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53

K44
K45
K46*
K47
K48
K49
K50
K51
K52
K53
K54
K55
K56
K57
K58
K59
K60
K61
K62
K63
K64
K65
K66
K67
K68
K69
K70
K71
K72
K73
K74
K75
K76
K77
K78
K79*
K80
K81
K82
K83
K84
K85
K86
K87
K88

54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99

K89
K90
K91
K92
K93
K94
K95
K96
K97
K98
K99
K100
K101
K102
K103
K104
K105
K106
K107
K108
K109
K110
K111
K112
K113
K114
K115
K116
K117
K118
K119
K120
K121
K122
K123
K124
K125
K126
K127
K128
K129
K130
K131
K132
K133

100
101
102
103
104
105
106
107
108
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145

K134
K135*
K136
K137*
K138
K139
K140
K141
K142
K143
K144
K145
K146*
K147
K148
K149
K150
K151
K152*
K153
K154
K155
K156
K157
K158
K159
K160
K161
K162
K163
K164
K165
K166
K167
K168
K169
K170
K171
K172
K173*
K174
K175*
K176
K177
K178

146
147
148
149
150
151
152
153
154
155
158
160
161
162
163
164
165
166
167
168
169
170
171
172
173
175
176
177
178
179
180
181
183
184
185
186
187
188
190
191
192
193
194
195
196

K179
K180
K181
K182
K183
K184*
K185
K186*
K187
K188
K189
K190
K191
K192
K193
K194
K195
K196
K197
K198
K199
K200
K201
K202
K203
K204*
K205
K206
K207
K208
K209
K210
K211
K212
K213
K214
K215
K216
K217
K218
K219
K220
K221
K222
K223

198
199
200
202
203
204
205
206
207
208
209
210
211
213
214
215
216
218
219
220
221
222
223
224
226
227
228
229
231
232
234
235
236
237
239
241
242
243
244
245
246
247
248
249
250

K224
K225
K226
K227
K228
K229
K230
K231
K232
K233
K234
K235
K236
K237
K238
K239
K240
K241
K242
K243*
K244
K245
K246
K247
K248*
K249*
K250
K251*
K252*
K253
K254
K255
K256
K257
K258
K259
K260
K261
K262
K263
K264
K265
K266
K267
K268

251
252
253
254
256
257
259
261
263
264
265
267
268
270
272
273
274
275
277
278
279
280
281
282
284
285
286
287
288
289
290
291
292
293
294
296
297
299
301
303
305
307
308
309
310

K269
K270
K271
K272
K273
K274
K275
K276
K277
K278
K279
K280
K281
K282
K283
K284
K285
K286
K287
K288
K289
K290
K291
K292
K293
K294
K295
K296*
K297
K298
K299
K300
K301
K302
K303
K304
K305
K306
K307
K308
K309
K310
K311
K312
K313

311
312
313
314
316
318
319
320
321
322
323
324
325
326
327
328
329
331
332
334
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
352
353
355
356
357
359
361
363
364
365

10

K314
K315
K316
K317
K318
K319
K320
K321
K322
K323
K324
K325
K326
K327
K328
K329*
K330
K331
K332
K333
K334
K335
K336
K337
K338
K339
K340
K341
K342
K343
K344
K345

366
368
369
370
371
373
374
375
377
378
379
381
382
384
385
386
387
388
389
391
392
393
394
395
396
400
401
402
403
404
405
406

11

K1
(ii) Every non-empty set of positive integers has a least member.
Since x is a rational number, qx is an integer for some strictly positive
integer. We can certainly find a k such that k + 1 > x k (ultimately
by the Axiom of Archimedes). But x is not an integer so k +1 > x > k.
Since mx, m and k are integers, m0 = mx mk is. Also
m0 x = mx2 mkx = mN k(mx)

is an integer since mx is. Since 1 > x k > 0 we have m > mx mk =

m0 > 0 and m > m0 1 contradicting the definition of m as least
strictly positive integer with mx an integer.

12

K2
Write cn = dn + d1
n . Then
d2n cn dn + 1 = 0
p
cn c2n 4
.
dn =
2
Since the product of the two roots of F is 1, we must take the bigger
root. Thus
p

cn + c2n 4
c + c2 4

dn =
2
2
where c is the limit of cn .

For the second paragraph, just take

d2n = (1 + k)/2 and d2n+1 = 2/(1 + k).
For the third paragraph, the condition |dn | 1 will do but we must
argue carefully. Look first to see which root of d2 cd + 1 = 0 lies
outside the unit circle.

13

K3
(Second part of problem.) Choose a1 = 2, b1 = 1, say.

14

K4
Take x = 0, f (t) = H(t), g(t) = 0. Take f (t) = 0, g(t) = H(t).
(H(t) = 1 for t > 0, H(t) = 0 for t 0.)

15

K5
(i), (ii) and (iii) are true and can be proved by, e.g. arguing by cases.
(iv) is false since f is continuous at 1/2. (f is, however, discontinuous
everywhere else.)
(v) is false. Take e.g. x = 161 + 171/2 , y = 31 171/2 .

16

K6
PN

n=1

by 1.

2n H(x qn ) is an increasing sequence in N bounded above

Observe that if x y then H(x qn ) H(y qn ) 0 for all n 0.

Thus
N
N
X
X
n
2 H(x qn )
2n H(y qn )
n=1

N
X
n=1

n=1

2n H(x qn ) H(y qn ) 0

for all N so f (x) f (y).

Observe that, if x > qm , then H(x qn ) H(qm qn ) 0 for all

n 0 and H(x qm ) H(qm qm ) = 1 so
N
X

n=1

H(x qn )

N
X
n=1

2n H(qm qn ) 2m

for all N m so f (x) f (qm ) + 2m for all x > qm . Thus f is not

continuous at qm .
If x is irrational, we can find an > 0 such that |qm x| > for all
m M and so

N
N
X

2n H(x qn )
2n H(y qn ) 2M

n=1

n=1

whenever |x y| < . Thus f is continuous at x.

17

K7
Observe that an+1 an is decreasing. Thus either
(a) an+1 an > 0 for all n and, if an is unbounded an , or
(b) There exists an N such that an+1 an 0 for all n N and, if
an is unbounded, an .
For the reverse inequality, observe that an satisfies the original
inequality.

18

K8
A decreasing sequence bounded below tends to a limit, so, for any
fixed x (0, 1),
xn = fn (x) y, say.

Since x > xn > 0 we have 1 > y 0. If y > 0 then by continuity

xn+1 = f (xn ) f (y)

and so y = f (y). Thus y = 0.

If
f (x) =

1
4
x
4

x
2

for x > 1/2,

for x 1/2,

then fn (x) 1/2 for 1 > x > 1/2.

We suppose our calculator works in radians. If an+1 = sin an then,
whatever a0 we start from, we have |a1 | 1. If a1 = b1 and bn+1 =
sin bn then an = bn for n 1. If 0 < a < 1, then 0 < sin a < a.
(If our calculator works in degrees, matters are simpler and less interesting.)

19

K9
(ii) we need only look at positive integers. If n 2(k + 1)
n
X
n j
n
(1 + ) =

j
j=0

k+1
k+1
(n/2)k+1 k+1

(k + 1)!
nk+1
k+1
k+1
2 (k + 1)!
so
nk (1 + )n n
as n .

k+1

2k+1 (k + 1)!

20

K10
Observe that

x1 + x 2 + + x n
x1 + x2 + + xn1
xn 1
n
=
(n 1)
n n
n
n1

x1 + x 2 + + x n
1 x1 + x2 + + xn1
=
1
n
n
n1
1 (1 0)1 = 0

as n .

Either there exists an N such that m(n) = m(N ) for all n N and
the result is immediate or m(n) and
xm(n)
xm(n)

0
n
m(n)
as n .
If > 1, then
x1 + x2 + + xn xm(n) 1
x1 + x2 + + xn
1 0 = 0.

n
n
n
If < 1, then
x1 + x2 + + xn xm(n) 1
x1 + x2 + + xn

.
n
n
n

21

K11
(i) True. If a + b = c and a, c Q then b = c a Q.

(ii) False. 0 2 = 0. (However the product of an irrational number

with a non-zero rational number is irrational.)
(iii) True. By the axiom of Archimedes we can find a strictly positive
integer N with N 1 < and an
integer M such that mN 1 x <

(m + 1)N 1 . Set y = mN 1 + 2/(2N ).

(iv) False. Take xn = 21/2 /N .

(v) False.
If a 2 is rational, then, since ais irrational,
the statement

is false. If a 2 is irrational, then, since (a 2 ) 2 = 2 the statement is

false.

22

K12
P
P
j
n1j
j n1j
|.
| |x y| n1
(i) |xn y n | = |(x y) n1
j=0 |x| |y|
j=0 x y
P
j
(ii) If P (t) = N
j=0 aj t , then
|P (x) P (y)| |x y|

N
X
j=0

aj |j|Rj1 .

(vi) An increasing sequence bounded above converges.

N
X

k=J+1

10k! 10(J+1)!

N
X

k=J+1

10k(J+1)

10
10(J+1)! .
9

23

K13*
No comments.

24

K14*
No comments.

25

K15
Apply the intermediate value theorem.
an converges (as in lion hunting) to c say.
Suppose f 0 (c) > 0. We can find an > 0 such that

f 0 (c)
f (t) f (c)
0
<

f
(c)

tc
2

and so

f (t) f (c)
f 0 (c)
>
tc
2
for all t with |t c| < . Thus f (t) > f (c) for c < t < c + and
f (t) < f (c) for c < t < c

Choose n such that 2n < . We have f (an ) < f (bn ) which is absurd.

26

K16
Repeat the lion hunting argument for the intermediate value theorem
with f (an ) g(an ), f (bn ) g(bn ). If an c, say, then, since f is
increasing this gives
g(bn ) f (bn ) f (c) f (an ) g(an ).

If an c, say, then since g is continuous g(an ) g(c) and so f (c) =

g(c).
First sentence of second paragraph. False. Consider [a, b] = [1, 1],
g(x) = 0, f (x) = x 1 for x 0, f (x) = x + 1 for x > 0.
Second sentence false. Consider [a, b] = [1, 1], g(x) = x + 3,
f (x) = x 1 for x 0, g(x) = x + 3, f (x) = x + 1 for x > 0.
Third sentence true. Apply intermediate value theorem to f g.

27

K17
Let > 0. Choose k > 0 such that |fk (x) f (c)| < for all
|x c| < k . Set = min1kN k . We claim that |g(x) g(c)| < for
all |x c| < .
To see this, suppose, without loss of generality, that fN (c) = g(c).
Then
and

g(x) fN (x) > fN (c) = g(c)

fj (x) fj (c) + g(c) +

for all 1 j N and so

for all |x c| < .

g(x) g(c) +

For the example we can take (a, b) = (1, 1), c = 0 and

if x 0,
0
fn (x) = nx
if 0 x < 1/n,

1
if 1/n x.

28

K18
(i) Observe that
{x R : E (, x] is countable}

is a non-empty bounded set and so has a supremum say. S

Observe
that En = E (, 1/n] is countable so E (, ) =
n=1 En
is countable and so E (, ] is countable. Thus E (, ) is
uncountable if and only if > . Similarly we can find a such that
E (, ) is uncountable if and only if < . Since E is uncountable,
> .
(ii) We have one of four possibilities.
(a) There exist and with > and
{e E : e < } and {e E : e > }

are uncountable if and only if < < .

(b) There exists an with

{e E : e < } and {e E : e > }

are uncountable if and only if < .

(c) There exists a with

{e E : e < } and {e E : e > }

are uncountable if and only if < .

(d) We have

{e E : e < } and {e E : e > }

uncountable for all .

All these possibilities can occur. Look at Z[1, 1], (0, ), (, 0)

and R.
(iii) Set E = {1/n : n 1, n Z}

29

K19
(i) False. Consider xj = 1.
(ii) False. Consider xj = j.

(iii) False. Consider xj = (1)j .

(iv) False. Consider xj = (1)j .

30

K20
(i) Let > 0. There exists an N such that an lim supr ar +
for n N . There exists an M such that n(p) N for p M . Now
lim supr ar + an(p) for p M . Since was arbitrary,
lim sup ar a.
r

(ii) Take an = (1 + (1)n )/2.

Take e.g. a2n +r = r/2n for 0 r < 2n , n 0.

(iii) Take an = bn = (1)n .

(True.) Let > 0. If n is large enough

lim sup ar + > an and lim sup br + > bn
r

so
lim sup ar + lim sup br + 2 > an + bn .
r

Thus
lim sup ar + lim sup br + 2 > lim sup(an + bn ).
r

But is arbitrary so
lim sup ar + lim sup br lim sup(an + bn ).
r

(True) Similarly if > 0 and n is large enough

bn lim inf br
r

and so
an + bn an + lim inf br .
r

Thus
lim sup(an + bn ) lim sup an + lim inf br
n

and, since was arbitrary,

lim sup(an + bn ) lim sup an + lim inf br .
n

31

K21

Thus

2
2

x
y
= kxk 2 x y + kyk

kxk2 kyk2
kxk2
kxkkyk kyk2
kyk2 2x y + kxk2
=
kxk2 kyk2
2

kx yk
.
=
kxkkyk

x
kx yk
y

kxk2 kyk2 = kxkkyk .

Ptolomeys result now follows from the triangle inequality for points of
the form kxk2 x.

32

K22
(i), (ii) and (iii) are true. Proof by applying the definitions.
(iv) is false. Take, for example, n = 1, U = {x : |x| < 1}.

33

K23*
No comments.

34

K24
(i) If (xn , yn ) E and k(xn , yn ) (x, y)k 0, then xn x and
yn y. Thus yn = 1/xn x and x = y so (x, y) E.
1 (E) = {x : x > 0} is not closed since 1/n 1 (E) and 1/n 0
/
1 (E) as n 0.
(ii) Let E = {(x, 1/x) : x > 0} {(1, 0)}. Then E is closed (union
two closed sets), 1 (E) = {x : x > 0} is not closed but 2 (E) = {y :
y 0} is.
(iii) If xn 1 (E) and xn x we can find yn such that (xn , yn ) E.
Since yn is a bounded sequence there exists a convergent subsequence
yn(j) y as j . Since xn(j) x the argument of (i) shows that
(x, y) E and x 1 (E). (It is worth looking at why this argument
fails in (ii).)

35

K25
(i) If G [K, K]n+m , then E [K, K]n .
(ii) Take n = m = 1, E = (0, ), f (x) = 1/x See question K.24.
(iii) Suppose k(xj , f (xj )) (x, y)k 0. Then kxj xk 0. Since
E is closed, x E. Since f is continuous, kf (xj ) f (x)k 0. Thus
y = f (x) and (x, y) G.
(iv) Take n = m = 1, E = R and f (x) = x1 for x 6= 0, f (0) = 0.

(v) Suppose xj E and kxj xk 0. Since G is closed and bounded

we can find a subsequence j(k) such that
k(xj(k) , f (xj(k) )) (x, y)k 0

for some (x, y) G. Since (x, y) G, we have y = f (x).

Observe that x E, so E is closed. If f was not continuous at
x, we could find a > 0 and xj E and kxj xk 0 such that
kf (xj ) f (x)k > and so
k(xj(k) , f (xj(k) )) (x, y)k >

for every j(k). Our result follows by reductio ad absurdum.

36

K26
If we write f (x) = 0 for x 6= 0 and f (0) = 1, then f is upper
semicontinuous but not continuous.
If f is not bounded we can find xn E such that f (xn ) n. Extract
a convergent subsequence and use upper semicontinuity to obtain a
contradiction. To show that the least upper bound is attained, take a
sequence for which f approaches its supremum, extract a convergent
subsequence and use upper semicontinuity again.
Take n = 1, K = [0, 1], f (x) = 1/x for x 6= 0 , f (0) = 0 to obtain
an upper semicontinuous function not bounded below.
Take n = 1, K = [0, 1], f (x) = x for x 6= 0 , f (0) = 1 to obtain an
upper semicontinuous function bounded below which does not attain
its infimum.
(As usual it is informative to see how the proof of the true result
breaks down when we loosen the hypotheses.)

37

K27
(i) f is continuous on [0, a] so is bounded and attains its bounds on
[0, a]. By periodicity f is everywhere bounded and attains its bounds.
(ii) Let f (x) = (x [x])1 ) if x is not an integer and f (x) = 0 if x is
an integer. (Here [x] is the integer part of x.)
(iii) By considering g(x) kx we see that we may suppose k = 0.
Given > 0 there exists an X > 0 such that |g(x + 1) g(x)| < (and
so |g(x + n) g(x)| < n for x > X. By hypothesis there exists a K
such that |g(y)| K for 0 y X + 1. Write n(x) for the integer
such that X < x n(x) X + 1. If x > X + 1 we have

g(x) g(x) g(x n(x)) g(x n(x))

x
x

n(x) K
+

x
x
as x . Since was arbitrary, g(x)/x 0 as x .

(iv) Any example for (ii) will work here.

(v) False. Consider, for example, h(x) = kx + x1/2 sin(x/2).

38

K28
No comments.

39

K29
Observe that xn can lie in at most two of the three intervals
[an1 , an1 + kn1 ], [an1 + kn1 , an1 + 2kn1 ]and [an1 + 2kn1 , bn1 ]
where kn1 = (bn1 an1 )/3.

40

K30
(i) Lots of different ways. Observe that the map J given by (x, y) 7
x y is continuous since
|(x + h) (y + k) x y| |x k| + |h y| + |h k|

khkkxk + kykkkk + khkkkk 0

as k(h, k)k 0. Since is continuous, the map A given by x 7 (x, x)

is continuous so, composing the two maps A and J, we see that the
given map is continuous.
Last part, continuous real valued function on a closed bounded set
attains a maximum.
(ii) Observe that, if h e = 0, then ke + hk2 = (1 + 2 ) so by (i)
and so
for all so

(1 + 2 )e (e) (e + h) ((e + h))

(e (h) + h (e + (e (e) h (h)) 0

e (h) + h (e) = 0.

(iv) If we assume the result true for Rn1 , then, since U has dimension n 1 and |U is self adjoint, U has an orthonormal basis e2 , e3 ,
. . . , en of eigenvectors for |U . Taking e1 = e, gives the desired result.

41

K31
(i) Observe that
|g(u) g(v)| kuvk

since, as simple consequence of the triangle inequality,

|kak kbk| ka bk.
Choose any z E. Set R = kz yk. The continuous function g
R) E at x0 , say.
attains a minimum on the closed bounded set B(y,
R) and g(x) g(x0 ) automatically, or
If x E then either x B(y,

x
/ B(y, R) and
g(x) > R = g(z) g(x0 ).

(ii) Let u = x0 y. Use the definition of x0 to show that

u u (u + h) (u + h)

whenever h E. By considering what happens when is small, deduce

that u h = 0.
If kx1 yk = kx0 yk and x1 E, then, setting h = x1 x0 , we
get h h = 0.
(iii) Since 6= 0 we can find a y
/ E. Set u = y x0 and b =
1
(kuk) u. Observe that
(x (b x)b) b = 0

so, since E has dimension n 1, we have

and
so
Set a = (b)b.

x (b x)b E

(x (b x)b) = 0
x = b xb.

42

K32
(i) Argument as in K31.
(ii) Suppose x E and x y > 0. Then, if is small and positive,
we have
x = (1 )0 + x E

but the same kind of argument as in K30 and K31 shows that kyk >
ky xk.
(iii) Translation.

43

K33
Write B for the closed unit ball. If kzk < 1 then
z = (1 kzk)z + kzk0

so z is not an extreme point.

On the other hand, if kxk = 1 and

z = x + (1 )y

with 0 < < 1 and x, y B, then, by the triangle inequality,

kzk kxk + k(1 )yk

with equality only if x, (1 )y and z are collinear. But

kxk + k(1 )yk + (1 ) = 1

and kzk = 1. Thus kxk = kyk = 1 and x, (1)y and z are collinear.
Inspection now shows that x = y = z.
A simpler pair of arguments gives the extreme points of the cube.
(ii) Since K is closed and bounded the continuous function x 7 x a
attains a maximum on K.
Suppose u0 is an extreme point of K 0 . If v, w K, 1 > > 0 and
then

x0 + u0 = v + (1 )w

x0 a = (x0 + u0 ) a

= v a + (1 )w) a

x0 a + (1 )x0 a
= x0 a.

The inequality in the last set of equations must thus be replaced by

equality and so v, w K 0 . Since u0 is an extreme point of K 0 , v =
w = u0 .
(iv) Use a similar argument to (ii) (or (ii) itself) to show that extreme
points of
{x K : T (x) = T (x0 )},

are extreme points of K.

44

K34*
No comments.

45

K35*
No comments

46

K36
(i) If neither K1 nor K2 have the finite intersection property, we can
find K1 , K2 , . . . KN , KN +1 , . . . , KN +M K such that
N
\

j=1

and so

Kj [a, c] = and
N\
+M
j=1

M
\

j=N +1

Kj [c, b] =

Kj [a, b] = .

(ii) We find [an , bn ] such that

Ln = {K [an , bn ] : K K}

has the finite intersection property

a = a 0 a1 a2 a n bn b 2 b1 b0 = b

and bn an = 2n (b a). We have an , bn for some [a, b].

We claim that K for each K K. For, if not, we can find K0 K
such that
/ K0 . Since K0 is closed, we can find a > 0 such that
(, +)K0 = . If N is sufficiently large, [aN , bN ] (, +)
so [aN , bN ] K0 = contradicting the finite intersection property of
LN .
T
Thus, by reductio ad absurdum, KK K.

47

K37*
No comments

48

K38
Observe that, if [aj , bj ] K, then
N
\

j=1

[aj , bj ] = [ max aj , min bj ] K

1jN

1jN

so, in particular, K has the finite intersection property.

T
If c KK K then c [a, b], that is to say a c b whenever
a E and b e for all e E. Thus c is a (and thus the) greatest
lower bound for E.

49

K39
(i) f10 (t) = 1 cos t 0 for all t. Thus f1 is everywhere increasing.
But f1 (0) = 0 so f1 (t) 0 for t 0 so t sin t for t 0.
(ii) f20 (t) = f1 (t) 0 for t 0 and f2 (0) = 0.

(iv) We have
2N
X
(1)j t2j+1
j=0

(2j + 1)!

sin t

2N
+1
X

(1)j t2j+1
(2j + 1)!

2N
+1
X

(1)j t2j+1
(2j + 1)!

j=0

for t 0 and, using the fact that sin(t) = sin t,

2N
X
(1)j t2j+1
j=0

for t 0.

(2j + 1)!

sin t

j=0

(v) Thus

as N .

j 2j+1
X
(1) t
|t|N +1

0
sin t

(2j + 1)! (2N + 3)!

j=0

50

K40
(i) Observe that g 0 is increasing since g 00 is positive. If g 0 (x1 ) > 0
then g 0 (t) g 0 (x1 ) > 0 for all t [x1 , x2 ] so 0 = g(x2 ) > g(x1 ) = 0
which is absurd. Thus g 0 (x1 ) 0 and g 0 (x2 ) 0. By the intermediate
value theorem there exist a c [x1 , x2 ] such that g 0 (c) = 0. Since g 0
is increasing g 0 (t) 0 and g is decreasing on [x1 , c] whilst g 0 (t) 0
and g is increasing on [c, x2 ]. Thus g(t) g(x1 ) = 0 on [x1 , c] and
g(t) g(x2 ) = 0. We use a similar argument to get the result on
[c, x2 ].
(ii) Look at g(t) = f (t) A Bt with A and B chosen to make the
hypotheses of (i) hold.
(iii) We may assume that 1 > n+1 > 0. The key algebraic manipulation in a proof by induction runs as follows.
n+1
!

!
n
X
X
j
Pn
j xj = f n+1 xn+1 + (1 n+1 )
f
xj

k
k=1
j=1
j=1
!
n
X
j
Pn
xj
n+1 f (xn+1 ) + (1 n+1 )f
k=1 k
j=1
n+1 f (xn+1 ) + (1 n+1 )

n+1
X

j f (xj )

n
X
j=1

Pn j

k=1

j=1

since
n
X
j=1

Pn j

k=1

= 1 and

n
X
k=1

k = (1 n+1 ).

(iv) Apply Jensen as suggested and take exponentials.

f (xj )

51

K41
The area of the inscribed quadrilateral is
=

4
X

a2 sin j cos j =

1
2

4
X

a2 sin 2j .

j=1

j=1

with a the radius of the circle. Also

P4

j=1 j

= .

Now sin00 t = sin t 0 for t [0, ], so sin is convex on [0, ] and

Jensens inequality gives
4
!
4
X
X
1
1
= 2a2
sin 2j 2a2 sin
(2j ) = 2a2 sin(/2) = 2a2 .
4
4
j=1

j=1

The area is attained when the j are all equal (and with a little thought,
observing that sin00 t < 0 on (0, )) only then.
Consider a circumscribing n-gon A1 A2 . . . An . If Aj Aj+1 touches the
circle at Xj let P
2j1 be the angle Aj1 OXj and 2j be the angle
Xj OAj . Then r=1 2nr = 2 and the area of the circumscribed
polygon is
2n
!
2n
X
X
= 21 a2
tan r 21 a2 2n tan
r /2n = na2 tan(/n)
r=1

r=1

(since tan is convex on (0, /2)). Equality is attained for a regular

polygon.

52

K42
Since g is continuous on [0, 1], it attains a maximum M say. Let
E = {x [0, 1] : f (x) = M }. Since E is non-empty it has a supremum
. Since f is continuous, g() = M . If = 0 or = 1 then M = 0. If
not then we can find a k with k, + k (0, 1) and
M = g() = 12 (g( k) + g( + k)) M

with equality if and only if g( k) = g( + k) = M . Thus + k E

contradicting the definition of . We have shown that M = 0 so g(t)
0 for all t. Similarly g(t) 0 for all t so g = 0.
Suppose that we drop the end conditions. By replacing g(t) by
G(t) = g(t) A Bt with G(0) = G(1) = 0, we can show that g
is linear.
[In higher dimensions the condition becomes the average of g over
the surface of any sufficiently small sphere equals the value at the
centre and this turns out to be equivalent to saying that g satisfies
Laplaces equation O2 g = 0. Observe that in one dimension this reduces to saying that g 00 = 0 so g is linear as we saw above. However in
higher dimensions the solutions of Laplaces equation are much more
diverse.]

53

K43
Observe that

f (h) f (0)
= |h sin h4 | |h| 0

h
as h 0.

f 0 (x) = 2x sin x4 4x2 cos x4 . Thus we have

If x 6= 0, then
f ((2n + 1))1/4 ) as n .

54

K44
Since g is continuous on [a, b], it attains a maximum at some c [a, b].
Since g 0 (c) = 0, we have c (a, b) and f 0 (c) = k.
For the final paragraph, note that f 00 (a) 0 f 00 (b).

55

K45
(i) The statement f must jump by at least 1 at z is not well
defined. If the reader feels that there should be some way of making this statement well defined in a useful way she should consider
f (x) = 21 sin(1/x) for x 6= 0, f (0) = 0.
However f can not be continuous at z. Observe that
max(|f (z) f (xn )|,|f (z) f (yn |)

12 (|f (z) f (xn )| + |f (z) f (yn |) 12 .

Thus we can find zn z with |f (z) f (Zn )| 1/2.

(ii) The behaviour of f 0 (zn(j) ) does not control the value of f 0 (z).
Proposed result is false, see K43.

56

K46*
No comments.

57

K47
Since g 0 is constant, the conclusions of intermediate value theorem
hold trivially for g 0 .

58

K48
Observe that
cos n = <(cos + i sin )n
X n
=
(1)r cosn2r sin2r
2r
02rn
X n
(1)r cosn2r (1 cos2 )r
=
2r
02rn

a real polynomial in cos of degree at most n. (Part (ii) shows that

the degree is exactly n.)
T0 (t) = 1, T1 (t) = t, T2 (t) = 2t2 1, T3 (t) = 4t3 3t.
(a) Observe that
cos n cos = (cos(n + 1) + cos(n 1))/2.

Thus tTn (t) = (Tn+1 (t) + Tn1 (t))/2 for all t [1, 1].

(b) But a polynomial of degree at most n + 1 which vanishes at n + 2

points is identically zero, so tTn (t) = (Tn+1 (t) + Tn1 (t))/2 for all t.
(c) Induction using (i).
(d) |Tn (cos )| = | cos n| 1.
(e) Tn vanishes at cos((r + 1/2)/n) with 0 r n 1.
For the last part make the change of variables x = cos with range
0 .
If m = 0 replace /2 by .

59

K49
Two polynomials with the same roots and the same leading coefficient are equal.
Use the fact that |Tn (t)| 1 for T [1, 1].
Choose and so that + = 1, + = 1. If P is the
interpolating polynomial of degree n for f on [a, b], then taking g and
Q such that
g(t) = f (t + ), Q(t) = P (t + ),
we see that Q is the interpolating polynomial of degree n for g on
[1, 1]. If |f n (x)| A on [a, b] then |g n (x)| An = A((b a)/2)n on
[1, 1] so
(b a)n A
|f (t) P (t)|
22n1
on [a, b].

60

K50
(ii) P (t) =

n
X
f j (a)
j=0

j!

tj .

(vii) Observe that

f (t)
f (n+1) (t )
= (n+1)
g(t)
g
(t )
with t , t a as t a and use continuity.

61

K51
If we set
P (t) =

n
X
j=0

Aj tj (1 t)n+1 +

n
X
j=0

Bj tn+1 (1 t)j ,

then the resulting system of equations for Aj is triangular with nonvanishing diagonal entries and thus soluble. The same holds for Bj .
If y (0, 1), set

xn+1 (1 x)n+1
.
y n+1 (1 y)n+1
F has vanishing r-th derivative at 0 and 1 for 0 r n and vanishes at
y. By Rolles theorem F 0 vanishes at least twice in (0, 1), F 00 vanishes
at least three times in (0, 1), F 000 vanishes at least four times in (0, 1),
. . . , F (n+1) vanishes at least n + 2 times in (0, 1), F (n+2) vanishes at
least n + 1 times in (0, 1), F (n+3) vanishes at least n times in (0, 1),
. . . , F 2n+2 vanishes at least once (at say) and this gives the result.
F (x) = f (x) P (x) E(y)

62

K52
(i) g(b) g(a) = g 0 (c)(b a) 6= 0 for some c (a, b).
(ii) A = (f (b) f (a))/(g(b) g(a)).
(iv) Observe that

with ct a as t a.

f 0 (ct )
f (t) f (a)
= 0
g(t) g(a)
g (ct )

63

K53
(i) We have
f (b) = f (a) + f 0 (a)h + f 00 (c)

h2
2!

f 0 (a)h = f (b) f (a) f 00 (c)

h2
2!

for some c (a, b).

(ii) Thus

and so
|f 0 (a)||h| 2M0 + 21 M2 |h|2

for all a and all h. Thus

2M0 M2
+
h
2h
for all h > 0 and, choosing h = 2(M0 M1 )1/2 , we have the required
result.
|f 0 (t)|

(iii) Take f (t) = t to see that (a) and (b) are false.
(c) is true by the method of (ii), provided that L 2(M0 M1 )1/2 .
(d) True. g(t) = sin t.
1/2

(e) True. Scaling (d), G(t) = M0 sin(M0

1/2

M2 t) will do.

(f) Fails scaling. If G as in (e) then

G0 (0)
= (M0 M1 )1/4
1/4
(M0 M2 )
can be made as large as we wish.

64

K54
(i) True. Choose a > 0 such that |f (t) f (s)| < 1 whenever
|t s| < 2 and an integer N > 2 + 1 . If x (0, 1) we can find
M N and x1 , x2 , . . . , xM (0, 1) such that x1 = x, xM = 1/2,
|xj+1 xj | for 1 j M 1. Since |f (xj+1 ) f (xj )| 1 we have
|f (x) f (1/2)| N and |f (x)| N + |f (1/2)|.
(ii) False. If f (t) = t then f does not attain its bounds.
(iii) False. Set f (t) = sin(1/t).

65

K55
(i) True. By the intermediate value theorem, we see that f (x) l
for some l as |x| . Given > 0, we can find a K such that
|f (x) l| < /2 for all |x| K. Since a continuous function on
a closed bounded set is uniformly continuous, we can find a with
1 > > 0 such that |f (x) f (y)| < for all x, y [K 2, K + 2]
with |x y| < . By going through cases, we see that |f (x) f (y)| <
for all x, y R with |x y| < .
(ii) False. Take f (t) = sin t.
(iii) False. Take f (t) = sin t2 .
(iv) False. Take f (t) = t.
(v) False. Take f (z) = exp(i|z|).
(vi) True. Observe that ||f (z)| |f (w)|| |f (z) f (w)|.
(vii) True. Let > 0. Choose > 0 such that |x y| < implies
||f (x)| |f (y)|| < /2.
Now suppose y > x and |x y| < . If f (x) and f (y) have the same
sign, then automatically, |f (x)f (y)| < /2. If they have opposite sign,
then by the intermediate value theorem, we can find u with y > u > x
and f (u) = 0. Since |x u|, |y u| < we have
|f (y) f (x)| |f (y) f (u)| + |f (u) f (x)| < /2 + /2 = .

(vii) False. Take f (x) = g(x) = x. If > 0 then, taking y = 8 1

and x = y + /2, we have |x y| < but |x2 y 2 | = (x + y)(x y) > 1
so x 7 x2 is not uniformly continuous.
(viii) True. Let > 0. We can find a > 0 such that |u w| <
implies |f (u) f (w)| < and an > 0 such that |x y| < implies
|g(x) g(y)| < . Thus |x y| < implies |g(f (x)) g(f (y))| < .

66

K56
Observe that
so that
Thus

k(x + u) yk kx yk + kuk
f (x + u) f (x) + kuk.
kf (a) f (b)k ka bk

and f is uniformly continuous.

If E is closed, then we can find yn E such that kx yn k

2) so we
f (x) + 1/n. The yn belong to the closed bounded set E B(x,
can find n(j) and a y with kyn(j) yk 0. Since E is closed,
y E. We observe that
f (x) + 1/n(j) kx yn k

kx yk ky yn k

kx yk f (x)

as j . Thus kx yk = f (x).

Conversely, if the condition holds, suppose yn E and kyn xk 0.

We have f (x) = 0 so there exists a y E with kx yk = f (x) = 0.
We have x = y E so E is closed.

67

K57
(i) Given > 0 we can find a > 0 such that, if u, v Q, then
|u v| < implies |f (u) f (v)| < . We can now find an N such
that |xn x| < /2 for n N . If n, m N then |xn xm | < so
|f (xn )f (xm )| < . Thus the sequence f (xn ) is Cauchy and converges.
(ii) Using the notation of (i), we can find an M such that |xm
x| < /2 and |ym x| < /2 for m M . Thus |xm ym | < and
|f (xm ) f (ym )| < for m M . Since was arbitrary, f (xm ) and
f (ym ) tend to the same limit.
(iii) Observe that (ultimately by the Axiom of Archimedes) any x
R is the limit of xn Q.
(iv) Take xn = x.
(v) We use the notation of (i). If |xy| < /3 we can find xn , yn Q
with xn x, yn y and |xn x| < /3, |yn y| < /3. Then
|xn yn | < for all n so |f (xn ) f (yn )| < and |F (x) F (y)| .

68

K58
PN
P
n
n
(i) If |z| < min(R, S) then N
n=0 bn z tends to limit
n=0 an z and
and so
N
N
N
X
X
X
bn z n
an z n +
(an + bn )z n =
n=0

n=0

n=0

tends to limit as N .

P
n
R < S If R
< |z| < S then
n=0 an z converges and
PSuppose
P

n
n
n=0 bn z diverges so
n=0 (an + bn )z diverges. Thus the sum has
radius of convergence R.
(ii) The argument of the first paragraph of (i) still applies.

(iii) We deal with the case > T R > 0. Let an = Rn + T n

and bn = Rn .
(iv) The radius of convergence is R unless = 0 in which case it is
infinite.
2

If |z| > 1 then, provided n is sufficiently large, |z|n (2R)n so

P(v)

n
If |z| < 1 then, provided n is sufficiently large,
n=0 an z diverges.
P
2
n
n
|z| (R/2) so n=0 an z n converges. The radius of convergence is
1.
If R = 0, then any radius of convergence with 0 1 is possible.
2
n
For = 0 take an = nn . For 0 < < 1 take an = n . For = 1
2
take an = n
where n tends to 1 very slowly from below.
n
Similar ideas for R = .

Pn
Pn
n
n
(vi) Let |cn | = max(|an |, |bn |). Since
j=0 |an z |
j=0 |cn z |
radius of convergence R0 at most R. Thus R0 min(R, S). But |cn |
|an | + |bn | so R0 min(R, S). Thus R0 = min(R, S).
If |cn | = min(|an |, |bn |) then similar arguments show that the radius
of convergence R00 max(R, S). But every value A max(R, S) is
possible for R00 . (If A < and R, S > 0 can take eg. a2n = R2n ,
b2n = A2n , a2n+1 = A2n1 , b2n = S 2n1 .

69

K59
Let > 0. We can find an N such that |h(n) | < for n N .
Observe that if n N then h(n) < + and so

an > 2 (+)+(1(+) n = 212(+)n .

Thus if |z| > 22(+)1 we have |an z n | . Thus the radius of convergence R satisfies R 22(+)1 . Since was arbitrary R 221 .
A similar argument shows that, if |z| < 22()1 |an z n | 0, and thus
R 221 so R = 221 .
Suppose 1 0. Define E inductively by taking an+1 = an /2 if
an /2 212 , an+1 = 2an otherwise. Then h(n) and |an z n | 1
whenever |z| = 221 and n is large. Thus we have divergence everywhere on the circle of convergence.
Suppose 1 0. Define E inductively by taking an+1 = 2an if
2an 2(12)n n2 , an+1 = an /2 otherwise. Then |an z n | n2 whenever |z| = 221 and n is large. Thus we have convergence everywhere
on the circle of convergence. When n is large we have
so
whence

an 4 2(12)n n2
2n2h(n)n 22+(12)n n2
(log 2)(n 2h(n)n) (log 2)(2 + (1 2)n 2 log n

and dividing by n log 2 and allowing n we see that

1 2 lim sup h(n) = 1 2
n

so lim supn h(n) = . A similar argument (but there are alternative

ways of proceeding) shows that lim inf n h(n) = so h(n) .

70

K60
If |an |1/n is unbounded, the radius of convergence is zero.

We deal with the case (lim supn |an |1/n )1 = R with 0 < R < .
The other cases are similar.
Suppose R > > 0. Then we can find an N such that |an |1/n
(R /2)1 for all n N . Thus if |z| < R we have
n

R
n
|an z | <
R /2
P
for n N and n=0 |an z n | converges by comparison with a convergent
geometric sum.
We can also find n(j) such that |an |1/n(j) (R + /2)1 . Thus,
if |z| > R + we have

n(j)
R+
n(j)
|an z | >
.
R + /2
Since was arbitrary, R is the radius of convergence as required.

71

K61
Write = lim supn an+1 /an . If > 0, we can find an N with
an+1 /an + for n N . Thus, by induction, an A( + )n for
n N where A = aN ( + )N . Thus, if n N ,
a1/n
A1/n ( + ) +
n

as n . Since was arbitrary

lim sup a1/n

lim sup an+1 /an .
n
n

Remaining inequalities similar or simpler.

All 8 possibilities can arise. Here are d.
(1) Set an = 1.
an+1
an+1
= lim inf a1/n
= lim sup a1/n
= lim sup
= 1.
lim inf
n
n
n
n
an
an
n
n
(2) Set a1 = 1, a2n +1 /a2n = 2, ar+1 /ar = 1 otherwise.
an+1
an+1
= lim inf a1/n
= lim sup a1/n
= 1 < 2 = lim sup
.
lim inf
n
n
n
n
an
an
n
n
(3) Set a1 = 1, a2n +1 /a2n = 2, a2n +2 /a2n +1 = 21 for n 2 ar+1 /ar =
1 otherwise.
an+1
an+1
lim inf
= 1/2 < lim inf a1/n
= lim sup a1/n
= 1 < 2 = lim sup
.
n
n
n
n
an
an
n
n
2n

2n+1

(4) Set a1 = 1, ar+1 /ar = 2 if 22 r < 22

, ar+1 /ar = 1
otherwise
an+1
an+1
= lim inf a1/n
= 1 < 2 = lim sup a1/n
= lim sup
.
lim inf
n
n
n
n
an
an
n
n
2n

2n+1

2n

2n+1

(5) Set a1 = 1, ar+1 /ar = 2 if 22 + 1 r < 22

, ar+1 /ar = 4 if
22n
r = 2 ar+1 /ar = 1 otherwise
an+1
an+1
= lim inf a1/n
= 1 < 2 = lim sup a1/n
< 4 = lim sup
.
lim inf
n
n
n
n
an
an
n
n
, ar+1 /ar = 4 if
(6) Set a1 = 1, ar+1 /ar = 2 if 22 + 2 r < 22
22n
22n
r = 2 , ar+1 /ar = 1/2 if r = 2 + 1 ar+1 /ar = 1 otherwise
an+1
an+1
< 4 = lim sup
lim inf
= 1 < 2 = lim sup a1/n
= 21 < lim inf a1/n
.
n
n
n
n
an
an
n
n
Other two obtained similarly.
Note lim sup formula will always give radius of convergence. Ratio
may not.

72

K62
(iii) Could write bn = cn + dn with dn = 0 for n large and |bn | .
Thus lim supn |Cn | .
(v) If bj = (1)j limit is 0.

(viii) Pick N (j), Mj , sj and rj inductively so that N (0) = 0, M (0) =

M (j)+1
1, s0 = 1/2. we pick rj+1 so that 0 < sj < rj+1 < 1 and 1 rj+1
<
N (j+1)
j
j
< 2 . Pick sj+1 so that
2 . Pick an N (j+1) > M (j) such that rj+1
N (j+1)
0 < rj+1 < sj+1 < 1 and sj+1
> 1 2j1 . Pick M (j + 1) > N (j + 1)
M (j+1)
so that sj+1
< 2j1 Set bn = 1 if N (j) n M (j), bn = 0
otherwise.
M (j)

Ar(j+1)

X
n=0

n
(1 rj+1 )rj+1
+

= (1

and

M (j)+1
rj+1 )

N (j)
rj+1

n=N (j+1)

n
(1 rj+1 )rj+1

< 2j + 2j = 2j+1

M (j+1)

As(j+1)

n=N (j+1)
N (j+1)

= sj+1

(1 sj+1 )snj+1
M (j+1)+1

sj+1

Thus Ar does not converge.

> 1 2j1 2j1 = 1 2j .

73

K63
(ii) C can be identified with R2 .
P
n
N +1
(iii) (a) If z 6= 1, N
)/(1 z) tends to a limit if
n=0 z = (1 z
and only if |z| < 1. The limit is 1/(1 z). If z = 1 we have divergence.
(b) Observe that, if z 6= 1,
Cn = (n + 1)

= (n + 1)1

k
n X
X

zj

k=0 j=0
n
X
k=0

1 z k+1
1z

1 z n+1
1
1
(n + 1) + z
=
n+1
1z
1z
z 1 z n+1
1
+
.
=
1z n+1 1z
Thus we have convergence if and only if |z| 1, z 6= 1 and the limit is
1/(1 z).

(c) Ar = (1 rz)1 . Abel sum exists if and only if |z| 1, z 6= 1

and is then 1/(1 z).

74

K64
(v) If k then, using part (i), Bk b. Since this is true for
every such sequence, B b as .
(vi) We seek to Borel sum z j . Observe that, if z 6= 1

X
1
n 1 z n+1
e
=
(1 ze(1z) ).
n!
1

z
1

z
n=0

We have convergence if and only if <(1 z) < 1 and the Borel sum is
then 1/(1 z).
If z = 1,

X
X
n X n
n
B =
(n + 1) e =
e +
e = 1 .
n!
n!
n!
n=0
n=0
n=0

75

K65
(i) True. Since the sequence is bounded, there exists a b and a strictly
increasing sequence n(j) such that bn(j) b. Set ujn(j) = 1, ujk = 0
otherwise.
(ii) True. We can extract subsequences with different limits and use
(i).
(iii) False. Let bj = 1 for 2n n j 2n 1, bj = 1 for
2 + 1 j 2n + n [n 4], bj = 0 otherwise Observe that, if unk = 1
n
for 2P
N k 2n , unk = 1 for 2n + 1 k 2n + N , then U G
but
k=0 ujk bk = 2N for j sufficiently large. But N is arbitrary.
n

(iv) True. Define N (r) and j(r) inductivelyP

as follows. Set N (0) =
r
j(0) = 1. Choose N (r + 1) > N (r) such that
k=N (r+1) |uj(r)k | < 2
PN (r+1)
and j(r + 1) > j(r) such that k=0 |uj(r+1)k | < 2r1 . Set bk =
(1)r+1 for N (r) k < N (r + 1). Then

X
(1)r+1
uj(r)k bk
k=0

= (1)r+1

as r .

X
k=0

X
k=0

N (r)

N (r+1)

uj(r)k bk +

k=0

uj(r)k 2

k=N (r)+1

N (r)

X
k=0

|uj(r)k | 2

uj(r)k 2r+3 1.

(v) False. Try bj = 0.

uj(r)k bk +

k=N (r+1)+1

k=N (r+1)+1

|uj(r)k |

uj(r)k bk

76

K66
(i) Try bk = 1 for k = N , bk = 0 otherwise. Try bk = 1 for all k.

77

K67
(i) implies (ii) since absolute convergence implies convergence.
For the converse observe that, writing xn = (xn1 , xn2 , . . . , xnm ), we
have
kxj k |xj1 | + |xj2 | + + |xjm |.
P
Thus, if j=1 kxj k diverges, we must be able to find a k with 1 k m
P
such that
j=1 |xjk | diverges. Choose j so that j xjk 0. Then
N
N

N
X
X
X

j x j
j xjk =
|xjk | .

j=1

as N .

j=1

j=1

78

K68
If

n=1

an converges, then

X
X
X
an
an
an
nS1 ,nN

nN

n=1

and, since an increasing sequence bounded above converges,

tends to a limit as N .
P
If nSj ,nN an tends to a limit so does
X
X
X
an =
an +
an .
nN

nS2 ,nN

nS1 ,nN

an

nS2 ,nN

Second paragraph does not depend on positivity aj so if remains

true. However, taking a2n1 = a2n = 1/n, S1 evens and S2 odds then
we see only if false.
1/2

1/2

Let S1 = {n : an /n an } and S2 = {n : an /n > an }.

P
P
1/2
nS1 an /n converges by comparison with
n=1 an . If n S2 then
so, since

n=1

n2

2
1 2
2
an = (a1/2
n ) < (n ) = n
P
1/2
converges, so does nS2 an /n. Hence the result.

By Cauchy Schwarz,
!2
N

N
X
X
X
X
X a1/2
1
1
n

an
an
2
n
n
n2
n=1
n=1
n=1
n=1
n=1

so, since an increasing sequence bounded above converges, we are done.

Observe that cn (an + bn )/2 and use comparison.

79

K69
(i) Diverges
2N
X
1 + (1)n

n=1

as N .

N
X
1
=

n
n=1

2
(ii) Converges by comparison since p2
n n .

(iii) Diverges by comparison. We have n1/n 1 (e.g. by taking

logarithms) so n1/n 1/2 for n large and so n11/n > 21 n1 for n
large.
PN (j)
P
(iv) If
n=1 an
n=1 an converges, then taking N (j) = j we have that
tends to a limit.
PN (j)
If n=1
an tends to a limit A, then for any M we can find a j with
N (j) > M so that
M
X
n=1

N (j)

an

X
n=1

an A

so, since an increasing sequence bounded above converges, we are done.

R
(v) If 1 f (t) dt converges then
Z N +1
Z
N Z n+1
N
X
X
f (t) dt =
f (t) dt
f (t) dt
an
n=1

n=1

P
since an increasing sequence bounded above converges
n=1 an converges.
P
For the converse observe that if
n=1 an converges then, if X N ,
Z X
Z N

N
1
N
1 Z n+1
X
X
X
an .
an
f (t) dt K
f (t) dt
f (t) dt =
1

n=1

n=1

n=1

Now use something like Lemma 9.4.

(vi) an = (1)n .
P N
(vii) Let M
a tend to b. Given > 0 we can find an N0 such
PM N n=1 n
that k n=1 an bk /2 for N N0 . We can also find an N1 such
that kan k /2M for n N1 . Let N2 = (M + 1)(N1 + N0 ). If
m N2 we can write m = N M + u with N N0 , 0 u M 1 and
M N + r N1 for all r 0. We then have
k

m
X
n=1

an bk k

MN
X
n=1

an bk +

MX
N +u

n=M N +1

kan k <

u
+

2 2M

80

as desired.
(viii) Set N (j) = 22j , an = j 1 if 22j n < 2 22j , an = j 1 if
2 22j n < 3 22j , an = 0 otherwise.

81

K70
(i) Observe that (an bn )1/2 (an + bn )/2 and use comparison.
(ii) Take bn = an+1 in (i).
(iii) Observe that an+1 (an an+1 )1/2 and use comparison.
(iv) a2n = 1, a2n+1 = n4 .

82

K71
(i) True. a4n 0 so there exists an N with |an | 1 for n N . Thus
|a5N | for n N and the result follows by comparison and the fact
that absolute converge implies convergence.
a4n

(ii) False. Take an = (1)n1/6 and use the alternating series test.
(iii) False. Take a2k = 2k , ar = 0 otherwise.
(iv) False. Same counterexample as (iii).
(v) True. Given > 0 we can find an N such that
for M N . Thus, if n 2N + 2
nan 2(n N 1)an 2

M
X

PM

j=N

aj < /2

aj < .

j=N

(vi) False. Take an = (n log n)1 for n 3 and use the integral
comparison test.
(vii) True. Abels test.
(viii) True. By CauchySchwarz,
N
!2
N
N

X
X
X
X
X
3/4
2
3/2
2
|an |n

|an |
n

|an |
n3/2 .
n=1

n=1

n=1

n=1

n=1

(ix) False. Take an = (1)n (log n)1 for n 3. Use alternating

series test and integral test.
(x) True. We must have an 0 so there exists a K such that
|an | K. Since |n5/4 an | Kn5/4 the comparison test tells us that
P

5/4
an is absolutely convergent and so convergent.
n=1 n

83

K72
akn 0 so there exists an N with an 1 for n N . Thus akn |ak+1
n |
for n N and the result follows by comparison.
Set f (n) = log(n + 1)n and observe that, if k 2
3N
X
j=1

akj

N
X
j=1

(2k 2)(log(n + 1))k ) .

(Use comparison and the fact that n (log n)k for n sufficiently large.)

84

K73
(i) Write
Sn =

n
X

1
j bj .

j=1

Then, by partial summation (or use Exercise 5.17),

n
n
X
X
j (Sj Sj1 )
bj =
j=1

j=1

n1
X
j=1

(j j+1 )Sj + n Sn

so that
n
n1
X X

bj
(j+1 j )kSj k + n kSn k

j=1

j=1

n1
X
j=1

(j+1 j )K + n K = (2n 1 )K 2kn

whence the result.

(ii) If > 0 then we can find an N such that k

n N . Thus, by (i),
1
n k

n
X
j=1

bj k

1
n

1
n

N
1
X
j=1

N
1
X
j=1

kbj k +

Pn

1
n k

j=N

n
X

j=N

1
j bj k for

bj k

kbj k2 2

as n . Kroneckers lemma follows.

(iii) ((a) fails) Let bj = j 4 , j = j 6 for j 6= 2r , b2r = 1 and 2r = 22r .

((b) fails) Let bj = j 2 , j = 1.

((c) fails) Let bj = j 4 , j = j 2 .

85

K74
Consider the set An of integers between 10n and 10n+1 1 inclusive
whos decimal
contain the integer 9. We observe
Sn expansion does not n+1
that {0} j=0 Aj contains (9/10)
10n+1 = 9n+1 elements. Thus
An contains at most 9n+1 elements and
X
j 1 10n 9n+1 .
jAn

The partial sums of the given series thus can not exceed
9n+1 = 90 and the series converges.

n=0

10n

86

K75
(i) Observe that 1/x 1/(n + 1) for n x n + 1 and integrate.
Tn Tn+1

=
n+1

n+1
n

1
dx.
x

Observe that T1 = 1 and

Z r+1
n1
X
1
1
Tn

dx .
r
x
r
r=1
Decreasing sequence bounded below tends to a limit.
(v) Given l we can choose integers pn , qn 1 such that pn+1 > 2pn ,
qn+1 > 2qn , pn /qn is strictly decreasing and log 2+(1/2) log(pn /qn ) l.
At the n-th stage, if the ratio of positive terms to negative exceeds
pn /qn use pn positive terms followed by qn + 1 terms if the ratio is less
than pn /qn , use pn + 1 positive terms followed by qn negative terms,
otherwise usepn positive terms followed by qn negative terms. Switch
from the n-th stage to the n + 1-th stage when the size of each unused
terms is less than (pn+1 + qn+1 + 1)1 2n and the ratio of positive terms
to negative differs from pn /qn by at most 2n .

87

K76
P
(ii) Observe that, if
j=0 bj converges, we can find an N () such
PN
that | j=M ) bj | for N M N () so

j
b
x

xM
j

j=M

for M N ().

(iii) If M N (), then

M 1

X
X
X X

bj xj
bj
|bj xj bj | +
bj +
bj xj

j=0

j=0

j=0

M
1
X
j=0

j=M

j=M

|bj xj bj | + 2 2

as x 1. Since is arbitrary the required result follows.

P
P
j
j
(iv) Observe that, if 0 < x < 1, then
j=0 bj x and
j=0 aj x ,
P
j
j=0 cj x are absolutely convergent so (using Exercise 5.38) we may
multiply them to get

X
X
X
j
j
aj x
c j xj .
bj x =
j=0

Now let x 1.

j=0

j=0

88

K77
P
P
n m
converges for |x| = |y| = , we have
(a) Since
m=0 cn,m x y
n=0
n+m
n m
|cn,m |
= |cn,m x y | 0 for n, m so there exists a K with
|cn,m | n+m < K for all n, m 0. Set = 1 .
P
P
P
(b) (i) nN, mN |xn y m | = nN |x|n mN |y|m .
E is the square (1, 1) (1, 1).
n+m n m
P
PP
r
|x y | = N
(ii)
r=0 (|x| + |y|)
n+mN
n

E is the square {(x, y) : |x| + |y| < 1}.

n+m 2n 2m
P
PP
2
2 r
|x y | = N
(iii)
r=0 (|x| + |y| ) .
n+mN
n

E is the circle {(x, y) : x2 + y 2 < 1}.

P
P
P
(iv) n, mN |xn y m /(n!m!)| = nN |x|n /n! mN |y|m /m!.
E = R2 .
P
P
P
(v) n mN |xn y m n!m!| = nN |x|n n! mN |y|m m!.

E = {0}.
P
P
(vi) nN |xn y n | = nN |xy|n .

E = {(x, y) : |xy| < 1} a set bounded by branches of hyperbola.

(c) Observe that |cnm xn y m | |cnm xn0 y0m | when |x| |x0 |, |y| |y0 |.

89

K78
(i) Choose Nj () so that |aj aj,n | < /M for n Nj () and set
N (, M ) = max1jM Nj ().
Observe that

X
j=1

aj

M
X
j=1

aj

M
X

aj,n

j=1

for all M .
P
(ii) If
j=1 aj converges with value A, then given > 0 we can find
P
an M such that M
j=1 aj > A . By (i) we can find N (, M ) such that

X
j=1

and so

X
j=1

aj,n

aj

X
j=1

M
X
j=1

aj

aj,n A 2

P
for all n N (, M ). Thus
j=1 aj,n A.
P
If
j=1 aj fails to converge, then given any K > 0, we can find
PM
an M such that
j=1 aj > 2K. But we can find an N such that
PM
PM
P
aj,n j=1 aj K for n N and so
j=1 P
j=1 aj,n K for n N .

Thus j=1 aj,n can not converge.

(iii) Yes we can drop the condition. Consider bj,n = aj,n aj,1 instead.

90

K79*
No comments

91

K80
It is possible to do this by fiddling with logarithms but P
I prefer to
nz
copy the proof P
of the radius of convergence. Show that if
n=0 bn e
converges then n=0 bn enw converges absolutely when <w < <z.
P
nz
Next observe that the sum
converges everywhere (set
n=0 bn e
X =P
) or nowhere (set X =P
) or we can find z1 and z2 such

nz1
nz2
that
b
e
converges
and
diverges. Explain why
n=0 n
n=0 bn e
this means that

X
bn enz converges}
X = sup{<z :
n=0

exists and has the desired properties.

For X = can take bn = 0, for X = can take bn = en , for

X = R with R finite can take bn = eRn .
P n nz
2
n nx
n=0 2 e /(n + 1) converges when z = log 2. Also 2 e /(n +
1)2 when x is real and x > log 2 so X = log 2.
P
inz
By the first paragraph there exists a Y 0 such that
n=0 cn e
converges for =z < YPand diverges for=z > Y . By taking complex coninz
converges for =z > Y and diverges
jugates we see that
n=0 cn e
for =z < Y .
P
P
P
Since
n=0 sn + tn diverges if exactly one
n=0 sn and
n=0 tn
Pof

inz
inz
converges
and
converges
if
both
converge,
c
(e
+
e
)
and
n=0 n
P
thus n=0 cn cos nz converges if =z < Y and diverges if =z > Y .

92

K81
(i) By integration by parts,
In+1 = [ sin
=n

x cos x]/2
0

/2
0

+n

/2

sinn1 x cos2 x dx
0

sinn1 x(1 sin2 x) dx

= nIn1 nIn+1 .
(ii) Observe that sinn1 x sinn x sinn+1 x for x [0, /2]. Integrating gives In+1 In In1 so
In+1
In
n
=

1.
n+1
In1
In1
so In /In1 1 as n .
(iii) I0 = /2, I1 = 1
(2n 1)(2n 3) . . . 1
2n(2n 2) . . . 2
I2n =
, andI2n+1 =
.
2n(2n 2) . . . 2 2
(2n + 1)(2n 1) . . . 1
Thus
n
Y

4k 2 2
I2n+1
=
1
2
4k 1
I2n
k=1
and the Wallis formula follows.

93

K82
(i) Power series or set f (x) = exp x x and observe f (0) = 0,
f 0 (x) 0 for all x 0.
(iii) Observe that

m
n
n
m

Y
Y
Y
Y

(1 + aj ) 1
(1 + aj ) = (1 + aj )
(1 + aj )

j=n+1
j=1
j=1
j=1
!
m
n
Y
Y
(1 + |aj |) 1 .
(1 + |aj |)

j=n+1

j=1

Now use the inequality (1 + |aj |) exp |aj |.

(iv) Use the general principle of convergence and (iv).

(v) We must have an 0. If |an | < 1/2 then

an
2|an |.
|bn | =
1 + an
P
Thus
n=1 bn converges absolutely.
n
n

Y
Y
Y
Y
1=
(1 + aj ) (1 + bj )
(1 + aj ) (1 + bj )
j=1

so

j=1 (1

+ aj )

j=1

j=1 (1

j=1

j=1

+ bj ) = 1.

(vi) Observe that

N
Y

(1 + aj /Rj ) =

j=1

as N and use (v).

N
Y
Rj1
j=1

Rj

R0
0
RN

94

K83
(i) Left to reader.
(ii) Let an = n , an,N = n if all the prime factors of N lie among
the first N primes, an,N = 0 otherwise. Observe that
0 an,N an,N +1 an ,
P
an as N and that
n=1 an converges.

that an,N
(iii) If

1
K
1 p1

1
K
1 p

pP, pN

for all N then

pP, pN

for all N and all < 1 so

pP

and

X
1
K
n
n=1

for all < 1. Thus

for all < 1 and all N so

for all

1
n=1 n

1
K
1 p

N
X
1
K

n
n=1
N
X
1
K
n
n=1

which is absurd.

(iv) Let E be a set of positive integers. Write

En = {r E : 2n r < 2n+1 }.

If En has M (n) elements and M (n) A2n then

X

rE,r2N

for all N .

N 1
N 1
N
1
X
1 X X 1 X j
A
=

2 M (j) A
2n(1) =
r
r
1 21
j=0 rE
j=0
j=0
j

95

K84
Q
Suppose that Pn > 1 for all n N . Then Pm = PN m
j=N (1 an ).
P
If j=1 an diverges then (see K82) Pm 0 which is impossible by the
P
definition of N . Thus
j=1 an converges and Pm tends to a limit. A
similar argument applies if Pn 1 for all n N .

Thus we need only consider the case Pn 1 changes sign (or is zero)
infinitely often. But an 0 and if (A) PN 1 1 0 PN 1 or
PN 1 1 0 PN 1 and (B) 0 an for n N 1 then we
have 1 + Pn 1 for n N . Thus Pn 1.
P
If
j=1 an diverges we can say nothing about l. (If l 1 consider
a1 = l 1, an = 0 for n 2. If l 1 consider a1 = 1, a2 = 1 l/2,
an = 0 for n 3.)

96

K85
Let an = z n , an,N = z n if 0 n 2N +1 1 and an,N = 0, otherwise.
Observe that 0 |an,N |an |, and that an,N an as N . By
dominated convergence (Lemma 5.25)
N
Y
j=1

2j

(1 + z ) =

X
n=0

an,N

X
n=0

an = (1 z)1 .

97

K86
Observe that
2 cot 2 =

cos2 sin2
= cot + tan
cos sin

and use induction.

Observe that

as n .

1 2n
1
1
cot
=
cos 2n
n
n
n
2
2
sin 2

98

K87
Very similar to K86. To obtain the formula for un , observe that
2 cos2 = 1 + cos 2 and cos(/4) = 21/2 .

99

K88
The square of a rational is rational.
(ii) Observe, by induction or Leibnizs rule, that f (k) is the sum of
powers of x with integral coefficients plus terms of the form
xr (1 x)s
M
(min(r, s))!
with M an integer and r, s 1. Thus f (k) (0) and f (k) (1) are always
integers. Since bn 2nr is an integer, G(0) + G(1) is always an integer.

as n .

0 < an f (x) sin x

an
0
n!

100

K89
(a)(i) Given c C we can find b B with g(b) = c and a A with
f (a) = b.
(ii) g f (a) = g f (a0 ) implies f (a) = f (a0 ) and so a = a0 .

(iii) f (a) = f (a0 ) implies g f (a) = g f (a0 ) and so a = a0 .

(iv) Given c C, we
can find a A with g f (a) = b. Observe that
f (a) B and g f (a) = c

(b) Let A = {a}, B = {a, b}, C = {a} with a 6= b f (a) = a,

g(a) = g(b) = a.

101

K90
If f satisfies the conditions of the first sentence of the second paragraph, we can write f = g J where J(x) = x and g satisfies the
conditions of the first paragraph. Thus
for all x E.

f 0 (x) = J 0 (x)g 0 (J(x)) = g 0 (J(x)) < 0

If ad bc = 0 then f is constant and f 0 = 0.

102

K91
(i) Suppose that f is continuous. If f is not strictly monotonic we
can find x1 < x2 < x2 so that f (x1 ), f (x2 ) and f (x3 ) are not a strictly
increasing or a strictly decreasing sequence. Without loss of generality
(or by considering f (1 x) and 1 f (x) if necessary), we may suppose
f (x3 ) f (x1 ) f (x2 ). If either of the inequalities are equalities then
f is not injective. Thus we may suppose f (x3 ) > f (x1 ) > f (x2 ). But
by the intermediate value theorem we can find c (x2 , x3 ) so that
f (c) = f (x1 ) and f is not injective.
Suppose f is strictly monotonic. Without loss of generality we suppose f increasing. We shall show that f is continuous at t (0, 1)
(the cases t = 0 and t = 1 are handled similarly). Observe that
f (1) > f (t) > f (0). Let > 0. Set 0 = min(1 f (t), f (t), )/2.
We can find s1 and s2 with f (s1 ) = f (t) 0 f (s2 ) = f (t) + 0 . If
O/ta = min(t s1 , t + s2 ), then |s t| > implies s2 > s > s1 so
f (s2 ) > f (s) > f (s1 ) and |f (s) f (t)| 0 < .
(ii) The same proof as in (i) shows that, if g is continuous and injective, it must be strictly monotonic. The example of g(t) = t/2 for
t 1/2, g(t) = t/2 + 1/2 shows that the converse is false.
(iii) The same proof as in (i) shows that, if g is strictly monotonic
and surjective, it must be continuous. The example g(t) = sin t shows
that the converse is false.

103

K92
Write f (x) = (log x)/x. By looking at f 0 we see that f is strictly
increasing from to e1 on the interval (0, e], has a maximum value
of e1 at e and is strictly decreasing from e1 to 0 on the interval
(0, ). The function f has a unique zero at 1,
We have xy = y x if and only if f (x) = f (y). The set
{(x, y) : xy = y x , x, y > 0}

is thus the union of the straight line {(x, x) : x > 0} and a curve with
reflection symmetry in that line, having asymptotes x = 1 and y = 1
and passing through (e, e).
If f (x) = f (y) and x < y, then x e so we need only examine the
cases m 2 to see that the integer solutions of nm = mn are n = m,
(m, n) = (2, 4) and (m, n) = (4, 2).
Since f () < f (e) we have e > e .

104

K93
Observe that
xy =

1
4

(x + y)2 (x y)2 .

105

K94
(i) Choose with > > 1. If we set f (x) = x x , then (recalling
that x = exp(x log ))
f 0 (x) = x1 x + x (log ) x = x1 x ( + x) > 0
for x sufficiently large (x x0 , say). Thus

x x = f (x)(/)x f (x0 )(/)x .

(ii) Set
(
1/2
1/2 )

1
1
1
, f3 (x) =
, f2 (x) = exp
f4 (x) = exp
log
x
x
x
and
f1 (x) =

1
log
x

Set y = 1/x so y as x 0+.

Observe that log f4 (x)/ log f3 (x) = 2y/(log y) as x .

Observe that log f3 (x)/ log f2 (x) = (log y)1/2 /2 as x .

Observe that log f2 (x)/ log f1 (x) = (log y)1/2 /(3 log log y) as
x .

106

K95
(ii) If Q has a root 6= 0 of order n 1 then differentiating the given
equation n 1 times and setting z = we get P ()Q(n) () = 0 so
P () = 0 and P and Q have a common factor. Thus Q(z) = Az N for
some N 1, A 6= 0 and it is easy to check that this too is impossible.
(iv) If the degree of P is no smaller than that of Q then
P (x)
log x .
Q(x)
If the degree of P is smaller than that of Q then
P (x)
log x 0.
Q(x)
If the degree of P is no smaller than that of Q then
P (x)
log x .
Q(x)

107

K96
Set f (x) = xlog(1+x). Then f (0) = 0 and f 0 (x) = 1(1+x)1 0
for 0 < x and f 0 (x) 0 for x > 0. so f (x) 0 and x log(1 x)
for 0 x 1.
Set g(x) = log(1 + x) x + x2 . Then g(0) = 0 and

g 0 (x) = (1 + x)1 1 + 2x = (x + 2x2 )(1 + x)1 0

for x > 0 and g 0 (x) 0 for 1/2 x 0 so g(x) 0 and log(1 + x)

x x2 for x 1/2.
Since
log
we have

kn
kn

Y
r
r X
log 1 2
1 2 =
n
n
r=1
r=1

kn
kn
kn
kn
X
X
Y
X
r
r
r
r2

log

.
1

2
2
2
4
n
n
n
n
r=1
r=1
r=1
r=1

But

and

as n so

kn
X
1
k2
kn(kn + 1)
k
r
k
+

=
=
n2
2n2
2
n
2
r=1
kn
X
r2
(kn)3
k

=
0
4
4
n
n
n
r=1
kn
Y
k2
r
log
1 2
n
2
r=1

and the result follows on applying exp.

108

K97
Suppose that x 1. Then xn 1 and

2n (xn 1) 2n+1 (xn+1 1) = 2n (x2n+1 2xn+1 + 1)

= 2n (xn+1 1)2 0.

Thus the sequence 2n (xn 1) is decreasing bounded below by 0 and so

tends to a limit. A similar argument applies if 1 > x > 0.
Since
2n (xn 1)
= xn 1
2n (1 1/xn )

2n (xn 1) and 2n (1 1/xn ) tend to the same limit.

(i) If x = 1, then xn = 1 so log 1 = 0.
(ii) If xn = 1 + n , y = 1 + n then
2 n
2 n

n n
(1 + n )(1 + n )
= 1+
.
1 + n + n
1 + n + n
Since 2n n tends to a limit we can find an A such that |n A2n .
Similarly we can find a B such that |n B2n . Thus we can find a
C such that

n n
2n

1 + n + n C2 .
It follows that

2n n
(1 + )(1 + ) 2n
X
2

n
n
C r 22nr
1

r=1 r
1 + n + n
n

2
X

2nr C r 22nr

r=1

2
X

2nr C r 2nr

r=1

as n .

C2n
0
1 C2n

By the mean value theorem

(b2
for some c (a, b) so

a2 ) = 2n c2

n 1

(b a)

2n ((1 + n )(1 + n ) (1 + n + n ) 0

109

that is to say
2n ((xy)2

so writing z = xy
and

(xn + yn 1)) 0

2n (zn 1) 2n (xn 1) 2n (yn 1) 0

log xy = log x + log y.

Result (ii) shows that we need only prove (iii), (iv), (v) and (vi) at
the point x = 1. To do this observe that a simple version of Taylors
theorem
|f (1 + ) f (1) f 0 (1)| sup |f 00 (1 + )|| 2 |
applied to f (x) = x

2n

||||

yields
n

|(1 + )2 1 2n | 2n 4 ||2

for all || 101 and n 3. (We can do better but we do not need
to.) Thus if || 101 and we set x = 1 + we obtain
for n 3 so that

|2n (xn 1) | 4||2

| log(1 + ) | 4||2 .

Thus log is continuous and differentiable at 1 with derivative 1.

Since log(x+t) = log x+log(1+t/x) the chain rule gives log 0 x = 1/x
and since log0 x > 0 we see that log is strictly increasing.
(vii) Since log 2 > log 1 = 0 we have log 2n = n log 2 so
log x as x and log y = log y 1 as y 0+. Since
log is continuous and strictly monotonic, part (vii) follows.

110

K98
If

f (x + )
f (x)

1/

then, taking logarithms,

log f (x + ) log f (x)
log l

so log f is differentiable and, by the chain rule, f = exp log f is differentiable at x. Since
d
f 0 (x)
log l =
log f (x) =
dx
f (x)
we have
f 0 (x)
l = exp
.
f (x)

111

K99
If
1
x1

x a
xa
tends to a limit as x a then (x a)f (x) 0. Since
f (x) =

as x a this gives

x a
a1
xa

1=0

so = .
If = then f (x) = F (x)/G(x) with
F (x) = x ax1 x + a

G(x) = x+1 a x ax + a+1

We find F (a) = F 0 (a) = 0, G(a) = G0 (a) = 0 and use LHopitals
rule to give
F 00 (a)
1
f (x) 00
=
.
G (a)
2a

112

K100
We only consider x when x is real and we define it as exp( log x)
where log is the real logarithm.

113

K101
(iii) Set f (x) = log g(x) and apply (i) to get g(x) = xb for some real
b. Easy to see that this is a solution.
(iv) Observe that, if x > 0, g(x) = g(x1/2 )2 > 0. Thus, by (iii)
g(x) = bx for all x > 0 and some b > 0.
Now g(1)2 = g(1) = 1 so g(1) = 1 and either g(x) = b|x| for all
x or g(x) = sgn(x)b|x| for all x. (Recall sgn x = 1 if x > 0, sgn x = 1
if x < 0.) It is easy to see that these are solutions.

114

K102
(i) We know that
(t/2)2 = t
but the equation (exp is)2 = exp iu, where u R, has two real solutions
in s with < s (the solutions being s/2 and one of s/2 ).
As a specific example, if (t) = exp 2it, then (3/4) = /2 but
(3/8) = 3/4.
However since is continuous we can find a > 0 such that |(t)
1| 1001 for |t| . If |t| then |(t)| /4 and |(t/2)| /4
so (t/2) = (t)/ (since |(t/2) | > /4).
(iii) The same ideas show that the continuous homomorphisms are
precisely the maps 7 n for some integer n.

115

K103
(i) f (0) = 2f (0) so f (0) = 0.
k
(ii) Set F (2k p1
j ) = 2 j whenever j and k are integers with j 1
and F (t) = 0 otherwise. (By the uniqueness of factorisation F , is well
defined.)

0.

1
Since F (p1
j ) and pj 0 as j , F is not continuous at

(iii)
f (t2n ) f (0)
f 0 (0)t
t2n
as n so f (t) = f 0 (0)t for all t 6= 0 and so for all t.
f (t) = 2n f (t2n ) = t

k 1
(iv) Set F (2k p1
j ) = 2 pj whenever j and k are integers with j 1
and F (t) = 0 otherwise. Since |F (t) F (0)| = |f (t)| |t|, we have F
continuous at 0. Since
F (p1
F (21/2 j 1 ) F (0)
j ) F (0)
0
=
1

1
but
21/2 j 1
p1
j

as j , F is not differentiable at 0.

(v) Observe that u(t) = u(t/2)2 0 for all t. If u(x) 0 then

n
u(2n x) = u(x)2 1 so by continuity u(0) = 1. Thus by continuity
there exists a > 0 such that u(t) > 0 for |t| < . Thus u(t) =
n
u(2n t)2 > 0 for |t| < 2n and so u is everywhere strictly positive.
Now set f (t) = log u(t) and apply part (iii).

116

K104
Consider the case x 6= 0. Observe that

1
= 1 + ()||
1/2
(1 + )
2
with () 0 as 0. (Use differentiation or Taylor series.) Observe
also that |x h| kxkkhk. Thus
x+h
f (x + h) =
(kxk2 + 2x h + khk2 )1/2
x+h
=
1/2

kxk 1 + (2x h + khk2 )/kxk2

2x h + khk2
x+h
+ 1 (h)khk
1
=
kxk
2kxk2
h
(x h)x
= f (x)

+ 2 (h)khk
kxk
kxk3
where kj (h)k 0 as khk 0. Thus f is differentiable at x and
Df (x)h =

(x h)x
h
.

kxk
kxk3

kxk2
hx
xh
= 0.
(Df (x)h) x =
kxk
kxk3
Since
kf (x) f (0)k = 1 9 0

as kxk 0, f is not even continuous at 0

Observe that f is constant along radii.

117

K105
Without loss of generality take z0 = 0.
(i)(ii) If f is complex differentiable at 0, then writing f 0 (0) =
A + Bi with A and B real, and taking h and k real
f (h + ik) = f (0) + (A + Bi)(h + ik) + (h + ik)|h + ik|
with (h + ik) 0 as |h + ik| 0. Taking real and imaginary parts,

1 (h, k)
Ah Bk
u(0.0)
u(h, k)
(h[ 2 + k 2 )1/2)
+
+
=
2 (h, k)
Ak + Bh
v(0, 0)
v(h, k)

with

1 (h, k)

2 (h, k) 0

as (h2 + k 2 )1/2 0. Thus F is differentiable at 0 with Jacobian matrix

A B
B A
If A = B = 0 take = 0 and = 0. Otherwise take = (A2 + B 2 )1/2
and such that cos = A, sin = B.
(ii)(iii)(iv) Just a matter of correct interpretation.
(iv)(i) Carefully reverse the first proof.

118

K106
(i) Observe that
g(x + h) = f (x + h, c x h)

= f (x, c x)f,1 (x, c x)h + f,2 (x, c x)(h) + (h, h)(h2 + h2 )1/2

= g(x) + f,1 (x, c x) f,2 (x, c x) + 21/2 (h, h)|h|

where (h, k) 0 as (h2 + k 2 )1/2 0. Thus g is differentiable with

derivative
g 0 (x) = f,1 (x, c x) f,2 (x, c x).
(ii) Observe that g = f u where

x
u(x) =
.
cx

Thus u has Jacobian matrix

1
.
1

By chain the rule g is differentiable with 1 1 Jacobian matrix

1
f,1 (x, c x) f,2 (x, x c)
= (f,1 (x, c x) f,2 (x, c x)).
1

Define g as stated. If f,1 = f,2 , then g 0 = 0 so, by the constant value

theorem, g is constant. Thus f (x, xc) = f (y, y c) for all x, y, c R.
Writing h(x + y) = f (0, x + y), we have the required result.

119

K107
(i) Differentiating with respect to and applying the chain rule, we
have
m
X
1
f (x) =
xj f,j (x).
j=1

Taking = 1 gives the required result.

(ii) We observe that v is differentiable with

m
m
X
X
0
1
v () =
xj f,j (x) =
xj f,j (x) = 1 v().
j=1

j=1

Thus

d
( v()) = 0
d
and by the constant value theorem
v() = v(1)
and f (x) = f (x).

120

K108
This is really just a question for thinking about.
(i) We seek to maximise the (square root of)

2
a b

cos

= (a cos + b sin )2 (c sin + b cos )2

f () =
c d
sin

and this we can do, in principle, by examining the points with f 0 () = 0.

(ii) We seek to maximise the (square root of)
m
!2
p
X
X
ars xs
r=1

s=1

subject to the constraint

m
X

x2s = 1

s=1

which can, in principle, be handled using Lagrange multipliers.

However, this involves solving an m m set of linear equations involving a parameter leading to polynomial of degree m in . The m
roots must then be found and each one inspected. This neither sounds
nor is a very practical idea even when m is quite small.

121

K109
If e1 , e2 , . . . , em is the standard basis (or any orthonormal basis)
and (aij ) is the associated matrix of , then
* n
+
m
X
X
aij =
arj er , ei
arj her , ei i =
r=1

r=1

= hej , ei i = hej , ei i = hei , ej i = aji

Conversely if aij = aji for all i, j, then essentially the same calculation
shows that
hej , ei i = hei , ej i

and so by linearity
! m
* n
+
X
X
xr er ,

ys e s
r=1

s=1

=
=
=

m X
m
X

r=1 s=1
m X
m
X

r=1 s=1
* n
X

xr ys her , es i
xr ys her , es i

xr er ,

r=1

m
X
s=1

ys e s

!+

Let u1 , u2 , . . . , em be an orthonormal basis of eigenvectors with

eigenvalues 1 , 2 , . . . m with |1 | |2 | |m |. Then
2
m
!2 m
m
m

X
X
X

2 2
2
x2j
xj j 1
xj j u j =
xj u j =

j=1

j=1

j=1

j=1

with equality when x2 = x3 = = xm = 0 (and possibly for other

values). This proves .
The matrix A given has eigenvalue 0 only. However kAk = 1.

122

K110
(ii) Let x =

Pn

j=1

xj ej . Then

yk =

n
X
j=1

unless x1 = 0.

kj xj
(

k
2 1/2
r=1 (r xr ) )

Pn

e1

If the largest eigenvalue is negative, then (in general) kxk k |1 |

and k(1)k yk u1 k 0 where u1 = e1 .
(iii) (c) The number of operations for each operation is bounded by
An2 for some A (A = 6 will certainly do),
Suppose |1 | > |j | for j 2. Then the e1 component of k x
will grow at a geometric rate j / relative to the other components.
(v) Look at (iii). If the two largest eigenvalues are very close the
same kind of thing will occur.
(iv) Suppose e1 the eigenvector with largest eigenvalue known. Use
the iteration
x 7 ((x x e1 )

or something similar. However, the accuracy to which we know e1 will

limit the accuracy to which we can find e2 and matters will get rapidly
worse if try to find the third largest eigenvalue and so on.
(vii) No problem in real symmetric case because eigenvectors orthogonal.

123

K111
(ii) hx, xi = hx, xi = kxk2 .
Thus
k kkxk2 = kxkk xk |hx, xi| = kxk2 kk2 kkxk2

for all x and so | k kk2 .

(iii) k kkk k k kk2 and so either = 0 (in which case

the result is trivial) or kk 6= 0 and k k kk. But = so kk =
k k kk. Thus kk = k k and kk2 = k kkk k k kk2
so k k = kk2 .

Since ( ) = = we have symmetric. If e = e

with e 6= 0 then
kek2 = he, ei = he, ei == he, ei = kek2 .

Thus 0.

(vi) Multiplication of A with A takes of the order of m3 operations

(without tricks) and dominates.

1 1

.
(vii) We have A A =
1 5
A has eigenvalues 1 and 2. A A has eigenvalues 4 51/2 . kAk =
(4 + 51/2 )1/2 .

124

K112
Follow the proof of Rolles theorem. If kc|| < 1 and f has a maximum
at c then f,1 (c) = f,2 (c) = 0 so Df (c) is the zero map.
g(x, y) = x2 will do. Observe that cos2 a cos b sin is not
identically zero (consider = /2 and other values of ).
We can not hope to have a Rolles theorem proof in higher dimensions.

125

K113
Since (x, y) 7 x y, (x, y) 7 xy, x 7 x1 and x 7 f (x) are all
continuous we see that F which can be obtained by composition of such
functions is continuous except perhaps at points (x, x).
If F is continuous at (x, x) then F (x + h, x) must tend to the limit
F (x, x) as h 0 so f is differentiable and F (x, x) = f 0 (x). Since
F (x + h, x + h) F (x, x) as h 0, f 0 must be continuous.
Conversely if f 0 exists and is continuous then, setting F (x, x) =
f 0 (x), we saw above that F is continuous at all points (x, y) with x 6= y.
If h 6= k the mean value theorem gives
f (x + h) f (x + k)
F (x + h, x + k) F (x, x) =
f 0 (x)
xy

0
= f x + (h,k h + (1 h,k )k) f 0 (x)

for some h,k with 0 < h,k < 1. But it is also true that, if we set
h,h = 1/2,
F (x + h, x + h) F (x, x) = f 0 (x + h) f 0 (x)

= f 0 x + (h,h h + (1 h,h )h) f 0 (x).

Thus

F (x + h, x + k) F (x, x) = f 0 x + (h,k h + (1 h,k )k) f 0 (x) 0

as (h2 + k 2 )1/2 0.

Suppose now that f is twice continuously differentiable. Much as

before, the chain rule shows that F is differentiable at all points (x, y)
with x 6= y. If f is twice continuously differentiable then the local form
of Taylors theorem shows us that
f 00 (x) 2
h + (h)h2
f (x + h) = f (x) + f 0 (x)h +
2
with (h) 0 as h 0. Thus, if h 6= k,
F (x + h, x + k)

(f (x) + f 0 (x)h + 12 f 00 (x)h2 + (h)h2 ) (f (x) + f 0 (x)k + 12 f 00 (x)k 2 + (k)k 2

hk
00 (x)
2 + (k)k 2
f
(h)h
= f 0 (x) +
(h + k) +
2
hk
f 00 (x)
= F (x, x) +
(h + k) + 0 (h, k)(h2 + k 2 )1/2
2

where 0 (h, k) 0 as (h2 + k 2 )1/2 0. The formula can be extended

to the case h = k by applying the appropriate Taylor theorem to f 0 .
Thus F is differentiable everywhere.

126

K114
If k 6= 0, then we can find an X > 0 such that |f 0 (t) k| < |k|/4 and
so |f 0 (t)| > 3|k|/4 for t X. Thus, using the mean value inequality,

f (x) f (x) F (X) x X F (X)

=

+
x
xX
x
x

f (x) f (X) x X f (X)

x x
xX

3|k| x X f (X)

4 x x
|k|
|k| |k|

2
4
4
1
whenever x max(3X, (4|f (X)|) ).
g(x) = x2 sin x4 will do.

127

K115
The local form of Taylors theorem
f 00 (x) 2
h + (h)h2
2
applied to log (or other arguments) show us that
f (x + h) = f (x) + f 0 (x)h +

log(1 h) = h

h2
+ (h)h2
2

with (h) 0 as h 0.
Thus since qn we can find an N such that

21 qn1 log(1 qn1 ) and 0 log(1 qn1 ) + qn1 2qn2 2n2

for all n N , so the results follow from the comparison test.

128

K116
If a = b = 0 maximum c.
If a = 0 maximum c if b 0, 10b + c if b 0
If a > 0 then maximum c if b/2a 5, 100a + 10b + c if b/2a 5.

If a < 0 then maximum c if b/2a 0, cb2 /(4a) if 0 b/2a 10,

100a + 10b + c if 10 b/2a.

129

K117
(i) Take x1 = 1, xj = 0 otherwise.
(ii) To see that B is positive definite if A is, take
1
1
x1 = (a1
11 a12 x2 + a11 a13 x3 + + a11 a1n xn ).

130

K118
(i) One way of seeing that the j are continuous is to solve the
characteristic equation. Now use the intermediate value theorem.

cos t sin t
will do.
(iii) B(t) =
sin t cos t

131

K119
The reduction to Figure K2 can not be rigorous unless we state
clearly what paths are allowed. However if we have a path which is not
symmetric under reflection in the two axes of symmetry joining mid
points of opposite sides then by considering the reflected path we can
reconstruct a shorter symmetric path.
[A more convincing route involves the observation that the shortest
path system for three points M , N , P is three paths M Z, N Z, P Z
making angle 2/3 to each other provided the largest angle of the
triangle M N P is less than 2/3 and consists of the two shortest sides
otherwise. (See Courant and Robbins What is Mathematics?.)]
One we have the diagram we see that the total path length is
f () = 2a sec + (a a tan )

where a is the length of the side of the square and is a base angle of
the isosceles triangles.
1 2 sin
f 0 () =
cos2
so f increases on [0, /6] and decreases on [/6, /4] so the best angle is
/6 (check that this gives an arrangement consistent with the statement
in square brackets). IN PARTICULAR = /4 is not best.
There are two such road patterns the other just being a rotation
through /4

132

K120
The time taken is
h x (1 + x2 )1/2
+
u
v
for 0 x h. We observe that
x
1
f 0 (x) = +
u v((1 + x2 )1/2
so f 0 is negative and f is decreasing for 0 x with 0 x (1
v 2 /u2 )1/2 and f 0 is positive and f is increasing for x (1v 2 /u2 )1/2 .
Thus if (1 v 2 /u2 )1/2 < h (i.e. (1 v 2 /u2 )h2 > 1) he should take
x = (1 v 2 /u2 )1/2 . However, if (1 v 2 /u2 )h2 < 1 his best course if
0 x h is to take x = h. Since choosing x outside this range is
clearly worse he should take x = h.
f (x) =

If h is large his initial run is almost directly towards the tree and his
greater land speed makes this desirable. If h is small his run is almost
perpendicular to the direction of the tree and does him very little good.

133

K121
We wish to find the maxima and minima of f (x, y) = y 2 13 x3 + ax
in the region x2 + y 2 1. We first examine the region x2 + y 2 < 1.
The stationary points are given by
0 = f,1 (x, y) = x2 + a, 0 = f,2 (x, y) = 2y.

Thus, if a < 0 or a 1, there are no interior stationary points and

both the global maximum and minimum must occur on the boundary.
If 0 < a < 1, then there are two interior stationary points at x =
a1/2 , y = 0. The Hessian is

2x 0
0
2

so (a1/2 , 0) is a saddle and (a1/2 , 0) a minimum. Thus the global

maximum must occur on the boundary and the global minimum may
occur at (a1/2 , 0) with value 2a3/2 /3 or may occur on the boundary.
If a = 0 then there is one stationary point at (0, 0), the Hessian is
singular there but inspection of the equation f (x, y) = y 2 13 x3 shows
that we have a saddle and both the global maximum and minimum
must occur on the boundary.
Now we look at the boundary x2 + y 2 = 1. Here
f (x, y) = g(x) = 1 x2 13 x3 + ax

with x [1, 1]. Notice that g is a cubic so we can draw sketches to

help ourselves.
Now g 0 (x) = 2x x2 + a and g 0 has roots at 1 (1 + a2 ). If |a| 1
then g is decreasing on (1, 1) so has maximum at x = 1 (which is
thus the global maximum for f at (1, 0)) and a minimum at x = 1
9which is thus the global minimum for f at (0, 1)).
If |a| < 1, g increases for x running from 1 to 1 (1 + a2 )
(which is thus a global maximum and f has its global maximum at

1/2
x = 1 + (1 + a2 ) y = 1 (1 + (1 + a2 ))2
) and decreases as x runs
2
from from 1 (1 + a ) to 1. Thus g has minima at x = 1 and 1 and
(by evaluating g at these points) a global minimum at 1 if a 1/3.
Since a 1/3 we know that this gives a global minimum for f at
(0, 1). If a 1/3, g has a global minimum at 1 and if 0 a 1/3
we know that it gives a global minimum for f at (0, 1).
If 1 > a > 0 we observe that we know that the global minimum
occurs when y = 0 and by looking at h(x) = f (x, 0) = 13 x3 + ax we
see that it occurs at (a1/2 , 0).

134

K122
The composition of differentiable functions is differentiable and the
product of differentiable functions (with values in R) are differentiable
so since the maps (x, y) 7 r and (x, y) 7 are differentiable except
at 0, f is. (To see that (x, y) 7 r is differentiable observe that r =
(x2 + y 2 )1/2 and use basic theorems again. To see that (x, y) 7 is
differentiable on
S = {(x, y) : |y| > 9x}

observe that on S, = tan1 x/y. To extend to R2 \ {0} either use

rotational symmetry or cover R2 \{0} with rotated copies of S on which
similar formulae hold.
Since
f (r cos , r sin ) f (0, 0)
g()
r
as r 0+, f has a directional derivative at 0 if and only if g() =
g().

so

If f is differentiable at 0 then
f (r cos , r sin ) f (0, 0)
cos f,1 (0, 0) + sin f,2 (0, 0)
r
g() = g(0) cos + g(/2) sin

ie
g() = A sin + B cos
for some constants A and B so
f (x, y) = Ax + By.
The necessary condition is sufficient by inspection.

135

K123
(i) Observe that.
f (t, t) = t2 ( a2 t)( b2 t)

which has a strict local minimum at t = 0.

However, if a > c > b, then

f (t, ct2 ) = (c a)(c b)t4

which has a strict local maximum at t = 0.

By looking at the behaviour of f along curves (x, y) = (t, t2 ) the

function f has no minima.
(ii) In part (i), f has Hessian

0 0
f,11 (0) f,12 (0)
=
0 2
f,21 (0) f,22 (0)

which is singular.

136

K124
(i) The second sentence holds by Exercise 8.22 which says that if f
is Riemann integrable so is |f |.
(ii) First sentence false. Take f = g = F .
Second sentence false. Take
f (x) = F (x), g(x) = 0

for x < 1/2

f (x) = 0,

for x 1/2.

g(x) = F (x)

(iii) By considering functions of the form

f (x) = F (x), g(x) = 0

for x < 1/2

f (x) = 0,

for x 1/2.

g(x) = F (x)

and of the form

f (x) = F (x), g(x) = 0

for x < 1/2

f (x) = 0,

for x 1/2.

g(x) = 1

and functions like f (1 x) and g(x) we see that the product of two
functions that are not Riemann integrable may or may not be Riemann
integrable and that the product of one that is Riemann integrable with
one that is not may or may not be Riemann integrable.

137

K125
(i) Consider the dissection DN = {r/N : 0 r N } where N M 2 .
At most 2M 2 intervals [(r 1)/N, R/N ] contain points of the form p/q
with p and q coprime and 1 q M . Thus
0 = s(f, DN ) S(f, DN ) 2M 2 N 1 + M 1 M 1

as
R 1 N . Since M is arbitrary f is Riemann integrable with
f (t) dt = 0.
0

(ii) f is continuous at irrational points x. Since there are only finitely

many points of the form p/q with p and q coprime and 1 q M and
x is not one of them we can find a > 0 such that (x , x + ) [0, 1]
contains no such points. Thus |f (x) f (y)| = f (y) M 1 for all
|x y| < .
f is discontinuous at rational points x. Choose xn irrational such
that xn x. Since 0 = f (xn ) 9 f (x), we are done.
(iii) Nowhere differentiable. By (ii) need only look at irrational
points x. If q is a prime we know that there exists a p such that
(p 1)/q x < p/q so
(f (p/q) f (x)
1/p

=1
p/q x
1/p
but choosing xn irrational such that xn x we have
(f (xn ) f (x)
= 0 0.
xn x

138

K126
Observe that, if 0 u < v 1, then un v n n(y x) (by the mean
value theorem or algebra). Write fn (x) = f (xn ). If D is a dissection
write
Dn = {y 1/n : y D}.

By the observation of the first sentence

S(fn , Dn ) s(fn , Dn 0 n S(f, D) s(f, D)

so, if f is Riemann integrable so is fn .

(ii) and so (i) is true

Suppose that |f (x)| K. Given > 0, we can find a 1 > > 0 such
that |f (x) f (0)| /2 for 0 x . Since 1/n 1 as n we
can find an N such that 1 1/n (K + |k|)1 /2 for n N .

If N n then |fn (x) k| /2 for x [0, 1/n ] and |fn (x)| K for
x [ 1/n , 1] so

Z 1/n
Z 1

f (x) dx k(1 1/n /2 and

f (x) dx k 1/n /2

0
1/n
so that

f (x) dx k .

(iii) is false. Take f (x) = 0 for x 6= 0 and f (0) = 1.

139

K127

Without loss of generality, suppose that f (a + b) 0. Observe

that, since f (a) = 0 and |f 0 (t)| K it follows that, by the mean value
inequality, |f (s)| K(s a) for a s (a + b)/2 with equality when
s = (b + a)/2 if and only if f (s) = K(s a) for all a s (a + b)/2.
Similarly |f (s)| K(b s) for b s (a + b)/2 with equality when
s = (b + a)/2 if and only if f (s) = K(b s) for all a s (a + b)/2.
Unless K = 0 (in which case the result is trivial) the conditions for
equality give f non-differentiable at (b + a)/2. Thus if we write

K(b a)/2 |f (b + a)/2 = 4(b a)

we have 4 > 0 and, by continuity, we can find a > 0 such that

|f (s)| (K )(s a) for (b + a)/2 s (b + a)/2
|f (s)| (K )(b s) for (b + a)/2 s (b + a)/2 +

Thus
Z b
Z
|f (s)|
a

(a+b)/2

(b+a)/2

+
(a+b)/2
2

K (b a)/2

< K(b a)2 /4

(a+b)/2+

+
(a+b)/2

b
(a+b)/2+
2

+ (K )( b a)/2

|f (s)| ds

(b a)/2

as required.

To see that this is best possible define

f (x) = K(x a)

f (x) = C (x (a + b)/2)2

for a x (a + b)/2 ,

for (a + b)/2 < x (a + b)/2,

and f ((a + b)/2 t) = f ((a + b)/2 + t) where C is chosen to make f

continuous and is chosen to make f differentiable at (a + b)/2
(thus we take = K/(2)).

140

K128
(v) Observe P
that once we know that f is Riemann integrable we
n1
ba
know that n
j=0 f (a + j(b a)/n) converges but the fact that
Pn1
ba
j=0 f (a + j(b a)/n) converges does not imply that f is Rien
mann integrable.

141

K129
For (A) note that, if Fj and Gj are positive bounded Riemann integrable functions with F1 G1 = F2 G2 then F1 + G2 = F2 + G1
and
Z b
Z b
Z b
F1 (t) dt +
G2 (t) dt =
F1 (t) + G2 (t) dt
a
a
a
Z b
=
F2 (t) + G1 (t) dt
a
Z b
Z b
G1 (t) dt
F2 (t) dt +
=
a

and so

b
a

F1 (t) dt

G1 (t) dt =
a

b
a

F2 (t) dt

G2 (t) dt.
a

The problems with (C) begin (I think) when we try to prove that if f
and g are Riemann integrable so is f + g. We also get problems (which,
I think have the same cause) when we try to show that anything which
is Riemann integrable under the old definition are Riemann integrable
under the new. One way forward is to show that anything which is
(A) integrable is (C) integrable. To do this we can first prove that a
bounded positive function f is Riemann integrable if and only if given
any > 0 we can find a dissection
D = {x0 , x1 , . . . , xN }

with a = x0 < x1 < < xN = b and a subset of {r : < 1 r N }

such that
and

|f (x) f (y)| for x, y [xr1 , xr ] when r .

X
(xr xr1 ) .
r
/

142

K130
(iii) Everything works well until we try to define the upper and lower
integrals but the set of s(D, f ) has no supremum in Q and S(D, f ) has
no infimum.

143

K131
(i) Since F is continuous on a closed bounded interval the infimum
and supremum are attained at and say. Now use the intermediate
value theorem.
(ii) Observe that
sup f (s)w(t) f (t)w(t)

s[a,b]

for all t [a, b] so

sup f (s)
s[a,b]

w(t) dt

f (t)w(t) dt
a

Using a similar result for the infimum we have

Z b
Z b
Z b
sup f (s)
w(t) dt
f (t)w(t) dt inf f (s)
w(t) dt.
s[a,b]

Now use part (i).

(iii) Take a = 1, b = 1, F (t) = w(t) = t.
If w everywhere negative, consider w.

s[a,b]

144

K132
The graph of f splits the rectangle [a, b] [f (a), f (b)] into two parts
whose areas correspond to the given integrals.
Consider a dissection of [a, b]
D = {x0 , x1 , . . . , xN }

with a = x0 < x1 < < xN = b. This gives rise to a dissection of

[f (a), f (b)]
D0 = {f (x0 ), f (x1 ), . . . , f (xN )}

with f (a) = F (x0 ) < f (x1 ) < < f (xN ) = f (b).

We have

S(D, f ) + s(D 0 , g) = bf (b) af (a)

so, taking an infimum over all D, we get

I (f ) + I (g) bf (b) af (a)

Similarly

s(D, f ) + S(D 0 , g) = bf (b) af (a)

so

I (f ) + I (g) bf (b) af (a).

Rb
R f (b)
But I (f ) = I (f ) = a f (t) dt and I (g) = I (g) = f (a) g(s) ds so
Z f (b)
Z b
g(s) ds = bf (b) af (a).
f (t) dt +
f (a)

(ii) Without loss of generality, assume Y F (X). Then

Z Y
Z X
Z f (X)
Z
X
f (x) dx +
g(s) ds =
f (x) dx +
g(s) ds +
0

g(s) ds
f (X)

XF (X) + (Y F (X))X = Y X

with equality if and only if Y = f (X).

(iii) Take f (x) = xp
1/q
X = Y 1/p .

1 1

, g(y) = y q

1 1

. Equality if and only if

145

K133
Since f is continuous at c, we can find a > 0 such that |f (c)
f (x)| < /4 when |x c| < . Thus if [, ] (c , c + )

I[,]
(f ) I[,] (f ) ( )/2.

146

K134
Suppose [a, b]. Then given 1 > > 0 we can find a > 0 such
that
|f (s) f ()|, |f (g(s)) f (g())|, |g 0 (s) g 0 ()| <

for all s [a, b] with |s | < .

Now suppose [, ] ( , + ). Without loss generality we

suppose g() g(). Then
Z
Z g()
g()

f (s) ds g() g() f (g())

|f (s) f (g())| ds

g()
g()

g() g()
sup |f (s) f (g())|
s[g(),g()]

g() g() sup |f (g(s)) f (g())|

t[,]

g() g()

sup |g 0 (t)|( )
t[,]

(|g 0 ()| + )( )
We also have, by a simpler argument
Z

f (g(t)) dt g() g() f (g()) ( ).

Thus

Z
Z g()

f (s) ds
f (g(t)) dt (1 + |g 0 ()| + )( )

g()

We can make as small as we want.

(2 + |g 0 ()|)( ).

147

K135*
No comments.

148

K136
The formula
sup
x[xj1 ,xj ]

f 0 (t)(xj xj1 ) f (xj ) f (xj1 )

inf

x[xj1 ,xj ]

f 0 (t)(xj xj1 )

is a version of the mean value inequality.

Consider a dissection of [a, b]
D = {x0 , x1 , . . . , xN }

with a = x0 < x1 < < xN = b. By summing the formula of the first

paragraph we get
so

S(D, f 0 ) f (b) f (a) s(D, f 0 )

I (f 0 ) f (b) f (a) I (f 0 )

and the result follows.

149

K137*
No comments.

150

K138
Let N 4. Take a dissection DN consisting of the points 1, 0, 1,
1 N 3 and rN 1 N 3 with 1 r N 1. Then

0 = s(DN , f ) S(DN , f ) N 1 + N 2N 3 = N 1 + 2N 2 0
R1
as N . Thus f is Riemann integrable and 1 f (t) dt = 0.

Similarly F (t) = 0 for all t. Thus F 0 (0) exists with value 0 = f (0).
Since f (1/n) = 1 9 0, f is not continuous at 0.

151

K139
(i)
n
X
r=1

n
X
(r + 1)r(r 1) r(r 1)(r 2)

r(r 1) =

r=1

n
X

r=

r=1

n
X

r =

r=1

n
X
(r + 1)r r(r 1)

r=1

n
X

r(r 1) +

r=1

n
X
r=1

r=

(n + 1)n(n 1)
.
3

(n + 1)n
2

n(n + 1)(2n + 1)
.
6

(ii)
n
X
r=1

n3 =

n2 (n + 1)2
.
4

(iii) If f (x) = xm
S(D, f ) =
and
s(D, f ) =

n
X

1
Pm (n)
1
r
+ m+1
n1 ( )m =
n
m+1
n
m+1
r=1

n1
X

r
1
Pm (n 1)
1
n1 ( )m =
(1 n1 )m+1 +

n
m+1
nm+1
m+1
r=0

so I f = I f = (m + 1)1 .
(iv) If f (x) = xm and F (x) = xm+1 /(m + 1) then F 0 = f and
Z b
f (x) dx = F (b) F (a).
a

152

K140
(v) We have |g(t) g(1)(t + 1)| K(t 1) so
Z 1

Z 1

g(t) dt g(1) K
(t 1) dt = 2K.

Scaling, we get
Z

arh

a(r1)h

f (t) dt f (a (r 1)h) Kh2

if |f (t)| K for t [a, b] and N h = b a. Thus, summing,

Z b

K(b a)h.
f
(t)
dt

S
(f
)
h

Taking a = 0, b = , N h = and F (t) = sin2 (nt), we see that the

result cannot be substantially improved.

153

K141
(ii) Let s = yt.
(viii) We have
tn+1
1
(1 + t + + tn ) =
.
1t
1t
So, integrating both sides from 0 to x, we have

Z x n+1
x2
t
xn+1
log(1 x) x +
=
+ +
dt 0
2
n+1
0 1t

as n for all |x| < 1. Thus

X
xr+1
log(1 x) =
r+1
r=1

for |x| < 1. [The result also holds if x = 1 but the argument above
does not prove this without extra work.]
(ix) By (viii)
log

1+x
1x

= log(1 + x) log(1 x)
=

X
(1)r xr+1
r=1

=2

r=1

for |x| < 1.

r+1

x2r+1
2r + 1

X
xr+1
r+1
r=1

If y > 0 and we seek an x with

y=
we obtain

1+x
1x

y1
y+1
so |x| < 1 and substitution in the formula of (ix) gives the desired
result.
x=

154

K142
Power series can be multiplied term by term within their radius of
convergence. We observe that

n1
n1
X
(1)r (1)nr
(1)n X 1
1
(1)n 2Sn1

=
+
.
=
r
nr
n r=1 r n r
n
r=1
The result is valid for |x| < 1.

155

K143
(i) Write
=
and observe that 1 > > 0, so

x3 x 2
x3 x 1

f (x1 ) + (1 )f (x3 ) f (x3 + (1 )x1 )

so that

(x3 x2 )f (x1 ) + (x2 x1 )f (x3 ) (x3 x1 )f (x2 )

and the result follows on rearrangement.

Observe, either similarly, or as a consequence that, if y1 < y2 < y3 ,

then
f (y2 ) f (y1 )
f (y3 ) f (y2 )
f (y3 ) f (y1 )
f (y3 ) f (y1 )

and

.
y3 y 1
y2 y 1
y3 y 2
y3 y 1
(ii) By (i), if h > 0, then
f (c) f (d)
f (c + h) f (c)

h
cd
where d is fixed with c > d > a. Since a non-empty set bounded below
has an infimum, f (c+) exists.
By definition, given > 0 we can find a > 0 such that
f (c + ) f (c)
.

If h > > 0 then (looking at the last paragraph of our discussion of (i))
f (c+) +

so

f (c+) +

f (c + h) f (c)
f (c + ) f (c)

f (c+)

h
f (c + h) f (c)
f (c+)
h

as h 0+.
(iii) We can find a > 0 such that, if > h > 0, then
f (c+) + 1

f (c + h) f (c)
f (c+)
h

and so
|f (c + h) f (c)| (|f (c+)| + 1)|h| 0

as h 0+. Similarly f (c + h) f (c) 0 as h 0 so we are done.

(iv) We have f (x) f (c) B(x c) for all x whenever f (c+)
B f (c).

156

(v) Observe that

g(t) tg(t) g(t)

and integrate to obtain c [, ].

so

We know that there exists a B such that f (x) f (c) B(x c) and

Now integrate.

(f (t) f (c))g(t) B(t c)g(t).

(vi) We use the notation of Exercise K40 (iii). Let c = j xj . We

know that there exists a B such that f (x) f (c) B(x c) and so
f (xj ) f (c) B(xj c)

whence
n
!
n
n
n
X
X
X
X
j f (xj ) f
j xj =
j (f (xj ) f (c))
j (xj c) = 0.
j=1

j=1

j=1

j=1

Similarly we know that there exists a B such that f (x) f (EX)

B(x EX) and so
f (X) f (EX) B(X EX).

Taking expectations gives the result.

(vii) Pick a0 , b0 so that a < a0 < c < b0 < b Observe that,

E = {(x, y) R2 : y f (x), x [a0 , b0 ]}

is a closed convex set. Applying Exercise 32 (iii) with y = (c, f (c)

n1 0 we see that there exist an , bn and cn such that
bn x + an y cn whenever (x, y) E

and bn c + an (f (c) n1 ) > cn . By considering what happens when y is

very large, we see that an 0 and the conditions given are incompatible
if an = 0. Thus an < 0 and, setting Bn = bn /an ,Cn = cn /an , we obtain
Bn x + y Cn whenever (x, y) E

and Bn c + (f (c) n1 ) < Cn .

In particular we have

Bn x + f (x) > Bn c + (f (c) n1 )

for all x [a0 , b0 ]. Thus

f (x) f (c) > Bn (x c) n1

for all x [a0 , b0 ]. If > 0 is such that [c , c + ] [a0 , b0 ] then

f (c + ) f (c) > Bn n1 and f (c) f (c + ) f (c) < Bn + n1 .

157

Thus Bn is bounded and we can find a convergent subsequence Bn(j)

B. We have
f (x) f (c) B(x c)

for all x [a0 , b0 ].

To extend the result to (a, b), observe that the result just proved
shows that that if a < a + n1 < c < b n1 < b we can find a Bn
such that
1

f (x) f (c) Bn (x c)

for all x [a + n , b n1 ]. A subsequence argument now gives the

existence of a B with
for all x (a, b).

f (x) f (c) B(x c)

158

K144
(ii) Suppose that < . If 0 < h, k < ( )/2 then
f ( h) f ( + h
f ( + k) f ()
f () f ( h)

h
hk
k

so

f () f (+).
Thus
N
X

j=1

f (cj +) f (cj )

= f (cN +)

N
X

j=1

f (f (cj ) f (cj1 +)) f (c1 )

f (cN +) f (c1 )
f (+) f ().

In particular the set

En = {c (a, b) : f (c) f (c+)}
contains at most [(f (+) f ())/N ] points (with [x] meaning the
integer part of x). Thus
E = {a, b}

Ej

j=1

is countable. If c
/ E, then f (c) = f (c+) = f (c), say, and
f (c + h) f (c)
f (c)
h
as h 0.
(iv) Enumerate the rational points in (1, 1) as q1 , q2 , . . . . If we set
fn (x) = n2 |x qn |
then fn is convex and
0 fn (x) 2n2
P
Thus, using the comparison test,
n=1 fn (x) converges everywhere to
f (x), say.

159

Noting that, by convexity, fm (x + h) + fm (x h) 2fm (x) 0 for

all m when x, x h, x + h (1, 1), we have
f (qn + h) + f (qn h) 2f (qn )
f (qn + h) f (qn ) f (qn ) f (qn h)

=
h
h
h
fn (qn + h) + f (qn h) 2f (qn )

h
2
= 2n 9 0
as h 0+, so f is not differentiable at qn

160

K145
(i) Given > 0, we can find a > 0 such that |f (x) f (c)| < for
|x c| < . If [an , bn ] (c , c + ), then

Z
Z
1

f (t) dt f (c) =
f (t) f (c) dt
|In |
|In | In
In
Z
1

|f (t) f (c)| dt
|In | In
Z
1

dt =
|In | In
(ii) The following is one generalisation (interpreting the notation
appropriately). Let B(c, r) be the ball centre c then (provided the
integrals exist)
Z
1
f (t) dV (t) f (c)
Vol B(c, r) B(c,r)

as r 0+ whenever f is continuous at c.

161

K146*
No comments.

162

K147
(i) The range of integration is not fixed.
(ii) If x is fixed the fundamental theorem of the calculus shows that
F,2 exists and
F,2 (x, y) = g(x, y)
so F,2 is continuous.
Theorem 8.57 tells us that, if x is fixed we may differentiate under
the integral to obtain
Z y
F,1 (x, y) =
g,1 (x, y) dx.
0

If (x0 , y0 ) is fixed Take R 2 + |y0 | + |x0 |. Since g,1 is continuous on

K = [R, R] [R, R] it is uniformly continuous and bounded. Thus
|F,1 (x, y) F,1 (x0 , y0 )|

|F,1 (x, y) F,1 (x0 , y)| + |F,1 (x0 , y) F,1 (x0 , y0 )|

|y| sup |g,1 (x, t) g,1 (x0 , t)| + |y y0 |

|t||y|

sup
(s,t),(s0 ,t)K,|ss0 ||xx

+ |y y0 |

0|

sup
(s,t),(s0 ,t)K

sup

|t|max(|y|,|y0 |)

|g,1 (x0 , t)|

|g,1 (s, t) g,1 (s0 , t)|

|g,1 (x0 , t)| 0

as k(x, y) (x0 , y0 )k 0. Thus F,2 is continuous and F is once differentiable.

(iii) The chain rule shows that G is differentiable and
Z x
0
G (x) = F1 (x, x) + F2 (x, x) =
g,1 (x, t) dt + g(x, x).
0

(iv) We have H(x) = F (x, h(x)) so the chain rule gives

Z h(x)
0
0
H (x) = F1 (x, h(x)) + F2 (x, h(x))h (x) =
g,1 (x, t) dt + h0 (x)g(x, h(x)).
0

163

K148
Observe that differentiating under the integral, using the symmetry
of partial derivatives and the fundamental theorem of the calculus

E 0 (t) =

=2

1
0

u
(x, t)
t

u
(x, t)
x

2 !

dx

u,2 (x, t)u,22 (x, t) + u,1 (x, t)u,12 (x, t) dx

0
1

u,2 (x, t)u,11 (x, t) + u,1 (x, t)u,21 (x, t) dx

2
2 !
Z 1
u
d
u
(x, t) +
(x, t)
=
dx
t
x
0 dx
"
2
2 #1
u
u
=
=0
(x, t) +
(x, t)
t
x
=2

so by the constant value theorem (and thus the mean value theorem)
E is constant.

164

K149
H is a smooth function with
H (t) = 0
H (t) = c

for t 0

for t 2

and 0 H (t) c for all t, where c is a strictly positive number.

Let L (t) = c1
(H (t) H (t + 1 2). If 0 < < 1/4 then
L (t) = 0

L (t) = 1
and 0 L (t) 1 for all t.

for t
/ [0, 1]

for t [2, 1 2]

Now let kn (x) = (1)[2nx] L/20 (2nx2[nx]) and follow Exercise 8.66.

165

K150
We seek to minimise
Z b
F (x, y(x), y 0 (x)) dx with F (u, v, w) = g(u, v)(1 + w 2 )1/2 .
a

The Euler-Lagrange equation gives

d
0 = F,2 (x, y(x), y 0 (x)) F,3 (x, y(x), y 0 (x))
dx
d g(x, y)y 0
2
= g,2 (x, y)(1 + y 0 )1/2
dx (1 + y 0 2 )1/2

g,1 (x, y)y 0 + g,2 (x, y)y 0 2 + g(x, y)y 00 g(x, y)y 0 2 y 00
+
.
= g,2 (x, y)(1 + y )
(1 + y 0 2 )1/2
(1 + y 0 2 )3/2
Multiplying through by 1 + y 0 2 gives the required result.
0 2 1/2

We seek to minimise
2

x ds.

We take g(x, y) = x and obtain

y 0

xy 00
=0
1 + y02

or

xy 0
=0
(1 + y 0 2 )1/2
(observe that g(x, y) has no direct dependence on y and recall Exercise 8.63). Thus
xy 0
=c
(1 + y 0 2 )1/2
whence
c
y0 = 2
(x c2 )1/2
and
x
y + a = c cosh1
c
for appropriate constants a and c.
d
dx

166

K151
Precisely as in the standard case, we consider
Z b
G() =
f (x, y(x) + g(x), y 0 (x) + g 0 (x), y 00 (x) + g 00 (x)) dx
a

and require
0

0 = G (0) =

f,2 (x, y, y 0 , y 00 ) + g 0 (x)f,3 (x, y, y 0 , y 00 ) + g 00 (x)f,4 (x, y, y 0 , y 00 ) dx

a

for all well behaved g with g(a) = g 0 (a) = 0, g(b) = g 0 (b) = 0.

Integrating by parts once and twice gives

Z b
d
g(x) f,2 (x, y(x), y 0 (x), y 00 (x)) f,3 (x, y(x), y 0 (x), y 00 (x))
0=
dx
a

2
d
+ 2 f,4 (x, y(x), y 0 (x), y 00 (x)) dx
dx
for all g as above, so we get an Euler-Lagrange type equation
d
0 = f,2 (x, y(x), y 0 (x), y 00 (x)) f,3 (x, y(x), y 0 (x), y 00 (x))
dx
d2
+ 2 f,4 (x, y(x), y 0 (x), y 00 (x)).
dx
In the case given this becomes
0 = 24 +

d2 00
y (x)
dx2

that is to say
y 0000 (x) = 24
so y = A + Bx + Cx2 + Dx3 + x4 . The boundary conditions give us
A = B = 0, 0 = C + D + 1, 4 = 2C + 3D + 4 so C = 3, D = 2 and
y(x) = 3x2 + 2x3 + x4 .

If we consider y(x) + x2 (1 x)2 sin N x, we can make small changes

in y increasing I so we can not have a maximum.

167

K152*
No comments.

168

K153
(i) g(1/2) = 1, g(x) = 0 otherwise.
(ii) Could take

g(x) =

1 m2n+3 |x r2n | if |x r2n | < m1 2n3 , 0 r n,

0
otherwise.

169

K154
Second sentence done exactly as the case h = 1 (Lemma 9.6).
R
If 1 f (t) dt converges then, since
Z Nh
N
1
N
1
X
X
f (nh),
f (t) dt h
f (nh)
h
n=0

we have (allowing N )
Z

X
f (nh)
h
n=0

and so

as h 0+.

n=1

f (t) dt h

f (nh),

n=1

f (t) dt hf (0) 0
f (nh)
h

n=1
0

Without loss of generality, suppose 1 > > 0. Choose K > 8N 1

and set
(
1 K|x rN 1 | if |x rN 1 | < K, 0 r n,
g(x) =
0
otherwise.
We could set

G(x) =

2k/2 (1 24k )|x r2k | if |x r2k | < 24k , 2k r 2k+1 1, k 2

0
otherwise.

170

K155
(i) Observe that
Z
Z x
0
f (t) dt f (n) +
f (x) f (n) +
n

x
0

|f (t)| dt f (n) +

for n x n + 1 and so
Z n+1
Z
f (x) dx f (n) +
n

n+1
n

|f 0 (t)| dt

n+1

|f 0 (t)| dt.

Similarly

n+1
n

f (x) dx f (n)

n+1
n

|f 0 (t)| dt.

It follows that,

Z
Z m+1
m
X
m+1

f (r)
f (x) dx +
|f 0 (x)| dx

n
n
r=n
R
Pn
and, if 1 f (x) dx converges, r=0 f (r) is a Cauchy sequence and converges by the general principle of convergence.
P
similar argument shows that
R nConversely, if r=0 f (r) converges, a P
f (x) dx tends to a limit. Further, if
r=0 f (r) converges we have
1
f (n) 0 and the arguments above show that
Z n+1
Z n+1
|f 0 (x)| dx 0
|f (x)| dx |f (n)| +
n
n
R
as n and so 1 f (x) dx converges.
(ii) If f is decreasing, positive and has a continuous derivative
Z X
Z X
0
f 0 (x) dx = f (1) f (X) f (1).
|f (x)| dx =
1

(iii) False. We consider a continuously differentiable function g such

that
g(x) =

(2n)1 22n
(2n + 1)1 22n1

if 22n + 21 x 22n+1 21 , n 0,
if 22n+1 + 21 x 22n+2 21 , n 0,

and, in addition, g is decreasing on [22n+1 21 , 22n+1 +21 ], increasing

on [22n 21 , 22n + 21 ] and g(2m ) = 0 for all m.

171

K156

n+1/2
n1/2

log x dx log n =

n+1/2

(log x log n) dx +

n
n+1/2

log(x/n) dx +
n
1/2

n
n1/2

(log x log n) dx

log(x/n) dx
n1/2

t
t
+ log 1
dt
log 1 +
=
n
n
0
making substitutions like x = t + n and x = n t.
Z

But by the mean value inequality,

t
t
1
1
2

log 1 +
sup
+ log 1

n
n
n t
t(0,1/2) t + n
2t
4
= 21 sup
2.
2
2
3n
t(0,1/2) n t
Thus by the comparison test

X
t
t
log 1 +
+ log 1
n
n
n=1

is absolutely convergent and so convergent to c, say. Thus

Z N +1/2
log x dx log N ! c.
1/2

But, integrating by parts,

Z N +1/2
Z
N +1/2
log x dx = [x log x]1/2
1/2

= (N + 12 ) log(N +

Thus

N +1/2

dx
1/2
1
) 21 ) log 21
2

(N + 1).

(N + 12 ) log(N + 12 ) N log N ! c0

for some constant C and the result follows on taking exponentials (and
noting that exp is continuous).

172

K157
With the notation of the previous question

2
n!
2n!
2n
n
e
e
C 1
(2n + 1/2)(2n+1/2)
(n + 1/2)(n+1/2)
where C is the constant of K156. Thus

1 2n
) 2n 1/2
2n (1 + 4n
C0
1 2n 2 n
n (1 + 2n )

and

2n
C 00
4 n
n
as n for appropriate constants C 0 and C 00 . In particular, we can
find constants K1 and K2 with K1 > K2 > 0 such that

2n
n 1/2
4n n1/2 .
K1 4 n

n
Thus


2n n

2n n
n
z 0 if |z| < 1/4 and n
z 0 if |z| > 1/4

n
n

P
2n n
z has radius of convergence 1/4.
so
n=0 n
n
n
z | K n1/2 for all |z| = 1/4 so,
If < 1/2 we have |n 2n
P n 2n n 1
z converges absolutely and so
by the comparison test,
n=0 n
n
converges everywhere on the circle of convergence.
n
P 2n n
If 1/2 we have |n 2n
z
z
|

K
for
all
|z|
=
1/4
so
2
n=0 n
n
n
diverges everywhere on the circle of convergence.

2n
If 1/2 then
n
K2 n1/2 and the comparison test tells
P 2n n n
us that n=0 n n z diverges when z = 1/4.
n 1/2

The only remaining case is |z| = 1/4,

z 6= 1/4 and 1/2 > 1/2.
n 2n
Let us write w = 4z and un = 4 n n . We know that un 0 so, if
we can showPthat un > un+1 for all sufficiently large n, Abels test will
n
tell us that
n=0 un w . But
1
)
(1 + 2n
un+1
=
1 1 < 1
un
(1 + n )

so the conditions of Abels test are satisfied. (Observe that

P (1 + x)
2n n
1 + x for 0 and x 0.) If 1/2 > 1/2 then n=0 n n z
converges everywhere on the circle of convergence except the point
z = 1/4.

173

K158
(i) Observe that
Z

g(X)

f (s) ds =
0

f (x)g 0 (x) dx
0

and that g(X) as X and, if g(h(Y )) as Y , then

h(Y ) as Y .
Only one limit so no problem of interchange.
(ii) (sin x)/x 1 as x 0 (by considering differentiation, if you
wish).
Observe that
0 (1)n+2

sin(x + n)
sin(x + (n + 1))
(1)n+1
x + (n + 1)
x + n

so
0 (1)

n+2

sin(x + (n + 1))
dx (1)n+1
x + (n + 1)

and, if we write
vn = (1)

n+1

(n+1)
n

sin(x + n)
dx
x + n

sin x
dx,
x

the vN are a decreasing sequence with

Z (n+1)
1
dx (n)1 0
0 vn
x
n
as n . Thus, by the alternating series test,
Z n
n1
X
sin x
dx =
(1)r vr
x
0
r=0

converges to a limit L with v0 L v0 v1 > 0.

Since

we have

as X .

(n+1)
n

sin x
1

x dx (n) 0,

sin x
dx L
x

(iii) Change of variable s = xt with t > 0 gives I(t) = I(1) = L.

Since (sin x)/x = (sin x)/x, I(t) = I(t).

174

(vi) By convexity sin x 2x/ for 0 x /2, so g(x) = g(x) =

2x/ 2 for 0 x /2 will do. Observe that

Z n
n1 Z (r+1)
X
sin x
sin x

dx =

x dx
x

r=1 r
n1 Z (r+1)
X
x r

dx
r
r
r=1
Z
n1
X
1
g(x) dx
=

r
0
r=1
as n .
R x
(v) If > 0, then the arguments above show that 1 sin
dx conx
verges. If 0, then

Z
(n+1) sin x Z (n+1)

| sin x| dx 9 0
dx

n
x
n
R x
so 1 sin
dx does not converge.
x
However, if 0 x 1 we have x sin x 2x/ so
sin x
x1 2x1 /
x
and comparison shows that
Z 1
sin x
dx
x

converges as 0+ if and only if < 2.

R x
Thus 0 sin
dx converges if and only if 2 > > 0.
x

175

K159
(i) Observe that
n
n
X
X
sin((n + 12 )x)
1 ei(2n+1)x
=
1+2
cos rx =
eirx = einx
1 eix
x
r=1
r=n

for |x| .

(ii) Change of variable t = x.

(iii) Observe that the integrals
Z (r+1)/(n+1/2)
r/(n+1/2)

sin (n + 12 )x
dx
sin 12 x

alternate in sign and decease in absolute value as r increases from

1 to n and as r decreases from ! to n. Observe further that if
and /(n + 1/2) then

Z
sin (n + 1 )x Z alpha sin (n + 1 )x

2
2
dx =
dx
0

1
1

(n + f rac12)| sin |

sin 2 x
sin 2 x
as n .
(iv) Write

2(sin 12 x 21 x)
1
2
=

x sin 21 x
x sin 12 x
and use LHopital or Taylor expansions.
(v) Given any > 0 we can find an > 0 with > such that

1
x sin x <
2

for |x| < and so

2 sin((n + 21 )x)
sin((n + 12 )x)

dx
dx 2 2 2 .

sin x
x

Thus, allowing n and using (ii) and (iii) we have

sin
x
2 2
2 2 .
,
dx

Since was arbitrary

and we are done.

sin x
, dx =
x

176

K160
(i) and (ii) (Here permits means permits but does not imply.)
f (n) = O(g(n)), permits f (n) = o(g(n)), permits f (n) = (g(n))
and permits f (n) g(n).
f (n) = o(g(n)), implies f (n) = O(g(n)), forbids f (n) = (g(n)) and
forbids f (n) g(n).
f (n) = (g(n)) permits f (n) = O(g(n)), permits f (n) g(n) and
forbids f (n) = o(g(n)).
f (n) g(n) implies f (n) = O(g(n)), implies f (n) = (g(n)) and
forbids f (n) = o(g(n)).
Suitable examples will be found amongst the pairs f (n) = g(n) = 1;
f (n) = n, g(n) = 1; f (n) = 1, g(n) = n.
(iii) f (n) = o(1) means f (n) 0 as n 0. f (n) = O(1) means f is
bounded.
(iv) Could take f (n) = 1 + n(1 + (1)n ), g(n) = 1 + n(1 + (1)n+1 ).

177

K161
(i) True. If 0 fj (n) Aj gj (n) with Aj 0 then

0 f1 (n) + f2 (A1 + A2 )(g1 (n) + g2 (n)).

(ii) False. Take f1 (n) = f2 (n) = g1 (n) = g2 (n).

(iii) True. If 0 fj (n) Aj gj (n) with Aj 0 then

0 f1 (n) + f2 (A1 + A2 ) max(g1 (n), g2 (n)).

2

(iv) True. n! nn (2n )n = 2n .

(v) True. Use LHopital or (I think better) observe that
cos x 1 + x2 /2 = O(x4 ), sin x = x + (x)x3 ,

with |(x)| < 1 when |x| is small so

(sin x)2 = x2 + 2(x)x4 + (x)2 x5

and
(sin x)2 x2 = O(x4 ).
(vi) False. Use LHopital or (I think better) observe that
cos x 1 + x2 /2 x4 /4! = O(x6 ), sin x = x x3 /6 + (x)x5 ,

with |(x)| < 1 when |x| is small so

sin2 x = x2 x4 /3 + 1 (x)x6 , and sin4 x = x4 + 2 (x)x6

where |1 (x)|, |2 (x)| 10 for |x| small (we can do better but we do
not need to) so
cos x 1 + sin2 x/2 sin4 x/4! + x4 /6 = O(x6 ).

178

K162
(i) If 1 < 0, then x is decreasing so
Z n+1
n
n Z r+1
n+1
X
X
X

r
x dx =
x dx
r .
r=1

r=1

r=2

Since

this gives

x dx =
1

n+1 1
,
+1

P
( + 1) nr=1 r 1 + 2n

n+1
n+1

as n .

If 0, then x is increasing so
Z
n Z r+1
n
X
X

x dx =
r
r=1

r=1

n+1
1

and much the same argument works.

x dx

n+1
X
r=2

(ii) If = 1, then much the same argument gives

n
X
1
log n.
r
r=1
(iv) No such modification, since

r=n

r1 diverges.

179

K163
Argument similar to that in K156. We have |g 00 (x)| Ax for
x R say. If |t| 12 and n R + 1, then applying the mean value
theorem twice gives
|(g(t + n + 12 ) g(n + 12 )) + (g(t + n + 12 ) g(n + 12 ))|
= |g 0 (s + n + 12 ) g 0 (s + n + 12 )|

= |2sg 0 (u + n + 12 )| A(n 12 )

for some s and u with

such that

1
2

> s > 0 and |u| < s. Thus we can find a B

|(g(t + n + 21 ) g(n + 12 )) + (g(t + n + 12 ) g(n + 12 ))| Bn

for n R and |t| 21 .

We have
Z

n+1
n

(g(x) g(n + 21 ) dx

Z 1

2
1
1
1
1

(g(t + n + 2 ) g(n + 2 )) + (g(t + n + 2 ) g(n + 2 )) dt

=

so

1
2

|(g(t + n + 12 ) g(n + 12 )) + (g(t + n + 12 ) g(n + 12 ))| dt

Bn
Z

n+1

(g(x) g(n + 21 )dx = O(n ).

We have
Z

n+1
n

g(x) dx g(n +

1
)
2

Bn

for n R so adding and using K162 or direct argument

Z

n
n+1

X
Bn+1

g(x) dx
g(r) C +

+ 1
r=1
R n+1
for some constant C. Dividing through by 1 g(x) dx gives the required result.

180

K164
Since sin x > 2x/ for 0 x /2, we have

0 log sin x log(/2) log x log(/2) + x1/2

R /2
when x is small and positive, so, by comparison, 0 log(sin x) dx converges. (Similar arguments apply to all the integrals of this question.)
Using the change of variables formula with t = 2x and symmetry we
have
Z /2
Z
Z /2
1
log(sin 2x) dx =
log(sin t) dt =
log(sin x) dx.
2 0
0
0
But, using symmetry again,
Z /2
Z /2
log(sin 2x) dx =
log(2 sin x cos x) dx
0

/2

log 2 dx +
0

/2

log(sin x) dx +
0

Z /2

= log 2 + 2
log(sin x) dx.
2
0
Combining our formulae gives the result.

/2

log(cos x) dx
0

181

K165

so

Ra
Suppose that 0 f (x) dx converges. Since f is decreasing
Z a
Z a
Z a
N
a X ar
f (x) dx
f
f (x) dx
f (x) dx

N r=1
N
0
a/N
0
Z a
N
ar
X
f
f (x) dx.

N
0
r=1

Ra
Suppose that 0 f (x) dx diverges. Then given any K we can find an
> 0 such that
Z a
f (x) dx K + 1.

However, since f is continuous

Z a
ar
X
f

f (x) dx
N

N rN

so we can find an N0 such that

Z a
ar
X

f (x) dx 1
f
N

N rN

and so

N
ar
X
K
f
N
r=1

for all N N0 . Thus

N
ar
X
f

N
r=1

as N .

(ii) Consider a = 1,
(
2n (1 24n |x 2n |) if |x 2n | 24n ,
f (x)
0
otherwise.
(iii) Observe that
(1 z)
so, dividing by z 1,

N
1
X
r=0

N
1
X
r=0

z =1z

z =1z

N
1
Y
r=0

N
1
Y
r=1

(z r )

(z r )

182

for z 6= 1. Since a polynomial of degree N 1 which vanishes at N

points vanishes identically we have
N
1
X
r=0

for all z.

z =1z

N
1
Y
r=1

(z r )

In particular, setting z = 1, we have

N=

N
1
Y
r=1

(1 ) =

1+2++(N 1)

r=1

where = /N . Thus
N =

N (N 1)/2

(2i)

N 1

N
1
Y

N
1
Y
r=1

( r r )

N
1
Y
r
r
N 1
sin
=2
sin .
N
N
r=1

(iv) Observe that

N 1
!
N
r
r
X

X
log sin
sin
log
=
2N r=1
N
2N
N
r=1

log N 1
2N
2
log N

N 1
log 2 +
log 2.
=
2N
2N
2
=

183

K166
Observe that
Ek = {x : f (x) = k}

is the union of a finite set of disjoint intervals (of the form [a, b)) of
total length (9/10)k1 /10. Thus (writing IA for the indicator function
of A)

!
X
X
kIEk + X 1
I Ek
fX =
1kX

1kX

is Riemann integrable with

!

Z 1
X
X
(9/10)k1 .
k(9/10)k1 + X 1 91
fX (x) dx = 101
0

1kX

1kX

Now writing [X] for the integer part of X,

!
X
(9/10)k1 = X(9/10)[X] ([X] + 1)(9/10)[X] 0
0 X 1 101
1kX

and using the Taylor expansion

(1 t)

(j + 1)tj

j=0

(Newtons binomial expansion) we have

Z 1
fX (x) dx 101 (1 9/10)2 = 10.
0

It makes no difference if we replace f by g with g(x) = f (x) except

at points where f (x) = . However we have to work a little harder to
show that gX is Riemann integrable. One way is the observe that
SM

gX (x) fX (x) = 0

for all x k=1 Ek so, given any > 0, we know that gX (x)fX (x) = 0
on the union of a finite set of disjoint intervals of total length 1 /2
and so
IfX X = I fX X I gX I gX I fX + X = IfX + X.

Since is arbitrary gX is integrable and IgX = IfX .

184

K167
(ii) Observe that if > 0 is fixed, we can find an R0 () such that
Z
Z b
fR,S (x) dx ab f (x) dx
a

for all R, S R0 (), and if R and S are fixed, we can find an 0 (R, S) >
0 such that
Z
Z b
b
fR,S (x) dx
a f (x) dx
a

for all 0 < < 0 (R, S).

(iii) If (A) is true, the argument of (i) shows that, if we write f + (x) =
max(f (x), 0) and f (x) = min(f (x), 0), then (A) remains true for both
f+ and f . There is this no loss of generality in assuming that f is
positive. The argument of (ii) now applies.
(iv) The substitution t = x1 suggests
f (t) =

sin t1
.
t

185

K168
(Note that, since (x, t) 7 etx is continuous on [0, X] [a, b], the
apparent singularity at x = 0 must be a trick of the notation.)
Z X ax
Z XZ b
e
ebx
etx dt dx
dx =
x
0
a
0
Z bZ X
=
etx dx dt
a
0
Z b
1 etX
=
dt.
t
a
Now
etX
eaX
0

t
t
for t [a, b] so

Z b tX
e
b
aX
0
0
dt e
log
t
a
a
as X 0 so
Z b
Z b tX
Z X ax
1
e
e
ebx
b
dx =
dt
dt log
x
t
a
a t
a
0
as x .

186

K169
By the fundamental theorem of the calculus

Z t Z d
Z d
d
f (u, v) dv du =
f (t, v) dv.
dt a
c
c
By differentiating under the integral sign and using the fundamental
theorem of the calculus

Z t

Z d
Z d Z t
Z d
d

f (u, v) du dv =
f (u, v) du dv =
f (t, v) dv.
dt c
a
a
c
c t
Thus
Z t Z d

Z d Z t
d
f (u, v) dv du
f (u, v) du dv = 0,
dt
a
c
c
a
and by the constant value theorem

Z t Z d
Z d Z t
f (u, v) dv) du
f (u, v) du dv
a
c
c
a

Z a Z d
Z d Z a
=
f (u, v) dv du
f (u, v) du dv = 0.
a

Z It is weaker because we say nothing about the existence or value of

f (x) dA.
R

187

K170
(i) We have
Z

for all x and

Z

so
Z

1
0

f (x, y) dx = 0
0

f (x, y) dy =
0

2 if 21 < y < 1,
0 otherwise,

f (x, y) dx dy = 0 6= 1 =

1
0

f (x, y) dy dx.
0

(ii) We have

for all x and

Z

Z
1

g(x, y) dy =
0

g(x, y) dx = 0
0

2u(2y 1) if 21 < y < 1,

0
otherwise,

so

1
0

g(x, y) dx dy = 0 6= 1 =

1
0

g(x, y) dy dx.
0

Theorem 9.20 demands continuity everywhere. (The underlying problem is that g is not bounded.)
(iii) We have
Z

1
0

1+y 2

y(w 2y 2 )
dw
w3
y2
1+y2

1
y
y3
=
+ 2
2
w w y2
y
=
2(1 + y 2 )2

1
h(x, y) dx =
2

so that
Z

1
0

1
0

1
h(x, y) dx dy =
4(1 + y 2 )

Observe that h(y, x) = h(x, y).

1
0

= 1/8.

188

K171
(ii) Observe that, if R 1, then
Z
Z
Z R

f (x, y) dx
f (x, y) dx
F

(1 +
+ y2)
tan1 R
=A
1 + y2
A( 2 tan1 R)

x2 )(1

dx

Thus given > 0 we can find an R such that

Z

Z R

/4
f
(x,
y)
dx

f
(x,
y)
dx

0
0
RR
for all y. Since 0 f (x, y) dx is a continuous function of y it follows
that given any y0 we can find a > 0 such that |y0 y| < implies

Z R
Z R

/2

f
(x,
y
)
dx
f
(x,
y)
dx

and so

Z
Z

f (x, y0 ) dx < .
f (x, y) dx

0
0
R
If follows that 0 f (x, y) dx is a continuous function of y. (This can
be done better using uniform convergence.)
R
Since 0 f (x, y) dx is continuous in y and
Z

A/2 ,
f
(x,
y)
dx

1 + y2
0
RR
comparison shows that 0 0 f (x, y) dx dy exists. A similar but simRRR
pler argument shows that 0 0 f (x, y) dx exists Equation F shows
that
Z Z

Z Z R

( tan1 R) A .
f
(x,
y)
dx
dy

f
(x,
y)
dx
dy
2

2
0
0
0
0
We also know that
Z Z R
Z

f (x, y) dx dy

so that
Z Z

f (x, y) dx dy

Similarly
Z Z
Z

f (x, y) dy dx

R
0

R
0

Z
Z

R
0

R
0

A
f (x, y) dx dy ( 2 tan1 R)
2

f (x, y) dx dy ( 2 tan1 R)A.

f (x, y) dy dx ( 2 tan1 R)A,

189

so, using Theorem 9.20 (Fubini on products of intervals),

Z Z

Z Z

2( tan1 R)A.
f
(x,
y)
dy
dx

f
(x,
y)
dx
dy
2

Allowing R gives the result.

190

K172
Just use the rectangles [0, 1] [0, N 1 ] and [0, 1] [N 1 , 1].
F2 (0) is not defined because x 7 f (x, 0) is not Riemann integrable.

191

K173*
No comments

192

K174
(iv) Observe that U is open, [2, 2] U and U does not have Riemann length so [3, 3] \ U is closed, bounded and does not have Riemann length.

193

K175*
No comments.

194

K176
(i) Observe that f (tj ) f (b). Recall that an increasing sequence
bounded above tends to limit.
(ii) Chose n(j) and xj so that n(1) = 1, x1 = tn(1) , n(2j) > n(2j 1),
x2j = sn(2j) > x2j1 , n(2j + 1) > n(2j), x2j+1 = tn(2j)+1 > x2j [j 1].
Consider the limit of f (xj ).
Let a = 0, b = 2, f (t) = 0 for t < 1, f (t) = 1 for t 1 tj = 1 j 1 ,
sj = 1.
(vii) If x
/ E {a}, then f (x) = f (x+). But f (x) f (x+) so
f (x) = f (x+). Thus f = f.
If x < y then, choosing x < xn < y < yn < b with xn , yn strictly
decreasing with xn x, yn y, we have f (xn ) f (yn ) so f(x)
f(y). A similar but simpler argument applies when y = b. Thus
f is increasing. If x [a, b) we can find xn x with xn strictly
decreasing and xn
/ E {a}. Thus f(xn ) = f (xn ) f(x) and f is
right continuous.

(ix) Observe that g (n)1 = 0 0 and g ((2n+ 12 ))1 = 1 1.

195

K177
(iii) The function f is Riemann integrable with respect to H if and
only if f (t) f (0) as t 0.
Observe that, if D is a dissection, then

SH (f, D) sH (f, D) SH (f, D {0}) sH (f, D {0})

= lim sup( sup f (x) inf f (x))
t0, t<0 x(t,0]

x(t,0]

so if I (f, H) = I (f, H) then f (t) f (0) as t 0.

Conversely if f (t) f (0) as t 0, then, taking DN = {N 1 , 0}

we have
SH (f, DN ), sH (f, DN ) f (0).

196

K178
(i) By adding a constant, we may suppose that u(t) 0 as t
We know that u is continuous except at a countable
P set of points xj and
and that, writing j = u(xj ) supt<xj u(t), j=1 j converges (indeed

X
j=1

j lim f (t).
t

Without loss of generality we may suppose 1 2 . . . . Define

un (t) so that
n
X
H(x xj ).
u(t) = un (t) +
j=1

We observe that un (t) un+1 (t) 0 so un (t) U (t) for some U :

R R. Since un is increasing for each n, U is increasing. We observe
that
0 inf un (t) un (s) n
s<x<t

and
0 un+1 (t) un+1 (s) un (t) un (s)

for any fixed t and s with t > s, so

inf U (t) U (s) = 0

s<x<t

and U is continuous.
A very much simpler argument shows that
for all x.

Pn

j=1

H(xxj ) converges

(ii) Suppose that f and g are bounded increasing left continuous

functions with 0 f (t) g(t) for all t and g is Riemann-Stieljes
integrable with respect to G. Then
I (f, G) I (f, G) ( lim G(t) lim G(t)).
t

By (i), we can write

f (t) = v(t) +

X
j=1

xj )
j H(t

where H(t)
= 0 for t 0, H(t)
= 1 for t > 0, j > 0 and
j=1 j
converges. By the first paragraph we need only show that the function
fN given by
fN (t) = v(t) +

N
X
j=1

xj )
j H(t

197

is Riemann-Stieljes integrable. We know that v is Riemann-Stieljes

xj ) is a Riemannintegrable so we need only show that t 7 H(t
Stieljes integrable function and this follows much the same pattern as
in K177.
(iii) f will be Riemann-Stieljes integrable if and only if no discontinuity of f coincides with that of G.
If f and G share a discontinuity x0 then
S(f, D) s(f, D) S(f, D {x0 }) s(f, D {x0 })

inf (G(t) G(s)) inf (f (t) f (s)) > 0

s<x0 <t

s<x0 <t

so f is not Riemann-Stieljes integrable with respect to G.

If f and G have no points of discontinuity in common, the argument
of (ii) will work if appropriately modified.

198

K179
(ii) The condition is G strictly increasing.
If G is not strictly increasing, choose a < b such that G(a) = G(b).
Take f (x) = max(0, 14(ba)1 |x(a+b)/2| to see that the theorem
fails.
Suppose G is strictly increasing. If f is bounded continuous positive
function with f (x0 ) 6= 0, observe that there exists an > 0 such that
|f (t) f (x0 )| f (x0 )/2 so f (t) f (x0 /2 for |x0 t| 2. Now
Z
(G(x0 + ) G(x0 ))f (x0 )
f (x) dG(x)
> 0.
2
R
(iii) Same condition as (ii).
Suppose G is strictly increasing. If f (x0 ) 6= 0 there exists an > 0
such that |f (t)f (x0 )| |f (x0 )|/2 for |x0 t| . Choose g continuous
so that g vanishes outside (x0 , x0 + ), g(t)f (x0 ) 0 for all t and
g(X0 )f (x0 ) = 1. By part (ii), g(t)f (t) = 0 for all t which is impossible.
Thus f is identically zero.

199

K180
(i) Observe that
n
n
X
X
|g(xj ) g(xj1 )| =
g(xj ) g(xj1 ) = g(b) g(a).
j=1

j=1

(ii) Observe that

n
n
X
X
|f (xj ) f (xj1 )|
|(f + g)(xj ) (f + g)(xj1 )|
j=1

j=1

n
X
j=1

|g(xj ) g(xj1 )|.

(iii) Observe that

|f (x)| |f (a)| + |f (b) f (x)| + |f (x) f (a)|.

(iv) Observe that, setting x0 = 0, xN +2 = 1 and xj = (N + 4 j +

1
1
) , we have
2
N
X
j=1

as N .

|f (xj ) f (xj1 )|

N
+5
X
j=5

j 1

200

K181
(ii) Given > 0, we can find a = x0 y0 x1 . . . xn yn = t
such that
n
X
F (yj ) F (xj ) F+ (t)
j=1

and a =

x00

y00

0
= t such that
x01 . . . xm ym
n
X
j=1

F (x0j ) F (yj0 ) F (t) .

Let a = z0 z1 . . . zN = t with

{z0 , z1 , z2 . . . zN } ={x0 , x1 , x2 . . . xn } {y0 , y1 , y2 . . . yn }

0
}
{x00 , x01 , x02 . . . x0m } {y00 , y10 , y20 . . . ym

Then
F (t) F (a) =

N
X
j=1

F (zj ) F (zj1 )
X

F (zj )F (zj1 )>0

F+ (t)

F (zj ) F (zj1 ) +

F (zj )F (zj1 )>0

F (zj )F (zj1 )<0

F (zj ) F (zj1 ),

F (zj ) F (zj1 ) F+ (t)

and
F (t)

F (zj )F (zj1 )>0

F (zj ) F (zj1 ) F (t)

so |F (t)F+ (t)+F+ (t)| 2. Since is arbitrary, F (t) = F+ (t)F (t).

The proof that VF = F+ (t) + F (t) is similar.
(iii) Observe that, if a = x0 y0 x1 . . . xn yn = t then
G+ (t) =

n
X

G+ (yj ) G+ (xj ) +

j=1

n
X

j=1
n
X

j=1

G+ (xj ) G+ (yj1 )

G+ (yj ) G+ (xj )

(f (yj ) f (xj )) + (G (yj ) G (xj ))

j=1
n
X
j=1

n
X

f (yj ) f (xj )

201

so G+ (t) F+ (t). Exactly the same calculation gives G+ (t) G (s)

F+ (t) F (s) for t s so G+ F+ is increasing.

202

K182
(ii) If F is right continuous at t then, since F = F (a) + F+ F ,
either both F+ and F are right continuous or neither is (which is
impossible by (i)).
(iii) Write
G+ (t) =

t
0

max(F (x), 0) dx, G (t) =

min(F 0 (x), 0) dx.

a

Then, by the fundamental theorem of the calculus, G+ is differentiable

with G0+ (t) = max(F 0 (t), 0) so G+ is increasing. Similarly G is decreasing. Further G0+ (t) G0 (t) = F 0 (t), so using the fundamental
theorem of the calculus, F = F (a) + G+ G .
Now let > 0. Since F 0 is continuous on [a, t], it is uniformly continuous so we can find an integer N such that
|F 0 (x) F 0 (y)| < whenever |x y| (t a)/N .

Let us an integer r with 1 r n good if there exists an x

[(r 1)/N, rN ] with |F 0 (x)| > . Let us call the other integers r with
1 r n bad. Observe that, if r is good, then F 0 is single signed on
x [(r 1)/N, rN ]. Thus, using the mean value theorem
N
X
X Z xj
|F (xj ) F (xj1 )|
|F 0 (x)| dx
j=1

j good

so that
VF (t)

xj1

|F 0 (x)| dx

a
t
a

j bad

(t a)/N

|F 0 (x)| dx (t a)

t
a

|F 0 (x)| dx = G+ (t) + G (t)

and F+ (t) = G+ (t), F (t) = G (t) as stated.

203

K183
Observe that f, is well behaved away from 0, so this exercise is
about behaviour near 0. The behaviour is different according as > 0,
= 0 or < 0.
(A) If > 0, we have
|x| sin(|x| ) = |x|+ + |x|+2 (|x|)

with (|x|) 0 as x 0.

If + > 1, f, is differentiable at 0 with derivative 0. We check

0
that f,
is continuous at 0 so f, is everywhere differentiable with
continuous derivative (so of bounded variation).
If 1 + > 0, f, is not differentiable at 0 but is continuous
0
there (so everywhere continuous). By inspection of f,
, or otherwise,
f, is of bounded variation.
If + = 0, f, is not continuous at 0 but is of bounded variation.
If 0 > + then f, (x) so f, is not continuous and not of
bounded variation.
(B) If = 0. f,0 = |x| sin 1. The function is differentiable everywhere with continuous derivative if and only if > 1, and continuous
everywhere if and only if > 0, of bounded variation if and only if
0.
(C) If < 0 then the function differentiable at 0 so everywhere if
and only if > 1. Derivative is then 0. Now
0
f,
(x) = x1 sin x + x+1 cos x

so derivative is continuous if and only if (consider x = (n/2)1 ) +

1 > 0.
The function is continuous at 0 (and so everywhere) if and only if
> 0.
If 0, inspection shows that f, is not of bounded variation. If
> 0 then by differentiating (or rescaling) we see that in each interval
In = [((n + 1))1/ , (n)1/ ] the function f, either increases and then
decreases or decreases and then increases. The total variation Vn of
f, in the interval In is thus 2 suptIn |f, (t)|. It follows that

2(n)/ Vn 2((n + 1))/

P
/
so f, is of bounded variation if and only if
converges, i.e.
n=1 n
/ < 1, i.e. > , i.e. + > 0.

204

K184*
No comments.

205

K185
(i) Observe that, in a self explanatory notation,
S(f, D, G + F ) = S(f, D, F ) + S(f, D, G)

and

S(f, D1 D2 , G + F ) = S(f, D1 D2 , F ) + S(f, D1 D2 , G)

S(f, D1 , F ) + S(f, D2 , G)

so

I (f, F + G) = I (f, F ) + I (f, G)

and similarly I (f, F + G) = I (f, F ) + I (f, G).
(ii) Rewrite as
Z
Z
Z
Z
f (x) dF1 (x) + f (x) dG2 (x) =
f (x) dF2 (x) + f (x) dG1 (x).
R

and observe that by part (i)

Z
Z
Z
f (x) dF1 (x) + f (x) dG2 (x) =
f (x) d(F1 + G2 )(x)
R

and

f (x) dF2 (x) +

f (x) dG1 (x) =

f (x) d(F2 + G1 )(x)

206

K186*
No comments.

207

K187
The main problem is the behaviour of (y) for y close to 1.
Observe that writing t = 1 s we have, using appropriate Taylor
theorems,
(1 (1 s)2 )1/2 = (2s s)1/2

= 21/2 s1/2 (1 s/2)1/2

21/2 s1/2 (1 (s)s)

R1
with (s) 0 as 0. Thus y (1t12 )1/2 dt converges and
Z 1y
Z 1
1
1
dt =
ds
2
1/2
(1 (1 s)2 )1/2
y (1 t )
0

= 21/2 (1 y)1/2 + (1 y)3/2 (1 y)

for 1 > y > 0, where (s) 0 as s 0.

This proves the existence of and shows that

sin( u) = 1 u2 /2 + (u)u2

for > u > 0 where (u) 0 as u 0. Thus in (iii) we know that

the extended function sin is differentiable at with sin0 () = 0. If
>u>0
sin0 ( u) = (1 (sin u)2 )
and

= (1 (1 u2 /2 + (u)u2 )2 )1/2 = u + (u)u

sin0 ( + u) = sin0 ( + u) = u (u)u

with (u) 0 as u 0+. Thus sin is twice differentiable at with

sin00 () = 1 = sin .

208

K188
The points at issue are very similar to those of K187.

209

K189
(iv) Observe that
Z R
Z
2

f (x) dx =
1

x2 dx = (1 R1 )

but

R
0

2 1/2

f (x)(1 + f (x) )
1

dx 2

as R .

R
1

x1 dx = 2 log R

(v) We have
Vol K =
but
0

Vol(K \ K) =

Z0

x1 dx =

(x1/2 + x1 )2 (x1/2 )2 dx

(x2 + 2x3/2 ) dx
0

= lim (1 R1 ) + 4(1 R1/2 ) = 5.

=

210

K190
We have
Z

f (x) ds =

=
=
=

f (x) ds
2

f (x) ds
3

f (x) ds

4
Z 1

cos(2t)(2 sin(2t), 2 cos(2t)) dt

Z 1
( cos(2t) sin(2t), cos2 (2t)) dt
= 2
0
Z 1
(sin(4t), 1 + cos(4t)) dt
=
0

= (0, )

Also

and

f (x) ds =

f (x) ds = 2

f (x) ds = (0, )
f (x) ds = (0, 2).

211

K191
(ii) We have
p(n k)
un (k + 1)
=
un k
(1 p)k

so un (k + 1) > un k if p(n k) > (1 p)k, i.e. if np > k and un (k + 1) <

un k if np < k (and, if np = k, then un (k + 1) = un k). We take k0 so
that np k0 1 and np k0 .
(ii) If k n(p + /2) then

p(n k)
p(1 p /2)
un (k + 1)
=

un k
(1 p)k
(1 p)(p + /2)

Pn
Now 1 =
j=0 un (j) un (k) so, writing k1 = k1 (n) for the least
integer greater than n(p + /2), we see that
kk1

p(1 p /2)
un (k)
(1 p)(p + /2)

for all k k1

If k2 = k2 (n) is the greatest integer less than n(p + /2) we have

X
X p(1 p /2) kk1
un (k)
(1 p)(p + /2)
knp+n
kk2
k2 (n)k1 (n)
1

p(1 p /2)
p(1 p /2)
1

0
(1 p)(p + /2)
(1 p)(p + /2)
as n .
Similarly
X

knpn

so

un (k) 0

Pr{|Nn np| n} =
as n .

|knp|n

un (k) 0

(iii) We have
EXj = p1 + (1 p)0 = p
and
EXj2 = p1 + (1 p)0 = p
so var Xj = p p2 = p(1 p).

212

The sum of the expectations is the expectation of the sum so

n
n
X
X
EXj .
Xj =
ENn = E
j=1

j=1

For independent random variables, the sum of variances is the variance

of the sum so
and Thchebychev gives

var Nn = np(1 p)

Pr{|Nn np| n}
as n .

var Nn
p(1 p)
=
0
2
(n)
2 n

213

K192
Recall that (A), (B) and (C) give
d(x, y) = (d(x, y) + d(y, x))/2 d(x, x)/2 = 0.
(i) Set x = y in (C)0 to obtain (B). (C) now follows from (B) and
(C)0 .
(ii) Setting d0 (x, y) = d(x, y), we see that d0 is a metric space. If
d(x, y) 0 for x, y X then X has exactly one point.

214

K193
By (A)0 , x x.
By (B), x y implies y x.
By (C) if x y and y z then

0 = d(x, y) + d(y, z) d(x, z) 0,

so d(x, z) = 0 and x z.

Observe x x0 , y y 0 gives

d(x, y) = d(x0 , x) + d(x, y) + d(y, y 0 ) d(x0 , y 0 )

and d(x0 , y 0 ) d(x, y) so d(x, y) = d(x0 , y 0 ). Thus d is well defined.

215

K194
(i) The result is true with A1 = 1. We use induction on n to prove
it. The result is trivially true for n = 1. Suppose it is true for n = m.
If Sm+1 then writing = (m + 1, (m + 1)) and 0 = 1 we see
that 0 fixes m + 1 so is generated by at most m elements of the form
(ij) (transpositions). Thus = 0 is is generated by at most m + 1
transpositions. Any shuffle of m cards needs at most m swaps.
(ii) Observe that (1i)(1j) = (ij) and use (i).
(iii) Observe that
(123 . . . n)k (12)(123 . . . n)k = (k k + 1)
so any transposition of the form (k k + 1) is a distance at most 2n + 1
from e. Next observe that
(k k + 1)(1 k)(k k + 1) = (1 k + 1)
so any transposition of the form (1 k) is a distance at most n(2n + 1)
from e. Thus, using (ii) and (i), the diameter of Sn is at most 2n2 (2n+1)
with respect to X3 .
(iv) Observe that ar b = bar . Let N be the integer part of (n 1)/2.
If N r then, by induction, the product of r elements of the form a, b,
a1 , must have one of the forms au with |u| r or bav with |v| r 1.
Looking at aN we see that the diameter of Dn is at least N .
(v) Sn has n! elements and Dn has 2n. There appears to be little
relationship between size and diameter.

216

K195
(i) We wish to prove the statement P (N ). If |x| 4k for some
integer k. and
x=

N
X
j=1

xj with |xj | 4n(j) for some integer n(j) 0 [1 j N ],

PN

2n(j) 2k .
P
If x = 1j=1 xj then x = x1 so P (1) is true.

then

j=1

Suppose now that P (r) is true for all 1 r N 1. If |x| 4k

and
N
X

x=

j=1

xj with |xj | 4n(j)

then if n(j) k for any j we are done. Thus we need only consider
the case n(j) k + 1 for all 1 j N . Let M be the smallest m such
that
m
X
|xj | 4k1 .
j=1

We know that
M
X
j=1

so

|xj |

N
X

M +1

M
1
X
j=1

|xj | + 4k1 2 4k1

|xj | |x|

M
X
j=1

|xj | 2 4k1

. Applying the inductive hypothesis to

0

x =

M
X
j=1

00

|xj | and x =

N
X

M +1

|xj |

we see that
N
X
j=1

2n(j) =

M
X
j=1

2n(j) +

N
X

j=M +1

2n(j) 2k1 + 2k1 = 2k .

Thus P (N ) is true and the induction is completed.

(ii) The key point is that d(x, 0) = 0 implies x = 0. But if x 6= 0
then |x| 4k for some k and so, by (i), d(x, 0) 2k .

217

P
(iii) If 4k+1 > |x| 4k then taking x = 1j=1 xj with x1 = x, we
see that 2k+1 d(x, 0). We also know, by (i), that d(x, 0) 2k .
Thus
4|x|2 d(x, 0) 41 |x|2 .
(iv) I think the graph is quite complex with f constant on intervals
of the form [4k (1 ), 4k ] (for some > 0) and on other intervals
which form a dense set (cf the devils staircase in K251).

218

K196
(i) and (iv) are false. Let X = R with usual metric, X1 = {x : x
0}, X2 = {x : x > 0}, f (x) = 0 for x X1 and f (x) = 1 for x X2 .
Parts (ii) and (iii) have one line solutions for the sufficiently sophisticated. Here are longer solutions.
(ii) True. If x X1 X2 then, without loss of generality, we may
suppose x X1 . Thus there exists a 1 > 0 such that, if z X and
d(x, z) < 1 , we have z X1 . Since f |X1 is continuous, we know that,
given > 0 there exists a 2 () > 0 such that d(x, z) < 2 () implies
(f (x), f (z)) < for all z X1 . Now d(x, z) < min(1 , 2 ()) implies
(f (x), f (z)) < for all z X.
(iii) True. If x X \ Xj , the argument of (ii) shows that f is
continuous at x. If x X1 X2 , then given > 0 there exists a
j () > 0 such that d(x, z) < j () implies (f (x), f (z)) < for all
z Xj and d(x, z) < minj=1,2 j () implies (f (x), f (z)) < for all
z X.

Set f (x) = g(x) when kg(x)k < 1, f (x) = kg(x)k1 g(x) otherwise.
Let
X1 = {x X : kg(x)k 1} and X2 = {x X : kg(x)k 1}

and apply part (iii).

219

K197
(i) Consider X = {1, 2} with the discrete metric and A = {1}.
(iii) By (ii), f is continuous. If A, B 6= choose a A, b B and
apply the intermediate value theorem to show that there exists a c with
f (c) = 1/2.
(iv) If A is closed, then {t [0, 1) : (1 t)a + tb A} is closed. If
A is open, then {t [0, 1) : (1 t)a + tb
/ A} is closed.

220

K198
Suppose A satisfies (i). Then, if ym A and ym y, we can find
Thus A F.
xm A with d(xm , ym ) < 1/m so xm y and y A.

But, if F A and F is closed, we have F A Thus A satisfies (ii).

Suppose A satisfies (ii). Then, since A is the intersection of closed
sets, A F and automatically F A whenever F F so A satisfies (iii).
Suppose A satisfies (iii). If xn A and xn x then xn A so since

Suppose x A,
then (writing B(x, 1/n) for the
A is closed x A.
open ball centre x radius 1/n) if B(x, 1/n) A = A \ B(x, 1/n) is a
strictly smaller closed set containing A contradicting our assumption.
Thus B(x, 1/n) A 6= and we can find xn B(x, 1/n) A. We have
xn A and xn x.
(iii0 B is an open set with B B such that, if U is open and
B U , we have B U .
Int Q = , Int Z = , Int{1/n : n 1} = , Int[a, b] = Int(a, b) =
Int[a, b) = (a, b).

Cl Q = R, Cl Z = Z, Cl{1/n : n 1} = {0} {1/n : n 1},

Cl[a, b] = Cl(a, b) = Cl[a, b) = [a, b].
f (A) for every A. If xn x but f (xn ) 9 f (x),
Suppose f (A)
then we can find a > 0 and n(j) such that (f (xn(j) ), f (x)) > .
Now take
A = {xn(j) : j 1}

to obtain a contradiction (x A but f (x)

/ f (A)). Thus f is continuous.
then we can find xn A
Conversely, if f is continuous and x A,

such that xn x and so, by continuity, f (xn ) f (x). Thus f (A)
f (A).
Let X = {1, 2} with d(1, 2) = d(2, 1) = 1 and d(1, 1) = d(2, 2) = 0.
Take y = 1.

221

K199
Observe that Cl U Int(Cl U )) and Cl U is closed so that
Cl U Cl(Int(Cl U )).

However, since U is open, and Cl U U , we have Int(Cl U )) U so

Cl(Int(Cl U )) Cl U

and Cl(Int(Cl U )) = Cl U .

It follows that, If B is any set,

Cl(Int(Cl(Int B))) = Cl(Int B)
and by taking complements, if C is any set
Int(Cl(Int(Cl C))) = Int(Cl C)
Since Cl Cl A = Cl B and Int Int A = Int B, the only possible distinct sets that we can produce are A, Int A, Cl A, Cl Int A, Int Cl A,
Int Cl Int A and Cl Int Cl A.
The set
A = {3} (2, 1] (Q [0, 1]) ((2, 4) \ {3})

shows that these 7 can all be distinct.

222

K200
(i) (C) Take = (max(2, kxk))1 , for example.
(iii) Since is absorbing we can define
n(x) = inf{t 0 : t1 x }.

We note that n(x) 0. Our object is to show that kxk = n(x) defines
the required norm.
If 0 the definition gives n(x) = n(x). The fact that is
symmetric now gives
for all R.

n(x) = ||n(x)

Condition (B) tells us that 0 , and then (D)0 together with

convexity tells us that, if x 6= 0, there exists a > 0 such that tx
/
for all t . Thus n(x) = 0 if and only if x = 0.
Suppose that x, y 6= 0. Then, if > 0,

(1 + )1 n(x)1 x, (1 + )1 n(y)1 y

and so, by convexity,

1
1
1
1
1 n(x) n(y)
(x + y)
(x + y) = (1 + )
(1 + )
n(x) + n(y)
n(x)1 + n(y)1
n(y)1
n(x)1
1
1
=
(1
+
)
n(x)
x
+
(1 + )1 n(y)1 y .
n(x)1 + n(y)1
n(x)1 + n(y)1
Since is arbitrary, n(x + y) n(x) + n(y).
(iv) A necessary and sufficient condition is
B KA

where KA = {Kx : x A } and C is the closed unit ball of the

norm k kC .

223

K201
Observe that
but that

xn = (1, 21 , 31 , . . . , n1 , 0, 0, . . . ) E

xn x = (1, 21 , 31 , . . . , m1 , (m + 1)1 , (m + 2)1 , . . . )

/ E.
If a(j) F and a(j) a, then

|a2m | = |a2m a(j)2m | ka a(j)k 0

as j 0 so a2m = 0 for all m. Thus a F .

Suppose that (V, k k) is a normed space and E is a finite dimensional

subspace. Let e1 , e2 , . . . , eN be a basis for E. Then
N

N
X

xj ej =
|xj |

j=1

j=1

is a norm on E. But all norms on a finite dimensional space are Lipschitz equivalent so there exists K 1 with Kkxk kxk K 1 kxk.
If yn E, y V and kyn yk 0, then yn is a Cauchy sequence in
(V, k k) so a Cauchy sequence in (E, k k) so, by Lipschitz equivalence, a
Cauchy sequence in (E, k k ) so, since (E, k k ) is complete, converges
in (E, k k ) to some z E. Since kyn zk 0, it follows, by
Lipschitz equivalence, that kyn zk 0. By the uniqueness of limits,
y = z E. Thus E is closed.

224

K202
(i) If
0 y + e0 = y + e
with , 0 R and e, e0 E then

(0 )y = e e0

so, applying T to both sides,

(0 )T y = 0

so = 0 and e = e0 .
Also

Tx
y E.
Ty

If T = 0 then N = E.
(ii) Observe that the statement that
B(y, ) N 6=

is equivalent to the statement that there exists an f N such that

ky f k <

and (setting e = f ) this is equivalent to saying that

for some e N and 6= 0.

ky + ek <

(iii) If T is continuous then since the inverse image of a closed set

under a continuous function is closed N is closed.
If N is closed pick a y
/ N and observe that there exists a > 0
such that
ky + ek |

for all and e N .

By part (i), if x E we have

x = y + e

with e N , so, by (ii),

Thus

kxk || whilst T x = T y.
kT xk 1 |T x|kxk

for all x and T is continuous.

(iv) We have N = {a s00 :

j=1

aj = 0}.

225

(a) If we use the norm k k , then N not closed. Take

if j = 1
1
1
ej (n) = n
if 2 j n + 1

0
otherwise.

Then e(n) N , but e (1, 0, 0, . . . )

/ N . Thus N is not closed.
P
1
(b) If we use the norm k kw , N not closed. Observe that
j=2 j
diverges so (since j 1 1) we can find N (n) such that
N (n)

n
Set A(n) =

PN (n)

X
j=2

j 1 n + 1.

j 1 . Take

1
ej (n) = A(n)1 j 1

j=2

if j = 1
if 2 j N (n)
otherwise.

Then e(n) N but e (1, 0, 0, . . . )

/ N . Thus N is not closed.
(c) If we use the norm k k1 , N is closed. If e(n) N and ke(n)
ek1 0 then (remember all sums over a finite number of non-zero
terms)

X X
X

ej (n) = ke(n) ek1 0

ej
ej =

j=1
j=1
j=1
P
so j=1 ej = 0.

(d) If we use the norm k2 , N not closed. Consider the same example
as (a).
(e) If we use the norm k ku then N is closed. We can use much the
same proof as (c).
[To see T not continuous in (a), (b) and (d) we can consider the
norms of e e(n) and T (e e(n)). In (c) and (e) simple inequalities
give kT k 1, indeed kT k = 1.]

226

K203
The key observation is that, since 21/2 is irrational, x + y21/2 =
x + y 0 21/2 for x, y, x0 , y 0 Q if and only if x = x0 and y = y 0 .
0

If pn , qn are positiveintegers with

pn /qn 21/2 (in the usual Eu
clidean metric) then N (pn , qn ) 0 but k(pn , qn )k .
On the other hand

N (x, y) = |x + y21/2 | |x| + 2|y| 2k(x, y)k

for (x, y) Q2 .

227

K204*
No comments.

228

K205
Pick xn Fn . If m n then

d(xn , xm ) diam(Fn ) 0

as n . Thus xn is Cauchy and we can find x X with d(xn , x) 0

as n .
Since
x Fm for each m so
T xn Fm for all n m and Fm isTclosed,

x j=1 Fj . On the other hand, if y j=1 Fj , then since x, y Fn

for each n,
d(x, y) = diam(Fn ) 0
T
so d(x, y) = 0 and x = y. Thus
j=1 Fj contains exactly one point.

(a) Take X = (0, 1] with the usual Euclidean metric d and Fn =

(0, 1/n].
(b) Take X = Z with the discrete metric d (so d(m, n) = 1 for
m 6= n). Let Fn = {m : m n}.
(c) Take Fn = {x : x n}.

229

K206
(i) If is a metric, then (x, y) > 0 for x 6= y so f is strictly
increasing.
If f is strictly increasing, then it is easy to check that is a metric.
(ii) If f is not continuous at x, then without loss of generality, we may
suppose that f is not left continuous and there exist x1 < x2 < . . . with
xj < x and xj x together with a > 0 such that f (x) f (xj )
for all j. Since
N
X
j=1

we have

(xj+1 , xj ) = f (xN +1 ) f (x1 ) < f (x) f (x1 ),

j=1

(xj+1 , xj ) convergent and xj Cauchy with respect to .

If t < x then there exists an N such that t > xN so (t, xj )

(t, xN ) > 0 for j N so (t, xj ) 9 0. If t > x, then (t, xj )
(t, x) > 0 for all j so (t, xj ) 9 0. If t = x, then (t, xj ) > 0
for all j so (t, xj ) 9 0. Thus the xj form a non-convergent Cauchy
sequence for .
If f (t) 9 then a similar argument shows that, if xj = j, the
xj form a non-convergent Cauchy sequence for . A similar argument
covers the case f (t) 9 .
(iii) We must have
g(t) + g(s) = d(t, 0) + d(0, s) d(t, s) = g(t + s)

for all t, s 0,

g(t) + g(s) = d(t, 0) + d(0, s) d(t, s) = g(t s)

for all t s 0 and

g(0) = d(0, 0) = 0.

We check that these conditions are sufficient.

(iv) If 2n1 x 2n , then
2n+1 g(x) 2n=1 g(2n1 ) =

n+1
2X

j=1

g(2n1 ) g

2n+1
X
j=1

2n1

= g(1)

so g(x) 2n1 g(1) x21 g(1). Take K = 21 g(1).

If we take g(x) = x1/2 we see that no relation d(x, y) L|x y| will

hold in this case.
(v) Since g is increasing and g(t) 0, g(t) l as t 0+ for some
l 0. If l > 0 then d(x, y) > l for all x 6= y so any Cauchy sequence
for d is eventually constant and so converges.

230

If l = 0 then, since g is decreasing, given any > 0 we can find and

> 0 such that > g(s) 0 implies |s| < . Thus if xn is Cauchy for
d, it follows that xn is Cauchy for the Euclidean metric so there exists
an x R with |xn x| 0 and so d(xn , x) = g(|xn x|) 0.
(vi) Sufficiency by direct verification. Not necessary. Let be the
discrete metric with (x, y) = 1 for x 6= y. Then g(t) = 0 for t 6= 1,
g(1) = 1 gives a metric but g(1/2) + g(1/2) = 0 and g(1) = 1.

231

K207
(i) Write ej for the vector with 1 in the jth place and 0 elsewhere.
Then
and so

kaj ej k kak
|aj | kakkej k1 .

232

K208
We have
N
X
j=1

N
X

(aj + bj )2

a2j

j=1

a2j

j=1

!1/2
!1/2

N
X

a2j

j=1

j=1

!1/2 2

!1/2 2
,
a2j

so,
sequence bounded above converges we have
Psince an increasing
2
j=1 (aj + bj ) convergent. Moreover

X
j=1

(aj + bj )2

a2j

j=1

!1/2

X
j=1

a2j

!1/2 2

(ii) To show completeness, suppose a(n) a Cauchy sequence in l 2 .

We have
|aj (n) aj (m)| ka(n) a(m)k2
so aj (n) is Cauchy with respect to the usual Euclidean distance and
there exists an aj R such that |aj (n) aj | 0.
Now observe that,
M
!1/2 M
!1/2 M
!1/2
X
X
X
a2j (n)

(aj aj (n))2
+
aj (n)2
j=1

j=1

M
X

M
X

j=1

j=1

j=1

(aj aj (n))2

!1/2

+ ka(n)k2

(aj aj (n))2

!1/2

+ sup ka(r)k2
r1

sup ka(r)k2
r1

as n . (Recall that a Cauchy sequence is bounded.) Thus

for all M and so a l2 .

M
X
j=1

a2j

!1/2

sup ka(r)k2
r1

233

Finally, if n, m N
!1/2
!1/2 M
!1/2 M
M
X
X
X
(a2j (n) aj (m))2
(a2j aj (m))2
+
(a2j aj (n))2

j=1

j=1

j=1

M
X

M
X

j=1

j=1

(a2j aj (m))2

!1/2

+ ka(n) a(m)k2

(a2j aj (m))2

!1/2

+ sup ka(s) a(r)k2

r,,sN

so, allowing m ,
M
!1/2
X
(a2j aj (n))2
sup ka(s) a(r)k2
r,sN

j=1

and, allowing M ,

ka a(n)k2 sup ka(s) a(r)k2

r,sN

for all n N . Thus

as n .

ka a(n)k2 0

(iii) We have (by Cauchy-Schwarz in N dimensions)

!1/2
!1/2 N
N
N
X
X
X
kak2 kbk
a2j
|aj bj |
a2j
j=1

j=1

j=1

so,
Psince an increasing sequence bounded above converges, we have
j=1 aj bj absolutely convergent and so convergent.

234

K209
(ii) we have
log

x y
+
p q

1
1
log x log y
p
q

for all x y > 0. Setting x = X p , y = Y Q , multiplying by 1 and taking

exponentials gives the inequality.
(iii) By (ii),
|f (t)|p |g(t)|q
+
.
|f (t)g(t)|
p
q
Now integrate.

235

K210
(i) Observe that (p 1)q = p, so, taking g(t) = |f (t)|p1 , we have
1/q
Z b
1/q
Z b
Z b
p
(p1)q
p
,
|f (t)| dt
=A
|f (t)|
|f (t)| dt A
a

whence the result.

(ii) We have
Z

b
a

|(f (t) + h(t))g(t)| dt

Z
1/p
b

|f (t)| dt

b
a

|f (t)g(t)| dt +
Z

b
p

|h(t)| dt

|h(t)g(t)| dt

a
1/p ! Z b
a

|g(t)| dt

1/q

Now use (i).

(iii) To see that (C([a, b]), k kp ) is not complete follow the proof that
(C([a, b]), k k1 ) is not complete.
(iv) We work in C([a, b]). kf gk1 kf kp kgkq . If kf gk1 Akgkq for
all g then kf kp A.

236

K211
It may be helpful to look at K209.
(iv) If k k is derived from an inner product, then
Taking

ka + bk2 + ka bk2 = 2(kak2 + (kak2 ).

a = (1, 0, 0, 0 . . . ), b = (0, 1, 0, 0, . . . ),

we see that this fails for k kp unless p = 2.

In the case C([0, 1], k kp ) we look at (for example)
Observe that

f (x) = g(1 x) = max(1 4x, 0).

Z

1
p

f (x) dx =
0

1/4(4t)p dt.
0

(v) We consider C([a, b]). It is easy to check that

kf gk1 kf k1 kgk

and (by setting g = 1) that if

kf gk1 Akgk

for all continuous g then kf k1 1. If f C([a, b]) and there exists

an x [a, b] with |f (x)| A + . Without loss of generality, we may
suppose f (x) A + . By continuity we can find a > 0 such that
f (t) A = /2 for t [a, b] (x , x + ). Choose a non-zero
g C([a, b]) with g(t) = 0 for all t
/ [a, b] (x , x + ). Then
|f gk1 > (A + /2)kgk1 .

Thus Holder and reverse Holder hold for p = 1, q = and p = ,

q = 1.

237

K212
(i) p = 1. A square with sides at /4 to the axes.
p = 2. A disc.
p = A square with sides parallel to the axes.
Other values of p give smooth boundary (as does p = 2).
(ii) If 1 < p < , then the chain rule shows that fp differentiable
except at (0, 0). Since
fp (x, 0)
= sgn x
x
does not tend to a limit as x 0, fp is not differentiable at (0, 0).
If x, y > 0 then f1 (x, y) = x + y so f1 is differentiable on {(x, y) :
x, y > 0}. On the other hand
f1 (h, y) f1 (0, y)
= sgn h
h
does not tend to a limit as h 0 so f1 is not differentiable at (0, y).
Similar arguments show that f1 is differentiable at the points (x, y)
with both x 6= 0 and y 6= 0 and only there.
If x > y 0 then f (x, y) = x so f1 is differentiable on {(x, y) :
x > y 0}. However
(
f (x + h, x) f (x, x)
1
if h > 0,
=
h
1 if h < 0,
so f is not differentiable at (x, x) if x > 0 and a similar argument
shows that f is not differentiable at (0, 0). Similar arguments show
that f is differentiable at the points (x, y) with x 6= y and only there.
(iii) If kxks = 1 then |xj | 1 and |xj |s |xj |r so
1 = kxkss kxkrr

so kxkr 1.
x = (1, 0, 0, . . . , 0).
(iv) We have
n
X
j=1

|xj |r

n
X
j=1

Now set rp = s (so q(s r) = s).

|xj |rp

!1/p

n1/q .

(vi) If s > r we can find an with r 1 > > s1 . If bj (N ) = j for

j N , bj (N ) = 0 otherwise, then kb(N )ks /kb(N )kr as N .

238

If bj = j , then b lr but b
/ ls .
(vii) Argument similar to (iii).
Consider gn = max(1 nx, 0).
Essentially same but there is scale change and we have the formula
(b a)1/r kf ks (b a)1/s kf kr .

239

K213
(ii) First look at maps from (Rn , k k ) to (Rm , k k ) Observe that
n

n
X

kT xk = max
aij xj max
|aij |kxk
1im
1im
j=1

j=1

and that, if we set yj (i) = sgn aij we have

kT y(i)k =
Thus kT k = max1im

Pn

n
X
j=1

j=1

|aij | =

n
X
j=1

|aij |ky(i)k .

|aij |.

Now look at maps from (Rn , k k1 ) to (Rm , k k ). Observe that

n

n
X

kT xk = max
aij xj max
|aij ||xj | max max |aij |kxk1
1im
1im 1jn
1im
j=1

j=1

and that, if we set yj (k) = 1 if j = k, yj (k) = 0, otherwise, then

kT y(k)k = max |aik | = max |aik |ky(k)k1
1im

1im

Thus kT k = max1im max1jn |aij |.

Now look at maps from (Rn , k k1 ) to (Rm , k k1 ). Observe that

n

m X

kT xk1 =
aij xj

i=1

=
=

j=1

n
m X
X

i=1 j=1
n X
m
X
j=1 i=1
n
X
j=1

|xj |

max |
1jn

|aij ||xj |

|aij ||xj |
m
X

i=1
m
X
i=1

|aij |

|aij |kxk1

and that, if we set yk (j) = 1 if k = j, yk (j) = 1, otherwise then

kT y(j)k1 =
Thus kT k = max1jn |

Pm

m
X
i=1

i=1

|aij | =

|aij |.

m
X
i=1

|aij ky(j)k1 .

240

If we look at maps from (Rn , k k ) to (Rm , k k1 ) we have the obvious

bound
m X
n
X
kT k
|aij |
i=1 j=1

but looking at the maps T1

1
1

and T2 with matrices

1 1
1
and
1 1
1

we see that, in the first case the bound is attained but in the second
case it is not since

1 1
x
x+y
=
1 1
y
xy
and |x + y| + |x y| 2 max(|x|, |y|). However, if we really need a
to find kT k we can obtain it as a solution of a linear programming
problem.

241

K214
Suppose a, b X. If > 0, we can find a y E such that
d(a, E) d(a, y) .

Thus

d(b, E) d(b, y) d(a, b) + d(a, y) d(a, b) + d(a, E) + .

Since is arbitrary

d(b, E) d(a, b) + d(a, E)

and, similarly, d(a, E) d(a, b) + d(b, E). Thus |f (a) f (b)| d(a, b)
and f is continuous.
Suppose k K, l L. In the previous paragraph we saw that
so, by definition,

d(k, M ) d(k, l) + d(l, M )

d(k, M ) d(k, l) + 0 (L, M )

so, allowing l to range freely,

d(k, M ) d(k, L) + 0 (L, M )

and so, by definition again,

so

d(k, M ) 0 (K, L) + 0 (L, M )

0 (K, M ) 0 (K, L) + 0 (L, M ).

It follows at once that obeys the triangle inequality. Also is

symmetric and positive. If (E, F ) = 0 then 0 (E, F ) = 0 and, if
e E, we can find fn F such that d(fn , e) 0. Since F is closed
e F so F E. Similarly E F so E = F . Thus is a metric.

The continuous image of a closed bounded set in Rn is closed and

bounded. If |fn (x) f (x)| < then
0 (f (Kn ), F (K)), 0 (f (K), F (Kn )) < .

No. Let K = [, ] and fn (x) = sin nx.

242

K215
The intersection of closed bounded sets is closed and bounded. Choose
xn Kn . Since K1 is closed and bounded, the theorem of Bolzano
Weierstrass says that we can pick n(j) such that xn(j) x for
some x K1 . Since xn(j) Km for j sufficiently large and Km is closed
x Km for all m so x K.
If (K, Kn ) 9 0 then we can find a > 0 such that there exists xn
Kn with d(xn , K) > (recall KN KN +1 ). By the argument of the
first paragraph we can find n(j) and x K with d(xn(j) , x) 0
which is absurd.

243

K216
(i) and (iii). It is helpful to observe that d(e, f ) = ({e}, {f }).
(ii) Take y Y if and only if d(y, E) 2/N . Observe that the set
of possible Y is finite.

244

K217
Since E is closed, r = inf{kz yk : y E} > 0. If 1 > 0 > 0 we
can find y0 E such that
kz y0 k > r(1 0 )

Set x0 = z y0 . Then kx0 k < r(1 + 0 ) and

for all e E.

kx0 ek = kz (y0 + e)k r

If we set x = kx0 k1 x0 k then kxk = 1 and

kx ek kx0 k1 r = (1 + 0 )1 > 1

for all e E if we choose 0 appropriately.

Let x1 be any vector of norm 1. Let E1 be the subspace generated by

x1 . Now proceed inductively choosing xn+1 of norm 1 with kxn+1 ek
1/2 for all e En and taking En+1 be the subspace generated by xn+1
and En .

245

K218
(i) If n 9 0 we can find a > 0 and n(j) with n(j) . If
un(j) (j) = , uk (j) = 0 if k 6= n(j), then u(j) E but the sequence
has no convergent subsequence.
Suppose n 0 and xn E. Set n(0, j) = n(j). Then since
[r , r ] is closed and bounded we can use BolzanoWeierstrass to
find sequences n(r, j) and yr [r , r ] such that
(a) n(r, j) is a subsequence of n(r, j 1) and n(r, j) > n(r 1, j).

(b) xr n(r, j) xr as j 0.

(c) |xj n(r, j) xj | 2r for all 1 j r.

Observe that x E. Now consider xn(r,r) . We have

kx(n(r, r)) xk max max |xj n(r, j) xj |, sup |j |

1jr

jr+1

max(2 , sup |j |) 0
jr+1

so x(n(r, r)) x as r .
P
PN (j)
(ii) If

diverges,
we
can
find
N
(j)
such
that
n
n=1
n=j n > 1.
PN (j)
If uk (r) = k /( n=j n ) for j k N (j), uk (r) = 0 otherwise, then
u(k) E but the sequence has no convergent subsequence.
The proof of the positive result resembles that in (i).

246

K219
(i) Take complements.
(ii) If not we can find xn such that B(xn , 1/n) does not lie in any U .
By the BolzanoWeierstrass property we can find n(j) 0 and x X
such that xn(j) x. Now there must exist a U U with x U . Since
U is open there exists a > 0 such that B(x, ) U . But we can find
a J such that n(J) > 2 1 and kxn(J ) xk < /2 so
B(xn(J) , 1/n(J)) B(x, ) U

which contradicts our initial assumption.

S
(iii) By Lemma 11.22 we can find y1 , y2 , . . . yM such that M
m=1 B(ym , ) =
X and by (ii) we can find Um U with B(ym , ) Um . Thus
SM
m=1 Um = X.

247

K220
(i) Take contrapositives.
(ii) If the space has property (B) then, taking U to be the collection of
S
B(y, y ), we can find y1 , y2 , . . . , yM such that X = M
m=1 B(ym , ym ).
Now set N = max1mM Nm . We have xN
/ B(ym , ym ) for each
1 m M and this gives a contradiction.
(iii) Combine the results of this question with that of K219.

248

K221
k kB is not a norm since k1kA = 0 but 1 6= 0. The rest can be
checked to be norms.
Recall that Lipschitz equivalent metrics are either both complete or
neither complete.
kf k kf kA 2kf k

so k k and k kA are equivalent. Looking at fn (x) = (n2 + (x 12 )2 )1/2

we see that fn is Cauchy. If it has a limit f in the uniform norm then
/ C 1 ([0, 1]).
it is a pointwise limit of fn so f (x) = |x 21 |. But f
kf kD kf kC 2kf kD

so k kC and k kD are equivalent. If fn is Cauchy in k kC then both

fn and fn0 are Cauchy in k k converge uniformly to continuous f and
F and standard theorems tell us that f is differentiable with f 0 = F .
Thus f C 1 ([0, 1]) and kfn f kC 0.
By considering fN (x) = sin N x we see that k kD is not Lipschitz
equivalent to k k or k k1 . By considering fN (x) = xN we see that
k k is not equivalent to k k1 (or observe that one norm is complete
and the other is not).

249

K222
(i) By the inverse function rule
d
1
sin1 x =
dx
(1 x2 )1/2
for |x| < 1. By the binomial expansion (or by some rigorous Taylor
expansion)
n
Y
X
(2r 1) n
1
=
1
+
t
(1 t)1/2
2r
n=1 r=1

for |t| < 1. Thus

Y
n
X
d
(2r 1) 2n
1
sin x = 1 +
x
dx
2r
n=1 r=1

for |x| < 1. But we may integrate term by term within the radius of
convergence so (since sin1 0 = 0)

Y

n
X
X
22n 2n 2n+1
(2r 1) x2n+1
1
sin x = x +
x
=
2r 2n + 1 n=0 2n + 1 n
n=1 r=1

for |x| < 1.

(ii) The power series has radius of convergence 1 so the only other
points to consider are x = 1. Taking logarithms (cf K96) shows that

n
n
Y
1
(2r 1) Y
1
=
n1/2
2r
2r
r=1
r=1
so

n
X
1 Y
(2r 1)2r
2n
+
1
n=1
r=1

converges by comparison with n3/2 so we actually have (by the Weierstrass M-test) uniform convergence on [1, 1] so the power series is
continuous on [1, 1] so, since sin1 is also continuous on [1, 1], we
have equality at x = 1 and x = 1 as well.

250

K223
(i) fn (x) = (n2 + x2 )1/2 (with positive square root) will do.
(ii) Use (i).
(iii) If fn is Cauchy in k k then both fn and fn0 are Cauchy in k k
converge uniformly to continuous f and F but then we know that f is
differentiable with f 0 = F . Thus f C 1 ([0, 1]) and kfn f kC 0.

251

K224
Let hn : [1, 1] R be given

2n+3 (x + 1)

1
hn (x) =

2n+3 x

h(x)

by
for
for
for
for

Let gn : [1, 1] R be given by

gn (x) =

x [1, 1 + 2n3 ],
x [1 + 2n3 , 2n3 ]
x [2n3 , 0]
x [0, 1].
1

hn (t) dt
1

By the fundamental theorem of analysis, gn is differentiable with continuous derivative hn so |gn0 (x)| 1 for all x and |gn (x)| = 1 outside
three intervals of total length 2n1 . We observe that gn (x) 1 |x|
uniformly on [1, 1]
We define kn : R R to be the 1-periodic function with kn (x) =
21 gn (2(x 1)) for x [1, 1] and define fn : [0, 1] R by fn =
n1 kn (nx) for x [0, 1].
Observe that fn A, kfn k 0 as n 0 but I(fn ) 0 and
I(0) = 1. Thus I is not continuous with respect to the uniform norm.
However if gn , g A and kgn gk 0 then gn0 (x) g(x) uniformly,
so (1 (gn0 (x))4 )2 (1 (g 0 (x))4 )2 uniformly, so
(1 (gn0 (x))4 )2 + gn (x)2 (1 (g 0 (x))4 )2 + (x)2

uniformly as n 0 so I(gn ) I(g) and I is continuous with respect

to k k .
(iii) Done in (i).
(iv) Since I is continuous it follows that, if fN f with respect to
kR k then Ifn If , so If = 0 which is impossible. (If If = 0 then
1
f (x)2 dx = 0 so f = 0 (since f is continuous) but I(0) = 1.)
1

252

K225
(i) If fN, (x) = sin 2N x then
Z 1

I(fN, ) =
1 2(2N )4 (cos 2N x)4
0

+ 2 (2N )8 (cos 2N x)8 + 2 (sin 2N x)2 dx

with

= 1 AN 4 + B8 N 8 + C2
A = 2(2)
B = (2)
C = (2)

Also

(cos 2x)4 dx > 0

0
1

(cos 2x)8 dx > 0

0
1

(sin 2x)2 dx > 0.

0

kfN, k = (1 + 2N ).

If we take gN = fN,N 4/3 then kgN k 0, and (when N is sufficiently

large) I(gN ) < 1 = I(0). If we take hN = f1,N 1 , then khN k 0, and
(when N is sufficiently large) I(hN ) > 1 = I(0).
(ii) Observe that g(t) = 6(t 1)(t 2) = 6t2 18t + 12 has zeros
at 1 and 2. Thus f (t) = 2t3 9t2 + 12t has stationary points at 1 and
2 so (from our knowledge of cubics or by looking more closely) has a
unique zero at 0. Setting P (t) = f (t)2 gives a function of the required
type.
(iii) There exists a > 0 such that P (t) > 0 for all 0 < |t| < .
Thus, if |f 0 (t)| < 1/2 , the function h : [0, 1] R defined by
h(x) = P (1 f 0 (x)2 ) + f (x)2 P (1)

is continuous and positive so

J(f ) I(0) =

1
0

h(x) dx 0

with equality only if h(x) = 0 for all x [0, 1] i.e. only if f = 0. Thus
if kf k < 1/2 we have J(f ) 0 with equality if and only if f = 0.
Use the fn of K224.
(iv) This is really no more complicated than finding a smooth function u : R R with 0 < u(x) < 1 for all t and inf xR u(x) = 0,
supxR u(x) = 1.
Set v(x) =

1
2

+ 1 tan1 x and G(s, t) = v(s).

253

K226
(iv) It remains true that d is a metric and that the continuous
functions form a closed subspace of B(E) and that the uniform limit
of continuous functions is continuous since the proofs do not use the
completeness of (Y, ).
Suppose that Y is not complete, so we can find a Cauchy sequence
y(n) in Y which does not converge. Let E = e and fy (e) = y. (Note
B(E) = C = C(E) and the members of C(E) are precisely the fz .)
Then d (fy , fz ) = (y, z) so the fy(n) do not converge. B(E), C(E)
and C(E) are not complete and the general principle of convergence
fails.

254

K227
Observe that, if 0 < v < u, then evx eux x1 0 as x 0 (exponentials beat powers) so we can find an A such that
0 eux x1 Aevx
R
for x 1 so, by comparison, 1 eux x1 dx converges. Also
so, by comparison,
converges.

R1
0

0 eux x1 x1

eux x1 dx converges. Thus

R
0

eux x1 dx

Our theorem on differentiating under a finite integral show that

Z n
Z n
d
ux 1
eux x dx.
e x
dx =
du 1/n
1/n

The argument of our proof on differentiating under an infinite integral

now shows that
Z
Z
d
ux 1
e x
dx =
eux x dx.
du 0
0
Now integrate by parts, to get
Z X
Z X
1 ux X
1
ux
eux x1 dx
+ u
e x dx = u x e

and, taking limits,

0 (u) = u1 (u).
Thus

d
u (u) = 0
du
and, by the constant value theorem,
(u) = A u .
(ii) We have
n (u) =

1
u .
(n 1)!

(iii) The arguments including the proofs of convergence for the appropriate integrals are similar to but simpler than those of (i).
Z
h
i
2
2
0
(u) =
xeux ex /2 dx = eux ex /2
u(u) = u(u)

so

and (u) = Aeu

2 /2

d
2
(u)eu /2 = 0
du
for some constant A.

255

(iv) To obtain (i) make the substitution s = ux. To obtain (iii) make
the substitution y = x u.

256

K228
Since f (n+1) g uniformly on the closed interval with end points a
and x, we have
Z x
g(t) dt = g(x) g(a).
a

Since g is continuous, the fundamental theorem of analysis gives

g(x) = g 0 (x)
so

d x
e g(x) = 0
dx
so, by the constant value theorem, g(x) = Cex for some C.
No, f (x) = 1 + ex has the property.

257

K229
(i) Since exponentials beat polynomials, fn (x) 0 for x 6= 0. But
fn (0) = 0 0, so fn 0 pointwise for all , > 0.
(ii) We observe that (if x (0, 1))

fn0 (x) = n x1 (1 x)n1 ( + n)x

so, examining the sign of fn0 , we see that there is a unique maximum
at x = /( + n). Thus

=n
1
0 f (x) f
+n
1 + /n
+ nr

so, remembering that (1 + n1 )n e , we see that fn 0 uniformly

if and only if > .
(iii) We have
Z 1
0

fn (x) dx

2/n

fn (x) dx
1/n
2/n

n (n )
1/n

=n
so

R1
0

2
1
n

2
1
n

dx

fn (x) dx 9 0 if + 1.

Now suppose < + 1. Choose with

.
1>>
+1
We have
Z 1
Z
fn (x) dx =
0

as n .

fn (x) dx +
0

n x dx +

n
Z 1
n

fn (x) dx
n (1 x)n dx

n(+1)
n
=
+
(1 n )n+1
+1
n+1

n
n(+1)
n(1)
+
(1 n ) (1 n )n
0,
=
+1
n+1

[An alternative approach is to show that, when = + 1,

Z 1
Z 1
> lim sup
fn (x) dx lim inf
fn (x) dx > 0
N

258

and deduce the other cases.]

259

K230
(i) Suppose that gn g uniformly and xn x. Let > 0. By the
continuity of g at x, we can find a > 0 such that |g(t)gn (t)| < /2 for
|x t| < and t [a, b]. Now there exists an N such that |xn x| <
for n N and an M such that kgn gk < /2 for n M . Thus if
n max(N, M )
|gn (xn ) g(x)| |gn (xn ) g(xn )| + |g(xn ) g(x)| < .

If gn 9 g uniformly then we can find a > 0 and n(j) such

that
|gn(j) (xn(j) ) g(xn(j) )| > .

By BolzanoWeierstrass we can find j(k) such that xn(j(k)) y

say. Set yn(j(k)) = xn(j(k)) , yr = y otherwise. Then yr y but, since
|gn(j) (yn(j) ) g(y)| |gn(j) (xn(j) ) g(xn(j) )| |g(xn(j) ) g(y)|

we have gr (yr ) 9 g(y).

|g(xn(j) ) g(y)|

(ii) Only if fails if g is not assumed continuous. Let [a, b] = [1, 1]

and gn (x) = g(x) = H(x) with H(x) = 0 for x 0 and H(x) = 1 for
x > 0. Then gn g uniformly but g(1/n) 9 g(0).

However, the if part still works. Suppose gn 9 g uniformly as

before and define > 0 as before. Now define n(j) and m(j) inductively
as follows. Set m(0) = 0. For each j 1 choose n(j) > m(j 1) and
xn(j) such that
|gn(j) (xn(j) ) g(xn(j) )| > .

Now we know that gn (xn(j) ) g(xn(j) ) so we can find m(j) > n(j)
such that
|gm(j) (xn(j) ) g(xn(j) )| < /2,

and so, in particular

|gn(j) (xn(j) ) gm(j) (xn(j) )| > /2.

By BolzanoWeierstrass we can find j(k) such that xn(j(k)) y

say. Set
yn(j(k)) = ym(j(k)) = xn(j(k)) ,
and yr = y otherwise. Since
|gn(j) (yn(j) ) gm(j) (ym(j) )| = |gn(j) (xn(j) ) gm(j) (xn(j) )| > /2 > 0

we know that gr (yr ) does not converge.

(ii) The proof only if in (i) still works. However if part fails (even
if gn is continuous). Let (a, b) = (0, 1) and gn (x) = max(1 nx, 0). If

260

x (0, 1), then we can find a > 0 such that [x , x + ] (0, 1).
Since gn 0 uniformly on [x , x + ] we apply part (i) to show
that xn x implies gn (xn ) g(x). However gn does not converge
uniformly on [0, 1].

261

K231
(i) Since f is continuous on [0, 1 ], it attains its bounds so there
exists a y [0, 1 ] with f (y) f (x) 0. Since 0 f (y) < 1, we
have
Z 1
f (x)n dx nf (y)n 0.
n
0

Observe that (if f 0 (1) 6= 0) f 0 (1) > 0 since f (1) > f (1 x) for
x (0, 1). Given any with f 0 (1) > > 0 we can find a 1 > > 0
such that
|f (x) 1 f 0 (0)(1 x)| = |f (x) f (1) f 0 (1)(1 x)| (1 x)
and so
1 (f 0 (1) )(1 x) f (x) 1 (f 0 (1) )(1 x)
for x [1 , 1]. Thus
Z 1
Z
0
n
(1 (f (1) )(1 x)) dx
1

f (x)n
1
1
1

(1 (f 0 (1) )(1 x))n dx.

Now
n

1
n

(1 A(1 x)) dx = n
=

(1 At)n dt

n
(1 (1 A)n+1 ) A1
A(n + 1)

so, using the first paragraph,

Z 1
n
f (x)n dx
0

1
.
f 0 (1)

(ii) Suppose g(y) = 1 for some y (0, 1). By the argument of

Rolles theorem, g 0 (y) = 0, so given any > 0, we can find a > 0
with [y , y + ] (0, 1) such that
g(t) > 1 |y t|
for all t with |y t| . Thus
Z 1
Z y+
n
n
g(x) dx n
(1 |y t|)n dt
0

2n
(1 (1 A)n+1 ) 1 .
(n + 1)

262

Since was arbitrary,

n
as n .

1
0

g(x)n dx

263

K232
Observe that, if is fixed with 1/3 > > 0,
Z 1
Z 1/2 /2
Sm, (x) dx
Sm, (x) dx
1/2 /2

but, if |x| 2 1/2 , then

1/2 /2

1/2 /2

3
1+
4

dx

|Sm, (x)| (1 3)n 0

as m . Thus Tm, (t) 0 uniformly on [1, 2 1/2 ] [2 1/2 , 1] as

m .

(vi) If F : [1, 1] R is continuous, then it is uniformly continuous

and, given any > 0, we can find an integer N such whenever |t s|
N 1 and t, s [1, 1] we have |F (t) F (s)| < /4. Thus if h is the
simplest piecewise linear function on [1, 1] with h(r/N ) = F (r/N )
for N r N we have |h(t) F (t)| < for all t [1, 1].

264

K233
Define F : [1, 1] R by
(
F (x) =

f (x)
if x [0, 1]
f (x) if x [1, 0].

Since F is continuous, we can find polynomials Qn with Qn (t) F (t)

uniformly on [1, 1]. Set Rn (t) = (Qn (t) + Qn (t))/2. Then
|Rn (t) F (t)| = |(Qn (t) + Qn (t))/2 (F (t) + F (t))/2|

|Qn (t) F (t)|/2 + |Qn (t) F (t)|/2 0

uniformly on [1, 1]. Also, since Rn (t) = Rn (t), we can find a polynomial Pn with Pn (t2 ) = Qn (t).
Let g : [1, 1] R be given by g(x) = x. Then

|g(1) Q(12 )| + |g(1) Q((12 ))| = |g(1) Q(1)| + |Q(1)) g(1)|

||g(1) g(1)| = 2.

265

K234
Write Xn (t) = Y1 + Y2 + + Yn where Yj = 1 if the jth trial is
a success and Yj = 0 otherwise. Observe that (since the expectation
of the sum of random variable is the sum of the expectations, and the
variance of the sum of independent random variables is the sum of the
variances)
EXn (t) =
var Xn (t) =

n
X

j=1
n
X
j=1

EYj = nt
var Yj = nt(1 t).

We know that there exists an M such that |f (t)| M for all t

[0, 1] and we know that given > 0 there exists a > 0 such that
|f (t)f (s)| whenever t, s [0, 1] and |ts| . By Tchebychevs
inequality

Xn (t)

Xn (t)
Xn (t)

Pr
t = Pr
E

n
n
n

var(Xn (t)/n)
2
t(1 t)
=
n 2
1
2
n
so, writing IA for the indicator function of A,

Xn (t)
Xn (t)
Xn (t)
Ef
f (t)E
f
f (t) I[t,t+][0,1]
n
n
n

Xn (t)
Xn (t)
+E
f
f (t) I[0,1]\[t,t+]
n
n

Xn (t)

Xn (t)
t + 2M Pr
t >
Pr
n
n
2M
+ 2.
n
Allowing n and observing that is arbitrary gives the required
result.
=

We have
n
n
r
r n
X
X
pn (t) =
Pr(Xn (t) = r) =
f
f
tr (1 t)nr ,
n
n
r
r=0
r=0

so pn is polynomial of degree n.

266

Since

and

n
X
n r
n r
nr
t (1 t)
0,
t (1 t)nr = 1
r
r
r=0
n
X
n r
r=0

r n

tr (1 t)nr = t

we can use Jensens inequality (or we could just quote K144 (iv)).

267

K235
(i) Suppose that fn does not converge uniformly to f . Since the fn (x)
are decreasing this means that there exists a > 0 and xn I such
that fn (xn ) > f (x) + . By the theorem of BolzanoWeierstrass we can
find a subsequence n(j) and an x I such that xn(j) x. Since
fn (x) f (x) we can find an N such that f (x) + /2 > fN (x) Since
f fN is continuous we can find an > 0 such that > fN (y) f (y)
for all y I with |y x| < . Choosing a J such that |xn(J) x| <
we obtain a contradiction.
(ii) False without (a). Let I = [0, 1], fn (1/r) = 1 for all r n,
fn (x) = 0 otherwise.
False without (b). Let I = [0, 1], fn (x) = xn .
False without (c). Witchs hat.
False without (d). Let I = (0, 1), fn = 1/(nx).
(iii) Easy to check n = 0.
If t [1, 1] and 0 pn1 (t) pn (t) |t| then
1
pn+1 (t) pn (t) = pn (t) pn1 (t) + (pn (t)2 pn1 (t)2 )
2

pn (t) + pn (t)
(pn (t) pn1 (t))
= 1
2
(1 |t|)(pn (t) pn1 (t) 0

and

t2 + |t|
|t|(|t| + 1)
=
|t|.
2
2
The result follows by induction.
pn+1 (t)

Since pn (t) is increasing and bounded above, pn (t) g(t) say. We

have
1
0 = pn (t) pn1 (t) (t2 pn1 (t))
2
1 2
1
g(t) g(t) (t g(t)) = (t2 g(t))
2
2
so (since g(t) 0) g(t) = |t|.
Setting f = g, fn = pn we see that part (i) applies and pn (t) |t|
uniformly as n .
(v) By induction pn is a polynomial.
(vi) Can extend to any bounded closed set I in Rn or any metric
space satisfying the BolzanoWeierstrass condition.

268

K236
(ii) Proceed inductively. Write pn (x) = xn . Set s0 = 1. When s0 , s1 ,
. . . , sn1 have been defined set
sn = p n

n1
X
hpn , sj i
j=0

hsj , sj i

(This is the Gramm-Schmidt technique.)

(iii) To prove uniqueness observe that
n
X

bj xj =

n
X

aj x j

j=0

j=0

implies bn = an and use induction. A similar induction proves existence.

P
j
Observe that Q = n1
j=0 cj x .

(iv) Write Pn for the set of polynomials of degree or less. Since the
polynomials are uniformly dense, we have, using (iii),

n
X

infn+1
aj sj f = inf kP f k2
P Pn

aR
j=0

(b a)1/2 inf kP f k 0
P Pn

as n .
(v) We have
k

n
X
j=0

bj s j

f k22

=
=

kf k22
kf k22

+
+

n
X

j=0
n
X
j=0

(b2j ksj k22 2bj hf, sj i)

(bj ksj k2

2
hf, sj iksj k1
2 )

n
X
(hf, sj iksj k1

2 )
j=0

2
2

n
n

X
X
hf, sj i
hf, sj i

2
= f
sj +
ksj k2 bj

ksj k22
ksj k22
j=0
j=0
2

2
n
n

X
X

= f
f (j)sj +
ksj k22 (bj f(j))2

j=0

j=0

The result of the first sentence can now be read off.

The result of the second sentence follows from part (iv).

269

(vi) Observe that sn (t)q(t)w(t) is single signed (that is always positive or always negative) and only zero at finitely many points. Thus
Z b
sn (t)q(t)w(t) dt 6= 0
a

and, by part (iii), q must be a polynomial of degree at least n. Thus

k n and we have exhibited n distinct zeros t1 , t2 , . . . , tn in (a, b).
Since sn can have at most n distinct zeros we are done.

270

K237
Observe (by Leibniz rule or induction or otherwise) that, if n > r
dr
0, then
(x2 1)n has (x2 1) as a factor and so vanishes when
dxr
x = 1. Thus, if n m 0, integrating by parts n times gives
Z 1 n
m
d
2
nd
(x 1)
(x2 1)m dx
n
m
dx
1 dx
1
n1
Z 1 n1
n
m+1
d
d
2
nd
2
n
2
nd

(x 1)
(x 1)
(x 1)
(x2 1)m dx
=
n1
n
n1
m+1
dx
dx
dx
1 dx
1
Z 1 n1
m+1
d
2
nd
(x

1)
(x2 1)m dx
=
n1
m+1
dx
dx
1
= ...
Z 1
dm+n
n
= (1)
(x2 1)n m+n (x2 1)m dx.
dx
1
Thus
Z

1
1

m
dn
2
nd
(x

1)
(x2 1)m dx =
dxn
dxm

0
if n 6= m
R1
2 n
(2n)! 1 (1 x ) dx if n = m.

Making the substitution x = sin , we have

Z 1
Z /2
2 n
(1 x ) dx =
cos2n+1 d.
1

If In =

R /2

/2

In =

/2

cosn d, and n 2 then integration by parts gives

/2

cosn1 cos d
/2

= cos

n1

= (n 1)

sin
Z

/2
/2

/2

+ (n 1)
/2

/2

cosn2 sin2 d
/2

cosn2 (1 cos2 ) d

= (n 1)In2 (n 1)In
so nIn = (n 1)In2 . Since I1 = 2 we have
n = (2n)!I2n+1 = (2n)! 2

n1
Y
j=0

2n 2j
2n+1 (n!)2n
=
.
2n + 1 2j
(2n + 1)!

271

By expanding (x2 1)n and differentiating n times we see that pn

is a polynomial of degree n with leading coefficient (2n)!/(n!). Take
cn = (n!)/(2n)!.
There are several ways of doing the last sentence including integrating by parts along the lines of the first part of the question.

272

K238
P
(i) Observe that P nj=1 P (xj )ej is a polynomial of degree at most
n 1 which vanishes at the n points xk and so must be the zero polynomial.

aj =

ej (t) dt.
1

(ii) Long division (more exactly, induction on the degree k of p)

shows that, if q is a polynomial of degree r, any polynomial p of degree
k can be written as p = sq + r where s has degree at most k s if k s
and is zero, otherwise, and r has degree at most k 1).
P
P
Observe that nj=1 aj Q(xj )pn (xj ) = nj=1 0 = 0.
Q
(iv) Let p(t) = nj=1 (t yj ). Then p is polynomial of degree exactly
n with leading coefficient 1 such that
Z 1
n
X
bj p(yj )q(yj ) = 0
p(t)q(t) dt =
1

j=1

whenever q has degree at most n1 (and so, in particular, when q = pr

with 0 r n 1). Thus p = pn .

273

K239
(i) The argument of K236 shows that sn + sn1 must change sign
at least n 1 times in (a, b). It changes sign only at zeros of odd order
and has exactly n zeros if multiple roots are counted multiply so, by
counting, it must have n simple zeros in (a, b).
(ii) Let the common root be y. We have P (x) = (x y)p(x) and
Q(x) = (x y)q(x). If p(y) = 0, then P has a multiple root at 0 and
we set = 1, = 0. If not then (q(y)/p(y))p q has a root at y and
(q(y)/p(y))P Q has a multiple root at y.
If sn and sn1 have a common root at y then there exist and
both non-zero (since sn and sn1 do not have multiple roots) such that
sn +sn1 and so sn +1 sn1 has multiple roots which is impossible.
(iii) Without loss of generality we may suppose that P has no real
roots in the open interval (x, y). Thus both P and Q are single signed
and, without loss of generality, we may suppose that P and Q are
positive in the open interval (x, y). Since a continuous function on a
closed bounded interval is bounded and attains its bounds we can find
s1 , s2 , s3 [x, y] such that
Q(s1 ) Q(s) Q(s2 ) > 0 and P (s3 ) P (s) 0

for s [x, y].

It follows that the set of real numbers

E = {t P (s3 )/Q(s1 ) : P (s) tQ(s) 0 for some s [x, y]}

is a non-empty and bounded above by P (s3 )/Q(s2 ). Thus E has a

supremum 0 say.We observe that, for each n 1 there exists an wn
[x, y] such that
P (wn ) (0 n1 )Q(wn ) 0.

By the BolzanoWeierstrass theorem there exists a sequence n(j)

and an w [x, y] such that wn(j) w. By continuity
P (w) 0 Q(w) = 0.

We note that this implies w (x, y). Since

P (t) 0 Q(t) 0

for all t [x, y] we see that w is a multiple root. Set = 0 .

Last sentence much as for (ii).

274

K240
Fix M 1 temporarily. Observe that (if |x| < M/2) we have

4
1

(x n)2 n2
P
2
so by Weierstrass M-test N
n=M (x n) converges uniformly. We write
FM,N (x) =

N
X

n=M

1
1
+
.
(x n)2 (x + n)2

Since the uniform limit of continuous functions is continuous, fM is

continuous.

2 2
8

(x n)3 + (x + n)3 n3

0
so by the Weierstrass M test fM,N
converges uniformly on [M/2, M/2]
to limit gM say. Thus fM is differentiable with derivative gM on
(M, M ). Since M was arbitrary we are done.

If y
/ Z we can find a > 0 such that B(y, 2)Z = and an integer
M such that |y| + 2 < M/4. Thus since the sum of two differentiable
functions is differentiable
M
1
X
1
1
+
F (x) = FM (x) +
(x n)2 (x + n)2
n=0

is well defined, continuous and differentiable on B(y, 2). Since y was

chosen arbitrarily, we are done.
(iii) Fix x
/ Z.
0=

N
X

n=N

N
+1
X
1
1

2 F (x) F (x + 1)
2
(x n)
(x
+
1)

n
n=N +1

as N 0, so F (x) = F (x + 1).

275

K241
(i) Since g is continuous on [0, 1] it is bounded, with |g(x)| K say
for all x [0, 1]. Now use periodicity.

If we set f (x) = x1 for x (0, 1] and define f (x) = f (x n)

whenever x n (0, 1] then f is periodic but not bounded. If we set
h(x) = x for all x then h is continuous but not bounded.
(ii) Observe that
1
|g(x)| =
4
and repeat.

x + 1
K
x

g
+ g

2
2
2

(iii) Use LHopital or power series manipulation

2
2
=
(sin x)2
(x (x)3 /3! + 1 (x)x3 )2
1
= 2
x (1 (x)2 /6 + 2 (x)x3 )2

2(x)2
1
3
= 2 1+
+ 3 (x)x
x
6

1
2
+
+ 3 (x)x
x2
3

where j (x) 0 as x 0.
Observe that g(x) g(0) as x 0 so g is continuous at 0. Now use
periodicity.
(iv) Observe that, if 2x
/ Z, then

x
x+1
+g
g
2
2

X
1
1
=
+ x
2 (cosec2 (x/2) cosec2 ((x + 1)/2)
x
1 2
2
( 2 n)
(2 n + 2)
n=

2
2
X
4
2
2 cos (x/2) sin (x/2)
+
(x

2n
+
1)

=
(x 2n)2
cos2 (x/2) sin2 (x/2)
n=
= 4g(x).

The result holds for all x by continuity.

By (i) and (ii) g(x) = 0 for all x. Thus if x
/ Z we have

1
= 2 cosec2 (x)
2
(x

n)
n=

276

and looking at x = 0,

X
1
2
= f1 (0) = 2
.
3
n2
n=1

277

K242
(i) fn0 (x) = x(x2 + n2 )1/2 . Thus

if x (0, 1],
1
0
fn (x) 0
if x = 0,

1 if x [1, 0).

We also have

fn |x| =

uniformly on [1, 1].

1
n2
0
2
2
1/2
|x| + (x + n )
n

1
R x(ii) We could take gn (x) = n max(1 |x (2n) |, 0) and fn (x) =
g (t) dt. fn tends pointwise to a (continuous from below at 0) Heav0 n
iside function.

(iii) Demand uniform convergence of the derivatives.

278

K243*
No comments.

279

K244
Observe that
|t|1/2 v(t) = |t|1/2

v(t) v(0)
0 v 0 (0) = 0.
t

(ii) Standard results (chain rule, product rule etc) show that g is
continuous except possibly at points (x0 , 0). Observe that since u is
non-zero only a bounded closed set we can find a K such that |u(t)| K
for all t. Thus, if y 6= 0
|g(x, y) g(x0 , 0)| = |g(x, y)| K|y 1/2 v(y)| 0

as y 0. Thus g is everywhere continuous.

(iii) Standard results (chain rule, product rule etc) show that g has
continuous partial derivatives on the open set
A = {(x, y) R2 : x, y 6= 0}.

Also we know that g is identically zero on the open set

B = {(x, y) R2 : x2 |y| < 1} {(x, y) R2 : x2 |y| > 3}

and so, trivially, has continuous partial derivatives there. The only
point not in A B is (0, 0). But g(x, 0) = 0 for all x and g(0, y) = 0
for all y so partial derivatives exist at (0, 0)
(iv) If y 6= 0, and |y| 1/9
Z 1
G(y) =
g(x, y) dx
0
Z 1
1/2
= |y|
v(y)
u(x|y|1/2 ) dx
0

= v(y)

y 1/2

u(s) ds = v(y),

and G(0) = 0 = v(0) so G0 (0) = v 0 (0).

Since (x, 0) B for x 6= 0, we have g,2 (x, 0) = 0 for x 6= 0 (and
for the reasons given in the last sentence of (iii), g,2 (0, 0) = 0 as well).
R1
Thus 0 g,2 (x, 0) dx = 0.

We do not have g,2 continuous at the origin. (More vividly, look at

the behaviour of g(x, y) when x is fixed and small and y runs from x2 /9
to x2 .)

280

K245
R
By comparison, 0 |fn (t)| dtR converges and so, since absolute con
fn (t) dt converges. Given > 0, we
vergence implies convergence,
0
R
can find an X such that X g(t) dt /2 and then
Z
Z
Z

|fn (t)| dt
g(t) dt /2.
fn (t) dt

X
X
X
R
Similarly 0 f (t) dt converges and
Z

/2.
f
(t)
dt

Thus

Z
Z

f
(t)
dt
f
(t)
dt

n
n

0
0

Z
Z
Z
Z

f (t) dt
fn (t) dt +

fn (t) dt
fn (t) dt +
X
X
X

ZX

fn (t) f (t) dt +

X

as n 0 and so, since was arbitrary, the result follows.

The proof above works under the conditions of the last sentence.

281

K246
(ii) If follows by comparison and (using Dinis theorem) dominated
convergence (K245).
R
To prove only if, Robserve that ifR any of the 0 fn (t) dt fails to con

verge then so does 0 f (t) dt. If 0 fn (t) dt converges to an then we

can find Xn such that
Z Xn
fn (t) dt an 2n
0

and so

Xn

f (t) dt an 2n
0
R
R
. Thus if 0 f (t) dt converges so do all the 0 fn (t) dt and the values an of the integrals form an increasing sequence bounded above so
converging to a limit A. We have
Z
Z X
Z X
A
fn (t) dt
fn (t) dt
f (t) dt
0
0
0
R
by
part
(i).
Since
X
is
arbitrary
f (t) dt converges. The fact that
0
R
R
fn (t) dt 0 f (t) dt now follows from the if part.
0

282

K247
(i) Observe that
0

2j dj (x, y)
2j
1 + dj (x, y)

so the sum converges and d is well defined. To prove the triangle

inequality, observe that, if d is metric,
d(y, z)
d(x, z)
d(x, y)
+

1 + d(x, y) 1 + d(y, z) 1 + d(x, z)

= A d(x, y)(1 + d(y, z))(1 + d(x, z)) + d(y, z)(1 + d(x, y))(1 + d(x, z))

d(x, z)(1 + d(x, y))(1 + d(y, z))

= A d(x, y) + d(y, z) d(x, z) + d(x, y)d(x, z) + d(y, z)d(x, z)

+ d(x, z)d(x, y)d(y, z)

0
for some A = A(x, y, z) > 0.
(ii) Suppose that xn is Cauchy for d. Then, since 2j d(a, b) dj (a, b),
the sequence xn is Cauchy for dj so there exists a zj X with dj (xj , zj )
0 for each j. We observe that dj+1 (xn , zj+1 ) dj (xn , zj+1 ) so dj (xj , zj+1 )
0 and by the uniqueness of limits zj+1 = zj . Thus
z1 = z2 = = zj = = z, say.

Since

0 d(xn , z)
=

N
X
j=1

N
X

2 dj (xn , z) +

j=1

2j

j=N +1

2j dj (xn , z) + 2N 2N

and N is arbitrary, d(xn , z) 0 as n 0.

(iii) Observe that any Cauchy sequence for (X, d ) is either eventually constant (and then converges to that constant value) or is a
subsequence of 1/n and so converges to 0 . Thus (X, d ) is complete.
However 1/n is a Cauchy sequence in (X, d) which does not converge.
(Observe that d(1/n, 1/m) 1/m when n 2m and d(1/n, 0 ) =
d(1/n, 0 ) = 2.)
Take d1 = d , dj = d otherwise.
(iv) Use the general principle of uniform convergence. If fj is Cauchy
then there is an fj such that f (j) fj . Now use the theorem on
interchanging derivative and limit.

283

(v) Could take

d(f, g) =

X
j=1

2j sup|z|1j 1 |f (z) g(z)|

.
1 + 2j sup|z|1j 1 |f (z) g(z)|

284

K248*
No comments.

285

K249*
No comments

286

K250
(i) Suppose that the result is false for some j. Then we can find
(j)
(j)
gk D such that kgk k 1/k but kgk k 1. Since kgk k 0 but gk
does not converge uniformly to 0 as k we have a contradiction.
(ii) Use the fact that we have a norm.

287

K251*
No comments.

288

K252*
No comments.

289

K253
(v) Observe that, given any > 0, we can find
0 = x0 < y0 < x1 < y1 < < xn1 < yn1 < xn < yn = 1
P
such that f (yj1 ) = f (xj ) for 1 j n and nj=0 (yj xj ) . Thus,
if |f (x)f (y)| K|xy| for all x, y [0, 1], we have 1 = f (1)f (0)
K which is impossible if is chosen sufficiently small.

290

K254
(i) Observe that rn (x) = xn is a strictly increasing, differentiable
function. It thus has a differentiable inverse r1/n whose derivative is
given by the rule for differentiation of inverses
rn1 (x)
rn1 (x)
1
=
=
.
(rn1 )0 (x) = 0 1
rn (rn (x))
rn (nrn1 (x))
nx
(ii) Take nn0 th powers of both sides.
(iii) (f) is trivial but must be checked. rn (x) = (r1 (x))n = xn .
(iv) r (x) = (r (x))1 .

291

K255
The key here is bounding |r (x) r (x)| with , Q. Since
|r (x) r (x)| = r (x)|1 r (x)|.

we have to bound r (x) and |1 r (x)| (again for Q).

If N we have N > 0 so rN (x) 1 so rN (x) r(x) 0
and so
for all 0 < x R.

0 r (x) rN (x) = xN RN

If 0 then r (1) = 1 and r is increasing so

|1 r (x)| max r (R) 1, 1 r (R1 )

for all R1 x R. We obtain a similar formula for 0 and

combining this with the previous paragraph we obtain

F
|r (x) r (x)| RN max r|| (R) 1, 1 r|| (R1 )
whenever N and R x R1 .

Since r1/n (x) 1 as n 0 (proof by, e.g. observing that (1 + )n

1 + n so 1 r1/n (1 + ) 1 + /n for > 0) it follows that r (x) 1
as 0 through values of Q (proof by e.g. observing that if
1/n 0 and x 1 then r1/n (x) r (x) 1).
(ii) Observe that r0 n r0 uniformly on [a, b].

292

K256
(v) Since
f 0 (y) = y 11/n y 11/(n+1) 0

for y 1, f is increasing on [1, ). But f (1) = 0 so f (y) 0 for

y 1. Thus
n(y 1/n 1) (n + 1)(y 1/(n+1) 1) 0

and so, since a decreasing sequence bounded below tends to a limit,

n(y 1/n 1) L.
(iii) If 1/n x 1/(n + 1) we also have
P (x) P (0)
P (1/(n + 1)) P (0)

x
x
n
P (1/(n + 1)) P (0)

n+1
1/(n + 1)
as n .

1L=L

(viii) To deal with 1 > a > 0 repeat argument or use chain rule on
the maps t 7 t and t 7 at .
If La = 0 then Pa is constant, if La > 0 then Pa is strictly increasing
and if La < 0 then Pa is strictly decreasing. It follows that L1 = 0,
La > 0 for a > 1 and La < 0 for a > 0.
(ix) By the chain rule,
d t d t
(a ) = a = La at = La (a )t
dt
dt
so La = La .
1

L2
Choose = L1
.
2 and set e = 2

(x) We step outside the context of the question and observe that
et = exp t so
La = Lelog a = log aLe = log a.

293

K257
(i) Observe that
ex + ex eix + eix
2
2
1 21/2 x
1/2 3
1/2 5
1/2 7
+ e2 x + e2 x + e2 x )
= e
4

cosh x cos x =

with = ei/4 .

(ii) The coefficient of x2n is

n
n
X
1 X 2n
(1)r
=
(1)r
(2r)!(2n

2r)!
(2n)!
2r
r=0
r=0

1 (1 + i)2n + (1 i)2n
=
=
(2n)!
2

0
22n
(2n)!

if n is odd
if n is even.

(iii) Set f (x) = cosh x cos x. Our main problem is to calculate f (n) (x)
and then to bound it. Using Leibnizs rule gives a calculation for f (r) (0)
which resembles the calculation for (ii). Also
n
X
n
(n)
cosh x 2n cosh R
|f (x)|
r
r=0
for |x| R. This gives an estimate for the remainder
(2R)n cosh R
|Rn (f, x)|
(n 1)!
for |x| R (or something similar) so Rn (f, x) 0.

(iv) Solve y (4) y = 0 with y(0) = 1, y 0 (0) = y 00 (0) = y (3) (0) = 0.

Remember to check the radius of convergence of the resulting power
series.

294

K258
P
n
n=0 an z has radius of convergence at least R, we know that
PSince

n
n=0 an z converges absolutely. Thus if |h| < R |z0 | we know that

X
n
|an |(|z0 | + |h|)n .
|z0 |nr |h|r =
|an |
r
n=0
r=0

n
X
X
n=0

Write cn,r = |an | nr |z0 |nr |h|r if r n, cn,r = 0 otherwise. By Fubinis
theorem for sums,
X

cn,r =

n=0 r=0

cn,r

r=0 n=0

so

n
X
X

n nr r
z h
an
an (z0 + h) =
r 0
n=0 r=0
n=0

X
X
n nr r
an
z h
=
r 0
r=0 n=r

X
X
n nr
r
z0
h
=
r
n=r
r=0
n

br h r

r=0

with

br =

X
n+m

m=0

Thus
P

r=0 br z
r

r=0 br h

z0m .

converges for |z| < R |z0 | and

n=0

an (z0 + h)n =

295

P
r
Let 2 = R |z0 |. Since
r=0 br converges we can find an M such
r
that |br | M . It follows that, if |w| /2,

g(z0 + w) g(z0 )
r1
=

b
w

r
1

w
r=2
|w|

X
k=0

|w|M

bk2 |w|k

|w|M 2
2

as |w| 0.

k=0

(w| 1 )k
2k

k=0

2M |w| 0

296

K259

(i) We could set A0 = 4, B0 = 1, C0 = /2, Bn = 2n+1 (n!)2 + 1 +

Pn1 (n)
n1
and take Cn so that sin(n) Cn = 1 for n 1.
r=0 |fr (x)|, An = Bn
Observe that

|fn(r) (x)| 2n1

P
(r)
for all n > r and all x so, by the Weierstrass M-test,
n=r+1 fn (x)
converges uniformly and

fn(r) (x) 2r .

n=r+1

P
P
(r)
Thus
f
(x)
converges
uniformly
for
each
r
and
f
=
n
n=1
n=1 fn is
infinitely differentiable with

X
(r)
f (x) =
fn(r) (x).
n=1

Also

(r)

(0) f

(r)

(0)

r1
X
n=1

|fn(r) (0)|

n=r+1

|fn(r) (0)| (n!)2 .

297

K260
(i) Just observe that

X
n=0

|an z |

X
n=0

|an |(R )n .

if |z| R .

(ii) Since |nan z n | |an z n |, the radius of convergence does not increase.
observe that, if |z| < R and we choose r = (2R+|z0 |)/3, we
PNow

have n=0 an rn convergent so an rn 0 so we can find an M 0 with

|an rn | M for all n and so |an | M rn . Now set = (R + 2|z0 |)/3.
Since r > , n(/r)n 0 as n 0 so there exists a K with |nan |
Kn for all n 0. Thus |nan z n | K(|z|/)n and, by comparison,
P

n
n=0 nan z converges absolutely and so converges.
P
P
n
n1
Thus
nz
n=0 nan z has radius of convergence
PR and so n=1 na
n2
folhas radius of convergence R. The result for n=2 n(n 1)an z
lows.
(iii) and (iv). Observe that

n2
n(n 1)
j

n(n 1)(n 2)!

n!
=

=
j!(n 2 j)!
j!(n j)!

n
.
j

so
n2

X n

n
n
n1
j nj
|(z + h) z nz h| =
z h

j
j=0
n2
X
n

|z|j |h|nj
j
j=0

n2
X
n2
2
n(n 1)|h|
|z|j |h|nj2
j
j=0
n(n 1)(|z| + |h|)n2 .

Suppose > 0 and |z| + |h| < r . By parts (ii) and (i), there exists
an A() independent of h and z such that

X
n=2

n(n 1)|an |(|z| + |h|)n A().

298

Thus
N

!
N
N
X

X
X

n
n
n1
a (z + h)
an z
h
nan z

n=0 n

n=0
n=1

=
an (z + h)n z n nz n1 h

n=0
|h|2

N
X
n=2

n(n 1)|an |(|z| + |h|)n

A()|h|2

and, allowing N ,

X
X
X

n
n1
n
nan z A()|h|2
an z
an (z + h)
h

n=0
n=1
n=0
(v) Divide by h and allow |h| 0.

299

K261
P
j
We try, formally, y =
j=0 aj x obtaining, on substituting in the
Legendre equation,

X
X
X
2
j2
j1
(1 x )
j(j 1)aj x 2x
jaj x + l(l + 1)
aj xj = 0.
j=2

j=1

j=0

Rearranging gives

j(j 1) 2j + l(l + 1) aj + (j + 2)(j + 1)aj+2 xj = 0

j=0

so we must have

(j + 1)(j + 2) l(l + 1) aj = (j + 2)(j + 1)aj+2

that is

aj+2 =

(j + l + 2)(j l)
aj .
(j + 2)(j + 1)

This gives
y(x) = a0

Qr
X

2k)(l + 2k + 1) 2r
x
(2r)!
Qr
X
k=0 (l 2k 1)(l + 2k + 2) 2r+1
x
+ a1
(2r + 1)!
r=0
k=0 (l

r=0

where a0 and a1 can be chosen freely. When l is not a positive integer

we observe that
(j + l + 2)(j l)
(1 + (l + 2)j 1 )(1 j 1 )
=
1
(j + 2)(j + 1)
(1 + 2j 1 )(1 + j 1 )
so careful application of the ratio test to the sums of even and odd
powers shows that the power series has radius of convergence 1 and
justifies our formal manipulation for |x| < 1.
If l is positive integer we see that one of the two infinite sums in
the first paragraph actually terminates to give a polynomial of degree
l. The other does not terminate and the argument above shows it has
radius of convergence 1,
If v(x) = (x2 1)n then v 0 (x) = 2nx(x2 1)n1 so
(x2 1)v 0 (x) = 2nxv(x)

as required. Using Leibnizs rule we have

dn+1
(1 x2 )v 0 (x)
dx
= (1 x2 )v (n+2) (x) 2(n + 1)xv (n+1) (x) + n(n + 1)v (n) (x)

300

and
dn+1
2nxv 0 (x) = 2nxv (n+1) (x) + 2n(n + 1)v (n) (x).
dx
Thus differentiating the final equation of the first sentence of this paragraph n + 1 times gives
(1 x2 )vn00 (x) 2xvn0 (x) + n(n + 1)v(x) = 0

as required. Observe that vn is the pn of K237.

301

K262
(this is a special case of Bessels equation.)
P
j
We try, formally, w =
j=0 aj z obtaining, on substituting in the
given equation,
z

X
j=2

j(j 1)aj z

j2

+z

X
j=1

jaj z

j1

+ (z 1)

aj z j = 0

j=0

that is to say
a0 +

j=2

Thus a0 = 0 and

(j 1)(j + 1)aj aj2 z j = 0

aj =

aj2
(j 1)(j + 1)
j

(1)
for j 2. It follows that a2j = 0 for all j 0 and a2j+1 = a1 22j (j!)
2 (j+1) .
Thus

X
(1)j
w(z) = a1
z 2j+1
2j
2
2 (j!) (j + 1)
j=0

Since
1
0
(j 1)(j + 1)
the ratio test shows that our power series has infinite radius of convergence so our solution is valid everywhere.
We note that we can only choose one constant a1 freely. Observe
that setting z = 0 in the original equation shows that automatically
w(0) = 0.
P
j
Let us try and find a solution w =
j=0 bj z for the equation
z2

d2 w
dw
+
z
+ (z 2 1)w = z 2 .
dz 2
dz

We obtain
b0 + (3b2 b0 1)z

j=3

(j 1)(j + 1)bj bj2 z j = 0

so, equating coefficients, b0 = 0, b2 = 1/3 and

bj =

bj2
(j 1)(j + 1)

302

for j 3. Thus
w(z) = b1 W (z) +

X
j=1

with
W (z) =

X
j=0

(1)j
z 2j
Qj1
2
(2j + 1) k=1 (2k + 1)

(1)j
z 2j+1 .
22j (j!)2 (j + 1)

(We can talk about a particular integral plus a complementary function.) As before the solution is valid everywhere.
P
j
If we try and find a solution w =
j=0 cj z for
d2 w
dw
+z
+ (z 2 1)w = z,
2
dz
dz
looking at the coefficient of z gives 0 = 1 which is impossible.
z2

303

K263
By Lemma 11.81 we have (provided |x| < 1)
(1 + x)1/2 =

an x n

n=0

with
n1

1 Y 1
an = (1)
( j).
n! j=0 2
n

We observe that (if z 6= 0)

| 12 n||z|
|an+1 z n+1 |
=
|z|
|an z n |
n
P
n
so, by the ratio test,
n=0 an z has radius of convergence 1. Since
power series can be multiplied within their circles of convergence

an z

n=0

!2

cn z n

n=0

for |z| > 1 with

1+x=

c n xn

n=0

for all |x| < 1 and x real. By the uniqueness of power series (or we
could use K268 below) we have c0 = c1 = 1 and cj = 0 otherwise so

an z

n=0

!2

=1+z

has radius of convergence infinity.

Take K > 1. Observe that (1 + z)1/2 has power series radius of
convergence 1 (proof much as above). Since (1 + K 1 z)1/2 has power
series has radius of convergence K the function (1+z)1/2 (1+K 1 z)1/2
has power series radius of convergence at least 1. Since power series
are continuous within their circle of convergence and
(1 z)1/2 (1 + K 1 z)1/2

304

P
n
1/2
(1 +
is not continuous at 1, the power series
n=0 cn z of (1 z)
1 1/2
K z) has radius of convergence exactly 1. Now

X
X
n
cn z n = (1 + z)1/2 (1 + z)1/2 (1 + K 1 z)1/2
an z
|
n=0

n=0

= (1 + K 1 z)1/2

X
an K n z n
=
n=0

for |z| < 1. So we have an example of two power series of radius

of convergence 1 whose product has radius of convergence K for any
K > 1. An example with radius of convergence infinity was given in the
first paragraph, An example with radius of convergence 1 is given by
considering (1 +P
z)1/2 (1 + z)1/2 . General result obtained by rescaling
j j
(i.e. looking at
j=0 bj R z for appropriate R.
[For enthusiasts. Observe that (if we ignore convergence) to obtain

X
X
X
j
j
uj z
vj z =
wj z j
j=0

j=0

j=0

we merely have to solve

n
X

unj vj = wn

j=0

Thus
u2n = (2n)! and v2n+1 = (2n + 1)! (so
P we j can ensure
P that
j
u
z
and
v
z
certainly
have radius of convergence 0 and
j
j=0 j
Pj=0

j
j=0 wj z is any series we please (so with any radius of convergence
we please).]
Repeating much the same arguments, if L > K > 1, then (1 +
z) (1+K 1 z)1/2 has power series with radius of convergence 1 and (1+
z)(1 + L1 z)1/2 has power series with radius of convergence L and their
product has power series with radius of convergence K. On the other
hand, if K > L > 1 then (1+z)1 (1+L1 z)1/2 (1+K 1 z)1/2 has power
series with radius of convergence 1 and (1 + z)(1 + L1 z)1/2 has power
series with radius of convergence L and their product has power series
with radius of convergence K. If K > 1 then (1 + z)1 (1 + K 1 z)1/2
has power series with radius of convergence 1 and (1 + K 1 z)1/2 has
has power series with radius of convergence K and their product has
power series with radius of convergence K. Remaining cases (infinite
radius of convergence) dealt with similarly.
1

305

K264
If we write
Sn (t) =

n
X

aj (t),

j=1

then
q
X

j aj (t) =

j=p

q
X
j=p

j (Sj (t) Sj1 (t))

= q+1 Sq (t) +

q
X
j=p

(j j+1 )Sj (t) p Sp1 (t)

and so

q
q

X
X

(j j+1 )K + p K = 2p K
j aj (t) q+1 K +

j=p

j=p

for all P
q p 1 and so, by the general principle of uniform convergence,
j=1 j aj (t) converges uniformly.

Observe that, since xn is a decreasing sequence,

q
q
q

X
X
X

j
p
j
j
bj x 2x sup
bj x 2 sup
bj x 0

qp
qp
j=p
j=p
j=p
P
j
as p , so by the general principle of uniform convergence,
j=1 bj x
converges uniformly.
We may thus integrate term by term to get
!
Z 1 X

Z 1

X
X
bn
n
n
bn x
dx =
bn x dx =
.
n+1
0
n=0
n=0 0
n=0

P
P
bn xn /(n+1)
If
n=0
n=0 bn /(n+1) converges, then the power series
P
n
has radius of convergence at least 1 so the
Ppowernseries n=0 bn x has
radius of convergence at least 1 and so n=0 bn x converges uniformly
on [0, 1 ] and
!
Z 1 X

1 Z 1

X
X
bn (1 )n+1
n
n
bn x dx =
dx =
bn x
n+1
0
n=0
n=0 0
n=0
for all 1 > > 0. By the result of the last paragraph but one,
P

j+1
/(j + 1) converges uniformly on [0, 1]. The function
j=1 bj x

xj+1
bj
j+1
j=1

306

is thus defined and continuous on [0, 1] so

X
X
bj
bj xj

j+1
j+1
j=1
j=1

as x 1. Thus

as 0+.

1
0

X
n=0

bn x

X
bj
dx
j+1
j=1

bj = (1)j we know,Pby the alternating series test, that

PTaking

j
1
for 0 x < 1.
j=0 bj /(j + 1) converges and
j=0 bj x = (1 + x)
Also
Z 1
1
dx = [log(1 + x)]10 = log 2.
0 1+x
j
Taking bP
series
2j = (1) , b2j+1 = 0 we know,
P by thej alternating

2 1
test, that j=0 bj /(j + 1) converges and
b
x
=
(1
+
x
)
for
j=0 j
0 x < 1. Also
Z 1
1 1
1
dx
=
tan x 0 = .
2
4
0 1+x

Not good methods. Rate of convergence painfully slow.

Observe that log a loga b = log b

X
X
j+1
(1) log2j+1 e =
j=0

j=0

log 2
(1)j+1
=
= 1.
(j + 1) log 2
log 2

307

K265
Can be done in several ways. Here is one. Write pn = Pr(X = n).
Observe that, since pn 0 and
1 tn
= 1 + t + + tn1 n
0
1t
for 1 > t > 0, we have
PM
PN
N
M
j
X
X
1 tj
j=0 pj
j=0 pj t
pj
jpj EX

1t
1t
j=0
j=0
for M N and so
N
X
j=0

pj

1 tj
X (1) X (t)

EX
1t
1t

for all N 0 and all 0 < t < 1. Allowing t 1 we have

N
X
j=0

jpj lim inf

t1

X (1) X (t)
X (1) X (t)
lim sup
EX.
1t
1t
t1

Allowing N gives the result.

308

K266
Just as in K82, if m n,
n

m
Y

(1 + aj (z)) (1 + aj (z))

j=1
j=1
!
!
!
m
n
X
X
|aj (z)| 1
|aj (z)|
exp
exp
j=n+1

j=1

exp

X
j=1

Mj

exp

j=n+1

Mj

as n
QN so, by the general principle of uniform convergence, it follows
that j=1 (1 + aj (z)) converges uniformly.

309

K267
(i) We have
|z 2 n2 | R2 n2

for all |z| R so, by the previous question, SN (z) converges uniformly
on D(R) = {z : |z| < R}. Since the uniform limit of continuous
functions is continuous, S is continuous on D(R) for each R > 0 and
so on all of C.
(ii) Observe that (if z is not an integer)
Q
(N !)2 N
SN (z + 1)
r=N (r (1 + z))
=
Q
SN (z)
(N !)2 N
r=N (r z)

N +1z
1 (z 1)/N
=
1.
N z
1 z/N
Thus S(z + 1) = S(z) for z not an integer. Extend to all z by
continuity.
P
RecallQfrom K82 that, if
1 for all
n=1 |an | converges, and an 6= Q

n then
(1
+
a
)
=
6
0.
Thus,
if
N
>
|z|
we
know
that
n
n=1
n=N (1
Q
N 1
2 2
2 2
z n ) 6= 0 and so S(Z) = 0 if and only if z n=1 (1 z n ) = 0 i.e.
if and only if z is an integer.
=

[We might guess that S(z) = A sin z. Near 0, S(z) behaves like z
so, if the guess is correct, A = 1 .]

310

K268
P
n
(i) Take R > r > 0 As usual we note that
n=0 an r converges,
n
n
so an r 0 and there exists an M such that |an | M r . Thus, if
|z| < r/2,

q
q

X
X

n
n
a
z
=
a
z

n
n

n=0

n=N

|aN z N |

= |z|

|z|N
N

q
X

n=N +1

|aN | |z|

|an z n |
q
X

n=N +1

|an z

|aN | M |z|r N 1

|z| (|aN | 2M |z|r

N 1

nN 1

q
X

(|z|r)nN 1

n=N +1

).

Thus if we set = min(r/2, |aN |rN +1 /4M ) we have

q
|a ||z|N
X

N
n
a
z

n=0
2
for all q N and so

X
|a ||z|N

N
n
.
an z

n=0

!
!

311

K269
(i) Proof of K268 works.
(ii) Exercise 5.91.
(iii) sin x =

X
(1)n x2n+1
n=0

(2n + 1)!

312

K270
(i) Observe that

n!
1
n
1
n
=
+
+
r
(r 1)!(n r)! n r + 1 r
r1

n!
n+1
=
(n + 1) =
.
(r!(n + 1 r)!
r
(ii) Result true for n = 0. If the result is true for n then
(x + y)n+1 = x(x + y)n + (x + y)n y

n
X
n
n
n+1
xr y n+1r + y n+1
+
=x
+
r
r

1
r=1
r
n+1
X
n+1
=
y n+1r
x
r=0

so the result is true by induction.

(iii) Result true for n = 0. If true for n then

Y
r1
n
nr1
n
Y
Y
X
n
(x j)
(x + y q) =
(x + y n)
(y k)
r
q=0
r=0
j=0
k=0

n
n
nr1
nr
r
r
X n Y
X n Y
Y
Y
(y k) +
=
(x j) (y k)
(x j)
r j=0
r j=0
r=0
r=0
k=0
k=0

n
n1
r
nr
n
Y
X
Y
Y
Y
n
n
=
(x j) +
+
(x j) (y k) +
(y k)
r
r

1
r=0
j=0
j=0
k=0
k=0

nr
r1
n+1
Y
X n+1 Y
(x j) (y k)
=
r
r=0
j=0
k=0

so the result is true by induction.

313

K271
(i) Since |w| < 1, we can find a > 1 such that |w| < 1. We can
now find an N 1 such that

x j M + j

= 1 + M j 1 <
j
j

for all |x| M and all j N 1. We observe that

X 1 n1

(x j) (N + M )N = K, say

n=0 n!

j=0
for all |x| M and all n N . Thus

X 1 n1
Y

(x j)w n K(w)n

n=0 n!
j=0

for all |x| M and all n 1 so by the Weierstrass M-test,

n1
X
1 Y
(x j)w n
n! j=0
n=0

converges uniformly. Since the uniform limit of continuous functions

is continuous, f is continuous on [M, M ]. Since N is arbitrary, f is
defined and continuous everywhere. Note that we have shown that the
sum is absolutely convergent.
(ii) By Cauchys lemma (Exercise 5.38) or considering the power
series in w we have

X
cn
f (x)f (y) =
j=0

with

cn =

n Y
r1
X
n
r=0

j=0

(x j)

nr1
Y
k=0

(y k).

Exercise K270 now gives f (x)f (y) = f (x + y).

Q
(iii) If x = 0 then x 0 = 0 and so n1
j=0 (x j) = 0. Thus f (0) = 1.
It follows that f (x)f (x) = f (0) = 1 so f (x) 6= 0. By the intermediate
value theorem, f (x) > 0 for all x.
(iv) Since f (x) > 0, we can define g(x) = log f (x) and, since f is
continuous, g is. Thus by K101, g(x) = ax and f (x) = eax = bx for
some b.
(v) Since b = f (1) = 1 + w, f (x) = (1 + w)x and we are done.

314

K272
(i) We prove that AN |aN | inductively. The result is trivial for
N = 0. If the result is true for N = k 1, we observe that
kak =

k1
k1 X
X

cn,m Pn,m (a1 , a2 , . . . , ak1 )

n=0 m=0

and
kAk =

k1
k1 X
X

Kn+m Pn,m (A1 , A2 , . . . , Ak1 )

n=0 m=0

where the Pn.m are multinomials with all coefficients positive. Thus
k|ak |

k1 X
k1
X

n=0 m=0

k1 X
k1
X

|cn,m |Pn,m (|a1 |, |a2 |, . . . , |ak1 |)

Kn+m Pn,m (A1 , A2 , . . . , Ak1 ) = kAk

n=0 m=0

and the induction can proceed.

(ii) Formally
X

X
dw
=K
(t)n (w)m
dt
n=0 m=0

!
!
X
X
=K
(t)n
(w)m
n=0

m=0

K
.
(1 t)(1 w)

To solve the differential equation we set, informally,

(1 w)dw =
obtaining
w(1 w/2) =

K
dt
1 t

K
log(1 t) + C.

Since w(0) = 0 this gives

w(1 w/2) =

K
log(1 t)

and
w(t) =

1 + (1 k 2 (log(1 t))2 )1/2

.

315

(iii) If we now fix w as in the previous paragraph then Exercise K79

tells us that w has a power series expansion

X
Aj tj
w(t) =
j=0

with radius of convergence > 0. (Note A0 = 0 since w(0) = 0.)

Since w satisfies FF our standard methods show that the P
An satisfy

n
F. By the arguments of (i), |an | An , so, by comparison,
n=0 an t
P
has radius of convergence at least . If we set u(t) = n=0 an tn , then,
since a0 = 0, we have u(0) = 0. By continuity, we can find with
0 < and |u(t)| < for |t| < . The required results can now be
read off.

316

K273
(i) Observe that, since f is continuous on [0, 4], it is uniformly continuous on [0, 4]. Since f is 2 periodic it is thus uniformly continuous
on R. Similarly g is bounded. Thus
Z

|f g(t + u) f g(t)| =
(f (t + u s) f (t s))g(s) ds
2
kgk

sup |f (x) f (y)| 0

|xy||u|

as u 0.
(iii) Observe, for example, that making the substitution x = t s
we have
Z
Z t+
1
1
f (t s)g(s) ds =
f (x)g(t x) dx = g f (t).
f g(t) =
2
2 t
Fubini gives
f (g h)(t) =
=
=
=
=

Z
Z
1
g(s u)h(u) du ds
f (t s)
2

Z Z
1
f (t s)g(s u)h(u) du ds
(2)2
Z Z
1
f (t s)g(s u)h(u) ds du
(2)2
Z Z u
1
f (t u v)g(v)h(u) dv du
(2)2 u
Z
1
f g(t u)h(u) du = (f g) h(t).
2

(iv) If we set G(s, t) = f (t s)g(s), then G has continuous partial

derivative G,2 (s, t) = f 0 (t s)g(s). We may thus apply our theorem
on differentiating under the integral to show that f g is differentiable
and (f g)0 = f 0 g.
(v) We have
Z

|un f (t) f (t)| =

un (s)(f (t s) f (t)) ds
2
Z

/n

=
un (s)(f (t s) f (t)) ds
2 /n

Z /n
1
un (s) ds
sup |f (u) f (v)|
2 /n
|uv|/n
=

sup

|uv|/n

|f (u) f (v)| 0

317

as n .
(vii) Take un to be a three times continuously differentiable function
in (v).

318

K274
(i) Fubini gives
Z

Z
1
1
int
e
f (t s)g(s) ds dt
f[
g(n) =
2
2
Z Z
1
eint f (t s)g(s) ds dt
2
(2)
Z Z
1
eint f (t s)g(s) dt ds
=
(2)2
Z

Z
1
1
ins
in(ts)
=
e
g(s)
f (t s)e
dt ds
2
2
g(n)f(n).
(iii) We have
|f(n) g(n)| = |f[
g(n)| kf gk .
(v) If e f = f for all f then

f(n) = e[
f (n) = e(n)f(n)

for all f and all n. Take f (t) = eimt to see that e(m) = 1 for all m
contradicting the Riemann-Lebesgue lemma.

319

K275
(i) We have
1
f1 (n) =
2

g1 (t)

(eit eit ) int

g1 (n + 1) g1 (n 1)
e
dt =
2i
2i

Thus, using Riemann-Lebesgue,

n
X
g1 (j + 1) g1 (j 1)
Sn (f1 , 0) =
2i
j=n
=

g1 (n + 1) + g1 (n 1) g1 (n + 1) g1 (n 1)
0.
2i

(ii) If we set

f2 (t)

if t/ is not an integer,
sin t
0
g2 (t) = f2 (0)
if t = 2n,

f 0 () if t = (2n + 1),
2

then g2 is automatically continuous except possibly at points n. But,

f2 (t) f2 ()
t

f20 () = g2 ()
g2 (t) =
t
sin(t )

as t so g2 is continuous at . Similarly g2 is continuous at all n.

(iii) Note that f4 CP (R). We have
Z
1

f4 (n) =
f3 (t/2)eint dt
2
Z
Z
1
1 2
i2ns
f3 (s)e
ds
f3 (s)ei2ns dt = f4 (2n).
=
2
2

Observe that f4 obeys the conditions of (iii).

(iv) Translate.

320

K276
(i) Let S(z) = z f (z). We have S 0 (z) = 1 f 0 (z) so |S 0 (z)| 1/2
for |z| and
|T z| |w| + |S(z) S(0)| |w| + |z| sup |S 0 (u)| /2 + /2 =
|u|

for all |z| .

(ii) Observe that, if z, u X,

|T z T u| = |Sz Su| |z u|/2.

Thus T is a contraction mapping and there is a unique z0 X with

T z0 = z0 i.e. with f (z0 ) = w.
(iii) Observe that the continuity of F 0 at 0 gives us a > 0 such that
|F 0 (z) 1| 1/2 for |z| .
(iv) Let F (z) = (g(z) g(0))/g 0 (0) and apply (iii).

(v) No. g = 0 is a counterexample. [However, in more advanced

work, it is shown that this is the only counterexample!]
(vi) No. If g(z) = z 2 , the equation g(z) = rei has two roots r 1/2 ei/2
and r1/2 ei/2 .

321

K277
(i) True. w is the solution.
(ii) False. f2 (z) is real so f2 (z) = i has no solution for > 0.
(iii) and (v) False. For (iii), observe <f3 (z) 0 so f3 (z) = has
no solution for > 0.
(iv) and (vi) True. For (vi), observe that we can find a with 1/6 >
> 0 such that |F 0 (z) 1| 1/3 for |z| 0 . Let X = {z : |z| }.
Set = 0 /3. If |w| write Sz = F (z) 1 and T z = Sz + |z|1/2 w.
Using the mean value inequality, we have
0

|T z| |w| + |Sz| + |z|2 0 /3 + sup |S 0 (u)||z| + |z|/3 0

|u|< 0

whenever |z| 0 and

|T z T u| |Sz Su| + ||z|2 |u|2 |

|z u|/3 + |z u|(|z| + |u|) = 2|z u|/3,

so T restricted to X is a contraction mapping on X. Since X is closed,

we are dealing with a complete metric space and we can find z0 X
with T z0 = z0 i.e. w = F (z0 ) + |z0 |2 .

322

K278
(i) Observe that
f (x)f 00 (x)
f 0 (x) f (x)f 00 (x)
=
+
f 0 (x)
f 0 (x)2
f 0 (x)2
so, by the mean value inequality,
T 0 (x) = 1

|T a T b| |||a b|.

Since R is complete under the usual metric, the contraction mapping

theorem applies.
(ii) Much as in (i),
|T x| = |T x T 0| |x| sup |T 0 (w)| |||x|
|w|a

so, if |x| a, we see, by induction that |T n x| |a| and |T n x|

||n |x| 0.
Observe that
T 0 (x) =

F (x)F 00 (x)
F 0 (x)2

and
F 00 (x) F (x)F 000 (x) 2F (x)(F 00 (x))2
+

,
F 0 (x)
F 0 (x)2
F 0 (x)3
so T (0) = T 0 (0) = 0 and T 00 (0) = F 00 (0)/F 0 (0). Thus

00

F
(0)
2
= |T x T (0) T 0 (0)x 1 T 00 (0)| (x)x2
T x
x
2

0
2F (0)
T 00 (x) =

with (x) 0 as as x 0. Choose so that (x) 1 for |x| .

(iii) We have quadratic convergence (roughly doubling of correct decimal places) with Newton rather than linear convergence (increasing the
number of correct decimal places by the same amount) as is implied by
the contraction mapping theorem. There may be problems if F 0 (0) is
small.

323

K279
Suppose f has a unique fixed point x. Then f (g(x)) = g(f (x)) =
g(x) and, since x is the unique fixed point of f , g(x) = x.
Let X = {1, 0, 1}, f (t) = t, g(t) = t. Then f g = gf = f , f has
only one fixed point but g has three.
Let X = {1, 1}, f (t) = t, g(t) = t. Then f g = gf = f , f has two
fixed points but g has none.

324

K280
We first prove uniqueness. If f (x) = x and f (y) = y then d(x, y) =
d(f (x), f (y)) Kd(x, y) so d(x, y) = 0 and x = y.
For existence choose x0 X and then choose inductively xn+1 X
such that that T xn+1 = xn (this uses surjectivity). We observe that
so

d(xn+1 , xn ) = d(T xn+2 , T xn+1 ) Kd(xn+2 , xn+1 )

K 1 d(xn+1 , xn ) d(xn+2 , xn+1 )

and precisely the same argument as in the contraction mapping theorem shows that xn z for some z X. Now z = T w for some w
and
d(z, w) d(xn , z) + d(xn , w) d(xn , z) + Kd(T xn , T w)
= d(xn , z) + Kd(xn1 , z) 0

as n . Thus d(z, w) = 0 so z = w and T z = z.

325

K281
(i) If T (a) = a and T (b) = b, then kT (a) T (b)k = ka bk so
a = b.
(ii) Uniqueness much as in (i). Existence not implied see e.g. second
paragraph of K281.
(iii) Let E = {1, 1} be a subset of R. The mapping T : E E
given by T x = x preserves length but has no fixed point. The map
T : E E given by T x = x preserve length and has two fixed points.

(The reader may prefer the example of a circle E in R2 . Non-trivial

rotations have no fixed points, reflection in a diameter has two.)
(iv) Observe that, since T decreases distance, it is continuous. The
map
x 7 kx T xk

is thus continuous and, since E is closed and bounded in Rn , thus

attains a minimum at x0 , say. By definition
so T x0 = x0 .

kx0 T x0 k kT x0 T 2 x0 k

(v) Let E be the closure of the set {T n c : n 0}. Then E is is

closed bounded set and T |E is a distance decreasing map from E to
itself. By (iv), T |E and so T has a fixed point.

326

K282
Observe that, if X = [0, ) with the usual metric, the map f given
by x 7 x + 1 is an isometry but we cannot find an isometry g : X 0
such that f g(x) = x.
We now restrict ourselves to bijective isometries. Since the bijective
maps on X form a group under composition we need only show we
have a subgroup.
If f and g are isometries then
d(f g(x), f g(y)) = d(g(x), g(y)) = d(x, y)

so the composition of bijective isometries is a bijective isometry. The

identity map is a bijective isometry.
If f is a bijective isometry, f 1 is well defined and
d(x, y) = d(f f 1 (x), f f 1 (y)) = d(f 1 (x), f 1 (y))

so f 1 is a bijective isometry.

(i) Take X to be the vertices of an isosceles triangle. G(X) is the

group of permutations of the vertices (isomorphic to S3 ) so non Abelian.
(ii) Take X to be the vertices of a triangle all of whose sides have
different length.
(iii) Take X to be a circle.
(iv) Take X to be a regular n-gon and T rotation about the centre
of symmetry through 2/n.

327

K283
(i) Observe that, since (X, d) has the BolzanoWeierstrass property,
we can find a sequence m(j) of strictly positive integers and an
X such that d(am(j) , ) 0. Take a sequence j(r) of strictly
positive integers such that m(j(r + 1)) > 3m(j(r)) + 1. Then setting
n0 (r) = m(j(r + 1)) m(j(r)) we have
d(an0 (r) , a) d(f m(j(r) an(r) , f m(j(r) a) = d(am(j(r+1)) , am(j(r)) )

as r .

d(am(j(r+1)) , ) + d(am(j(r)) , ) 0

Taking n(j) to be subsequence of the n0 (j) with

d(bn(r) , b) 0,

we have d(an(r) , a), d(bn(r) , b) 0. Thus

d(a, b) d(f (a), f (b)) d(f n(r) (a), f n(r) (b)))

d(a, b) + d(an(r) , a) + d(bn(r) , b) d(a, b)

as r so d(a, b) = d(f (a), f (b)).

Examples Y = R, usual metric, g(x) = 2x, h(x) = 1+2x for x 0,

h(x) = 1 + 2x if x < 0.
(ii) Since f (x) = f (y) implies d(x, y) = d(f (x), f (y)) = 0 implies
x = y, f is injective.
To show surjectivity take any a X. As before we can find a
subsequence n(r) with n(r) 2 such that d(an(r) , a) 0. Let br =
an(r)1 . By the BolzanoWeierstrass property we can find b X and
r(j) such that d(br(j) , b) 0. Thus

d(a, f (b)) d(a, an(r(j)) ) + d(f (br(j) , f (b)) = d(a, an(r(j)) ) + d(br(j) , b) 0
as j .

Example, Y = [0, ) with usual metric and f (x) = x + 1.

328

K284
Observe that g f : X X is expansive so, by K283, g f is bijective and an isometry. Thus g is surjective and (since it is expansive)
injective. Thus f is bijective. Since
d(x, y) (f (x), f (y)) d(g f (x), g f (y)) = d(x, y),

f is an isometric bijection so (Y, ) has the BolzanoWeierstrass property and g is an isometric bijection.
If X = Y = {x R2 : kxk = 1} with the usual metric, then, taking
f = g to be rotation through /4, we see that f and g need not be
inverses.
If X = Y = [0, ) with the usual metric and we set f (x) = 2x,
g(x) = x + 1 then f is a distance increasing bijection but not distance
preserving and g is distance preserving but not a bijection.

329

K285
We note that x 7 f (x) and x 7 kf (x)k are continuous maps on a
closed bounded subset of Rn . Thus there exists an M with kf (x)k M
for all x A.
By convexity, g maps A into A. We have
kg(a) g(b)k = (1 )kf (a) f (b)k (1 )ka bk
so g is a contraction mapping and so, since A is closed and we are
dealing with a complete metric space, g has a fixed point z, say. We
observe that
kf (z) zk = kf (z) gzk

kf (x)k + ka0 k

(M + ka0 k).
Thus

inf{kf (a) ak : a A} = 0.
But x 7 kf (x) xk is a continuous map on a closed bounded subset of
Rn so attains its infimum. Thus kf (y) yk = 0 for some y A which
is thus a fixed point.
Let V = R, A = [2, 1] [1, 2] and T x = x. Then A is closed
and bounded and T is distance non-increasing with no fixed point.
Let V = R, A = (0, 1) and T x = x/2. Then A is convex and bounded
and T is distance decreasing with no fixed point.
Let V = A = R and T x = x + 1. Then A is closed and convex and
T is distance non-increasing with no fixed point.
In final paragraph take
A = {q R :
n

n
X
i=1

qi = 1 and qi 0 for 1 i n}

with
and f (q) = q
qj =

n
X
i=1

qi pij .

330

We then have A closed, convex and bounded, f maps A to A and

n

n X
n
n X

X
X

kf (a) f (b)k1 =
|ai bi |pij

(ai bi )pij

j=1 i=1
n
n X
X

i=1 j=1
n
X
i=1

|ai bi |pij

|ai bi | = ka bk1 .

j=1 i=1

331

K286
(i) Note that
|d(x, T x) d(y, T y) d(x, y) + d(T x, T y) 2d(x, y)

so S is continuous. Thus since our space has the BolzanoWeierstrass

property S must attain a minimum at z say. If T z 6= z then

S(T N (z,T z) z) = d T N (z,T z) z, T (T N (z,T z) z)

= d T N (z,T z) z, T N (z,T z) (T z)
< d(z, T z) = S(z)

which is absurd. Thus T z = z.

(ii) Write fn (x) = d(T n x, z). Then fn is continuous (by direct proof
or by composition of continuous functions) so Un = fn1 ((, )) is
open. Since d(T n+1 x, z) = d(T (T n x), T z) d(T n x, z) we have Un
Un+1 .
Observe that F is a closed subset of (X, d) and so inherits the
BolzanoWeierstrass property. If T x U then T x Un for some
n so x Un+1 U . Thus, if F 6= we can apply (i) to T |F : F F
to obtain w F with T w = w contradicting uniqueness.
(iii) Thus U = X and, given x X, we can find an n such that
x Un so d(T m x, z) < for m n. Since was arbitrary, the required
result follows.

332

K287
Observe that if q X then
m
X
i=1

qi pij 0 and

m X
m
X

qi pij =

j=1 i=1

m X
m
X

qi pij = 1.

i=1 j=1

(i) Note all norms on Rm Lipschitz equivalent. X is a closed (e.g.

because the
of closed sets of form Ei = {x : xi 0} and
Pintersection
m
E = {x :
i=1 xi = 1}) and bounded (if x E then kxk 1).
(ii) We have

m X

kT u T vk1 =
(ui vi )pij

j=1

i=1

m
m X
X

j=1 i=1
m X
m
X
i=1 j=1

|ui vi |pij

|ui vi |pij = ku vk1 .

(iii) Show that

(n+1)
pij

m
X

pnik pkj

k=1

and use induction.

(n)

Probabilistically, pij is the probability that, starting at i we are at

j after n steps.
(iv) Repeat the calculation of (ii) but observe that if u 6= v there
must exist I and I 0 such that uI vI and uI 0 vI 0 are non-zero and
have opposite signs. Thus

m
m
X
X

N
|ui vi |pN
(u

v
)p
<

i
i ij
ij

i=1

i=1

for each j.

(vi) T = gives 1 = 2 so 1 = 2 = 1/2. T n q if and only

(2n+1)
(2n)
if q = . Observe p11
= 0 and p12 = 0 for all n so there is no n
with pnij 6= 0 for all i, j.

(vii) T = for all X. We have T n q = q for all q X.

(n)
Observe that p12 = 0 for all n.
(viii) T = has as solutions all X with 1 = 2 and 3 = 4 .

333

Given q X set 1 = 2 = (q1 + q2 )/2 and 3 = 4 = (q3 + q4 )/2.

Then, T q = and (if n 1) T n q = .

334

K288
We have
X

X
i=1 j=1

|xi pij | =

X
i=1 j=1

|xi |pij =

X
i=1

|xi | = kxk1

so Fubinis
P theorem tells us that i=1 xi pij converges for each j. Further i=1 |xi |pij converges and

X
X
X
X

|xi |pij
xi pij

j=1

i=1

j=1 i=1
X

X
i=1 j=1

|xi |pij = kxk1 .

Thus T x is a well defined point of l 1 and kT xk1 kxk1 . (Note that

this only shows that kT k 1.) Since n and n imply
n + n + , it is easy to check that T is linear.
Further, using Fubini again, if q X
X

X
X
qi pij =
qi pij = 1
j=1 i=1

and, trivially,

j=1 qi pij

i=1 j=1

0 so T q X.

(ii) If u, v X and u 6= v then we can find I and I 0 such that

hat uI vI and uI 0 vI 0 are non-zero and have opposite signs. If
k = max{I 0 , I} then as in (iv) of K287, d(T N (k) u, T N (k) v) < d(u, v).
(iii) If T u = u and T v = v then (ii) shows that u = v.
If T h = h and hi 0 for all i then either h = 0 or setting k =
khk1 h we have k X and T k = k. If T = has a solution in X
then the required solutions are with 0. If not, the only solution
is 0.
(iv) We must have I > 0 for some I. Take k = max{I, i}

X
i =
r pN (k) pri I pN (k) pIi > 0.
r=1

(v)(a) We have for i 2

i = 21 i+1 + 14 i + 41 i1 .

so
2i+1 3i + i1 = 0

and n = An + B n with and roots of the auxiliary equation

22 3 + 1 = 0. Thus n = A + B(1/2)n .

335

Treating the previous paragraph as purely exploratory we observe

that n = 2n is a solution.
(b) Proceeding as in (a) we get
i+1 3i + 2i1 = 0

with general solution i = A + B2n . There are no constants A and B

giving X.
(c) Proceeding as in (a) we get
i+1 2i + i1 = 0

with general solution i = A + Bn. There are no constants A and B

giving X.

336

K289
(i) Observe that J > 0 and set h = J1 .
(ii) If e(j)k = 1 if k = j and e(j)k = 0 otherwise, then the e(k)
form a sequence in X with no convergent subsequence. (Observe that
ke(k) e(l)k1 = 2 for k 6= l.) Thus X does not have the Bolzano
Weierstrass property. However (see Exercise K218)
Y = {y : hi |yi | for i 1}

does so, since X is closed, so does XJ = X Y . Observe that

= khk1
1 h XJ .

Also if q XJ then T q X and

X
(T q)j =
qi pij
i=1

hi pij = hj

i=1

so T q XJ . Thus by Exercise K186 T n q for all q XJ .

(iii) In particular T n e(J) . But J is arbitrary.

P
Now if q X and > 0 we can find M such that M
r=1 qr > 1 .
PM
Write u = q r=1 qr e(r). Then

!
M
M
M

X
X

qr + kT n uk1
kT n q k1 T n
qr e(r)
qr + 1

r=1

r=1
r=1
1

M
X
r=1

M
X
r=1

qr kT n e(r) k1 + + kuk1

qr kT n e(r) k1 + 2

as n . Since was arbitrary, kT n q k1 0 as required.

337

K290
For uniqueness. Suppose T x = T y. Then T n x = T n1 T x =
T
T y = T n y so, by the contraction mapping theorem, x = y.
n1

For existence. By the contraction mapping theorem we can find an

x0 with T n x0 = x0 . Now T n (T x0 ) = T (T n x0 ) = T x0 so, by uniqueness,
T x0 = x0 .
In question K282, T n has a fixed point but is not a contraction
mapping.
Desired result trivially true for n = 0. If true for n then

Z t

n+1
n+1

|(T
h)(t) (T
k)(t)| (h(s), s) (k(s), s) ds
a
Z t

|(h(s), s) (k(s), s)| ds

a
Z t

|M (h(s) k(s))| ds
a
Z t
M n+1

kh kk
(s a)n ds
n!
a
n+1
M
kh kk (t a)n+1 .

(n + 1)!
The induction can proceed.
We thus have
kT n h T n kk

1 n
M (b a)n kh kk .
n!

Since
1 n
M (b a)n 0
n!
(observe that un+1 /un 0) we can find an N such that
1
kT N h T N kk kh kk
2
for all h, k C([a, b]). Thus there exists a unique f C([a, b]) with
Z t
f (t) = g(t) +
(f (s), s) ds
un =

and so (if g(t) = c) a unique solution of

f 0 (t) = (f (t), t)
with initial condition f (a) = c.

338

K291
Consider Rn with its usual Euclidean norm and recall that it is
complete. Set
n
X
T x = y, with yi =
aij f (xj ) + bi
j=1

then, using Cauchy-Schwarz inequality and the mean value inequality,

we have

!2 1/2
n
n
X
X
aij (f (uj ) f (vj )
kT u T vk =
j=1

i=1

n
n
X
X

n
n
X
X

a2ij

j=1

= M ku vk
with K 0 and

j=1

i=1

i=1

a2ij

= Kku vk
2

n
X
j=1

K =M

n
X
(f (uj ) f (vj ))2
j=1

n
X
(M |uj vj |2 )
j=1

a2ij

!1/2

n
X

a2ij .

j=1

If K < 1 the contraction mapping theorem applies.

!!1/2

!!1/2

339

K292
A little fiddling around (we could write y dy = dx obtaining, in a
purely exploratory, non-rigorous manner (1 )y 1 = x + A) shows
that, if 0 < < 1, we have y(t) = 0 for all t, and y(t) = 0 for t < s,
y(t) = (1 )1/(1) (t s)1/(1) for t s are solutions. (We need
s 0 to satisfy y(0) = 0).
If 1 then using the mean value theorem

||y1 | |y2 | | = |z|1 ||y1 | |y2 || (max(|y1 |, |y2 |))1

and this Lipschitz condition shows that the obvious solution y(t) = 0
for all t is unique.

340

K293
(i) If u > 0, then f (t, u) 0 for t > 0 so, by continuity, f (u, 0) 0.
Similarly f (t, u) 0 so f (u, 0) 0 Thus f (u, 0) = 0 for u > 0.
Similarly f (0, u) = 0 for u < 0 so f (0, 0) = 0 by continuity.
Observe, just as in the proof of Rolles theorem, that, if y is constant
on [0, c], then y(t) = 0 for t [0, c]. Otherwise, since a continuous function on a closed bounded interval is bounded and attains its bounds, y
must have a maximum or minimum at some (0, c) with y() 6= 0.
Without loss of generality, let be a maximum. Then y() > 0 so,
using the fact that at an interior maximum the derivative is zero,
0 = y 0 () = f (, y()) > 0
which is impossible. Thus y(t) = 0 for t [0, c] for all c > 0 so y(t) = 0
for t 0 and, similarly, for all t. Thus F has at most one solution.
Since y = 0 is a solution it is the unique solution.
(ii) If E = {(x, t) : k(x, t)k } with > 0 then E is closed and the
restriction of f is continuous on each of the three closed sets,
and

E1 = {(x, t) E : x t2 }, E2 = {(x, t) E : |x| t2 }

E3 = {(x, t) E : x t2 }

S
and so the restriction f |E is continuous on E = 3i=1 Ei . Thus f is
continuous at every (x, t) with k(x, t)k > with > 0 and so at every
(x, t) 6= (0, 0) Since |f (t, u)| 2 max(|x|, |t|), f is continuous at (0, 0)
(Lots of other ways to do this, many probably better.)
Now
y1 (t) =

t
2

f (s, s ) ds =
0

t
0

(2s) ds = y0 (t)

for t 0 and a similar calculation applies when t < 0. Thus yn =

(1)n y0 .

341

K294
Suppose, if possible, that y( ) 0 for some > 0. Let
= inf{s 0 : y(s) 0}.

By continuity, y() = 0. By the mean value theorem, we can find an s

with 0 < s < such that
|y(s)| y 0 (s) = s1 (y(s) y(0)) < 0

which is absurd. Thus y(t) > 0 for all t, so y 0 (t) > 0 for all t, so y is
strictly increasing, so y 0 (t) y(0) = 1 so y(t) 1 + t for all t 0.
On (tj , tj+1 ), we have
y 0 (t) y(t) y((tj ) = 2j

so, by the mean value inequality,

(tj+1 tj )2j y(tj+1 ) y(tj )) = 2j

and so

tj+1 tj 2j(1) .

Thus
tr =

r1
X
j=0

for all r and so

tj+1 tj

r1
X
j=0

2j(1)

1
1 21

1
.
1 21

(ii) Observe that, if y 0 = 1 + y 2 , then y 0 y 2 .

R
(iii) Recall that (w log w)1 dw = log log w + a. We have y(t) =
exp(exp t) as a solution.

342

K295
Observe that
y((r + 1)h) y(rh)
y 0 (rh).
h
Observe that
|f (t, u) f (t, v)| sup |f,2 (t, s)||u v| K|u v|
s

so we have a (global) Lipschitz condition.

By the chain rule y 00 exists and
y 00 (t) = f,1 (t, y(t)) + y 0 (t)f,2 (t, y(t)) = f,1 (t, y(t)) + f (t, y(t))f,2 (t, y(t)).
Thus |y 00 (t)| L for all t and so
|y(nh + s) y(nh) y 0 (nh)s|

Ls2
2

whence
|y((n + 1)h) y(nh) f (nh, y(nh))h|

Lh2
2

and
|y((n + 1)h) yn+1 | |y(nh) yn | + |y((n + 1)h) yn+1 y(nh) + yn |

|y(nh) yn | + |y((n + 1)h) y(nh) f (nh, y(nh))h| + |f (nh, y(nh))h y n+1 + yn |

|y(nh) yn | + |y((n + 1)h) y(nh) f (nh, y(nh))h| + |f (nh, y(nh))h f (nh, y n )h|
Lh2
+ Kh|y(nh) yn |
2
Lh2
.
= (1 + Kh)|y(nh) yn | +
2

|y(nh) yn | +

Thus if
|y(nh) yn |
it follows that

Lh
(1 + Kh)n+1 1 ,
2K

Lh2
LhN
n+1
|y((n + 1)h) yn+1 | (1 + Kh)
(1 + Kh)
1 +
2K
2

Lh
(1 + Kh)n+2 1 .

2K
Since the result is true for n = 0 it is true inductively.
In particular
|y(N h) yN |

Lh
(1 + Kh)N +1 1 .
2K

343

K296*
No comments.

344

K297
As in K295,
|y((n + 1)h) y(nh) y 0 (nh)h|

Lh2
,
2

and then, much as in K295,

|y((n + 1)h) yn+1 | (1 + Kh)|y(nh) yn | +

Lh2
+
2

so, by induction,
|y(nh) yn |
Thus

Lh
+
2K Kh

|y(a) yN |

(1 + Kh)n+1 1 .

Lh
+
2K Kh

(eaK 1).

The error due to machine inaccuracy increases directly with the number of computations.
(ii) and (iii) Recall that

1/2
2
A
A
1/2 1/2
B h
+ 2A1/2 B 1/2
+ Bh =
h
h1/2

takes its minimum value 2A1/2 B 1/2 when h = (A/B)1/2 . We want

h = (2/L)1/2 giving minimum error (2L)1/2 K 1 1/2 .

345

K298
(i) Set f (t, y(t)) = g(t), M = L. Either redo calculations with K = 0
or observe that

Z
N
1

a
X
LhN eaK 1

[hN ]
g(rh)
=
|y(a)

y
|

g(t)
dt

0
2
K
n=0

aM h
aLhN
=
2
2

as K 0+. Take a = M a/2.

(ii) We have
Z
Z a
N
1
X
G(rh) =
G(t) dt h
0

n=0

1 + sin 2N t

dt = = kGk .
2
2
2

(iii) No. Consider f (t, u) = G(t) in (ii).

346

K299
(i) The result is trivially true for n = 0. If it is true for n then
(uv)(n+1) (x) = ((uv)0 )(n) (x) = (u0 v + uv 0 )(n) (x)
n
n
X
X
n (n+1r)
n (nr)
(r)
=
u
(x)v (x) +
u
(x)v (r+1) (x)
r
r
r=0
r=0

n
X
n
n
(n+1)
u(n+1r) (x)v (r) (x) + u(x)v (n+1) (x)
+
=u
(x)v(x) +
r

1
r
r=1

n+1
X n+1
=
u(n+1r) (x)v (r) (x)
r
r=0

so the induction can proceed.

(ii) Differentiating we get

x(1 + x2 )1/2 y(x) + (1 + x2 )1/2 y 0 (x) =

1 + x(1 + x2 )1/2
x + (1 + x2 )1/2

so, multiplying through by (1 + x2 )1/2 we have

(1 + x2 )y 0 (x) + xy(x) = 1.
Differentiating n times, using Leibnizs formula, we have
(1 + x2 )y (n+1) (x) + 2nxy (n) (x) + n(n 1)y (n1) (x) + xy (n) (x) + ny (n1) (x) = 0

for n 1. Setting x = 0, we have

y (n+1) (0) + n2 y (n1) (0) = 0

Since y(0) = 0 and y 0 (0) = 1 (by setting x = 0 in the first two displayed
equations in the statement of (ii)) we have y (2n) (0) = 0 and
y (2n+1) (0) = 2n n!.
Set un = y (n) (0)/n!. Then
|u2n+1 x2 | x2
|u2n1 | 2n + 1
which tends to zero if |x| 1 but diverges otherwise. Thus the Taylor
series for y has radius 1.
P
n
Now
term by term with in its circle
j=0 un x may be differentiated
P
n
of convergence so u(x) =
j=0 un x is a solution for the given differential equation with the same value as y at 0. Since the Lipschitz
conditions apply, y(t) = u(t) for |t| < 1 so

X
y (n) (0) n
y(x) =
x
n!
j=0
for |x| < 1.

347

K300
Since K is continuous on the closed bounded subset [0, 1]2 of R2 , it
is bounded so there exists an M with M K(s, t) for (s, t) [0, 1]2 .
If f C([0, 1]), set
(T f )(t) = g(t) +

K(s, t)f (s) ds.

0

We observe that T f C([0, 1]) since K is uniformly continuous (because K is continuous on the closed bounded subset [0, 1]2 ) and so

Z 1

|(T f )(u) (T f )(v)| = (K(u, s) K(v, s))f (s) ds

0

kf k

as u v.

sup

kxyk|uv|

|K(x) K(y)| 0

Further, if || < M 1 , T is a contraction mapping on the complete

normed space (C([0, 1]), k k ) since

Z 1

K(s, t)(f (s) h(s) ds

|(T f )(t) (T h)(t)| = ||
0
Z 1
||
|K(s, t)(f (s) h(s)| ds
0

||M kf hk

so kT f T hk M kf hk . The contraction mapping theorem now

gives the required result.

348

K301
(ii) Observe that

u (t)
d 01
u1 (t)
dt 00
u1 (t)
0
u (t)
1
= u01 (t)
u00 (t)
1

u2 (t)
u02 (t)
u00
2 (t)
u02 (t)
u02 (t)
u00
2 (t)

u3 (t)
u03 (t)

u00
3 (t)

u03 (t) u1 (t)
0
u3 (t) + u00
1 (t)
u00 (t)
u00
3 (t)
1

u2 (t)
u00
2 (t)
u00
2 (t)

u1 (t)
u2 (t)
u3 (t)
0
0
0

u2 (t)
u3 (t)
= 0 + 0 + u1 (t)
u000 (t) u000 (t) u000 (t)
1
2
3

u
(t)
1

u01 (t)
=
a(t)u00 (t) b(t)u0 (t) c(t)u1 (t)
1
1
0

u (t) u0 (t) u0 (t)

2
3
01

0
0

= a(t) u1 (t) u2 (t) u3 (t)

u00 (t) u00 (t) u00 (t)
1

so W (t) = a(t)W (t).

u3 (t) u1 (t)
00
u3 (t) + u01 (t)
u000 (t)
u00
3 (t)
1

u2 (t)
u02 (t)
u000
2 (t)

u3 (t)
0
u3 (t)

u000
3 (t)

u2 (t)
u02 (t)
0
a(t)u00
2 (t) b(t)u2 (t) c(t)u2 (t)

u3 (t)

u03 (t)

00
0
a(t)u3 (t) b(t)u3 (t) c(t)u3 (t)

349

K302
Observe that, since W (a) 6= 0, we have g(a) 6= 0. Similarly g(b) 6= 0.
Thus, if g has no zeros in (a, b), g has no zeros in [a, b]. Thus f /g
is well defined continuous function on [a, b] which is differentiable on
(a, b). By Rolles theorem there exists a c (a, b) with
0
f
f 0 (c)g(c) g 0 (c)f (c)
W (c)
0=
(c) =
=
.
g
(g 0 (c))2
(g 0 (c))2

Thus W (c) = 0 contrary to the hypothesis. Thus there must be a zero

of g between any two zeros of f and, similarly, a zero of f between any
two zeros of g. [Moreover the zeros must be simple, if f (c) = f 0 (c) = 0,
then W (c) = 0.]

350

K303
(i) Without loss of generality suppose that 1 6= 0. Then without
loss of generality we may suppose 1 = 1 and set 2 = . Then

f1 f2
W (f1 , f2 )(x) = det
f1 f20

f1 + f2 f2
= det 0
f1 + f20 f20

0 f2
= 0.
= det
0 f20
However, if g1 (x) = (x 1)2 for x 1 and g2 (x) = g1 (x) (we could
play the same trick with infinitely differentiable functions) then g1 and
g2 are continuously differentiable and their Wronskian vanishes everywhere but they are not linearly dependent.
(ii) By Exercise 12.24 applied to the vectorial equation

d y(t)
v1 (t)
,
=
a1 (t)v(t) a2 (t)y1 (t)
dt v(t)
it follows that if a1 and a2 are continuous (so bounded on [a, b]), the
equation
y 00 (x) + a1 (x)y 0 (x) + a2 y(x) = 0
has exactly one solution with y(a) = A and y 0 (a) = B.
If W (f1 , f2 )(x) never vanishes then setting

1
f1 (x) f2 (x)
a1 (x) = W (f1 , f2 )(x)
det 00
f1 (x) f200 (x)

and

1
f1 (x) f20 (x)
a2 (x) = W (f1 , f2 )(x)
det 00
f1 (x) f200 (x)

we have a1 and a2 continuous. Thus the equation

F

y 00 (x) + a1 (x)y 0 (x) + a2 y(x) = 0

has exactly one solution with y(a) = A and y 0 (a) = B.

But F is equivalent to
W (y, f1 , f2 ) = 0
and so has f1 and f2 as solutions. Since W (f1 , f2 )(a) 6= 0 the equations
1 f1 (a) + 2 f2 (a) = y(a)
1 f10 (a) + 2 f20 (a) = y 0 (a)

351

have a solution, so, if y is a solution, the uniqueness result of the

previous paragraph shows that
y(x) = 1 f1 (x) + 2 f2 (x)
for all x so f1 and f2 are indeed a basis.
Recall from e.g. Lemma 12.8 that if f1 and f2 are solutions of
y 00 (x) + b1 (x)y 0 (x) + b2 y(x) = 0
then if W (f1 , f2 ) vanishes at one point it vanishes everywhere.

352

K304
Observe that
0 = (u00 v + 2v 0 u0 + v 00 u) + p(u0 v + v 0 u) + qvu
= v(u00 + pu0 + qu) + v 00 u + v 0 (2u0 + pu) = v 00 u + v 0 (2u0 + pu)
so, writing w = v 0 we have

w0 (x)u(x) + w(x)(2u0 (x) + p(x)u(x)) = 0

Observe that is linear so, if w is a solution, so is Bw. Thus

Z x
y(x) = u(x)v(x) = u(x) A + B
w(t) dt
0

is a solution. Since it contains two arbitrary constants we would expect

it to be the general solution.
(ii) We try y(x) = v(x)(exp x) obtaining
v 00 (x) = (2x 1)v 0 (x)

so v 0 (x) = B(exp(x2 x)) and

y(x) = A exp(x) + B exp(x)

t
0

exp(x2 x) dx.

(iii) y(x) = (A + Bx) exp x.

(iv) If u is a solution, try y = uv, obtaining
0 = (u000 v + 3v 0 u00 + 3v 00 u0 + v 000 u) + p1 (u00 v + 2u0 v 0 + v 00 u) + p2 (u0 v + v 0 u) + p3 vu
= w00 u + w0 (3u0 + p1 u) + w(3u00 + 2u0 p1 + p2 )
a second order equation for w = v 0 .

353

K305
(i) We have
f = (u001 y1 + 2u01 y10 + u1 y100 + p(u01 y1 + u1 y10 ) + qu1 y1 )
+ (u002 y2 + 2u02 y20 + u2 y200 ) + p(u02 y2 + u2 y20 ) + qu2 y2
= 2u01 y10 + u001 y1 + pu01 y1 + 2u02 y20 + u002 y2 + pu02 y2 .
(ii) If we differentiate , then we get

0 = u001 y1 + u01 y1 + u002 y2 + u02 y2

so becomes

f = u01 y10 + u02 y20 .

In other words, we have the pair of equations

u01 y10 + u02 y20 = f
u01 y1 + u02 y2 = 0
so, since W never vanishes,
u01 = W 1 f y2 , and u02 = W 1 f y1 .
(iii) Performing the integration we get
y(x) = A1 y1 (x) + A2 y2 (x)
Z a
Z
1
+ y1 (x)
W (s)f (s)y2 (s) dx + y2 (x)
x

W 1 (s)f (s)y1 (s) dx

b

and, if y1 and y2 are chosen as in our discussion of Greens function, we

obtain our standard Greens function solution plus the general solution
of the equation with f = 0.
(iv) In the case given, we can take y1 (x) = ex and y2 (x) = ex ,
obtaining W (x) = 2 and
u01 (x) =

ex
ex
ex
x
0
=
+
e

1
and
u
(x)
=
.
2
1 + ex
1 + ex
1 + ex

Thus
u1 (x) = A1 + log(1 + ex ) ex + x and u2 (x) = A2 + log(1 + ex )

and we have the general solution

y(x) = A1 ex + A2 ex + (ex ex ) log(1 + ex ) ex + 1.

(v) Set y = y1 u1 + y2 u2 + y3 u3 subject to
0 = y1 u01 + y20 u02 + y3 u03
0 = y10 u01 + y20 u02 + y30 u03 .

354

We get
f = y100 u1 + y200 u2 + y300 u3 .
Since the Wronskian never vanishes, we can solve the last three equations to find u01 , u02 and u03 . Integrating now gives u1 , u2 , u3 and thus
the solution.

355

K306
Observe that since K is continuous on the closed bounded set S it
is uniformly continuous. Thus
Z 1

|(TK (f ))(s) (TK (f ))(t)| =

K(s, y) K(t, y)f (y) dy
0

kf k

sup

kzwk|st|

|K(z) K(w)| 0

as s t. Thus Tk maps C(I) to C(I). We easily check linearity and

|(TK (f ))(s)| kf k kKk

so TK is continuous and kT k kKk .

(i) True, since kTKn TK k = kTKn K k kKn Kk .

(ii) False. If Kn (x, y) = max(0, 1 n2 (x2 + y 2 )) then kKn k = 1 for

all n but
Z 1/n
|(TKn (f ))(s)|
|f (y)| dy kf k /n
0

so kTKn k n

0.

(iii) True. Since TK TL = TKL , it suffices to show that TK = 0

implies K = 0. Suppose K 6= 0. Then we can find (x, y) S such that
K(x, y) 6= 0. Without loss of generality suppose K(x, y) = m > 0. By
continuity we can find a > 0 such that (s, t) S and |sx|, |ty|
implies K(s, t) m/2. Set f (t) = max(0, 1 2 1 |t y|). Then
TK f (x) > 0 so TK 6= 0.
(iv) False. Let T f (x) = f (1/2) for all [x [0, 1]. Then T is linear
and continuous with kT k 1. We claim that T 6= TK for all K. For
consider fn (x) = max(0, 1 n|x 12 |). T fn = 1 for all n but, if n 2,
Z 1 +1/n
2
|TK fn |
kKk ds = 2kKk /n 0
as n .

1
1/n
2

356

K307
(ii) We apply Exercise 12.24 to the vectorial equation

d y1 (t)
v1 (t)
=
a(t)v1 (t) b(t)y1 (t)
dt v1 (t)
to obtain the existence and uniqueness of y1 .
R1
Rt
(iii) We have (noting that t u(x) dx = 1 u(x) dx)
Z 1
Z t
0
,2 (s, t)f (s) ds
t)f (t) +
H
H,2 (s, t)f (s) ds H(t,
y (t) = H(t, t)f (t) +
t
0
Z t
Z 1
=
H,2 (s, t)f (s) ds +
H,2 (s, t)f (s) ds
0

and

,2 (t, t)f (t) +

y (t) = H,2 (t, t)f (t) +
H,22 (s, t)f (s) ds H
0
Z t
Z t
,22 (s, t)f (s) ds.
H
H,22 (s, t)f (s) ds +
= f (t) +
00

,22 (s, t)f (s) ds

H

Summing gives

y 00 (t) + a(t)y 0 (t) + b(t)y(t)

Z t
= f (t) +
(b(t)H,2 + a(t)H,2 (s, t) + H,22 (s, t))f (s) ds
0
Z t
,2 + a(t)H
,2 (s, t) + H
,22 (s, t))f (s) ds
+
(b(t)H
0

= f (t).

Further
y(0) =

and similarly y(1) = 0.

1
0

1)f (s) ds =
H(s,

0 ds = 0
0

357

K308
We seek a twice continuously differentiable function y(x) = G(x, t)
which is four times differentiable (using left and right derivatives at
end points) on [0, t) and (t, 1] and satisfies y(0) = y 0 (0) = 0,
y (4) (x) k 4 y(x) = 0 for x [0, t)

and y(1) = y 0 (1) = 0,

y (4) (x) k 4 y(x) = 0 for x (t, 1]

whilst
This gives

y 000 (t+) y 000 (t) = 1.

y(x) = a sinh kx + b sin kx + c cosh x + d cos x for x [0, t]

(remember we have continuity at t) and

0 = c + d, 0 = b + d
so
y(x) = A(sinh kx sin kx) + B(cosh kx cos kx) for x [t, 1]

and, using appropriate symmetries

y(x) = C(sinh k(1 x) sin k(1 x)) + D(cosh k(1 x) cos k(1 x)) for x (t, 1].
The continuity of the zeroth, first
condition on the discontinuity for the
in the next paragraph)

M A B C D = 0 0 0 1 so

where

sinh kt sin kt
k cosh kt cos kt

M =
k2 sinh kt sin kt

3
k cosh kt + cos kt

cosh kt cos kt
k sinh kt + sin kt
k2 cosh kt cos kt
k3 sinh kt + sin kt

and second derivatives and the

third gives (subject to the check

A B C D

T
= M 1 0 0 0 1

sinh k(1 t) + sin k(1 t))

k cosh
k(1 t) + cos k(1 t))

k2 sinh k(1 t) sin k(1 t))

k3 cosh k(1 t) cos k(1 t))

cosh k(1 t) cos k(1 t)
k sinh k(1 t) + sin k(1 t)
.

k 2 cosh k(1 t) + cos k(1 t)

3
k sinh k(1 t) sin k(1 t)

We must check that the four functions

sinh kx sin kx , cosh kx + cos kx

and

sinh k(1 x) sin k(1 x) , cosh k(1 x) + cos k(1 x)

are indeed linearly independent. If

A sinh kx sin kx + B cosh kx cos kx

+ C sinh k(1 x) sin k(1 x) + D cosh k(1 x) cos k(1 x) = 0,

then writing

y(x) = A sinh kx sin kx + B cosh kx cos kx

358

we must have and y(1) = y 0 (1) = 0 that is to say

A sinh k sin k + B cosh k cos k = 0

A cosh k cos k + B sinh k + sin k = 0.

But

sinh
k

sin
k
cosh
k

cos
k

det
cosh k cos k
sinh k + sin k

= sinh2 k sin2 k cosh2 k cos2 k

= 1 + cos2 k sin2 k = cos 2k 1.

Thus if k 6= n for some integer n, we have A = B = 0 (and so

C = D = 0).
If k = n then, by direct calculation, the Wronskian of our four
functions (which is a multiple of det M ) vanishes at 1 so det M = 0
everywhere and our method fails. (The method will also fail for simpler
reasons in the case k = 0.) [In fact we are in a situation like the last
example of Chapter 12 and there is no solution.]

359

K309
We seek a continuous function y(x) = G(x, t) which is twice differentiable (using left and right derivatives at end points) on [0, t) and
(t, 1]) and satisfies y(0) = y 0 (0) = 0,
y 00 (x) + 2y 0 (x) + 2 y(x) = 0 for x [0, t)

and

y 00 (x) + 2y 0 (x) + 2 y(x) = 0 for x (t, )

whilst

y 0 (t+) y 0 (t) = 1.

This gives

y(x) = 0 for x [0, t]

and []

y(x) = Ae(xt) sinh((x t)) + Be(xt) cosh((x t))

for x [t, ).

Continuity gives B = 0 and the condition on the jump of the first

derivative gives
1 = A.
Thus
G(x, t) =

0
if x t
1 (xt)
e
sinh((x t))

so, if > , the general solution on [0, ) of

y 00 (x) + 2y 0 (x) + 2 y(x) = f (x)

is
y(t) =

G(s, t)f (s) ds =

0
2

f (s)e(ts) sinh((t s)) ds

with the positive square root of 2 .

If > the argument proceeds as far as [] in the paragraph above.
We then get
y(x) = Ae(xt) sin((x t)) + Be(xt) cos((x t)) + for x [t, ),
where is the positive square root of 2 2 . Arguing much as above
we get B = 0, A = 1 and the general solution on [0, ) of
y 00 (x) + 2y 0 (x) + 2 y(x) = f (x)

as
y(t) =

t
0

f (s)e(ts) sin((t s)) ds.

360

If = the argument again proceeds as far as []. We then get

y(x) = A(x t)e(xt) + Be(xt) .

Arguing much as before we get B = 0, A = 1 and the general solution

on [0, ) of
y 00 (x) + 2y 0 (x) + 2 y(x) = f (x)

as
y(t) =

(t s)f (s)e(ts) ds.

If is large the pointer takes a long time to return to close to equilibrium. If is small the pointer overshoots and we have violent oscillation so it also takes a long time before the pointer remains close to
equilibrium. As a rule of thumb close to is preferred.

361

K310
(i) Observe that, if m n
m
m
X
X
kT kn+1
r
0
kSn Sm k
kT k
kT kr
1

kT
k
r=n+1
r=n+1

as n , so the sequence Sn is Cauchy. Thus we can find an S

L(U, U ) with kSn Sk 0.
Now
kSk kSn Sk + kSn k kSn Sk +
kSn Sk + (1 kT k)

n
X
r=0

kT kr

(1 kT k)1

so kSk (1 kT k)1 . A similar calculation gives

kI Sk kT k(1 kT k)1 .

(ii) Observe that

kS(I T ) Ik kSn (I T ) Ik + k(S Sn )(I T )k
= kT n+1 k + k(S Sn )(I T )k

kT kn+1 + kS Sn kkI T k 0

as n 0, so kS(I T ) Ik = 0 and S(I T ) = I. Similarly

(I T )S = I. Thus I T is invertible with inverse S. Setting T = I A
gives the result.
(iii) Observe that
kA Ik = kB 1 B B 1 Ck kB CkkB 1 k < 1

so A is invertible. Thus

A1 B 1 C = I and CA1 B 1 = BB 1 CA1 B 1 = BB 1 = I

so C is invertible with inverse A1 B 1 .
We observe that
kC 1 k kB 1 kkA1 k kB 1 k(1 kI Ak)1 = kB 1 k(1 kI Ak)1

but
so

1 > kB CkkB 1 k kA Ik 0
kC 1 k kB 1 k(1 kB CkkB 1 k)1 .

Similar arguments give the second inequality.

(iv) If B E, then part (ii) shows that

= {C L(U, U ) : kB Ck < kB 1 k1 } E

362

so E is open. If C then

k(C) (B)k kBk1 kkB Ck(1 kBk1 kB Ck)1 0

as kC Bk 0. Thus is continuous.
(iv) Calculations as in (ii).

(v) Use the chain rule or (essentially the same) observe that
(B + H) = (B + H)1 = (B(I + B 1 H))1
= (I + B 1 H)1 B 1
= (I B 1 H + (B 1 H)kHk)B 1
= B 1 B 1 HB 1 + (H)kHk

= (B) + B (H) + (H)kHk.

where

k(H)k = kB 1 (B 1 H)B 1 k kB 1 k2 k(B 1 H)k 0

as kHk 0.

If n = 1, this reduces to the statement that

d
cx1 = cx2 .
dx

363

K311
(ii) If kAk = 0, then A = 0 and there is nothing to prove so we
assume kAk 6= 0. exists since any non-empty set of real numbers
bounded below has an infimum. Now choose > 0. We can find a j
such that kAj k1/j + . Thus
1/(jk+r)

kAjk+r k1/(jk+r) kAj kk kAkr

( + )jk/(jk+r) kAkr/(jk+r) +

for all 0 r j 1. Thus

lim inf kAn k1/n lim sup kAn k1/n +

n

n 1/n

and since > 0 kA k

(iii) and (iv) If (A) < 1, set = ((A) + 1)/2. There exists an
n
N such that kAn k1/n and so kAkP
n for n N . The kind

of arguments used in K310 show that j=0 Aj converges and that the
limit is the inverse of I A.

n 1/n
If (A) > 1 then there exists an N such
1 for all
P thatj kA k
n N so, by the easy part of the GPC, j=0 A fails to converge.

364

K312
(i) k()n k1/n = ||kn k1/n
(ii) Take a basis ej and let be the unique linear map with ej =
ej+1 for 1 j m 1, em = 0. We have () = 0.
(iii) The
Pm eigenvectors ej with
Q eigenvalues j are linearly independent. (If m

e
=
0
apply
j=1 j j
j6=k ( j ) to both sides.) They thus
form a basis. If the standard basis is uj , let be the unique linear map
with ej = uj .
Observe that
k(1 )n k1/n = kn 1 k1/n k1 k1/n kn k1/n kk1/n ()

so (1 ) (). Similarly

() = ( 1 (1 )) ( 1 )

so (1 ) = (). But if is a diagonal matrix with respect to the

standard basis, has norm equal to the absolute value of the largest
diagonal entry so ( 1 ) = max |j |.
(iv) If u1 and u2 form the standard basis let , be the linear maps
given by
u1 = u2 , u2 = 0, u2 = u1 , u1 = 0.

365

K313
(i) This is just the formula for change of basis.
(ii) Here and elsewhere we use the fact that
m2 sup |aij | kk sup |aij |
1i,jm

1i,jm

when (aij ) is the matrix of with respect to the matrix. (In fact we
merely use
K sup |aij | kk K 1 sup |aij |
1i,jm

1i,jm

for some K and this follows from the fact that all norms on a finite
dimensional space are Lipschitz equivalent.)
Given any > 0, we can certainly find cij with cij = bij for i 6= j,
|cii bii | < and the cii all distinct.
(iii) Use the notation of (i). Use (ii) to find n with m distinct
eigenvalues and kn k 0. Let n = 1 n . We know that n
has the same eigenvalues as n and
k n k = k1 ( n )k k 1 kk n kkk 0

as n 0.

(iv) If has m distinct eigenvalues we know that we can find a basis

of eigenvectors. With respect to this basis, is a diagonal matrix D
with entries the eigenvalues so
m
Y
(t) = det(tI D) =
(t j )
j=1

and (D) = 0 the zero m m matrix. Thus () = 0.

2

(v) We know that the map from L(Cm , Cm ) to C m given by 7 (aij )

is continuous so the map 7 (A) is continuous. But the inverse
map (aij ) 7 is also continuous so the map 7 () is continuous.
By the previous parts of the question, given any we can find n
with k n k 0 with n (n ) = 0, so, by continuity, () = 0
(vi) We need non singular. One (among many ways) of
doing this is as follows. Take e1 , e2 , . . . , em to be a basis of Cm . Let
be the linear map with ej = ej+1 for 1 j m 1, em = e1
and be the linear map with ej = jej for 1 j m.

366

K314
(i) We know that, given any > 0, we can find a() 1 and b() 1
such that
kr k a()(() + )r and k r k b()(() + )r

for all r 0. Thus

X n

k( + )n k =
j nj

j
j=0
n
X
n
kj kk nj k

j
j=0
n
X
n
a()b()
(() + )j (() + )nj
j
j=0
= a()b()(() + () + 2)n

so, taking n-th roots and allowing n ,

( + ) (() + () + 2).

Since > 0 this gives the result.

(ii) Observe that, given an upper triangular matrix A, we can find

a diagonal matrix B with entries in absolute value less than any give
> 0 such that the diagonal entries of A + B are all distinct. Observe
that A and B commute. Now argue as in K313 parts (i) to (iii).
(iii) If has m distinct eigenvalues, then we may find a basis e1 ,
e2 , . . . of eigenvalues vectors with corresponding eigenvalues j with
|1 | |2 | . . . . Since kn e1 k/ke1 k = |1 |n , we have () |1 |.
However, if > |1 | we have

m
m
X
X
j

n
n

xj ej
kej k 0

j=1

j=1

so () . Thus () = |1 |.

If does not have n distinct eigenvalues, then, by (ii), we can find

r such that the eigenvalues of and + r differ by less than 1/r,
kr k 1/r and and r commute.
Observe that, by (i),
and

( + r ) () + (r ) () + kr k = () + 1/r
() = ( + r r ) ( + r ) + 1/r

so ( + r ) () and the desired result follows.

367

(iv) If is represented by A with respect to one basis and by B

with respect to another, we can find an invertible such that 1 is
represented by B with respect to the first basis. Now
k(1 )n k = kn 1 k kkkn kk1 k,

so, taking n-th roots and letting n , we have ( 1 ) () and

similarly
() = ( 1 1 ) ( 1 )

so (1 ) = () and there is no ambiguity.

368

K315
(i) Observe that
n

xn = x0 +

n1
X

j b

j=0

so xn ( ) b.

(ii) Observe that the map : Cm Cm given by x = b + x

satisfies
k n x n yk = kn (x y)k kn kkx yk.

For large enough n we have n a contraction mapping so xn c where

c is the unique fixed point given by
c = c
i.e. by c = b.
P
j
Suppose () > 1. If n1
j=0 b diverges, then take x0 = 0. If not,
take x0 to be an eigenvector corresponding to a largest (in absolute
value eigenvalue).
Pn1 j
Suppose () = 1. If
j=0 b diverges, then take x0 = 0. If
not, then either all of the largest in absolute value eigenvalues are 1
and we consider x0 = e with e an associated eigenvector to get two
fixed points (with necessarily different limits) or one of the largest in
absolute value eigenvalues, is not 1. Take x0 = e with e an associated
eigenvector to get a non-convergent sequence.

369

K316
We require (I A) < 1. If kI Ak is small, convergence will be
rapid with the error roughly multiplied by kI Ak at each step.

Jacobi needs (D 1 (U +L)) < 1 and Gauss needs ((DL)1 U ) < 1.

Variation needs (H) < 1 with

H = (D L)1 ((1 )D + U )).

Noting that the determinant of a triangular matrix is the product of

its diagonal elements we have
det H = (det(D L))1 det((1 )D + U ) = (det D)1 det((1 )D)
= (det D)1 (1 )n det D = (1 )n .

Now the determinant of a matrix is the product of its eigenvalues (multiple roots counted multiply) (consider the coefficient of t0 in det(tI
A)) so, if | det H| > 1, we have (H) > 1. Thus the scheme fails for
< 0 or > 2.

370

K317
From K310 we know that the map 7 1 is continuous on the open
subset of L(U, U ) where the inverse is defined. Thus thus composition
of the maps
is continuous.

x 7 Df (x), 7 1

371

K318
(i) Observe that, if m n,

m
m
n
X r X
r
X
kkr

r=0 r!
r! r=n+1 r!
r=0

P
Pn
r
r
(since m
r=0 (x /r!) converges for all x) so, by completeness,
r=0 r!
converges.

(ii) Observe that

n

n
n
X r X

r s
r
r
r s
X
X
X

( + )

r=0 r! r=0 r!

r!
r! s! 0r+sn r,s0 r! s!
r=0
0r,sn

r s
X

r!
s!
r+s>n, 0r,sn

kkr kks
r! s!
0r,sn

r+s>n,

kkr kks
r! s!
r,s0

2nr+s>n,

2n
X
1
=
(kk + |k)k 0
k!
k=n+1

[Or use results from Chapter 3.]

(iii) Observe that, if 0 < khk < kk/2 we have

n
n
X
X
r
r
h
h

h =

r!
r!
r=2
r=0

n
X
|h|kkr
r=2

r!

|h| kk

X
r=0

2r1 = |h|2 kk2 .

Thus
exp(h) = + h + h2 (h)h2
where k (h)k kk2 for 0 < khk < kk/2. Similarly
exp(h) = + h + h2 (h)h2
where k (h)k kk2 for 0 < khk < kk/2.

372

We thus have
kh2 (exp(h) exp(h) exp(h) exp(h)) ( )k

= kh2 ( + h + h2 (h)h2 )( + h + h2 (h)h2 )

( + h + h2 (h)h2 )( + h + h2 (h)h2 )
( )k

2|h|kkk (h)k + kkk(h) k + 2h2 k(h) kk (h)k 0

as h 0. Thus, if 6= , we have exp(h) exp(h) 6= exp(h) exp(h)

for h small.
If exp exp = exp( + ) and exp exp = exp( + ) then
exp exp = exp exp .
(iv) Observe by considering the various terms in the expansion or by
induction that
n
X
n
n
n
k( + ) k
kkj kknj
j
j=1

n
X
n1
kkj kknj

n
j

1
j=1

n1
X
n1
= kk
n
kkr kkn1r
r
r=0
= nkk(kk + kk)n1

(We shall do something similar in K319.)

Thus
N

N
N
X ( + )n X
()n

X 1

k( + )n n k

n=0

n!
n!
n!
n=0
n=1
kk
= kk
so that

N
X
n
(kk + kk)n1
n!
n=1

N
1
X
n=0

1
(kk + kk)n
n!

= kk exp(kk + kk)
k exp( + ) exp()k kk exp(kk + kk) 0

as kk 0. Thus exp is continuous.

373

K319
(ii) If L(U, U ) and L(U, U ) write

n ()() = n1 + n2 + 2 n3 + + n1 .

It is easy to check that n () : L(U, U ) L(U, U ) is a continuous

linear function with
kn ()k nkkn1 ,
The next paragraph (not surprisingly) uses K260 (iii). Observe by
considering the various terms in the expansion or by induction that
n
X
n
n
n
k( + ) n ()()k
kkj kknj
j
j=2

n
X
n2
kkj kknj
n(n 1)

j2
j=2

n2
X
n2
2
kkr kkn2r
= n(n 1)kk
r
r=0
= n(n 1)kk2 (kk + kk)n2

Thus n is differentiable at with derivative n ().

(iii) Since kn ()k nkkn1 we have
N
X
kn ()k
j=1

PN

n ()
j=1 n!

n!

N
X
nkkn1
j=1

n!

ekk

so
converges in the space of linear maps from L(U, U ) to
L(U, U ) in the appropriate operator norm. Call the limit ().
We have

N
N
N
N
X ( + )n X
()n X n ()
X 1

k( + )n n n ()()k

n=0
n!
n!
n!
n!
n=0
n=1
n=1

N
X
1

n(n 1)kk2 (kk + kk)n2

n!
n=2

= kk2
2

so

N
2
X
n=0

1
(kk + kk)n
n!

kk exp(kk + kk)
k exp( + ) exp() ()k kk2 exp(kk + kk)

and exp is differentiable at with derivative ().

374

K320
Use the fact that
m2 sup |aij | kk sup |aij |
1i,jm

1i,jm

when (aij ) is the matrix of with respect to the matrix. (Or merely
use
K sup |aij | kk K 1 sup |aij |
1i,jm

1i,jm

for some K and this follows from the fact that all norms on a finite
dimensional space are Lipschitz equivalent.)
If A is upper triangular then Ar is upper triangular with (j, j) th
entry the rth power of ajj the (j, j) th entry of A. Thus exp A is upper
triangular with (j, j) th entry exp ajj . Thus
Y
X
det(exp A) =
exp ajj = exp
ajj = exp Trace A

and, since any has an upper triangular representation and det and
Trace depend on and not the representation chosen,
det(exp ) = exp Trace .

375

K321
(ii) Note, for example, that, writing r(B) for the rank of B, we have
2 r(A) r(A) r(A2 ) r(A2 ) r(A3 ) r(A3 ) r(A4 ).
Since A4 = 0 and A2 6= 0 we have r(A2 ) r(A3 ) 1 so 2 3 which is
absurd.
(iii) We have
a2 + bc = 1
b(a + d) = 0
c(a + d) = 0
d2 + bc = 1
so either a + d 6= 0, in which case b = c = 0 and a = d = 1 so A = I
or a + d = 0 and, either a2 = 1 in which case bc = 0 and

1 0
1
or A =
A=
1
0 1
for some or a2 6= 1 and
A=

a
b
b1 (1 a2 ) a

for some a and b.

Observe that
1 = det I = det A2 = (det A)2
so det A = 1.
By inspection the only such matrices with det A = 1 are I and the
only such matrices with diagonal entries are

1 0
.
A=
0 1
(iv) Observe that S(I) = I and that S has a continuous derivative
and that DS(I)(C) = 2C so DS(I) is invertible. The result now follows
from the inverse function theorem.
(v) The result fails for Y = 0 since

1 0
1
is a solution for all .

376

We show that DS(Y ) is not invertible so the the hypotheses of the

inverse function theorem fail. Observe that if

u v
H=
w x
then

DS(Y )H = Y H + Y H

u v
u
v
+
=
w x
w x

u 0
=2
0 x

so DS(Y ) is not bijective. (Consider, for example, u = x = 0, v = w =

1. Then H 6= 0 but DS(Y )H = 0.)
(vi) Inspection of cases (along the lines of (v)) shows that, unless
B = I then, if B 2 = I, there exist Cn with Cn 6= B, kCn Bk 0
and Cn2 = I. If B = I then the argument of (iv) (or argument from
the result of (iv)) shows that G and H exist.

377

K322
We have
hx, xi = hx, T xi = hx, xi = hx, xi.

So hx, xi = 0.

If 2 x = x, then
hx, xi = hx, xi = h2 x, xi = hx, xi 0.

Since a real cubic has at least one real root, the characteristic equation det(t ) = 0 must have at least one real root. Thus has an
eigenvector e3 of norm 1. By the first paragraph e3 = 0. We observe
that e3 is an eigenvector of 2 and so since 2 is symmetric we can find
e1 and e2 such that the ej form an orthonormal basis. We know that
e1 is perpendicular to e1 . Also
he1 , e3 i = he1 , e3 i = 0

Thus e1 = e2 for some R.

Similarly e2 is perpendicular to e2 and e3 whilst

so e2 = e1 .

he2 , e1 i = he2 , e1 i =

With respect to the basis ej the linear map exp has the
representation
P (1)r 2r+1

n P (1)r 2r

0 0
r=0
r=0 (2r+1)!
(2r)!
X 1
P
P (1)
r 2r
(1)r 2r+1
0 0 =

r=0 (2r+1)!
r=0
(2r)!
n!
n=0
0 0 0
0
0

cos sin 0
= sin cos 0
0
0
1

so exp is a rotation through about e3 .

matrix
0

0
1

378

K323
We have
= ( + T )/2 + ( T )/2.

The first term is symmetric the second antisymmetric. Note that the
decomposition is unique (if = + with symmetric and antisymmetric then ( + T )/2 = ).
If is as stated
k( T )/2k k( )/2k + k( )T /2k = k k <

so = + + with antisymmetric, kk symmetric and kk 1.

Now
= T = ( + + )( + ) = + ( ) 2 2 + 2 + 2

so

But
so

2 = 2 + ( ) + 2 2 .
k( + T )/2k k( )/2k + k( )T /2k = k k <
2kk kk2 + 2kkk + 2 2 2 + 22 + 2 = 42

so kk 22 . [We can reuse this estimate to show that kk = 2 /2 +

O(3 ).]

379

K324
We call matrices U with U U T orthogonal. If, in addition, det U = 1
we say it is special orthogonal.
(i) Observe that
1 = det U U T = det U det U T = (det U )2 .
(ii) Observe that
!T
N
X Bj
j=0

j!

N
X
(B T )j
j=0

j!

N
X
(B)j
j=0

j!

so, since the map A 7 AT is continuous, (exp B)T = exp(B) =

(exp B)1 . (Recall that, if C and D commute, exp(C+D) = exp C exp D.)
(iii) The maps h 7 hB, A 7 exp A and C 7 det C are continuous
so their composition is. Note if B T = B we have (hB)T = hB so
exp hB is orthogonal so det(exp hB) = 1. Since h 7 det(exp hB) is
continuous, the intermediate value theorem tells us that that the map
is constant so (1) = (0) and det(exp B) = 1.
(iv) By (ii) and (iii) exp hB is special orthogonal so A(exp hB) =
(exp hB)A. But, as in K318, we have
kh1 (I hB exp hB)k 0

so
so

kA(h1 (I hB exp hB)) (h1 (I hB exp hB)Ak 0

kAB BAk 0

i.e. AB = BA.

(v) If n 2, we can have 1 I, J n with I 6= J. Considering B

with matrix given by bIJ = bJI = 1, bij = 0 otherwise, we see that
aII = aJJ and aIJ = aJI for all I 6= J.
If n 3, we can have 1 I, J n with I, J and K distinct
Considering B with matrix given by bIJ = bJI = 1, bJK = bKJ = 1,
bij = 0 otherwise, we see that
n
n
X
X
aIs bsK = aIK .
bIr arK =
aJK =
r=1

s=1

On the other hand, if we consider B with matrix given by bIJ = bJI =

1, bJK = bKJ = 1, bij = 0 otherwise, we see that
n
n
X
X
aIs bsK = aIK .
bIr arK =
aJK =
r=1

s=1

380

The two last results show that aJK = 0 for all J 6= K.

Thus A = I. Conversely, if A = I, it is easy to see that A is
isotropic.
(vi) When n = 2 we have shown that

a b
A=
b a

Let k be the positive square root of a2 + b2 . Then

cos sin
,
A=k
sin cos

that is A is product of a dilation and a rotation. But the special

orthogonal matrices are precisely the rotations so the isotropic matrices
are precisely those which are a product of a dilation and a rotation.
The isotropic matrices of determinant 1 are the orthogonal matrices.
(vii) If n = 1 the special orthogonal matrices consist of one example
(1) so all matrices are isotropic.
(viii) If we omit the restriction det U = 1, we get a stronger condition
so the new isotropic matrices are a subset of the old. If n 3 or n = 1
this makes no difference, by inspection.
If n = 2, we observe that

1 0
cos sin
1 0
cos sin
=
0 1
sin cos
0 1
sin cos

so under the new definition only matrices of the form I can be isotropic
(and it is easily checked that they are).

381

K325
The area of the inscribed n-gon is given by
n
n
X
k
r2 X
k
2
=
A=
sin k
r sin cos
2
2
2 k=1
k=1

subject to

n
X

k = 2.

k=1

We form the Lagrangian

n
n
X
r2 X
k = 2
sin k
L=
2 k=1
k=1

and observe that, at a stationary point,

L
0=
= cos k .
k
Thus 1 = 2 = = n . Inspection (this is methods question) shows
that we have the maximum.
The area of the inscribed n-gon is given by
n
X
k
A=
r2 tan
2
k=1

subject to

n
X

k = 2.

k=1

We form the Lagrangian

n
n
X
X
k
2
k = 2
L=
r tan
2
k=1
k=1

and observe that, at a stationary point,

L
1
0=
=
k
2 cos2

k
2

Thus 1 = 2 = = n . Inspection (this is methods question) shows

that we have the maximum.

382

K326
Observe that the set
E = {x Rn : xj 0 for all j and

xpj = c}

j=1

P
p
is closed (observe that the map x 7
j=1 xj is continuous) and
bounded (if x E then c1/p xP
j 0.) Thus the continuous function f : E R given by f (x) = j=1 xj yj attains a maximum. This
maximum does not occur if any of the xj = 0 (for, if xJ = 0, then
if delta is very small and positive and we take x0J = and take so
that taking x0r = xr if xr 6= 0, x0r = 0 if xr = 0 and r 6= J we
have x0 E, we see that = O( p ) and f (x0 ) > f (x)). Thus, if the
Lagrange method gives us a unique stationary point with xj 6= 0 for
all j, it will indeed be the maximum.
Form the Lagrangian
L=

n
X
j=1

xj yj

n
X

xpj .

j=1

At a stationary point,
0=

L
= yj pxjp1 ,
xj

so we have indeed got the maximum. Rewriting the last equations we

obtain
yj = pxjp1
so
xpj = kyjq
1
P
P
n
q
for some constant k. Since nj=1 xpj = c, we have k = c
y
j=1 j
and
!1/p
! n

n
n
X
X
X
1+q/p
xj yj = c1/p
yj
yjq
j=1

j=1

=c

c1/p

1/p

n
X

j=1

yjq

yjq

!1/q

j=1

n
X
j=1

n
X
j=1

yjq

!1/p

383

Thus
c1/p
Pn

n
X

Thus

n
X
j=1

tpj

tpj

!1/q

n
X

t j yj

j=1

= c with equality if and only if tpj = kyjq .

with equality if and only if

yjq

j=1

p
j=1 tj

whenever tj 0 and

n
X
j=1

yjq

!1/q

kyjq .

n
X

t j yj

j=1

We assumed that all the yj where non-zero but a little reflection

shows that if tj 0 and yj 0 (and t, y 6= 0) then
!1/q
n ! n
n
X p
X
X q

tj
t j yj
yj
j=1

with equality if and only if

j=1

j=1

tpj

kyjq .

Thus if t, y 6= 0 then
n
! n
!1/q
n
X
X
X
p
q
|tj |
|yj |

|tj yj |
j=1

j=1
p

j=1

with equality if and only if |tj | = k|yj | .

384

K327
(iii) The sum of the squares of the four sides of a parallelogram equals
the sum of the squares of the two diagonals.
(iv) Observe that
kxk kx yk + kyk

so

kxk kyk kx yk.

Now interchange x and y.

Observe that

4hx, yi = kx + yk2 + kx yk2

(kxk + kyk)2 + (kxk kyk)2

= 4kxkkyk

(v) Observe, for example, that the parallelogram law fails for the
uniform norm when we consider f (x) = max(1 4x, 0) and g(x) =
f (1 x).

385

K328
(ii) We have
ku + v + wk2 + ku + v wk2 = k(u + v) + wk2 + k(u + v) wk2
= 2ku + vk2 + 2kwk2 .

Thus
4(h(u + w, vi + h(u w, vi)

= ku + w + vk2 ku + w vk2 + ku w + vk2 ku w vk2

= 2ku + vk2 + 2kwk2 2ku vk2 2kwk2

= 8hu, vi
so (1) holds.

Setting w = u, we obtain (2). Now set u = (x + y)/2 and w =

(x y)/2 to obtain
hx, vi + hy, vi = hu + w, vi + hu w, vi
= 2hu, vi

= h2u, vi

= hx + y, vi.
(iii) A simple induction now gives hnx, yi = nhx, yi for n a positive
integer. The remark that
hx, yi = 41 (kx yk2 kx + yk2 = hx, yi

now gives the result for all integer n. Now we have

so

nhmn1 x, yi = hmx, yi = mhx, yi

hmn1 x, yi = mn1 hx, yi

for all integer m and n with n 6= 0.

Now observe that the Cauchy-Schwarz inequality holds (see K327 (iv))
so
|hx, yi hk x, yi| = |h( k )x, yi| k( k )xkkyk 0

as k 0. Thus allowing k through rational values of k gives

the full result.
(vi) Observe that if a norm obeys the parallelogram law on a dense
subset it must obey it on the whole space.
(vii) In parts (ii) and (iii) start by proving the identities for <hx, yi.

386

K329*
No comments.

387

K330
(i) Since |d(e, t) d(e, s)| d(s, t), the function fe is continuous.
Since |d(e, t) d(e0 , t)| d(e0 , e) it is bounded.
(ii) We have
and

|fu (t) fv (t)| = |d(u, t) d(v, t)| d(u, v)

so kfu fv k = d(u, v).

|fu (u) fv (u)| = d(u, v)

(iii) Since C(E) is complete, the closed subset Y is complete under

the uniform norm. We now observe that is a distance preserving
such that (E) is dense in Y .
mapping from (E, d) to (Y, d)

388

K331
(ii) We have xn1 = ME (n1 , x) = M (n1 , x) M (0, x) as n 0
by the continuity of M so M (0, x) = 0 for all x 6= 0. In particular
M (0, 1) = M (0, 2), which is incompatible with group laws (ab = cb
implies a = b).
(iii) As for (ii).
(iv) If xn is Cauchy for d, then d(xn , 1) is bounded so there exist
A > 1 such that A1 xn A. Observe that (by the mean value
theorem) there exists a K > 1 such that K 1 |xy| d(x, y) K|xy|
for all |x|, |y| [A1 , A]. Take M (x, y) = xy.

389

K332
(i) If E = {0} then we are done. If not, then since x E implies
x E, we know that {x E : x > 0} is non-empty.
Let = inf{x E : x > 0}. If = 0 then we can find xn E with
0 < xn < 1/n. We have
E

n=1

{mxn : m Z}

so, since E is closed, E = R.

If > 0, then since E is closed E. Given e E we can find

k Z such that
> e k 0.
Since e k E we have e k = 0 so E Z so E = Z.

(ii) A similar argument shows that either E = S 1 or E = {exp 2in/N :

0 n N 1} for some positive integer N .

(iii) If E is a subset of a one dimensional subspace of R2 then (i)

tells that either
E = {xu : x R} or E = {nu : n Z}.
If not either
(A) xn 6= 0 with kxn k 0, or (B) not.
In case (A) either
(A)0 There exists an N such that all the xn lie in a one dimensional
subspace or (A)00 not.
In case (A)00 , E will be dense in R2 and so E = R2 .
In case (A)0 the argument of (i) tells us that there is a subspace
U = {xu : x R} E.
We are considering the case U 6= E. Consider
= inf{kxk : x E \ U }.

If = 0, E will be dense in R2 and so E = R2 . If 6= 0 then since

U is a subgroup of E, ky zk whenever y E \ U and z U .
Thus, since E is closed, we can find v E \ U with kvk = . Suppose
e E. By simple geometry we can find x R and k Z such that
ke xu kvk kvk/2 so e xu kv U so e kv U . Thus
E = {xu + kv : x R, k Z}.

390

In case (B) let u be an element of smallest norm E \ {0}, write

U = {lu : l Z}

and let v an element of smallest norm in E \ U . If e E then by

simple geometry we can find integers l and k such that e lu kv lies
in the parallelogram with vertices (u v)/2 and so is 0. Thus
E = {lu + kv : l Z, k Z}.

391

K333
Second paragraph. Let X = [1, 1] with the usual metric d. Let
E = X \ {0} and let Y = E with the restriction of d to Y . Let
f (x) = x so f is certainly uniformly continuous. Suppose f : (X, d)
(Y, ) is continuous and f(x) = f (x) for all x E. If f(x) = y then
|x y| = (f(x), f(0)) 0

as |x| = d(x, 0) 0 (x 6= 0) which is absurd.

We have the following theorem. If (x, d) is a metric space and E a

dense subset of X then if (Y, ) is a complete metric space any uniformly
continuous function f : E Y can be extended to a continuous (indeed
uniformly continuous) function f : X Y .
The proof follows a standard pattern. Suppose x X. Then we can
find xn E with d(xn , x) 0. Since f is uniformly continuous on E,
given any > 0 we can find a () > 0 such that d(u, v) < implies
(f (u), f (v)) < . Choose N such that d(xn , x) < /2 for n N .
Then d(xn , xm ) < and (f (xn ), f (xm )) < for n, m N . Thus
f (xn ) is Cauchy and so converges in (Y, ) to limit lx , say. A similar,
but simpler, argument shows that if yn E with d(yn , x) 0 then
(f (yn ), lx ) 0. Thus we may define f(x) = lx without ambiguity.
Taking xn = x, we see that f(x) = f (x) for all x E.
Now suppose > 0, u, v X and d(u, v) < ()/3. We can find
un , vn E with d(un , u), d(vn , v) < ()/3 and d(un , u), d(vn , v) 0.
By the triangle inequality d(un , vn ) < () so (f (un ), f (vn )) < and
(f(u), f(v)) . Thus f is uniformly continuous.

392

K334
Observe that
f (x) f (y)
0
xy
as y x. Thus f is everywhere differentiable with derivative zero and
so is constant.
If we replace R by Q, then f is uniformly continuous on Q and so
can be extended to a continuous function f on R. If x, y R and
x 6= y choose xn , yn Q with xn x and yn y. Since
we have
so, allowing n ,

|f (xn ) f (yn )| (xn yn )2

|f(xn ) f(yn )| (xn yn )2

|f(x) f(y)| (x y)2

whence f is constant and so f is.

The function in Example 1.3 satisfies the condition:Given x Q there exists a (x) > 0 with

|f (x) f (y)| (x y)2

for all x, Q with |x y| < (x) but the (x) is not uniform.

393

K335
This is the statement that a complete metric space is totally bounded
(see e.g. Section 11.2).
S
S
If Fn is finite set such that xFn B(x, 1/n) = X, then E =
n=1 Fn
is a countable dense subset.

If we give Z the usual metric, then, since the sequence xn = n has no

convergent subsequence, (|xn xm | 1 for m 6= n), Z does not have
the BolzanoWeierstrass property. However Z is a countable dense
subset of itself and since any Cauchy sequence is eventually constant
it is complete.
The space Q with the usual metric is not complete but is a countable
dense subset of itself.
If we give R the discrete metric (d(x, y) = 1 if x 6= y) then B(x, 1/2) =
{x} so the only dense subset of R is R itself which is uncountable. Since
any Cauchy sequence is eventually constant R with the discrete metric
is complete.

394

K336
(i) and (ii). Let xn be a sequence of points forming a dense subset.
If U is open, consider xn . If xn
/ U , we do nothing, If xn U , then
put n . Since U is open we can find a n0 > 0 such that d(xn , y) > n0
for all y
/ U . If U = X set n = 1. Otherwise inf yU
/ d(xn , y) is a well
defined strictly positive number. Set n = inf yU
/ d(xn , y)/2.
S
We claim that Sn B(xn , n ) = U . Observe that if n then
B(xn , n ) U so n B(xn , n ) U . Now suppose x U . Since U
is open we can find a > 0 with 1 > such that B(x, ) U . Since
the sequence xn is dense we can find an N such that d(xN , x) < /4.
Automatically
inf yU
/ d(xN , y) > 3/4 so N 3/8 and x B(xN , N ).
S
Thus n B(xn , n ) = U .
(iii) The space R with the discrete metric d(x, y) = 1 if x 6= y is
complete. Any open ball of the form B(x, r) with r > 1 is R itself. Any
open ball of the form B(x, r) with r < 1 is {x}. Thus the countable
unions of open balls are countable subsets of R and R itself. But every
set is open so any uncountable set (e.g. {x : x > 0}) which is not the
whole space is a counterexample.
S
n

)r). Since the count(iv) Observe that B(x, r) =

n=1 B(x, (1 2
able union of countable sets is countable, part (ii) shows that every
open set
Sj U is the countable union of closed balls L1 , L2 , . . . say. Set
Kj = r=1 Lr .

395

K337
(i) If x U , we can find a > 0 such that (x , x + ) U . Thus
x x.
If x y, then x, y (a, b) U so y x
If x y and y z, then x, y (a, b) U and y, z (c, d) U .
We then have (a, b) (c, d) an open interval (since both (a, b) and (c, d)
contain y) lying within U so x z.
(ii) Since x [x], [x] is non-empty. If [x] is bounded above look at
= sup[x]. If U then we can find > 0 such that (2, +2)
U . We can find [x] such that so x , and
+ . Thus x + contradicting the definition of . By
reductio ad absurdum,
{y [x] : y x} [x, ).

But we can find n 1/n with n x, so we can find (an , cn ) U

with a, n (an , cn ). Thus

[
[
{y [x] : y x}
[x, cn )
[x, n ) = [x, ).
n=1

n=1

Thus

{y [x] : y x} = [x, ).
If [x] is unbounded above we can find n n with n x so we can
find (an , cn ) U with x, n (an , cn ). Thus

[
[
{y [x] : y x}
[x, cn )
[x, )
n=1

n=1

so

{y [x] : y x} = [x, ).
(iii) Take C to be the set of equivalence classes.
(v) Every open interval contains a rational. Thus each element of C
contains a rational which belongs to no other element so C is countable.
(Or say that the map f Q C given by f : (q) = [q] is surjective.)
If D is an open disc {x : kx x0 k = r} and kx x0 k = r then any
open disc containing y intersects D but y
/ D.

396

K338
(i) Observe that

n
X
j=1

|(f + g)(xj ) (f + g)(xj1 )|

n
X
j=1

|f (xj ) f (xj1 )| +

n
X
j=1

|g(xj ) g(xj1 )|

so f, g BV implies f + g BV and V (f + g) V (f ) + V (g).

Observe also that if V (f ) = 0 then

0 |f (t) f (0)| |f (1) f (t)| + |f (t) f (0)| V (f ) = 0

so f (t) = f (0) for all t.
The proof that BV is complete follows a standard form of argument.
Suppose fn is Cauchy in BV . Arguing as in the previous paragraph

|fp (t) fq (t)| |fp (0) fq (0)| + V (fp V (fq ) = kfp fq kBV
so fn (t) is Cauchy in R for each t. Thus fn f (t) for some f (t). We
now show that f BV . Since any Cauchy sequence is bounded we
can find a K with K kfn kBV for all n. Now suppose 0 = x0 x1
xN = 1, Then
N
X
j=1

|f (xj ) f (xj1 )|

N
X
j=1

N
X
j=1

|(f fn )(xj ) (f fn )(xj1 )| +

N
X
j=1

|(f fn )(xj ) (f fn )(xj1 )| + K

|fn (xj ) fn (xj1 )|

397

as n . Thus f BV . Finally we use an irrelevant m argument

to observe that
|(f fn )(0)| +

N
X
j=1

|(f fm )(0)| +
+

N
X
j=1

|(f fn )(xj ) (f fn )(xj1 )|

N
X
j=1

|(f fm )(xj ) (f fm )(xj1 )| + |(fm fn )(0)|

|(fm fn )(xj ) (fm fn )(xj1 )|

|(f fm )(0)| +
|(f fm )(0)| +

N
X
j=1

N
X
j=1

|(f fm )(xj ) (f fm )(xj1 )| + kfn fm kBV

|(f fm )(xj ) (f fm )(xj1 )| + sup kfp fq kBV
p,qn

sup kfp fq kBV

p,qn

as m . Thus kfn f kBV 0 as n .

(ii) If f (t) = 0 for 0 t 1/2, f (t) = 1 for 1/2 < t 1 then
fP BV but f
/ C. If g(t) = t cos t then g C but, examining
n
/ BV .
j=1 |g(1/j) g(1/(j + 1))|, we see that g

C BV is the intersection of two vector spaces and so a vector

subspace of BV . If f BV C then arguments like those in (i)
show that kf k kf kBV . Thus if fn BV C and f BV and
kfn f kBV 0 we know that fn is Cauchy in (BV, k kBV ) and so fn
is Cauchy in (C, k k ). Thus fn converges uniformly to some g C.
We now observe that
|fn (t) f (t)| kf fn kBV 0
and
|fn (t) g(t)| kg fn k 0
and so f (t) = g(t) for each t [0, 1]. Thus f = g and f BV C.
Thus BV C is closed in (BV, k kBV ).

The mean value inequality shows that CR1 is a subspace of BV . We

t
observe that if g is continuous and G(t) = 1/2 g(x) dx then
|G(xj ) G(xj1 )|

sup

t[xj1 ,xj ]

|g(t)||xj xj1 |.

398

Thus if we take

if 0 x 21 21 n1 ,
1
gn (x) = 4n(x 1/2) if 21 21 n1 < x < 21 + 21 n1 ,

1
if 21 + 21 n1 x 1,
Rt
and set Gn = 1/2 gn (x) dx, G(t) = |t 21 | we have Gn C 1 , G BV
and kG Gn kBV 0 as n but G
/ C 1.

(iii) Since g C 1 there exists an M such that |g 0 (t)| M for all t.

Let > 0. By the definition of V (g). 0 = x0 x1 xn = 1 such
that
n
X
j=1

|g(xj ) g(xj1 )| V (g) .

Let m be a strictly positive integer. By the staircase property of the

function f we can find a finite collection I of disjoint closed sets of
total length 1 (2/3)m such that f is constant on each I I. We can
choose 0 = y0 y1 yN = 1 so that
xj {y0 , y1 , . . . , yN }
for each 0 j n and
either [yr1 , yr ] I for some I I or (yr1 , yr ) I = for all I I
whenever 1 r N . We observe that
N
X
r=1

|g(yr ) g(yr1 )|

n
X
j=1

|g(xj ) g(xj1 )| V (g) .

Let us say that r A if (yr1 , yr ) I = for all I I. We have

X
X
X
|(f + g)(yr ) (f + g)(yr1 )|
|f (yr ) f (yr1 )|
|g(yr ) g(yr1 )|
rA

rA

rA

rA

(f (yr ) f (yr1 ))

=1
1
1

X
rA
/

(f (yr ) f (yr1 ))

X
rA

X
rA

rA

|g(yr ) g(yr1 )|
X
rA

|g(yr ) g(yr1 )|

|g(yr ) g(yr1 )|
M |yr yr1 |

1 M (2/3)m = V (f ) M (2/3)m

399

and
X
rA
/

|(f + g)(yr ) (f + g)(yr1 )|

Thus

X
r

|g(yr ) g(yr1 )|

V (g) M (2/3)m .

V (f + g)

X
rA

X
rA

X
rA
/

|g(yr ) g(yr1 )|

|g(yr ) g(yr1 )|

|(f + g)(yr ) (f + g)(yr1 )| +

V (f ) + V (g) 2M (2/3)m .

X
rA
/

X
rA
/

X
rA
/

|f (yr ) f (yr1 )|

(f (yr ) f (yr1 ))

|(f + g)(yr ) (f + g)(yr1 )|

Since and m can be chosen freely, V (f + g) V (f ) + V (g) and so

V (f + g) = V (f ) + V (g).
In particular, if g C 1 we have
1

kg f kBV V (f g) V (f ) + V (g) V (f ) = 1

so C is not dense in BV .

400

K339
(iii) f : R (1, 1) given by f (x) = (2/) tan1 x will do.
(iv) f : I J given by f (x) = exp(ix) will do. Observe that
(1, 1) is a dense subset of [1, 1] which is complete. Observe that J
is a dense subset of {z C : |z| = 1} which is complete.
The Cauchy sequences xn = 1 1/n and yn = 1 + 1/n have no
limits in I. But |xn yn | 9 0 so they can not have the same limit in
the completion.

401

K340
(i) Suppose, if possible, that f 1 is not uniformly continuous. Then
we can find an > 0 and un , vn Y such that (un , vn ) 0 but
d(f 1 (un ), f 1 (vn )) > . By the BolzanoWeierstrass property of X,
applied twice to obtain a subsequence and then a subsequence of that
subsequence, we can find n(j) 0 and , X such that
d(f 1 (un(j) ), ) 0 and d(f 1 (vn(j) ), ) 0.

By the continuity of f , we have (un(j) , f ()) 0 and (vn(j) , f ())

0 so (since (un , vn ) 0) we have f () = f (). Since f is bijective,
= so
< d(f 1 (un(j) ), f 1 (vn(j) )) d(f 1 (un(j) ), ) + d(f 1 (vn(j) ), ) 0

which is absurd.

Suppose yn Y . Then f 1 (yn ) X and by the BolzanoWeierstrass

property of X we can find x X and n(j) such that d(f 1 (yn(j) ), x)
0 and so, by the continuity of f ,

(yn(j) , f (x)) = f (f 1 (yn(j) )), f (x) 0

as j . Thus Y has the BolzanoWeierstrass property.
(ii) The fact that f 1 is continuous.

(iii) If f (x) = x1/3 , the mean value theorem shows that |f (x)
f (y)| |x y|/3 if |x|, |y| 1. Since f is continuous, it is uniformly
continuous on [2, 2] so f is uniformly continuous on R. However
f 1 (x) = x3 so f (x + ) f (x) 3x 2 as x for all > 0,
so f 1 is not uniformly continuous.
(iv) Observe that (x, y) < 1/2 implies x = y implies d(f (x), f (y)) =
0 but that d(1/n, 0) 0 and (f 1 (1/n), f 1 (0)) = (1/n, 0) = 1 9 0.

402

K341
Since (X, d) has the BolzanoWeierstrass property, there exists a K
such that d(x, y) K for all x, y X (direct
or use total boundP proof
j
edness), thus, by the Weierstrass M-test, j=1 (2 d(x, xj ))2 converges
and f (x) l2 .
Observe that
kf (x) f (y)k22 =

so f is continuous.

(2j (d(x, xj ) d(y, xj ))2

j=1

X
j=1

(2j d(x, y))2 d(x, y)2

If x 6= y then d(x, y) > 0. Set 4k = d(x, y). We can find an N with

d(x, xN ) < k and so with d(y, xN ) > 3k whence d(x, xN ) 6= d(y, xN )
and f (x) 6= f (y). Thus f is continuous. It follows that f : X f (X) is
a bijective continuous function and, by K339 (i) is a homeomorphism.

403

K342
(i) We can have (X, d) complete but (Y, ) not. Let X = R with the
usual metric d, Y = (/2, /2) with the usual metric , and f (x) =
tan1 (x). Using the mean value inequality |f (x) f (y)| |x y|.
We can have (Y, ) complete but (X, d) not. Let X = (4, 4) with
the usual metric d and y = [1, 1] with the usual metric . Take
f (x) = sin x and observe that by the mean value inequality |f (u)
f (v)| |u v|.
(ii) If (X, d) is complete (Y, ) must be. Let yn be Cauchy in Y .
Since f is surjective we can find xn X with f (xn ) = yn . Since
d(xn , xm ) K 1 (f (xn ), f (xm ) = (yn , ym ),

the xn are Cauchy so we can find x X with d(xn , x) 0. By

continuity, (yn , f (x)) = (f (xn 0, f (x)) 0 as n .

We can have (Y, ) complete but (X, d)) not. Let Y = R with
the usual metric , and X = (/2, /2) with the usual metric d
f (x) = tan(x). Using the mean value theorem, |f (x) f (y)| |x y|.

404

K343
Observe that
4 4 4

so d(, ) + d(, ) d(, ).

Suppose n = (an , bn ) forms a Cauchy sequence with respect to d.

If there exists a c > 0 such that bn an > c for all n, then we know
that there exists an N such that d(n , m ) < c/2 for n, m N . Thus,
provided n, m N , we know that n m 6= so |an am | d(n , m ).
It follows that an is Cauchy in the standard Euclidean metric, so an a
for some a. Similarly bn b for some b. Since bn an > c we have
b a c > 0 and, writing = (a, b), we have d(n , ) 0.
If there does not exist a c > 0 such that bn an > c for all n, then
we can find n(j) such that bn(j) an(j) 0. If is any open
interval of length ||, d(, n(j) ) ||. Thus (X, d) is not complete.

However if we take X = X {} and define d by d (, ) = d(, )

for , X, d (, ) = d (, ) = || if X and d (, ) = 0
then d is a metric and a Cauchy sequence which consists from some
point on of intervals of length greater than some fixed c converges by
the arguments above and (since a Cauchy sequence with a convergent
subsequence converges) one which does not converges to .
[The reader may prefer to replace by .]

405

K344
(ii) Observe that the open unit ball E centre 1 consists of the odd
numbers. If r E and s
/ E then d(r, s) = 1 so E is closed.

(iii) d(n + 2k , n) 2k 0 so the complement of {n} is not closed

and {n} is not open.

(iv) d(r, s) 2k implies r s mod 2k so r2 s2 mod 2k and

d(r2 , s2 ) 2k . Thus f is continuous.
k

(v) d(n + 2k , n) 0 as k but d(2n+2 , 2n ) = 2n 9 0. Thus g

is nowhere continuous.
P
(vi) Let xn = nr=0 22r . If 0 k 2m then (think of the binary
expansion of x k) d(xn , y) 2m2 for all n k + 2 so the sequence
xn can not converge. However d(xn , xm ) 22r for all n, m r so the
sequence is Cauchy.

406

K345
Recall that we can find E : R R infinitely differentiable with
E(x) = 0 for x < 0 and E strictly increasing on [0, ). Thus, if we
consider F (x, y) = E(x + 1)E(x + 1)E(y + 1)E(y + 1), we have F
infinitely differentiable, F (0, 0) 6= 0 and F (x, y) = 0 if |x| 1 and/or
|y| 1.
Set (xn , yn ) = (n2 , n4 ) and n = (10n)8 . Note that, if we write
An = [xn 2n , xn + 2n ] [yn 2n , yn + 2n , ]

no line through the origin can intersect both An and Am for n 6= m.

Set
G(x, y) =

m=1

1
1
(y ym )).
n1 F (4m
(x xm ), 4m

For each (u, v) 6= 0 we can find a > 0 and an n such that

1
1
F (4m
(x xm ), 4m
(y ym )) = 0

for all m 6= n and all |u x|, |v y| < . Thus G is well defined

1
1
everywhere (observe that F (4m
(xm ), 4m
(ym )) = 0) and infinitely
differentiable except at (0, 0).
Since n (10n)8 we have G(x, y) n1 for |x|, |y| (10n)1 so
G is continuous at (0, 0). Since no line through the origin can intersect
both An and Am for n 6= m, G(t, t) = 0 for sufficiently small |t|, so
G has directional derivative zero at the origin.
However,
n1 E(1)4
|G(xn , yn ) G(0, 0)|
=
nE(1)4 /2 ,
k(xn , yn )k
k(xn , yn )k
so G is not differentiable at the origin.