Solutions Sheet 1, Limits Questions 9-15

9) To prove limx→+∞ H (x) = +∞, let K > 0 be given. Choose X = K.

Assume x > X.
Remember, we hope to prove H (x) > K so we look for lower bounds on
H (x). Then
1
H (x) = 2 + x > x,
x +1
where we are simplifying the expression by “throwing away” the complicated
part 1/ (x2 + 1) > 0 . Continuing,

H (x) > x > X = K.

Thus we have verified the definition of limx→+∞ H (x) = +∞.

To prove limx→−∞ H (x) = −∞ let L < 0 be given. Choose X = L − 1.

Assume x < X.
We hope to prove H (x) < L and so we look for upper bounds on H (x).
This means that we can’t throw away the 1/ (x2 + 1) term. Instead we use
the fact that 1/ (x2 + 1) < 1 for any x ∈ R. Then
1
H (x) = + x < 1 + x < 1 + X = L.
x2 +1
Thus we have verified the definition of limx→−∞ H (x) = −∞.

The graph of H (x) is

3

-3 -2 -1 0 1 2 3
x
-1

-2

-3

1
9) i) Assume limx→0 f (x) exists and call it L, say. So

∀ε > 0, ∃δ > 0 : ∀x, 0 < |x| < δ ⇒ |f (x) − L| < ε. (1)

Take ε = 1/3 and let δ > 0 be given by (1).

Consider the deleted neighbourhood 0 < |x| < δ. In any such region we
can find a rational x1 . Then f (x1 ) = 1 while |f (x1 ) − L| < 1/3 by (1) , so

|1 − L| < 1/3.

In the region 0 < |x| < δ we can also find an irrational x2 . Then f (x2 ) =
0 while |f (x2 ) − L| < 1/3 again by (1) , so

|0 − L| < 1/3.

This means that L is both close to 0 and 1. To get a contradiction, note

1 = |1 − L + L| ≤ |1 − L| + |L| < 2/3,

having used the triangle inequality. This is a contradiction so the initial

assumption is false, thus limx→0 f (x) does not exist.

ii) We might guess that the limit is 0.

Let ε > 0 be given. Choose δ = ε. Then for x : 0 < |x − 0| < δ we have

|xf (x) − 0| = |xf (x)| ≤ |x| since |f (x)| ≤ 1,

< δ = ε.

So we have verified the definition of limit, thus limx→0 xf (x) = 0.

11) i) The rational function

3x2 + 4x + 1
x2 + 2x + 3
is well-defined at 0 (in particular the denominator is not 0) so by the Quotient
Rule
3x2 + 4x + 1 limx→0 (3x2 + 4x + 1) 1
lim 2 = 2
= .
x→0 x + 4x + 3 limx→0 (x + 4x + 3) 3

2
 ii) Divide top and bottom by the largest power of x to get

3x2 + 4x + 1 3 + 4/x + 1/x2

lim = lim
x→∞ x2 + 4x + 3 x→∞ 1 + 4/x + 3/x2

limx→∞ (3 + 4/x + 1/x2 )

= by the Quotient Rule
limx→∞ (1 + 4/x + 3/x2 )
= 3.

iii) The rational function is not well-defined at x = −1, the denominator

is 0. This means that the denominator must factorise with a factor x + 1,
and we find that x2 + 4x + 3 = (x + 1) (x + 3) . The limit can only exist if
the numerator is also zero at x = −1, i.e. also has a factor x + 1. In fact
3x2 + 4x + 1 = (x + 1) (3x + 1). Hence

3x2 + 4x + 1 (x + 1) (3x + 1)
lim = lim
x→−1 x2 + 4x + 3 x→−1 (x + 1) (x + 3)

3x + 1
= lim
x→−1 x + 3

limx→−1 (3x + 1)
= by the Quotient Rule
limx→−1 (x + 3)
= −1.

12) i) You may only use the product rule for limits when both limits exist.
Here we know from the notes that limx→0 sin (π/x) does not exist, so we
cannot apply the Product Rule (even if the answer it gives is correct!)

ii) We might guess that the limit is 0.

Let ε > 0 be given. Choose δ = min (1, ε). Then for x : 0 < |x − 0| < δ
we have
 π     π      π 
 3
x sin − 0 = x3 sin  ≤ x 3  since sin  ≤ 1,
    
x x x
< |x| since |x| < δ ≤ 1,
< ε since |x| < δ ≤ ε.

Hence we have verified the definition of

π 
3
lim x sin = 0.
x→0 x

3
13) i)

sinh x ex − e−x
=
x 2x
ex − 1 + 1 − e−x 1 ex − 1 e−x ex − 1
   
= = +
2x 2 x 2 x

1 + e−x
  x 
e −1
= .
2 x

Now use the product rule for limits along with the results that limx→0 (ex − 1) /x =
1 and
1 1
lim e−x = lim x
= by the Quotient Rule
x→0 x→0 e limx→0 ex
= 1,

to get
sinh x
lim = 1.
x→0 x
ii)
tanh x sinh x 1
= × .
x x cosh x
So we need only note that
ex + e−x 2
cosh x = → =1
2 2
as x → 0. Combined with part i) this gives
tanh x
lim = 1.
x→0 x
iii) Apply the same idea as used for (cos x − 1)/x2 in lectures: For x 6= 0,

cosh2 x − 1
 
cosh x − 1 cosh x − 1 cosh x + 1
= × = 2
x2 x2 cosh x + 1 x (cosh x + 1)
 2
sinh x 1
= since cosh2 x − sinh2 x = 1,
x cosh x + 1

1
→ 12 × as x → 0, by part 1.
2
4
Thus
cosh x − 1 1
lim 2
= .
x→0 x 2
14) i)

θ θ cos θ
lim = lim
θ→0 tan θ θ→0 sin θ

θ
= lim × lim cos θ,
θ→0 sin θ θ→0

by product rule, allowable since both limits exist,

1
= sin θ
× lim cos θ,
limθ→0 θ
θ→0

by quotient rule, allowable since both limits exist,

= 1.

ii) a)
sin 3θ 3 sin 3θ 3
lim = lim = lim f (g (θ))
θ→0 2θ 2 θ→0 3θ 2 θ→0
where g (θ) = 3θ and f (η) = (sin η) /η.

Note that g (0) = 0 and f is not defined at η = 0. So to be able to use

the Theorem on the Composition Rule for Limits, we need only note that
g (θ) 6= 0 in some deleted neighbourhood of θ = 0. (For if g (θ) where zero
for some θ 6= 0 then f (g (θ)) would not be defined for such θ.) In fact,
g (θ) = 3θ 6= 0 for all θ 6= 0.
The conclusion of the Composition Theorem states that

lim f (g (θ)) = lim f (η) = 1.

θ→0 η→0

Hence
sin 3θ 3
lim = .
θ→0 2θ 2
b) Write
sin θ2


= f (g (θ))
θ2
where g (θ) = θ2 and f (η) = (sin η) /η.

5
 Then limθ→0 g (θ) = 0 and limη→0 f (η) = 1. Note that g (0) = 0 and
f is not defined at η = 0. So to be able to use the Theorem on the Com-
position Rule for Limits, we need only note that g (θ) 6= 0 in some deleted
neighbourhood of θ = 0. In fact, g (θ) = θ3 6= 0 for all θ 6= 0. Hence

sin θ2


sin (η)
lim 2 = lim = 1.
θ→0 θ η→0 η

iii) Use the Sum Rule

sin bx − sin ax sin bx sin ax

lim = lim − lim
x→0 x x→0 x x→0 x
sin bx sin ax
= b lim − a lim
x→0 bx x→0 ax

= b lim f (g (x)) − a lim f (h (x))

x→0 x→0

where f (x) = (sin x) /x, h (x) = bx and h (x) = ax. Then the composite
rules for limits gives

lim f (g (x)) = lim f (x) = 1

x→0 x→0

and
lim f (h (x)) = lim f (x) = 1.
x→0 x→0

Hence
sin bx − sin ax
lim = b − a.
x→0 x
15) Assume limx→+∞ sin (πx) = L exists. The definition tells us that

∀ε > 0, ∃X ∈ R : ∀x, x > X =⇒ |sin (πx) − L| < ε. (2)

Take ε = 1/3 and let X be the real number that (2) tells us exists.
Take x = n where n ∈ N such that n > X. Since n is an integer sin (πn) =
0 and since n > X we have |sin (πn) − L| < 1/3 by (2) . Combining we have
|L| < 1/3.
Consider now x = 2n + 1/2 chosen also to be > X. Now
π 
sin πx = sin + 2nπ = 1
2

6
and so |sin (πx) − L| < 1/3 becomes |1 − L| < 1/3. Thus L is close to both
1 and 0!

Rigorously,

1 = |1 − L + L| ≤ |1 − L| + |L| < 2/3,

having used the triangle inequality. This is a contradiction so the initial

assumption is false, therefore limx→+∞ sin (πx) does not exist.

2011/12