Topic 2: Limits

2.1. Tangent Lines and Rates of Change

A tangent line to the function
f
at the point
a x
is a line that just touches the graph of the
function at the point in question and is moving in the same direction as the graph at that point.
Example 1: Find the tangent line to
2
2 15 ) ( x x f at . 1 x
Solution: We know from algebra that to find the equation of a line, we need a single point on
the line and the slope of the line. The tangent line and the graph of the function must touch at
, 1 x
so the point
) 13 , 1 ( )) 1 ( , 1 ( f
must be on the tangent line.
At this point in time all that were going to be able to do is to get an estimate for the slope of
the tangent line. Lets call the given point
). 13 , 1 ( P
Then lets pick another point on the graph
of the function, say
)). ( , ( x f x Q
For the sake of argument, lets take choose 2 x and so the
second point will be
). 7 , 2 ( Q
1
We can see from this graph that the secant and tangent lines are somewhat similar and so the
slope of the secant line should be somewhat close to the actual slope of the tangent line. So, as
an estimate of the slope of the tangent line we can use the slope of the secant line, lets call it
.
PQ
m
. 6
1 2
13 7

PQ
m
To get a better estimate, we can take an x that is closer to 1 x and find the slope of the new
secant line. In other words, as we take Q closer and closer to P, the slope of the secant line
connecting Q and P should be getting closer and closer to the slope of the tangent line.
We only looked at Qs that were to the right of P, but we could have just as easily used Qs that
were to the left of P and we would have received the same results. In fact, we should always
take a look at what is happening on both sides of the point in question when doing this kind of
process.
Lets get a formula for the slope of the line between P and
)). ( , ( x f x Q
.
1
) 1 ( 2
1
13 2 15
1
13 ) (
2 2

x
x
x
x
x
x f
m
PQ
Now lets pick some values of x getting closer and closer to
, 1 x
and plug in to get some
slopes.
x
PQ
m x
PQ
m
2 -6 0 -2
1.5 -5 0.5 -3
1.1 -4.2 0.9 -3.8
1.01 -4.02 0.99 -3.98
1.001 -4.002 0.999 -3.998
1.0001 -4.0002 0.9999 -3.9998
So, if we take xs to the right of 1 and move them in very close to 1, it appears that the slope of
the secant lines appears to be approaching 4. Likewise, if we take xs to the left of 1 and move
them in very close to 1, the slope of the secant lines again appears to be approaching 4. So we
will estimate that the slope of the tangent line is also 4.
Finally, the equation of the tangent line to of
2
2 15 ) ( x x f at 1 x is
) 1 ( 4 13 x y
or
. 17 4 + x y
There are a couple of important points to note. First, never assume that what is happening on
one side of a point will also be happening on the other side as well. Secondly, we should never
try to determine a trend based on a couple of points that arent really all that close to the point
in question. Thirdly, while we couldnt actually plug 1 x into the formula of the secant slope,
we were able to determine some information about what was happening at 1 x by simply
looking at what was happening around . 1 x
2
The next problem we are going to consider is the rate of change problem. Consider a function
f
that represents some quantity that varies as x varies. What we want to do here is determine
just how fast
f
is changing at some point
. a x
This is called the instantaneous rate of
change of
f
at a.
At this point, all we can do is to estimate the rate of change. If we think of
f
as something
that is changing in time and x being the time measurement, we can find the average rate of
change. To compute the average rate of change of
f
at a, all we need is another point, say x,
and we compute
.
) ( ) (
in Change
in Change
A.R.C
a x
a f x f
x
f

Then to estimate the instantaneous rate of change at a, all we need to do is to choose values
of x getting closer and closer to a from both sides and compute the values of ARC.
Example 2: Suppose the amount of air in a balloon after t hours is given by
. 35 6 ) (
2 3
+ t t t V Estimate the instantaneous rate of change of the volume after 5 hours.
Solution:
.
5
25 6
5
) 5 ( ) (
in Change
in Change
A.R.C
2 3

t
t t
t
V t V
t
V
Here is a table of values of t that are getting closer and closer to 5 from both sides, and the
average rate of change for those values.
t A.R.C. t A.R.C.
6 25.0 4 7.0
5.5 19.75 4.5 10.75
5.1 15.91 4.9 14.11
5.01 15.0901 4.99 14.9101
5.001 15.009001 4.999 14.991001
5.0001 15.00090001 4.9999 14.99910001
From this table it looks like the average rate of change is approaching 15, so we can estimate
that the instantaneous rate of change is 15 at . 5 t
So, just what does this tell us about the volume at this point? We have estimated that at
, 5 t
the volume is changing at a rate of 15 cm
3
/hr. The rate being positive means that the balloon is
increasing in volume at the instant . 5 t But this doesnt mean the balloon is increasing in
volume all the time.
3
2.2. The Limit
Definition: We say that the limit of
f
is L as x approaches a, and write this as
L x f
a x

) ( lim
provided we can make
) (x f
as close to L as we want for all x sufficiently close to a, from both
sides, without actually letting x be a.
This is not the exact, precise definition of a limit. The precise mathematical definition of a limit
is given in at the end of this topic. So just what does this definition mean?
Suppose we know that the limit exists. According to our definition, we can decide how close
to L that we would like to make
). (x f
Now suppose we want to make
) (x f
no more that
0.001 away from L. This means that we want one of the following:
, 001 . 0 ) ( < L x f
if
L x f > ) (
, 001 . 0 ) ( < x f L
if
L x f < ) (
Now according to our definition, if we get x sufficiently close to a, we can make one of the
statements above true. However, this interpretation is inaccurate. It actually says that there is a
value of x, say X, so that for all xs that are closer to a than X, one of the statements above will
be true.
It is important to note that we must look at values of x that are on both sides of
. a x
We also
note that we are not allowed to use
a x
in the definition. We use the information that limits
give us to get information about what is going around
, a x
but the limit itself is not
concerned with what is actually going on at
. a x

Example 1: Estimate the value of the limit .
2
12 4
lim
2
2
2
x x
x x
x

+

Solution: We choose values of x that are getting closer and closer to 2 x and plug these
values into the function. This gives the following table of values:
x ) (x f x ) (x f
2.5 3.4 1.5 5.0
2.1 3.857142857 1.9 4.157894737
2.01 3.985074627 1.99 4.015075377
2.001 3.998500750 1.999 4.001500750
2.0001 3.999850007 1.9999 4.000150008
2.00001 3.999985000 1.99999 4.000015000
Notice that we cannot plug in 2 x into the function as this would give us a division by zero
error. This is not a problem since the limit doesnt care what is happening at . 2 x
4
From the table, it appears that
) (x f
is going to 4 as x goes to 2. We guess that
. 4
2
12 4
lim
2
2
2

x x
x x
x
The graph of the function in the range of xs that we are interested in is shown below.
The two arrows on the graph that are moving in towards 2 x show us what y value is
) (x f

approaching. In our case we can see that as x moves in towards 2 (from both sides), the
function is approaching
4 y
even though the function itself is not defined at . 2 x
Example 2: Estimate the value of the limit
), ( lim
2
x g
x
where

'

2 , 6
2 ,
2
12 4
) (
2
2
x
x
x x
x x
x g
Solution: First note that
. 6 ) 2 ( g
However,
. 4 ) ( lim
2

x g
x

We sometimes see situations like the previous example where the limit at a point and the
function value at a point are different. But there are times when the function value and the
5
limit at a point are equal and we will eventually see some examples of those. It is important
however, to not get excited about things when the function and the limit do not take the same
value at a point.
Example 3: Estimate .
cos 1
lim
0
x
x
x

Solution: First note that if we plug in

, 0 x
we get 0/0. Building a table of values like
Example 1 gives us
x
) (x f
x
) (x f
1 0.45969769 -1 -0.45969769
0.1 0.04995835 -0.1 -0.04995835
0.01 0.00499996 -0.01 -0.00499996
0.001 0.00049999 -0.001 -0.0004999
From the table, we guess . 0
cos 1
lim
0

x
x
x

Example 4: Estimate
.

cos lim
0

,
_

t
t
Solution: Lets build a table of values.
t
) (t f
t
) (t f
1 -1 -1 -1
0.1 1 -0.1 1
0.01 1 -0.01 1
0.001 1 -0.001 1
We are tempted to guess that the limit is 1. However, this answer is wrong. Consider the
following function evaluations:
1
2001
1

,
_

f 0
2001
2

,
_

f
2
2
4001
4

,
_

f
In all three evaluations, we evaluated the function at a number that is less that 0.001 and got
three totally different numbers! To see what is happening, a graph of the function would be
convenient.
6
From this graph, we can see that as we move in towards
, 0 t
the function starts oscillating
wildly and in fact the oscillations increases in speed the closer to 0 that we get. This function
clearly does not settle towards a single number, so this limit does not exist!
Example 4: Estimate the value of the limit
), ( lim
0
t h
t
where

'

<

0 , 0
0 , 1
) (
x
x
t h
Solution: Below is the graph of this function.
We can see that as t approaches 0 from the right, the function moves in towards 1. And as t
approaches 0 from the left, the function moves in towards 0. According to our definition of the
limit, the function needs to move in towards a single value as t move in towards 0. This isnt
happening here and so in this example we will also say that the limit doesnt exist.
7
2.3. One-sided Limits
Definition: Right-handed limit
We say
L x f
a x

+

) ( lim
provided we can make
) (x f
as close to L as we want for all
a x >

sufficiently close to a, and without actually letting x be a.
Definition: Left-handed limit
We say
L x f
a x

) ( lim
provided we can make
) (x f
as close to L as we want for all
a x <

sufficiently close to a, and without actually letting x be a.
The difference between normal and one-sided limits is that the function only needs to settle
down to a single number on either the right side or left side of
a x
depending on the one-
sided limit we are dealing with.
Example 1: Estimate
) ( lim
0
t h
t

and
), ( lim
0
t h
t
+

where

'

<

0 , 0
0 , 1
) (
x
x
t h
Solution:
0 ) ( lim
0

t h
t
and
. 1 ) ( lim
0

+

t h
t
So
) ( lim
0
t h
t
does not exist.
Example 2: Estimate
,
_

t
t

cos lim
0
and
.

cos lim
0

,
_

t
t
Solution: From the graph of this problem, both of the one-sided limits suffer the same problem
as the normal limit. The function does not settle down to a single number on either side of
. 0 t Therefore, both of the one-sided limits does not exist.
Theorem: Given a function
, f

L x f
a x

) ( lim
if and only if
. ) ( lim ) ( lim L x f x f
a x a x

+

8
Example 3: Given the following graph of
, f
compute each of the following.
a)
) 4 ( f
b)
) ( lim
4
x f
x


c)
) ( lim
4
x f
x
+

d)
) ( lim
4
x f
x
e)
) 1 ( f
f)
) ( lim
1
x f
x

g)
) ( lim
1
x f
x

h)
) ( lim
1
x f
x
i)
) 6 ( f
j)
) ( lim
6
x f
x


k)
) ( lim
6
x f
x
+

l)
) ( 6 x f
2.4. Limit Properties
We now introduce some properties of limits that help us to compute limits without building a
table of values.
Assume that
M x f
a x

) ( lim
and
N x g
a x

) ( lim
and that c is any constant. Then the following
properties are true.
1.
cM x cf
a x

)] ( [ lim
2.
N M x g x f
a x
t t

)] ( ) ( [ lim
3.
MN x g x f
a x

)] ( ) ( [ lim
4.
, / )] ( / ) ( [ lim N M x g x f
a x

provided 0 N
5.
, )] ( [ lim
n n
a x
M x f

where n is any real number

6.
n
n
a x
M x f

) ( lim
(this is a special case of the property 5)
7.
c c
a x

lim
8.
a x
a x

lim
(easily verified by graphing the line
x y
)
9.
n n
a x
a x

lim
(this is a special case of property 5 using
x x f ) (
)
Note that these properties also hold for one-sided limits as well. The proofs of these properties
require the precise mathematical definition of limits which we will learn later.
9
Example 1: Compute the limit
). 9 5 3 ( lim
2
2
+

x x
x
Solution:
9 lim 5 lim 3 lim ) 9 5 3 ( lim
2 2
2
2
2
2
+ +
x x x x
x x x x
Property 2
9 lim lim 5 lim 3
2 2
2
2
+
x x x
x x
Property 1
9 ) 2 ( 5 ) 2 ( 3
2
+ Property 7 to 9
. 7

For Example 1, observe that if we had defined , 9 5 3 ) (
2
+ x x x p then
). 2 ( ) ( lim
2


p x p
x
In
other words, we see that the limit is the value that we would get by just evaluating the function
at the point in question. This seems to violate one of the main concepts about limits that weve
seen to this point. Previously, we made a big deal about the fact that limits do not care about
what is happening at the point in question. They only care about what is happening around the
point. Despite appearances, the limit still doesnt care about what the function is doing
at . 2 x In this case, the function that weve got is simply nice enough so that what is
happening around the point is exactly the same as what is happening at the point. Later we will
formalize just what is meant by nice enough.
In general, if
) (x f
is nice enough, then
). ( ) ( lim a f x f
a x

Here are some functions that are

nice enough:
Polynomials are nice enough for all xs.

) ( / ) ( ) ( x q x p x f
is nice enough if p and q are nice enough, and we dont get division
by zero at the point we are evaluating.
cos x and sin x are nice enough for all xs.
sec x and tan x are nice enough everywhere that cos x is not zero.
csc x and cot x are nice enough everywhere that sin x is not zero.

n
x is nice enough for all xs if n is odd.

n
x is nice enough for 0 x if n is even.

x
b is nice enough for all xs ( 1 b is a positive real number).

x
b
log
is nice enough for x > 0.
Any algebraic combinations of the functions above will also be nice enough (as long as
the function value at the point of interest is defined).
Example 2:
.
6
13
1 ) ( 7 ) 1 ( 2
) 1 ( 10 ) 1 ( 3 6
1 7 2
10 3 6
lim
3 4
2
3 4
2
1

+ +
+

+ +
+

x x x
x x
x

Example 3:
. 1854 . 8 3 cos 3 sin
3 ln 1
3 cos sin
ln 1
lim
3
5 5
3
+ +
+
+

,
_

+ +
+
+

e
x x
x
e
x
x
x

10
2.4. Computing Limits
Example 1: Evaluate .
2
12 4
lim
2
2
2
x x
x x
x

+

Solution:

) 2 (
) 6 )( 2 (
lim
2
12 4
lim
2
2
2
2

+

+

x x
x x
x x
x x
x x
Factorize
x
x
x
) 6 (
lim
2
+

Cancel 2 x (we can do so because 2 x )

. 4
2
6 2

If we plug in 2 x without factorizing and cancelling, we get 0/0. This is called an

indeterminate form.
Example 2: Evaluate .
18 ) 3 ( 2
lim
2
0
h
h
h

Solution:
h
h h
h
h h
h
h h
h
h
h h h h
12 2
lim
18 18 12 2
lim
18 ) 9 6 ( 2
lim
18 ) 3 ( 2
lim
2
0
2
0
2
0
2
0

. 12 ) 12 2 ( lim
0

h
h

Example 3: Evaluate .
4
4 3
lim
4
t
t t
t

+

Solution:
) 4 3 (
) 4 3 (
4
) 4 3 (
lim
4
4 3
lim
4 4
+ +
+ +

+

t t
t t
t
t t
t
t t
t t
Rationalize the numerator
) 4 3 )( 4 (
4 3
lim
2
4
+ +
+

t t t
t t
t
) 4 3 )( 4 (
) 1 )( 4 (
lim
4
+ +
+

t t t
t t
t
Factorize
) 4 3 )( 4 (
) 1 )( 4 (
lim
4
+ +
+

t t t
t t
t
Rewrite the factor in the denominator
4 3
1
lim
4
+ +
+

t t
t
t
Cancel the common factor
.
8
5

11
Example 4: Given

'

< +

2 , 3 1
2 , 5
) (
2
y y
y y
y g
, compute the limits
) ( lim
6
y g
y
and
). ( lim
2
y g
y
Solution:
17 ) 3 1 ( lim ) ( lim
6 6


y y g
y y
since the values on both sides of
6 y
are greater than 2.
9 5 ) 2 ( ) 5 ( lim ) ( lim
2 2
2 2
+ +


y y g
y y
since

2 y implies
. 2 < y
7 ) 2 ( 3 1 ) 3 1 ( lim ) ( lim
2 2

+ +

y y g
y y
since
+
2 y implies
. 2 > y
Since the one-sided limits are not equal, then the normal limit does not exist.
Note that a very simple change to the function in Example 4 will make the limit at
2 y

exist, so dont think that limits at cutoff points in piecewise functions dont ever exist.
Example 5: Given

'

< +

2 , 3 3
2 , 5
) (
2
y y
y y
y g
, compute the limit
). ( lim
2
y g
y
Solution:
9 5 ) 2 ( ) 5 ( lim ) ( lim
2 2
2 2
+ +


y y g
y y
since

2 y implies
. 2 < y
9 ) 2 ( 3 3 ) 3 3 ( lim ) ( lim
2 2

+ +

y y g
y y
since
+
2 y implies
. 2 > y
Therefore,
. 9 ) ( lim
2


y g
y

Property: If
) ( ) ( x g x f
for all x on
] , [ b a
(except possibly at
c x
) and
, b c a
then
). ( lim ) ( lim x g x f
c x c x

Squeeze Theorem: Suppose that

) ( ) ( ) ( x g x h x f
for all x on
] , [ b a
(except possibly at
c x
) and . b c a If
, ) ( lim ) ( lim L x g x f
c x c x


then
. ) ( lim L x h
c x

If we assume that all three functions

, f
g, and h are nice enough (not required by the
theorem), then following figure illustrates what is happening in the Squeeze Theorem.
12
Example 6: Evaluate .
1
cos lim
2
0
x
x
x
Solution: Firstly, we know the following fact about cosine:
. 1 cos 1 x
Since we are taking the limit at
, 0 x
then
, 0 x
and we can say
. 1 ) / 1 cos( 1 x
Multiplying by ,
2
x we get
. ) / 1 cos(
2 2 2
x x x x
Since
, 0 lim lim
2
0
2
0


x x
x x
then by the Squeeze Theorem, . 0
1
cos lim
2
0

x
x
x

We can verify Example 6 graphically as follows:
13
2.5. Infinite Limits
Definition: We say

) ( lim x f
a x
if we can make
) (x f
as large as we want for all x
sufficiently close to a from both sides, without actually letting
. a x
Definition: We say

) ( lim x f
a x
if we can make
) (x f
as large as we want and negative for
all x sufficiently close to a from both sides, without actually letting
. a x
These definitions can be appropriately modified for the one-sided limits as well.
Example 1: Evaluate each of the following limits:
x
x
1
lim
0

x
x
1
lim
0
+

x
x
1
lim
0
Solution:
In all three cases notice that we cant just plug in . 0 x We proceed by plugging in some
points and see what value the function is approaching. Although we said that we were no
longer going to do this, but in this case it is a good way to illustrate just whats going on with
this function.
x x / 1 x x / 1
-0.1 -10 0.1 10
-0.01 -100 0.01 100
-0.001 -1000 0.001 1000
-0.0001 -10000 0.0001 10000
From the table of values, we can estimate the value of the two one-sided limits and once we
have done that, we can use the fact that the normal limit will exist only if the two one-sided
limits exist and have the same value.
We can make the function as large and positive as we want for all xs sufficiently close to zero
while staying positive (i.e. on the right). Likewise, we can make the function as large and
negative as we want for all xs sufficiently close to zero while staying negative (i.e. on the
left). So, we have

x
x
1
lim
0

+

x
x
1
lim
0
Here is a quick sketch of the graph of 1/x. Although we mentioned that we wont do this too
often as most functions are not something we can just quickly sketch out, this graph is easy to
sketch and accuracy will not be an issue.
14
We can see from this graph that the function does behave much as we predicted that it would
from our table values. Finally, the normal limit will not exist since the two one-sided have
different values.
Example 2: Evaluate each of the following limits:
2
0
6
lim
x
x

2
0
6
lim
x
x
+

2
0
6
lim
x
x
Solution: Lets take a look at the right-hand limit first. As we take smaller and smaller values
of x, while staying positive, squaring them will only make them even smaller (recall squaring a
number between zero and one will make it smaller) and of course it will stay positive. So we
have a positive constant divided by an increasingly small positive number. The result should
then be an increasingly large positive number. Therefore,

+

2
0
6
lim
x
x
Now, lets take a look at the left hand limit. As we take smaller and smaller values of x, while
staying negative, squaring them will only make them smaller as well, and upon squaring, the
result is now positive. So, we have a positive constant divided by an increasingly small
positive number. The result, as with the right hand limit, will be an increasingly large positive
number. Therefore,

2
0
6
lim
x
x
It follows that the normal limit will exist and be infinity since the two one-sided limits both
exist and have the same value. Here is a quick graph verifying the limits.
15
Example 3: Evaluate each of the following limits:
2
4
lim
2
+

x
x
2
4
lim
2
+

+

x
x
2
4
lim
2
+

x
x
Solution: Lets start with the right-hand limit. We know that
. 0 2 2 2 > + >
+
x x x
As x gets closer and closer to 2, 2 + x will be getting closer and closer to zero while staying
positive. So well have a negative constant divided by an increasingly small positive number.
The result will be an increasingly large and negative number. Therefore,
.
2
4
lim
2

+

+

x
x
For the left-hand limit, we have
. 0 2 2 2 < + <

x x x
As x gets closer and closer to 2, 2 + x will be getting closer and closer to zero while staying
negative. So well have a negative constant divided by an increasingly small negative number.
The result will be an increasingly large and positive number. Therefore,
.
2
4
lim
2

+

x
x
Finally, since two one sided limits are not the same the normal limit wont exist. Here is a
quick graph of the function for verification purposes.
16
At this point we should briefly acknowledge the idea of vertical asymptotes. Each of the three
previous graphs has one. A vertical asymptote is a vertical line in which the graph will go
towards infinity and/or minus infinity on one or both sides of the line. So for Example 3, the
vertical asymptote is the line . 2 x
Definition: The function
f
will have a vertical asymptote at
a x
if we have any of the
following limits at
. a x
t

) ( lim x f
a x
t
+

) ( lim x f
a x
t

) ( lim x f
a x
Example 4: Evaluate each of the following limits:
3
2
lim
3

x
x
x
3
2
lim
3

+

x
x
x
3
2
lim
3

x
x
x
Solution: In this example, the numerator is not a constant as in the previous examples. Lets
start with the right-hand limit. We have
0 3 3 3 > >
+
x x x for the denominator
6 2 3
+
x x for the numerator
So, as we let x get closer and closer to 3 from the right, the numerator, while not a constant, is
getting closer and closer to a positive constant while the denominator is getting closer and
closer to zero, and will be positive since we are on the right side. So the result should be an
increasingly larger positive number. The right-hand limit should then be positive infinity.
Similarly, we can justify that the left-hand limit is negative infinity. Finally, we have

3
2
lim
3
x
x
x

3
2
lim
3
x
x
x
3
2
lim
3

x
x
x
D.N.E
Here is a quick graph for verification purposes.
17
Lets generalize what we have developed from the examples. Given the functions
f
and
, g

suppose we have

) ( lim x f
a x
and
L x g
a x

) ( lim
for some real numbers a and L. Then

)] ( ) ( [ lim x g x f
a x

) ( ) ( lim x g x f
a x
if L > 0

) ( ) ( lim x g x f
a x
if L < 0

0
) (
) (
lim

x f
x g
a x
Note that the above set of facts also holds for one-sided limits. They will also hold if

) ( lim x f
a x
with a change of sign on the infinities in the first three parts.
2.6. Limits at Infinity
We will now look at these limits:
) ( lim x f
x
and
). ( lim x f
x
For many of the limits that were
going to be looking at, we will need the following two facts.

, 0 lim

r
x
x
c
where c is any real number and r is a positive rational number

, 0 lim

r
x
x
c
where c is any real number, r is a positive rational number, and
r
x is
defined for x < 0.
Note that the sign of c will not affect the answer. Regardless of the sign of c well still have a
constant divided by a very large number which will result in a very small fraction, and the
larger x gets, the smaller the fraction gets. The sign of c will affect which direction the fraction
approaches zero (i.e. from the positive or negative side) but it still approaches zero.
18
Example 1: Evaluate each of the following limits.
a)
) 8 2 ( lim
2 4
x x x
x


b)
,
_

+ +

8 2
3
1
lim
2 3 5
x x x
x
Solution (a): If we just plug in, we get
. ) 8 2 ( lim
2 4


x x x
x
We are probably
tempted to say that the answer is zero or
,
but we have no way to be sure. This is one of
those indeterminate forms.
So, we need a way to get around this problem. What well do here is factor the largest power
of x out of the whole polynomial:
. 2
8 1
2 lim lim
8 1
2 lim ) 8 2 ( lim
3 2
4
3 2
4 2 4

,
_

1
]
1

,
_

x x
x
x x
x x x x
x x x x
Solution (b): Similarly, factoring the largest power of x out of the whole polynomial gives us
.
3
1 8 1 2
3
1
lim 8 2
3
1
lim
5 3 2
5 2 3 5

1
]
1

,
_

+ +

,
_

+ +

x x x
x x x x
x x

In general, if
0 1
1
1
) ( a x a x a x a x p
n
n
n
n
+ + + +

is a polynomial of degree n (
0
n
a
), then
, lim ) ( lim
n
n
x x
x a x p

. lim ) ( lim
n
n
x x
x a x p

Example 2: Evaluate each of the following limits.

a)

,
_

4
2 4
5 7
8 2
lim
x
x x x
x
b)

,
_

4
2 4
5 7
8 2
lim
x
x x x
x
Solution (a): If we just plug in, we get
.
5 7
8 2
lim
4
2 4

,
_

x
x x x
x
We are probably tempted
to say that the answer is
,
zero, or 1, but we have no way to be sure. This is one of
those indeterminate forms as well.
What well do is identify the largest power of x in the denominator (and yes, we only look at
the denominator for this) and then factor this out of both the numerator and denominator.
Doing this for the first limit gives
.
5
2
5 0
0 0 2
5 / 7
/ 8 / 1 2
lim
) 5 / 7 (
) / 8 / 1 2 (
lim
5 7
8 2
lim
4
3 2
4 4
3 2 4
4
2 4

,
_

1
]
1

,
_

x
x x
x x
x x x
x
x x x
x x x
19
Solution (b): Similar factoring gives us
.
5
2
5 0
0 0 2
5 / 7
/ 8 / 1 2
lim
) 5 / 7 (
) / 8 / 1 2 (
lim
5 7
8 2
lim
4
3 2
4 4
3 2 4
4
2 4

,
_

1
]
1

,
_

x
x x
x x
x x x
x
x x x
x x x
Notice that it doesnt matter which infinity we are going towards, we get the same value for the
limit. However, in Example 1, the direction of infinity affects the answer.
Example 3: Evaluate each of the following limits.
a)

,
_

+

x
x
x
2 5
6 3
lim
2
b)

,
_

+

x
x
x
2 5
6 3
lim
2
Solution (a): The largest power of x in the denominator is x. So we have
.
) 2 / 5 (
) / 6 3 (
lim
) 2 / 5 (
) / 6 3 (
lim
2 5
6 3
lim
2 2 2 2
2
1
1
]
1

1
1
]
1

,
_

+

x x
x x
x x
x x
x
x
x x x
This is where we need to be really careful with the square root in the problem. We write
.
) 2 / 5 (
) / 6 3 ( | |
lim
) 2 / 5 (
) / 6 3 (
lim
2 2 2
1
1
]
1

1
1
]
1

+

x x
x x
x x
x x
x x
Now we cant just cancel out x because of the bars. Recall that
.
0 ,
0 ,
| |

'

<

x x
x x
x
Since we are
going to infinity (which is positive), we can just drop the bars. Now we have
.
2
3
2 0
0 3
2 / 5
) / 6 3 (
lim
) 2 / 5 (
) / 6 3 (
lim
) 2 / 5 (
) / 6 3 ( | |
lim
2 2 2

,
_

1
1
]
1

1
1
]
1

+

x
x
x x
x x
x x
x x
x x x
Solution (b): Similar factoring gives us
.
2
3
2 0
0 3
2 / 5
) / 6 3 (
lim
) 2 / 5 (
) / 6 3 (
lim
) 2 / 5 (
) / 6 3 ( | |
lim
2 5
6 3
lim
2
2 2 2

,
_

+

1
1
]
1

1
1
]
1

,
_

+


x
x
x x
x x
x x
x x
x
x
x
x x x
Note that when we dropped the bars, we wrote x since we are going to minus infinity.
20
Definition: A function
f
will have a horizontal asymptote at
L y
if either one of the
following is true:
L x f
x


) ( lim
or
. ) ( lim L x f
x


Here is a graph showing horizontal asymptotes at
1 y
and
, 1 y
as well as a vertical
asymptote at . 2 / 5 x
Example 4:
.
5 0
0
5 / 1
/ 4
lim
) 5 / 1 (
) / 4 (
lim
5 1
4
lim
3
3
3 3
3 3
3
6 2

+

x
x x
x x
x x x
x
x x
x x x
.
5 0
0
5 / 1
/ 4
lim
) 5 / 1 (
) / 4 (
lim
5 1
4
lim
3
3
3 3
3 3
3
6 2

+

x
x x
x x
x x x
x
x x
x x x
. 0
0 2
0 0 0
/ 3 2
/ 9 / 5 / 1
lim
) / 3 2 (
) / 9 / 5 / 1 (
lim
3 2
9 5
lim
4 3 2
4
4 3 2 4
3 4
2

+


x
x x x
x x
x x x x
x x
x x
x x x

Example 5: These limits can be easily verified.
. lim

x
x
e
. 0 lim

x
x
e
. 0 lim

x
x
e
. lim

x
x
e

Example 6:
2
8 4 2
lim
x x
x
e


The limit in the exponent is
. 8 ) 8 / 4 / 2 ( lim 8 4 2 lim
2 2 2


x x x x x
x x
Let . 8 4 2
2
x x y Since the exponent goes to minus infinity as x goes to infinity, then
. 0 lim lim
2
8 4 2




y
y
x x
x
e e

21
Example 7:
x
x
e
/ 1
0
lim
+

The limit in the exponent is

. / 1 lim
0

+

x
x
Let
. / 1 x y
Since the exponent goes to infinity as x goes to zero from the right, then
. lim lim
/ 1
0


+
y
y
x
x
e e

Example 8:
) 9 2 3 4 ( lim
15 2 6 10 x x x x x
x
e e e e e


+ +
Plugging in gives us
0 0 ) 9 2 3 4 ( lim
15 2 6 10
+ + + +


x x x x x
x
e e e e e
which is an
indeterminate form.
What we do now is factor out the function with the largest exponent that is causing problems.
The function with the largest exponent is .
10x
e Factorizing gives us
) 9 2 3 4 1 ( lim ) 9 2 3 4 ( lim
25 12 9 4 10 15 2 6 10 x x x x x
x
x x x x x
x
e e e e e e e e e e




+ + + +
Now we have
. ) 0 0 0 0 1 ( ) 9 2 3 4 1 ( lim
25 12 9 4 10
+ + + +


x x x x x
x
e e e e e

Example 9:
x x x
x x
x
e e e
e e

3 2 8
2 6
lim
2 4
2 4
As with the previous example, we look at the function with the largest positive exponent in the
denominator and we factor out this function from the numerator and denominator.
.
4
3
0 0 8
0 6
3 2 8
2 6
lim
) 3 2 8 (
) 2 6 (
lim
3 2 8
2 6
lim
5 2
6
5 2 4
6 4
2 4
2 4

x x
x
x
x x x
x x
x
x x x
x x
x
e e
e
e e e
e e
e e e
e e

Example 10: These limits can be easily verified.
. ln lim
0

+

x
x
. ln lim

x
x

Example 11:
) 1 7 ln( lim
2 3
+

x x
x
The limit in the argument is
+

) 1 7 ( lim
2 3
x x
x
Let . 1 7
2 3
+ x x y Since the argument goes to infinity as x goes infinity, then
. ln lim ) 1 7 ln( lim
2 3
+

y x x
y x

22
Example 12:
,
_

x x
x
5
1
ln lim
2
Let .
5
1
2
x x
y

Since , 0
0 1
0
/ 5 1
/ 1
lim lim
2

x
x
y
x x

then
. ln lim
5
1
ln lim
0
2

,
_

+

y
x x
y x
This is because as x goes to negative infinity, it values are all negative, and so y is always
positive, thus y approaches 0 from the right.
2.7. Continuity
Definition 1: A function
f
is continuous at
a x
if
). ( ) ( lim a f x f
a x

Definition 2: A function
f
is continuous from the left at
a x
if
). ( ) ( lim a f x f
a x

A function
f
is continuous from the right at
a x
if
). ( ) ( lim a f x f
a x

+

Definition 3: A function
f
is continuous on the interval I if it is continuous at all points of
. I If I contains endpoints, then
f
is continuous from the left or right at the endpoints.
Example 1: Given the graph of
f
below, determine if
f
is continuous at
, 2 x

, 0 x
and
. 3 x
Solution:
f
is discontinuous at 2 x (jump discontinuity).
f
is continuous at . 0 x
f
is discontinuous at 3 x (removable discontinuity).
23
From this example, we can make the observation that a function is continuous if its graph has
no holes or breaks in it.
Example 2: Determine where the function
15 2
10 4
) (
2

+

t t
t
t h is not continuous.
Solution: Rational functions are continuous everywhere except where we have division by
zero. The denominator is zero when
, 3 t
and . 5 t
Theorem 1: If
f
and
g
are continuous at
, a x
then the following are also continuous at
. a x
Assume c is a constant and n is a positive integer.

g f t

cf

fg

, / g f
provided
0 ) ( a g

n
x f )] ( [
Theorem 2:
Polynomial functions are continuous on R.
Rational functions of the form
q p /
are continuous for all x for which
. 0 ) ( x q
Trigonometric functions are continuous on their domains.
Exponential functions of the form
x
b are continuous on R.
Logarithmic functions are continuous on
). , 0 (
Theorem 3: If
g
is continuous at
a
and
f
is continuous at
), (a g
then the composite function
g f
is continuous at
. a
)). ( ( )) ( ( lim a g f x g f
a x

Example 3: Evaluate the limit

. lim
sin
0
x
x
e

Solution: Since we know that exponential and sine functions are continuous everywhere, then
. 1 lim
0 0 sin sin
0

e e e
x
x

Example 4: Evaluate the limits (a)
1 cos
1 cos
lim
2
0

x
x
x
and (b)
. ln lim
0
x
x
Solution:
(a) . 2 1 cos lim
1 cos
) 1 )(cos 1 (cos
lim
1 cos
1 cos
lim
0 0
2
0
+

x
x
x x
x
x
x x x
(b) x ln is continuous on
). , 0 (
However, 0 ln x only for
). , 1 [
This means x ln is
undefined for . 1 < x So
x
x
ln lim
0
does not exist.
24
Intermediate Value Theorem: Suppose that
f
is continuous on
] , [ b a
and let M be any
number between
) (a f
and
). (b f
Then there exists at least one number c such that

b c a < <

M c f ) (
Below is the graph of a continuous function that illustrates the Intermediate Value Theorem.
If we pick any value of M that is between the value of
) (a f
and the value of
), (b f
and draw a
line straight out from this point, the line will cut the graph at least once.
Example 5: Show that 5 10 5 2 ) (
2 3
+ x x x x p has a root somewhere in
]. 2 , 1 [
Solution: We want to find c such that 2 1 < < c and
. 0 ) ( c p
This is actually the conclusion
of the Intermediate Value Theorem. So all we need to do is to set up the conditions to use the
theorem.

p
is a polynomial so it is continuous everywhere.

8 ) 1 ( p
and
, 19 ) 2 ( p
so
). 1 ( 0 ) 2 ( < < p p

2.7. The Precise Definition of Limits
Definition 1: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that
L x f
a x

) ( lim
if for every number
, 0 >
there is a number 0 > such that
. | ) ( | | | 0 < < < L x f a x
Just what does it mean? Lets take a look at the following graph.
25
For any number
, 0 >
we can draw two horizontal lines + L and . L Then there is a
number
, 0 >
which we need to find, that will allow us to draw two vertical lines + a and
. a Observe that if we pick any x between + a and
, a
but not at a, this means
. | | 0 < < a x
Then
) (x f
will lie between + L and
, L
which means
. | ) ( | < L x f
We can actually choose from an infinite number of possible . The idea is to choose to
make the horizontal yellow region as thin as we like. Moreover, we are only interested in
values of
f
at points close to a, but not at a. So we dont care if
) (a f
is defined or not.
Example 1: Use the definition of the limit to prove
. 0 lim
2
0

x
x
Solution: . 0 a L Let 0 > be any number. We want to find a number 0 > such that
. | | | | 0
2
< < < x x
One way to find is to look at the right hand expression and do some simplification:
. | | when | |
2
< < x x
So is a possible . Lets check. We start from the left hand expression and work to get the
right hand expression:
. | | | | | | 0
2 2
< < < < x x x
Example 2: Use the definition of the limit to prove
. 6 4 5 lim
2

x
x
Solution: 6 L and . 2 a Let 0 > be any number. We want to find a number 0 > such
that
. | 6 ) 4 5 ( | | 2 | 0 < < < x x
Look at the right hand expression and do some simplification:
.
5
| 2 | when | 2 || 5 | when | ) 2 ( 5 | when | 10 5 | when | 6 ) 4 5 ( |

< < < < < x x x x x
So 5 / is a possible . Lets check. We start from the left hand expression and work to get
the right hand expression:
. | 10 5 | | ) 2 ( 5 | | 2 | 5
5
| 2 | 0

< < < < < x x x x

26
Example 3: Use the definition of the limit to prove
. 9 11 lim
2
4
+

x x
x
Solution: Let 0 > be any number. We want to find a number 0 > such that
. | 20 | | 4 | 0
2
< + < < x x x
Look at the right hand expression and do some simplification:
. | 5 || 4 | when | ) 5 )( 4 ( | when | 20 |
2
< + < + < + x x x x x x
The problem here is the term
. | 5 | + x
Note that whatever number we choose for
,
it must be
small. Suppose we assume
. 1 | 4 | < x
Then
. 10 5 8 5 3 1 4 1 1 | 4 | < + < < < < < < x x x x
So assuming
1 | 4 | < x
gives us
. 10 | 5 | 10 5 10 10 5 8 < + < + < < + < x x x
Then we have
, | 5 || 4 | 10 | 4 | < + < x x x
which implies . | 5 || 4 |
10
| 4 |

< + < x x x Hence we choose to be the smaller of the

two assumptions we made,
1 | 4 | < x
and ,
10
| 4 |

< x and we write
). 1 , 10 / min(

Definition 2: For the right hand limit, we say that
L x f
a x

+

) ( lim
if for every number
, 0 >
there is a number 0 > such that
. | ) ( | ) or ( 0 < + < < < < L x f a x a a x
Definition 3: For the left had limit, we say that
L x f
a x

) ( lim
if for every number
, 0 >
there is a number 0 > such that
. | ) ( | ) or ( 0 < < < < < L x f a x a a x
Definition 4: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that

) ( lim x f
a x
if for every number
, 0 > M
there is a number 0 > such that
. ) ( | | 0 M x f a x > < <
Definition 5: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that

) ( lim x f
a x
if for every number
, 0 < N
there is a number 0 > such that
. ) ( | | 0 N x f a x < < <
27
Lets illustrate Definition 4.
What the definition is telling us is that no matter how large we choose M to be we can always
find an interval around
a x
so that as long as we stay within that interval, the graph of the
function will be above the line
. M y
Note as well that the larger M is, the smaller were
probably going to need to make .
Example 4: Use the definition of the limit to prove .
1
lim
2
0

x
x
Solution: Let 0 > M be any number. We want to find a number 0 > such that
. / 1 | | 0
2
M x x > < <
As before, look at the right hand expression and do some simplification:
.
1
| | when
1
when
1
2
2
M
x
M
x M
x
< < >
So
M / 1
is a possible . Lets check:
.
1 1 1
| |
1
| | 0
2
2 2
M
x M
x
M
x
M
x > < < < <

Definition 6: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that
L x f
x


) ( lim
if for every number
, 0 >
there is a number 0 > M such that
. | ) ( | < > L x f M x
Definition 7: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that
L x f
x


) ( lim
if for every number
, 0 >
there is a number 0 < N such that
28
. | ) ( | < < L x f N x
Lets illustrate Definition 6.
What the definition is telling us is that is that no matter how close to L we want to get
(mathematically this is given by
< | ) ( | L x f
for any chosen

), we can find another number

M such that the graph of the function to the right of M will be in the region between + L and
L (the shaded region above), Finally, note that the smaller we make
,
the larger well
probably need to make M.
Example 5: Use the definition of the limit to prove . 0
1
lim

x
x
Solution: Let 0 > be any number. We want to find a number 0 < N such that
. | / 1 | < < x N x
As before, look at the right hand expression and do some simplification:
.
1
or
1
when
1
| | when
| |
1
when
1

< > > < < x x x
x x
So / 1 is a possible . Lets check:
.
1
| |
1 1
| |
1
| |
1


< < > > <
x x
x x x
Note that when we take the absolute value of both sides, we change both sides from negative
numbers to positive numbers, so we have to change the direction of inequality as well.
Definition 8: Let
f
be a function defined on an interval that contains
, a x
except possibly
at
. a x
Then we say that


) ( lim x f
x
if for every number
, 0 > N
there is a number 0 > M such that
. ) ( N x f M x > >
The other three are almost identical except for the signs and the directions of inequality.
29
Definition 9: A function
f
is continuous at
a x
if for every number
, 0 >
there is a
number 0 > such that
. | ) ( ) ( | | | < < a f x f a x
30