Integrales

Chapter 2
Iterated, Line, and Surface

Integrals
2.1 Iterated Integrals

• Iterated Double Integrals • Double Integrals over General Regions • We must confess in all humil-
Changing the Order of Integration • Triple Integrals • Triple Integrals ity that, while number is a
product of our mind alone,
over General Regions
space has a reality beyond
the mind whose rules we can-
Iterated Double Integrals not completely prescribe.
Let f be a real-valued function of two variables x, y defined on a rectangular region -Carl Gauss
R = {(x, y) : a ≤ x ≤ b, c ≤ y ≤ d}
where a, b, c, d are real numbers. We partition [a, b] and [c, d] as follows.
Let m, n be positive integers, let a = x0 , x1 , . . . , xm = b be a partition of [a, b],
and let c = y0 , y1 , . . . , yn = d be a partition of [c, d]. For 1 ≤ i ≤ m and 1 ≤ j ≤ n,
consider the rectangular subregions
Ri,j = {(x, y) : xi−1 ≤ x ≤ xi , yj−1 ≤ y ≤ yj } .
The collection ∆ = {Ri,j : i = 1, . . . , m and j = 1, . . . , n} is called a partition of
R into sub-rectangles. Let ∆xi = xi − xi−1 , ∆yj = yj − yj−1 , and let �∆� be the
norm of ∆ which is defined as the maximum of all the ∆xi ’s and ∆yj ’s. We denote
the area of rectangle Ri,j by
∆Ai,j = ∆xi ∆yj .
Let L be a real number. We symbolically write
n �
� m
L = lim f (xi,j , yi,j ) ∆Ai,j
�∆�→0
i=1 j=1
if for every ε > 0 there exists δ > 0 such that

� �
� n m �
��
� f (xi,j , yi,j ) ∆Ai,j − L�� < ε
�
� i=1 j=1 �
for all (xi,j , yi,j ) in Ri,j whenever �∆� < δ. If such a limit L exists, we say f is
Riemann integrable or integrable on R, and we write
� � �n � m
f (x, y)dA = lim f (xi,j , yi,j ) ∆Ai,j .
R �∆�→0
i=1 j=1
59
60 CHAPTER 2. ITERATED, LINE, AND SURFACE INTEGRALS
We quote a theorem from advanced calculus about Riemann integrable functions.

We omit the proof of the theorem since it is beyond the scope of this book.1
Theorem 2.1 Riemann Integrable Functions
Let z = f (x, y) be a bounded real-valued function on a rectangular region
R = {(x, y) : a ≤ x ≤ b, c ≤ y ≤ d} .
If the points of discontinuities of f in R lie on a finite union of graphs of continuous

functions of one independent variable, then f is Riemann integrable on R.
In particular, every continuous real-valued function

�� on a rectangular region R
is integrable on R. However, in order to evaluate R
f (x, y)dA, we introduce the
concept of an iterated integral.
If we fix a value of y, then f (x, y) is a function of x only, and we have definite
�b
integral a f (x, y)dx. To illustrate, we assume y is constant in the integration below:
� 4 �x=4
1 2 �
x sin y dx = x sin y ��
0 2 x=2
� � � �
1 2 1 2
= (4) sin y − (2) sin y
2 2
= 6 sin y.
�d
Likewise, c f (x, y)dy is a definite� integral
� if x is held constant. The next theorem
implies that a double integral R
f (x, y)dA may be evaluated as an iterated
integral if f is continuous on R. Also, we omit the proof of the theorem that is
usually discussed in advanced calculus texts.
Theorem 2.2 Fubini’s Theorem and Iterated Integrals
If z = f (x, y) is the function in Theorem 2.1, then the identity below holds:
� � � b � d � d � b
f (x, y)dA = f (x, y)dydx = f (x, y)dxdy
R a c c a
In particular, if f is continuous on R, the above identity applies.
1A bounded real-valued function on a rectangular region R ⊆ R2 whose set of discontinuities

is of measure zero is Riemann integrable. Also, if y = f (x), a ≤ x ≤ b, is a continuous function,
then {(x, f (x)) : a ≤ x ≤ b} is a set of measure zero in R2 .
2.1. ITERATED INTEGRALS 61
Example 1 Evaluating an Iterated Integral

� 3� π
Evaluate the integral x2 cos(xy)dydx.
1 0
Solution For the inner integral, we have

�
1
cos(xy)dy = sin xy + C.
x
Consequently, we evaluate the iterated integral as follows:
� 3� π � 3 �y=π
�
2
x cos(xy)dydx = x sin(xy)�� dx
1 0 1 y=0
� 3
= x sin(πx)dx.
1
Then we integrate by parts using the substitutions:
u = x dv = sin(πx)dx
1
du = dx v = − cos(πx)
π
� �
Since udv = uv − vdu, we obtain
� 3 �x=3 �
x � 1 3
x sin(πx)dx = − cos(πx)�� + cos(πx)dx
1 π x=1 π 1
�x=3
2 1 � 2
= + 2 sin(πx)�� = .
π π x=1 π
� 3 � π
2
Thus, we find x2 cos(xy)dydx = .
1 0 π
✷
Try This 1
� 4 � 9 √
Evaluate xdydx.
0 0
Double Integrals over General Regions

Let φ1 , φ2 , ψ1 , ψ2 be continuous real-valued functions satisfying φ1 (x) ≤ φ2 (x) and
ψ1 (y) ≤ ψ2 (y) for all x in [a, b], and y in [c, d]. We evaluate double integrals of
a real-valued continuous function z = f (x, y) over elementary regions in the
xy-plane, see Figures 1-2. We classify these regions as either of type Rx or Ry , or
possibly of both types, depending on whether the boundary of the region are graphs
of functions of x or y, respectively:
A) Rx = {(x, y) : a ≤ x ≤ b, φ1 (x) ≤ y ≤ φ2 (x)}
B) Ry = {(x, y) : c ≤ y ≤ d, ψ1 (y) ≤ x ≤ ψ2 (y)}
y y
y�Φ2�x� x�Ψ1�y�
d
y�Φ1�x�
x�Ψ2�y�
c
x x
a b
Figure 1 Region of type Rx Figure 2 Region of type Ry

��
To evaluate Rx
f (x, y)dA, let R be a rectangular region that contains Rx . We
extend f to a function F on R:
�
f (x, y) if (x, y) ∈ Rx
F (x, y) =
0 if (x, y) ∈
/ Rx
Since f is bounded and continuous on Rx , the function F is bounded

�� and possibly
discontinuous only on points in the graph of φ1 or φ2 . Then R
F (x, y)dA exists
by Theorem 2.1. Suppose we choose R such that is consists of points (x, y) satisfying
a ≤ x ≤ b and p ≤ y ≤ q. Notice, for any x in [a, b], we find F (x, y) = 0 if y does
not belong to the interval [φ1 (x), φ2 (x)]. Thus,
� q � φ2 (x) � φ2 (x)
F (x, y)dy = F (x, y)dy = f (x, y)dy.
p φ1 (x) φ1 (x)
and the latter integral exists by the continuity of f on Rx . By Fubini’s Theorem,

we obtain
� � � b� q � b � φ2 (x)
F (x, y)dA = F (x, y)dydx = f (x, y)dydx.
Rx a p a φ1 (x)
We have a similar identity for double integrals over a type Ry region.
Theorem 2.3 Integrating over Regions of Type Rx or Ry
If z = f (x, y) is a continuous real-valued function on a region of type Rx or Ry ,

then
� � � b � φ2 (x)
a) f (x, y)dA = f (x, y)dydx
Rx a φ1 (x)
� � � d� ψ2 (y)
b) f (x, y)dA = f (x, y)dxdy
Ry c ψ1 (y)
� �
In Theorem 2.3, if f (x, y) = 1 for all x, y, then dA represents the area of
R
region R where R = Rx or R = Ry .
Also, we may apply double integrals to define the volume of a solid.
Definition 1 Volume of a Solid and Double Integrals
Let R be a region in the xy-plane of either type Rx or Ry . Let z = f (x, y) be

a nonnegative real-valued continuous function defined on R. Let S be the solid
consisting of the points (x, y, f (x, y)) in 3-space where (x, y) ∈ R. The volume of
S is defined as � �
Volume = f (x, y)dA.
R
Example 2 Evaluating the Volume of a Solid

1
Find the volume of the solid that lies below the plane f (x, y) = (6 − 2x − 2y) and
3
above the region
R = {(x, y) : 0 ≤ y ≤ 2, 0 ≤ x ≤ y/2}
in the xy-plane, see Figure 3.
Figure 3
A solid bounded above by a
Solution The region R is both type Rx and Ry , see Figure 4. We choose to
plane, and with a triangular
evaluate the integral for the volume as a type Ry region.
base in the xy-plane
� �
Volume = f (x, y)dA
R y
� 2� y/2
1
= (6 − 2y − 2x)dxdy
0 0 3 2
� �x=y/2
1 2 �
2� x�y�2
= (6 − 2y)x − x � dy
3 0 x=0
� 2 � �
1 y2 x
= 3y − y 2 − dy 1
3 0 4
Figure 4
8 The base of the above solid
Volume = units3 in the xy-plane.
9
✷
Try This 2
A solid S is bounded above by the plane z = y and bounded below by the region
R = {(x, y) : 0 ≤ x ≤ 1, 0 ≤ y ≤ 1 − x2 }.
Find the volume of S.
Figure 5
The solid for Try This 2.
Changing the Order of Integration

��
To be able to integrate R
f (x, y)dA, there may be an advantage to evaluating
one iterated integral over another. The reason is the order of integration, i.e.,
integrating with respect to x then integrating with respect to y, may be easier than
integrating first with respect to y, then with respect to x secondly.
y Example 3 Switching the Order of Integration
x�y�4 Sketch the region R of integration of

4 � 4� 1
sin(πx2 )dxdy.
0 y/4
Then switch the order of integration, and evaluate the resulting integral.
x Solution In Figure 6, we see the triangular region R. To switch the order of

1 integration to dydx, we partition the interval [0, 1] in the x-axis into smaller subin-
Figure 6 tervals. If the order of integration is dxdy, partition [0, 4] in the y-axis.
Region of integration Then draw a typical rectangle such that its base is a subinterval in the partition,
and its length extends from the lower boundary to the upper boundary of the region.
Since the order of integration is dydx, each point (x, y) in a typical rectangle satisfies
0 ≤ x ≤ 1 and 0 ≤ y ≤ 4x.
Then we integrate as follows:
� 4� 1 � 1 � 4x
sin(πx2 )dxdy = sin(πx2 )dydx
0 y/4 0 0
� 1 �y=4x
�
= y sin(πx )��
2
dx
0 y=0
� 1
= 4x sin(πx2 )dx
0
y �x=1
2 �
2 �
= − cos(πx )�
π x=0
2 � 4 � 1
4
sin(πx2 )dxdy =
0 y/4 π
✷
x
Figure 7
Try This 3
Region of integration
for Try This 3.
Sketch the region R of integration of
� 4� 2
x
� dydx.
0 x/2 y3 +1
Then switch the order of integration, and evaluate the resulting integral.
Triple Integrals
Let w = g(x, y, z) be a real-valued function that is defined on a Cartesian product
S = [a, b] × [c, d] × [e, f ]
= {(x, y, z) : a ≤ x ≤ b, c ≤ y ≤ d, e ≤ z ≤ f }
We partition each of [a, b], [c, d], and [e, f ] into finitely many subintervals. Let
[xi , xi+1 ], [yj , yj+1 ], and [zk , zk+1 ] be subintervals in the partitions, and we denote
their lengths by ∆xi , ∆yj , ∆zk , respectively. We consider a box
Sijk = [xi , xi+1 ] × [yj , yj+1 ] × [zk , zk+1 ]
whose volume is denoted by ∆Vijk = ∆xi ∆yj ∆zk . Let pijk be a point in Sijk , and
let �∆� be the maximum of the norms of the partitions of [a, b], [c, d], and [e, f ].
We define g to be Riemann integrable or integrable on the box S if there
is a number L ∈ R such that for each ε > 0 there exists a positive number δ > 0
satisfying � �
� �
��
� g pijk ∆Vijk − L�� < ε
�
� i j k �
for all pijk in Sijk whenever �∆� < δ. In the above sums, we evaluate over all the
values of i, j, k. In such a case, we write
� � � ��
g(x, y, z)dV = lim g pijk ∆Vijk = L.
S �∆�→0
i j k
We state an integrability condition for� g� similar

� to Theorem 2.1. Also, we state a
Fubini’s theorem for triple integrals S
g(x, y, z)dV . We omit the proofs since
they are usually discussed in advanced calculus texts.
Theorem 2.4 Riemann Integrable Functions
Let w = g(x, y, z) be a bounded real-valued function on a box
S = [a, b] × [c, d] × [e, f ].
If the points of discontinuities of g in S lie on a finite union of graphs of continuous

functions of two independent variables, then g is integrable on S.
Theorem 2.5 Fubini’s Theorem for Triple Integrals
If w = g(x, y, z) is the function in Theorem 2.4, then

� � � � b � d � f
g(x, y, z)dV = g(x, y, z)dzdydx
S a c e
Also, the six iterated triple integrals exist and are equal.
Example 4 Evaluating a Triple Integral

� 2� 3� 2
� �
Evaluate the integral x + z 2 − y 2 dxdydz.
1 0 0
Solution We integrate as follows:

� 2� 3� 2 � � � �
� � 2 3
x2 � � �x=2
x + z 2 − y 2 dxdydz = + x z 2 − y 2 �� dydz
1 0 0 1 0 2 x=0
� 2 � 3 � � ��
= 2 + 2 z 2 − y 2 dydz
1 0
� 2 � 3 � �
= 2 1 + z 2 − y 2 dydz
1 0
� � �
2 � � y 3 �y=3
= 2 y 1 + z2 − � dz
1 3 �y=0
� 2 � � � �
= 2 3 1 + z 2 − 9 dz
1
� 2 � �
= 6 z 2 − 2 dz
1
� 2� 3� 2 � �
x + z 2 − y 2 dxdydz = 2.
1 0 0
✷
Try This 4
� 2 � 1 � 1
Evaluate the triple integral yzexy dxdydz.
0 0 0
Triple Integrals over General Regions

��
We wish to extend and evaluate triple integrals S
g(x, y, z)dV over elementary
regions S ⊆ R3 that we are about to describe. We consider an elementary region
in the xy-plane:
Rx = {(x, y) : a ≤ x ≤ b, φ1 (x) ≤ y ≤ φ2 (x)}
where φ1 , φ2 are continuous functions of x such that φ1 (x) ≤ φ2 (x). Then we
associate an elementary region in 3-space such as
Syx = {(x, y, z) : (x, y) ∈ Rx , Φ1 (x, y) ≤ z ≤ Φ2 (x, y)}.
where Φ1 , Φ2 are continuous functions on Rx satisfying Φ1 (x, y) ≤ Φ2 (x, y). Using
similar ideas leading to Theorem 2.3, we have
� � � � b � φ2 (x) � Φ2 (x,y)
g(x, y, z)dV = g(x, y, z)dzdydx
Syx a φ1 (x) Φ1 (x,y)
provided g is continuous on Syx .

Similarly, if an elementary region in the xy-plane is given by

Ry = {(x, y) : c ≤ y ≤ d, ψ1 (y) ≤ x ≤ ψ2 (y)}
where ψ1 , ψ2 are continuous functions of y such that ψ1 (y) ≤ ψ2 (y), we associate
an elementary region
Sxy = {(x, y, z) : (x, y) ∈ Ry , Ψ1 (x, y) ≤ z ≤ Ψ2 (x, y)}.
where Ψ1 , Ψ2 are continuous functions on Ry satisfying Ψ1 (x, y) ≤ Ψ2 (x, y). Like-
wise, we obtain
� � � � d � ψ2 (y) � Ψ2 (x,y)
g(x, y, z)dV = g(x, y, z)dzdxdy
Sxy c ψ1 (y) Ψ1 (x,y)
whenever g is continuous on Sxy . There are six possible iterated triple integrals,
and another one of these is
� � � � f � γ2 (z) � Γ2 (y,z)
g(x, y, z)dV = g(x, y, z)dxdydz
S e γ1 (z) Γ1 (y,z)
where γ1 , γ2 are continuous functions of z satisfying γ1 (z) ≤ γ2 (z), and Γ1 , Γ2 are

continuous functions of (y, z) in some elementary region in the yz-plane such that
Γ1 (y, z) ≤ Γ2 (y, z).
We summarize and define an elementary region S ⊆ R3 in 3-space. For any
point in S, two of its coordinates lie in an elementary region in a plane such as
the xy-, yz-, or xz-plane, and the third coordinate lies between two continuous
functions of the first two variables. Specifically, let x1 , x2 , x3 be a permutation of
x, y, z, and let p < q be real constants. Let f1 , f2 , F1 , F2 be continuous functions
satisfying f1 (x3 ) ≤ f2 (x3 ) whenever p ≤ x3 ≤ q, and further F1 (x2 , x3 ) ≤ F2 (x2 , x3 )
if f1 (x3 ) ≤ x2 ≤ f2 (x3 ). A point lies in S if its coordinates satisfy p ≤ x3 ≤ q,
f1 (x3 ) ≤ x2 ≤ f2 (x3 ), and F1 (x2 , x3 ) ≤ x1 ≤ F2 (x2 , x3 ). Following the idea of
Theorem 2.3, we have the following theorem:
Theorem 2.6 Triple Integrals over Elementary Regions
If g is a real-valued continuous function on an elementary region S ⊆ R3 , then

� � � � q � f2 (x3 ) � F2 (x2 ,x3 )
g dV = g dx1 dx2 dx3
S p f1 (x3 ) F1 (x2 ,x3 )
� � �
In Theorem 2.6, if g(x, y, z) = 1 for all x, y, then dV is the volume of S.
S
Example 5 Triple Integral

Let S ⊆ R3 be a region in the first octant that is bounded above by the plane
� � �
2x + y + z = 2, and bounded below by the xy-plane. Then evaluate 6xdV .
S
Solution In Figure 8, the base of solid S in the xy-plane may be expressed as

Rx = {(x, y) : 0 ≤ x ≤ 1, 0 ≤ y ≤ 2 − 2x}.
Then region S is given by Figure 8
Solid of integration
Syx = {(x, y, z) : (x, y) ∈ Rx , 0 ≤ z ≤ 2 − 2x − y}. for Example 5.
We apply Theorem 2.6, and evaluate the inner-most integral.
� � � � 1 � 2−2x � 2−2x−y
6xdV = 6x dzdydx
S 0 0 0
� 1 � 2−2x
= 6x(2 − 2x − y)dydx
0 0
� 1 � 2−2x
= (12x(1 − x) − 6xy) dydx
0 0
Then we evaluate the inner-integral with respect to y:
� � � � 1 �y=2−2x
� �
6xdV = 12x(1 − x)y − 3xy 2 �� dx
S 0 y=0
� 1 � �
= 24x(1 − x)2 − 12x(1 − x)2 dx
0
� 1
= 12x(1 − x)2 dx
0
� � �
6xdV = 1.
S
Try This 5
A solid S ⊆ R3 lies in the first octant, bounded by the surfaces z = 1 − x2 and

��
x + y = 1, and the coordinate planes. Then evaluate S
10xdV .
Figure 9
The solid of integration 2.1 Check-It Out
for Try This 5.
Evaluate the iterated integral.
� � � 3 � 1 � 2 � x � y
1. 10(2x + y)3 dxdy 2. xy 2 dydx 3. (x − y)dzdydx
0 0 0 0 0
Rewrite the integral by switching the order of integration.

� 1 � x � 2 � y/2
4. f (x, y)dydx 5. f (x, y)dxdy
0 0 0 0
True or False. If false, revise the statement to make it true or explain.
� b � d � b � d
1. f (x, y)dydx = f (x, y)dxdy
a c a c
� 2 � 2 � 2 � 2
2. (x + y)dydx = xdx + ydy
0 0 0 0
� 1 � 1 � 1
1
3. xyz dzdydx =
0 0 0 8
� �
4. If R = {(x, y) : x + y ≤ r }, then 2 2 2
dA = πr2 .
R
�
5. The volume of the solid bounded by the hemisphere z = 4 − x2 − y 2 and
� 2� 2�
the plane z = 0 is given by 4 − x2 − y 2 dydx.
−2 −2
Exercises for Section 2.1
In Exercises 1-8, evaluate the iterated integrals.

� π/2 � 1 � 1/3 � 1
1. 6x sin(xy)dydx 2. πy sec2 (πxy)dxdy
π/6 0 1/4 0
� 2 � 1 � 3 � 2 � z � y
3. (4xy + 2z)dzdxdy 4. 12xdxdydz
1 0 1 0 0 0
� 1 � x � e � y
dxdy
5. 6ex+y dydx 6.
0 −x 1 0 x2 + y 2
� 2� z� y/2 � e� 4 � 2x
4 y
7. � dxdydz 8. √ dzdydx
1 0 0 y2 − x2 1 2 x z z2 − x2
In Exercises 9-14, we see a region R of integration that is bounded by the graph of the indicated equation.
Express the double integral of f (x, y) over R using dydx. Then switch the order of integration to dxdy.
√
9. y = 3x 10. y = x 11. y = x/2
y y y
3
2
1
x x x
1 4 2
For No. 9 For No. 10 For No. 11

√
12. y = x2 13. y= 1 − x2 14. (x − 2)2 + y 2 = 4
y y y
4 2
x
2
x
�1 1
x
2
For No. 12 For No. 13 For No. 14
In Exercises 15-26, switch the order of integration, and evaluate the integral with the new limits.
� � � � √ � �
1 2x 4 8 x 2 1
15. 2xydydx 16. dydx 17. 8xydxdy
0 0 0 x2 0 y/2
� � √ � � � �
4 x 4 2 2 x2
18. dydx 19. √
dydx 20. 2xdydx
0 0 0 x 0 0
� 2 � 4 � 1 � 1 � π/2 � 1
21. 2xdydx 22. dydx 23. cos(x)dydx
0 x2 −1 |x| 0 sin x
� 4 � √
8 x √ � 1 � 2 � 1 � 1 � �
7 y y2 πx2
24. dydx 25. 4e dydx 26. cos dxdy
0 x2 192 0 2x 0 y 4
In Exercises 27-34, find the volume of the solid that is bounded by the coordinate planes, and the graph of
the equations in 3-space.
27. z = 1 − y 2 , x = 2, y = 1 28. x + y + z = 3, x = 2, y = 1
For No. 27 For No. 28
29. z = 3, y = 4 − x2 30. x+y+z =1

√
31. z = x, x = 4, y = 3 32. 2x + y + z = 2
33. x2 + z 2 = 1, y 2 + z 2 = 1 34. z = 1 − y 2 , z = 1 − x2
Miscellaneous Problems
35. Find the volume of the solid that is bounded by the plane x + 4y + 16z = 8,
and the coordinate planes.
36. Find the volume of the solid bounded by the plane 2x + y + z = 2,

37. Find the volume of the solid in the first octant bounded by the plane x + 4y + z = 4,
38. Find the volume of the solid in the first octant bounded by the plane 2x + y + z = 4,
39. Find the volume of a solid in the first octant that is bounded by the surfaces
z = 1 − y 2 and x = 2.
40. Find the volume of a solid in the first octant that is bounded by the graphs
of z = 1 − y 2 , y = 2x, x = 0, and z = 0.
� � � �
x 1 π
41. Evaluate √ dA where R = (x, y) : 0 ≤ x ≤ , arcsin x ≤ y ≤ .
R 1 − x2 2 6
� �
sin x
42. Evaluate dA where R = {(x, y) : 0 ≤ y ≤ 1, y ≤ x ≤ 1}.
R x
� � �
43. Evaluate (xy − z)dV where S is the solid in 3-space that is bounded
S
by the graphs of x = 1, y = 4, z = 1, and the coordinate planes.
� � �
√
44. Evaluate xyzdV where S is the solid in 3-space that is bounded
S
by the graphs of x = 1, y = 1, z = 1, and the coordinate planes.
45. Find the volume of a solid that is bounded by the circular cylinder x2 + y 2 = 1,
the plane z + 2y = 2, and the xy-plane.
� � �
46. Evaluate 16dV where S is the solid in the first octant bounded by the paraboloid z = x2 + y 2 ,
S
the plane z = 1, and the coordinate planes.
√
47. A solid S consists of the points (x, y, z) satisfying 0 ≤ x, y, z ≤ 1 and 2 xy ≤ z, see Figure for No. 47.
Then find the volume of S. In terms of probability, if x, y, z are independent uniform random
variables on [0, 1], then the volume of S is the probability that the solutions t of the quadratic
equation xt2 + zt + y = 0 are real numbers..
Figure for No. 47

2.2. CHANGE OF VARIABLES IN INTEGRATION 73
2.2 Change of Variables in Integration

• Jacobian Determinant • Cylindrical Coordinates • Spherical Coordi-
nates
Jacobian Determinant v
The substitution method is an important technique in the integration of a function
of one variable, namely,
� x(b) � b
d
f (x)dx = f (x(u))x� (u)du. D
x(a) a
c
We extend the above method to the integration of multivariable functions.
u
Let D = {(u, v) ∈ R2 : a ≤ u ≤ b, c ≤ u ≤ d} be a rectangular region in the a b
uv-plane, see Figure 1. Let x = x(u, v) and y = y(u, v) be real-valued functions Figure 1
on D with continuous first partial derivatives. Let T be a transformation from D Region S in the uv-plane
into the xy-plane defined by T (x, y) = (x(u, v), y(u, v)). Further, suppose T is a
one-to-one function on D, i.e., T maps distinct points in D to distinct points in the
xy-plane.
y
Let C be the image of D under the transformation T , i.e,
C = T (D) = {(x, y) ∈ R2 : x = x(u, v), y = y(u, v), (u, v) ∈ D}. T�a,d�

If the sides of rectangle D are small, we approximate the area of C. The idea2 C
is to use the linear approximation of T near (a, c) as described in (36), page 39.
Moreover, we use a result from linear algebra claiming that a one-to-one linear
T�b,c�
transformation maps a rectangle to a parallelogram. T�a,c�
x
In Figure 2, let v be the vector from point T (a, c) to T (b, c), and let w be the Figure 2
vector from T (a, c) to T (a, d). We may approximate v and w by the tangent vectors Region T (R) in the xy-plane
along the boundary of C, respectively, i.e.,
� �
� �
∂T � ∂T �
v ≈ ∆u � and w ≈ ∆v �
∂u � ∂v �
(a,c) (a,c)
if ∆u = b − a and ∆v = d − c are small. The area of D is ∆u∆v, and the area of C

is approximately the area of the parallelogram defined by v and w. Recall, the area
of a parallelogram is the magnitude of the cross product of the vectors defining the
sides of the parallelogram. Then
Area of C ≈ �v × w�
� � � �
� �
� ∂T ��
∂T � �
≈ � × � (Area of D)
� ∂u ��
∂v � �
� (a,c) (a,c)
�
�   �
� i j k �
� �
�   �
�  ∂x ∂y  �
�  0  �
= � Det  ∂u ∂u  � (Area of D)
�   �
�   �
� ∂x ∂y �
� 0 �
∂v ∂v
2 The linear approximation of T near x = (a, c) is defined by T � (x) = A(x − x ) + T (x )

0 0 0
where A is a 2 by 2 matrix, and x, x0 ∈ R2 are realized as column vectors.
where the partial derivatives are evaluated at (a, c). Consequently, we find
� �
� ∂(x, y) �
Area of C ≈ �� (Area of D)
∂(u, v) �
where  ∂x ∂y 
∂(x, y)  ∂u ∂u 
= Det 

.

∂(u, v) ∂x ∂y
∂v ∂v
∂(x, y)
The factor is called the Jacobian determinant of the transformation T .
∂(u, v)
Let z = f (x, y) be a real-valued continuous function on an elementary region
R in the xy-plane. We sketch a proof of the change of variables theorem. For a
detailed proof, consult an advanced calculus textbook.
� � n �
� m
f (x, y)dxdy ≈ f (xi , yj )∆xi ∆yj
R i=1 j=1
n �
m � �
� � ∂(x, y) �
≈ �
f (x(ui , vj ), y(ui , vj )) � � ∆u∆v
∂(u, v) �
i=1 j=1
� � � �
� ∂(x, y) �
= �
f (x(u, v), y(u, v)) � � dudv
S ∂(u, v) �
We summarize the discussion into a theorem.
Theorem 2.7 Change of Variables
Let z = f (x, y) be a continuous real-valued function defined on an elementary

region R in the xy-plane. Let (x, y) = T (u, v) be a one-to-one function defined on
an elementary region S in the uv-plane. Suppose the components of T , namely,
x = x(u, v) and y = y(u, v) have continuous partial derivatives and T (S) = R.
Then � � � � � �
� ∂(x, y) �
f (x, y)dxdy = �
f (x(u, v), y(u, v)) � � dudv.
R S ∂(u, v) �
y
1 R
Example 1 Applying a Change of Variables
x � � �
1.5
�0.5 Evaluate x2 − y 2 dA where R is the rectangular region with vertices
R
1 1
Figure 3 at (0, 0), (1, 1), ( 32 , 12 ), and ( 12 , − 12 ). Let x = (u + v) and y = (u − v).
2 2
Region in xy-plane
Solution Solving for u and v, we find u = x + y and v = x − y. The four vertices v

of R correspond to four points in the uv-plane:
S
1
(x, y) (u, v)
u
(0, 0) (0, 0) 2
(1, 1) (2, 0)
( 32 , 12 ) (2, 1) Figure 4
Region in uv-plane
( 12 , − 12 ) (0, 1)
The region R is mapped to region S in the uv-plane, see Figure 2. Since x = (u+v)/2
and y = (u − v)/2. The Jacobian determinant is given by
� ∂x ∂x � � 1 1 �
∂(x, y) ∂u ∂v 2 2 1
= Det = Det =−
∂(u, v) ∂y ∂y 1
− 1 2
∂u ∂v 2 2
Notice, x2 − y 2 = uv. Applying Theorem 2.7, we obtain
� � � � 2 � 1 � �
√ � ∂(x, y) �
x2 − y2 dA = �
uv � � dvdu
R 0 0 ∂(u, v) �
��2√ � �� 1 �
1 √
= udu vdv
2 0 0
√
4 2
=
9
✷
Try This 1
Evaluate the same integral in Example 1, but let x = u + v and y = u − v.
y
Example 2 Applying Polar Coordinates
� �
2 2
Evaluate e−(x +y ) dA where R is the region between the circles x2 + y 2 = 9
R
and x2 + y 2 = 4, where y ≥ 0. See Figure 5.
Solution We apply the change of variables x = r cos θ and y = r sin θ, where x
2 3
r ≥ 0. These equations arise from trigonometry, as shown below.
Figure 5
r A region S between two
y concentric circles.
Θ
x
Figure 6 Right triangle trigonometry
�
Θ Solving for r and θ, we have r = x2 + y 2 and tan θ = y/x if x �= 0. In
particular, r is the distance between point (x, y) and the origin. In Figure 5, we
Π deduce that r satisfies 2 ≤ r ≤ 3. Also, θ is the standard angle that the line through
the origin and point (x, y) makes with the positive x-axis. From Figure 5, we see
θ satisfies 0 ≤ θ ≤ π. The region S in the rθ-plane corresponding to region R is
S
shown in Figure 7.
Since x = r cos θ and y = r sin θ, the Jacobian determinant is given by
r
1 2 3 � ∂x ∂x � � �
Figure 7 ∂(x, y) ∂r ∂θ cos θ −r sin θ
= Det = Det
Rectangular region S ∂(r, θ) ∂y ∂y
sin θ r cos θ
in the rθ-plane. ∂r ∂θ
� �
= r cos2 θ + sin2 θ = r.
Applying r2 = x2 + y 2 and the change of variables, we obtain

� � � � � �
2 2 2 � ∂(x, y) �
e−(x +y ) dA = e−r �� drdθ
R S ∂(u, v) �
� π � 3
2
= re−r drdθ
0 2
� π �r=3
1 �
−r 2 �
= − e � dθ
2 0 r=2
� �
2
+y 2 ) π � −4 �
e−(x dA = e − e−9 .
R 2
✷
Try This 2
� � �
Evaluate x2 + y 2 dA where R is the circular region bounded by x2 + y 2 = 1.
R
Apply the transformation defined by x = r cos θ and y = r sin θ.
Next, we discuss the change of variables for triple integrals. Let T be a one-to-
one transformation from a rectangular box
B = {(u, v, w) : a1 ≤ u ≤ b1 , a2 ≤ v ≤ b2 , a3 ≤ w ≤ b3 }
into R3 where ai < bi are real constants, i = 1, 2, 3. We investigate the effect of T

on the volume of B.
Let x = x(u, v, w), y = y(u, v, w), z = z(u, v, w) be real-valued functions on B
with continuous partial derivatives such that (x, y, z) = T (u, v, w). The derivative
T � of T at (a1 , a2 , a3 ) provides the best linear approximation3 of T near (a1 , a2 , a3 ).
3 The linear approximation of T near x = (a , a , a ) is defined by T � (x) = A(x − x ) + T (x )

0 1 2 3 0 0
where A is a 3 by 3 matrix, and x, x0 ∈ R3 are realized as column vectors. See (36), page 39.
We apply a result from linear algebra stating that an injective linear transfor-
mation in 3-space should map B onto a parallelepiped C. If all the sides of B are
small, then the volume of T (B) is approximately the volume of the parallelepiped
T � (B) whose three defining edges are vectors
∂T ∂T ∂T
v 1 = ∆u , v 2 = ∆v , v 3 = ∆w .
∂u ∂v ∂w
where the partial derivatives are evaluated at (a, b, c), ∆u = b1 − a1 , ∆v = b2 − a2 ,
and ∆w = b3 − a3 . By Theorem 1.2, the volume of T � (B) is the absolute value of
the triple scalar product of v 1 , v 2 , and v 3 . Applying identity (1) in page 152, the
triple scalar product is the determinant of the 3 by 3 matrix whose rows are the
vectors. That is,
∂(x, y, w)
(v 1 × v 2 ) · v 3 = Vol(B)
∂(u, v, w)
where the volume of B is Vol(B) = ∆u∆v∆w, and the Jacobian determinant is

 ∂x ∂x ∂x 
∂u ∂v ∂w
 
∂(x, y, z)  ∂y ∂y ∂y 
= Det 
 ∂u ∂v ∂w 

∂(u, v, w)  
∂z ∂z ∂z
∂u ∂v ∂w
and the partial derivatives are evaluated at (a, b, c). Then an effect of transformation
T on the volume of B is given by
� �
� ∂(x, y, w) �
Vol(T (B)) ≈ Vol(T � (B)) = �� Vol(B).
∂(u, v, w) �
Furthermore, let A = F (x, y, z) be a real-valued function. We sketch a proof of the
change of variables for triple integrals.
� � � �n � m � p
F (x, y, z)dV ≈ F (xi , yj , zk )∆xi ∆yj ∆zk
M i=1 j=1 k=1
n �
� p
m � � �
� ∂(x, y, z) �
≈ �
(F ◦ T )(ui , vj , wk ) � � ∆u∆v∆w
i=1 j=1 k=1
∂(u, v, w) �
� � � � �
� ∂(x, y, z) �
= �
(F ◦ T )(u, v, w) � � dV
N ∂(u, v, w) �
A detailed proof of the change of variables is found in advanced calculus books.
Theorem 2.8 Change of Variables II
Let A = F (x, y, z) be a continuous real-valued function defined on an elementary

region M in the xyz-space. Let (x, y, z) = T (u, v, w) be a one-to-one function
defined on an elementary region N in the uvw-space. Suppose the components of
T , namely, x = x(u, v, w), y = y(u, v, w) and z = y(u, v, w) have continuous partial
derivatives and T (N ) = M . Then
� � � � � � � �
� ∂(x, y, z) �
F (x, y, z)dV = (F ◦ T )(u, v, w) �� dV.
M N ∂(u, v, w) �
Example 3 Triple Integral and Change of Variables

� � � �
Evaluate x2 + y 2 dV where M is the solid bounded by the circular
M
2 2
cylinder x + y = 4, and the planes z = 0 and z = 1, see Figure 8.
Apply the transformation x = r cos θ, y = r sin θ, and z = z.
Figure 8
Cylindrical solid of radius 2 Solution The base of the solid M lies in the xy-plane. Notice, x2 + y 2 = r2 . Then
each point (x, y, 0) in the base satisfies x2 + y 2 ≤ 4, or equivalently 0 ≤ r ≤ 2 and
0 ≤ θ ≤ 2π. In addition, the z-component of any point in M satisfies 0 ≤ z ≤ 1.
Then the solid defined by
N = {(r, θ, z) : 0 ≤ r ≤ 2, 0 ≤ θ ≤ 2π, 0 ≤ z ≤ 1}
is mapped into M by the transformation, see Figure 9. The Jacobian determinant

of the transformation is given by
 ∂x ∂x ∂x   
∂r ∂θ ∂z cos θ −r sin θ 0
 
∂(x, y, z)  ∂y ∂y ∂y   

r cos θ 0 
Figure 9 
= Det  ∂r ∂θ ∂z   =  sin θ =r
Rectangular solid N
∂(r, θ, z)  ∂z ∂z ∂z   
∂r ∂θ ∂z
0 0 1
Applying the change of variables theorem, we obtain

� � � � � � � √ � �
� ∂(x, y, z) �
2 2
x + y dV = 2
r �� dV
M N ∂(r, θ, z) �
� 1 � 2π � 2
= r2 drdθdz
0 0 0
16π
= .
3
✷
Try This 3
� � �
Evaluate ez dV where M is the solid between two circular cylinders
M
x + y = 1 and x2 + y 2 = 4, and the planes z = 0 and z = 1. See Figure 10,

2 2
Figure 10
Washer M between two and apply the same transformations used in Example 3.
cylindrical cylinders
Example 4 Changing the Variables in a Triple Integral

� � �
dV
Evaluate 2 + y2 + z2
where the solid S is the unit ball bounded by
S 1 + x
x2 + y 2 + z 2 = 1. Apply the transformation x = ρ sin φ cos θ, y = ρ sin φ sin θ, and
z = ρ cos φ where 0 ≤ ρ ≤ 1, 0 ≤ φ ≤ π, and 0 ≤ θ ≤ 2π.
Solution We find
x2 + y 2 + z 2 = (ρ sin φ cos θ)2 + (ρ sin φ sin θ)2 + (ρ cos φ)2
= (ρ sin φ)2 (cos2 θ + sin2 θ) + (ρ cos φ)2 Figure 11

Solid ball S of radius 1.
= (ρ sin φ)2 + (ρ cos φ)2
= ρ2 .
Observe, ρ is the distance between (x, y, z) and the origin. Also, θ is the angle
between the x-axis and the vector from the origin to (x, y, 0) for tan θ = y/x. The
dot product of (x, y, z) with k satisfies
(ρ sin φ cos θ, ρ sin φ sin θ, ρ cos φ) · k = ρ cos φ.
In particular, φ is the angle between vector (x, y, z) and k.
The solid ball S is mapped into a rectangular solid T , see Figure 12. The
Jacobian determinant of the transformation is given below but we postpone the
Figure 12
proof to page 83 of the section: Rectangular solid T
 ∂x ∂x ∂x 
∂ρ ∂φ ∂θ
 
∂(x, y, z)  ∂y ∂y ∂y 
= Det 
 ∂ρ ∂φ ∂θ  = ρ2 sin φ.

∂(ρ, φ, θ)  
∂z ∂z ∂z
∂ρ ∂φ ∂θ
Applying the change of variables, we obtain

� � � � � � � �
dV 1 � ∂(x, y, z) �
= � �
1 + x 2 + y2 + z2 1 + ρ2 � ∂(ρ, φ, θ) � dρdφdθ
S T
� 2π � π � 1
ρ2 sin φ
= dρdφdθ
0 0 0 1 + ρ2
� 1
ρ2
= 4π dρ
0 1 + ρ2
� 1 � �
1
= 4π 1− dρ
0 1 + ρ2
�ρ=1
�
= 4π(ρ − arctan ρ��
ρ=0
� � �
dV
= π(4 − π)
S 1 + x2 + y 2 + z 2
✷
Try This 4
� � �
Evaluate dV where S is the unit ball that is bounded by the sphere
S
x2 + y 2 + z 2 = 1. Apply the transformation x = ρ sin φ cos θ, y = ρ sin φ sin θ, and
z = ρ cos φ where 0 ≤ ρ ≤ 1, 0 ≤ φ ≤ π, and 0 ≤ θ ≤ 2π.
Cylindrical Coordinates
The cylindrical coordinates of a point (x, y, z) ∈ R3 in Cartesian coordinates are
(r, θ, z) where
y
r2 = x2 + y 2 , tan θ = if x �= 0.
x
In particular, θ is the angle between the positive x-axis and the line segment joining
the origin to (x, y, 0). Usually, we require 0 ≤ θ < 2π. Also, (r, θ) represent the polar
coordinates of (x, y). Note, r is a real number representing the directed distance
from the origin to (x, y). The cylindrical coordinates of a point are not unique.
For instance, the cylindrical coordinates (2, π/6, 3), (2, 13π/6, 3), and (−2, 7π/6, 3)
Figure 13 represent the same point in 3-space. The identities below
Cylindrical coordinates
(r, θ, z) of point P .
x = r cos θ, y = r sin θ, z = z.
are helpful when converting to Cartesian coordinates from cylindrical coordinates.
Example 5 Switching between Cartesian and Cylindrical Coordinates

√
Find the cylindrical coordinates of the point P (x, y, z) = (2, −2 3, 4) in Cartesian
� �
coordinates. Then find the Cartesian coordinates of the point Q(r, θ, z) = 4, 5π6 ,1
given in cylindrical coordinates.
√
Solution Since P (x, y, z) = (2, −2 3, 4), we find
� √
r = x2 + y 2 = 4 + 12 = 4.
√
From the identity tan θ = y/x = − 3, we may choose θ = − π3 , see Figure 14. Then
the cylindrical coordinates of point P are
� π � � 2π
�
Figure 14 (r, θ, z) = 4, − + 2kπ, 4 , −4, + 2kπ, 4
3 3
Cylindrical point
� π � where k is an integer.
(r, θ, z) = 4, − , 4 . � �
3
Using the cylindrical coordinates Q(r, θ, z) = 4, 5π
6 , 1 , we find
5π √
x = r cos θ = 4 cos = −2 3
6
5π
y = r sin θ = 4 sin = 2.
6
Thus, the Cartesian coordinates of Q are
√
Q(x, y, z) = (−2 3, 2, 1)
since z = 1.
✷
Try This 5
Find the cylindrical coordinates of point P (x, y, z) = (−1, 1, 2). Also, find the
� �
Cartesian coordinates of point 3, 3π2 , 1 given in cylindrical coordinates.
Example 6 Finding the Volume of a Solid

Find the volume of the solid S bounded by the cone x2 + y 2 = z 2 and the
sphere x2 + y 2 + z 2 = 2 where z ≥ 0, see Figure 15.
��
Solution The volume of S is S
dV . In cylindrical coordinates, we have the
identity x2 + y 2 = r2 . Then an equation of the cone x2 + y 2 = z 2 in cylindrical
coordinates is z = r. The cylindrical coordinates for the sphere is r2 + z 2 = 2.
To find where the surfaces intersect, substitute the former equation into the later Figure 15
one. Then 2z 2 =�2 and z = 1 for�z ≥ 0. Thus, every point (x, y, z) in S satisfies Solid S bounded between
a cone and a sphere.
x2 + y 2 ≤ 1 and x2 + y 2 ≤ z ≤ 2 − x2 − y 2 .
Since x2 + y 2 ≤ 1, we have 0 ≤ r ≤ 1 and 0 ≤ θ ≤ 2π, see the base of the solid in
Figure 16. The solid T corresponds to solid S under the change in coordinates from
rectangular to cylindrical coordinates. In Example 3, the Jacobian determinant
associated to the change of variables is
∂(x, y, z)
= r.
∂(r, θ, z)
� � √
From x2 + y 2 ≤ z ≤ 2 − x2 − y 2 , we obtain r ≤ z ≤ 2 − r2 . Applying the
change of variables theorem for integration, we find
� � � � � � � �
� ∂(x, y, z) �
dV = � � dV
� �
S T ∂(r, θ, z) Figure 16
√
Solid T between z = r2 ,
� 2π � 1 � 2−r 2 r2 + z 2 = 2, and θ = 2π
= rdzdrdθ in the first octant.
0 0 r
� 2π � 1 � � �
= r 2 − r2 − r2 drdθ
0 0
4π √
Volume = ( 2 − 1).
3
✷
Try This 6
Find the volume of the solid bounded by the ellipsoid x2 + y 2 + 4z 2 = 4.

Figure 17
Express the volume as a triple integral that uses cylindrical coordinates.
The ellipsoid
See Figure 17.
x2 + y 2 + 4z 2 = 4.
Spherical Coordinates
The spherical coordinates of a point (x, y, z) ∈ R3 in Cartesian coordinates are
(ρ, θ, φ) where
� y
ρ = x2 + y 2 + z 2 , tan θ = if x �= 0, and ρ cos φ = (x, y, z) · k.
x
The coordinate 0 ≤ θ < 2π is the angle between the positive x-axis and a vector
from the origin to point (x, y, 0), see Figure 18. The coordinate ρ ≥ 0 is the distance
Figure 18 between (x, y, z) and the origin. Further, we require 0 ≤ φ ≤ π and φ is the angle
Spherical coordinates between the positive z-axis and the line segment joining the origin to (x, y, z). The
(ρ, θ, φ) of point P . identities below
x = ρ sin φ cos θ, y = ρ sin φ sin θ, z = ρ cos φ
are useful when converting to Cartesian coordinates from spherical coordinates.
Example 7 Switching between Cartesian and Spherical Coordinates

� √ �
Find the spherical coordinates of point A(x, y, z) = 23 , 23 , 1 .
� �
Then find the Cartesian coordinates of the point B(ρ, θ, φ) = 8, 7π π
4 , 6 .
� √ �
Solution Since A(x, y, z) = 32 , 23 , 1 , we find
�
� 9 3
ρ= x2 + y2 + z2 = + + 1 = 2.
4 4
y √1 , π
Using the identity tan θ = x = 3
we choose θ = 6, see Figure 19. Also, we find
ρ cos φ = (x, y, z) · k
� √ �
3 3
2 cos φ = , ,1 · k = 1
2 2
π
φ = .
Figure 19 3
Spherical coordinates
� � Then the spherical coordinates of point A are
of point A 2, π6 , π3 .
� π π�
A(ρ, θ, φ) = 2, , .
6 3
� �
Next, we determine the Cartesian coordinates of B(ρ, θ, φ) = 8, 7π π
4 , 6 .
π 7π √
x = ρ sin φ cos θ = 8 sin cos = 2 2
6 4
π 7π √
y = ρ sin φ sin θ = 8 sinsin = −2 2
6 4
π √
z = ρ cos φ = 8 cos = 4 3
6
√ √ √
Thus, the Cartesian coordinates are B(x, y, z) = (2 2, −2 2, 4 3).
✷
Try This 7
√
Find the spherical coordinates of point C(x, y, z) = (3, 3 3, 6). Also, find
� �
√ 3π π
the rectangular (or Cartesian) coordinates of D (ρ, θ, φ) = 2 2, , .
4 3
For the purpose of evaluating triple integrals, we evaluate the Jacobian determi-
nant of the transformation to Cartesian from spherical coordinates. The spherical
coordinates (ρ, φ, θ) of a point (x, y, z) satisfy
x = ρ sin φ cos θ, y = ρ sin φ sin θ, z = ρ cos φ.
We compute the determinant below by expanding the minors in the third row, i.e.,
 ∂x ∂x ∂x   
∂ρ ∂φ ∂θ sin φ cos θ ρ cos φ cos θ −ρ sin φ sin θ
   
∂(x, y, z)  ∂y ∂y ∂y   sin φ sin θ ρ cos φ sin θ ρ sin φ cos θ 
= Det 
 ∂ρ ∂φ ∂θ  = 
 
∂(ρ, φ, θ)  ∂z ∂z ∂z   
∂ρ ∂φ ∂θ
cos φ −ρ sin φ 0
= Det[A3,1 ] cos φ − Det [A3,2 ] (−ρ sin φ)
where the 2 by 2 matrices A3,j are obtained from the above 3 by 3 matrix by crossing
out the third row and jth column, j = 1, 2. Namely, the matrices A3,j are
   
ρ cos φ cos θ −ρ sin φ sin θ sin φ cos θ −ρ sin φ sin θ
A3,1 =   , A3,2 =  
ρ cos φ sin θ ρ sin φ cos θ sin φ sin θ ρ sin φ cos θ
Then
Det[A3,1 ] = ρ2 sin φ cos φ cos2 θ + ρ2 sin φ cos φ sin2 θ
= ρ2 sin φ cos φ(cos2 θ + sin2 θ) = ρ2 sin φ cos φ.
Using a similar calculation, we find
Det[A3,2 ] = ρ sin2 φ.
Consequently, the Jacobian determinant for the change of variables to Cartesian

coordinates from spherical coordinates is given by
∂(x, y, z)
= Det[A3,1 ] cos φ − Det [A3,2 ] (−ρ sin φ)
∂(ρ, φ, θ)
= (ρ2 sin φ cos φ) cos φ − (ρ sin2 φ)(−ρ sin φ)
= ρ2 sin φ. (1)
Example 8 Triple Integral and Spherical Coordinates

� � �
Evaluate zdV where S is the solid between the two
S
spheres of radii 1 and 2, and centered at the origin. See Figure 20.
Solution In spherical coordinates, we have z = ρ cos φ. Also, the Jacobian

determinant for spherical coordinates is ρ2 sin φ, see (1).
Analyzing Figure 20, the spherical coordinate ρ satisfies 1 ≤ ρ ≤ 2 since the
Figure 20 distance between a point in S and the origin is a value between 1 and 2, inclusive.
A solid S bounded by The coordinate φ satisfies 0 ≤ φ ≤ π/2 for the angle between vector k and a vector
two hemispheres. from the origin to a point in S is a value from 0 to π/2 radians. Also, the polar
angle of any point is S satisfies 0 ≤ θ < 2π. In Figure 21, we see that the spherical
coordinates of S describe a rectangular box.
Then by the change of variables theorem, we obtain
� � � � 2π � π/2 � 2
zdV = ρ3 sin φ cos φ dρ dφ dθ
S 0 0 1
� 2π � π/2 �ρ=2
ρ4 ��
= sin φ cos φ dφ dθ
0 0 4 �ρ=1
� 2π � π/2
15
= sin φ cos φ dφ dθ
4 0 0
Figure 21
The solid S in � 2π �φ=π/2
spherical coordinates. 15 1 �
= sin (φ)��
2
dθ
4 0 2 φ=0
� 2π
15
= dθ
8 0
� � �
15π
zdV = .
S 4
Try This 8
Set up a triple integral in spherical coordinates for the volume of the part of
solid S in Figure 20 that lies in the first octant. Then evaluate the integral.
2.2 Check-It Out
Evaluate the integral by applying the indicated transformation.

� �
2
+y 2
1. ex dA where R is the circular region of radius 1 and centered at the origin.
R
Apply the transformations x = r cos θ and y = r sin θ.

� � � �
2. x2 + y 2 + z 2 dV where S is the unit ball, i.e., the solid bounded by the unit
S
sphere x2 + y 2 + z 2 = 1. Apply a change of variables by using spherical coordinates, see Example 8.
∂(x, y)
1. If x = u − v and y = u + v, the Jacobian determinant satisfies = 2.
∂(u, v)
∂(x, y)
2. If x = r cos θ and y = r sin θ, the Jacobian determinant satisfies = r2 .
∂(r, θ)
� � √ � 2π � 5
x2 +y 2
3. If R = {(x, y) : x + y ≤ 25}, then2 2
e dA = rer drdθ.
R 0 0
∂(x, y, z)
4. The Jacobian determinant for spherical coordinates satisfies = ρ sin φ.
∂(ρ, φ, θ)
5. If x = 2u and y = 3v, then by the change of variables we have
� 3� 2 � 1� 1
f (x, y)dxdy = f (2u, 3v)dudv.
0 0 0 0
In Exercises 1-8, evaluate the integral by applying the indicated change of variables. The region of integration
R in Cartesian coordinates is shown. The region S to where R is mapped by the change of variables is shown,
too. In addition, state the Jacobian determinant of the change of variables.
� �
1. ydA where R is the quadrilateral region with vertices at (0, 0), (4, 0), (1, 2), and (5, 2).
R
v−u
Let x = and y = v.
2
y v
2
2
S
R
u
1 4 5
x �8
Figure for 1
� �
2. (y −x)dA where R is the quadrilateral region bounded by the lines y = x+1, y = x+2, y = 2x−1,
R
and y = 2x + 2. Let x = u − v and y = 2u − v.

y v
5
2
R
2 u
1 2
x �1
�1 2 3
Figure for 2
� �
x−y
3. dA where R is the quadrilateral region with vertices at (1, 0), (2, 0), (1, 1), and ( 12 , 12 ).
R x+y
u+v u−v
Let x = and y = .
2 2
y v
2
1
1
2 R 1
S
x
1
2
1 2
u
1 2
Figure for 3
� �
� �2
4. x2 + y 2 dA where R is the circular unit circle centered at the origin.
R
Let x = r cos θ and y = r sin θ.

y Θ
2Π
R
x S
1
r
1
Figure for 4
� � �
5. zdV where R is the solid in the first octant bounded by the surfaces z = x2 + y 2 ,
R
x2 + y 2 = 1, and the coordinate planes. Let x = r cos θ, y = r sin θ, and z = z.
Figure for 5
� � �
2
+y 2 +z 2 )3/2
6. e(x dV where R is the solid bounded by the spheres x2 + y 2 + z 2 = 1 and
R
x + y + z 2 = 4. Change the variables to spherical coordinates.

2 2
Figure for 6
� � � � �
7. 2zdV where R is the solid bounded by the surfaces z = 4 − x2 − y 2 , z = 2 − x2 − y 2 ,
R
x = 1, y = x, and y = 0. Change the variables to cylindrical coordinates.
Figure for 7
� � � �
8. zdV where R is the solid bounded by the surfaces z = x2 + y 2 ,
R
x2 + y 2 = 1, and z = 0. Change the variables to spherical coordinates.
Figure for 8
In Exercises 9-14, if a point is defined in Cartesian coordinates, express the point in cylindrical coordinates.
If the point is given in cylindrical coordinates, change to Cartesian coordinates.
� �
√ 5π √
9. (x, y, z) = (2 3, 2, 1) 10. (r, θ, z) = 2, ,3 11. (x, y, z) = (−6, 2 3, 2)
3
� � � � � √ �
3π √ 2π 3 3
12. (r, θ, z) = 2, ,4 13. (r, θ, z) = 2, , −2 14. (x, y, z) = − , − , −3
2 3 2 2
In Exercises 15-20, if a point is given in Cartesian coordinates, express the point in spherical coordinates.
If the point is expressed in spherical coordinates, convert to Cartesian coordinates.
� � � √ �
√
√ 2π 5π 3 3
15. (x, y, z) = (1, 3, 2 3) 16. (ρ, θ, φ) = 4, , 17. (x, y, z) = ,− , −1
3 6 2 2
� � � �
7π 2π 3π � √ √ �
18. (ρ, θ, φ) = 2, , 19. (ρ, θ, φ) = 2, π, 20. (x, y, z) = 6 3, −6, 4 3
6 3 4
In Exercises 21-24, evaluate the integral by applying the indicated change of variables. Sketch the indicated
region R. Also, sketch the image S of the transformation defined by the change of variables.
� �
21. (x − y)dA where R is a quadrilateral region with vertices (0, 0), (1, 1), (−5, 1), and (−6, 0).
R
Apply the change of variables x = 2u + v and y = v.

��
22. R
(y − x)dA where R is a quadrilateral region with vertices (0, 0), (2, 0), (3, 1), and (1, 1).
Apply the change of variables x = −u + v and y = v.

� �
23. dA where R is the region bounded by the lines y = x, y = x + 1, y = −x, and y = −x + 1.
R
−u + v u+v
Let x = and y = .
2 2
� � �
24. y + 2x dA where R is the region bounded by the lines y = 4x + 2, y = 4x + 5,
R
−u + v u + 2v
y = −2x + 3, and y = −2x + 1. Let x = and y = .
6 3
In Exercises 25-28, solve for x and y in the given substitution. Then evaluate the integral by applying a
change of variables. Sketch the region R in the Cartesian plane, and the image S of the transformation
defined by the change of variables.
� �
√
25. x − y dA where R is a trapezoid with vertices (1, 0), (2, 0), (0, −2) and (0, −1).
R
Substitute u = x + y and v = x − y.
� �
26. xy dA where R is the region in the first quadrant that is bounded by the graphs of y = x, y = 3x,
R
xy = 1, and xy = 3. Substitute u = xy and v = y.

� �
27. (x + y) dA where R is the triangular region with vertices (0, 0), (2, 1), and (1, 2).
R
Substitute u = −x + 2y and v = x + y.
� �
28. (x − y)(1 + x + y)dA where R is the triangular region with vertices (0, 0), (1, 0), and (1, 1).
R
Substitute u = x + y and v = x − y.
� �
29. (x2 − y 2 )dA where R is a quadrilateral region bounded by the lines y = x, y = x − 2, y = −x,
R
u+v −u + v
and y = −x + 3. Substitute x = and y = .
2 2
� � � �
2 2 2
30. Verify the identity e−x −y dA = π 1 − e−n where Rn = {(x, y) : x2 + y 2 ≤ n2 }, n ≥ 1.
Rn
31. We adopt the notation of Exercise 30. Applying the Monotone Convergence Theorem4 , we have
� � � �
2 2 2 2
e−x −y dA = lim e−x −y dA.
R2 n→∞ Rn
Moreover, by Fubini’s theorem, we have

� � � ∞ � ∞
−x2 −y 2 2
−y 2
e dA = e−x dxdy.
R2 −∞ −∞
� ∞
2 √
Then verify the identity e−x dx = π
−∞
4 The Monotone Convergence Theorem and a version of Fubini’s Theorem used in Exercise 31
are proved in a course such as integration theory or real analysis. The proof is beyond the scope
of this text.
32. A solid is bounded by the surfaces x2 + y 2 = 1, z = 0, and z = 2. Sketch the solid.

Then find the volume of the solid by evaluating a triple integral in cylindrical coordinates.
In Exercises 33-36, evaluate the integral by changing the variables to spherical coordinates.
Include a graph of solid R.
� � �
� �
33. 1 − x2 − y 2 − z 2 dV where R is the unit ball of radius 1 and centered at the origin.
R
� � �
34. z dV where R is a solid in the first octant, and R is one-eighth of a solid ball of radius 2
R
� � �
π
35. dV which is the volume of the solid R inside the sphere ρ = 1 and above the cone φ = 3.
R
� � �
36. dV which is the volume of the solid R inside the sphere ρ = 2, above the plane z = 0, and
R
π
below the cone φ = 4.
In Exercises 37-38, evaluate the integral by changing the variables to cylindrical coordinates.
Also, sketch the solid R.
� � �
37. xy dV where R is the solid in the first octant bounded by x2 + y 2 = 4, z = 0, and z = 1.
R
� � � �
z
38. 2 2
dV where R is the solid bounded by z = x2 + y 2 and z = 1.
R 1+x +y
2.3. LINE INTEGRALS AND A FUNDAMENTAL THEOREM 91
2.3 Line Integrals and a Fundamental Theorem

• The Arc Length Function • Line Integral • Re-parametrizing a Curve
• Fundamental Theorem for Line Integrals • Path Integrals
The Arc Length Function

Geometrically, a curve is an arc of a graph such as a parabola, circle, line, and z
others. In this section we study integrals of functions that are defined on curves. r'!t"
We begin with a definition of the parametrization of a curve.
r!t"
Definition 2 Parametrization of a Curve

y
Let r be a continuous function from a closed interval [a, b] to Rn where n = 2, 3. x
Suppose r satisfies the two conditions below. Figure 1
A position vector r(t) and
a) r is one-to-one on [a, b], or r is one-to-one on [a, b) and r(a) = r(b). a tangent vector r� (t)
b) r � (t) is continuous on (a, b) such that �r � (t)� and �r �1(t)� are bounded.
In such a case, r parametrizes the smooth curve C defined by
C = {r(t) | a ≤ t ≤ b}.
For simplicity, we may write that a smooth curve is a curve. If r is one-to-one

on [a, b], we say C is a non-intersecting curve. If r is one-to-one on [a, b) and
r(a) = r(b), we say C is a simple closed curve.
Suppose n = 3 and r(t) = (x(t), y(t), z(t)) is a vector in standard position, i.e.,
the initial point of r(t) is the origin, and the terminal point is (x(t), y(t), z(t)).
If a < t < b and ∆t > 0 is a small, the difference quotient
r (t + ∆t) − r(t)
∆t
is a vector whose initial point is the terminal point of r(t). As ∆t → 0, the limit
r � (t) is a vector that is tangent to the curve C at point r(t), see Figure 1.
To define the arc length of C, partition [a, b] into m subintervals. Let t0 = a,
∆t = (b − a)/m, and tk = tk−1 + ∆t for k = 1, . . . , m − 1. Using the partition,
subdivide C into m subarcs, and find the sum of the distances between the initial
and terminal points of each subarc.
The distance between the endpoints of r(tk ) and r(tk−1 ) satisfies
�
2 2 2
�r (tk ) − r(tk−1 )� = x� (pk ) + y � (qk ) + z � (vk ) ∆t
because of the Mean Value Theorem, where tk−1 < pk , qk , vk < tk . We recall
�
�r � (t)� = (x� (t))2 + (y � (t))2 + (z � (t))2 .
Since x� , y � , z � are continuous, we estimate each of pk , qk , vk by a single value t∗k
in (tk−1 , tk ). Consequently,
�r (tk ) − r(tk−1 )� ≈ �r � (t∗k )� ∆t.
Then a reasonable approximation for the arc length of C is the Riemann sum
m−1
� m−1
�
�r (tk ) − r(tk−1 )� ≈ �r � (t∗k )� ∆t.
k=0 k=0
�
By definition, �r � is bounded and continuous, and consequently, integrable.
We let m → ∞, and define the limit of the Riemann sums as the arc length of C.
� b
Length(C) = �r � (t)� dt.
a
The arc length of C is independent of the parametrization; this follows from a
modification of Theorem 2.9 to be discussed later in the section. Moreover, the
arc length function is defined by
� t
s(t) = �r � (w)� dw. (2)
a
Example 1 Arc Length Function

z
Find the arc length function for the curve parametrized by
� �
3/2 3/2 3t 3
r(t) = t , 2t , , 0≤t≤ .
2 5
Then find the arc length from r(0) to r(3/5), see Figure 2.
y Solution The derivative is

� �
� 3√ √ 3
r (t) = t, 3 t, .
x 2 2
Figure 2 Then the magnitude of derivative is
A curve emanating
� �
� 9 9 45 9
from the origin. �r (t)� = t + 9t + = t+
4 4 4 4
3√
= 5t + 1.
2
Thus, the arc length function satisfies
� t � t
3√
s(t) = �r � (w)� dw = 5w + 1dw
0 0 2
�w=t
3 2�
= (5w + 1)3/2 ��
10 3 w=0
1� �
= (5t + 1)3/2 − 1
5
Hence, the arc length from r(0) to r(3/5) is given by

� �
3 1 � 3/2 � 7
s = 4 −1 v = .
5 5 5
✷
Try This 1
Find the arc length function for
r(θ) = (cos θ, sin θ, θ), 0 ≤ θ ≤ 2π.
The graph of r is a helix, see Figure 3.
The Line Integral

�b
A line integral generalizes the concept of an integral a
f (x)dx of a real-valued
function f defined on [a, b]. In a line integral, we integrate a vector field F that is
defined on a curve.
Figure 3
The helix r lies on a
Definition 3 Line Integral circular cylinder of radius 1.
Let C ⊆ Rn be a curve parametrized by a function r defined on I = [a, b] where

n = 2, 3. Let F be a vector field defined on C such that F ◦ r is continuous from I
�
to Rn . The line integral of F along C is denoted by C F · dr, and defined by
� � b
F · dr = F (r(t)) · r � (t)dt
C a
Later in the section, we show the line integral of a vector field along a curve is
independent of the parametrization provided the parameterization is orientation
preserving, see Theorem 2.10.
We discuss an application of line integrals. Suppose a vector field F represents
a force that causes a particle to move along a curve C parametrized by r. That is,
�
the particle is at position r(t) at time t. We show the line integral C F · dr
represents the work done by the force on the particle.
The unit tangent vector to r(t) is defined by
1
T (t) = r � (t). (3)
�r � (t)�
Since the differential of the arc length function is ds = �r � (t)� dt, we write
� � b
r � (t)
F · dr = F (r(t)) · �r � (t)�dt
C a �r � (t)�
� b
= F (r(t)) · T (t) ds. (4)
a
Recall, the projection of vector v = F (r(t)) onto unit vector w = T (t) satisfies
Projw (v) = (F (r(t)) · T (t)) T (t)
see identity (15) in page 15. The scalar F (r(t)) · T (t) ds in the integrand of (4)
represents the work done by the force F along the tangential component T (t),
as the particle moves through an arc length of ds. Then the line integral (4) is
the work done by a force F acting on a particle that is moving along curve C.
z Example 2 Evaluating a Line Integral
Let F (x, y, z) = (yz, x, 2x) be a vector field. Let C be the curve parametrized by
� � �
r(t) = t, t2 , t3 , 0 ≤ t ≤ 1, see Figure 4. Then evaluate C F · dr.
Solution Let r(t) = (x(t), y(t), z(t)) where x(t) = t, y(t) = t2 , and z(t) = t3 .
y Evaluating the vector field, we obtain
x F (r(t)) = F (x(t), y(t), z(t))
Figure 4 � � = (y(t)z(t), x(t), 2x(t))

The curve r(t) = t, t2 , t3 , �5 �
0≤t≤1 = t , t, 2t .
Also, the derivative of the curve is

� �
r � (t) = 1, 2t, 3t2 .
Applying Definition 3, the line integral is given by

� � 1
F · dr = F (r(t)) · r � (t)dt
C 0
� 1 � � � �
= t5 , t, 2t · 1, 2t, 3t2 dt
0
� 1 � �
= t5 + 2t2 + 6t3 dt
0
� �t=1
t6 2t3 3t4 ��
= + +
6 3 2 �t=0
�
7
z F · dr = .
C 3
✷
Try This 2
� �
Let F (x, y, z) = (x, z, y) be a vector field. Let r(t) = t, 2t, t2 , 0 ≤ t ≤ 2, be a
x �
y parametrization of curve C. Then evaluate the line integral C F · dr.
Figure 5
For Try This 2
Re-parametrizing a Curve
We study the effect of re-parametrizing a curve has on a line integral. Let C be a

curve that is parametrized by a function r defined on [a, b]. The orientation of C
with respect to r is the direction along C that the points r(t) trace as t increases.
For instance, the orientation may be counter clockwise on a simple closed curve.
Let α be a one-to-one continuous function from [c, d] onto [a, b] such that the
1
derivative α� and α� are bounded and continuous on (c, d). We say the composite
function r ◦ α is a re-parametrization of r.
Notice, r ◦ α parametrizes C. We say r ◦ α is orientation preserving if the
orientations of C with respect to r and to r ◦ α are the same. Otherwise, we
say r ◦ α is orientation reversing.
The line integral of vector field F using the parametrization r1 = r ◦ α satisfies
� d � d
F (r1 (t)) · r1 � (t) dt = F (r(α(t)) · r � (α(t)) α� (t) dt Chain Rule
c c
� α(d)
= F (r(w)) · r � (w) dw where w = α(t), dw = α� (t)dt
α(c)
� b
= ± F (r(w)) · r � (w) dw (5)
a
�b �a
for a
f (x)dx = − b
f (x)dx. The sign in (5) depends on whether α(c) = a or
α(c) = b. We choose the plus sign if α(c) = a, i.e., r1 is orientation preserving.
Otherwise, we choose the negative sign if r1 is orientation reversing.
Moreover, any two parametrizations of a curve C with the same initial points are
re-parametrizations of each other, see theorem below.
Theorem 2.9 Parametrization of Curves
Let r1 and r2 be functions defined on I1 = [a, b] and I2 = [c, d], respectively, that
parametrize a curve C. If r1 (a) = r2 (c), then r1 is a re-parametrization of r2 .
Proof Let t ∈ (a, b), and choose s ∈ (c, d) such that r1 (t) = r2 (s). Then the
function given by s = f (t), f (a) = c, and f (b) = d is a bijection from I1 onto I2 .
By definition, r1 = r2 ◦ f . We show f is differentiable. If ∆t �= 0 is small enough,
choose ∆s �= 0 such that r1 (t + ∆t) = r2 (s + ∆s). Then
r1 (t + ∆t) − r1 (t) = r2 (s + ∆s) − r2 (s).
By definition, we have f (t + ∆t) = s + ∆s.

Then
f (t + ∆t) − f (t) ∆s
f � (t) = lim = lim
∆t→0 ∆t ∆t→0 ∆t
�r1 (t + ∆t) − r1 (t)�

= lim ∆t .
∆t→0 �r2 (s + ∆s) − r2 (s)�
∆s
By the continuity of the curves, ∆t → 0 if and only if ∆s → 0. Consequently,
�r1 � (t)�
f � (t) = . (6)
�r2 � (s)�
By definition, ri � and r1i � are bounded for i = 1, 2. Then f � and f1� are bounded
on (a, b). Hence, r1 = r2 ◦ f , i.e., r1 is a re-parametrization of r2 .
✷
A curve C has exactly two orientations. If we fix an orientation for C, the other
orientation is called the opposite orientation and which we denote by −C.
Let C be a non-intersecting oriented curve, and let F be a continuous vector field
on C. Applying identity (5) and Theorem 2.9, the line integral of F along C is
independent of any orientation preserving parametrization r of C.
Next, we analyze the line integral on a simple closed curve C. Note, the initial
points of two functions r1 and r2 that parametrize C may not be the same. We
r2�c��
y assume the orientations of C with respect to r1 and r2 are the same.
r2�d��
r1�w�
Suppose r1 is defined on [a, b], and r2 is defined on [c, d]. Choose w, v such that
r1 (w) = r2 (c) = r2 (d), and r2 (v) = r1 (a) = r1 (b), as in Figure 6. Then
r2�v�
� � �
x b w b
�r1�a�
F (r1 (t)) · r1 � (t)dt = F (r1 (t)) · r1 � (t)dt + F (r1 (t)) · r1 � (t)dt
�r1�b�
a a w
� d � v
= F (r2 (t)) · r2 � (t)dt + F (r2 (t)) · r2 � (t)dt
v c
Figure 6
Parametrizations r1 � d
and r2 with different = F (r2 (t)) · r2 � (t)dt

c
initial points.
where in the middle equation we applied identity (5) with the plus sign. Thus, the
line integral of a vector field along an oriented simple closed curve is independent
of the orientation preserving parametrization of the curve. We summarize below.
Theorem 2.10 Line Integrals and Re-parametrization of Curves

Let C be an oriented curve, and let F be a continuous vector field defined on C.
�
Then the line integral C F · dr is independent of the orientation preserving
parametrization of C. Also,
� �
F · dr = − F · dr.
−C C
Moreover, the lengths of curves C and −C are the same.
Proof We only have to prove the last statement. Let r be a parametrization

of C, and let r1 = r ◦ α be a re-parametrization where α is a one-to-one function
from [c, d] onto [a, b] such that α� and 1/α� are continuous and bounded on (a, b).
Since α is one-to-one, α� is entirely nonpositive or entirely nonnegative. Then
� d � d
�r1 � (t)� dt = �r � (α(t))� |α� (t)| dt
c c
 � d

 �r � (α(t))� α� (t) dt if α is increasing

 c
= �

 d

 − �r � (α(t))� α� (t) dt if α is decreasing
c
 � b

 �r � (w)� dw if w = α(t), and α is increasing


a
= �

 a

 − �r � (w)� dw if α is decreasing
b
� d � b
�r1 � (t)� dt = �r � (t)� dt
c a
Hence, the arc lengths of C and −C are equal.

✷
Given a parametrization r of a curve C defined on [a, b], there is a natural orientation-

reversing parametrization r ∗ of −C. Namely, r ∗ is defined on [a, b] and
r ∗ (t) = r (a + b − t) . (7)
Example 3 Evaluating a Line Integral

z
Let F (x, y, z) = (sin πx, cos πz, y) be a vector field. Let C be the oriented line
�
segment from A(1, 0, 0) to B(0, 2, 1), see Figure 7. Then evaluate C F · dr. 1
Solution The vector from A to B is (−1, 2, 1). Then a parametric equation of 1

2 y
the line segment from A to B is x
r(t) = (−1, 2, 1)t + (1, 0, 0) = (1 − t, 2t, t) , 0 ≤ t ≤ 1. Figure 7
� Line segment from
Since r (t) = (−1, 2, 1), we obtain A(1, 0, 0) to B(0, 2, 1).
� � 1
F · dr = (sin π(1 − t), cos πt, 2t) · (−1, 2, 1)dt
C 0
� 1
= (− sin π(1 − t) + 2 cos πt + 2t) dt
0
�1
1 2 �
= − cos(π(1 − t)) + sin(πt) + t2 ��
π π 0
�
2
F · dr = 1− .
C π
✷
Try This 3
Evaluate the line integral in Example 3 where C is the oriented line segment
from A(2, 0, 0) to B(0, 4, 2).
Example 4 Evaluating a Line Integral
y Let F (x, y) = (−y, x) be a vector field. Let C be a circle of radius 2, centered at

�
the origin, and oriented counter clockwise. Then evaluate the line integral C F ·dr.
Solution The standard equation of circle C is x2 + y 2 = 4, see Figure 8. Using

x
2 the identity cos2 θ + sin2 θ = 1, the parametrization
r(θ) = (2 cos θ, 2 sin θ) , 0 ≤ θ ≤ 2π
traces the circle in a counter clockwise motion. The derivative is
Figure 8
A counter-clockwise path r � (θ) = (−2 sin θ, 2 cos θ) .
around a circle of radius 2
According to Theorem 2.10, the line integral is independent of the orientation
preserving parametrization. Thus,
� � 2π
F · dr = F (2 cos θ, 2 sin θ) · (−2 sin θ, 2 cos θ) dθ
C 0
� 2π
= (−2 sin θ, 2 cos θ) · (−2 sin θ, 2 cos θ) dθ
0
� 2π � 2π
� �
= 4 sin2 θ + 4 cos2 θ dθ = 4 dθ
0 0
�
F · dr = 8π.
C
✷
Try This 4
Let F (x, y) = (−y, 1) be a vector field. Let C be the unit circle oriented counter
�
clockwise. Then evaluate the line integral C F · dr.
We extend the definition of a line integral of a vector field. Let C be a finite union
of smooth curves Ci , i = 1, . . . , n.
C = C1 ∪ · · · ∪ Cn
�
We denote the line integral of F along C by C F · ds, and define it by
� � �
F · dr = F · dr + · · · + F · dr. (8)
C C1 Cn
Using Theorem 2.10, it can be shown that the left side is independent of the finite
union of smooth curves Ci . We say C is a piecewise-smooth curve.
Example 5 Line Integral Along a Triangular Image
Let F (x, y, z) = (xz, 2y, yz) be a vector field. Let C be a simple closed triangular
path that bounds the plane 2x + 2y + z = 4, and the coordinate planes. The
�
orientation is counter clockwise as shown in Figure 9. Then evaluate C F · dr.
Solution We parametrize the segment C1 from (2, 0, 0) to (0, 2, 0). Since y is z
increasing from y = 0 to y = 2, let y = t where 0 ≤ t ≤ 2. Notice, the plane 4

x + y = 2 contains C1 . Then x = 2 − t, and we obtain a parametrization of C1 :
r3 r2
r1 (t) = (2 − t, t, 0), 0 ≤ t ≤ 2.
Applying Definition 3 and F (x, y, z) = (xz, 2y, yz), we find 2
� � 2 2 y
F · dr = (0, 2t, 0) · (−1, 1, 0) dt = 4. r1
x
C1 0
Figure 9
Secondly, consider the segment C2 from (0, 2, 0) to (0, 0, 4). Then z is increasing A triangular path in the
z = 0 to z = 4. Let z = t for 0 ≤ t ≤ 4. Note, the plane 2y + z = 4 contains C2 . plane 2x + 2y + z = 4.
Then y = (4 − t)/2, and we have a parametrization for C2 :
� �
4−t
r2 (t) = 0, , t , 0 ≤ t ≤ 4.
2
The line integral along C2 satisfies
� � 4� � � �
4t − t2 1
F · dr = 0, 4 − t, · 0, − , 1 dt
C2 0 2 2
� 4
1 � � 4
= −t2 + 5t − 4 dt = .
2 0 3
Similarly, a parametrization of the segment C3 from (0, 0, 4) to (2, 0, 0) is given by
r3 (t) = (t, 0, 4 − 2t) , 0 ≤ t ≤ 2.
Then the line integral along C3 satisfies
� � 2
F · dr = F (t, 0, 4 − 2t) · (1, 0, −2) dt
C3 0
� 2 � � 8
= 4t − 2t2 , 0, 0 · (1, 0, −2) dt = .
0 3
Finally, we obtain
� � � �
F · dr = F · dr + F · dr + F · dr
C C1 C2 C3
4 8
= 4+ +
3 3
= 8.
✷
1 Try This 5
Let F (x, y) = (2y, 1) be a vector field. Let C be a counter clockwise closed curve
�
bounded y = x2 and y = x, see Figure 10. Then evaluate the line integral C F · dr.
x
1
Figure 10
An oriented closed path
along y = x2 and y = x. Fundamental Theorem for Line Integrals
�
If a vector field F is conservative, the line integral C
F · dr depends only on F ,
and the initial and terminal points of C, as the next theorem shows.
Theorem 2.11 Fundamental Theorem for Line Integrals
Let F be a continuous vector field satisfying F = ∇f for some real-valued

function f . Let C be an oriented curve that is parametrized by a function r
with initial point r(a) and terminal point r(b). Then
�
F · dr = f (r(b)) − f (r(a)) .
C
� �
∂f ∂f
Proof For simplicity, let F be a function of two variables. Let F = , .
∂x ∂y
By the chain rule and the Fundamental Theorem of Calculus, we find
� � b� � � �
∂f �� ∂f ��
F · dr = , · r � (t)dt
C a ∂x �r (t) ∂y �r (t)
� b
�
= (f ◦ r) (t) dt
a
= f (r(b)) − f (r(a)) .
✷
We recall a characterization of conservative vector fields, see Theorem 1.10.
Let G be a vector field defined on R3 with continuous first partial derivatives
except for finitely many points. Then curl(G) = 0 if and only if G = ∇g for
some real-valued function g.
If G = (M, N, P ) where M, N, P are real-valued functions of x, y, z, then the
proof of Theorem 1.10 shows
� � � � � �
∂P ∂N ∂P ∂M ∂N ∂M
curl G = − i− − j+ − k.
∂y ∂z ∂x ∂z ∂x ∂y
Thus, curl(G) = 0 if and only if

∂P ∂N ∂P ∂M ∂N ∂M
= , = , = (9)
∂y ∂z ∂x ∂z ∂x ∂y
Often, we may denote the line integral of G = (M, N, P ) along C by

� �
G · dr = (M dx + N dy + P dz) .
C C
For vector fields F = (M, N ) of two variables x and y, we may denote the line
integral by � �
F · dr = (M dx + N dy) .
C C
Example 6 Fundamental Theorem for Line Integrals

Let C be a curve that is parametrized by r(t) = (2 sin t, 2 cos t), 0 ≤ t ≤ π2 . Then
�
evaluate C (ydx + (x − 1)dy) by the Fundamental Theorem for Line Integrals.
Solution We apply (9) to show that F is the gradient of a real-valued function.

We write F = (M, N, P ) where M = y, N = x − 1, and P = 0. Since M, N are
functions of x and y only, we find curl(F ) = 0 exactly when
∂N ∂M
= .
∂x ∂y
Clearly, the above statement is true. Then curl(F ) = 0. Now, we may eyeball and
directly find a function f whose gradient is F . In fact,
∇(xy − y) = (y, x − 1) = F
and we let f (x, y) = xy − y + C where C is a constant.

∂f ∂f
We present a different method for finding f . Notice, ∂x = y and ∂y = x − 1.
Integrating with respect to x, we find
f (x, y) = xy + φ(y)
∂f
where φ is a function of y only. Since ∂y = x + φ� (y), we obtain
x + φ� (y) = x−1
φ� (y) = −1
φ(y) = −y + C.
In any case, we have
f (x, y) = xy + φ(y) = xy − y + C.
Hence, by the Fundamental Theorem for Line Integrals we find

�
F · dr = f (r(π/2)) − f (r(0)) = f (2, 0) − f (0, 2) = 2.
C
✷
Try This 6
Let C be an oriented curve with initial point (1, 0) and terminal point (2, 2).
� � �
Then evaluate C y 2 dx + 2xydy by the Fundamental Theorem for Line Integrals.
Example 7 Fundamental Theorem for Line Integrals

Let C be an oriented path from point A(3, 1, 1) to B(1, 1, 2) where y �= 0. If
� �
z xz x
F (x, y, z) = , − 2, −1
y y y
�
is a vector field, evaluate C F · dr by the Fundamental Theorem for Line Integrals.
Solution We find a function f satisfying ∇f = F , but we omit the details in

verifying (9). We may eyeball f and choose f (x, y, z) = xz/y − z.
Alternatively, since ∂f /∂x = z/y, we integrate with respect to x. Then
xz
f (x, y, z) = + φ(y, z)
y
∂f x ∂φ ∂f x
where φ is a function of y and z only. Note, ∂z = y + ∂z . Also, ∂z = y −1
for ∇f = F . Thus,
x ∂φ x
+ = −1
y ∂z y
∂φ
= −1.
∂z
Then φ(y, z) = −z + ψ(y) where ψ is a function of y only. Thus,
xz
f (x, y, z) = − z + ψ(y). (10)
y
∂f ∂f
Observe, ∂y = − xz �
y 2 + ψ (y) and ∂y = − xz
y 2 for ∇f = F . Combining, we obtain
xz xz
− + ψ � (y) = −
y2 y2
ψ(y) = C
where C is a constant. Then we rewrite (10) as follows:

xz
f (x, y, z) = − z + C.
y
Hence, by the Fundamental Theorem for Line Integrals we find
�
F · dr = f (1, 1, 2) − f (3, 1, 1) = 2 − 0
C
= 2.
✷
Try This 7
Apply the Fundamental Theorem for Line Integrals in evaluating

� � �
1 y
dx + dy − 2 dz
C z z
where C is an oriented smooth path from A(2, 2, 2) to B(4, 4, 2) where z �= 0.
Path Integrals
Let C be a curve parametrized by a function r. Recall, the differential of the arc
length function in (2) satisfies
ds = �r � (t)� dt.
Next, we define the path integral of real-valued function f along C.
Definition 4 Path Integral
Let C be a smooth curve parametrized by a function r defined on [a, b]. Let f be

a real-valued function such that f ◦ r is continuous on [a, b]. The path integral of
�
f along C is denoted by C f (x)ds, and defined by
� � b
f (x)ds = f (r(t)) �r � (t)� dt.
C a
z
The line integral of f along C is independent of the parametrization r of C.
The proof of which is similar to that of Theorem 2.10. That is, if r1 and r2 are
two functions that parametrize C, and defined on [a1 , b1 ], [a2 , b2 ], respectively, then
� b1 � b2 x y
f (r1 (t)) �r1 � (t)� dt = f (r2 (t)) �r2 � (t)� dt.
a1 a2
Figure 11
� f (x) ≥ 0, the path integral
If
The above identity holds even if r1 is orientation reversing of r2 . C
f (x)ds is the area of a
fence built on a curve.
Geometrically, if f (r(t)) ≥ 0 represents the height of a fence at r(t), the path
�
integral C f (x)ds is the area of the fence, see Figure 11. The product f (x)ds is
the area of a rectangle with base ds and height f (x).
A parameter s for a parametrizing function r is called arc length parameter
if �r � (s)� = 1 for all s except possibly at the endpoints. Consequently, the arc
length of r in subinterval [c, d] is d − c. To re-parametrize r(t) by using the arc
length parameter s, we suggest the following guidelines.
Parametrizing by the Arc Length Parameter
Let r be a smooth curve on [a, b].

� t
1. Evaluate the arc length function, s = φ(t) = �r � (t)� dt.
a
2. Since s = φ(t) is an increasing function, solve for the inverse function t = φ−1 (s).
� �
3. The arc length parametrization is r1 (s) = r φ−1 (s) .
Let C be a curve parametrized by r(s) where s is the arc length parameter.

From definition (3), the unit tangent vector to the curve C at r(s) satisfies
r � (s)
T (s) = = r � (s)
�r � (s)�
see (3). The curvature κ of C is a function that is defined on C where
κ(x) = �T � (s)� (11)
and x = r(s). The curvature κ is independent of the parametrization of C.

In particular, let r1 (s1 ) and r2 (s2 ) be two functions that parametrize C where s1
and s2 are the arc length parameters. If r1 and r2 have the same orientations,
then r1 = r2 ◦ f where f � (s1 ) = 1, see identity (6). By the chain rule, we find
dr1 dr2 � dr2

= f (s1 ) = .
ds1 ds2 ds2
If r1 and r2 have opposite orientations, r1 and r2 ∗ have the same orientations
where r2 ∗ (s2 ) = r2 (a2 + b2 − s2 ), see (7). Similarly, r1 = r2 ∗ ◦ f and
� � �
dr1 dr2 ∗ �� dr2 ∗ �� dr2 ��
= f (s1 ) = =− .
ds1 ds2 �s2 =f (s1 ) ds2 �s2 =f (s1 ) ds2 �a2 +b2 −s2
Thus, we find
dr1 dr2
=± .
ds1 ds2
where the sign depends on whether r1 and r2 are orientation preserving or
reversing. In any case, the magnitudes of the second derivatives are equal, i.e.,
� 2 � � 2 �
� d r1 � � d r2 �
� � � �
� ds2 � = � ds2 � .
1 2
Thus, the curvature κ is independent of the parametrization.

Example 8 Evaluating a Path Integral

Find the curvature function κ of the curve C parametrized by
1
r(s) = √ (cos s, sin s, s)
2
�
where 0 ≤ s ≤ 2π. Then evaluate the path integral C κds.
Solution Indeed, s is the arc length parameter for

1
r � (s) = √ (− sin s, cos s, 1)
2
and
1 1 �√ �
�r � (s)� = √ � (− sin s, cos s, 1) � = √ 2 = 1.
2 2
Since T (s) = r � (s), the curvature κ at x = r(s) satisfies
� �
� 1 �
κ(x) = �T (s)� = �r (s)� = � √ (− cos s, − sin s, 0)�
� ��
�
2
1
= √ .
2
Applying Definition 4 and since �r � (s)� = 1, the path integral of κ along C
is given by
� � 2π � 2π
1
κ(x)ds = κ(r(t)) �r � (t)� dt = √ dt
C 0 0 2
√
= 2π.
Try This 8
Find the curvature function κ for the circle C defined by

� �s� � s ��
r(s) = R cos , sin
R R
�
where 0 ≤ s ≤ 2πR, and R > 0. Then evaluate the path integral C
κds.
2.3 Check-It Out
1. Find the arc length function for the curve r(t) = (t, mt + b).
2. Let F (x, y) = (−y, 3x) be a vector field. Let C be the curve parametrized
�
by r(t) = (cos t, sin t), 0 ≤ t ≤ π2 . Then evaluate C F · dr.
3. Let F (x, y) = (y, x + 1) be a vector field. Let C be the curve parametrized by r(t) = (3 cos t, 3 sin t),
�
0 ≤ t ≤ π3 . Then evaluate C F · dr by the Fundamental Theorem for Line Integrals.
4. Let G(x, y, z) = (x + y, y, z) be a vector field, and let C be the line segment that joins point
�
A(2, 1, 0) to B(1, 2, 2). Then evaluate C F · dr.
1. A parametrization of the unit circle that is oriented in the counter clockwise direction is given by
r(θ) = (sin θ, cos θ), 0 ≤ θ ≤ 2π.
2. The line segment from (1, 0, 0) to (0, 1, 1) is parametrized by r(t) = (1 − t, t, t) where 0 ≤ t ≤ 1.
3. The unit tangent vector to a curve parametrized by r(t), a ≤ t ≤ b, is given by r � (t).
4. If C is a curve parametrized by r(t) = (t, 0, 1) where 0 ≤ t ≤ 1, then the line integral of a vector
� � 1
field F on C satisfies F · dr = F (t, 0, 1) · i dt
C 0
� b
5. A definite integral f (x)dx is a line integral.
a
6. If C is a curve parametrized by r(t) = (t, 0, 1) where 0 ≤ t ≤ 1, then the line integral of a vector field
� � 1
F on C satisfies F · dr = F (t, 0, 1) · i dt
C 0
�
7. If F (x, y) = (y, 2x), the Fundamental Theorem for Line Integrals applies to F · dr.
C
8. Let r(t) be a parametrization of a curve where a ≤ t ≤ b . Then t is the arc length parameter if there
is a constant c > 0 satisfying �r� (t)� = c for all a < t < b.
9. Let r(t) be a parametrization of a curve where t is the arc length parameter. Then the curvature of
the curve is �r�� (t)�.
10. Let C be an oriented curve with initial point (a, b) and terminal point (c, d). If F (x, y) = (y, x),
�
then F · dr = cd − ab.
C
In Exercises 1-4, evaluate the line integral of the vector field along the curve C parametrized by r.
�
π
1. F · dr where F (x, y) = (2x, y), and r(t) = (sin t, cos t), 0 ≤ t ≤ .
C 6
�
π
2. (xdx + xydy) where r(t) = (2 sin t, cos t), 0 ≤ t ≤ .
C 6
�
3. G · dr where G(x, y, z) = (y, y + z, x), and r(t) = (2 − t, t + 1, t), 0 ≤ t ≤ 2.
C
�
π
4. (sin y dx + cos z dy − sin x dz) where r(t) = (t, 2t, t), 0 ≤ t ≤ .
C 3
In Exercises 5-12, parametrize the path C by a curve r that is one-to-one except possibly at the endpoints.
Then evaluate the line integral of the indicated vector field F or G along the path C.
�
5. F · dr where F (x, y) = (−y, x), and C is the directed line segment from (1, 0) to (3, 4).
C
�
6. F · dr where F (x, y) = (y, x), and C is the directed line segment from (1, 1) to (2, 3).
C
�
7. (dx + xdy) where C is the oriented line segment from the origin to (2, 4).
C
�
8. (3xdx + (x + y)dy) where C is the line segment from the point (2, 2) to (0, 4).
C
� � �
9. F · dr where F (x, y) = (−4y, 4x), C is a path along the unit circle that joins (1, 0) to √1 , √1
2 2
C
in the counter clockwise direction.
�
10. F · dr where F (x, y) = (x + y, x − y), and C is a path along a circle of radius 2, centered
C
√
at the origin, and joining (2, 0) to (1, 3) in the counter clockwise direction.
�
11. G · dr where G(x, y, z) = (x + 2y, y − 2z, x + z), and C is the line segment from (1, 2, 3) to (3, 6, 5).
C
�
12. (ydx + z 2 dy + xdz) where C is the line segment from (0, 1, 1) to ( 12 , 2, 0).
C
In Exercises 13-26, evaluate the line integral by the Fundamental Theorem for Line Integrals. State the
function f such that the gradient ∇f is the conservative vector field in the integral.
�
13. (y 2 dx + 2xydy) where C is a path from the origin to (−1, 3).
C
�
y
14. (x2 dx − dy) where C is an oriented path from the (1, 2) to (3, 6).
C 2
� � �
t π
15. (cos x cos y dx − sin x sin y dy) where C is the path defined by r(t) = t, ,0≤t≤ 3.
C 2
�
16. F · dr where C is a path from point A(1, 0) to B(0, 1), and F (x, y) = (yexy + 1, xexy ).
C
�
17. F ·dr where C is the image of the curve r(t) = (arccos t, arcsin t), 0 ≤ t ≤ 1, F (x, y) = (ey cos x, ey sin x).
C
�
� �
18. 6ye2x dx + 3e2x dy where C is the path defined by r(t) = t, t2 + 2 , 0 ≤ t ≤ 1.
C
�
19. 3 cos(2y)dx − 6x sin(2y)dy where C is the path defined by r(t) = (t, (t − 1)π), 1 ≤ t ≤ 2.
C
�
π
20. F · dr where F (x, y) = (−2ey sin(x), 2ey cos x), and C is defined by r(t) = (t, 6t/π), 0 ≤ t ≤ .
C 3
�
� � π
21. F · dr where F (x, y) = y cos(x)ey sin x , sin(x)ey sin x , and C is defined by r(t) = (t, 2t), 0 ≤ t ≤ .
C 3
� � �y�
�
x √
22. arcsin dx + � dy where C is a path joining point (1, 0) to ( 3, 1) with y �= ±2.
C 2 4 − y2
�
� �
23. yz 2 dx + xz 2 dy + 2xyzdz where C is a path from point P (2, 2, 1) to Q (3, 1, 2).
C
�
�π � �π �
24. (ey cos x dx + ey sin x dy + dz) where C is a path from point A 2 , 0, 1 to B 3 , 1, 2 .
C
�
25. F ·dr where C is a path from point P (0, 2, 1) to Q(2, 1, 2), and F (x, y, z) = (yexy−z , xexy−z . − exy−z ).
C
� � �
√ z
26. F · dr where C is a path from point P (1, 0, 1) to Q( 3, 2, 1), and F (x, y, z) = , 1, arctan x .
C 1 + x2
In Exercises 27-34, evaluate the path integral of the given real-valued function along the smooth curve C
parametrized by r. See Definition 4.
�
1
27. f (x, y)ds where f (x, y) = xy − x2 , r(t) = (3t, 4t), 0 ≤ t ≤ .
C 3
�
4 1
28. f (x, y)ds where f (x, y) = (x − y), r(t) = (12t, 5t), 0 ≤ t ≤ .
C 13 2
�
π
29. (x + y)ds where r(t) = (4 sin 2t, 4 cos 2t), 0 ≤ t ≤ .
C 8
�
√ π
30. f (x, y, z)ds where f (x, y, z) = x − y + z, r(t) = (cos t, sin t, 3 t), 0 ≤ t ≤ .
C 2
�
√ π
31. f (x, y, z)ds where f (x, y, z) = xy, r(t) = (2 cos t, 2 sin t, 5 t), 0 ≤ t ≤ .
C 6
�
2 cos x � � π
32. f (x, y, z)ds where f (x, y, z) = , r(t) = 8t, 2t2 , t3 /3 , 0 ≤ t ≤ .
C 16 + y 16
�
33. g(x, y, z)ds where g(x, y, z) = yz − x2 , r(t) = (t, 2t, 2t), 0 ≤ t ≤ 2.
C
�
�√ √ √�
34. g(x, y, z)ds where g(x, y, z) = xyz, r(t) = t, 2 t, 2 t , 0 ≤ t ≤ 3.
C
In Exercises 35-40, find the unit tangent vector T (t) to the smooth curve
parametrized by r. See identity (3) in page 93.

� �
35. r(t) = t, t2 36. r(t) = (8t, 4t, 8t)
37. r(t) = (a cos t, a sin t), a > 0 38. r(t) = (b cos t, b sin t, t), b > 0
� � � �√ �
39. r(t) = t, 32 t2 , t3 40. r(t) = 2 t, et , e−t
In Exercises 41-44, find the curvature of the curve parametrized by r at the
indicated point P . By the chain rule, the curvature κ satisfies
�dT /dt� �T � (t)�

κ = �T � (s)� = = � .
ds/dt �r (t)�
� �
41. r(t) = t, t2 , P (0, 0) 42. r(t) = (t, mt), P (t, mt)
� √�
43. r(t) = (3 cos t, 4 sin t), P (0, 4) 44. r(t) = t, 2 t , P (1, 2)
In Exercises 45-46, a fence is built on a curve C parametrized by r. The height of the fence at r(t) is given
by f (r(t)) where f is the indicated real-valued function. Find the area of the fence.
45. r(t) = (cos t, sin t), 0 ≤ t ≤ 2π, f (x, y) = 2 + xy 46. r(t) = (t, 3t), 0 ≤ t ≤ 1, f (x, y) = 1 + y − x
y
x
Figure for 45 Figure for 46
In Exercises 47-48, a curve C is parametrized by a function r. Find the work
done by the given force F on a particle that is moving along C.

� �
� � ln x
47. r(t) = t2 , t , 0 ≤ t ≤ 1, F (x, y) = (y, −x) 48. r(t) = (et , t), 0 ≤ t ≤ 2, F (x, y) = , x + y2
y
49. Let r be a smooth curve defined on [a, b], and let T (t) be the unit tangent vector at r(t),
see definition (3). Let s be the arc length parameter where ds/dt = �r � (t)�.
ds
a) Verify the identities r � (t) = T (t), and
dt
d2 s ds
r �� (t) = T (t) + T � (t).
dt2 dt
b) Using the fact that T (t) × T (t) = 0, prove

� �2
ds � �
� ��
r (t) × r (t) = T (t) × T � (t) .
dt
c) Using the fact that T (t) · T (t) = 1, prove T (t) is perpendicular to T � (t).
d) Show
� �2
ds � � �
� ��
�r (t) × r (t)� = �T (t)� .
dt
�T � (t)�
e) Using the fact that curvature satisfies κ(t) = , prove the identity
�r� (t)�
�r � (t) × r �� (t)�
κ(t) = 3 . (12)
�r � (t)�
� �
50. Let C be a parabolic curve parametrized by r = 0, t, t2 .
2
a) Applying the curvature identity (12), verify κ(t) = .
(1 + 4t2 )3/2
� ∞
b) Evaluate κ(t)dt.
−∞
�
c) Evaluate the path integral κds.
C
51. Let C be a hyperboloid curve parametrized by r = (sinh t, cosh t, 0).
3/2
a) Applying the curvature identity (12), verify κ(t) = (sech 2t) .
�
b) Evaluate the path integral κds.
C
52. Let r be a smooth curve defined on [a, b], and let T (t) be the unit tangent vector at r(t).
a) The normal vector to the curve at r(t) is defined by
T � (t)
N (t) = � �
�T � (t)� . (13)
b) Using the fact that T (t) · T (t) = 1, prove N (t) is perpendicular to T (t).
c) The binormal vector to the curve at r(t) is defined by
B(t) = T (t) × N (t). (14)
Using the fact that B(t) · B(t) = 1, show B � (t) · B(t) = 0.
d) Using the fact that B(t) · T (t) = 0 = B(t) · T � (t), prove B � (t) · T (t) = 0.
e) Applying c) and d), prove B � (t) is multiple of N (t).
53. If r(t) = (t, t2 , t2 ), find the unit tangent vector T (t), the normal vector N (t), and the binormal
vector B(t) at r(t).
2.4. SURFACE INTEGRALS 111
2.4 Surface Integrals

• Area of a Surface • Surface Integrals of Vector Fields • Surface Integrals
of Real-Valued Functions
Area of a Surface
In Section 2.3, we discussed line integrals along smooth curves. In this section, we
study surface integrals, or integrals of vector fields defined on surfaces. When the
surface is a subset of the xy-plane, the surface integral is a double integral as we
will see. We begin with the definition of a parametrized surface.
Definition 5 Parametrization of a Surface Figure 1

Surface parametrized by two
Let R ⊆ R2 be an elementary region5 . Let r be a continuous function from R to independent variables.
R3 that satisfies the two conditions below except possibly on the boundary of R.
a) r is one-to-one with continuous partial derivatives, and

b) the functions � �
� ∂r ∂r �
φ(u, v) = � ×
� ∂u ∂v �
� (15)
1
and are nonzero and bounded.
φ(u, v)
In such a case, we say r parametrizes the surface defined by
M = {r(u, v) | (u, v) ∈ R}.
The function φ in (15) is integrable on R, and we omit its proof since it is

similar to the application of Theorem 2.1 to Theorem 2.3 in pages 60 and 62.
Next, we define the surface area of M . For the moment, let R be a rectangular
region. Partition R into sub-rectangles Rij with uniform width ∆u and uniform
height ∆v. Let (ui , vj ) be the vertex of Rij nearest the origin, as in Figure 1.
We approximate the image r(Rij ) by a rectangular region Aij whose sides are the
r ��
vectors ∆u ∂∂u
�
and ∆v ∂∂vr �(ui ,vj ) . The area of rectangle Aij is the magnitude
(ui ,vj )
of the cross product of the vectors that define the rectangle. That is,
��
� ∂r �� ∂r ��
� �
Area (Aij ) = � ∆u × ∆v �
� ∂u (ui ,vj ) ∂v �(ui ,vj ) �
� �
� ∂r �� ∂r
�
� �
� � � �
= � � × � � ∆u∆v
� ∂u (ui ,vj ) ∂v (ui ,vj ) �
5 Elementary regions are discussed in page 61

The Riemann sum of the areas of rectangles Aij approaches a limit as the norm of
the partition of R approaches zero, due to the boundedness and continuity in (15).
The limit which is an integral is a reasonable definition for the surface area of M .
� �
Notice, the integral of � ∂ r × ∂ r � over R exists even if R is an elementary region.
∂u ∂v
In general, let r be a function satisfying Definition 5. Let M be the corresponding

surface parametrized by r. We define the surface area of M as follows:
� � � �
� ∂r ∂r �
(Surface Area of M ) = � × � dA (16)
� ∂v �
R ∂u
where dA = dudv or dA = dvdu. The surface area of M is independent of the

parametrization r in Definition 5; the proof is left as an exercise and depends
on the change of variables theorem.
We consider a special case where a surface M is the graph of a real-valued
function z = f (x, y) with continuous partial derivatives. We parametrize M by
the one-to-one function
r(x, y) = (x, y, f (x, y)).
The cross product of the partial derivatives of r is given by
� �
� i j k �
∂r ∂r � �
� 0 ∂f �
× = � 1 ∂x �
∂x ∂y � �
� 0 1 ∂f �
∂y
� �
∂f ∂f
= − , − , 1 .
∂x ∂y
Applying (16), the area of the surface defined by z = f (x, y), (x, y) ∈ R, satisfies
� � � �
� ∂r ∂r �
(Surface Area of M ) = � × � dA
� ∂y �
R ∂x
� � �2 � �2
� �
∂f ∂f
= 1+ + dA. (17)
R ∂x ∂y
z
Example 1 Evaluating a Surface Area
Find the surface area of the portion of the plane x + 2y + z = 2 in the first octant,
see Figure 2.
1 y
x 2
Solution Solving for z, we find and write
Figure 2 z = f (x, y) = 2 − x − 2y.
The plane x + 2y + z = 2
in the first octant. Let M be the triangular region in the first octant defined by z = f (x, y). The
subset R of the xy-plane below the triangular region satisfies
� �
� 2−x
�
R = (x, y) 0 ≤ x ≤ 2, 0 ≤ y ≤ .
2
We apply identity (17) to find the surface area of M . Notice,

� � �2 � � 2
∂f ∂f � √
1+ + = 1 + (−1)2 + (−2)2 = 6.
∂x ∂y
Then the surface area of M is given by

� � �2 � �2
� �
∂f ∂f
(Surface Area of M ) = 1+ + dA
R ∂x ∂y
√ � 2 � (2−x)/2
= 6 dydx
0 0
√
= 6 (Area of R)
√
= 6 sq. units
since the area of triangle R is 1 square unit. ✷
Try This 1
Find the surface area of the portion of the plane 2x + 2y + z = 4 in the first
octant above the triangular region in the xy-plane with vertices (0, 0, 0), (1, 0, 0),
and (1, 1, 0).
Example 2 Surface Area
Find the surface area of the part of the sphere
x2 + y 2 + z 2 = 25
where x2 + y 2 ≤ 16 and z ≥ 0, see Figure 3.

Figure 3
A dome above a circle
Solution We solve for z and write of radius 4.
�
z = f (x, y) = 25 − x2 − y 2 .
Evaluating the partial derivatives, we find

∂f −x ∂f −y
a) =� b) =�
∂x 25 − x2 − y 2 ∂y 25 − x2 − y 2
Then we obtain
� � �2 � �2
∂f ∂f 5
1+ + =� .
∂x ∂y 25 − x2 − y 2
Let M be the upper part of the sphere that lies above circular region
� � �
R = (x, y) � x2 + y 2 ≤ 16 .
We apply identity (17) to find the surface area of M .

� � �2 � �2
� �
∂f ∂f
(Surface Area of M ) = 1+ + dA
R ∂x ∂y
� �
5 dA
= � .
R 25 − x2 − y 2
We apply a change of variables to polar coordinates. Let
a) x = r cos θ b) y = r sin θ, c) dA = rdrdθ
where 0 ≤ r ≤ 4 and 0 ≤ θ ≤ 2π. Since x2 + y 2 = r2 , we obtain

� 4 � 2π
5r dθdr
(Surface Area of M ) = √
0 0 25 − r2
� 4
rdr
= 10π √
0 25 − r2
� �r=4
�
= −10π 25 − r ��
2
r=0
(Surface Area of M ) = 20π sq. units
Try This 2
1� 2 �
Find the area of the surface z = x + y 2 that lies below the plane z = 2.
2
See Figure 4.
Figure 4
For Try This 2
Next, we evaluate the area of a surface of revolution. Let C1 be a curve in the

yz-plane that is parametrized by
r1 (t) = (0, y(t), z(t)), a ≤ t ≤ b (18)
where y(t) ≥ 0, as in Figure 5. A surface of revolution M is generated when C1

Figure 5 is revolved about the z-axis. Then surface M is parametrized by
The curve z = f (y) is
revolved about the z-axis. r(θ, t) = (y(t) cos θ, y(t) sin θ, z(t)) (19)
where 0 ≤ θ < 2π, t ∈ [a, b]. We leave the verification of (19) as an exercise.
Moreover, the cross product of the partial derivatives of r satisfies
� �
� i j k �
� �
∂r ∂r � �
× = � −y(t) sin θ y(t) cos θ 0 ��
∂θ ∂t �
� �
� y � (t) cos θ y � (t) sin θ z � (t) �
= (y(t)z � (t) cos θ, y(t)z � (t) sin θ, −y(t)y � (t)) .

Since y(t) ≥ 0, we find
� �
� ∂r ∂r � �
� � [y � (t)]2 + [z � (t)]2
� ∂θ × ∂t � = y(t)
= y(t) �r1 � (t)� (20)

Applying (16), the area of the surface of revolution M defined by (18) and (19) is
� b � 2π � �
� ∂r ∂r �
� �
(Area of Surface of Revolution) = � ∂θ × ∂t � dθdt
a 0
� b
= 2π y(t) �r1 � (t)� dt. (21)
a
Example 3 Surface Area of a Sphere
Parametrize the sphere S of radius 2 that is centered at the origin.

Then find the surface area of S.
Solution Let C1 be the semicircle of radius 2 in the yz-plane given by

π π
r1 (ψ) = (0, 2 cos ψ, 2 sin ψ), − ≤ ψ ≤ .
2 2 Figure 6
Semicircle generates
The sphere S is generated when C1 is revolved about the z-axis, see Figure 6.
a sphere by a rotation.
Applying (19) we parametrize S as follows
r(θ, ψ) = (2 cos ψ cos θ, 2 cos ψ sin θ, 2 sin ψ)
where 0 ≤ θ < 2π. From the definition of r1 (ψ), we find �r1 � (ψ)� = 2 and let
y(ψ) = 2 cos ψ be the y-component of r1 (ψ). Using (21), we obtain
� b
(Surface Area of S) = 2π y(ψ) �r1 � (ψ)� dψ
a
� π/2
= 2π (2 cos ψ)(2) dψ
−π/2
�π/2
�
= 8π sin ψ ��
−π/2
= 16π sq. units

✷
Try This 3
Parametrize a right circular cone S whose base is a circle of radius 2 and the height
is 2 units. Assume the base lies on the xy-plane and is centered at the origin, see
Figure 7. Then find the surface area of the cone S.
Figure 7
Right circular cone with
radius 2 and height 2 units. Surface Integrals of Vector Fields
Let r be a function from an elementary region R ⊆ R2 into R3 as in Definition 5.
Let M ⊆ R3 be the surface that is parametrized by r. Let F be a vector-valued
function defined on M with values in R3 such that F ◦ r is continuous. In such
a case, we say F is a continuous vector field defined on M .
Definition 6 Surface Integral of a Vector Field
Let M ⊆ R3 be a surface parametrized by a function r satisfying Definition 5.

Let F be a continuous vector field defined on M , where the values of F lie in
��
R3 . The surface integral of F on M denoted by M
F · dS is defined by
� � � � � �
∂r ∂r
F · dS = F (r(u, v)) · × dA
M R ∂u ∂v
where dA = dudv or dA = dvdu, and R is the domain of definition of r.
An orientation on a surface M is a continuous function that assigns to each

interior point p of M a unit vector N (p) that is perpendicular to the tangent plane
��
at p. We claim the surface integral M
F · dS is independent of the orientation
preserving parametrization r of M , as we briefly explain below.
By Definition 5, r is defined on an elementary region R. Given p = r(u, v) where
(u, v) lies in the interior of R, the unit vector below
� �
1 ∂r ∂r
�
Nr (p) = ∂ r ∂ r � ×
� × � ∂u ∂v
∂u ∂v
is normal to the tangent plane at p, where the partial derivatives are evaluated at
(u, v). For the details, see Exercise 25. We say r is orientation preserving if
Nr (p) = N (p)
for all interior points p ∈ M . While r is orientation reversing if
Nr (p) = −N (p)
for all interior points p ∈ M .

y
Example 4 Evaluating a Surface Integral
2
Let F (x, y, z) = (0, 0, y) be a vector field. Let R be a triangular region in the
xy-plane with vertices at (1, 2), (0, 2), and the origin, see Figure 8. Let M be
y�2x
the surface that is oriented, and parametrized by
� � �
r(x, y) = x, y, x2 + y 2 + 1 , (x, y) ∈ R. x
1
��
Then evaluate the surface integral M
F · dS. Figure 8
Triangular region R
�
Solution The surface M is a portion of the larger surface z = x2 + y 2 + 1 that
lies above the region R in the xy-plane, see Figure 9. The cross product of the
partial derivatives of r satisfies
� �
� i j k �
� �
� x �
∂r ∂r � � �
× = �� 1 0 �
x2 + y 2 + 1 �
∂x ∂y � y �
� 0 1 � �
� �
x2 + y 2 + 1
� �
−x −y
= � , � , 1 .
x2 + y 2 + 1 x2 + y 2 + 1
Figure 9
We apply Definition 6 to evaluate the surface integral. Notice, F (r(x, y)) = (0, 0, y). Surface M above
� � � � � � region R in Figure 8
∂r ∂r
F · dS = F (r(x, y)) · × dA
M R ∂x ∂y
� � � �
1 2
−x −y
= (0, 0, y) · � , � , 1 dydx
0 2x x2 + y 2 + 1 x2 + y 2 + 1
� 1 � 2 � 1 �y=2 � 1
y 2 ��
= ydydx = � dx = 2 − 2x2 dx
0 2x 0 2 y=2x 0
� �
4
F · dS = .
M 3
✷
Try This 4
Let F (x, y, z) = (0, 0, 1) be a vector field. Let R be a triangular region in the

xy-plane with vertices at (0, 12 ), ( 12 , 0), and the origin. Let M be the surface
parametrized by
� � �
r(x, y) = x, y, 1 − x2 − y 2 , (x, y) ∈ R.
��
F · dS.

Let F (x, y, z) = (x, y, z) be a vector field on the unit sphere centered at (0, 0, 0).
Assume the sphere is oriented by unit vectors that point away from the origin.
��
F · dS.
Solution In Example 3, we parametrized a sphere of radius 2 centered at the

origin. Likewise, we obtain a parametrization r of the unit sphere, namely,
r(θ, ψ) = (cos ψ cos θ, cos ψ sin θ, sin ψ)
where 0 ≤ θ < 2π and − π2 ≤ ψ ≤ π

2. For the cross product, we simplify and find
� �
� i j k �
� �
∂r ∂r � �
× = � − cos ψ sin θ cos ψ cos θ 0 ��
∂θ ∂ψ �
� �
� − sin ψ cos θ − sin ψ sin θ cos ψ �
� �
= cos2 ψ cos θ, cos2 ψ sin θ, cos ψ sin ψ .
r , we observe the cross product vector
Analyzing the third component of ∂∂θr × ∂∂ψ
points away from the origin. In particular, r is orientation preserving.
Moreover, we find
� �
∂r ∂r
F (r(θ, ψ)) · × =
∂θ ∂ψ
� �
(cos ψ cos θ, cos ψ sin θ, sin ψ) · cos2 ψ cos θ, cos2 ψ sin θ, cos ψ sin ψ =
cos3 ψ(cos2 θ + sin2 θ) + cos ψ sin2 ψ =
cos3 ψ + cos ψ sin2 ψ =
cos ψ(cos2 θ + sin2 θ) = cos ψ.
Applying Definition 6, we evaluate the surface integral as follows:

� � � � � �
∂r ∂r
F · dS = F (r(θ, ψ)) · × dA
M R ∂θ ∂ψ
� π/2 � 2π
= cos ψ dθdψ
−π/2 0
� π/2
= 2π cos ψ dψ
−π/2
� �
F · dS = 4π.
M
✷
Try This 5
The orientation on the unit sphere M are unit vectors that point away from the
��
origin. If F (x, y, z) = (0, 0, z), evaluate the surface integral M
F · dS.
� � In the short sequel, we give a geometric interpretation of a surface integral

M
F · dS. Suppose a fluid is flowing through a surface M such that the fluid’s
direction and velocity at point p in S is given by vector F (p) ft/sec. The component
r × ∂ r normal to S at p is given by the dot product
of F (p) in the vector ∂∂x ∂y
� �
1 ∂r ∂r
F (p) · ∂ r ∂ r × ft/sec
� ∂x × ∂y � ∂x ∂y
r × ∂ r �dxdy ft2 is the
We partition M into rectangular-like subregions where � ∂∂x ∂y
area of a subregion. When we multiply the above dot product to the area of a
subregion, we obtain
� �
∂r ∂r
F (p) · × dxdy ft3 /sec
∂x ∂y
that describes the volume of fluid flowing through the subregion per second. Sum-
ming up and taking the limit� � as the norm of the partition approaches zero, we
obtain the surface integral M
F · dS that is the volume of fluid flowing through
surface M per second.
Surface Integrals of Real-Valued Functions
Definition 7 The Integral of a Real-Valued Function Defined on a Surface
Let f be a real-valued function on a surface M . Suppose M is parametrized by a

function r, as in Definition 5, such that f ◦ r is continuous. We denote the integral
��
of f on M by M
f dS, and define it by
� � � � � �
� ∂r ∂r �
f dS = �
f (r(x, y)) � × � dA (22)
M R ∂x ∂y �
where R is an elementary region in the xy-plane on which r is defined.
The surface integral of a real-valued function f on M is independent of the

parametrization r of M . That is, whether r is orientation preserving or reversing,
the integral (22) is invariant. See Exercise 25 for a proof.
In particular, if f is a nonnegative function, the surface integral of f describes
the volume of a solid. Moreover, the base of the solid is the surface M and the
height of the solid at point r(x, y) is f (r(x, y)). The integrand in (22) represents
� �
� �
the volume of a rectangular box whose base has area � ∂ r × ∂ r � dA and the height
∂x ∂y
of the box is f (r(x, y)).

Let M be the rectangular region in the plane z = 1 − y where 0 ≤ x, y ≤ 1.
Let f (x, y, z) = x be a real-valued function that is defined on M .
��
f dS.
Solution To parametrize M where z = 1 − y, we let
r(x, y) = (x, y, 1 − y)
where (x, y) lies in a rectangular region R in the xy-plane defined by 0 ≤ x, y ≤ 1.

For the cross product of the partial derivatives of r, we find
� �
� i j k ��
�
∂r ∂r � �
× = �� 1 0 0 ��
∂x ∂y � �
� 0 1 −1 �
= (0, 1, 1) .
Using the given function f , we obtain
f (r(x, y)) = f (x, y, 1 − y) = x.
Applying Definition 7, the surface integral satisfies

� � � � � �
� ∂r ∂r �
f dS = f (r(x, y)) �� × � dA
M R ∂x ∂y �
Figure 10 � 1 � 1 √
The base of a solid is = x 2 dxdy
the surface M in the 0 0
plane z = 1 − y.
√
2
= .
2
��
The surface integral M
f dS represents the volume of a solid in Figure 10. The
base of the solid is the surface M , and the height at (x, y, z) ∈ M of the solid is
f (x, y, z).
✷
Try This 6
Let M be the rectangular region in the plane z = 1 − y where 0 ≤ x, y ≤ 1, i.e., M

is the base of the solid in Figure 10. If f (x, y, z) = y + z is defined on M , evaluate
��
the surface integral M
f dS.
2.4 Check-It Out
In Exercises 1-3, let M be the triangular part of the plane x + y + z = 1 in the first octant.
Suppose M is oriented by unit vectors normal to M that point away from the origin.
1. Find the surface area of M .

��
2. If F (x, y, z) = (0, 0, y + z), evaluate the surface integral M
F · dS
��
3. If f (x, y, z) = y + z, then evaluate the surface integral M
f dS
1. The plane x + y + z = 4 in the first octant is parametrized by r(x, y) = (x, y, 4 − x − y)

where 0 ≤ x, y ≤ 4.
2. Let M be a surface parametrized by r(s, t) = (s, t, 1) where 0 ≤ s, t ≤ 1.

� � � 1� 1
If F is vector field on M , then F · dS = F (s, t, 1) · k dsdt
M 0 0
� � �2 � �2
� �
∂f ∂f
3. The surface area of z = f (x, y) is given by 1+ + dA
R ∂x ∂y
for some region R in the xy-plane.
� �
4. The surface area of M is given by �r(s, t)� dA where r is a function that
R
parametrizes M , and r is defined on some region R in the st-plane.
5. If F (x, y, z) = (x, y, z), and M is parametrized by r(x, y) = (x, y, 1 − x − y),

� � � �
then F · dS = (1 + x + y)dA for some region R in the xy-plane.
M R
6. If f (x, y, z) = x2 + y 2 + z, and M is parametrized by r(s, t) = (s, t, s2 + t2 ),

� � � � �
then f dS = (s2 + t2 ) 1 + 4s2 + 4t2 dA for some region R in the st-plane.
M R
7. If F is a vector field on a surface M defined by z = f (x, y), then

� � � � � �
∂f ∂f
F · dS = F (x, y, f (x, y)) · − , − , 1 dA for some region R in the xy-plane.
M R ∂x ∂y

In Exercises 1-6, find a parametrization of the indicated surface. Then find the area of the surface.
1. The quadrilateral region defined by 2x + y + z = 4 where 0 ≤ x ≤ 1 and 0 ≤ y ≤ 2.

√
3 2
2. The surface in the parabolic cylinder z = 2 x where 0 ≤ x ≤ 1 and 0 ≤ y ≤ x.
3. The elliptical-shaped surface x + y + z = 3 such that x2 + y 2 ≤ 1.
4. The dome-shaped surface x2 + y 2 + z 2 = 4 such that x2 + y 2 ≤ 1 and z > 0.
5. The surface consisting of points (x, y, z) satisfying x2 + y 2 = 1 and 0 ≤ z ≤ 1 − y.

�
6. The cone defined by z = x2 + y 2 where x2 + y 2 ≤ 16. The top is not included.
��
In Exercises 7-16, evaluate the surface integral M
F · dS for the indicated vector field F defined on
the oriented surface M . Assume M is oriented by unit vectors pointing upward.
7. Let F (x, y, z) = (y, 0, z), and let M be parametrized by r(t, θ) = (2t cos θ, 2t sin θ, 1) where
0 ≤ θ ≤ 2π and 0 ≤ t ≤ 1.
8. Let F (x, y, z) = (0, y + z, x), and let M be parametrized by r(t, θ) = (t cos θ, t sin θ, 1 − t sin θ)
where 0 ≤ θ ≤ 2π and 0 ≤ t ≤ 1.
9. Let F (x, y, z) = (y, −x, z), and let M be parametrized by r(ψ, θ) = (sin ψ cos θ, sin ψ sin θ, cos ψ)
where 0 ≤ ψ ≤ π and 0 ≤ θ ≤ 2π. Note, M is the unit sphere.
� � � � �
10. Let F (x, y, z) = −x/z, −y/z, 1/z 2 , and let M be parametrized by r(x, y) = x, y, 1 + x2 + y 2
where 0 ≤ x ≤ a, 0 ≤ y ≤ b.
11. Let F (x, y, z) = (0, −x, 2y), and let M be the triangular planar region defined by x + y + 2z = 2 in the
first octant.
�
12. Let F (x, y, z) = (0, 0, 2), and let M be upper unit hemisphere z = 1 − x2 − y 2 .
13. Let F (x, y, z) = (2y − 2z, −1, 2x), and let M be the triangular planar region defined by x − y + z = 1
in the octant where x, z > 0 and y < 0.
14. Let F = (0, z, 0), and let M be the portion of the plane 2x + 2y + z = 4 in the first octant.
15. Let F = curl(0, 0, y), and let M be the portion of the sphere x2 + y 2 + z 2 = 25 in the first octant.
�
16. Let F = curl(0, x2 , 0), and let M be the hemisphere z = 4 − x2 − y 2 .
��
In Exercises 17-24, evaluate the surface integral M
f dS for the indicated real-valued function f
defined on the given surface M .
17. Let f (x, y, z) = 3z, and let M be the surface parametrized by r(θ, t) = (t cos θ, t sin θ, t)
√
where 0 ≤ t ≤ 2 and 0 ≤ θ ≤ 2π.
18. Let f (x, y, z) = x2 , and let M be the surface parametrized by r(θ, t) = (t cos θ, t sin θ, t + 1)
where 0 ≤ t ≤ 2 and 0 ≤ θ ≤ 2π.
�
19. Let f (x, y, z) = x2 + y 2 , and let M be the surface parametrized by r(x, y) = (x, y, x2 + y 2 )
such that x2 + y 2 ≤ 1.
�
20. Let f (x, y, z) = z x2 + y 2 , and let M be the disk of radius 2 defined by x2 + y 2 ≤ 4 and z = 3.
21. Let f (x, y, z) = 2x + y + z, and let M be the plane 2x + y + z = 4 in the first octant.
22. Let f (x, y, z) = 2y, and let M be the plane x + y + z = 3 in the first octant.
23. Let f (x, y, z) = 6xy, and let M be the triangular region with vertices (1, 0, 0), (0, 2, 0), and (0, 0, 2).
24. Let f (x, y, z) = 6z, and let M be the triangular region with vertices (2, 0, 0), (0, 0, 4), and (1, 1, 1).
The Effects of Re-parametrizations on Surface Integrals
25. Let r1 , r2 be two functions that parametrize a surface M ⊆ R3 , as in Definition 5. Suppose r1 , r2 are
defined on elementary regions R1 , R2 ⊆ R2 , respectively. To each interior point (s, t) in R2 , choose an
interior point (u, v) in R1 such that
r1 (u, v) = r2 (s, t).
a) Prove there exists a function h that is one-to-one and differentiable in the interior of R2 such that
(u, v) = h(s, t). We sketch a proof below.
r2 ��
Let r2 (s0 , t0 ) = r1 (u0 , v0 ) = (x0 , y0 , z0 ) ∈ M . Since the cross product ∂∂s r2 ��
× ∂∂t is
(s 0 ,t0 ) (s 0 ,t0 )
nonzero, without loss of generality we assume the third component of the cross product is nonzero,
 ∂x2 ∂x2 
 = ∂x2 ∂y2 − ∂x2 ∂y2 �= 0
∂s ∂t
Det 
∂y2 ∂y2 ∂s ∂t ∂t ∂s
∂s ∂t
where r2 (s, t) = (x2 , y2 , z2 ). We apply the Inverse Function Theorem to the system of equations:
�
x2 (s, t) = x
y2 (s, t) = y
Since the above determinant is nonzero, it is possible to express (s, t) as a differentiable function
of (x, y) near (x0 , y0 ). That is, (s, t) = f (x, y) for some invertible function f defined on an open
disk containing (x0 , y0 ) with continuous partial derivatives.
r1 ��
Similarly, since ∂∂u ×
�
∂ r1 �
is nonzero, we write (u, v) = g(x, y), (u, v) = g(x, z), or
(u ,v )
0 0 ∂v (u ,v )
0 0
(u, v) = g(y, z) for some function g with continuous partial derivatives in an open disk containing
(x0 , y0 ), (x0 , z0 ), or (y0 , z0 ), respectively. In any case, we obtain (s, t) = f ◦ g −1 (u, v).
b) Apply the Chain Rule to verify

� ��
∂r2 ∂r2 ∂x ∂y ∂x ∂y ∂r1 ∂r1
× = − ×
∂s ∂t ∂s ∂t ∂t ∂s ∂x ∂y
� �
∂(x, y) ∂r1 ∂r1
= ×
∂(s, t) ∂x ∂y
∂(x,y)
where is the Jacobian determinant of the transformation from (s, t) ∈ R1 to (x, y) ∈ R2 .
∂(s,t)
r2 × ∂ r2
Thus, the tangent plane to the surface M is well-defined since the cross-products ∂∂s ∂t
r1 × ∂ r1 are parallel.
and ∂∂x ∂y
c) Let F be a continuous vector field defined on M and with values in R3 .

Applying part a) and the change of variables theorem, we find
� � � � � � � � ��
∂r2 ∂r2 ∂(x, y) ∂r1 ∂r1 � ∂(s, t) �
F (r2 (s, t)) · × dsdt = F (r1 (x, y)) · × � �
∂s ∂t ∂(s, t) ∂x ∂y � ∂(x, y) � dxdy
R2 R1
� � � � � �
∂r1 ∂r1 ∂(x, y) � ∂(s, t) �
= F (r1 (x, y)) · × � �
∂x ∂y ∂(s, t) � ∂(x, y) � dxdy
R1
� � � �
∂r1 ∂r1
= ± F (r1 (x, y)) · × dxdy
R1 ∂x ∂y
where the sign is positive if the Jacobian determinant ∂(x,y)

∂(s,t) is positive, and
the sign is negative if the Jacobian determinant is negative.
The Jacobian determinant is positive when r1 and r2 are orientation preserv-
ing. Then the integral of F on surface M is independent of the orientation-
preserving parametrization r of M .
d) Applying the change of variable theorem, we find

� � � � � � � � ��
� ∂r2 ∂r2 � � ∂(x, y) ∂r1 ∂r1 � � ∂(s, t) �
f (r2 (s, t)) �
� ∂s × � dsdt = f (r1 (x, y)) � × � � �
� ∂(x, y) � dxdy
R2 ∂t � R1
� ∂(s, t) ∂x ∂y �
� � ��
� ∂r1 ∂r1 �
= f (r1 (x, y)) �
� ∂x × � dxdy.
R1 ∂y �
�In� particular, the integral of the scalar-valued function f over M , namely,

M
f dS, is independent of the parametrization r of M . Moreover, the path
integral is independent of the orientation of r, i.e., the orientation may be
preserving or reversing.
2.5 Integral Theorems of Green, Stokes, and Gauss

• Stokes’ Theorem • Stokes’ Theorem Applied to Surfaces that are Graphs
of Functions • Green’s Theorem • Gauss’ Divergence Theorem • Proofs
of the Integral Theorems
Stokes’ Theorem
Stokes’ Theorem relates a surface integral to a line integral along the boundary of
the surface. In the short sequel, we develop the relationship.
Let M be an oriented surface parametrized by a function r satisfying Definition 5
in page 111. Consequently, there are unit vectors N (p) that are normal to M at
p such that N (p) varies continuously as a function of p. Furthermore, suppose the
boundary ∂M of M is a simple closed curve. We define a positive orientation for
Figure 1 the curve ∂M using the orientation of M , see Figure 1. That is, using your right
An oriented surface M , and hand, let your thumb point to the direction of the unit vectors N (p), and the
its boundary ∂M is oriented
positively. direction along the curve ∂M as you curl your fingers is the positive orientation
for ∂M .
In words, Stokes’ Theorem states that the line integral of a differentiable vector
field F in 3-space along a positively oriented boundary ∂M is the equal to the
surface integral of the curlF on the surface M .
Theorem 2.12 Stokes’ Theorem
Let M be an oriented surface such that its boundary ∂M is a simple closed curve.
Suppose ∂M has the positive orientation. Let F be a vector field defined on M
with values in R3 . If F has continuous partial derivatives, then
� � �
F · dr = curlF · dS.
∂M M
z Example 1 Illustrating Stokes’ Theorem
2 N Verify Stokes’ Theorem if F (x, y, z) = (y, −z, −x), M is the plane x + y + z = 2

in the first octant, and M is oriented by unit vectors that point away from the
y
origin. Notice, the boundary of M is oriented counter-clockwise, see Figure 2.
2 2
Solution Let r 1 parametrize the straight path C1 from (2, 0, 0) to (0, 2, 0).
x
Figure 2 r 1 (t) = (2 − t, t, 0), 0 ≤ t ≤ 2.
The boundary is oriented
positively.
Likewise, let r 2 parametrize the straight path C2 from (0, 2, 0) to (0, 0, 2).
r 2 (t) = (0, 2 − t, t), 0 ≤ t ≤ 2.
Similarly, let r 3 parametrize the straight path from (0, 0, 2) to (2, 0, 0).
r 3 (t) = (t, 0, 2 − t), 0 ≤ t ≤ 2.
2.5. INTEGRAL THEOREMS OF GREEN, STOKES, AND GAUSS 127
The boundary of M is the union of the oriented lines, i.e., ∂M = C1 ∪ C2 ∪ C3 .

Then the line integral of F along ∂M satisfies
� � � �
F · dr = F · dr + F · dr + F · dr (23)
∂M C1 C2 C3
Since F (x, y, z) = (y, −z, −x), the line integral along r 1 is given by
� � 2
F · dr = F (r 1 (t)) · r �1 (t)dt
C1 0
� 2
= (t, 0, t − 2) · (−1, 1, 0)dt
0
� 2
= − tdt = −2.
0
Similarly, the line integral along r 2 satisfies

� � 2
F · dr = F (r 2 (t)) · r �2 (t)dt
C2 0
� 2 � 2
= (2 − t, −t, 0) · (0, −1, 1)dt = tdt = 2.
0 0
�
Likewise, we find C3
F · dr = 2. Applying (24), the sum of the line integrals equals
�
F · dr = 2. (24)
∂M
Since the plane is z = 2 − x − y, we parametrize M by

r(x, y) = (x, y, 2 − x − y)
where (x, y) lies in the xy-plane directly below M , see Figure 2a. Then
� �
� i j k ��
�
∂r ∂r � �
× = �� 1 0 −1 �� = (1, 1, 1).
∂x ∂y � �
� 0 1 −1 �
The curl of F satisfies y
� �
� i j k �
� � 2
� ∂ �
curlF = ∇ × F = �� ∂x ∂
∂y
∂ �
∂z �
= (1, 1, −1). x�2�y
� �
� y −z −x �
R
Using the region of integration R in Figure 2a, the surface integral is given by
� � � 2 � 2−y x
2
curlF · dS = (1, −1, −1) · (1, 1, 1)dxdy
M 0 0 Figure 2a
� 2� 2−y
The region R in the
= dxdy = (Area of R) = 2. xy-plane below M .
0 0
Hence, the surface integral agrees with the line integral (24), thereby, verifying
Stokes’ Theorem.
✷
z
z�2 N
Try This 1
Let M be a circular region of radius 1 in the plane z = 2, and centered at (0, 0, 2).
y
1 Assume M is oriented by unit vectors N that point upward, as seen in
x Figure 2b. Verify Stokes’ Theorem for the vector field F (x, y, z) = (−y/2, x/2, 0).
Figure 2b
A unit disk centered at
(0, 0, 2), and oriented by
vector N = k.
Stokes’ Theorem Applied to Surfaces that are Graphs of Functions
Let M be a surface that is the graph of a function, i.e., M is parametrized by
r(x, y) = (x, y, f (x, y)) (25)
where f is a function on an elementary region R in the xy-plane with continuous

partial derivatives. Suppose the boundary ∂M of M is a simple closed curve.
Notice, M is oriented by unit vectors pointing away from the origin. Then the
positive orientation on ∂M is the counter clockwise direction in the xy-plane.
In Section 2.4, we have seen
 
i j k
∂r ∂r   � ∂f ∂f
�
 ∂f 
× = Det  1 0 ∂x  = − ,− ,1
∂x ∂y   ∂x ∂y
∂f
0 1 ∂y
Let F (x, y, z) be a vector field defined on M and with values in R3 . Suppose

curlF (r(x, y)) = (L, N, P ) where L, N, P are functions of (x, y).
Then the surface integral of curlF on M satisfies
� � � � � �
∂r ∂r
curlF · dS = curlF (r(x, y)) · × dA
M R ∂x ∂y
� � � �
∂f ∂f
= −L −N + P dA (26)
R ∂x ∂y
where dA = dxdy or dA = dydx. Thus, if M is a surface parametrized by (25),

∂M is a simple closed curve oriented in the counter clockwise motion, and
curlF = (L, N, P ), then we may rewrite Stokes’ Theorem as follows:
� � �
F · ds = curlF · dS
∂M M
� � � �
∂f ∂f
= −L −N +P dA (27)
R ∂x ∂y
z
Example 2 Surfaces Defined by Functions
N
2 2
Let C be the intersection of the surfaces x + y = 1 and z = 3 − x − y, as in
Figure 3. Assume curve C is oriented by the counterclockwise direction on the
xy-plane. Apply Stokes’ Theorem in evaluating the line integral
� 1 y
y 3 dx + x3 dy + (y 3 + z)dz. x
C
Figure 3
The region R in the
Solution
xy-plane below M .
The vector field in the line integral is
� �
F (x, y, z) = y 3 , x3 , y 3 + z .
We evaluate the curl of F :
� �
� i j �k
� �
� ∂ � � �
curlF = ∇ × F = �� ∂x ∂
∂y
∂
∂z
� = 3y 2 , 0, 3x2 − 3y 2 .
�
� �
� y3 x3 y3 + z �
Notice, the surface M that is enclosed by C is the graph of f (x, y) = 3 − x − y
where x2 + y 2 ≤ 1. If (L, N, P ) = curlF , then
� �
∂f ∂f � �
−L −N + P = 3y 2 − 0 + (3x2 − 3y 2 ) = 3x2 .
∂x ∂y
In applying Stokes’ Theorem, we use the special case (27).
� � �
3 3 3
y dx + x dy + (y + z)dz = curlF · dS
C M
� � � �
∂f ∂f
= −L −N +P dA
R ∂x ∂y
� �
= 3x2 dA.
R
The region of integration R is the unit disk of radius 1 centered at the origin.
Using a change of variables to polar coordinates with x = r cos θ and dA = rdrdθ,
we obtain
� � � 2π � 1
3x2 dA = 3r3 cos2 θ drdθ
R 0 0
� 2π
3
= cos2 θ dθ
4 0
3π
= .
4
Hence, by Stokes’ Theorem we obtain
�
3π
y 3 dx + x3 dy + (y 3 + z)dz = .
C 4
✷
z
N
Try This 2
Let M be an oriented rectangular region with vertices (0, 0, 1), (1, 0, 1), (1, 2, 0),
1 2 y and (0, 2, 0), see Figure 4. Let the boundary ∂M of M be oriented positively.
x �
Then apply Stokes’ Theorem in evaluating the line integral ∂M (xzdx + ydz).
Figure 4
A 2 by 1 rectangular region
and a normal vector N .
Green’s Theorem
We analyze a special case of Stokes’ Theorem. In the parametrization (25), let
f (x, y) = 0. In particular, each point in M has the form (x, y, 0) where the set of
of (x, y)’s lie in an elementary region R in the xy-plane, see Figure 5.
We assume M is oriented by the unit vector k, and the boundary ∂M is a simple
closed curve oriented with the counter clockwise direction.
Moreover, we consider vector fields of the form F (x, y, 0) = (f1 (x, y), f2 (x, y), 0)
where f1 , f2 are real-valued functions defined on R. Then the curl is given by
� �
� i j k �
� �
� �
curlF = ∇ × F = � � ∂ ∂ ∂ �
∂x ∂y ∂z �
� �
� f (x, y) f (x, y) 0 �
1 2
� �
∂f2 ∂f1
z = 0, 0, − .
∂x ∂y
Applying a special case of Stokes’ Theorem, i.e., (27), we obtain
k
� � � � �
∂f2 ∂f1
(f1 (x, y), f2 (x, y), 0) · ds = − dA.
y ∂M R ∂x ∂y
M
x Since z = 0, the vector field F may be realized as having values in R2 .
Then Stokes’ Theorem reduces to the following:
Figure 5
A surface M in the xy-plane
oriented by the unit vector k.
Theorem 2.13 Green’s Theorem
Let M be an elementary region in the xy-plane whose boundary ∂M is a simple

closed curve, oriented by the counter clockwise motion. If f1 , f2 are real-valued
functions defined on M with continuous partial derivatives, then
� � � � �
∂f2 ∂f1
(f1 (x, y)dx + f2 (x, y)dy) = − dA.
∂M M ∂x ∂y
�
Notice, the line integral ∂M
(f1 (x, y)dx + f2 (x, y)dy) reduces to the area of M
∂f2 ∂f1
if ∂x − ∂y = 1 for all (x, y) ∈ M .
y
Example 3 Applying Green’s Theorem
y�2x
� 2
Apply Green’s Theorem in evaluating the line integral y 2 dx + 3xydy where C is
C
a triangular path from the origin to point (1, 0) to point (1, 2) and to the origin. M
x
1
Solution Let M be the triangular region enclosed by C, see Figure 6. Notice,
C = ∂M , i.e, the boundary of M , is oriented in the counter clockwise direction. Figure 6
Applying Green’s Theorem, we obtain Triangular region with
� � hypotenuse y = 2x.
� � �
∂(3xy) ∂(y 2 )
y 2 dx + 3xydy = − dA
C M ∂x ∂y
� �
= (3y − 2y) dA
M
� 1 � 2x
= ydydx
0 0
� 1 �y=2x
1 2 ��
= y dx
0 2 �y=0
� 1
= 2x2 dx
0
� y
2 2
y dx + 3xydy = .
C 3
✷
1
M y�x�2
Try This 3 x
2
�
Apply Green’s Theorem in evaluating the line integral sin x dx + 3xy 2 dy where Figure 7
C Triangular path with
C is a triangular path from (0, 0) to (2, 1) to (0, 1) to (0, 0). See Figure 7. hypotenuse y = x/2.
y
Example 4 Illustrating Green’s Theorem
� �
Verify Green’s Theorem if F (x, y) = − 13 y 3 , 13 x3 and M is a circular region of M
x
radius 1 and centered at the origin. Suppose the unit circle ∂M , i.e., boundary 1
of M , is oriented in the counter clockwise direction. See Figure 8.
Solution The unit circle ∂M is parametrized in the counter clockwise motion by

Figure 8
r(t) = (cos t, sin t) Counter clockwise
circular path
where 0 ≤ t ≤ 2π. We evaluate the line integral of F along ∂M as follows:
� � 2π
F · dr = F (r(t)) · r � (t) dt
∂M 0
� 2π
1 � �
= − sin3 t, cos3 t · (− sin t, cos t) dt
3 0
� 2π
1 � �
= cos4 t + sin4 t dt
3 0
� �� 2 � �2 �
2π
1 1 + cos 2t 1 − cos 2t
= + dt
3 0 2 2
� 2π � �
1 2 + 2 cos2 2t
= dt
3 0 4
� 2π
1
= (3 + cos 4t) dt
12 0
� �t=2π
1 sin 4t ��
= 3t +
12 4 �t=0
�
π
F · dr = .
∂M 2
To verify Green’s Theorem, we evaluate a double integral as in Theorem 2.13.

We apply a change of variables by changing from Cartesian to polar coordinates.
Notice, M is the circular region centered at the origin and radius 1.
� � � � � � ��
∂ 1 3 ∂ 1 3 � 2 �
x − − y dA = x + y 2 dA
M ∂x 3 ∂y 3 M
� 2π � 1
= r3 drdθ
0 0
� 2π
1 π
= dθ = .
0 4 2
since x2 + y 2 = r2 and dA = rdrdθ. Then the double integral and line integral
have the same values. This completes the verification of Green’s Theorem.
✷
Try This 4
� �
Verify Green’s Theorem if F (x, y) = − y2 , x2 and M is the circular region of
radius 1 and centered at (1, 1). Assume circle ∂M , i.e., the boundary of M
is oriented in the counter clockwise direction.
Gauss’ Divergence Theorem

We consider a solid M in R3 whose boundary ∂M consists of finitely many surfaces
that are oriented by unit vectors pointing outward from the solid. An example
of such a solid is a cube, and its boundary consists of six planar faces. Another
example of a solid is the ball x2 + y 2 + z 2 ≤ r2 , and its boundary is the sphere z
x2 + y 2 + z 2 = r2 where r > 0. See Figures 9 and 10. N
Theorem 2.14 Gauss’ Divergence Theorem N
Let M be a solid region in R3 such that its boundary ∂M consists of finitely many
surfaces oriented by unit vectors pointing outward from M . Let F be a vector field
y
defined on M and with values in R3 . If F has continuous partial derivatives, then
x
� � � � � Figure 9
F · dS = div(F ) dV. A cube oriented by unit
∂M M vectors pointing outward.
Example 5 Illustrating Gauss’ Divergence Theorem
Let M be the ball x2 + y 2 + z 2 ≤ r2 of radius r > 0. Verify the Divergence Theorem

for F (x, y, z) = (0, 0, z). Assume the boundary ∂M of M is oriented by unit vectors
pointing outward from the ball, see Figure 10.
Solution We parametrize ∂M using spherical coordinates, i.e.,

r(φ, θ) = (r sin φ cos θ, r sin φ sin θ, r cos φ)
where 0 ≤ φ ≤ π and 0 ≤ θ < 2π, see Section 2.2. Furthermore, we find
� �
� i j k �
� �
∂r ∂r � � Figure 10
× = � r cos φ cos θ r cos φ sin θ −r sin φ ��
∂φ ∂θ � Sphere of radius r and
� �
� −r sin φ sin θ r sin φ cos θ 0 � outward pointing vectors.
� �
= r2 sin2 φ cos θ, r2 sin2 φ sin θ, r2 sin φ cos φ
Notice, the above cross product is a vector that points away from the ball.
Applying Definition 6 in Section 2.4, the surface integral of F on ∂M satisfies
� � � π � 2π � �
∂r ∂r
F · dS = F (r(φ, θ)) · × dθdφ Ρ
∂M 0 0 ∂φ ∂θ
r
� π � 2π � �
∂r ∂r
= (0, 0, r cos φ) · × dθdφ
∂φ ∂θ Π
0 0
Φ Θ
� π � 2π
= r3 cos2 φ sin φ dθdφ 2Π
0 0 Figure 11
� π Rectangular solid T
3
= 2πr cos2 φ sin φ dφ
0
� �
4πr3
F · dS = .
∂M 3
Next, we evaluate the triple integral as in Theorem 2.14. Since div(F ) = 1, we find
� � � � � �
div(F ) dV = dV.
M M
We apply a change of variables and use the spherical coordinates x = ρ sin φ cos θ,
y = ρ sin φ sin θ, and z = ρ cos φ where 0 ≤ ρ ≤ r, 0 ≤ φ ≤ π, and 0 ≤ θ < 2π.
The ball M is mapped into a rectangular solid T , see Figure 11, by the trans-
formation that sends (x, y, z) ∈ M into (φ, θ, ρ) ∈ T . In page 83, the Jacobian
determinant of the transformation is given by
 ∂x ∂x ∂x 
∂ρ ∂φ ∂θ
 
∂(x, y, z)  ∂y ∂y ∂y 
= Det 
 ∂ρ ∂φ ∂θ  = ρ2 sin φ.

∂(ρ, φ, θ)  
∂z ∂z ∂z
∂ρ ∂φ ∂θ
Applying the change of variables, we obtain

� � � � � � � �
� ∂(x, y, z) �
dV = � �
� ∂(ρ, φ, θ) � dρdφdθ
M T
� 2π � π � r
= ρ2 sin φ dρdφdθ
0 0 0
� π
2πr3 4πr3
= sin φ dφ =
3 0 3
Hence, in the case of Example 5, we have verified
� � � �
F · dS = div(F ) dV.
∂M M
Try This 5
Let M be the cube in the first octant bounded by x = s, y = s, z = s where s > 0

as in Figure 9. Let ∂M be oriented by unit vectors that point away from M . If
��
F (x, y, z) = (2x, 0, 0), evaluate ∂M
F · dS by Gauss’ Divergence Theorem.
Example 6 Applying Gauss’ Divergence Theorem
Let M be the solid circular cylinder bounded by x2 + y 2 = 4, z = 0 and z = 1.

Let the boundary ∂M of M be � oriented by� unit vectors that point away from M ,
see Figure 12. If F (x, y, z) = x3 , 2x2 y, x2 z , apply Gauss’ Divergence Theorem in
evaluating the surface integral
� �
F · dS.
∂M
Solution The divergence of F is given by

� � ∂ 3 ∂ ∂ 2
div x3 , 2x2 y, x2 z = (x ) + (2x2 y) + (x z)
∂x ∂y ∂z
= 3x2 + 2x2 + x2 = 6x2 .
Applying the Gauss’ Divergence Theorem, we rewrite the surface integral.

� � � � � � � �
Figure 12
F · dS = div(F ) dV = 6x2 dV. A disk oriented by unit
∂M M M
vectors pointing outward.
To evaluate the above triple integral, we use cylindrical coordinates
x = r cos θ, y = r sin θ, z = z
as discussed in Section 2.2. Notice, 0 ≤ r ≤ 2, 0 ≤ θ ≤ 2π, and 0 ≤ z ≤ 1.

The Jacobian determinant of the transformation from cartesian to cylindrical
coordinates is given by
 ∂x ∂x ∂x
  
∂r ∂θ ∂z cos θ −r sin θ 0
   
∂(x, y, z)  ∂y ∂y ∂y   sin θ r cos θ 0 
= Det 
 ∂r ∂θ ∂z =
  =r
∂(r, θ, z)   
∂z ∂z ∂z 0 0 1
∂r ∂θ ∂z
Consequently, we obtain
� � � � 1 � 2π � 2 � �
� ∂(x, y, z) �
6x2 dV = 6(r cos θ)2 �� drdθdz
M 0 0 0 ∂(r, θ, z) �
� 1� 2π � 2
= 6r3 cos2 θ drdθdz
0 0 0
� 1� 2π �r=2
3r4 ��
= cos2 θ dθdz
0 0 2 �r=0
� 1 � 2π
= 24 cos2 θ dθdz
0 0
� 1 � 2π
= 12 (1 + cos 2θ) dθdz
0 0
� 1 � �θ=2π
1 �
= 12 θ + sin 2θ �� dz
0 2 θ=0
� 1
= 12 2π dz
0
� �
6x2 dV = 24π.
M
Hence, we have � �
F · dS = 24π.
∂M
✷
Try This 6
Let M be the solid bounded by x2 + y 2 = 1, z = −2 and z = 2. Suppose the

boundary
� 3 � ∂M is oriented by unit vectors pointing away from M . If F (x, y, z) =
z, y , x , apply Gauss’ Divergence Theorem in evaluating the surface integral
� �
F · dS.
∂M
Proof of Green’s Theorem

y
We prove Green’s Theorem for the case when M ⊂ R2 is an elementary region
that is both of type Rx and Ry , see page 61. That is, we may express M as
C2
M = {(x, y) : a ≤ x ≤ b, g(x) ≤ y ≤ f (x)} Type Rx
M
= {(c, d) : c ≤ y ≤ d, m(y) ≤ x ≤ n(y)} Type Ry
C1
x where f, g and m, n are continuous real-valued functions on [a, b] and [c, d],
a b
respectively. Assume the boundary ∂M is a simple closed curve oriented by
Figure 13a
g(x) ≤ f (x) for a ≤ x ≤ b the counter clockwise motion. In Figure 13a, let ∂M = C1 ∪ C2 where C1 and
−C2 are parametrized by
r 1 (x, y) = (x, g(x))
y
r ∗2 (x, y) = (x, f (x))
and a ≤ x ≤ b. If f1 is a real-valued function defined on M , the line integral of
d C3 the vector field (f1 , 0) along ∂M satisfies
M � � �
C4
c f1 dx = f1 dx − f1 dx
∂M C1 −C2
x
� b � b
= f1 (x, g(x))dx − f1 (x, f (x))dx
Figure 13b a a
m(y) ≤ n(y) for c ≤ y ≤ d
� b � f (x)
∂f1
= − dydx
a g(x) ∂y
� � �
∂f1
f1 dx = − dA.
∂M M ∂y
Similarly, if f2 is defined on M , the line integral of (0, f2 ) along ∂M satisfies
� � �
∂f2
f2 dy = dA.
∂M M ∂x
Adding the last two line integrals, we obtain

� � � � �
∂f2 ∂f1
f1 dx + f2 dy = − dA.
∂M M ∂x ∂y
This proves Green’s Theorem when M is an elementary region that is of both
type Rx and Ry .
Proof of Stokes’ Theorem

We prove Stokes’ Theorem for surfaces M that are graphs of functions. Following
(25), we parametrize M by
r(x, y) = (x, y, f (x, y))
where f is a function with continuous partial derivatives defined on an elementary

region R in the xy-plane . In Section 2.4, we proved
� �
∂r ∂r ∂f ∂f
× = − ,− ,1 . (28)
∂x ∂y ∂x ∂y
Notice, M is oriented by unit vectors pointing away from the origin. Assume the
boundary ∂M of M is a simple closed curve. Then the positive orientation on ∂M
is the counter clockwise direction in the xy-plane.
Let F be a vector field defined on M such that the values of F lie in R3 . We write
F = (P, Q, R) where P, Q, R are real-valued functions with continuous partial
derivatives defined on R. Then the curl is given by
� � � � � �
∂R ∂Q ∂R ∂P ∂Q ∂P
curlF = − i− − j+ − k (29)
∂y ∂z ∂x ∂z ∂x ∂y
see (30) in Section 1.3.

On the other hand, the boundary ∂M of M is parametrized by
c(t) = (x(t), y(t), f (x(t), y(t))
where x(t), y(t) are certain real-valued function defined on [a, b]. Then the line
integral of F along the boundary ∂M satisfies
� � b
F · dr = (F ◦ c)(t) · c� (t)dt
∂M a
 � � 
� b � �
� � ∂f � � ∂f �
= 
((P, Q, R) ◦ c)(t) · x (t), y (t), � x (t) + � y � (t) dt
a ∂x � ∂y �
c(t) c(t)
where the chain rule is used in the last line. Evaluating the dot product, we find
 � 
� � b �
F · dr = P (c(t))x� (t) + R(c(t)) ∂f �� x� (t) dt +
∂M a ∂x �
c(t)
 � 
� b �
Q(c(t))y � (t) + R(c(t)) ∂f �� y � (t) dt
a ∂y �
c(t)
� � � � �
∂f ∂f
= P +R dx + Q + R dy
∂M ∂x ∂y
where we used the definition of the line integral in the previous line.
By Green’s Theorem, chain rule, and the product rule, we obtain
� � � � � � � ��
∂ ∂f ∂ ∂f
F · dr = Q+R − P +R dA
∂M R ∂x ∂y ∂y ∂x
� � � �
∂Q ∂Q ∂f ∂R ∂f ∂R ∂f ∂f ∂2f
= + + + +R dA
R ∂x ∂z ∂x ∂x ∂y ∂z ∂x ∂y ∂x∂y
� � � �
∂P ∂P ∂f ∂R ∂f ∂R ∂f ∂f ∂2f
− + + + +R dA
R ∂y ∂z ∂y ∂y ∂x ∂z ∂y ∂x ∂y∂x
� � � �
∂Q ∂Q ∂f ∂R ∂f
= + + dA
R ∂x ∂z ∂x ∂x ∂y
� � � �
∂P ∂P ∂f ∂R ∂f
− + + dA.
R ∂y ∂z ∂y ∂y ∂x
∂f ∂f
Grouping and factoring ∂x and ∂y , we find
� � � ��
∂R ∂Q ∂f ∂P ∂R ∂f
F · dr = − − + − − +
∂M R ∂y ∂z ∂x ∂z ∂x ∂y
� ��
∂Q ∂P
− dA
∂x ∂y
� � � � � �
∂R ∂Q ∂P ∂R ∂Q ∂P ∂f ∂f
= − , − , − · − , − , 1 dA
R ∂y ∂z ∂z ∂x ∂x ∂y ∂x ∂y
� � � �
∂r ∂r
= (curlF ) · × dA
R ∂x ∂y
where in the last line we used the curl of F in (29), and the cross product in (28).
Note, the right side of the last line is the definition of the surface integral of curlF
on surface M . Finally, we obtain
� � �
F · dr = curlF · dS.
∂M M
This completes the proof of Stokes’ Theorem for the case when the surface M is
the graph of a function z = f (x, y).
Proof of Gauss’ Divergence Theorem

Let M be an elementary solid in 3-space, see page 66 in Section 2.1. Assume M
is oriented by unit vectors pointing away from M . Let F = (P, Q, R) be a vector
field where P, Q, R are real-valued functions on M . We express the surface
integral of F on the boundary ∂M of M as a sum:
� � � � � � � �
F · dS = (P i) · dS + (Qj) · dS + (Rk) · dS.
∂M ∂M ∂M ∂M
Using the definition of the divergence, we obtain

� � � � � � � � � � � �
∂P ∂Q ∂R
divF dV = dV + dV + dV.
M M ∂x M ∂y M ∂z
The Gauss’ Divergence Theorem will follow if the three statements below are true.
� � � � �
∂P
(P i) · dS = dV (30)
∂M M ∂x
� � � � �
∂Q
(Qj) · dS = dV (31)
∂M M ∂y
� � � � � Figure 14
∂R
(Rk) · dS = dV (32) Solid with three faces
∂M M ∂z F1 , F2 , F3 , and outward
pointing vectors.
We only prove (32), and in a special case. Suppose ∂M consists of three faces
Fi for i = 1, 2, 3 such that F3 is a lateral surface, and k is perpendicular to the
normal vectors to F3 , see Figure 14.
If r3 is an orientation preserving parametrization of F3 defined on D3 ⊆ R2 , then
� �
∂r3 ∂r3
k· × = 0.
∂x ∂y
Consequently, by the definition of the surface integral we obtain

� � � � � �
∂r3 ∂r3
Rk · dS = Rk · × dA = 0. (33)
F3 D3 ∂x ∂y
Also, we know the surface integral of Rk on ∂M is the sum of the surface integrals
on F1 , F2 , F3 . Combined with (33), we obtain
� � � � � �
Rk · dS = Rk · dS + Rk · dS. (34)
∂M F1 F2
Let D be an elementary region in the xy-plane, and let Φ1 , Φ2 be continuous real-

valued functions on D satisfying
M = {(x, y, z) : (x, y) ∈ D, Φ1 (x, y) ≤ z ≤ Φ2 (x, y)}.
For i = 1, 2, we parametrize Fi as follows:
ri (x, y) = (x, y, Φi (x, y)), (x, y) ∈ D.
Since F1 is oriented by upward pointing unit vectors, we have

� � � � � �
∂r1 ∂r1
Rk · dS = Rk · × dA
F1 D ∂x ∂y
� � � �
∂Φ1 ∂Φ1
= Rk · − ,− , 1 dA
D ∂x ∂y
� �
= R(x, y, Φ1 (x, y))dA.
D
We have oriented F2 by downward pointing unit vectors. Notice, the cross product
� �
∂r2 ∂r2 ∂Φ2 ∂Φ2
− × = , , −1
∂x ∂y ∂x ∂y
is downward pointing and normal to F2 . Then we obtain

� � � � � �
∂r2 ∂r2
Rk · dS = Rk · − × dA
F2 D ∂x ∂y
� �
= − R(x, y, Φ2 (x, y))dA.
D
Then we rewrite (34) as follows:

� � � � �z=Φ1 (x,y))
∂R(x, y, z) ��
Rk · dS = � dA
∂M D ∂z z=Φ2 (x,y))
� � �
∂R
= dV.
M ∂z
Thus, we have proved (32). The proofs of (30)-(31) are similar when we assume M
has a symmetry in its parametrizations:
M = {(x, y, z) : (x, z) ∈ R� , Φ�1 (x, z) ≤ y ≤ Φ�2 (x, z)}
= {(x, y, z) : (y, z) ∈ R�� , Φ��1 (y, z) ≤ x ≤ Φ��2 (y, z)}
where Φ�1 , Φ�2 and Φ��1 , Φ��2 are defined on subsets of the xz-plane and yz-plane,
respectively. This completes a sketch of the proof of Gauss’ Divergence Theorem.
Theorem 2.15 Path Independence and Conservative Vector Fields
Let F be a vector field from R3 into R3 with continuous partial derivatives

except at finitely many points. Then F is a conservative vector field if and
�
only if C F · ds = 0 for all simple closed curves C in R3 .
Proof Let M be an oriented surface in R3 such that C = ∂M has the positive

orientation. By definition, if F is a conservative vector field then curlF = 0, see
page 37. Applying Stokes’ Theorem, we obtain
� � �
F · dr = curlF · dS
C M
� �
= 0 · dS = 0.
M
�
To prove the converse, suppose C
F · dr = 0 for any simple closed curve C. Let
C1 and C2 be two paths from points P to Q such that C = C1 ∪ (−C2 ) is a simple
� � �
closed path. Since C F · dr = C1 F · dr − C2 F · dr, we find
� �
F · dr = F · dr.
C1 C2
That is, the line integral of F is independent of the path from P to Q. Let
f (x, y, z) denote the value of the line integral of F from (0, 0, 0) to point (x, y, z).
For simplicity, suppose x, y, z > 0. Let r1 (t1 ) = (t1 , 0, 0), r2 (t2 ) = (x, t2 , 0), and
r3 (t2 ) = (x, y, t3 ).
Notice, the union or the sum r = r1 + r2 + r3 is a path from the origin to
(x, y, z). If F = (F1 , F2 , F3 ), then
�
f (x, y, z) = F · dr
r
� � �
= F · dr + F · dr + F · dr
r1 r2 r3
� x � y � z
= F1 (t1 , 0, 0)dt1 + F2 (x, t2 , 0)dt2 + F3 (x, y, t3 )dt3 .
0 0 0
Applying the Fundamental Theorem of Calculus, we obtain

∂f
= F3 .
∂z
∂f ∂f
Similarly, by choosing different paths, we find ∂x = F1 and ∂y = F2 . Thus,
∇f = F . From the identity curl(∇f ) = 0, we find F is a conservative vector field.
✷
2.5 Check-It Out
�
1. Apply Stokes’ Theorem in evaluating the line integral F · dr where M is
∂M
the part of the plane 2x+y+z = 2 in the first octant, and F (x, y, z) = (z, y, 0).
�
2. Apply Green’s Theorem in evaluating the line integral F · dr. where C is
C
the boundary of the rectangular region 0 ≤ x ≤ 2, 0 ≤ y ≤ 1, and F = (ey , ex ).
3. Apply Gauss’ Divergence Theorem in evaluating the surface integral

� �
F · dS where F (x, y, z) = (x, −y, z), and M is the solid cube bounded
∂M
by the planes x = 1, y = 1, z = 1, and the coordinate planes.

1. The line integral of a vector field F along the boundary ∂M of surface M is equal
to the surface integral of the curlF on M .
2. The surface integral of a vector field F along the boundary ∂M of a solid M is equal
to the triple integral of divF on M .
3. Let R be a region that is enclosed by a simple closed curve C that is oriented in the counter
� � �
2
clockwise direction. If F (x, y) = (−x , xy), then F · dr = (y − 2x)dA.
C R
4. Let M be parametrized by r(x, y) = (x, y, x2 + y 2 ) where (x, y) lies in a region R. Assume

the boundary ∂M is oriented in the counter clockwise motion as seen from above the z-axis.
� � �
If curlF (x, y, z) = (x, −y, −1), then F · dr = (x, −y, −1) · (2x, 2y, 1)dA.
C R
5. Let F (x, y, z) = (x, y, z), and let M be the ball defined x2 + y 2 + z 2 ≤ r2 and oriented by
unit vectors that point away from the origin. Applying Gauss’ Divergence Theorem, we find
� �
F · dS is equal to the volume of M .
∂M
In Exercises 1-8, apply Stokes’ Theorem in evaluating a line integral of a vector field F along the
boundary ∂M of a surface M . Assume M is oriented by upward pointing unit normal vectors, and
∂M is oriented positively by the counter clockwise direction.
�
1. F · dr where F (x, y, z) = (0, z, x), M is the part of the surface z = y 2 in the first octant
∂M
above the rectangular region in the xy-plane with vertices at the origin, (2, 0, 0), (2, 1, 0) and (0, 1, 0).
�
2. (ez − 2y)dx + ey dy + ex dz where M is the part of the plane z = 2 that lies above the rectangular
∂M
region in the xy-plane satisfying 0 ≤ x ≤ 2, 0 ≤ y ≤ 3.
�
3. F · dr where F (x, y, z) = (y, z, −4x), and M is the part of the plane 2x + 3y + 3z = 6 in
∂M
the first octant.

�
4. F · dr where M is the plane 2x + 2y + z = 2 in the first octant, and F (x, y, z) = (−x, −z 2 , −y).
∂M
�
5. F · dr where F (x, y, z) = (0, ex , ey ), and M is the part of the plane x + y + z = 1 in the first octant.
∂M
�
6. �−xy, z − y, z 2 � · dr where M is the part of the plane z = 4 in the first octant that lies above
∂M
the interior of the quarter of the unit circle in the xy-plane centered at the origin.
� �
7. F · dr where M is the upper hemisphere z = 1 − x2 − y 2 and F (x, y, z) = (−x2 y, 0, 0).
∂M
� �
8. F · dr where F (x, y, z) = (1, y + x3 , z), and M is the part of the hemisphere z = 4 − x2 − y 2
∂M
that is bounded by the cylinder x2 + y 2 = 1.
In Exercises 9-16, apply Green’s Theorem in evaluating the line integral of F along the indicated curve C
oriented in the counter clockwise direction.
�
9. y 2 dx + 4xydy where C is a triangular path from the origin to point (1, 0) to point (1, 4)
C
and to the origin.
�
10. F · dr where F (x, y) = (x − y 2 , 2x + y), C is a triangular path from the origin to point (1, 1)
C
to point (0, 2) and to the origin.

�
� 3 �
11. −y dx + x3 dy where C is the circle of radius 2, and centered at the origin.
C
�
� �
12. x3 + y dy where C is the unit circle centered at the origin.
C
�
� � √
13. F · dr where F (x, y) = x2 , 20xy , and C is a closed curve formed by y = x and y = x2 .
C
�
14. F · dr where F (x, y) = (x − y, x + y), and C is a closed curve formed by y = 4 and y = x2 .
C
�
1 � 2 2�
15. F · dr where F (x, y) = y , x , C is the parallelogram with vertices
C 2
(0, 0), (2, 0), (3, 1), and (1, 1).
�
� �
16. F · dr where F (x, y) = x2 , ex−y , and C is a parallelogram with vertices
C
(0, 0), (1, 0), (2, 1), and (1, 1).

In Exercises 17-24, apply Gauss’ Divergence Theorem in evaluating the surface integral of F on the
boundary ∂M of a solid M that is oriented by unit vectors that point away from M .
� �
17. F · dS where F (x, y, z) = (x, y 2 , z 3 ) and M is the solid bounded by the planes x = 3, y = 2,
∂M
z = 1, and the coordinate planes.
� �
18. F · dS where F (x, y, z) = (z, y, x) and M is the solid bounded by the cylinder x2 + y 2 = 9,
∂M
the planes z = 0, and z = 2.
� �
19. F · dS where F (x, y, z) = (x, x2 y, y 2 z) and M is the solid bounded by the cylinder x2 + y 2 = 1,
∂M
the planes z = 1, and z = 2.
� �
20. F · dS where F (x, y, z) = (xz 2 , x2 z, xy 2 ), and M is the sphere x2 + y 2 + z 2 = 1.
∂M
� �
21. F · dS where F (x, y, z) = (x3 , y 3 , z 3 ), and M is the sphere x2 + y 2 + z 2 = r2 with r > 0.
∂M
� �
22. F · dS where F (x, y, z) = (y, x, z 2 ), and M is the solid bounded by the cylinder x2 + y 2 = 4,
∂M
�
the hemisphere z = 16 − x2 − y 2 , and the plane z = 0.
� �
23. F · dS where F (x, y, z) = (y, x2 , z 3 ), and M is the solid bounded
∂M
by the ellipsoid x2 + y 2 + 4z 2 = 4.
� � � �
y3
24. F · dS where F (x, y, z) = z, , x , and M is the solid bounded
∂M 18
by the ellipsoid x2 + 9y 2 + 4z 2 = 36.
25. Let r(x, y, z) = (x, y, z) be the position vector field, and let
r(x, y, z)
F (x, y, z) = .
�r(x, y, z)�3
If M is a solid that does not contain (0, 0, 0) and for which Gauss’ Divergence
� �
Theorem applies, verify F · dS = 0.
∂M
26. One of Gauss’ Several Lemmas Let F be the vector field in Exercise 25.
Let M be a sphere centered at (0, 0, 0), and of any radius ε > 0. Prove
� �
F · dS = 4π.
M
27. If F is the vector field in Exercise 25, show F �= curlG for any vector field G.
28. Let F be a vector field from R3 to R3 with continuous first partial

derivatives. If divF = 0, prove F = curlG for some vector field G.
Notice, F should be defined on all of R3 , see Exercise 27.
Chapter 2 Multiple Choice Test

� �
1. A unit tangent vector to the curve r(t) = t, t2 , −t2 at the origin is given by
A. i−k B. i+k C. j−k D. −i
2. The curvature to r(t) = (cos t, sin t, 2t) at point P (1, 0, 0) is equal to

√
1 5 √
A. B. C. 5 D. 5
5 5
3. Let C be a triangular path from the �origin to point (1, 0) to point (1, 2) and
to the origin. Then the line integral ydx + 3xdy equals
C
A. 1 B. 2 C. 4 D. 8
4. If u = x + y and v = x − y, the absolute value of the Jacobian determinant

∂(x,y)
∂(u,v) is equal to
1 1
A. 4 B. 2 C. D.
4 2
5. Let C be directed line segment from the origin to point (1, 3). Then the line
�
integral dx + dy equals
C
A. 2 B. 3 C. 4 D. 5
� 2 � 3
6. The iterated integral (2x + y 2 )dydx equals
0 0
A. 30 B. 21 C. 13 D. 7
� � √
4 x
7. The iterated integral f (x, y)dydx equals
0 x/2
� 4 � 2y � 2 � √
y
A. √
f (x, y)dxdy B. f (x, y)dxdy
0 y 0 2y
� 2 � 2y � 4 � y2
C. f (x, y)dxdy D. √
f (x, y)dxdy
0 y2 0 y
8. Let C be a rectangular path that joins the points from (0, 0) to (2, 0) to
(2, 1) to (0, 1), and to (0, 0). Applying Green’s Theorem, the line integral
�
(xy, x2 ) · dr is equal to
C
� 2 � 1 � 2 � 1
A. (2x + y)dydx B. (2x − y)dydx
0 0 0 0
� 2 � 1 � 2 � 1
C. (y + 2x)dydx D. (y − 2x)dydx
0 0 0 0
9. Let C be a path from point (0, 1) to ( π4 , 2). Applying the Fundamental

�
Theorem of Line Integrals, 6y cos(2x)dx + 3 sin(2x)dy is equal to
C
A. 12 B. 9 C. 6 D. 3
� � �
2 2
10. If R = {(x, y) : x + y ≤ 4}, the double integral x2 + y 2 dA equals
R
� 2π � 4 � 2π � 4
A. r3 drdθ B. r2 drdθ
0 0 0 0
� 2π � 2 � 2π � 2
C. rdrdθ D. r2 drdθ
0 0 0 0
11. Let M be the portion of the plane x + y + z = 2 that lies in the fist octant.
� �
If F (x, y, z) = (x, y, z), then the surface integral F (x, y, z) · dS equals
M
A. 2 B. 4 C. 8 D. None of the given
12. Let S be the solid in the first octant that is bounded by the surfaces
z = 1 − y 2 , y = 1, x = 0, z = 0, and y = x. The volume of S equals
2 1 5
A. B. C. D. None of the given
3 3 12
13. Let M be the planar surface x + y + z = 1 in the first octant. Assume M is

oriented by unit vectors that point away from the origin, and the boundary
�
∂M has the positive orientation. By Stokes’ Theorem, F · dr is equal to
∂M
� 1 � 1−x � 1 � 1
A. curlF · (−1, −1, 1)dydx B. curlF · (−1, −1, 1)dxdy
0 0 0 0
� 1 � 1 � 1 � 1−x
C. curlF · (1, 1, 1)dxdy D. curlF · (1, 1, 1)dydx
0 0 0 0
14. Let F (x, y, z) = (x, y, z), let M be the disk x2 + y 2 + z 2 ≤ 1, and let ∂M be
the sphere x2 + y 2 + z 2 = 1. Assume ∂M is oriented by unit vectors that
point away from the origin. By Gauss’ Divergence Theorem, the surface
� �
integral F · dS is equal to
∂M
� 1 � 2π � π � 1 � 2π � π
2
A. 3ρ sin φ dφdθdρ B. 3ρ sin φ dφdθdρ
0 0 0 0 0 0
� 1 � 2π � π � 1 � 2π � π
C. 2ρ2 sin φ dφdθdρ D. 2ρ sin φ dφdθdρ
0 0 0 0 0 0
√
15. A point (x, y, z) = ( 3, −1, 2) is given in Cartesian coordinates.
The cylindrical coordinates of the point are equal to
A. (2, 7π
6 , 2) B. (2, − π6 , 2) C. (4, 11π
6 , 2) D. (4, − π6 , 2)
16. A point (ρ, θ, φ) = (4, π3 , π6 ) is given in spherical coordinates.

The Cartesian coordinates (x, y, z) of the point are equal to
√ √ √ √ √ √
A. (1, 3, 2 3) B. (4, 4 3, 8 3) C. (12, 4 3, 8) D. (3, 3, 2)
Investigation Projects
Maxwell’s Equations
Let E, H, and J be vector-valued functions of (x, y, z, t) such that the values of
E, H, and J lie in R3 . Let ρ be a real-valued function of (x, y, z, t). In [1], a set of
Maxwell’s equations is given as follows:
divE = ρ (Gauss’ law) (35)
divH = 0 (36)
∂H
curlE + = 0 (Faraday’s law) (37)
∂t
∂E
curlH − = J (Ampere’s law) (38)
∂t
The divergence and curl of a vector field is evaluated by fixing t. If in component

form, we write E(x, y, z, t) = (E1 (x, y, z, t), E2 (x, y, z, t), E3 (x, y, z, t)), the
divergence is defined in the usual way:
∂E1 ∂E2 ∂E3
divE(x, y, z, t) = + + .
∂x ∂y ∂z
For each t, the curl of E is
 
i j k
 
curlE(x, y, z, t) = Det 

∂
∂x
∂
∂y
∂
∂z
.

E1 E2 E3
If φ is a real-valued function of (x, y, z, t), the Laplacian of φ is denoted ∇2 φ,

and defined by
∂2φ ∂2φ ∂2φ
∇2 φ = div(∇φ) = + 2 + 2.
∂x2 ∂y ∂z
Likewise, if A = (A1 , A2 , A3 ) is a vector-valued function of (x, y, z, t) with values in
R3 , the Laplacian of A is defined by
� �
∇2 A = ∇2 A1 , ∇2 A2 , ∇2 A3 .
Let J and ρ be given, and suppose A and φ satisfy

∂φ
divA + = 0 (39)
∂t
∂2φ
∇2 φ = −ρ (40)
∂t2
∂2A
∇2 A = − J. (41)
∂t2
We assume t lies in an open interval I, and A, H, J are defined for all (x, y, z) in
R3 . Also, for each t ∈ I, assume E, ρ, φ are everywhere in R3 except for finitely
many (x, y, z)’s. We assume A and φ have continuous second partial derivatives.
In Exercises 1-5, we outline a proof that the vector-valued functions defined by

∂A
E = −∇φ − and H = curlA (42)
∂t
satisfy Maxwell’s equations (35)-(38).
�∂ � ∂
1. Verify divE = ρ. Hint: Apply (39) and div ∂t A = ∂t (divA).
2. Prove divH = 0. Hint: div(curlA) = 0.
H = 0. Hint: curl(∇φ) = 0 and curl � ∂ A� =

3. Show curlE + ∂∂t ∂
(curlA).
∂t ∂t
4. Prove curl(curlA) = ∇(divA) − ∇2 A.
∂
5. Verify curlH − ∂t E = J . Hint: Apply Exercise 4.
In Exercises 6-7, let fi (x) and gj (x) be functions that are selected
from {sin x, cos x} for i, j = 1, 2, 3. We consider the following functions:
φ(x, y, z, t) = f1 (nx)g1 (nt) (43)

� � �
A(x, y, z, t) = f1 (nx)g1� (nt), f2 (mz)g2 (mt), f3 (py)g3 (pt) (44)
where n, m, p ∈ R are constants.
6. If J ≡ 0 and ρ ≡ 0 are the zero functions, prove φ and A as defined

by (43)-(44) satisfy (39), (40), and (41).
7. Let J ≡ 0 and ρ ≡ 0. If φ and A are defined by (43)-(44), evaluate

the functions E and H in (42). Notice, E and H satisfy Maxwell’s
equations because of Exercises 1-5.

Integrales

Uploaded by

Copyright:

Available Formats

Integrales

Uploaded by

Document Information

Original Description:

Original Title

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

Integrales

Uploaded by

Copyright:

Available Formats

Chapter 2

Iterated, Line, and Surface

2.1 Iterated Integrals

if for every ε > 0 there exists δ > 0 such that

We quote a theorem from advanced calculus about Riemann integrable functions.

Theorem 2.1 Riemann Integrable Functions

Let z = f (x, y) be a bounded real-valued function on a rectangular region

If the points of discontinuities of f in R lie on a finite union of graphs of continuous

In particular, every continuous real-valued function

Theorem 2.2 Fubini’s Theorem and Iterated Integrals

In particular, if f is continuous on R, the above identity applies.

1A bounded real-valued function on a rectangular region R ⊆ R2 whose set of discontinuities

Example 1 Evaluating an Iterated Integral

Solution For the inner integral, we have

Then we integrate by parts using the substitutions:

Double Integrals over General Regions

Figure 1 Region of type Rx Figure 2 Region of type Ry

Since f is bounded and continuous on Rx , the function F is bounded

and the latter integral exists by the continuity of f on Rx . By Fubini’s Theorem,

We have a similar identity for double integrals over a type Ry region.

Theorem 2.3 Integrating over Regions of Type Rx or Ry

If z = f (x, y) is a continuous real-valued function on a region of type Rx or Ry ,

Also, we may apply double integrals to define the volume of a solid.

Definition 1 Volume of a Solid and Double Integrals

Let R be a region in the xy-plane of either type Rx or Ry . Let z = f (x, y) be

Example 2 Evaluating the Volume of a Solid

Find the volume of S.

Changing the Order of Integration

y Example 3 Switching the Order of Integration

x�y�4 Sketch the region R of integration of

x Solution In Figure 6, we see the triangular region R. To switch the order of

S = [a, b] × [c, d] × [e, f ]

Sijk = [xi , xi+1 ] × [yj , yj+1 ] × [zk , zk+1 ]

We state an integrability condition for� g� similar

Theorem 2.4 Riemann Integrable Functions

Let w = g(x, y, z) be a bounded real-valued function on a box

S = [a, b] × [c, d] × [e, f ].

If the points of discontinuities of g in S lie on a finite union of graphs of continuous

Theorem 2.5 Fubini’s Theorem for Triple Integrals

If w = g(x, y, z) is the function in Theorem 2.4, then

Example 4 Evaluating a Triple Integral

Solution We integrate as follows:

Triple Integrals over General Regions

provided g is continuous on Syx .

Similarly, if an elementary region in the xy-plane is given by

where γ1 , γ2 are continuous functions of z satisfying γ1 (z) ≤ γ2 (z), and Γ1 , Γ2 are

Theorem 2.6 Triple Integrals over Elementary Regions

If g is a real-valued continuous function on an elementary region S ⊆ R3 , then

Example 5 Triple Integral

Solution In Figure 8, the base of solid S in the xy-plane may be expressed as

We apply Theorem 2.6, and evaluate the inner-most integral.

Then we evaluate the inner-integral with respect to y:

A solid S ⊆ R3 lies in the first octant, bounded by the surfaces z = 1 − x2 and

Rewrite the integral by switching the order of integration.

True or False. If false, revise the statement to make it true or explain.

Exercises for Section 2.1

In Exercises 1-8, evaluate the iterated integrals.

For No. 9 For No. 10 For No. 11