Chapter 2

2
Limits and Continuity
OVERVIEW Mathematicians of the seventeenth century were keenly interested in the

study of motion for objects on or near the earth and the motion of planets and stars. This
study involved both the speed of the object and its direction of motion at any instant, and
they knew the direction at a given instant was along a line tangent to the path of motion.
The concept of a limit is fundamental to finding the velocity of a moving object and the
tangent to a curve. In this chapter we develop the limit, first intuitively and then formally.
We use limits to describe the way a function varies. Some functions vary continuously;
small changes in x produce only small changes in ƒ(x). Other functions can have values
that jump, vary erratically, or tend to increase or decrease without bound. The notion of
limit gives a precise way to distinguish between these behaviors.
2.1 Rates of Change and Tangents to Curves

Calculus is a tool that helps us understand how a change in one quantity is related to a
change in another. How does the speed of a falling object change as a function of time?
How does the level of water in a barrel change as a function of the amount of liquid poured
into it? We see change occurring in nearly everything we observe in the world and universe,
and powerful modern instruments help us see more and more. In this section we introduce
the ideas of average and instantaneous rates of change, and show that they are closely
related to the slope of a curve at a point P on the curve. We give precise developments of
these important concepts in the next chapter, but for now we use an informal approach so
you will see how they lead naturally to the main idea of this chapter, the limit. The idea of
a limit plays a foundational role throughout calculus.
Average and Instantaneous Speed

HISTORICAL BIOGRAPHY* In the late sixteenth century, Galileo discovered that a solid object dropped from rest (not
Galileo Galilei moving) near the surface of the earth and allowed to fall freely will fall a distance proportional
(1564–1642) to the square of the time it has been falling. This type of motion is called free fall. It assumes
negligible air resistance to slow the object down, and that gravity is the only force acting on
the falling object. If y denotes the distance fallen in feet after t seconds, then Galileo’s law is
y = 16t 2,
where 16 is the (approximate) constant of proportionality. (If y is measured in meters, the
constant is 4.9.)
A moving object’s average speed during an interval of time is found by dividing the
distance covered by the time elapsed. The unit of measure is length per unit time: kilome-
ters per hour, feet (or meters) per second, or whatever is appropriate to the problem at hand.
*To learn more about the historical figures mentioned in the text and the development of many major
elements and topics of calculus, visit www.aw.com/thomas.
41
42 Chapter 2: Limits and Continuity
EXAMPLE 1 A rock breaks loose from the top of a tall cliff. What is its average speed
(a) during the first 2 sec of fall?
(b) during the 1-sec interval between second 1 and second 2?
Solution The average speed of the rock during a given time interval is the change in
distance, ∆y, divided by the length of the time interval, ∆t. (Increments like ∆y and ∆t
are reviewed in Appendix 3, and pronounced “delta y” and “delta t.”) Measuring distance
in feet and time in seconds, we have the following calculations:
∆y 16(2)2 - 16(0)2 ft
(a) For the first 2 sec: = = 32 sec
∆t 2 - 0
∆y 16(2)2 - 16(1)2 ft
(b) From sec 1 to sec 2: = = 48 sec
∆t 2 - 1
We want a way to determine the speed of a falling object at a single instant t0, instead of
using its average speed over an interval of time. To do this, we examine what happens
when we calculate the average speed over shorter and shorter time intervals starting at t0 .
The next example illustrates this process. Our discussion is informal here, but it will be
made precise in Chapter 3.
EXAMPLE 2 Find the speed of the falling rock in Example 1 at t = 1 and t = 2 sec.
Solution We can calculate the average speed of the rock over a time interval 3 t0, t0 + h4 ,
having length ∆t = h, as
∆y 16(t0 + h)2 - 16t0 2

= . (1)
∆t h
We cannot use this formula to calculate the “instantaneous” speed at the exact moment t0
by simply substituting h = 0, because we cannot divide by zero. But we can use it to cal-
culate average speeds over increasingly short time intervals starting at t0 = 1 and t0 = 2.
When we do so, by taking smaller and smaller values of h, we see a pattern (Table 2.1).
TABLE 2.1 Average speeds over short time intervals 3 t0, t0 + h4

∆y 16(t0 + h)2 - 16t0 2
Average speed: =
∆t h
Length of Average speed over Average speed over

time interval interval of length h interval of length h
h starting at t0 = 1 starting at t0 = 2
1 48 80
0.1 33.6 65.6
0.01 32.16 64.16
0.001 32.016 64.016
0.0001 32.0016 64.0016
The average speed on intervals starting at t0 = 1 seems to approach a limiting value

of 32 as the length of the interval decreases. This suggests that the rock is falling at a speed
of 32 ft > sec at t0 = 1 sec. Let’s confirm this algebraically.
2.1 Rates of Change and Tangents to Curves 43
If we set t0 = 1 and then expand the numerator in Equation (1) and simplify, we find
that
∆y 16(1 + h)2 - 16(1)2 16(1 + 2h + h2) - 16

= =
∆t h h
32h + 16h2
= = 32 + 16h.
h
For values of h different from 0, the expressions on the right and left are equivalent and the
average speed is 32 + 16h ft>sec. We can now see why the average speed has the limiting
value 32 + 16(0) = 32 ft>sec as h approaches 0.
Similarly, setting t0 = 2 in Equation (1), the procedure yields
∆y
= 64 + 16h
∆t
for values of h different from 0. As h gets closer and closer to 0, the average speed has the
limiting value 64 ft > sec when t0 = 2 sec, as suggested by Table 2.1.
The average speed of a falling object is an example of a more general idea which we
discuss next.
Average Rates of Change and Secant Lines

Given any function y = ƒ(x), we calculate the average rate of change of y with respect to
y
y = f (x) x over the interval [x1, x2] by dividing the change in the value of y, ∆y = ƒ(x2) - ƒ(x1),
by the length ∆x = x2 - x1 = h of the interval over which the change occurs. (We use
Q(x 2, f(x 2 )) the symbol h for ∆x to simplify the notation here and later on.)
Secant
DEFINITION The average rate of change of y = ƒ(x) with respect to x over the
Δy
P(x1, f (x1)) interval [x1, x2] is
∆y ƒ(x2) - ƒ(x1) ƒ(x1 + h) - ƒ(x1)
Δx = h = x2 - x1 = , h ≠ 0.
∆x h
x
0 x1 x2
FIGURE 2.1 A secant to the graph Geometrically, the rate of change of ƒ over [x1, x2] is the slope of the line through the
y = ƒ(x). Its slope is ∆y> ∆x, the points P(x1, ƒ(x1)) and Q(x2, ƒ(x2)) (Figure 2.1). In geometry, a line joining two points of a
average rate of change of ƒ over the curve is a secant to the curve. Thus, the average rate of change of ƒ from x1 to x2 is identi-
interval [x1, x2]. cal with the slope of secant PQ. Let’s consider what happens as the point Q approaches the
point P along the curve, so the length h of the interval over which the change occurs
approaches zero. We will see that this procedure leads to defining the slope of a curve at a
point.
Defining the Slope of a Curve

P We know what is meant by the slope of a straight line, which tells us the rate at which it
L rises or falls—its rate of change as a linear function. But what is meant by the slope of a
curve at a point P on the curve? If there is a tangent line to the curve at P—a line that just
O touches the curve like the tangent to a circle—it would be reasonable to identify the slope
of the tangent as the slope of the curve at P. So we need a precise meaning for the tangent
at a point on a curve.
FIGURE 2.2 L is tangent to the For circles, tangency is straightforward. A line L is tangent to a circle at a point P if L
circle at P if it passes through P passes through P perpendicular to the radius at P (Figure 2.2). Such a line just touches the
perpendicular to radius OP. circle. But what does it mean to say that a line L is tangent to some other curve C at a point P?
To define tangency for general curves, we need an approach that takes into account
the behavior of the secants through P and nearby points Q as Q moves toward P along the
curve (Figure 2.3). Here is the idea:
1. Start with what we can calculate, namely the slope of the secant PQ.
2. Investigate the limiting value of the secant slope as Q approaches P along the curve.
(We clarify the limit idea in the next section.)
3. If the limit exists, take it to be the slope of the curve at P and define the tangent to the
curve at P to be the line through P with this slope.
HISTORICAL BIOGRAPHY This procedure is what we were doing in the falling-rock problem discussed in Example 2.
Pierre de Fermat The next example illustrates the geometric idea for the tangent to a curve.
(1601–1665)
Tangent Secants
P P
Q Tangent
Secants
Q
FIGURE 2.3 The tangent to the curve at P is the line through P whose slope is the limit of
the secant slopes as Q S P from either side.
EXAMPLE 3 Find the slope of the parabola y = x2 at the point P(2, 4). Write an
equation for the tangent to the parabola at this point.
Solution We begin with a secant line through P(2, 4) and Q(2 + h, (2 + h)2) nearby.
We then write an expression for the slope of the secant PQ and investigate what happens to
the slope as Q approaches P along the curve:
∆y (2 + h)2 - 22 h2 + 4h + 4 - 4
Secant slope = = =
∆x h h
h2 + 4h
= = h + 4.
h
If h 7 0, then Q lies above and to the right of P, as in Figure 2.4. If h 6 0, then Q lies to the
left of P (not shown). In either case, as Q approaches P along the curve, h approaches zero
and the secant slope h + 4 approaches 4. We take 4 to be the parabola’s slope at P.
y
y = x2 (2 + h) 2 − 4
Secant slope is = h + 4.
h
Q(2 + h, (2 + h) 2)
Tangent slope = 4
Δy = (2 + h)2 − 4
P(2, 4)
Δx = h
x
0 2 2+h
NOT TO SCALE
FIGURE 2.4 Finding the slope of the parabola y = x2 at the point P(2, 4) as
the limit of secant slopes (Example 3).
The tangent to the parabola at P is the line through P with slope 4:

y = 4 + 4(x - 2) Point-slope equation
y = 4x - 4.
Instantaneous Rates of Change and Tangent Lines

The rates at which the rock in Example 2 was falling at the instants t = 1 and t = 2 are
called instantaneous rates of change. Instantaneous rates and slopes of tangent lines are
closely connected, as we see in the following examples.
EXAMPLE 4 Figure 2.5 shows how a population p of fruit flies (Drosophila) grew
in a 50-day experiment. The number of flies was counted at regular intervals, the counted
values plotted with respect to time t, and the points joined by a smooth curve (colored blue
in Figure 2.5). Find the average growth rate from day 23 to day 45.
Solution There were 150 flies on day 23 and 340 flies on day 45. Thus the number of
flies increased by 340 - 150 = 190 in 45 - 23 = 22 days. The average rate of change
of the population from day 23 to day 45 was
∆p 340 - 150 190

Average rate of change: = = ≈ 8.6 flies>day.
∆t 45 - 23 22
350
Q(45, 340)
300
Number of flies
250 Δp = 190
200 Δp
P(23, 150) ≈ 8.6 flies!day
Δt
150
Δt = 22
100
50
t
0 10 20 30 40 50
Time (days)
FIGURE 2.5 Growth of a fruit fly population in a controlled

experiment. The average rate of change over 22 days is the slope
∆p> ∆t of the secant line (Example 4).
This average is the slope of the secant through the points P and Q on the graph in
Figure 2.5.
The average rate of change from day 23 to day 45 calculated in Example 4 does not
tell us how fast the population was changing on day 23 itself. For that we need to examine
time intervals closer to the day in question.
EXAMPLE 5 How fast was the number of flies in the population of Example 4 grow-
ing on day 23?
Solution To answer this question, we examine the average rates of change over increas-
ingly short time intervals starting at day 23. In geometric terms, we find these rates by
calculating the slopes of secants from P to Q, for a sequence of points Q approaching P
along the curve (Figure 2.6).
p
Slope of PQ = 𝚫p , 𝚫t
B(35, 350)
Q (flies , day) 350
Q(45, 340)
300
340 - 150
Number of flies
(45, 340) ≈ 8.6 250
45 - 23
330 - 150 200
(40, 330) ≈ 10.6 P(23, 150)
40 - 23 150
310 - 150 100

(35, 310) ≈ 13.3
35 - 23 50
265 - 150 t
(30, 265) ≈ 16.4 0 10 20 30 40 50
30 - 23 A(14, 0) Time (days)
FIGURE 2.6 The positions and slopes of four secants through the point P on the fruit fly graph (Example 5).
The values in the table show that the secant slopes rise from 8.6 to 16.4 as the t-coordinate
of Q decreases from 45 to 30, and we would expect the slopes to rise slightly higher as t
continued on toward 23. Geometrically, the secants rotate counterclockwise about P and
seem to approach the red tangent line in the figure. Since the line appears to pass through
the points (14, 0) and (35, 350), it has slope
350 - 0
= 16.7 flies>day (approximately).
35 - 14
On day 23 the population was increasing at a rate of about 16.7 flies > day.
The instantaneous rates in Example 2 were found to be the values of the average
speeds, or average rates of change, as the time interval of length h approached 0. That is,
the instantaneous rate is the value the average rate approaches as the length h of the inter-
val over which the change occurs approaches zero. The average rate of change corre-
sponds to the slope of a secant line; the instantaneous rate corresponds to the slope of the
tangent line as the independent variable approaches a fixed value. In Example 2, the inde-
pendent variable t approached the values t = 1 and t = 2. In Example 3, the independent
variable x approached the value x = 2. So we see that instantaneous rates and slopes of
tangent lines are closely connected. We investigate this connection thoroughly in the next
chapter, but to do so we need the concept of a limit.
Exercises 2.1
Average Rates of Change 5. R(u) = 24u + 1; 30, 24
31, 24
In Exercises 1–6, find the average rate of change of the function over
6. P(u) = u 3 - 4u 2 + 5u;
the given interval or intervals.
1. ƒ(x) = x3 + 1 Slope of a Curve at a Point
a. 32, 34 b. 3- 1, 14 In Exercises 7–14, use the method in Example 3 to find (a) the slope
of the curve at the given point P, and (b) an equation of the tangent
2. g(x) = x2 - 2x line at P.
a. 31, 34 b. 3- 2, 44 7. y = x2 - 5, P(2, - 1)
3. h(t) = cot t 8. y = 7 - x2, P(2, 3)
a. 3p>4, 3p>44 b. 3p>6, p>24 9. y = x2 - 2x - 3, P(2, - 3)
4. g(t) = 2 + cos t 10. y = x2 - 4x, P(1, - 3)
a. 30, p4 b. 3- p, p4 11. y = x3, P(2, 8)
12. y = 2 - x3, P(1, 1) b. What is the average rate of increase of the profits between
2012 and 2014?
13. y = x3 - 12x, P(1, - 11)
c. Use your graph to estimate the rate at which the profits were
14. y = x3 - 3x2 + 4, P(2, 0)
changing in 2012.
Instantaneous Rates of Change T 18. Make a table of values for the function F(x) = (x + 2)>(x - 2)
15. Speed of a car The accompanying figure shows the time-to- at the points x = 1.2, x = 11>10, x = 101>100, x = 1001>1000,
distance graph for a sports car accelerating from a standstill. x = 10001>10000, and x = 1.
s a. Find the average rate of change of F(x) over the intervals
650
P 31, x4 for each x ≠ 1 in your table.
600 b. Extending the table if necessary, try to determine the rate of
Q4
change of F(x) at x = 1.
T 19. Let g(x) = 2x for x Ú 0.
500 Q3
Distance (m)
400 Q2
a. Find the average rate of change of g(x) with respect to x over
300 the intervals 31, 24, 31, 1.54 and 31, 1 + h4.
Q1 b. Make a table of values of the average rate of change of g with
respect to x over the interval 31, 1 + h4 for some values of h
200
100 approaching zero, say h = 0.1, 0.01, 0.001, 0.0001, 0.00001,

and 0.000001.
t
0 5 10 15 20 c. What does your table indicate is the rate of change of g(x)
Elapsed time (sec) with respect to x at x = 1?
a. Estimate the slopes of secants PQ1, PQ2, PQ3, and PQ4, d. Calculate the limit as h approaches zero of the average rate of
arranging them in order in a table like the one in Figure 2.6. change of g(x) with respect to x over the interval 31, 1 + h4.
What are the appropriate units for these slopes? T 20. Let ƒ(t) = 1>t for t ≠ 0.
b. Then estimate the car’s speed at time t = 20 sec. a. Find the average rate of change of ƒ with respect to t over the
16. The accompanying figure shows the plot of distance fallen versus intervals (i) from t = 2 to t = 3, and (ii) from t = 2 to t = T.
time for an object that fell from the lunar landing module a dis- b. Make a table of values of the average rate of change of ƒ with
tance 80 m to the surface of the moon. respect to t over the interval 32, T4 , for some values of T
a. Estimate the slopes of the secants PQ1, PQ2, PQ3, and PQ4, approaching 2, say T = 2.1, 2.01, 2.001, 2.0001, 2.00001,
arranging them in a table like the one in Figure 2.6. and 2.000001.
b. About how fast was the object going when it hit the surface? c. What does your table indicate is the rate of change of ƒ with
y respect to t at t = 2?
80
d. Calculate the limit as T approaches 2 of the average rate of
P
Q4 change of ƒ with respect to t over the interval from 2 to T. You
Distance fallen (m)
60 Q3 will have to do some algebra before you can substitute T = 2.
Q2
21. The accompanying graph shows the total distance s traveled by a
40 bicyclist after t hours.
Q1
20 s
t 40
0 5 10
Distance traveled (mi)
Elapsed time (sec) 30

T 17. The profits of a small company for each of the first five years of
20
its operation are given in the following table:
10
Year Profit in $1000s
t
0 1 2 3 4
2010 6
Elapsed time (hr)
2011 27
2012 62
2013 111 a. Estimate the bicyclist’s average speed over the time intervals
2014 174 30, 14, 31, 2.54 , and 32.5, 3.54 .
b. Estimate the bicyclist’s instantaneous speed at the times t = 12,
t = 2, and t = 3.
a. Plot points representing the profit as a function of year, and c. Estimate the bicyclist’s maximum speed and the specific time
join them by as smooth a curve as you can. at which it occurs.
22. The accompanying graph shows the total amount of gasoline A in a. Estimate the average rate of gasoline consumption over the
the gas tank of an automobile after being driven for t days. time intervals 30, 34, 30, 54, and 37, 104 .
A b. Estimate the instantaneous rate of gasoline consumption at
the times t = 1, t = 4, and t = 8.
16
Remaining amount (gal)
c. Estimate the maximum rate of gasoline consumption and the

12 specific time at which it occurs.
t
0 1 2 3 4 5 6 7 8 9 10
Elapsed time (days)
2.2 Limit of a Function and Limit Laws

In Section 2.1 we saw that limits arise when finding the instantaneous rate of change of a
function or the tangent to a curve. Here we begin with an informal definition of limit and
show how we can calculate the values of limits. A precise definition is presented in the
next section.
HISTORICAL ESSAY
Limits Limits of Function Values
Frequently when studying a function y = ƒ(x), we find ourselves interested in the func-
if c is an irrational number, like p or 22, whose values can only be approximated by

tion’s behavior near a particular point c, but not at c. This might be the case, for instance,
y
“close” rational numbers at which we actually evaluate the function instead. Another situ-
ation occurs when trying to evaluate a function at c leads to division by zero, which is
2 undefined. We encountered this last circumstance when seeking the instantaneous rate of
change in y by considering the quotient function ∆y>h for h closer and closer to zero.
2
y = f(x) = x − 1 Here’s a specific example in which we explore numerically how a function behaves near a
1 x− 1
particular point at which we cannot directly evaluate the function.
x
−1 0 1 EXAMPLE 1 How does the function
y x2 - 1
ƒ(x) =
x - 1
behave near x = 1?
2
y=x+1 Solution The given formula defines ƒ for all real numbers x except x = 1 (we cannot
1 divide by zero). For any x ≠ 1, we can simplify the formula by factoring the numerator
and canceling common factors:
x
−1 0 1 (x - 1)(x + 1)
ƒ(x) = = x + 1 for x ≠ 1.
x - 1
FIGURE 2.7 The graph of ƒ is
identical with the line y = x + 1 The graph of ƒ is the line y = x + 1 with the point (1, 2) removed. This removed point is
except at x = 1, where ƒ is not shown as a “hole” in Figure 2.7. Even though ƒ(1) is not defined, it is clear that we can make
defined (Example 1). the value of ƒ(x) as close as we want to 2 by choosing x close enough to 1 (Table 2.2).
2.2 Limit of a Function and Limit Laws 49
Generalizing the idea illustrated in Example 1, suppose ƒ(x) is defined on an open

TABLE 2.2 As x gets closer to interval about c, except possibly at c itself. If ƒ(x) is arbitrarily close to the number L (as
1, ƒ(x ) gets closer to 2. close to L as we like) for all x sufficiently close to c, we say that ƒ approaches the limit L
as x approaches c, and write
x2 −1
x ƒ(x) =
x−1 lim ƒ(x) = L,
xSc
0.9 1.9
which is read “the limit of ƒ(x) as x approaches c is L.” For instance, in Example 1 we
1.1 2.1
would say that ƒ(x) approaches the limit 2 as x approaches 1, and write
0.99 1.99
1.01 2.01 x2 - 1
lim ƒ(x) = 2, or lim = 2.
xS1 xS1 x - 1
0.999 1.999
1.001 2.001 Essentially, the definition says that the values of ƒ(x) are close to the number L whenever x
0.999999 1.999999 is close to c (on either side of c).
Our definition here is “informal” because phrases like arbitrarily close and sufficiently close
1.000001 2.000001
are imprecise; their meaning depends on the context. (To a machinist manufacturing a piston,
close may mean within a few thousandths of an inch. To an astronomer studying distant galaxies,
close may mean within a few thousand light-years.) Nevertheless, the definition is clear enough to
enable us to recognize and evaluate limits of many specific functions. We will need the precise
definition given in Section 2.3, however, when we set out to prove theorems about limits or study
complicated functions. Here are several more examples exploring the idea of limits.
EXAMPLE 2 The limit value of a function does not depend on how the function
is defined at the point being approached. Consider the three functions in Figure 2.8. The
function ƒ has limit 2 as x S 1 even though ƒ is not defined at x = 1. The function g has
limit 2 as x S 1 even though 2 ≠ g(1). The function h is the only one of the three
functions in Figure 2.8 whose limit as x S 1 equals its value at x = 1. For h, we have
limx S 1 h(x) = h(1). This equality of limit and function value is of special importance, and
we return to it in Section 2.5.
y y y
y 2 2 2
y=x
1 1 1
c
x x x
−1 0 1 −1 0 1 −1 0 1
x x2 − 1 , x ≠ 1
c
2
(a) f (x) = x − 1 (b) g(x) = x− 1 (c) h(x) = x + 1
x− 1
1, x=1
(a) Identity function
y FIGURE 2.8 The limits of ƒ(x), g(x), and h(x) all equal 2 as x approaches 1. However, only h(x)
has the same function value as its limit at x = 1 (Example 2).
k y=k
EXAMPLE 3
(a) If ƒ is the identity function ƒ(x) = x, then for any value of c (Figure 2.9a),
x lim ƒ(x) = lim x = c.

0 c xSc xSc
(b) If ƒ is the constant function ƒ(x) = k (function with the constant value k), then for
(b) Constant function
any value of c (Figure 2.9b),
FIGURE 2.9 The functions in
lim ƒ(x) = lim k = k.
Example 3 have limits at all points c. xSc xSc
For instances of each of these rules we have

lim x = 3 and lim (4) = lim (4) = 4.
xS3 x S -7 xS2
We prove these rules in Example 3 in Section 2.3.
A function may not have a limit at a particular point. Some ways that limits can fail to
exist are illustrated in Figure 2.10 and described in the next example.
y y y
0, x < 0 1, x≠0
y= y= x 1
1, x ≥ 0
0, x = 0
1
x x x
0 0 0
0, x≤0
y=
sin 1x , x > 0
–1
(a) Unit step function U(x) (b) g(x) (c) f(x)
FIGURE 2.10 None of these functions has a limit as x approaches 0 (Example 4).
EXAMPLE 4 Discuss the behavior of the following functions, explaining why they
have no limit as x S 0.
(a) U(x) = e
0, x 6 0
1, x Ú 0
1
x, x≠0
(b) g(x) = W
0, x = 0
0, x … 0
(c) ƒ(x) = c
1
sin x , x 7 0
Solution
(a) It jumps: The unit step function U(x) has no limit as x S 0 because its values jump
at x = 0. For negative values of x arbitrarily close to zero, U(x) = 0. For positive
values of x arbitrarily close to zero, U(x) = 1. There is no single value L approached
by U(x) as x S 0 (Figure 2.10a).
(b) It grows too “large” to have a limit: g(x) has no limit as x S 0 because the values of
g grow arbitrarily large in absolute value as x S 0 and do not stay close to any fixed
real number (Figure 2.10b). We say the function is not bounded.
(c) It oscillates too much to have a limit: ƒ(x) has no limit as x S 0 because the func-
tion’s values oscillate between +1 and -1 in every open interval containing 0. The
values do not stay close to any one number as x S 0 (Figure 2.10c).
The Limit Laws

To calculate limits of functions that are arithmetic combinations of functions having
known limits, we can use several fundamental rules.
THEOREM 1—Limit Laws If L, M, c, and k are real numbers and

lim ƒ(x) = L and lim g(x) = M, then
xSc xSc
1. Sum Rule: lim (ƒ(x) + g(x)) = L + M

xSc
2. Difference Rule: lim (ƒ(x) - g(x)) = L - M

xSc
3. Constant Multiple Rule: lim (k # ƒ(x)) = k # L

xSc
4. Product Rule: lim (ƒ(x) # g(x)) = L # M

xSc
ƒ(x) L
5. Quotient Rule: lim = , M ≠ 0
x S c g(x) M
6. Power Rule: lim 3 ƒ(x) 4 n = L n, n a positive integer
lim 2ƒ(x) = 2L = L 1>n, n a positive integer

xSc
n n
7. Root Rule:
xSc
(If n is even, we assume that lim ƒ(x) = L 7 0.)

xSc
In words, the Sum Rule says that the limit of a sum is the sum of the limits. Similarly, the
next rules say that the limit of a difference is the difference of the limits; the limit of a con-
stant times a function is the constant times the limit of the function; the limit of a product
is the product of the limits; the limit of a quotient is the quotient of the limits (provided
that the limit of the denominator is not 0); the limit of a positive integer power (or root) of
a function is the integer power (or root) of the limit (provided that the root of the limit is a
real number).
It is reasonable that the properties in Theorem 1 are true (although these intuitive
arguments do not constitute proofs). If x is sufficiently close to c, then ƒ(x) is close to L
and g(x) is close to M, from our informal definition of a limit. It is then reasonable that
ƒ(x) + g(x) is close to L + M; ƒ(x) - g(x) is close to L - M; kƒ(x) is close to kL;
ƒ(x)g(x) is close to LM; and ƒ(x)>g(x) is close to L>M if M is not zero. We prove the Sum
Rule in Section 2.3, based on a precise definition of limit. Rules 2–5 are proved in Appen-
dix 4. Rule 6 is obtained by applying Rule 4 repeatedly. Rule 7 is proved in more advanced
texts. The Sum, Difference, and Product Rules can be extended to any number of func-
tions, not just two.
EXAMPLE 5 Use the observations limx S c k = k and limx S c x = c (Example 3) and

the fundamental rules of limits to find the following limits.
(a) lim ( x3 + 4x2 - 3 )
xSc
x4 + x2 - 1
(b) lim
xSc x2 + 5
(c) lim 24x2 - 3

x S -2
Solution
(a) lim ( x3 + 4x2 - 3 ) = lim x3 + lim 4x2 - lim 3 Sum and Difference Rules
xSc xSc xSc xSc
= c3 + 4c2 - 3 Power and Multiple Rules
lim ( x4 + x2 - 1 )
x4 + x2 - 1 xSc
(b) lim = Quotient Rule
xSc x2 + 5 lim ( x2 + 5 )
xSc
lim x4 + lim x2 - lim 1

xSc xSc xSc
= Sum and Difference Rules
lim x2 + lim 5
xSc xSc
c4 + c2 - 1
= Power or Product Rule
c2 + 5
(c) lim 24x2 - 3 = 2 lim ( 4x2 - 3 ) Root Rule with n = 2
x S -2 x S -2
= 2 lim 4x2 - lim 3 Difference Rule

x S -2 x S -2
= 24(-2)2 - 3 Product and Multiple Rules
= 216 - 3
= 213
Theorem 1 simplifies the task of calculating limits of polynomials and rational functions.
To evaluate the limit of a polynomial function as x approaches c, merely substitute c for x
in the formula for the function. To evaluate the limit of a rational function as x approaches
a point c at which the denominator is not zero, substitute c for x in the formula for the
function. (See Examples 5a and 5b.) We state these results formally as theorems.
THEOREM 2—Limits of Polynomials

If P(x) = an xn + an - 1 xn - 1 + g + a0, then
lim P(x) = P(c) = an cn + an - 1 cn - 1 + g + a0.
xSc
THEOREM 3—Limits of Rational Functions

If P(x) and Q(x) are polynomials and Q(c) ≠ 0, then
P(x) P(c)
lim = .
xSc Q(x) Q(c)
EXAMPLE 6 The following calculation illustrates Theorems 2 and 3:

3 2
x3 + 4x2 - 3 (-1) + 4(-1) - 3 0
lim = = = 0
x S -1 x2 + 5 (-1)2 + 5 6
Identifying Common Factors

It can be shown that if Q(x) is a poly- Eliminating Common Factors from Zero Denominators
nomial and Q(c) = 0, then (x - c) is
a factor of Q(x). Thus, if the numerator Theorem 3 applies only if the denominator of the rational function is not zero at the limit
and denominator of a rational function point c. If the denominator is zero, canceling common factors in the numerator and
of x are both zero at x = c, they have denominator may reduce the fraction to one whose denominator is no longer zero at c. If
(x - c) as a common factor. this happens, we can find the limit by substitution in the simplified fraction.
y EXAMPLE 7 Evaluate
2
y= x + x−2
x2 − x x2 + x - 2
(1, 3) lim .
3 xS1 x2 - x
Solution We cannot substitute x = 1 because it makes the denominator zero. We test

the numerator to see if it, too, is zero at x = 1. It is, so it has a factor of (x - 1) in com-
x mon with the denominator. Canceling this common factor gives a simpler fraction with the
−2 0 1
same values as the original for x ≠ 1:
(a)
x2 + x - 2 (x - 1)(x + 2) x + 2
= = x , if x ≠ 1.
y x2 - x x(x - 1)
y=x+2 Using the simpler fraction, we find the limit of these values as x S 1 by Theorem 3:
x
3 (1, 3)
x2 + x - 2 x + 2 1 + 2
lim = lim x = = 3.
xS1 x2 - x xS1 1
See Figure 2.11.
x
−2 0 1
Using Calculators and Computers to Estimate Limits

(b)
When we cannot use the Quotient Rule in Theorem 1 because the limit of the denominator
FIGURE 2.11 The graph of is zero, we can try using a calculator or computer to guess the limit numerically as x gets
ƒ(x) = ( x2 + x - 2 ) > ( x2 - x ) in closer and closer to c. We used this approach in Example 1, but calculators and computers
part (a) is the same as the graph of can sometimes give false values and misleading impressions for functions that are unde-
g(x) = (x + 2)>x in part (b) except fined at a point or fail to have a limit there. Usually the problem is associated with round-
at x = 1, where ƒ is undefined. The ing errors, as we now illustrate.
2x2 + 100 - 10
functions have the same limit as x S 1
(Example 7).
EXAMPLE 8 Estimate the value of lim .
xS0 x2
Solution Table 2.3 lists values of the function obtained on a calculator for several points
approaching x = 0. As x approaches 0 through the points {1, {0.5, {0.10, and {0.01,
the function seems to approach the number 0.05.
As we take even smaller values of x, {0.0005, {0.0001, {0.00001, and {0.000001,
the function appears to approach the number 0.
Is the answer 0.05 or 0, or some other value? We resolve this question in the next
example.
2x 2 + 100 - 10
TABLE 2.3 Computed values of ƒ(x) = near x = 0
x2
x ƒ(x)
{1 0.049876
{0.5 0.049969
t approaches 0.05?
{0.1 0.049999
{0.01 0.050000
{0.0005 0.050000
{0.0001 0.000000
t approaches 0?
{0.00001 0.000000
{0.000001 0.000000
Using a computer or calculator may give ambiguous results, as in the last example.
The calculator could not keep track of enough digits to avoid rounding errors in computing
the values of ƒ(x) when x is very small. We cannot substitute x = 0 in the problem, and the
numerator and denominator have no obvious common factors (as they did in Example 7).
Sometimes, however, we can create a common factor algebraically.
EXAMPLE 9 Evaluate
2x2 + 100 - 10
lim .
xS0 x2
Solution This is the limit we considered in Example 8. We can create a common factor
2x2 + 100 + 10 (obtained by changing the sign after the square root). The preliminary
by multiplying both numerator and denominator by the conjugate radical expression
algebra rationalizes the numerator:
2x2 + 100 - 10 2x2 + 100 - 10 2x2 + 100 + 10

2x2 + 100 + 10
2
= #
x x2
x2 + 100 - 100
x2 ( 2x2 + 100 + 10 )
=
x2
x2( 2x2 + 100 + 10)
= Common factor x2
2x + 100 + 10
1
= 2
. Cancel x2 for x ≠ 0.
Therefore,
2x2 + 100 - 10
2x + 100 + 10
1
lim = lim
xS0 x2 xS0 2
Denominator not 0 at
202 + 100 + 10
1
= x = 0; substitute.
1
= = 0.05.
20
This calculation provides the correct answer, in contrast to the ambiguous computer
results in Example 8.
We cannot always algebraically resolve the problem of finding the limit of a quotient
where the denominator becomes zero. In some cases the limit might then be found with
y
the aid of some geometry applied to the problem (see the proof of Theorem 7 in Section 2.4),
or through methods of calculus (illustrated in Section 7.5). The next theorems give helpful
h
tools by using function comparisons.
f
L
g The Sandwich Theorem

The following theorem enables us to calculate a variety of limits. It is called the Sandwich
x
0 c Theorem because it refers to a function ƒ whose values are sandwiched between the val-
ues of two other functions g and h that have the same limit L at a point c. Being trapped
FIGURE 2.12 The graph of ƒ is sand- between the values of two functions that approach L, the values of ƒ must also approach L
wiched between the graphs of g and h. (Figure 2.12). You will find a proof in Appendix 4.
THEOREM 4—The Sandwich Theorem Suppose that g(x) … ƒ(x) … h(x) for
all x in some open interval containing c, except possibly at x = c itself. Suppose
also that
lim g(x) = lim h(x) = L.
xSc xSc
Then limx S c ƒ(x) = L.
y 2 The Sandwich Theorem is also called the Squeeze Theorem or the Pinching Theorem.
y=1+ x
2
2
EXAMPLE 10 Given that
y = u(x)
x2 x2
1 - … u(x) … 1 + for all x ≠ 0,
1
4 2
2
y=1− x
4 find limx S 0 u(x), no matter how complicated u is.
x
−1 0 1 Solution Since
FIGURE 2.13 Any function u(x) lim ( 1 - ( x2 >4 ) ) = 1 and lim ( 1 + ( x2 >2 ) ) = 1,
xS0 xS0
whose graph lies in the region between
y = 1 + (x2 >2) and y = 1 - (x2 >4) the Sandwich Theorem implies that limx S 0 u(x) = 1 (Figure 2.13).
has limit 1 as x S 0 (Example 10).
EXAMPLE 11 The Sandwich Theorem helps us establish several important limit rules:
(a) lim sin u = 0 (b) lim cos u = 1
uS0 uS0
(c) For any function ƒ, lim 0 ƒ(x) 0 = 0 implies lim ƒ(x) = 0.
y
y = 0u0
xSc xSc
y = sin u
1 Solution
(a) In Section 1.3 we established that - 0 u 0 … sin u … 0 u 0 for all u (see Figure 2.14a).
Since limu S 0 ( - 0 u 0 ) = limu S 0 0 u 0 = 0, we have
u
−p p
−1 y = −0 u 0
lim sin u = 0.
uS0
(a) (b) From Section 1.3, 0 … 1 - cos u … 0 u 0 for all u (see Figure 2.14b), and we have
limu S 0 (1 - cos u) = 0 or
y
lim cos u = 1.
y = 0u0 uS0
(c) Since - 0 ƒ(x) 0 … ƒ(x) … 0 ƒ(x) 0 and - 0 ƒ(x) 0 and 0 ƒ(x) 0 have limit 0 as x S c, it
2
1 follows that limx S c ƒ(x) = 0.

y = 1 − cos u
u Another important property of limits is given by the next theorem. A proof is given in
−2 −1 0 1 2
the next section.
(b)
FIGURE 2.14 The Sandwich Theorem

confirms the limits in Example 11. THEOREM 5 If ƒ(x) … g(x) for all x in some open interval containing c, except
possibly at x = c itself, and the limits of ƒ and g both exist as x approaches c,
then
lim ƒ(x) … lim g(x).
xSc xSc
Caution The assertion resulting from replacing the less than or equal to ( … ) inequality by
the strict less than ( 6 ) inequality in Theorem 5 is false. Figure 2.14a shows that for u ≠ 0,
- 0 u 0 6 sin u 6 0 u 0 . So limu S 0 sin u = 0 = limu S 0 0 u 0 , not limu S 0 sin u 6 limu S 0 0 u 0 .
Exercises 2.2
Limits from Graphs d. lim ƒ(x) exists at every point c in (-1, 1).
xSc
1. For the function g(x) graphed here, find the following limits or e. lim ƒ(x) exists at every point c in (1, 3).
explain why they do not exist. xSc
a. lim g(x) b. lim g(x) c. lim g(x) d. lim g(x) y

xS1 xS2 xS3 x S 2.5
y = f (x)
y 1
x
y = g(x) −1 1 2 3
1 −1
x −2
1 2 3
2. For the function ƒ(t) graphed here, find the following limits or Existence of Limits
explain why they do not exist. In Exercises 5 and 6, explain why the limits do not exist.
x 1
xS0 0 x 0
a. lim ƒ(t) b. lim ƒ(t) c. lim ƒ(t) d. lim ƒ(t) 5. lim 6. lim
t S -2 t S -1 tS0 t S -0.5 xS1 x - 1
s 7. Suppose that a function ƒ(x) is defined for all real values of x

except x = c. Can anything be said about the existence of
limx S c ƒ(x)? Give reasons for your answer.
8. Suppose that a function ƒ(x) is defined for all x in 3-1, 1]. Can
1
s = f (t)
t anything be said about the existence of limx S 0 ƒ(x)? Give reasons

−2 −1 0 1 for your answer.
−1 9. If limx S 1 ƒ(x) = 5, must ƒ be defined at x = 1? If it is, must
ƒ(1) = 5? Can we conclude anything about the values of ƒ at
x = 1? Explain.
3. Which of the following statements about the function y = ƒ(x) 10. If ƒ(1) = 5, must limx S 1 ƒ(x) exist? If it does, then must
graphed here are true, and which are false? limx S 1 ƒ(x) = 5? Can we conclude anything about limx S 1 ƒ(x)?
Explain.
a. lim ƒ(x) exists.
xS0
b. lim ƒ(x) = 0 Calculating Limits
xS0
Find the limits in Exercises 11–22.
c. lim ƒ(x) = 1
xS0
11. lim ( x2 - 13 ) 12. lim ( - x2 + 5x - 2 )
d. lim ƒ(x) = 1 x S -3 xS2
xS1
e. lim ƒ(x) = 0 13. lim 8(t - 5)(t - 7) 14. lim ( x3 - 2x2 + 4x + 8 )
tS6 x S -2
xS1
f. lim ƒ(x) exists at every point c in (-1, 1). 2x + 5
xSc 15. lim 16. lim (8 - 3s)(2s - 1)
g. lim ƒ(x) does not exist. x S 2 11 - x3 s S 2>3
xS1 y + 2
17. lim 4x(3x + 4)2 18. lim 2
x S -1>2 yS2 y + 5y + 6
20. lim 2z2 - 10
y
y = f (x)
19. lim (5 - y)4>3
1 y S -3 zS4
25h + 4 - 2
23h + 1 + 1
3
21. lim 22. lim
x hS0 hS0 h
−1 1 2
Limits of quotients Find the limits in Exercises 23–42.
−1
x - 5 x + 3
23. lim 2 24. lim 2
x S 5 x - 25 x S -3 x + 4x + 3
x2 + 3x - 10 x2 - 7x + 10
4. Which of the following statements about the function y = ƒ(x) 25. lim 26. lim
x S -5 x + 5 xS2 x - 2
graphed here are true, and which are false?
t2 + t - 2 t 2 + 3t + 2
a. lim ƒ(x) does not exist. 27. lim 2 28. lim 2
xS2 tS1 t - 1 t S -1 t - t - 2
b. lim ƒ(x) = 2 -2x - 4 5y3 + 8y2
xS2 29. lim 3 30. lim
c. lim ƒ(x) does not exist. x S -2 x + 2x 2 y S 0 3y4 - 16y2
xS1
1
+ 1 53. Suppose limx S c ƒ(x) = 5 and limx S c g(x) = -2. Find
x-1 - 1 x - 1 x + 1
31. lim 32. lim a. lim ƒ(x)g(x) b. lim 2ƒ(x)g(x)
xS1 x - 1 xS0 x xSc xSc
u4 - 1 y3 - 8 ƒ(x)
33. lim 3 34. lim 4 c. lim (ƒ(x) + 3g(x)) d. lim
uS1 u - 1 ƒ(x) - g(x)
2x - 3
y S 2 y - 16 xSc xSc
4x - x2 54. Suppose limx S 4 ƒ(x) = 0 and limx S 4 g(x) = -3. Find

x S 4 2 - 2x
35. lim 36. lim
xS9 x - 9 a. lim (g(x) + 3) b. lim xƒ(x)
2x2 + 8 - 3
xS4 xS4
x - 1
x S 1 2x + 3 - 2
37. lim 38. lim g(x)
x S -1 x + 1 c. lim (g(x))2 d. lim
2x2 + 12 - 4
xS4 xS4 ƒ(x) - 1
x + 2
2x2 + 5 - 3
39. lim 40. lim 55. Suppose limx S b ƒ(x) = 7 and limx S b g(x) = -3. Find
x - 2
b. lim ƒ(x) # g(x)
2 - 2x2 - 5
xS2 x S -2
a. lim (ƒ(x) + g(x))
xSb xSb
4 - x
x S 4 5 - 2x 2 + 9
41. lim 42. lim c. lim 4g(x) d. lim ƒ(x)>g(x)
x S -3 x + 3 xSb xSb
56. Suppose that limx S -2 p(x) = 4, limx S -2 r(x) = 0, and
Limits with trigonometric functions Find the limits in Exercises
limx S -2 s(x) = - 3. Find
43–50.
a. lim (p(x) + r(x) + s(x))
43. lim (2 sin x - 1) 44. lim sin2 x x S -2
xS0 x S p>4
b. lim p(x) # r(x) # s(x)
45. lim sec x 46. lim tan x x S -2
xS0 x S p>3
c. lim (- 4p(x) + 5r(x))>s(x)
1 + x + sin x x S -2
47. lim 48. lim (x2 - 1)(2 - cos x)
xS0 3 cos x xS0
49. lim 2x + 4 cos (x + p) 50. lim 27 + sec2 x Limits of Average Rates of Change
x S -p xS0 Because of their connection with secant lines, tangents, and instanta-
neous rates, limits of the form
Using Limit Rules
ƒ(x + h) - ƒ(x)
51. Suppose limx S 0 ƒ(x) = 1 and limx S 0 g(x) = -5. Name the lim
hS0 h
rules in Theorem 1 that are used to accomplish steps (a), (b), and
(c) of the following calculation. occur frequently in calculus. In Exercises 57–62, evaluate this limit
lim (2ƒ(x) - g(x)) for the given value of x and function ƒ.
2ƒ(x) - g(x) xS0
lim = (a) 57. ƒ(x) = x2, x = 1
xS0 (ƒ(x) + 7)2>3 lim (ƒ(x) + 7)2>3
xS0 58. ƒ(x) = x2, x = - 2
lim 2ƒ(x) - lim g(x) 59. ƒ(x) = 3x - 4, x = 2
xS0 xS0
= (b)
a lim ( ƒ(x) + 7 ) b
2>3 60. ƒ(x) = 1>x, x = - 2
xS0 61. ƒ(x) = 2x, x = 7
2 lim ƒ(x) - lim g(x)
xS0 xS0
62. ƒ(x) = 23x + 1, x = 0
= (c)
a lim ƒ(x) + lim 7b
2>3
Using the Sandwich Theorem

63. If 25 - 2x2 … ƒ(x) … 25 - x2 for - 1 … x … 1, find
xS0 xS0
(2)(1) - (-5) 7
= = limx S 0 ƒ(x).
(1 + 7) 2>3 4
64. If 2 - x2 … g(x) … 2 cos x for all x, find limx S 0 g(x).
52. Let limx S 1 h(x) = 5, limx S 1 p(x) = 1, and limx S 1 r(x) = 2.
65. a. It can be shown that the inequalities
Name the rules in Theorem 1 that are used to accomplish steps
(a), (b), and (c) of the following calculation. x2 x sin x
lim 25h(x)
1 - 6 6 1
25h(x)
6 2 - 2 cos x
xS1
lim = (a) hold for all values of x close to zero. What, if anything, does
xS1 p(x)(4 - r(x)) lim (p(x)(4 - r(x)))
xS1
4xlim
this tell you about
5h(x)
=
S1
(b) x sin x
lim ?
a lim p(x)b a lim ( 4 - r(x) ) b
xS0 2 - 2 cos x
xS1 xS1
4 xS1
Give reasons for your answer.
T b. Graph y = 1 - ( x2 >6 ) , y = (x sin x)>(2 - 2 cos x), and
5lim h(x)
= (c) y = 1 together for - 2 … x … 2. Comment on the behavior
a lim p(x)b a lim 4 - lim r(x)b of the graphs as x S 0.
xS1 xS1 xS1
2(5)(5) 5
= =
(1)(4 - 2) 2
66. a. Suppose that the inequalities 71. Let ƒ(x) = ( x2 - 1 ) > ( 0 x 0 - 1 ) .

a. Make tables of the values of ƒ at values of x that approach
1 x2 1 - cos x 1 c = -1 from above and below. Then estimate limx S -1 ƒ(x).
- 6 6
2 24 x2 2
b. Support your conclusion in part (a) by graphing ƒ near
hold for values of x close to zero. (They do, as you will see in c = -1 and using Zoom and Trace to estimate y-values on
Section 9.9.) What, if anything, does this tell you about the graph as x S -1.
1 - cos x c. Find limx S -1 ƒ(x) algebraically.
72. Let F(x) = ( x2 + 3x + 2 ) > ( 2 - 0 x 0 ) .
lim ?
xS0 x2
Give reasons for your answer.
a. Make tables of values of F at values of x that approach
T b. Graph the equations y = (1>2) - ( x2 >24 ) , c = -2 from above and below. Then estimate limx S -2 F(x).
y = (1 - cos x)>x2, and y = 1>2 together for -2 … x … 2.
b. Support your conclusion in part (a) by graphing F near
Comment on the behavior of the graphs as x S 0.
c = -2 and using Zoom and Trace to estimate y-values on
the graph as x S -2.
Estimating Limits c. Find limx S -2 F(x) algebraically.
T You will find a graphing calculator useful for Exercises 67–74. 73. Let g(u) = (sin u)>u.
67. Let ƒ(x) = (x2 - 9)>(x + 3). a. Make a table of the values of g at values of u that approach
a. Make a table of the values of ƒ at the points x = -3.1, u0 = 0 from above and below. Then estimate limu S 0 g(u).
- 3.01, -3.001, and so on as far as your calculator can go. b. Support your conclusion in part (a) by graphing g near
Then estimate limx S -3 ƒ(x). What estimate do you arrive at u0 = 0.
if you evaluate ƒ at x = -2.9, -2.99, -2.999, c instead?
74. Let G(t) = (1 - cos t)>t 2.
b. Support your conclusions in part (a) by graphing ƒ near
a. Make tables of values of G at values of t that approach t0 = 0
c = -3 and using Zoom and Trace to estimate y-values on
from above and below. Then estimate limt S 0 G(t).
the graph as x S - 3.
b. Support your conclusion in part (a) by graphing G near
c. Find limx S -3 ƒ(x) algebraically, as in Example 7.
68. Let g(x) = ( x2 - 2 ) >(x - 22).
t0 = 0.
a. Make a table of the values of g at the points x = 1.4, 1.41, Theory and Examples
75. If x4 … ƒ(x) … x2 for x in 3- 1, 14 and x2 … ƒ(x) … x4 for
of 22. Estimate limx S 22 g(x).
1.414, and so on through successive decimal approximations
x 6 -1 and x 7 1, at what points c do you automatically know
c = 22 and using Zoom and Trace to estimate y-values on

b. Support your conclusion in part (a) by graphing g near limx S c ƒ(x)? What can you say about the value of the limit at
the graph as x S 22.

these points?
76. Suppose that g(x) … ƒ(x) … h(x) for all x ≠ 2 and suppose that
c. Find limx S 22 g(x) algebraically.
lim g(x) = lim h(x) = -5.
69. Let G(x) = (x + 6)> ( x2 + 4x - 12 ) . xS2 xS2
a. Make a table of the values of G at x = -5.9, - 5.99, -5.999, Can we conclude anything about the values of ƒ, g, and h at
and so on. Then estimate limx S -6 G(x). What estimate do x = 2? Could ƒ(2) = 0? Could limx S 2 ƒ(x) = 0? Give reasons
you arrive at if you evaluate G at x = - 6.1, -6.01, for your answers.
- 6.001, cinstead? ƒ(x) - 5
77. If lim = 1, find lim ƒ(x).
b. Support your conclusions in part (a) by graphing G and using xS4 x - 2 xS4
Zoom and Trace to estimate y-values on the graph as ƒ(x)

78. If lim 2 = 1, find
x S - 6. x S -2 x
c. Find limx S -6 G(x) algebraically.

a. lim ƒ(x)
70. Let h(x) = ( x2 - 2x - 3 ) > ( x2 - 4x + 3 ) . x S -2
a. Make a table of the values of h at x = 2.9, 2.99, 2.999, and ƒ(x)

b. lim x
so on. Then estimate limx S 3 h(x). What estimate do you x S -2
arrive at if you evaluate h at x = 3.1, 3.01, 3.001, c ƒ(x) - 5

79. a. If lim = 3, find lim ƒ(x).
instead? xS2 x - 2 xS2
b. Support your conclusions in part (a) by graphing h near ƒ(x) - 5

b. If lim = 4, find lim ƒ(x).
c = 3 and using Zoom and Trace to estimate y-values on the xS2 x - 2 xS2
graph as x S 3.
c. Find limx S 3 h(x) algebraically.
2.3 The Precise Definition of a Limit 59
ƒ(x) x4 - 16
80. If lim = 1, find 83. lim
x S 0 x2 xS2 x - 2
a. lim ƒ(x) x3 - x2 - 5x - 3
xS0 84. lim
x S -1 (x + 1)2
21 + x - 1
ƒ(x)
b. lim x 3
xS0 85. lim
xS0 x
T 81. a. Graph g(x) = x sin (1>x) to estimate limx S 0 g(x), zooming in
2
x - 9
2x2 + 7 - 4
on the origin as necessary.
86. lim
xS3
b. Confirm your estimate in part (a) with a proof.
T 82. a. Graph h(x) = x2 cos (1>x3) to estimate limx S 0 h(x), zooming 1 - cos x
87. lim
in on the origin as necessary. xS0 x sin x
b. Confirm your estimate in part (a) with a proof. 2x2
88. lim
xS0 3 - 3 cos x
COMPUTER EXPLORATIONS
Graphical Estimates of Limits
In Exercises 83–88, use a CAS to perform the following steps:
a. Plot the function near the point c being approached.
b. From your plot guess the value of the limit.
2.3 The Precise Definition of a Limit

We now turn our attention to the precise definition of a limit. We replace vague phrases
like “gets arbitrarily close to” in the informal definition with specific conditions that can
be applied to any particular example. With a precise definition, we can avoid misunder-
standings, prove the limit properties given in the preceding section, and establish many
important limits.
To show that the limit of ƒ(x) as x S c equals the number L, we need to show that the
gap between ƒ(x) and L can be made “as small as we choose” if x is kept “close enough”
to c. Let us see what this would require if we specified the size of the gap between ƒ(x)
and L.
EXAMPLE 1 Consider the function y = 2x - 1 near x = 4. Intuitively it appears

y that y is close to 7 when x is close to 4, so limx S 4 (2x - 1) = 7. However, how close to
y = 2x − 1 x = 4 does x have to be so that y = 2x - 1 differs from 7 by, say, less than 2 units?
Upper bound: Solution We are asked: For what values of x is 0 y - 7 0 6 2? To find the answer we
first express 0 y - 7 0 in terms of x:
y=9
9
To satisfy
this 7 0 y - 7 0 = 0 (2x - 1) - 7 0 = 0 2x - 8 0 .
5
Lower bound:
The question then becomes: what values of x satisfy the inequality 0 2x - 8 0 6 2? To find
y=5 out, we solve the inequality:
x
0 2x - 8 0 6 2
0 3 4 5 -2 6 2x - 8 6 2
Restrict 6 6 2x 6 10
to this
3 6 x 6 5 Solve for x.
FIGURE 2.15 Keeping x within 1 unit
-1 6 x - 4 6 1. Solve for x - 4.
of x = 4 will keep y within 2 units of
y = 7 (Example 1). Keeping x within 1 unit of x = 4 will keep y within 2 units of y = 7 (Figure 2.15).
y In the previous example we determined how close x must be to a particular value c to

ensure that the outputs ƒ(x) of some function lie within a prescribed interval about a limit
1
value L. To show that the limit of ƒ(x) as x S c actually equals L, we must be able to show
L+ that the gap between ƒ(x) and L can be made less than any prescribed error, no matter how
10
f (x)
f (x) lies small, by holding x close enough to c.
L in here
1 Definition of Limit
L−
10
for all x ≠ c
Suppose we are watching the values of a function ƒ(x) as x approaches c (without taking
in here
on the value of c itself). Certainly we want to be able to say that ƒ(x) stays within one-
d d
x tenth of a unit from L as soon as x stays within some distance d of c (Figure 2.16). But that
x
0 c−d c c+d in itself is not enough, because as x continues on its course toward c, what is to prevent
ƒ(x) from jittering about within the interval from L - (1>10) to L + (1>10) without
FIGURE 2.16 How should we define tending toward L?
d 7 0 so that keeping x within the interval We can be told that the error can be no more than 1>100 or 1>1000 or 1>100,000.
(c - d, c + d) will keep ƒ(x) within the
Each time, we find a new d@interval about c so that keeping x within that interval satisfies
interval aL - b?
1 1 the new error tolerance. And each time the possibility exists that ƒ(x) jitters away from L
,L +
10 10 at some stage.
The figures on the next page illustrate the problem. You can think of this as a quarrel
between a skeptic and a scholar. The skeptic presents P@challenges to prove that the limit
does not exist or, more precisely, that there is room for doubt. The scholar answers every
challenge with a d@interval around c that keeps the function values within P of L.
How do we stop this seemingly endless series of challenges and responses? We can
do so by proving that for every error tolerance P that the challenger can produce, we can
present a matching distance d that keeps x “close enough” to c to keep ƒ(x) within that
P@tolerance of L (Figure 2.17). This leads us to the precise definition of a limit.
DEFINITION Let ƒ(x) be defined on an open interval about c, except possibly at c

itself. We say that the limit of ƒ(x) as x approaches c is the number L, and write
L+P
lim ƒ(x) = L,
f (x) lies xSc
L in here
f (x) if, for every number P 7 0, there exists a corresponding number d 7 0 such
that for all x,
0 6 0x - c0 6 d 0 ƒ(x) - L 0 6 P.
L−P
1
for all x ≠ c
in here
d d
One way to think about the definition is to suppose we are machining a generator
x shaft to a close tolerance. We may try for diameter L, but since nothing is perfect, we must
x be satisfied with a diameter ƒ(x) somewhere between L - P and L + P. The d is the mea-
0
c−d c c+d
sure of how accurate our control setting for x must be to guarantee this degree of accuracy
FIGURE 2.17 The relation of d and P in the diameter of the shaft. Notice that as the tolerance for error becomes stricter, we may
in the definition of limit. have to adjust d. That is, the value of d, how tight our control setting must be, depends on
the value of P, the error tolerance.
Examples: Testing the Definition

The formal definition of limit does not tell how to find the limit of a function, but it
enables us to verify that a conjectured limit value is correct. The following examples show
how the definition can be used to verify limit statements for specific functions. However,
the real purpose of the definition is not to do calculations like this, but rather to prove gen-
eral theorems so that the calculation of specific limits can be simplified, such as the theo-
rems stated in the previous section.
y y y y
y = f(x) y = f(x) y = f(x) y = f(x)

1 1
L+ L+
10 10 1 1
L+ L+
100 100
L L L L
1 1
1 1 L− L−
L− 100 100
10 L−
10
x x x x
0 c 0 c 0 c 0 c
c − d1/10 c + d1/10 c − d1/100 c + d1/100
The challenge: Response: New challenge: Response:
Make 0 f(x) − L 0 < P = 1 0 x − c 0 < d1/10 (a number) Make 0 f(x) − L 0 < P = 1 0 x − c 0 < d1/100
10 100
y y
y = f(x) y = f (x)
1 1
L+ L+
1000 1000
L L
1 1
L− L−
1000 1000
x x
0 c 0 c
New challenge: Response:

P = 1
1000 0 x − c 0 < d1/1000
y y y
y = f(x) y = f (x) y = f(x)
1 1
L+ L+
100,000 100,000
L+P
L L L
1 1 L−P
L− L−
100,000 100,000
x x x
0 c 0 c c
0
New challenge: Response: New challenge:
0 x − c 0 < d1/100,000
1
P =
100,000 P = ...
EXAMPLE 2 Show that

lim (5x - 3) = 2.
xS1
Solution Set c = 1, ƒ(x) = 5x - 3, and L = 2 in the definition of limit. For any given
P 7 0, we have to find a suitable d 7 0 so that if x ≠ 1 and x is within distance d of
c = 1, that is, whenever
0 6 0 x - 1 0 6 d,
it is true that ƒ(x) is within distance P of L = 2, so
0 ƒ(x) - 2 0 6 P.
y We find d by working backward from the P@inequality:

0 (5x - 3) - 2 0 = 0 5x - 5 0 6 P
y = 5x − 3
50x - 10 6 P
2+P
2 0 x - 1 0 6 P>5.
Thus, we can take d = P>5 (Figure 2.18). If 0 6 0 x - 1 0 6 d = P>5, then
0 (5x - 3) - 2 0 = 0 5x - 5 0 = 5 0 x - 1 0 6 5(P>5) = P,
2−P
x
which proves that limx S 1(5x - 3) = 2.
The value of d = P>5 is not the only value that will make 0 6 0 x - 1 0 6 d imply
0 1−P 1 1+P
5 5
0 5x - 5 0 6 P. Any smaller positive d will do as well. The definition does not ask for a
“best” positive d, just one that will work.
−3 EXAMPLE 3 Prove the following results presented graphically in Section 2.2.

NOT TO SCALE (a) lim x = c
xSc
FIGURE 2.18 If ƒ(x) = 5x - 3, then (b) lim k = k
0 6 0 x - 1 0 6 P>5 guarantees that
(k constant)
xSc
0 ƒ(x) - 2 0 6 P (Example 2).
Solution
(a) Let P 7 0 be given. We must find d 7 0 such that for all x
0 6 0x - c0 6 d implies 0 x - c 0 6 P.
y
The implication will hold if d equals P or any smaller positive number (Figure 2.19).
y=x
This proves that limx S c x = c.
c+P
(b) Let P 7 0 be given. We must find d 7 0 such that for all x
0 6 0x - c0 6 d 0 k - k 0 6 P.
c+d
c implies
c−d
Since k - k = 0, we can use any positive number for d and the implication will hold
c−P (Figure 2.20). This proves that limx S c k = k.
x
0 c−d c c+d
Finding Deltas Algebraically for Given Epsilons
FIGURE 2.19 For the function In Examples 2 and 3, the interval of values about c for which 0 ƒ(x) - L 0 was less than P
ƒ(x) = x, we find that 0 6 0 x - c 0 6 d was symmetric about c and we could take d to be half the length of that interval. When
will guarantee 0 ƒ(x) - c 0 6 P whenever such symmetry is absent, as it usually is, we can take d to be the distance from c to the
d … P (Example 3a). interval’s nearer endpoint.
EXAMPLE 4 For the limit limx S 5 2x - 1 = 2, find a d 7 0 that works for P = 1.

That is, find a d 7 0 such that for all x
0 6 0x - 50 6 d 0 2x - 1 - 2 0 6 1.
y
1
y=k
k+P
Solution We organize the search into two steps.
1. Solve the inequality 0 2x - 1 - 2 0 6 1 to find an interval containing x = 5 on
k
k−P
which the inequality holds for all x ≠ 5.

0 2x - 1 - 2 0 6 1
-1 6 2x - 1 - 2 6 1
x
1 6 2x - 1 6 3
0 c−d c c+d
FIGURE 2.20 For the function

ƒ(x) = k, we find that 0 ƒ(x) - k 0 6 P 1 6 x - 1 6 9
for any positive d (Example 3b). 2 6 x 6 10
3 3 The inequality holds for all x in the open interval (2, 10), so it holds for all x ≠ 5 in
( ) x
this interval as well.
2 5 8 10
2. Find a value of d 7 0 to place the centered interval 5 - d 6 x 6 5 + d (centered
FIGURE 2.21 An open interval of ra- at x = 5) inside the interval (2, 10). The distance from 5 to the nearer endpoint of
dius 3 about x = 5 will lie inside the open (2, 10) is 3 (Figure 2.21). If we take d = 3 or any smaller positive number, then the
interval (2, 10). inequality 0 6 0 x - 5 0 6 d will automatically place x between 2 and 10 to make
0 2x - 1 - 2 0 6 1 (Figure 2.22):
y 0 6 0x - 50 6 3 1 0 2x - 1 - 2 0 6 1.
y = "x − 1
3
How to Find Algebraically a D for a Given ƒ, L, c, and E + 0
2 The process of finding a d 7 0 such that for all x
0 6 0x - c0 6 d 1 0 ƒ(x) - L 0 6 P
1 can be accomplished in two steps.
1. Solve the inequality 0 ƒ(x) - L 0 6 P to find an open interval (a, b) contain-
3 3
x ing c on which the inequality holds for all x ≠ c.

0 1 2 5 8 10
2. Find a value of d 7 0 that places the open interval (c - d, c + d) centered
at c inside the interval (a, b). The inequality 0 ƒ(x) - L 0 6 P will hold for all
NOT TO SCALE
FIGURE 2.22 The function and inter-

x ≠ c in this d@interval.
vals in Example 4.
EXAMPLE 5 Prove that limx S 2 ƒ(x) = 4 if
ƒ(x) = e
x2, x≠2
1, x = 2.
Solution Our task is to show that given P 7 0 there exists a d 7 0 such that for all x
0 6 0x - 20 6 d 0 ƒ(x) - 4 0 6 P.
y
1
y = x2
1. Solve the inequality 0 ƒ(x) - 4 0 6 P to find an open interval containing x = 2 on
4+P
which the inequality holds for all x ≠ 2.
For x ≠ c = 2, we have ƒ(x) = x2, and the inequality to solve is 0 x2 - 4 0 6 P:
(2, 4)
0 x2 - 4 0 6 P
4
4−P -P 6 x2 - 4 6 P
4 - P 6 x2 6 4 + P
(2, 1) 24 - P 6 0 x 0 6 24 + P Assumes P 6 4; see below.
24 - P 6 x 6 24 + P.
x An open interval about x = 2
0 2
"4 − P "4 + P
that solves the inequality
The inequality 0 ƒ(x) - 4 0 6 P holds for all x ≠ 2 in the open interval ( 24 - P,

24 + P ) (Figure 2.23).
FIGURE 2.23 An interval containing
x = 2 so that the function in Example 5
satisfies 0 ƒ(x) - 4 0 6 P.
interval ( 24 - P, 24 + P ) .
2. Find a value of d 7 0 that places the centered interval (2 - d, 2 + d) inside the
Take d to be the distance from x = 2 to the nearer endpoint of ( 24 - P, 24 + P ) .

In other words, take d = min 5 2 - 24 - P, 24 + P - 2 6 , the minimum (the
smaller) of the two numbers 2 - 24 - P and 24 + P - 2. If d has this or any

smaller positive value, the inequality 0 6 0 x - 2 0 6 d will automatically place x
between 24 - P and 24 + P to make 0 ƒ(x) - 4 0 6 P. For all x,
0 6 0x - 20 6 d 1 0 ƒ(x) - 4 0 6 P.
This completes the proof for P 6 4.
the interval ( 0, 24 + P ) . In other words, take d = min 5 2, 24 + P - 2 6 . (See

If P Ú 4, then we take d to be the distance from x = 2 to the nearer endpoint of
Figure 2.23.)
Using the Definition to Prove Theorems

We do not usually rely on the formal definition of limit to verify specific limits such as
those in the preceding examples. Rather, we appeal to general theorems about limits, in
particular the theorems of Section 2.2. The definition is used to prove these theorems
(Appendix 5). As an example, we prove part 1 of Theorem 1, the Sum Rule.
EXAMPLE 6 Given that limx S c ƒ(x) = L and limx S c g(x) = M, prove that
lim (ƒ(x) + g(x)) = L + M.
xSc
Solution Let P 7 0 be given. We want to find a positive number d such that for all x
0 6 0x - c0 6 d 1 0 ƒ(x) + g(x) - (L + M) 0 6 P.
Regrouping terms, we get
0 ƒ(x) + g(x) - (L + M) 0 = 0 (ƒ(x) - L) + (g(x) - M) 0 Triangle Inequality:
… 0 ƒ(x) - L 0 + 0 g(x) - M 0 . 0a + b0 … 0a0 + 0b0
Since limx S c ƒ(x) = L, there exists a number d1 7 0 such that for all x
0 6 0 x - c 0 6 d1 1 0 ƒ(x) - L 0 6 P>2.
Similarly, since limx S c g(x) = M, there exists a number d2 7 0 such that for all x
0 6 0 x - c 0 6 d2 1 0 g(x) - M 0 6 P>2.
Let d = min 5d1, d2 6 , the smaller of d1 and d2. If 0 6 0 x - c 0 6 d then 0 x - c 0 6 d1,
so 0 ƒ(x) - L 0 6 P>2, and 0 x - c 0 6 d2, so 0 g(x) - M 0 6 P>2. Therefore
0 ƒ(x) + g(x) - (L + M) 0 6 P + P = P.
2 2
This shows that limx S c (ƒ(x) + g(x)) = L + M.
Next we prove Theorem 5 of Section 2.2.
EXAMPLE 7 Given that limx S c ƒ(x) = L and limx S c g(x) = M, and that ƒ(x) … g(x)
for all x in an open interval containing c (except possibly c itself), prove that L … M.
Solution We use the method of proof by contradiction. Suppose, on the contrary, that
L 7 M. Then by the limit of a difference property in Theorem 1,
lim (g(x) - ƒ(x)) = M - L.
xSc
Therefore, for any P 7 0, there exists d 7 0 such that

0 (g(x) - ƒ(x)) - (M - L) 0 6 P whenever 0 6 0 x - c 0 6 d.
Since L - M 7 0 by hypothesis, we take P = L - M in particular and we have a num-
ber d 7 0 such that
0 (g(x) - ƒ(x)) - (M - L) 0 6 L - M whenever 0 6 0 x - c 0 6 d.
Since a … 0 a 0 for any number a, we have
(g(x) - ƒ(x)) - (M - L) 6 L - M whenever 0 6 0 x - c 0 6 d
which simplifies to
g(x) 6 ƒ(x) whenever 0 6 0 x - c 0 6 d.
But this contradicts ƒ(x) … g(x). Thus the inequality L 7 M must be false. Therefore
L … M.
Exercises 2.3
Centering Intervals About a Point 9. 10.
In Exercises 1–6, sketch the interval (a, b) on the x-axis with the
f(x) = " x y
f (x) = 2" x + 1
point c inside. Then find a value of d 7 0 such that for all
x, 0 6 0 x - c 0 6 d 1 a 6 x 6 b.
y c=1
L=1 c=3
1. a = 1, b = 7, c = 5 P = 1 y = "x L=4
5 4 P = 0.2
2. a = 1, b = 7, c = 2 4
1 y = 2" x + 1
3. a = - 7>2, b = -1>2, c = -3 3 4.2
4. a = - 7>2, b = -1>2, c = -3>2 4 4
3.8
5. a = 4>9, b = 4>7, c = 1>2
6. a = 2.7591, b = 3.2391, c = 3 x 2
0 9 1 25
16 16
x
−1 0 2.61 3 3.41
Finding Deltas Graphically
In Exercises 7–14, use the graphs to find a d 7 0 such that for all x NOT TO SCALE
0 6 0 x - c 0 6 d 1 0 ƒ(x) - L 0 6 P. 11. 12.

7. 8.
y y
y
y f (x) = − 3 x + 3 x2
f (x) = f(x) = 4 − x 2
2 c = −1
y = 2x − 4 c = −3 c=2
L=4 L=3
L = 7.5 P = 0.25 3.25
6.2 f (x) = 2x − 4 P = 0.15 P=1
6 c=5
5.8 L=6 y = −3 x + 3 y = 4 − x2 3
2 y = x2 5
P = 0.2 7.65
7.5 4 2.75
x 7.35 3
0 5
4.9 5.1
x
0 2
"3 "5
NOT TO SCALE
x NOT TO SCALE
−3 0
−3.1 −2.9
x
NOT TO SCALE " 5 −1 " 3 0
− −
2 2
NOT TO SCALE
13. 14. Using the Formal Definition

y
Each of Exercises 31–36 gives a function ƒ(x), a point c, and a posi-
y
tive number P. Find L = lim ƒ(x). Then find a number d 7 0 such
xSc
f(x) = 2 that for all x
"−x f (x) = 1x
c = −1 0 6 0x - c0 6 d 1 0 ƒ(x) - L 0 6 P.
L=2 c=1
2.01 2
P = 0.5 L=2 31. ƒ(x) = 3 - 2x, c = 3, P = 0.02
P = 0.01
y= 2 2 32. ƒ(x) = - 3x - 2, c = -1, P = 0.03
"−x
2.5 2
x - 4
1.99 33. ƒ(x) = , c = 2, P = 0.05
x - 2
2
y = 1x x2 + 6x + 5
1.5 34. ƒ(x) = , c = -5, P = 0.05
x + 5
35. ƒ(x) = 21 - 5x, c = - 3, P = 0.5
36. ƒ(x) = 4>x, c = 2, P = 0.4

x
x 0 1 Prove the limit statements in Exercises 37–50.
16 −1 16 0 1 1
− − 2 37. lim (9 - x) = 5 38. lim (3x - 7) = 2
9 25 2.01 1.99
39. lim 2x - 5 = 2 40. lim 24 - x = 2

xS4 xS3
NOT TO SCALE
xS9 xS0
41. lim ƒ(x) = 1 if ƒ(x) = e

2
x, x ≠ 1
xS1 2, x = 1
Finding Deltas Algebraically
42. lim ƒ(x) = 4 if ƒ(x) = e
Each of Exercises 15–30 gives a function ƒ(x) and numbers L, c, and x 2, x ≠ -2
P 7 0. In each case, find an open interval about c on which the inequal- x S -2 1, x = -2
ity 0 ƒ(x) - L 0 6 P holds. Then give a value for d 7 0 such that for
all x satisfying 0 6 0 x - c 0 6 d the inequality 0 ƒ(x) - L 0 6 P
1 1 1
x S 23
43. lim x = 1 44. lim =
xS1 x2 3
holds.
15. ƒ(x) = x + 1, L = 5, c = 4, P = 0.01 x2 - 9 x2 - 1
45. lim = -6 46. lim = 2
x S -3 x + 3 xS1 x - 1
16. ƒ(x) = 2x - 2, L = -6, c = -2, P = 0.02
17. ƒ(x) = 2x + 1, 47. lim ƒ(x) = 2 if ƒ(x) = e
4 - 2x, x 6 1
L = 1, c = 0, P = 0.1
18. ƒ(x) = 2x,
xS1 6x - 4, x Ú 1
L = 1>2, c = 1>4, P = 0.1
19. ƒ(x) = 219 - x, 48. lim ƒ(x) = 0 if ƒ(x) = e
2x, x 6 0
L = 3, c = 10, P = 1
20. ƒ(x) = 2x - 7,
xS0 x>2, x Ú 0
L = 4, c = 23, P = 1 1
49. lim x sin x = 0
21. ƒ(x) = 1>x, L = 1>4, c = 4, P = 0.05 xS0
22. ƒ(x) = x ,2
L = 3, c = 23, P = 0.1 y
23. ƒ(x) = x2, L = 4, c = - 2, P = 0.5
24. ƒ(x) = 1>x, L = -1, c = -1, P = 0.1
2
25. ƒ(x) = x - 5, L = 11, c = 4, P = 1
26. ƒ(x) = 120>x, L = 5, c = 24, P = 1 1 1 y = x sin 1x
−
27. ƒ(x) = mx, m 7 0, L = 2m, c = 2, P = 0.03 2p 2p
x
1 1
28. ƒ(x) = mx, m 7 0, L = 3m, c = 3, P = c 7 0 −p p
29. ƒ(x) = mx + b, m 7 0, L = (m>2) + b,
c = 1>2, P = c 7 0
30. ƒ(x) = mx + b, m 7 0, L = m + b, c = 1,
P = 0.05
1 When Is a Number L Not the Limit of ƒ(x) as x u c?

50. lim x2 sin x = 0
xS0 Showing L is not a limit We can prove that limx S c ƒ(x) ≠ L by
providing an P 7 0 such that no possible d 7 0 satisfies the condition
y
for all x, 0 6 0 x - c 0 6 d 1 0 ƒ(x) - L 0 6 P.
1 y = x2
We accomplish this for our candidate P by showing that for each
d 7 0 there exists a value of x such that
0 6 0x - c0 6 d and 0 ƒ(x) - L 0 Ú P.
y= x 2 sin 1
x
x
−1 2 0 2 1 y
−p p
y = f(x)
L+P
L
−1 y = −x 2
L−P
f (x)
Theory and Examples

x
51. Define what it means to say that lim g(x) = k. 0 c− d c c+ d
xS0
52. Prove that lim ƒ(x) = L if and only if lim ƒ(h + c) = L. a value of x for which
0 < 0 x − c 0 < d and 0 f(x) − L 0 ≥ P
xSc hS0
53. A wrong statement about limits Show by example that the

following statement is wrong.
The number L is the limit of ƒ(x) as x approaches c
57. Let ƒ(x) = e
x, x 6 1
if ƒ(x) gets closer to L as x approaches c.
x + 1, x 7 1.
Explain why the function in your example does not have the
given value of L as a limit as x S c.
54. Another wrong statement about limits Show by example that y
the following statement is wrong. y=x+1
The number L is the limit of ƒ(x) as x approaches c if, given any
P 7 0, there exists a value of x for which 0 ƒ(x) - L 0 6 P.
2
Explain why the function in your example does not have the y = f (x)
given value of L as a limit as x S c. 1
T 55. Grinding engine cylinders Before contracting to grind engine
cylinders to a cross-sectional area of 9 in2, you need to know how x
1
much deviation from the ideal cylinder diameter of c = 3.385 in.
you can allow and still have the area come within 0.01 in2 of the
required 9 in2. To find out, you let A = p(x>2)2 and look for the y=x
interval in which you must hold x to make 0 A - 9 0 … 0.01.
What interval do you find?
56. Manufacturing electrical resistors Ohm’s law for electrical a. Let P = 1>2. Show that no possible d 7 0 satisfies the fol-
circuits like the one shown in the accompanying figure states that lowing condition:
V = RI. In this equation, V is a constant voltage, I is the current For all x, 0 6 0 x - 1 0 6 d 1 0 ƒ(x) - 2 0 6 1>2.
in amperes, and R is the resistance in ohms. Your firm has been
asked to supply the resistors for a circuit in which V will be 120 That is, for each d 7 0 show that there is a value of x such
volts and I is to be 5 { 0.1 amp. In what interval does R have to that
0 6 0x - 10 6 d 0 ƒ(x) - 2 0 Ú 1>2.
lie for I to be within 0.1 amp of the value I0 = 5?
and
− This will show that limx S 1 ƒ(x) ≠ 2.

V I R
+ b. Show that limx S 1 ƒ(x) ≠ 1.
c. Show that limx S 1 ƒ(x) ≠ 1.5.
x 2, x 6 2 60. a. For the function graphed here, show that limx S -1 g(x) ≠ 2.
58. Let h(x) = c 3, x = 2 b. Does limx S -1 g(x) appear to exist? If so, what is the value of
2, x 7 2. the limit? If not, why not?
y
y
2
y = h(x)
4
y = g(x)
3 1
y=2
2
1 y = x2 x
−1 0
x
0 2
COMPUTER EXPLORATIONS
Show that In Exercises 61–66, you will further explore finding deltas graphi-
cally. Use a CAS to perform the following steps:
a. lim h(x) ≠ 4
xS2 a. Plot the function y = ƒ(x) near the point c being approached.
b. lim h(x) ≠ 3
xS2 b. Guess the value of the limit L and then evaluate the limit sym-
c. lim h(x) ≠ 2 bolically to see if you guessed correctly.
xS2
c. Using the value P = 0.2, graph the banding lines y1 = L - P
59. For the function graphed here, explain why
and y2 = L + P together with the function ƒ near c.
a. lim ƒ(x) ≠ 4
xS3 d. From your graph in part (c), estimate a d 7 0 such that for all x
b. lim ƒ(x) ≠ 4.8
xS3 0 6 0x - c0 6 d 1 0 ƒ(x) - L 0 6 P.
c. lim ƒ(x) ≠ 3 Test your estimate by plotting ƒ, y1, and y2 over the interval
0 6 0 x - c 0 6 d. For your viewing window use c - 2d …
xS3
x … c + 2d and L - 2P … y … L + 2P. If any function val-

y ues lie outside the interval 3L - P, L + P], your choice of d
was too large. Try again with a smaller estimate.
e. Repeat parts (c) and (d) successively for P = 0.1, 0.05, and 0.001.
4.8
x4 - 81 5x3 + 9x2
4 61. ƒ(x) = , c = 3 62. ƒ(x) = , c = 0
y = f (x) x - 3 2x5 + 3x2
3 sin 2x x(1 - cos x)
63. ƒ(x) = , c = 0 64. ƒ(x) = , c = 0
3x x - sin x
2x - 1
3
65. ƒ(x) = , c = 1
x - 1
x
3x2 - (7x + 1) 2x + 5
0 3 66. ƒ(x) = , c = 1
x - 1
2.4 One-Sided Limits

In this section we extend the limit concept to one-sided limits, which are limits as x
approaches the number c from the left-hand side (where x 6 c) or the right-hand side
(x 7 c) only.
Approaching a Limit from One Side

To have a limit L as x approaches c, a function ƒ must be defined on both sides of c and its
values ƒ(x) must approach L as x approaches c from either side. That is, ƒ must be defined
in some open interval about c, but not necessarily at c. Because of this, ordinary limits are
called two-sided.
2.4 One-Sided Limits 69
y If ƒ fails to have a two-sided limit at c, it may still have a one-sided limit, that is, a
limit if the approach is only from one side. If the approach is from the right, the limit is a
y= x right-hand limit. From the left, it is a left-hand limit.
0x0
1 The function ƒ(x) = x> 0 x 0 (Figure 2.24) has limit 1 as x approaches 0 from the right,
and limit -1 as x approaches 0 from the left. Since these one-sided limit values are not the
same, there is no single number that ƒ(x) approaches as x approaches 0. So ƒ(x) does not
x have a (two-sided) limit at 0.
0
Intuitively, if ƒ(x) is defined on an interval (c, b), where c 6 b, and approaches arbi-
trarily close to L as x approaches c from within that interval, then ƒ has right-hand limit
−1 L at c. We write
lim ƒ(x) = L.
x S c+
The symbol “x S c+ ” means that we consider only values of x greater than c.
FIGURE 2.24 Different right-hand and Similarly, if ƒ(x) is defined on an interval (a, c), where a 6 c and approaches arbi-
left-hand limits at the origin. trarily close to M as x approaches c from within that interval, then ƒ has left-hand limit M
at c. We write
lim ƒ(x) = M.
x S c-
The symbol “x S c- ” means that we consider only x-values less than c.
These informal definitions of one-sided limits are illustrated in Figure 2.25. For the
function ƒ(x) = x> 0 x 0 in Figure 2.24 we have
lim ƒ(x) = 1 and lim ƒ(x) = -1.
x S 0+ x S 0-
y y
f (x) M
L f(x)
x x
0 c x 0 x c
(a) lim+ f (x) = L (b) lim _ f(x) = M
x: c x: c
FIGURE 2.25 (a) Right-hand limit as x approaches c. (b) Left-hand limit as x

approaches c.
The domain of ƒ(x) = 24 - x2 is 3 -2, 24 ; its graph is the semicircle

y
EXAMPLE 1
in Figure 2.26. We have
y = "4 − x 2
lim 24 - x2 = 0 and lim 24 - x2 = 0.
x S -2 + x S 2-
The function does not have a left-hand limit at x = -2 or a right-hand limit at x = 2. It
does not have a two-sided limit at either -2 or 2 because each point does not belong to an
x open interval over which ƒ is defined.
−2 0 2
One-sided limits have all the properties listed in Theorem 1 in Section 2.2. The right-hand
ƒ(x) = 24 - x2 has right-hand limit 0
FIGURE 2.26 The function
limit of the sum of two functions is the sum of their right-hand limits, and so on. The theorems
for limits of polynomials and rational functions hold with one-sided limits, as do the Sandwich
at x = - 2 and left-hand limit 0 at x = 2
Theorem and Theorem 5. One-sided limits are related to limits in the following way.
(Example 1).
THEOREM 6 A function ƒ(x) has a limit as x approaches c if and only if it has

left-hand and right-hand limits there and these one-sided limits are equal:
lim ƒ(x) = L 3 lim ƒ(x) = L and lim ƒ(x) = L.
xSc x S c- x S c+
y EXAMPLE 2 For the function graphed in Figure 2.27,

y = f(x)
2 At x = 0: limx S 0+ ƒ(x) = 1,
limx S 0- ƒ(x) and limx S 0 ƒ(x) do not exist. The function is not de-
1 fined to the left of x = 0.
At x = 1: limx S 1- ƒ(x) = 0 even though ƒ(1) = 1,
x
0 1 2 3 4 limx S 1+ ƒ(x) = 1,
limx S 1 ƒ(x) does not exist. The right- and left-hand limits are not
FIGURE 2.27 Graph of the function
in Example 2. equal.
At x = 2: limx S 2- ƒ(x) = 1,
limx S 2+ ƒ(x) = 1,
limx S 2 ƒ(x) = 1 even though ƒ(2) = 2.
y At x = 3: limx S 3- ƒ(x) = limx S 3+ ƒ(x) = limx S 3 ƒ(x) = ƒ(3) = 2.
At x = 4: limx S 4- ƒ(x) = 1 even though ƒ(4) ≠ 1,
limx S 4+ ƒ(x) and limx S 4 ƒ(x) do not exist. The function is not
defined to the right of x = 4.
L+P At every other point c in 3 0, 44, ƒ(x) has limit ƒ(c).

f (x)
f (x) lies
L in here Precise Definitions of One-Sided Limits
L−P The formal definition of the limit in Section 2.3 is readily modified for one-sided
limits.
for all x ≠ c
in here
d DEFINITIONS We say that ƒ(x) has right-hand limit L at c, and write
x
x
0 c c+d lim ƒ(x) = L (see Figure 2.28)
x S c+
FIGURE 2.28 Intervals associated with if for every number P 7 0 there exists a corresponding number d 7 0 such that
the definition of right-hand limit. for all x
c 6 x 6 c + d 1 0 ƒ(x) - L 0 6 P.
We say that ƒ has left-hand limit L at c, and write
y lim ƒ(x) = L (see Figure 2.29)
x S c-
if for every number P 7 0 there exists a corresponding number d 7 0 such that
for all x
c - d 6 x 6 c 1 0 ƒ(x) - L 0 6 P.
L+P
f (x)
f (x) lies
L in here EXAMPLE 3 Prove that
L−P lim 2x = 0.
x S 0+
for all x ≠ c
in here Solution Let P 7 0 be given. Here c = 0 and L = 0, so we want to find a d 7 0 such
that for all x
0 2x - 0 0 6 P,
d
x 0 6 x 6 d 1
x
0 c−d c
or
2x 6 P.
FIGURE 2.29 Intervals associated with
the definition of left-hand limit. 0 6 x 6 d 1
y Squaring both sides of this last inequality gives

f (x) = "x x 6 P2 if 0 6 x 6 d.
If we choose d = P2 we have
2x 6 P,
P
f(x) 0 6 x 6 d = P2 1
or
0 2x - 0 0 6 P.
x
L=0 x d = P2
0 6 x 6 P2 1
FIGURE 2.30 lim 1x = 0 in Example 3. According to the definition, this shows that limx S 0+ 2x = 0 (Figure 2.30).
x S 0+
The functions examined so far have had some kind of limit at each point of interest. In
general, that need not be the case.
EXAMPLE 4 Show that y = sin (1>x) has no limit as x approaches zero from either
side (Figure 2.31).
x
0
y = sin 1x
−1
FIGURE 2.31 The function y = sin (1>x) has neither a right-

hand nor a left-hand limit as x approaches zero (Example 4). The
graph here omits values very near the y-axis.
Solution As x approaches zero, its reciprocal, 1>x, grows without bound and the values
of sin (1>x) cycle repeatedly from -1 to 1. There is no single number L that the function’s
values stay increasingly close to as x approaches zero. This is true even if we restrict x to
positive values or to negative values. The function has neither a right-hand limit nor a left-
hand limit at x = 0.
Limits Involving (sin U) , U

A central fact about (sin u)>u is that in radian measure its limit as u S 0 is 1. We can see
this in Figure 2.32 and confirm it algebraically using the Sandwich Theorem. You will see
the importance of this limit in Section 3.5, where instantaneous rates of change of the
trigonometric functions are studied.
1
y = sin u (radians)
u
u
−3p −2p −p p 2p 3p
NOT TO SCALE
FIGURE 2.32 The graph of ƒ(u) = (sin u)>u suggests that the right-
and left-hand limits as u approaches 0 are both 1.
y
THEOREM 7—Limit of the Ratio sin U , U as U u 0
T sin u
lim = 1 (u in radians) (1)
1 uS0 u
tan u
1 Proof The plan is to show that the right-hand and left-hand limits are both 1. Then we
sin u will know that the two-sided limit is 1 as well.
To show that the right-hand limit is 1, we begin with positive values of u less than
u cos u p>2 (Figure 2.33). Notice that
x
O Q A(1, 0)
Area ∆OAP 6 area sector OAP 6 area ∆OAT.
1
We can express these areas in terms of u as follows:
FIGURE 2.33 The figure for the proof of
Theorem 7. By definition, TA>OA = tan u, 1 1 1
Area ∆OAP = base * height = (1)(sin u) = sin u
but OA = 1, so TA = tan u.
2 2 2
1 2 1 u
Area sector OAP = r u = (1)2u = (2)
2 2 2
1 1 1
Area ∆OAT = base * height = (1)(tan u) = tan u.
2 2 2
Thus,
Equation (2) is where radian measure 1 1 1

sin u 6 u 6 tan u.
comes in: The area of sector OAP is u>2 2 2 2
only if u is measured in radians.
This last inequality goes the same way if we divide all three terms by the number
(1>2) sin u, which is positive, since 0 6 u 6 p>2:
u 1
1 6 6 .
sin u cos u
Taking reciprocals reverses the inequalities:

sin u
1 7 7 cos u.
u
Since limu S 0+ cos u = 1 (Example 11b, Section 2.2), the Sandwich Theorem gives
sin u
lim = 1.
u S 0+ u
To consider the left-hand limit, we recall that sin u and u are both odd functions (Sec-
tion 1.1). Therefore, ƒ(u) = (sin u)>u is an even function, with a graph symmetric about
the y-axis (see Figure 2.32). This symmetry implies that the left-hand limit at 0 exists and
has the same value as the right-hand limit:
sin u sin u
lim = 1 = lim+ ,
u S 0- u uS0 u
so limu S 0 (sin u)>u = 1 by Theorem 6.
cos h - 1 sin 2x 2
EXAMPLE 5 Show that (a) lim = 0 and (b) lim = .
hS0 h xS0 5x 5
Solution
(a) Using the half-angle formula cos h = 1 - 2 sin2 (h>2), we calculate
cos h - 1 2 sin2 (h>2)

lim = lim -
hS0 h hS0 h
sin u
= - lim sin u Let u = h>2.
uS0 u
Eq. (1) and Example 11a
= -(1)(0) = 0. in Section 2.2
(b) Equation (1) does not apply to the original fraction. We need a 2x in the denominator,
not a 5x. We produce it by multiplying numerator and denominator by 2>5:
sin 2x (2>5) # sin 2x

lim = lim
x S 0 5x x S 0 (2>5) # 5x
2 sin 2x Now, Eq. (1) applies

= lim with u = 2x.
5 x S 0 2x
2 2
= (1) =
5 5
tan t sec 2t
EXAMPLE 6 Find lim .
tS0 3t
Solution From the definition of tan t and sec 2t, we have

tan t sec 2t 1 1 sin t 1
lim = lim # t # cos t #
tS0 3t tS0 3 cos 2t
1 sin t # 1 # 1
= lim
3 t S 0 t cos t cos 2t
1 1 Eq. (1) and Example 11b
= (1)(1)(1) = . in Section 2.2
3 3
Exercises 2.4
Finding Limits Graphically y
1. Which of the following statements about the function y = ƒ(x) y = f (x)
graphed here are true, and which are false?
y 2
y = f (x)
1 1
x x
−1 0 1 2 −1 0 1 2 3
a. lim ƒ(x) = 1 b. lim ƒ(x) = 0

x S -1+ x S 0- a. lim ƒ(x) = 1 b. lim ƒ(x) does not exist.
x S -1+ xS2
c. lim- ƒ(x) = 1 d. lim- ƒ(x) = lim+ ƒ(x)
xS0 xS0 xS0 c. lim ƒ(x) = 2 d. lim ƒ(x) = 2
xS2 x S 1-
e. lim ƒ(x) exists. f. lim ƒ(x) = 0
xS0 xS0 e. lim ƒ(x) = 1 f. lim ƒ(x) does not exist.
x S 1+ xS1
g. lim ƒ(x) = 1 h. lim ƒ(x) = 1
xS0 xS1 g. lim+ ƒ(x) = lim- ƒ(x)
xS0 xS0
i. lim ƒ(x) = 0 j. lim ƒ(x) = 2
xS1 x S 2- h. lim ƒ(x) exists at every c in the open interval (-1, 1).
xSc
k. lim ƒ(x) does not exist. l. lim ƒ(x) = 0
x S -1- x S 2+ i. lim ƒ(x) exists at every c in the open interval (1, 3).
xSc
2. Which of the following statements about the function y = ƒ(x)
j. lim ƒ(x) = 0 k. lim ƒ(x) does not exist.
graphed here are true, and which are false? x S -1- x S 3+
3 - x, x 6 2 6. Let g(x) = 2x sin(1>x).

3. Let ƒ(x) = c x y
+ 1, x 7 2.
2 1 y = "x
y = "x sin 1x
y=3−x
3 1
y= x+1
2 2p x
0 1 2 1
p p
x
0 2 4
a. Find limx S 2+ ƒ(x) and limx S 2- ƒ(x).

b. Does limx S 2 ƒ(x) exist? If so, what is it? If not, why not?
−1 y = −"x
c. Find limx S 4- ƒ(x) and limx S 4+ ƒ(x).
d. Does limx S 4 ƒ(x) exist? If so, what is it? If not, why not? a. Does limx S 0+ g(x) exist? If so, what is it? If not, why not?
b. Does limx S 0- g(x) exist? If so, what is it? If not, why not?
3 - x, x 6 2
c. Does limx S 0 g(x) exist? If so, what is it? If not, why not?
2, x = 2
4. Let ƒ(x) = d
7. a. Graph ƒ(x) = e
x x 3, x≠1
, x 7 2.
2 0, x = 1.
b. Find limx S 1- ƒ(x) and limx S 1+ ƒ(x).
y
c. Does limx S 1 ƒ(x) exist? If so, what is it? If not, why not?
8. a. Graph ƒ(x) = e
y=3−x 1 - x 2, x≠1
3 2, x = 1.
b. Find limx S 1+ ƒ(x) and limx S 1- ƒ(x).
y= x
2 c. Does limx S 1 ƒ(x) exist? If so, what is it? If not, why not?
x
−2 0 2
Graph the functions in Exercises 9 and 10. Then answer these questions.
a. Find limx S 2+ ƒ(x), limx S 2- ƒ(x), and ƒ(2). a. What are the domain and range of ƒ?
b. Does limx S 2 ƒ(x) exist? If so, what is it? If not, why not? b. At what points c, if any, does limx S c ƒ(x) exist?
c. Find limx S -1- ƒ(x) and limx S -1+ ƒ(x). c. At what points does only the left-hand limit exist?
d. Does limx S -1 ƒ(x) exist? If so, what is it? If not, why not? d. At what points does only the right-hand limit exist?
21 - x2, 0 … x 6 1
0, x … 0
5. Let ƒ(x) = c 9. ƒ(x) = c 1, 1 … x 6 2
1
sin x , x 7 0. 2, x = 2
x, - 1 … x 6 0, or 0 6 x … 1
y
10. ƒ(x) = c 1, x = 0
1 0, x 6 - 1 or x 7 1
Finding One-Sided Limits Algebraically

x Find the limits in Exercises 11–18.
0
x S -0.5- A x Ax + 2
0, x≤0 x + 2 x - 1
y= 1
11. lim 12. lim+
sin x , x > 0 + 1 xS1
lim a ba b
−1 x 2x + 5
13.
x S -2+ x + 1 x2 + x
14. lim- a ba x ba b
1 x + 6 3 - x
a. Does limx S 0+ ƒ(x) exist? If so, what is it? If not, why not? xS1 x + 1 7
b. Does limx S 0- ƒ(x) exist? If so, what is it? If not, why not? 2h2 + 4h + 5 - 25
15. lim+
c. Does limx S 0 ƒ(x) exist? If so, what is it? If not, why not? hS0 h
2.5 Continuity 75
26 - 25h2 + 11h + 6 41. lim

tan u
42. lim
u cot 4u
16. lim- u 2 cot 3u sin2 u cot2 2u
hS0 h uS0 uS0
0x + 20 0x + 20 Theory and Examples

17. a. lim +(x + 3) b. lim -(x + 3)
x S -2 x + 2 x S -2 x + 2
22x (x - 1) 22x (x - 1)
43. Once you know limx S a+ ƒ(x) and limx S a- ƒ(x) at an interior point
of the domain of ƒ, do you then know limx S a ƒ(x)? Give reasons
0x - 10 0x - 10
18. a. lim+ b. lim- for your answer.
xS1 xS1
44. If you know that limx S c ƒ(x) exists, can you find its value by cal-
Use the graph of the greatest integer function y = : x ; , Figure 1.10 in culating limx S c+ ƒ(x)? Give reasons for your answer.
Section 1.1, to help you find the limits in Exercises 19 and 20. 45. Suppose that ƒ is an odd function of x. Does knowing that
:u ; :u ; limx S 0+ ƒ(x) = 3 tell you anything about limx S 0- ƒ(x)? Give rea-
19. a. lim+ b. lim-
uS3 u uS3 u sons for your answer.
20. a. lim+(t - : t ; ) b. lim-(t - : t ; ) 46. Suppose that ƒ is an even function of x. Does knowing that
tS4 tS4
limx S 2- ƒ(x) = 7 tell you anything about either limx S -2- ƒ(x) or
limx S -2+ ƒ(x)? Give reasons for your answer.
sin U
Using lim = 1
U Uu 0
Formal Definitions of One-Sided Limits
sin 22u x lies in I, then 2x - 5 6 P. What limit is being verified and
47. Given P 7 0, find an interval I = (5, 5 + d), d 7 0, such that if
sin kt
22u
21. lim 22. lim t (k constant) what is its value?
uS0 tS0
x lies in I, then 24 - x 6 P. What limit is being verified and

sin 3y h 48. Given P 7 0, find an interval I = (4 - d, 4), d 7 0, such that if
23. lim 24. lim-
y S 0 4y hS0sin 3h
tan 2x 2t what is its value?
25. lim x 26. lim tan t
xS0 tS0 Use the definitions of right-hand and left-hand limits to prove the
x csc 2x limit statements in Exercises 49 and 50.
27. lim 28. lim 6x2(cot x)(csc 2x)
x S 0 cos 5x xS0 x x - 2
xS0 0 x 0 xS2 0 x - 2 0
49. lim- = -1 50. lim+ = 1
x + x cos x x2 - x + sin x
29. lim 30. lim
x S 0 sin x cos x
51. Greatest integer function Find (a) limx S 400+ : x ; and (b)
xS0 2x
31. lim
1 - cos u
32. lim
x - x cos x limx S 400- : x ; ; then use limit definitions to verify your findings.
u S 0 sin 2u xS0 sin2 3x (c) Based on your conclusions in parts (a) and (b), can you say
sin (1 - cos t) sin (sin h) anything about limx S 400 : x ; ? Give reasons for your answer.
33. lim 34. lim
tS0 1 - cos t hS0 sin h
Let ƒ(x) = e
x2 sin (1>x), x 6 0
2x,
sin u sin 5x 52. One-sided limits
35. lim 36. lim x 7 0.
u S 0 sin 2u x S 0 sin 4x
37. lim u cos u 38. lim sin u cot 2u Find (a) limx S 0+ ƒ(x) and (b) limx S 0- ƒ(x); then use limit defini-
uS0 uS0
tions to verify your findings. (c) Based on your conclusions in
tan 3x sin 3y cot 5y parts (a) and (b), can you say anything about limx S 0 ƒ(x)? Give
39. lim 40. lim
xS0 sin 8x yS0 y cot 4y reasons for your answer.
2.5 Continuity
y When we plot function values generated in a laboratory or collected in the field, we often
500 connect the plotted points with an unbroken curve to show what the function’s values are
Q4
likely to have been at the points we did not measure (Figure 2.34). In doing so, we are
Distance fallen (m)
375 Q3 assuming that we are working with a continuous function, so its outputs vary regularly and
Q2 consistently with the inputs, and do not jump abruptly from one value to another without
250
taking on the values in between. Intuitively, any function y = ƒ(x) whose graph can be
Q1
125 sketched over its domain in one unbroken motion is an example of a continuous function.
Such functions play an important role in the study of calculus and its applications.
t
0 5 10
Elapsed time (sec) Continuity at a Point
FIGURE 2.34 Connecting plotted points To understand continuity, it helps to consider a function like that in Figure 2.35, whose
by an unbroken curve from experimental limits we investigated in Example 2 in the last section.
data Q1, Q2, Q3, c for a falling object.
EXAMPLE 1 At which numbers does the function ƒ in Figure 2.35 appear to be not
continuous? Explain why. What occurs at other numbers in the domain?
y Solution First we observe that the domain of the function is the closed interval 3 0, 44,
so we will be considering the numbers x within that interval. From the figure, we notice
y = f (x) right away that there are breaks in the graph at the numbers x = 1, x = 2, and x = 4. The
2
breaks appear as jumps, which we identify later as “jump discontinuities.” These are num-
bers for which the function is not continuous, and we discuss each in turn.
1
Numbers at which the graph of ƒ has breaks:

x
0 1 2 3 4
At x = 1, the function fails to have a limit. It does have both a left-hand limit,
FIGURE 2.35 The function is not limx S 1- ƒ(x) = 0, as well as a right-hand limit, limx S 1+ ƒ(x) = 1, but the limit values are
continuous at x = 1, x = 2, and x = 4 different, resulting in a jump in the graph. The function is not continuous at x = 1.
(Example 1). At x = 2, the function does have a limit, limx S 2 ƒ(x) = 1, but the value of the func-
tion is ƒ(2) = 2. The limit and function values are not the same, so there is a break in the
graph and ƒ is not continuous at x = 2.
At x = 4, the function does have a left-hand limit at this right endpoint, limx S 4- ƒ(x) = 1,
but again the value of the function ƒ(4) = 12 differs from the value of the limit. We see
again a break in the graph of the function at this endpoint and the function is not continu-
ous from the left.
Numbers at which the graph of ƒ has no breaks:
At x = 0, the function has a right-hand limit at this left endpoint, limx S 0+ ƒ(x) = 1,
and the value of the function is the same, ƒ(0) = 1. So no break occurs in the graph of the
function at this endpoint, and the function is continuous from the right at x = 0.
At x = 3, the function has a limit, limx S 3 ƒ(x) = 2. Moreover, the limit is the same
value as the function there, ƒ(3) = 2. No break occurs in the graph and the function is
continuous at x = 3.
At all other numbers x = c in the domain, which we have not considered, the func-
example, limx S 5>2 ƒ(x) = ƒ1 52 2 = 32 . No breaks appear in the graph of the function at any
tion has a limit equal to the value of the function at the point, so limx S c ƒ(x) = ƒ(c). For
of these remaining numbers and the function is continuous at each of them.
The following definitions capture the continuity ideas we observed in Example 1.
DEFINITIONS Let c be a real number on the x-axis.

The function ƒ is continuous at c if
lim ƒ(x) = ƒ(c).
xSc
The function ƒ is right-continuous at c (or continuous from the right) if
lim ƒ(x) = ƒ(c).
x S c+
Continuity Two-sided The function ƒ is left-continuous at c (or continuous from the left) if
from the right continuity Continuity
from the left lim ƒ(x) = ƒ(c).
x S c-
y = f(x)
x From Theorem 6, it follows immediately that a function ƒ is continuous at an interior

a c b
point c of its domain if and only if it is both right-continuous and left-continuous at c (Fig-
FIGURE 2.36 Continuity at points a, b, ure 2.36). We say that a function is continuous over a closed interval 3 a, b4 if it is right-
and c. continuous at a, left-continuous at b, and continuous at all interior points of the interval.
2.5 Continuity 77
y This definition applies to the infinite closed intervals 3 a, q) and (- q, b4 as well, but only
y = "4 − x 2 one endpoint is involved. If a function is not continuous at an interior point c of its domain,
2 we say that ƒ is discontinuous at c, and that c is a point of discontinuity of ƒ. Note that a
function ƒ can be continuous, right-continuous, or left-continuous only at a point c for
which ƒ(c) is defined.
x
The function ƒ(x) = 24 - x2 is continuous over its domain 3 -2, 24

−2 0 2
FIGURE 2.37 A function that EXAMPLE 2

is continuous over its domain (Figure 2.37). It is right-continuous at x = -2, and left-continuous at x = 2.
(Example 2).
EXAMPLE 3 The unit step function U(x), graphed in Figure 2.38, is right-continuous
at x = 0, but is neither left-continuous nor continuous there. It has a jump discontinuity at
x = 0.
y
We summarize continuity at an interior point in the form of a test.
y = U(x)
1
Continuity Test
x
0 A function ƒ(x) is continuous at a point x = c if and only if it meets the follow-
ing three conditions.
FIGURE 2.38 A function
that has a jump discontinuity 1. ƒ(c) exists (c lies in the domain of ƒ).
at the origin (Example 3). 2. limx S c ƒ(x) exists (ƒ has a limit as x S c).
3. limx S c ƒ(x) = ƒ(c) (the limit equals the function value).
For one-sided continuity and continuity at an endpoint of an interval, the limits in

parts 2 and 3 of the test should be replaced by the appropriate one-sided limits.
EXAMPLE 4 The function y = : x ; introduced in Section 1.1 is graphed in Figure 2.39.

y
It is discontinuous at every integer because the left-hand and right-hand limits are not
equal as x S n:
lim : x ; = n - 1 lim : x ; = n .
4
and
Since : n ; = n, the greatest integer function is right-continuous at every integer n (but not
x S n- x S n+
y = :x;
3
2 left-continuous).
The greatest integer function is continuous at every real number other than the inte-
1 gers. For example,
−1 1 2 3 4
x lim : x ; = 1 = : 1.5 ; .
x S 1.5
In general, if n - 1 6 c 6 n, n an integer, then
−2 lim : x ; = n - 1 = : c ; .
xSc
FIGURE 2.39 The greatest integer Figure 2.40 displays several common types of discontinuities. The function in Figure
function is continuous at every noninte- 2.40a is continuous at x = 0. The function in Figure 2.40b would be continuous if it had
ger point. It is right-continuous, but not ƒ(0) = 1. The function in Figure 2.40c would be continuous if ƒ(0) were 1 instead of 2.
left-continuous, at every integer point The discontinuity in Figure 2.40c is removable. The function has a limit as x S 0, and we
(Example 4). can remove the discontinuity by setting ƒ(0) equal to this limit.
The discontinuities in Figure 2.40d through f are more serious: limx S 0 ƒ(x) does not
exist, and there is no way to improve the situation by changing ƒ at 0. The step function in
Figure 2.40d has a jump discontinuity: The one-sided limits exist but have different val-
ues. The function ƒ(x) = 1>x2 in Figure 2.40e has an infinite discontinuity. The function
in Figure 2.40f has an oscillating discontinuity: It oscillates too much to have a limit as
x S 0.
y y y y
2
y = f (x) y = f(x) y = f (x)
y = f (x)
1 1 1 1
x x x x
0 0 0 0
(a) (b) (c) (d)

y y
y = sin 1x
y = f (x) = 12 1
x
x
−1 0
x
0
−1
(e) (f)
FIGURE 2.40 The function in (a) is continuous at x = 0; the functions in (b) through (f ) are not.
Continuous Functions
Generally, we want to describe the continuity behavior of a function throughout its entire
domain, not only at a single point. We know how to do that if the domain is a closed interval.
In the same way, we define a continuous function as one that is continuous at every point in
its domain. This is a property of the function. A function always has a specified domain, so if
we change the domain, we change the function, and this may change its continuity property
as well. If a function is discontinuous at one or more points of its domain, we say it is a
discontinuous function.
EXAMPLE 5
(a) The function y = 1>x (Figure 2.41) is a continuous function because it is continuous
at every point of its domain. It has a point of discontinuity at x = 0, however, because
y
it is not defined there; that is, it is discontinuous on any interval containing x = 0.
(b) The identity function ƒ(x) = x and constant functions are continuous everywhere by
y = 1x Example 3, Section 2.3.
Algebraic combinations of continuous functions are continuous wherever they are defined.
x
0
THEOREM 8—Properties of Continuous Functions If the functions ƒ and g are
continuous at x = c, then the following algebraic combinations are continuous
at x = c.
1. Sums: ƒ + g
FIGURE 2.41 The function y = 1>x 2. Differences: ƒ - g
is continuous over its natural domain. It
3. Constant multiples: k # ƒ, for any number k
has a point of discontinuity at the origin,
so it is discontinuous on any interval 4. Products: ƒ#g
containing x = 0 (Example 5). 5. Quotients: ƒ>g, provided g(c) ≠ 0
6. Powers: ƒn, n a positive integer
2ƒ, provided it is defined on an open interval
n
7. Roots:
containing c, where n is a positive integer
2.5 Continuity 79
Most of the results in Theorem 8 follow from the limit rules in Theorem 1, Section 2.2.
For instance, to prove the sum property we have
lim (ƒ + g)(x) = lim (ƒ(x) + g(x))
xSc xSc
= lim ƒ(x) + lim g(x) Sum Rule, Theorem 1
xSc xSc
= ƒ(c) + g(c) Continuity of ƒ, g at c
= (ƒ + g)(c).
This shows that ƒ + g is continuous.
EXAMPLE 6
(a) Every polynomial P(x) = an xn + an - 1xn - 1 + g + a0 is continuous because
lim P(x) = P(c) by Theorem 2, Section 2.2.
xSc
(b) If P(x) and Q(x) are polynomials, then the rational function P(x)>Q(x) is continuous
wherever it is defined (Q(c) ≠ 0) by Theorem 3, Section 2.2.
EXAMPLE 7 The function ƒ(x) = 0 x 0 is continuous. If x 7 0, we have ƒ(x) = x,

a polynomial. If x 6 0, we have ƒ(x) = -x, another polynomial. Finally, at the origin,
limx S 0 0 x 0 = 0 = 0 0 0 .
The functions y = sin x and y = cos x are continuous at x = 0 by Example 11 of
Section 2.2. Both functions are, in fact, continuous everywhere (see Exercise 70). It fol-
lows from Theorem 8 that all six trigonometric functions are then continuous wherever
they are defined. For example, y = tan x is continuous on g ∪ (-p>2, p>2) ∪
(p>2, 3p>2) ∪ g.
Composites
All composites of continuous functions are continuous. The idea is that if ƒ(x) is continuous
at x = c and g(x) is continuous at x = ƒ(c), then g ∘ ƒ is continuous at x = c (Figure 2.42).
In this case, the limit as x S c is g(ƒ(c)).
g f
˚
Continuous at c
f g
Continuous Continuous
at c at f(c)
c f(c) g( f (c))
FIGURE 2.42 Composites of continuous functions are continuous.
THEOREM 9—Composite of Continuous Functions If ƒ is continuous at c and

g is continuous at ƒ(c), then the composite g ∘ ƒ is continuous at c.
Intuitively, Theorem 9 is reasonable because if x is close to c, then ƒ(x) is close to

ƒ(c), and since g is continuous at ƒ(c), it follows that g(ƒ(x)) is close to g(ƒ(c)).
The continuity of composites holds for any finite number of functions. The only
requirement is that each function be continuous where it is applied. For an outline of a
proof of Theorem 9, see Exercise 6 in Appendix 4.
EXAMPLE 8 Show that the following functions are continuous on their natural
domains.
(a) y = 2x2 - 2x - 5
x2>3
(b) y =
1 + x4
(c) y = ` ` (d) y = ` `
x - 2 x sin x
x2 - 2 x2 + 2
y
Solution
(a) The square root function is continuous on 3 0, q) because it is a root of the continu-
0.4
ous identity function ƒ(x) = x (Part 7, Theorem 8). The given function is then the
g(t) = 2t , and is continuous on its natural domain.

0.3 composite of the polynomial ƒ(x) = x2 - 2x - 5 with the square root function
0.2
(b) The numerator is the cube root of the identity function squared; the denominator is an
(c) The quotient (x - 2)>(x2 - 2) is continuous for all x ≠ { 22, and the function
0.1 everywhere-positive polynomial. Therefore, the quotient is continuous.
x
−2p −p 0 p 2p is the composition of this quotient with the continuous absolute value function
(Example 7).
FIGURE 2.43 The graph suggests that
y = 0 (x sin x)>(x2 + 2) 0 is continuous
(d) Because the sine function is everywhere-continuous (Exercise 70), the numerator term
x sin x is the product of continuous functions, and the denominator term x2 + 2 is an
(Example 8d). everywhere-positive polynomial. The given function is the composite of a quotient of
continuous functions with the continuous absolute value function (Figure 2.43).
Theorem 9 is actually a consequence of a more general result, which we now state
and prove.
THEOREM 10—Limits of Continuous Functions If g is continuous at the point b

and limx S c ƒ(x) = b, then
limx S c g(ƒ(x)) = g(b) = g(limx S c ƒ(x)).
Proof Let P 7 0 be given. Since g is continuous at b, there exists a number d1 7 0

such that
0 g(y) - g(b) 0 6 P whenever 0 6 0 y - b 0 6 d1.
Since limx S c ƒ(x) = b, there exists a d 7 0 such that
0 ƒ(x) - b 0 6 d1 whenever 0 6 0 x - c 0 6 d.
If we let y = ƒ(x), we then have that
0 y - b 0 6 d1 whenever 0 6 0 x - c 0 6 d,
which implies from the first statement that 0 g(y) - g(b) 0 = 0 g(ƒ(x)) - g(b) 0 6 P whenever
0 6 0 x - c 0 6 d. From the definition of limit, this proves that limx S c g(ƒ(x)) = g(b).
EXAMPLE 9 As an application of Theorem 10, we have the following calculation:
lim cos a2x + sin a + xb b = cos a lim 2x + lim sin a + xb b

3p 3p
x S p/2 2 x S p/2 x S p/2 2
= cos (p + sin 2p) = cos p = -1.
2.5 Continuity 81
Intermediate Value Theorem for Continuous Functions

Functions that are continuous on intervals have properties that make them particularly use-
ful in mathematics and its applications. One of these is the Intermediate Value Property. A
function is said to have the Intermediate Value Property if whenever it takes on two
values, it also takes on all the values in between.
THEOREM 11—The Intermediate Value Theorem for Continuous Functions If ƒ is

a continuous function on a closed interval 3 a, b4, and if y0 is any value between
ƒ(a) and ƒ(b), then y0 = ƒ(c) for some c in 3 a, b4.
y
y = f (x)
f(b)
y0
f(a)
x
0 a c b
3
Theorem 11 says that continuous functions over finite closed intervals have the Inter-
mediate Value Property. Geometrically, the Intermediate Value Theorem says that any
2
horizontal line y = y0 crossing the y-axis between the numbers ƒ(a) and ƒ(b) will cross
the curve y = ƒ(x) at least once over the interval 3 a, b4 .
1 The proof of the Intermediate Value Theorem depends on the completeness property
of the real number system (Appendix 7) and can be found in more advanced texts.
x The continuity of ƒ on the interval is essential to Theorem 11. If ƒ is discontinuous at
0 1 2 3 4
even one point of the interval, the theorem’s conclusion may fail, as it does for the func-
FIGURE 2.44 The function tion graphed in Figure 2.44 (choose y0 as any number between 2 and 3).
ƒ(x) = e
2x - 2, 1 … x 6 2
A Consequence for Graphing: Connectedness Theorem 11 implies that the graph of a
3, 2 … x … 4
function continuous on an interval cannot have any breaks over the interval. It will be
does not take on all values between
connected—a single, unbroken curve. It will not have jumps like the graph of the greatest
ƒ(1) = 0 and ƒ(4) = 3; it misses all the
integer function (Figure 2.39), or separate branches like the graph of 1>x (Figure 2.41).
values between 2 and 3.
A Consequence for Root Finding We call a solution of the equation ƒ(x) = 0 a root of
the equation or zero of the function ƒ. The Intermediate Value Theorem tells us that if ƒ is
continuous, then any interval on which ƒ changes sign contains a zero of the function.
In practical terms, when we see the graph of a continuous function cross the horizon-
tal axis on a computer screen, we know it is not stepping across. There really is a point
where the function’s value is zero.
EXAMPLE 10 Show that there is a root of the equation x3 - x - 1 = 0 between 1 and 2.
Solution Let ƒ(x) = x3 - x - 1. Since ƒ(1) = 1 - 1 - 1 = -1 6 0 and ƒ(2) =

23 - 2 - 1 = 5 7 0, we see that y0 = 0 is a value between ƒ(1) and ƒ(2). Since ƒ is
continuous, the Intermediate Value Theorem says there is a zero of ƒ between 1 and 2.
Figure 2.45 shows the result of zooming in to locate the root near x = 1.32.
5 1
1 1.6
−1 2
−2 −1
(a) (b)
0.02 0.003
1.320 1.330 1.3240 1.3248
−0.02 −0.003
(c) (d)
FIGURE 2.45 Zooming in on a zero of the function ƒ(x) = x3 - x - 1. The zero is near
x = 1.3247 (Example 10).
EXAMPLE 11 Use the Intermediate Value Theorem to prove that the equation
22x + 5 = 4 - x2
y
has a solution (Figure 2.46).
4 y = 4 − x2
Solution We rewrite the equation as
22x + 5 + x2 = 4,
3
and set ƒ(x) = 22x + 5 + x2 . Now g(x) = 22x + 5 is continuous on the interval
2
y = " 2x + 5 1 3 -5>2, q) since it is the composite of the square root function with the nonnegative linear
function y = 2x + 5. Then ƒ is the sum of the function g and the quadratic function y = x2,
and the quadratic function is continuous for all values of x. It follows that ƒ(x) = 22x + 5
x
0 c 2
+ x2 is continuous on the interval 3 -5>2, q). By trial and error, we find the function values
ƒ(0) = 25 ≈ 2.24 and ƒ(2) = 29 + 4 = 7, and note that ƒ is also continuous on the
y = 22x + 5 and y = 4 - x2
FIGURE 2.46 The curves
finite closed interval 3 0, 24 ⊂ 3 -5>2, q). Since the value y0 = 4 is between the numbers
22x + 5 = 4 - x2 (Example 11).

have the same value at x = c where 2.24 and 7, by the Intermediate Value Theorem there is a number c∊[0, 24 such that
ƒ(c) = 4. That is, the number c solves the original equation.
Continuous Extension to a Point

Sometimes the formula that describes a function ƒ does not make sense at a point x = c.
It might nevertheless be possible to extend the domain of ƒ, to include x = c, creating a
new function that is continuous at x = c. For example, the function y = ƒ(x) = (sin x)>x
is continuous at every point except x = 0, since the origin is not in its domain. Since
y = (sin x)>x has a finite limit as x S 0 (Theorem 7), we can extend the function’s
domain to include the point x = 0 in such a way that the extended function is continuous
at x = 0. We define the new function
sin x
, x ≠ 0
F(x) = W x
1, x = 0.
2.5 Continuity 83
The function F(x) is continuous at x = 0 because

sin x
lim x = F(0),
xS0
so it meets the requirements for continuity (Figure 2.47).
y y
(0, 1) (0, 1)
f (x) F(x)
a− , b a , b a− , b a , b
p 2 p 2 p 2 p 2
2 p 2 p 2 p 2 p
p p x p p x
− 0 − 0
2 2 2 2
(a) (b)
FIGURE 2.47 The graph (a) of ƒ(x) = (sin x)>x for - p>2 … x … p>2 does not include
the point (0, 1) because the function is not defined at x = 0. (b) We can remove the discon-
tinuity from the graph by defining the new function F(x) with F(0) = 1 and F(x) = ƒ(x)
everywhere else. Note that F(0) = limx S 0 ƒ(x).
More generally, a function (such as a rational function) may have a limit at a point
where it is not defined. If ƒ(c) is not defined, but limx S c ƒ(x) = L exists, we can define a
new function F(x) by the rule
F(x) = e
ƒ(x), if x is in the domain of ƒ
L, if x = c.
The function F is continuous at x = c. It is called the continuous extension of ƒ to
x = c. For rational functions ƒ, continuous extensions are often found by canceling com-
mon factors in the numerator and denominator.
EXAMPLE 12 Show that

x2 + x - 6
ƒ(x) = , x ≠ 2
y x2 - 4
x2 + x − 6 has a continuous extension to x = 2, and find that extension.
2 y=
x2 − 4
1 Solution Although ƒ(2) is not defined, if x ≠ 2 we have
x x2 + x - 6 (x - 2)(x + 3) x + 3
−1 0 1 2 3 4 ƒ(x) = = = .
x2 - 4 (x - 2)(x + 2) x + 2
(a)
y The new function
x+3
y= x + 3
2 x+2 F(x) =
5 x + 2
4 1
is equal to ƒ(x) for x ≠ 2, but is continuous at x = 2, having there the value of 5>4. Thus
−1 0 1 2 3 4
x F is the continuous extension of ƒ to x = 2, and
(b) x2 + x - 6 5
lim = lim ƒ(x) = .
xS2 2
x - 4 x S 2 4
FIGURE 2.48 (a) The graph
of ƒ(x) and (b) the graph of The graph of ƒ is shown in Figure 2.48. The continuous extension F has the same graph
its continuous extension F(x) except with no hole at (2, 5>4). Effectively, F is the function ƒ with its point of disconti-
(Example 12). nuity at x = 2 removed.
Exercises 2.5
Continuity from Graphs 7. a. Is ƒ defined at x = 2? (Look at the definition of ƒ.)
In Exercises 1–4, say whether the function graphed is continuous on b. Is ƒ continuous at x = 2?
3- 1, 34. If not, where does it fail to be continuous and why?
8. At what values of x is ƒ continuous?
1. 2.
9. What value should be assigned to ƒ(2) to make the extended
y y function continuous at x = 2?
y = f(x) y = g(x) 10. To what new value should ƒ(1) be changed to remove the discon-
2 2 tinuity?
1
1 Applying the Continuity Test
x x At which points do the functions in Exercises 11 and 12 fail to be con-
−1 0 1 2 3 −1 0 1 2 3 tinuous? At which points, if any, are the discontinuities removable?
Not removable? Give reasons for your answers.
3. 4.
11. Exercise 1, Section 2.4 12. Exercise 2, Section 2.4
y y
At what points are the functions in Exercises 13–30 continuous?
y = k(x) y = k(x)
2 1 1
2 13. y = - 3x 14. y = + 4
x - 2 (x + 2)2
1 1
x + 1 x + 3
15. y = 16. y =
x x x2 - 4x + 3 x2 - 3x - 10
−1 0 1 2 3 −1 0 1 2 3
17. y = 0 x - 1 0 + sin x
1 x2
0x0 + 1
18. y = -
2
Exercises 5–10 refer to the function
cos x x + 2
x2 - 1, -1 … x 6 0 19. y = x 20. y = cos x
2x, 0 6 x 6 1 px
21. y = csc 2x 22. y = tan
ƒ(x) = e 1, x = 1 2
-2x + 4, 1 6 x 6 2 x tan x 2x4 + 1
0, 2 6 x 6 3 23. y = 24. y =
x2 + 1 1 + sin2 x
25. y = 22x + 3 26. y = 2

4
graphed in the accompanying figure. 3x - 1
y 27. y = (2x - 1)1>3 28. y = (2 - x)1>5

y = f (x)
2 (1, 2) x2 - x - 6
, x ≠ 3
29. g(x) = c x - 3
y = 2x y = −2x + 4 5, x = 3
(1, 1)
x3 - 8
x , x ≠ 2, x ≠ -2
−1 0 1 2 3 x2 - 4
30. ƒ(x) = d
3, x = 2
y = x2 − 1 −1
4, x = -2
The graph for Exercises 5–10.
5. a. Does ƒ(- 1) exist? Limits Involving Trigonometric Functions
Find the limits in Exercises 31–38. Are the functions continuous at the
b. Does limx S -1+ ƒ(x) exist? point being approached?
c. Does limx S -1+ ƒ(x) = ƒ(- 1)? p
31. lim sin (x - sin x) 32. lim sina cos (tan t)b
d. Is ƒ continuous at x = - 1? xSp tS0 2
6. a. Does ƒ(1) exist? 33. lim sec (y sec2 y - tan2 y - 1)

yS1
b. Does limx S 1 ƒ(x) exist?
cos ( sin x1>3 ) b
p
c. Does limx S 1 ƒ(x) = ƒ(1)? 34. lim tana
xS0 4
d. Is ƒ continuous at x = 1?
2.5 Continuity 85
48. For what values of a and b is

35. lim cos a b
219 - 3 sec 2t
p
tS0
ax + 2b, x … 0
36. lim 2csc2 x + 5 13 tan x g(x) = c x2 + 3a - b, 0 6 x … 2
x S p/6
3x - 5, x 7 2
A
cos2 x - cos x
37. lim sin x
xS0 continuous at every x?
38. lim sec a b

p(sin 2x - sin x)
T In Exercises 49–52, graph the function ƒ to see whether it appears to
xS0 3x have a continuous extension to the origin. If it does, use Trace and Zoom
to find a good candidate for the extended function’s value at x = 0. If
Continuous Extensions the function does not appear to have a continuous extension, can it be
39. Define g(3) in a way that extends g(x) = (x2 - 9)>(x - 3) to be extended to be continuous at the origin from the right or from the left? If
continuous at x = 3. so, what do you think the extended function’s value(s) should be?
10 0 x 0 - 1
40. Define h(2) in a way that extends h(t) = (t 2 + 3t - 10)>(t - 2)
10 x - 1
to be continuous at t = 2. 49. ƒ(x) = x 50. ƒ(x) = x
41. Define ƒ(1) in a way that extends ƒ(s) = (s3 - 1)>(s2 - 1) to be
sin x
0x0
continuous at s = 1. 51. ƒ(x) = 52. ƒ(x) = (1 + 2x)1>x
42. Define g(4) in a way that extends
g(x) = (x2 - 16)> (x2 - 3x - 4) Theory and Examples
to be continuous at x = 4. 53. A continuous function y = ƒ(x) is known to be negative at
x = 0 and positive at x = 1. Why does the equation ƒ(x) = 0
43. For what value of a is
have at least one solution between x = 0 and x = 1? Illustrate
ƒ(x) = e
x2 - 1, x 6 3 with a sketch.
2ax, x Ú 3 54. Explain why the equation cos x = x has at least one solution.
continuous at every x? 55. Roots of a cubic Show that the equation x3 - 15x + 1 = 0
has three solutions in the interval 3-4, 4].
44. For what value of b is
56. A function value Show that the function F(x) = (x - a)2 #
g(x) = e 2
x, x 6 -2 (x - b)2 + x takes on the value (a + b)>2 for some value of x.
bx , x Ú - 2
are values c for which ƒ(c) equals (a) p; (b) - 23; (c) 5,000,000.
57. Solving an equation If ƒ(x) = x3 - 8x + 10, show that there
continuous at every x?
45. For what values of a is 58. Explain why the following five statements ask for the same infor-
mation.
a2x - 2a, x Ú 2 a. Find the roots of ƒ(x) = x3 - 3x - 1.
ƒ(x) = b
12, x 6 2
b. Find the x-coordinates of the points where the curve y = x3
continuous at every x? crosses the line y = 3x + 1.
46. For what value of b is c. Find all the values of x for which x3 - 3x = 1.
x - b d. Find the x-coordinates of the points where the cubic curve
, x 6 0 y = x3 - 3x crosses the line y = 1.
g(x) = c b + 1
x2 + b, x 7 0 e. Solve the equation x3 - 3x - 1 = 0.
continuous at every x? 59. Removable discontinuity Give an example of a function ƒ(x)
that is continuous for all values of x except x = 2, where it has
47. For what values of a and b is
a removable discontinuity. Explain how you know that ƒ is dis-
- 2, x … -1 continuous at x = 2, and how you know the discontinuity is
ƒ(x) = c ax - b, -1 6 x 6 1 removable.
3, x Ú 1 60. Nonremovable discontinuity Give an example of a function
g(x) that is continuous for all values of x except x = -1, where it
continuous at every x? has a nonremovable discontinuity. Explain how you know that g
is discontinuous there and why the discontinuity is not removable.
61. A function discontinuous at every point 68. The sign-preserving property of continuous functions Let ƒ
a. Use the fact that every nonempty interval of real numbers be defined on an interval (a, b) and suppose that ƒ(c) ≠ 0 at
contains both rational and irrational numbers to show that the some c where ƒ is continuous. Show that there is an interval
function (c - d, c + d) about c where ƒ has the same sign as ƒ(c).
69. Prove that ƒ is continuous at c if and only if
ƒ(x) = e
1, if x is rational
0, if x is irrational lim ƒ(c + h) = ƒ(c).
hS0
is discontinuous at every point. 70. Use Exercise 69 together with the identities
b. Is ƒ right-continuous or left-continuous at any point? sin (h + c) = sin h cos c + cos h sin c,
62. If functions ƒ(x) and g(x) are continuous for 0 … x … 1, could
ƒ(x)>g(x) possibly be discontinuous at a point of 30, 14? Give
cos (h + c) = cos h cos c - sin h sin c
reasons for your answer. to prove that both ƒ(x) = sin x and g(x) = cos x are continuous
63. If the product function h(x) = ƒ(x) # g(x) is continuous at x = 0, at every point x = c.
must ƒ(x) and g(x) be continuous at x = 0? Give reasons for
your answer. Solving Equations Graphically
64. Discontinuous composite of continuous functions Give an T Use the Intermediate Value Theorem in Exercises 71–76 to prove that
example of functions ƒ and g, both continuous at x = 0, for each equation has a solution. Then use a graphing calculator or com-
which the composite ƒ ∘ g is discontinuous at x = 0. Does this puter grapher to solve the equations.
contradict Theorem 9? Give reasons for your answer. 71. x3 - 3x - 1 = 0
65. Never-zero continuous functions Is it true that a continuous 72. 2x3 - 2x2 - 2x + 1 = 0
function that is never zero on an interval never changes sign on 73. x(x - 1)2 = 1 (one root)
74. 2x + 21 + x = 4
that interval? Give reasons for your answer.
66. Stretching a rubber band Is it true that if you stretch a rubber
75. cos x = x (one root). Make sure you are using radian mode.
band by moving one end to the right and the other to the left,
some point of the band will end up in its original position? Give 76. 2 sin x = x (three roots). Make sure you are using radian
reasons for your answer. mode.
67. A fixed point theorem Suppose that a function ƒ is continuous
on the closed interval 30, 14 and that 0 … ƒ(x) … 1 for every x
in 30, 14 . Show that there must exist a number c in 30, 14 such
that ƒ(c) = c (c is called a fixed point of ƒ).
2.6 Limits Involving Infinity; Asymptotes of Graphs

y In this section we investigate the behavior of a function when the magnitude of the inde-
pendent variable x becomes increasingly large, or x S { q . We further extend the con-
4 cept of limit to infinite limits, which are not limits as before, but rather a new use of the
term limit. Infinite limits provide useful symbols and language for describing the behavior
3
of functions whose values become arbitrarily large in magnitude. We use these limit ideas
2 y = 1x
to analyze the graphs of functions having horizontal or vertical asymptotes.
1
x Finite Limits as x u t H
−1 0 1 2 3 4
−1
The symbol for infinity (q) does not represent a real number. We use q to describe the
behavior of a function when the values in its domain or range outgrow all finite bounds.
For example, the function ƒ(x) = 1>x is defined for all x ≠ 0 (Figure 2.49). When x is
positive and becomes increasingly large, 1>x becomes increasingly small. When x is
negative and its magnitude becomes increasingly large, 1>x again becomes small. We
summarize these observations by saying that ƒ(x) = 1>x has limit 0 as x S q or
FIGURE 2.49 The graph of y = 1>x x S - q, or that 0 is a limit of ƒ(x) = 1>x at infinity and negative infinity. Here are pre-
approaches 0 as x S q or x S - q . cise definitions.
2.6 Limits Involving Infinity; Asymptotes of Graphs 87
DEFINITIONS
1. We say that ƒ(x) has the limit L as x approaches infinity and write
lim ƒ(x) = L
xS q
if, for every number P 7 0, there exists a corresponding number M such that
for all x
x 7 M 1 0 ƒ(x) - L 0 6 P.
2. We say that ƒ(x) has the limit L as x approaches minus infinity and write
lim ƒ(x) = L
xS -q
if, for every number P 7 0, there exists a corresponding number N such that
for all x
x 6 N 1 0 ƒ(x) - L 0 6 P.
Intuitively, limx Sq ƒ(x) = L if, as x moves increasingly far from the origin in the positive
direction, ƒ(x) gets arbitrarily close to L. Similarly, limx S -q ƒ(x) = L if, as x moves
increasingly far from the origin in the negative direction, ƒ(x) gets arbitrarily close to L.
The strategy for calculating limits of functions as x S { q is similar to the one for
finite limits in Section 2.2. There we first found the limits of the constant and identity
functions y = k and y = x. We then extended these results to other functions by applying
Theorem 1 on limits of algebraic combinations. Here we do the same thing, except that the
starting functions are y = k and y = 1>x instead of y = k and y = x.
The basic facts to be verified by applying the formal definition are
No matter what
y 1
positive number P is, lim k = k and lim = 0. (1)
the graph enters x S {q x S {q x
1
this band at x = P We prove the second result in Example 1, and leave the first to Exercises 87 and 88.
y = 1x
and stays.
y=P
EXAMPLE 1 Show that
P 1 1
1 (a) lim x = 0 (b) lim x = 0.
N = −P x Sq x S -q
x
0
M = 1P Solution
y = –P (a) Let P 7 0 be given. We must find a number M such that for all x
–P
` x - 0 ` = ` x ` 6 P.
1 1
No matter what x 7 M 1
positive number P is,
the graph enters
1 The implication will hold if M = 1>P or any larger positive number (Figure 2.50).
this band at x = − P
and stays. This proves limx Sq (1>x) = 0.
(b) Let P 7 0 be given. We must find a number N such that for all x
FIGURE 2.50 The geometry behind the
` x - 0 ` = ` x ` 6 P.
argument in Example 1. 1 1
x 6 N 1
The implication will hold if N = -1>P or any number less than -1>P (Figure 2.50).
This proves limx S - q (1>x) = 0.
Limits at infinity have properties similar to those of finite limits.
THEOREM 12 All the Limit Laws in Theorem 1 are true when we replace
limx S c by limx S q or limx S - q . That is, the variable x may approach a finite
number c or { q.
EXAMPLE 2 The properties in Theorem 12 are used to calculate limits in the same
way as when x approaches a finite number c.
lim a5 + x b = lim 5 + lim x

1 1
(a) Sum Rule
xS q xS q xS q
= 5 + 0 = 5 Known limits
p 23
= lim p 23 # x # x
1 1
(b) lim
xS - q x2 xS - q
= lim p 23 # lim x # lim x

1 1
Product Rule
xS - q xS - q xS - q
= p 23 # 0 # 0 = 0 Known limits
y 2
y = 5x +2 8x − 3
Limits at Infinity of Rational Functions
3x + 2
2 To determine the limit of a rational function as x S { q, we first divide the numerator
and denominator by the highest power of x in the denominator. The result then depends on
Line y = 5
1 3 the degrees of the polynomials involved.
x EXAMPLE 3 These examples illustrate what happens when the degree of the numera-
−5 0 5 10 tor is less than or equal to the degree of the denominator.
−1 5 + (8>x) - (3>x2)
5x2 + 8x - 3 Divide numerator and
(a) lim = lim denominator by x2 .
−2
xS q 3x2 + 2 xS q 3 + (2>x2)
NOT TO SCALE
5 + 0 - 0 5
= = See Fig. 2.51.
FIGURE 2.51 The graph of the func- 3 + 0 3
tion in Example 3a. The graph approaches
the line y = 5>3 as 0 x 0 increases.
11x + 2 (11>x2) + (2>x3) Divide numerator and
(b) lim = lim denominator by x3 .
x S - q 2x 3 - 1 xS - q 2 - (1>x3)
0 + 0
= = 0 See Fig. 2.52.
y 2 - 0
8
11x + 2
Cases for which the degree of the numerator is greater than the degree of the denomi-
y= nator are illustrated in Examples 9 and 13.
2x 3 − 1
6
Horizontal Asymptotes
4
If the distance between the graph of a function and some fixed line approaches zero as a
2 point on the graph moves increasingly far from the origin, we say that the graph approaches
the line asymptotically and that the line is an asymptote of the graph.
x Looking at ƒ(x) = 1>x (see Figure 2.49), we observe that the x-axis is an asymptote
−4 −2 0 2 4 6
of the curve on the right because
−2
1
lim x = 0
−4 xS q
and on the left because

−6
1
−8
lim x = 0.
xS - q
We say that the x-axis is a horizontal asymptote of the graph of ƒ(x) = 1>x.
FIGURE 2.52 The graph of the
function in Example 3b. The graph
approaches the x-axis as 0 x 0 increases.
DEFINITION A line y = b is a horizontal asymptote of the graph of a func-
tion y = ƒ(x) if either
lim ƒ(x) = b or lim ƒ(x) = b.
xS q xS - q
The graph of the function

5x2 + 8x - 3
ƒ(x) =
3x2 + 2
sketched in Figure 2.51 (Example 3a) has the line y = 5>3 as a horizontal asymptote on
both the right and the left because
5 5
lim ƒ(x) = and lim ƒ(x) = .
xS q 3 xS - q 3
EXAMPLE 4 Find the horizontal asymptotes of the graph of

y
x3 - 2
0x03 + 1
2 ƒ(x) = .
y=1
x Solution We calculate the limits as x S { q.

0
y = −1 x3 - 2 x3 - 2 1 - (2>x3)
xS q 0 x 0 3 + 1
For x Ú 0: lim = lim = lim = 1.
x3 − 2 x S q x3 + 1 x S q 1 + (1>x 3)
0x03 + 1
f(x) =
−2
x3 - 2 x3 - 2 1 - ( 2>x3 )
xS - q 0 x 0 + 1
For x 6 0: lim = lim = lim = -1.
FIGURE 2.53 The graph of the 3 x S - q ( -x ) + 1
3 x S - q -1 + ( 1>x 3 )
function in Example 4 has two
horizontal asymptotes. The horizontal asymptotes are y = -1 and y = 1. The graph is displayed in Figure
2.53. Notice that the graph crosses the horizontal asymptote y = -1 for a positive value
of x.
EXAMPLE 5 Find (a) lim sin (1>x) and (b) lim x sin (1>x).
xS q x S {q
Solution
(a) We introduce the new variable t = 1>x. From Example 1, we know that t S 0+ as
x S q (see Figure 2.49). Therefore,
1
lim sin x = lim+ sin t = 0.
y xS q tS0
1 (b) We calculate the limits as x S q and x S - q:
1 sin t 1 sin t
lim x sin x = lim+ t = 1 and lim x sin x = lim- t = 1.
xS q tS0 xS - q tS0
1
y = x sin x
The graph is shown in Figure 2.54, and we see that the line y = 1 is a horizontal
x
−1 1 asymptote.
FIGURE 2.54 The line y = 1 is a The Sandwich Theorem also holds for limits as x S { q. You must be sure, though,
horizontal asymptote of the function that the function whose limit you are trying to find stays between the bounding functions
graphed here (Example 5b). at very large values of x in magnitude consistent with whether x S q or x S - q.
EXAMPLE 6 Using the Sandwich Theorem, find the horizontal asymptote of the curve
sin x
y = 2 + x .
y Solution We are interested in the behavior as x S { q. Since
0 … ` x ` … `x`
sin x 1
y = 2 + sinx x
and limx S {q 0 1>x 0 = 0, we have limx S {q (sin x)>x = 0 by the Sandwich Theorem.
2
1 Hence,
lim a2 + x b = 2 + 0 = 2,
sin x
x
−3p −2p −p 0 p 2p 3p x S {q
FIGURE 2.55 A curve may cross one of and the line y = 2 is a horizontal asymptote of the curve on both left and right (Figure 2.55).
its asymptotes infinitely often (Example 6). This example illustrates that a curve may cross one of its horizontal asymptotes many
times.
We can investigate the behavior of y = ƒ(1>x) as x S 0 by investigating y = ƒ(t) as

t S { q , where t = 1>x.
Find lim+ x j x k .
1
EXAMPLE 7
xS0
y Solution We let t = 1>x so that

y=t
lim x j x k = lim t : t ;
3 1 1
xS0 + tS q
:t;
2
y=t −1 From the graph in Figure 2.56, we see that t - 1 … … t, which gives
1
1 - t … t :t; … 1
t 1 1 1
−2 −1 1 2 3 Multiply inequalities by t 7 0.
−2 It follows from the Sandwich Theorem that

lim t : t ; = 1,
1
FIGURE 2.56 The graph of the greatest tS q
integer function y = : t ; is sandwiched so 1 is the value of the limit we seek.

between y = t - 1 and y = t.
EXAMPLE 8 Find lim 1 x - 2x2 + 16 2.
Solution Both of the terms x and 2x2 + 16 approach infinity as x S q, so what hap-
xS q
pens to the difference in the limit is unclear (we cannot subtract q from q because the
symbol does not represent a real number). In this situation we can multiply the numerator
and the denominator by the conjugate radical expression to obtain an equivalent algebraic
result:
x + 2x2 + 16
lim 1 x - 2x2 + 16 2 = lim 1 x - 2x2 + 16 2
xS q xS q x + 2x2 + 16
x2 - (x2 + 16)
+ 2x + 16 + 2x2 + 16
-16
= lim 2
= lim .
xS q x xS q x
As x S , the denominator in this last expression becomes arbitrarily large, so we see that
q
the limit is 0. We can also obtain this result by a direct calculation using the Limit Laws:
16
- x
x S q x + 2x 2 + 16 1 + 21 + 0
-16 0
lim = lim = = 0.
A x2
xS q 2
x 16
1 + +
x2
2
Oblique Asymptotes
y= x −3=x+1+ 1
2x − 4 2 2x − 4 If the degree of the numerator of a rational function is 1 greater than the degree of the
y denominator, the graph has an oblique or slant line asymptote. We find an equation for
The vertical distance
between curve and the asymptote by dividing numerator by denominator to express ƒ as a linear function plus
6 line goes to zero as x : ∞ a remainder that goes to zero as x S { q.
5
4 Oblique EXAMPLE 9 Find the oblique asymptote of the graph of
x=2
3 asymptote
2 y= x+1 x2 - 3
2 ƒ(x) =
2x - 4
1
x in Figure 2.57.
−1 0 1 2 3 4 x
−1 Solution We are interested in the behavior as x S { q . We divide (2x - 4) into
−2 (x2 - 3):
−3
x
+ 1
2
FIGURE 2.57 The graph of the function 2x - 4) x2 - 3
in Example 9 has an oblique asymptote. x2 - 2x
2x - 3
2x - 4
1
This tells us that
x2 - 3 x 1
ƒ(x) = = ¢ + 1≤ + ¢ ≤.
2x - 4 2()* 2x - 4
(1)1*
linear g(x) remainder
As x S { q , the remainder, whose magnitude gives the vertical distance between the
graphs of ƒ and g, goes to zero, making the slanted line
x
g(x) = + 1
2
an asymptote of the graph of ƒ (Figure 2.57). The line y = g(x) is an asymptote both to the
right and to the left. The next subsection will confirm that the function ƒ(x) grows arbitrarily
large in absolute value as x S 2 (where the denominator is zero), as shown in the graph.
y
You can get as high Notice in Example 9 that if the degree of the numerator in a rational function is greater
as you want by
taking x close enough
than the degree of the denominator, then the limit as 0 x 0 becomes large is + q or - q,
to 0. No matter how depending on the signs assumed by the numerator and denominator.
high B is, the graph
B goes higher.
Infinite Limits
y = 1x Let us look again at the function ƒ(x) = 1>x. As x S 0+, the values of ƒ grow without
x bound, eventually reaching and surpassing every positive real number. That is, given any
x
0 x positive real number B, however large, the values of ƒ become larger still (Figure 2.58).
No matter how
low −B is, the Thus, ƒ has no limit as x S 0+. It is nevertheless convenient to describe the behavior of ƒ
graph goes lower. by saying that ƒ(x) approaches q as x S 0+. We write
You can get as low as −B
1
you want by taking lim ƒ(x) = lim+ x = q.
x close enough to 0. x S 0+ xS0
FIGURE 2.58 One-sided infinite limits: In writing this equation, we are not saying that the limit exists. Nor are we saying that there
1 1 is a real number q, for there is no such number. Rather, we are saying that limx S 0+ (1>x)
lim+ x = q and lim- x = - q.
xS0 xS0 does not exist because 1>x becomes arbitrarily large and positive as x S 0+.
As x S 0 -, the values of ƒ(x) = 1>x become arbitrarily large and negative. Given
any negative real number -B, the values of ƒ eventually lie below -B. (See Figure 2.58.)
We write
1
lim ƒ(x) = lim- x = - q.
x S 0- xS0
y Again, we are not saying that the limit exists and equals the number - q. There is no real
number - q. We are describing the behavior of a function whose limit as x S 0- does not
y= 1 exist because its values become arbitrarily large and negative.
x−1
1
1 1
EXAMPLE 10 Find lim+ and lim .
x xS1 x - 1 x S 1- x - 1
−1 0 1 2 3
Geometric Solution The graph of y = 1>(x - 1) is the graph of y = 1>x shifted 1
unit to the right (Figure 2.59). Therefore, y = 1>(x - 1) behaves near 1 exactly the way
y = 1>x behaves near 0:
1 1
lim = q and lim = - q.
FIGURE 2.59 Near x = 1, the func- x S 1+ x - 1 x S 1- x - 1
tion y = 1>(x - 1) behaves the way the
function y = 1>x behaves near x = 0. Its Analytic Solution Think about the number x - 1 and its reciprocal. As x S 1+, we
graph is the graph of y = 1>x shifted have (x - 1) S 0+ and 1>(x - 1) S q. As x S 1-, we have (x - 1) S 0 - and
1 unit to the right (Example 10). 1>(x - 1) S - q.
EXAMPLE 11 Discuss the behavior of
1
ƒ(x) = as x S 0.
y x2
Solution As x approaches zero from either side, the values of 1>x2 are positive and
No matter how become arbitrarily large (Figure 2.60). This means that
B high B is, the graph
goes higher. 1
lim ƒ(x) = lim = q.
xS0 xS0 x2
f (x) = 12
x The function y = 1>x shows no consistent behavior as x S 0. We have 1>x S q if
x S 0+, but 1>x S - q if x S 0-. All we can say about limx S 0 (1>x) is that it does not
x 0 x
x exist. The function y = 1>x2 is different. Its values approach infinity as x approaches zero
from either side, so we can say that limx S 0 (1>x2) = q.
FIGURE 2.60 The graph of ƒ(x) in
Example 11 approaches infinity as x S 0.
EXAMPLE 12 These examples illustrate that rational functions can behave in various
ways near zeros of the denominator.
(x - 2)2 (x - 2)2 x - 2
(a) lim = lim = lim = 0
2
xS2 x - 4 x S 2 (x - 2)(x + 2) xS2 x + 2
x - 2 x - 2 1 1
(b) lim = lim = lim =
x2 - 4 x S 2 (x - 2)(x + 2) x S 2 x + 2 4
xS2
x - 3 x - 3 The values are negative

(c) lim+ 2 = lim+ = -q for x 7 2, x near 2.
xS2 x - 4 x S 2 (x - 2)(x + 2)
x - 3 x - 3 The values are positive

(d) lim- = lim = q for x 6 2, x near 2.
xS2 x2 - 4 x S 2- (x - 2)(x + 2)
x - 3 x - 3
(e) lim = lim does not exist. See parts (c) and (d).
xS2 x2 - 4 x S 2 (x - 2)(x + 2)
2 - x -(x - 2) -1
(f ) lim 3 = xlim = lim = -q
xS2 (x - 2) S 2 (x - 2)3 x S 2 (x - 2)2
In parts (a) and (b) the effect of the zero in the denominator at x = 2 is canceled
because the numerator is zero there also. Thus a finite limit exists. This is not true in part
(f ), where cancellation still leaves a zero factor in the denominator.
2x5 - 6x4 + 1
EXAMPLE 13 Find lim .
xS - q 3x2 + x - 7
y
Solution We are asked to find the limit of a rational function as x S - q , so we divide
the numerator and denominator by x2 , the highest power of x in the denominator:
2x5 - 6x4 + 1 2x3 - 6x2 + x-2
y = f(x) lim = lim
x S - q 3x 2 + x - 7 x S - q 3 + x -1 - 7x-2
2x2 (x - 3) + x-2
B = lim
x S - q 3 + x -1 - 7x-2
= - q, x-n S 0, x - 3 S - q
because the numerator tends to - q while the denominator approaches 3 as x S - q .

x
0 c
c−d c+d
Precise Definitions of Infinite Limits
Instead of requiring ƒ(x) to lie arbitrarily close to a finite number L for all x sufficiently
FIGURE 2.61 For c - d 6 x 6 c + d, close to c, the definitions of infinite limits require ƒ(x) to lie arbitrarily far from zero.
the graph of ƒ(x) lies above the line y = B. Except for this change, the language is very similar to what we have seen before. Figures 2.61
and 2.62 accompany these definitions.
DEFINITIONS
c−d c+d
c
x
1. We say that ƒ(x) approaches infinity as x approaches c, and write
0
lim ƒ(x) = q,
xSc
if for every positive real number B there exists a corresponding d 7 0 such

that for all x
0 6 0x - c0 6 d
−B
1 ƒ(x) 7 B.
2. We say that ƒ(x) approaches minus infinity as x approaches c, and write
y = f(x)
lim ƒ(x) = - q,
xSc
if for every negative real number -B there exists a corresponding d 7 0

FIGURE 2.62 For c - d 6 x 6 c + d,
such that for all x
the graph of ƒ(x) lies below the line 0 6 0x - c0 6 d 1 ƒ(x) 6 -B.
y = - B.
The precise definitions of one-sided infinite limits at c are similar and are stated in the
exercises.
1
EXAMPLE 14 Prove that lim = q.
xS0 x2
Solution Given B 7 0, we want to find d 7 0 such that
1
0 6 x - 0 6 d implies 7 B.
x2
Now,
1 1
7 B if and only if x2 6
x2 B
or, equivalently,
0x0 6 1 .
2B
Thus, choosing d = 1> 2B (or any smaller positive number), we see that
0 x 0 6 d implies 1 1
7 2 Ú B.
x2 d
Therefore, by definition,
1
lim = q.
xS0 x2
Vertical Asymptotes
y Notice that the distance between a point on the graph of ƒ(x) = 1>x and the y-axis
approaches zero as the point moves vertically along the graph and away from the origin
Vertical asymptote (Figure 2.63). The function ƒ(x) = 1>x is unbounded as x approaches 0 because
1 1
lim = q and lim = - q.
y = 1x x S 0+ x x S 0- x
Horizontal 1
asymptote We say that the line x = 0 (the y-axis) is a vertical asymptote of the graph of ƒ(x) = 1>x.
x
0 1 Horizontal Observe that the denominator is zero at x = 0 and the function is undefined there.
asymptote,
y=0
DEFINITION A line x = a is a vertical asymptote of the graph of a function
Vertical asymptote, y = ƒ(x) if either
x=0
lim ƒ(x) = { q or lim ƒ(x) = { q.
x S a+ x S a-
FIGURE 2.63 The coordinate axes are
asymptotes of both branches of the hyper-
bola y = 1>x. EXAMPLE 15 Find the horizontal and vertical asymptotes of the curve
x + 3
y = .
x + 2
Solution We are interested in the behavior as x S { q and the behavior as x S -2,
where the denominator is zero.
The asymptotes are quickly revealed if we recast the rational function as a polynomial
with a remainder, by dividing (x + 2) into (x + 3):
1
x + 2) x + 3
x + 2
1
y This result enables us to rewrite y as:

Vertical
asymptote, 6
1
x = −2 5 x+3 y = 1 + .
y= x + 2
4 x+2
1 As x S { q , the curve approaches the horizontal asymptote y = 1; as x S -2, the curve
3 =1+
Horizontal x+2
asymptote, approaches the vertical asymptote x = -2. We see that the curve in question is the graph
2
y=1 of ƒ(x) = 1>x shifted 1 unit up and 2 units left (Figure 2.64). The asymptotes, instead of
1
being the coordinate axes, are now the lines y = 1 and x = -2.
x
−5 −4 −3 −2 −1 0 1 2 3
−1 EXAMPLE 16 Find the horizontal and vertical asymptotes of the graph of
−2
8
−3 ƒ(x) = - .
x2 - 4
−4
Solution We are interested in the behavior as x S { q and as x S {2, where the

FIGURE 2.64 The lines y = 1 and denominator is zero. Notice that ƒ is an even function of x, so its graph is symmetric with
x = - 2 are asymptotes of the curve in respect to the y-axis.
Example 15.
(a) The behavior as x S { q. Since lim x Sq ƒ(x) = 0, the line y = 0 is a horizontal
asymptote of the graph to the right. By symmetry it is an asymptote to the left as well
(Figure 2.65). Notice that the curve approaches the x-axis from only the negative side
y
(or from below). Also, ƒ(0) = 2.
8 8
y=− (b) The behavior as x S {2. Since
7 x2 − 4
6 lim ƒ(x) = - q and lim ƒ(x) = q,
5 x S 2+ x S 2-
Vertical
Vertical 4 asymptote, x = 2 the line x = 2 is a vertical asymptote both from the right and from the left. By sym-
asymptote, 3 Horizontal metry, the line x = -2 is also a vertical asymptote.
x = −2 2 asymptote, y = 0
1 There are no other asymptotes because ƒ has a finite limit at all other points.
x
−4−3−2−1 0 1 2 3 4 EXAMPLE 17 The curves
1 sin x
y = sec x = cos x and y = tan x = cos x
both have vertical asymptotes at odd-integer multiples of p>2, where cos x = 0 (Figure 2.66).
FIGURE 2.65 Graph of the function
y y
in Example 16. Notice that the curve y = sec x y = tan x
approaches the x-axis from only one side.
Asymptotes do not have to be two-sided.
1 1
x x
3p −p p 0 p p 3p 3p −p p 0 p p 3p
− − − − −1
2 2 2 2 2 2 2 2
FIGURE 2.66 The graphs of sec x and tan x have infinitely many vertical asymptotes
(Example 17).
Dominant Terms
In Example 9 we saw that by long division we could rewrite the function
x2 - 3
ƒ(x) =
2x - 4
as a linear function plus a remainder term:
ƒ(x) = a + 1b + a b.
x 1
2 2x - 4
This tells us immediately that
For 0 x 0 large,
x 1
ƒ(x) ≈ + 1 is near 0.
2 2x - 4
1
ƒ(x) ≈ For x near 2, this term is very large in absolute value.
y 2x - 4
20 If we want to know how ƒ behaves, this is the way to find out. It behaves like
y = (x>2) + 1 when |x| is large and the contribution of 1>(2x - 4) to the total value of ƒ
15 is insignificant. It behaves like 1>(2x - 4) when x is so close to 2 that 1>(2x - 4) makes
the dominant contribution.
10 We say that (x>2) + 1 dominates when x is numerically large, and we say that
f(x) 1>(2x - 4) dominates when x is near 2. Dominant terms like these help us predict a
5
g(x) = 3x 4 function’s behavior.
x
−2 −1 0 1 2 EXAMPLE 18 Let ƒ(x) = 3x4 - 2x3 + 3x2 - 5x + 6 and g(x) = 3x4. Show that
although ƒ and g are quite different for numerically small values of x, they are virtually
identical for 0 x 0 very large, in the sense that their ratios approach 1 as x S q or x S - q .
−5
(a)
Solution The graphs of ƒ and g behave quite differently near the origin (Figure 2.67a),
y but appear as virtually identical on a larger scale (Figure 2.67b).
We can test that the term 3x4 in ƒ, represented graphically by g, dominates the poly-
500,000
nomial ƒ for numerically large values of x by examining the ratio of the two functions as
x S { q. We find that
300,000 ƒ(x) 3x4 - 2x3 + 3x2 - 5x + 6
lim = lim
x S {q g(x) x S {q 3x4
= lim a1 - + b
100,000 2 1 5 2
+ -
x S {q 3x x2 3x3 x4
x
−20 −10 0 10 20
= 1,
−100,000
which means that ƒ and g appear nearly identical when 0 x 0 is large.
(b)
FIGURE 2.67 The graphs of ƒ and Summary

g are (a) distinct for 0 x 0 small, and
(b) nearly identical for 0 x 0 large
In this chapter we presented several important calculus ideas that are made meaningful and
precise by the concept of the limit. These include the three ideas of the exact rate of change of
(Example 18). a function, the slope of the graph of a function at a point, and the continuity of a function. The
primary methods used for calculating limits of many functions are captured in the algebraic
Limit Laws of Theorem 1 and in the Sandwich Theorem, all of which are proved from the pre-
cise definition of the limit. We saw that these computational rules also apply to one-sided limits
and to limits at infinity. However, to calculate complicated limits such as
1 x
lim a1 + x b ,
x - sin x
lim , and lim x1>x ,
xS0 xS0 x3 xS0
techniques other than simple algebra are required. The derivative is one of the tools we
need to calculate limits such as these, and this notion is the central subject of our next
chapter.
Exercises 2.6
Finding Limits 2 - t + sin t r + sin r
11. lim 12. lim
1. For the function ƒ whose graph is given, determine the following t S -q t + cos t r Sq 2r + 7 - 5 sin r
limits.
Limits of Rational Functions
a. lim ƒ(x) b. lim ƒ(x) c. lim ƒ(x) In Exercises 13–22, find the limit of each rational function (a) as
xS2 x S -3 + x S -3 -
d. lim ƒ(x) e. lim+ ƒ(x) f. lim- ƒ(x) x S q and (b) as x S - q.
x S -3 xS0 xS0
2x + 3 2x3 + 7
g. lim ƒ(x) h. lim ƒ(x) i. lim ƒ(x) 13. ƒ(x) = 14. ƒ(x) =
xS0 x Sq x S -q 5x + 7 x - x2 + x + 7
3
y x + 1 3x + 7
15. ƒ(x) = 2 16. ƒ(x) = 2
x + 3 x - 2
3 7x3 9x4 + x
17. h(x) = 18. h(x) =
2 x3 - 3x2 + 6x 2x4 + 5x2 - x + 6
f
1 10x5 + x4 + 31 x3 + 7x2 - 2
19. g(x) = 20. g(x) =
x x6 x2 - x + 1
−6 −5 −4 −3 −2 −1 1 2 3 4 5 6
−1 3x7 + 5x2 - 1 5x8 - 2x3 + 9
21. f(x) = 22. h(x) =
−2 6x3 - 7x + 3 3 + x - 4x5
−3
Limits as x u H or x u − H
The process by which we determine limits of rational functions applies
2. For the function ƒ whose graph is given, determine the following equally well to ratios containing noninteger or negative powers of x:
limits. Divide numerator and denominator by the highest power of x in the
a. lim ƒ(x) b. lim+ ƒ(x) c. lim- ƒ(x) denominator and proceed from there. Find the limits in Exercises 23–36.
xS4 xS2 xS2
x Sq A 2x 2 + x
d. lim ƒ(x) e. lim + ƒ(x) f. lim ƒ(x) 8x2 - 3 x2 + x - 1 1>3
xS2 x S -3 x S -3 - 23. lim 24. lim ¢ ≤
x S -q 8x2 - 3
g. lim ƒ(x) h. lim+ ƒ(x) i. lim- ƒ(x)
x S -3 xS0 xS0
A x3 + x - 2
j. lim ƒ(x) k. lim ƒ(x) l. lim ƒ(x) 1 - x3 5 x2 - 5x
xS0 x Sq x S -q 25. lim ¢ ≤ 26. lim
xS - q x2 + 7x xS q
2 2x + x-1 2 + 2x
y
2 - 2x
27. lim 28. lim
xS q 3x - 7 xS q
2x - 2x
3
f 3 5
x-1 + x-4
2x + 2x
2
29. lim 3 5
30. lim
1 xS - q xS q x-2 - x-3
x 2x - 5x + 3
3
2x5>3 - x1>3 + 7
x S q x 8>5 + 3x + 2x
−6 −5 −4 −3 −2 −1 1 2 3 4 5 6 31. lim 32. lim
−1 xS - q 2x + x2>3 - 4
2x2 + 1 2x2 + 1
−2
−3 33. lim 34. lim
xS q x + 1 xS - q x + 1
x - 3 4 - 3x3
24x2 + 25 2x6 + 9
35. lim 36. lim
xS q xS - q
In Exercises 3–8, find the limit of each function (a) as x S q and
(b) as x S - q. (You may wish to visualize your answer with a Infinite Limits
graphing calculator or computer.) Find the limits in Exercises 37–48.
2 2 1 5
3. ƒ(x) = x - 3 4. ƒ(x) = p - 37. lim+ 38. lim-
x S 0 3x x S 0 2x
x2
3 1
1 1 39. lim- 40. lim+
5. g(x) = 6. g(x) = xS2 x - 2 xS3 x - 3
2 + (1>x) 8 - (5>x2) 2x 3x
41. lim + 42. lim -
-5 + (7>x) 3 - (2>x) x S -8 x + 8 x S -5 2x + 10
4 + ( 22>x2)
7. h(x) = 8. h(x) =
3 - (1>x2) 4 -1
43. lim 44. lim 2
x S 7 (x - 7)2 x S 0 x (x + 1)
2 2
sin 2x cos u 45. a. lim+ b. lim-
9. lim x 10. lim xS0 3x1>3 xS0 3x1>3
x Sq u S -q 3u
2 2 Graphing Simple Rational Functions

46. a. lim+ b. lim-
xS0 x1>5 xS0 x1>5 Graph the rational functions in Exercises 63–68. Include the graphs
4 1 and equations of the asymptotes and dominant terms.
47. lim 48. lim
xS0 x2>5 xS0 x2>3 1 1
63. y = 64. y =
x - 1 x + 1
Find the limits in Exercises 49–52. 1 -3
49. lim tan x 50. lim sec x 65. y = 66. y =
2x + 4 x - 3
x S (p>2)- x S (-p>2) +
x + 3 2x
51. lim- (1 + csc u) 52. lim (2 - cot u) 67. y = 68. y =
uS0 uS0 x + 2 x + 1
Inventing Graphs and Functions
1
53. lim 2 as In Exercises 69–72, sketch the graph of a function y = ƒ(x) that satis-
x - 4 fies the given conditions. No formulas are required—just label the
a. x S 2+ b. x S 2- coordinate axes and sketch an appropriate graph. (The answers are not
c. x S - 2+ d. x S - 2- unique, so your graphs may not be exactly like those in the answer
x section.)
54. lim 2 as
x - 1 69. ƒ(0) = 0, ƒ(1) = 2, ƒ(-1) = - 2, lim ƒ(x) = -1, and
x S -q
a. x S 1+ b. x S 1-
lim ƒ(x) = 1
c. x S - 1+ d. x S - 1- x Sq
70. ƒ(0) = 0, lim ƒ(x) = 0, lim+ ƒ(x) = 2, and
55. lim a - x b as
x2 1 x S {q xS0
2 lim ƒ(x) = -2
x S 0-
a. x S 0+ b. x S 0- 71. ƒ(0) = 0, lim ƒ(x) = 0, lim- ƒ(x) = lim + ƒ(x) = q,
c. x S 2 3
x S {q xS1 x S -1
2 d. x S - 1 lim ƒ(x) = - q, and lim ƒ(x) = - q
xS1 + x S -1-
2
x - 1 72. ƒ(2) = 1, ƒ(-1) = 0, lim ƒ(x) = 0, lim+ ƒ(x) = q,
56. lim as x Sq xS0
2x + 4
lim ƒ(x) = - q, and lim ƒ(x) = 1
a. x S -2+ b. x S -2- x S 0- x S -q
c. x S 1+ d. x S 0- In Exercises 73–76, find a function that satisfies the given conditions

and sketch its graph. (The answers here are not unique. Any function
x2 - 3x + 2 that satisfies the conditions is acceptable. Feel free to use formulas
57. lim as
x3 - 2x2 defined in pieces if that will help.)
a. x S 0+ b. x S 2+ 73. lim ƒ(x) = 0, lim ƒ(x) = q, and lim ƒ(x) = q
x S {q x S 2- x S 2+
c. x S 2- d. x S 2
74. lim g(x) = 0, lim- g(x) = - q, and lim+ g(x) = q
e. What, if anything, can be said about the limit as x S 0? x S {q xS3 xS3
2 75. lim h(x) = - 1, lim h(x) = 1, lim- h(x) = -1, and

x - 3x + 2 xS - q xS q xS0
58. lim as lim+ h(x) = 1
x3 - 4x xS0
a. x S 2+ b. x S -2+ 76. lim k(x) = 1, lim- k(x) = q, and lim+ k(x) = - q
c. x S 0- d. x S 1+ x S {q xS1 xS1
77. Suppose that ƒ(x) and g(x) are polynomials in x and that
e. What, if anything, can be said about the limit as x S 0?
limx S q (ƒ(x)>g(x)) = 2. Can you conclude anything about
Find the limits in Exercises 59–62. limx S - q (ƒ(x)>g(x))? Give reasons for your answer.
b as
3 78. Suppose that ƒ(x) and g(x) are polynomials in x. Can the graph of
59. lima2 -
t 1>3 ƒ(x)>g(x) have an asymptote if g(x) is never zero? Give reasons
a. t S 0+ b. t S 0- for your answer.
79. How many horizontal asymptotes can the graph of a given ratio-
1
60. lima + 7b as nal function have? Give reasons for your answer.
t 3>5
a. t S 0+ b. t S 0- Finding Limits of Differences When x u t H
b as
80. lim ( 2x + 9 - 2x + 4 )
1 2
61. lima 2>3 +
x (x - 1)2>3
81. lim ( 2x2 + 25 - 2x2 - 1 )
xS q
a. x S 0+ b. x S 0-
( 2x2 + 3 + x )
xS q
c. x S 1+ d. x S 1-
82. lim
24x2 + 3x - 2 )
xS - q
b as
1 1
62. lima - 83. lim ( 2x +
84. lim ( 29x2 - x - 3x )
x1>3 (x - 1)4>3 xS - q
a. x S 0+ b. x S 0- xS q
c. x S 1+ d. x S 1-
Chapter 2 Questions to Guide Your Review 99
85. lim ( 2x2 + 3x - 2x2 - 2x ) 96. lim-

1
= -q 97. lim+
1
= q
86. lim ( 2x2 + x - 2x2 - x )
xS q xS2 x - 2 xS2 x - 2
1
xS q 98. lim- = q
x S 1 1 - x2
Using the Formal Definitions
Use the formal definitions of limits as x S { q to establish the limits
in Exercises 87 and 88. Oblique Asymptotes
Graph the rational functions in Exercises 99–104. Include the graphs
87. If ƒ has the constant value ƒ(x) = k, then lim ƒ(x) = k.
x Sq and equations of the asymptotes.
88. If ƒ has the constant value ƒ(x) = k, then lim ƒ(x) = k.
x S -q x2 x2 + 1
99. y = 100. y =
Use formal definitions to prove the limit statements in Exercises 89–92. x - 1 x - 1
x2 - 4 x2 - 1
-1 1 101. y = 102. y =
0x0
89. lim = -q 90. lim = q x - 1 2x + 4
xS0 x2 xS0
x2 - 1 x3 + 1
-2 1 103. y = x 104. y =
91. lim = -q 92. lim = q x2
x S 3 (x - 3)2 x S -5 (x + 5)2
93. Here is the definition of infinite right-hand limit. Additional Graphing Exercises
T Graph the curves in Exercises 105–108. Explain the relationship
between the curve’s formula and what you see.
x
24 - x2
We say that ƒ(x) approaches infinity as x approaches c from the 105. y =
right, and write
24 - x2
lim ƒ(x) = q, -1
106. y =
x S c+
if, for every positive real number B, there exists a correspond- 1

107. y = x2>3 + 1>3
ing number d 7 0 such that for all x x
108. y = sin a 2 b
c 6 x 6 c + d 1 ƒ(x) 7 B. p
x + 1
Modify the definition to cover the following cases. T Graph the functions in Exercises 109 and 110. Then answer the follow-
ing questions.
a. lim ƒ(x) = q
x S c-
a. How does the graph behave as x S 0+?
b. lim+ ƒ(x) = - q
xSc b. How does the graph behave as x S { q?
c. lim- ƒ(x) = - q c. How does the graph behave near x = 1 and x = -1?
xSc
Use the formal definitions from Exercise 93 to prove the limit state- Give reasons for your answers.
ments in Exercises 94–98.
ax - x b a b
3 1 2>3 3 x 2>3
1 1 109. y = 110. y =
94. lim+ x = q 95. lim- x = - q 2 2 x - 1
xS0 xS0
Chapter 2 Questions to Guide Your Review

1. What is the average rate of change of the function g(t) over the 5. What function behaviors might occur for which the limit may fail
interval from t = a to t = b? How is it related to a secant line? to exist? Give examples.
2. What limit must be calculated to find the rate of change of a func- 6. What theorems are available for calculating limits? Give exam-
tion g(t) at t = t0? ples of how the theorems are used.
3. Give an informal or intuitive definition of the limit 7. How are one-sided limits related to limits? How can this relation-
ship sometimes be used to calculate a limit or prove it does not
lim ƒ(x) = L.
xSc exist? Give examples.
Why is the definition “informal”? Give examples. 8. What is the value of limu S 0 ((sin u)>u)? Does it matter whether u
4. Does the existence and value of the limit of a function ƒ(x) as x is measured in degrees or radians? Explain.
approaches c ever depend on what happens at x = c? Explain
and give examples.
9. What exactly does limx S c ƒ(x) = L mean? Give an example in 15. What are the basic types of discontinuity? Give an example of
which you find a d 7 0 for a given ƒ, L, c, and P 7 0 in the pre- each. What is a removable discontinuity? Give an example.
cise definition of limit. 16. What does it mean for a function to have the Intermediate Value
10. Give precise definitions of the following statements. Property? What conditions guarantee that a function has this
a. limx S 2- ƒ(x) = 5 b. limx S 2+ ƒ(x) = 5 property over an interval? What are the consequences for graph-
ing and solving the equation ƒ(x) = 0?
c. limx S 2 ƒ(x) = q d. limx S 2 ƒ(x) = - q
17. Under what circumstances can you extend a function ƒ(x) to be
11. What conditions must be satisfied by a function if it is to be con-
continuous at a point x = c? Give an example.
tinuous at an interior point of its domain? At an endpoint?
18. What exactly do limx Sq ƒ(x) = L and limx S -q ƒ(x) = L mean?
12. How can looking at the graph of a function help you tell where Give examples.
the function is continuous?
19. What are limx S {q k (k a constant) and limx S {q (1>x)? How do
13. What does it mean for a function to be right-continuous at a you extend these results to other functions? Give examples.
point? Left-continuous? How are continuity and one-sided conti-
nuity related? 20. How do you find the limit of a rational function as x S { q?
Give examples.
14. What does it mean for a function to be continuous on an interval?
Give examples to illustrate the fact that a function that is not con- 21. What are horizontal and vertical asymptotes? Give examples.
tinuous on its entire domain may still be continuous on selected
intervals within the domain.
Chapter 2 Practice Exercises
limx S 0 ƒ(x) = 1>2 and limx S 0 g(x) = 22. Find the limits as
Limits and Continuity 4. Suppose the functions ƒ(x) and g(x) are defined for all x and that
1. Graph the function
x S 0 of the following functions.
1, x … -1
-x, -1 6 x 6 0 a. - g(x) b. g(x) # ƒ(x)
ƒ(x) = e 1, x = 0 c. ƒ(x) + g(x) d. 1>ƒ(x)
-x, 0 6 x 6 1 ƒ(x) # cos x
e. x + ƒ(x) f.
1, x Ú 1. x - 1
Then discuss, in detail, limits, one-sided limits, continuity, and In Exercises 5 and 6, find the value that limx S 0 g(x) must have if the
one-sided continuity of ƒ at x = -1, 0, and 1. Are any of the given limit statements hold.
4 - g(x)
discontinuities removable? Explain. 5. lim a x b = 1 6. lim ax lim g(x)b = 2
x 0
S x S -4 xS0
2. Repeat the instructions of Exercise 1 for
7. On what intervals are the following functions continuous?
0, x … -1 a. ƒ(x) = x1>3 b. g(x) = x3>4
1>x, 0 6 0x0 6 1
ƒ(x) = d c. h(x) = x -2>3
d. k(x) = x-1>6
0, x = 1
8. On what intervals are the following functions continuous?
1, x 7 1.
a. ƒ(x) = tan x b. g(x) = csc x
3. Suppose that ƒ(t) and ƒ(t) are defined for all t and that limt S t0
ƒ(t) = - 7 and limt S t0 g(t) = 0. Find the limit as t S t0 of the cos x sin x
c. h(x) = x - p d. k(x) = x
following functions.
a. 3ƒ(t) b. (ƒ(t))2
Finding Limits
ƒ(t) In Exercises 9–28, find the limit or explain why it does not exist.
c. ƒ(t) # g(t) d.
g(t) - 7 x2 - 4x + 4
f. 0 ƒ(t) 0
9. lim
e. cos (g(t)) x3 + 5x2 - 14x
a. as x S 0 b. as x S 2
g. ƒ(t) + g(t) h. 1>ƒ(t)
Chapter 2 Practice Exercises 101
x2 + x Continuous Extension
10. lim
x + 2x4 + x3
5 31. Can ƒ(x) = x(x2 - 1)> 0 x2 - 1 0 be extended to be continuous at
x = 1 or -1? Give reasons for your answers. (Graph the func-
a. as x S 0 b. as x S - 1
1 - 2x
tion—you will find the graph interesting.)
x 2 - a2
11. lim 12. lim 32. Explain why the function ƒ(x) = sin (1>x) has no continuous
xS1 1 - x x S a x 4 - a4
extension to x = 0.
(x + h)2 - x2 (x + h)2 - x2
13. lim 14. lim T In Exercises 33–36, graph the function to see whether it appears to have
hS0 h xS0 h
a continuous extension to the given point a. If it does, use Trace and
1 1 Zoom to find a good candidate for the extended function’s value at a. If
- (2 + x)3 - 8
2 + x 2 the function does not appear to have a continuous extension, can it be
15. lim x 16. lim x
xS0 xS0
extended to be continuous from the right or left? If so, what do you
x1>3 - 1 x2>3 - 16 think the extended function’s value should be?
2x - 1 2x - 8
17. lim 18. lim
x - 1
x - 2
xS1 x S 64
33. ƒ(x) = 4
, a = 1
tan (2x) x
19. lim 20. lim- csc x 5 cos u
x S 0 tan (px) xSp
34. g(u) = , a = p>2
4u - 2p
21. lim sin a 35. h(t) = ( 1 + 0 t 0 ) 1>t, a = 0
x
+ sin xb 22. lim cos2 (x - tan x)
xSp 2 xSp
x
36. k(x) = , a = 0
1 - 20 x 0
8x cos 2x - 1
23. lim 24. lim
xS0 3 sin x - x xS0 sin x
In Exercises 25–28, find the limit of g(x) as x approaches the indi- Limits at Infinity
cated value. Find the limits in Exercises 37–46.
25. lim+ (4g(x))1>3 = 2 2x + 3 2x2 + 3
xS0
37. lim 38. lim
x S q 5x + 7 xS - q 5x2 + 7
1 x2 - 4x + 8 1
x S 25
26. lim = 2 39. lim 40. lim 2
x + g(x) xS - q 3x3 x S q x - 7x + 1
3x2 + 1 q x2 - 7x x4 + x3
27. lim = 41. lim 42. lim
xS1 g(x) x -
S q x + 1 x S q 12x3 + 128
5 - x2 sin x (If you have a grapher, try graphing the function
x S -2 2g(x)
xS q : x ;
28. lim = 0 43. lim
for - 5 … x … 5.)
(If you have a grapher, try graphing
cos u - 1
T Roots 44. lim ƒ(x) = x (cos (1>x) - 1) near the origin to
uS q u
x + sin x + 2 2x
29. Let ƒ(x) = x3 - x - 1. “see” the limit at infinity.)
a. Use the Intermediate Value Theorem to show that ƒ has a 45. lim
zero between - 1 and 2. xS q x + sin x
b. Solve the equation ƒ(x) = 0 graphically with an error of x2>3 + x-1
46. lim 2>3
magnitude at most 10-8. xS q x + cos2 x
c. It can be shown that the exact value of the solution in part (b) is
269 1>3 269 1>3
Horizontal and Vertical Asymptotes
a + b + a - b .
1 1 47. Use limits to determine the equations for all vertical asymptotes.
2 18 2 18 x2 + 4
a. y =
Evaluate this exact answer and compare it with the value you x - 3
found in part (b). x2 - x - 2
b. ƒ(x) = 2
T 30. Let ƒ(u) = u 3 - 2u + 2. x - 2x + 1
a. Use the Intermediate Value Theorem to show that ƒ has a x2 + x - 6
c. y = 2
zero between - 2 and 0. x + 2x - 8
48. Use limits to determine the equations for all horizontal asymptotes.
2x + 4
b. Solve the equation ƒ(u) = 0 graphically with an error of
1 - x2
2x + 4
magnitude at most 10-4.
a. y = b. ƒ(x) =
c. It can be shown that the exact value of the solution in part (b) is x2 + 1
2x2 + 4
A 27 A 27 B 9x2 + 1
x2 + 9
a - a
1>3 1>3
19 19 c. g(x) = d. y =
- 1b + 1b . x
Evaluate this exact answer and compare it with the value you
found in part (b).
Chapter 2 Additional and Advanced Exercises
T 1. Assigning a value to 00 The rules of exponents tell us that 4. Controlling the flow from a draining tank Torricelli’s law
a0 = 1 if a is any number different from zero. They also tell us says that if you drain a tank like the one in the figure shown, the
that 0n = 0 if n is any positive number. rate y at which water runs out is a constant times the square root
If we tried to extend these rules to include the case 00, we of the water’s depth x. The constant depends on the size and
would get conflicting results. The first rule would say 00 = 1, shape of the exit valve.
whereas the second would say 00 = 0.
We are not dealing with a question of right or wrong here.
Neither rule applies as it stands, so there is no contradiction. We
could, in fact, define 00 to have any value we wanted as long as
we could persuade others to agree.
What value would you like 00 to have? Here is an example
that might help you to decide. (See Exercise 2 below for another x
Exit rate y ft3 min
example.)
a. Calculate xx for x = 0.1, 0.01, 0.001, and so on as far as
your calculator can go. Record the values you get. What
Suppose that y = 2x>2 for a certain tank. You are trying to
pattern do you see?
b. Graph the function y = xx for 0 6 x … 1. Even though the maintain a fairly constant exit rate by adding water to the tank
function is not defined for x … 0, the graph will approach with a hose from time to time. How deep must you keep the water
the y-axis from the right. Toward what y-value does it seem if you want to maintain the exit rate
a. within 0.2 ft3 >min of the rate y0 = 1 ft3 >min?
to be headed? Zoom in to further support your idea.
b. within 0.1 ft3 >min of the rate y0 = 1 ft3 >min?

T 2. A reason you might want 00 to be something other than 0 or 1
As the number x increases through positive values, the numbers
1>x and 1 > (ln x) both approach zero. What happens to the number 5. Thermal expansion in precise equipment As you may know,
most metals expand when heated and contract when cooled. The
1 1>(ln x)
ƒ(x) = a x b
dimensions of a piece of laboratory equipment are sometimes so
critical that the shop where the equipment is made must be held
as x increases? Here are two ways to find out. at the same temperature as the laboratory where the equipment is
to be used. A typical aluminum bar that is 10 cm wide at 70°F
a. Evaluate ƒ for x = 10, 100, 1000, and so on as far as your will be
calculator can reasonably go. What pattern do you see?
b. Graph ƒ in a variety of graphing windows, including win- y = 10 + (t - 70) * 10-4
dows that contain the origin. What do you see? Trace the centimeters wide at a nearby temperature t. Suppose that you are
y-values along the graph. What do you find? using a bar like this in a gravity wave detector, where its width
3. Lorentz contraction In relativity theory, the length of an must stay within 0.0005 cm of the ideal 10 cm. How close to
object, say a rocket, appears to an observer to depend on the t0 = 70°F must you maintain the temperature to ensure that this
speed at which the object is traveling with respect to the observer. tolerance is not exceeded?
If the observer measures the rocket’s length as L 0 at rest, then at 6. Stripes on a measuring cup The interior of a typical 1-L mea-
speed y the length will appear to be suring cup is a right circular cylinder of radius 6 cm (see accom-
panying figure). The volume of water we put in the cup is there-
B
y2 fore a function of the level h to which the cup is filled, the
L = L0 1 - .
c2 formula being
This equation is the Lorentz contraction formula. Here, c is the V = p62h = 36ph.
speed of light in a vacuum, about 3 * 108 m>sec. What happens
to L as y increases? Find limy S c- L. Why was the left-hand limit How closely must we measure h to measure out 1 L of water
needed? (1000 cm3) with an error of no more than 1% (10 cm3)?
Chapter 2 Additional and Advanced Exercises 103
17. A function continuous at only one point Let
ƒ(x) = e
x, if x is rational
0, if x is irrational.
Stripes
a. Show that ƒ is continuous at x = 0.
about b. Use the fact that every nonempty open interval of real num-
1 mm bers contains both rational and irrational numbers to show
wide that ƒ is not continuous at any nonzero value of x.
(a) 18. The Dirichlet ruler function If x is a rational number, then x
can be written in a unique way as a quotient of integers m>n
where n 7 0 and m and n have no common factors greater than
r = 6 cm
1. (We say that such a fraction is in lowest terms. For example,
6>4 written in lowest terms is 3>2.) Let ƒ(x) be defined for all x
in the interval 30, 14 by
ƒ(x) = e
1>n, if x = m>n is a rational number in lowest terms
Liquid volume 0, if x is irrational.
h V = 36ph
For instance, ƒ(0) = ƒ(1) = 1, ƒ(1>2) = 1>2, ƒ(1>3) = ƒ(2>3) =
1>3, ƒ(1>4) = ƒ(3>4) = 1>4, and so on.
a. Show that ƒ is discontinuous at every rational number in 30, 14 .
(b) b. Show that ƒ is continuous at every irrational number in 30, 14 .
(Hint: If P is a given positive number, show that there are only
finitely many rational numbers r in 30, 14 such that ƒ(r) Ú P.)
A 1-L measuring cup (a), modeled as a right circular cylinder (b)
c. Sketch the graph of ƒ. Why do you think ƒ is called the
of radius r = 6 cm
“ruler function”?
Precise Definition of Limit 19. Antipodal points Is there any reason to believe that there is
In Exercises 7–10, use the formal definition of limit to prove that the always a pair of antipodal (diametrically opposite) points on
function is continuous at c. Earth’s equator where the temperatures are the same? Explain.
7. ƒ(x) = x2 - 7, c = 1 8. g(x) = 1>(2x), c = 1>4 20. If limx S c (ƒ(x) + g(x)) = 3 and limx S c (ƒ(x) - g(x)) = -1, find
9. h(x) = 22x - 3, c = 2 10. F(x) = 29 - x, c = 5 limx S c ƒ(x)g(x).
11. Uniqueness of limits Show that a function cannot have two dif- 21. Roots of a quadratic equation that is almost linear The equa-
ferent limits at the same point. That is, if limx S c ƒ(x) = L 1 and tion ax2 + 2x - 1 = 0, where a is a constant, has two roots if
limx S c ƒ(x) = L 2, then L 1 = L 2. a 7 -1 and a ≠ 0, one positive and one negative:
12. Prove the limit Constant Multiple Rule: -1 + 21 + a - 1 - 21 + a
r+(a) = a , r-(a) = a ,
lim kƒ(x) = k lim ƒ(x) for any constant k.
xSc xSc
a. What happens to r+(a) as a S 0? As a S -1+?
13. One-sided limits If limx S 0+ ƒ(x) = A and limx S 0- ƒ(x) = B,
b. What happens to r-(a) as a S 0? As a S -1+?
find
c. Support your conclusions by graphing r+(a) and r-(a) as
a. limx S 0+ ƒ(x3 - x) b. limx S 0- ƒ(x3 - x)
functions of a. Describe what you see.
c. limx S 0+ ƒ(x2 - x4) d. limx S 0- ƒ(x2 - x4)
d. For added support, graph ƒ(x) = ax2 + 2x - 1 simultane-
14. Limits and continuity Which of the following statements are ously for a = 1, 0.5, 0.2, 0.1, and 0.05.
true, and which are false? If true, say why; if false, give a counter-
22. Root of an equation Show that the equation x + 2 cos x = 0
example (that is, an example confirming the falsehood).
has at least one solution.
a. If limx S c ƒ(x) exists but limx S c g(x) does not exist, then
23. Bounded functions A real-valued function ƒ is bounded from
limx S c(ƒ(x) + g(x)) does not exist.
above on a set D if there exists a number N such that ƒ(x) … N
b. If neither limx S c ƒ(x) nor limx S c g(x) exists, then for all x in D. We call N, when it exists, an upper bound for ƒ on
limx S c (ƒ(x) + g(x)) does not exist. D and say that ƒ is bounded from above by N. In a similar man-
c. If ƒ is continuous at x, then so is 0 ƒ 0 . ner, we say that ƒ is bounded from below on D if there exists a
d. If 0 ƒ 0 is continuous at c, then so is ƒ. number M such that ƒ(x) Ú M for all x in D. We call M, when it
exists, a lower bound for ƒ on D and say that ƒ is bounded from
In Exercises 15 and 16, use the formal definition of limit to prove that below by M. We say that ƒ is bounded on D if it is bounded from
the function has a continuous extension to the given value of x. both above and below.
x2 - 1 x2 - 2x - 3 a. Show that ƒ is bounded on D if and only if there exists a
number B such that 0 ƒ(x) 0 … B for all x in D.
15. ƒ(x) = , x = -1 16. g(x) = , x = 3
x + 1 2x - 6
b. Suppose that ƒ is bounded from above by N. Show that if sin (x2 - x - 2) sin (x2 - x - 2)
limx S c ƒ(x) = L, then L … N. c. lim = lim #
x S -1 x + 1 x S -1 (x 2 - x - 2)
c. Suppose that ƒ is bounded from below by M. Show that if
(x2 - x - 2) (x + 1)(x - 2)
limx S c ƒ(x) = L, then L Ú M. lim = 1 # lim = -3
x + 1 x + 1
24. Max 5a, b6 and min 5a, b6
x S -1 x S -1
sin 1 1 - 2x 2 sin 1 1 - 2x 2 1 - 2x
1 - 2x
a. Show that the expression d. lim = lim =
x - 1 x - 1
0a - b0
xS1 x 1
S
max 5a, b6 = 11 - 2x 21 1 + 2x 2
a + b
+
2 2 1 - x
(x - 1)1 1 + 2x 2 - 1)1 1 + 2x 2
1
1 # lim = lim = -
equals a if a Ú b and equals b if b Ú a. In other words, xS1 x S 1 (x 2
max 5a, b6 gives the larger of the two numbers a and b.
b. Find a similar expression for min 5a, b6, the smaller of a
sin (1 - cos x) sin x
x S 0 sin 2x
and b. 25. lim 26. lim+
xS0 x
sin U
Generalized Limits Involving sin (sin x) sin ( x2 + x )
U 27. lim 28. lim
xS0 x xS0 x
sin 1 2x - 3 2
The formula limu S 0 (sin u)>u = 1 can be generalized. If limx S c
ƒ(x) = 0 and ƒ(x) is never zero in an open interval containing the sin ( x2 - 4 )
point x = c, except possibly c itself, then 29. lim 30. lim
xS2 x - 2 xS9 x - 9
sin ƒ(x)
lim = 1.
xSc ƒ(x) Oblique Asymptotes
Here are several examples. Find all possible oblique asymptotes in Exercises 31–34.
2x3>2 + 2x - 3
2x + 1
sin x2 1
a. lim = 1 31. y = 32. y = x + x sin x
x S 0 x2
sin x2 sin x2 x2 33. y = 2x2 + 1 34. y = 2x2 + 2x

b. lim x = lim 2 lim x = 1 # 0 = 0
xS0 xS0 x xS0
Chapter 2 Technology Application Projects
Mathematica/Maple Modules:
Take It to the Limit
Part I
Part II (Zero Raised to the Power Zero: What Does It Mean?)
Part III (One-Sided Limits)
Visualize and interpret the limit concept through graphical and numerical explorations.
Part IV (What a Difference a Power Makes)
See how sensitive limits can be with various powers of x.
Going to Infinity
Part I (Exploring Function Behavior as x u H or x u − H )
This module provides four examples to explore the behavior of a function as x S q or x S - q .
Part II (Rates of Growth)
Observe graphs that appear to be continuous, yet the function is not continuous. Several issues of continuity are explored to obtain results that you
may find surprising.

Chapter 2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 2

Uploaded by

Copyright:

Available Formats

2

Limits and Continuity

OVERVIEW Mathematicians of the seventeenth century were keenly interested in the

2.1 Rates of Change and Tangents to Curves

Average and Instantaneous Speed

∆y 16(t0 + h)2 - 16t0 2

TABLE 2.1 Average speeds over short time intervals 3 t0, t0 + h4

Length of Average speed over Average speed over

The average speed on intervals starting at t0 = 1 seems to approach a limiting value

∆y 16(1 + h)2 - 16(1)2 16(1 + 2h + h2) - 16

Average Rates of Change and Secant Lines

Defining the Slope of a Curve

The tangent to the parabola at P is the line through P with slope 4:

Instantaneous Rates of Change and Tangent Lines

∆p 340 - 150 190

FIGURE 2.5 Growth of a fruit fly population in a controlled

310 - 150 100

100 approaching zero, say h = 0.1, 0.01, 0.001, 0.0001, 0.00001,

60 Q3 will have to do some algebra before you can substitute T = 2.

Elapsed time (sec) 30

c. Estimate the maximum rate of gasoline consumption and the

2.2 Limit of a Function and Limit Laws

if c is an irrational number, like p or 22, whose values can only be approximated by

Generalizing the idea illustrated in Example 1, suppose ƒ(x) is defined on an open

x lim ƒ(x) = lim x = c.

For instances of each of these rules we have

(a) Unit step function U(x) (b) g(x) (c) f(x)

The Limit Laws

THEOREM 1—Limit Laws If L, M, c, and k are real numbers and

1. Sum Rule: lim (ƒ(x) + g(x)) = L + M

2. Difference Rule: lim (ƒ(x) - g(x)) = L - M

3. Constant Multiple Rule: lim (k # ƒ(x)) = k # L

4. Product Rule: lim (ƒ(x) # g(x)) = L # M

lim 2ƒ(x) = 2L = L 1>n, n a positive integer

(If n is even, we assume that lim ƒ(x) = L 7 0.)

EXAMPLE 5 Use the observations limx S c k = k and limx S c x = c (Example 3) and

(c) lim 24x2 - 3

= c3 + 4c2 - 3 Power and Multiple Rules

lim x4 + lim x2 - lim 1

= 2 lim 4x2 - lim 3 Difference Rule

= 24(-2)2 - 3 Product and Multiple Rules

THEOREM 2—Limits of Polynomials

THEOREM 3—Limits of Rational Functions

EXAMPLE 6 The following calculation illustrates Theorems 2 and 3:

Identifying Common Factors

Solution We cannot substitute x = 1 because it makes the denominator zero. We test

Using Calculators and Computers to Estimate Limits

algebra rationalizes the numerator:

2x2 + 100 - 10 2x2 + 100 - 10 2x2 + 100 + 10

g The Sandwich Theorem

Then limx S c ƒ(x) = L.

1 follows that limx S c ƒ(x) = 0.

FIGURE 2.14 The Sandwich Theorem

a. lim g(x) b. lim g(x) c. lim g(x) d. lim g(x) y

s 7. Suppose that a function ƒ(x) is defined for all real values of x

t anything be said about the existence of limx S 0 ƒ(x)? Give reasons

4x - x2 54. Suppose limx S 4 ƒ(x) = 0 and limx S 4 g(x) = -3. Find

Using the Sandwich Theorem

66. a. Suppose that the inequalities 71. Let ƒ(x) = ( x2 - 1 ) > ( 0 x 0 - 1 ) .

c = 22 and using Zoom and Trace to estimate y-values on

the graph as x S 22.

Zoom and Trace to estimate y-values on the graph as ƒ(x)

c. Find limx S -6 G(x) algebraically.

a. Make a table of the values of h at x = 2.9, 2.99, 2.999, and ƒ(x)