This action might not be possible to undo. Are you sure you want to continue?

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

David A. SANTOS dsantos@ccp.edu

January 15, 2009 Version

Ð ÙÐÙ×

ree to photocopy and distribute

Mathesis iuvenes tentare rerum quaelibet ardua semitasque non usitatas pandere docet.

ii Copyright c 2007 David Anthony SANTOS. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled “GNU Free Documentation License”.

Contents

Preface 1 Vectors and Parametric Curves 1.1 Points and Vectors on the Plane . . . . . . . . 1.2 Scalar Product on the Plane . . . . . . . . . . 1.3 Linear Independence . . . . . . . . . . . . . . 1.4 Geometric Transformations in two dimensions 1.5 Determinants in two dimensions . . . . . . . 1.6 Parametric Curves on the Plane . . . . . . . . 1.7 Vectors in Space . . . . . . . . . . . . . . . . 1.8 Cross Product . . . . . . . . . . . . . . . . . . 1.9 Matrices in three dimensions . . . . . . . . . 1.10 Determinants in three dimensions . . . . . . . 1.11 Some Solid Geometry . . . . . . . . . . . . . . 1.12 Cavalieri, and the Pappus-Guldin Rules . . . . 1.13 Dihedral Angles and Platonic Solids . . . . . . 1.14 Spherical Trigonometry . . . . . . . . . . . . 1.15 Canonical Surfaces . . . . . . . . . . . . . . . 1.16 Parametric Curves in Space . . . . . . . . . . 1.17 Multidimensional Vectors . . . . . . . . . . . 2 Differentiation 2.1 Some Topology . . . . . . . . . . . . 2.2 Multivariable Functions . . . . . . . 2.3 Limits . . . . . . . . . . . . . . . . . 2.4 Deﬁnition of the Derivative . . . . . . 2.5 The Jacobi Matrix . . . . . . . . . . 2.6 Gradients and Directional Derivatives 2.7 Levi-Civitta and Einstein . . . . . . . 2.8 Extrema . . . . . . . . . . . . . . . . 2.9 Lagrange Multipliers . . . . . . . . . 3 Integration 3.1 Differential Forms . . . . . 3.2 Zero-Manifolds . . . . . . . 3.3 One-Manifolds . . . . . . . 3.4 Closed and Exact Forms . . 3.5 Two-Manifolds . . . . . . . 3.6 Change of Variables . . . . 3.7 Change to Polar Coordinates 3.8 Three-Manifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v 1 1 11 15 17 25 31 41 50 58 62 64 68 70 74 79 87 90 97 97 99 100 104 107 116 122 124 129

133 . 133 . 136 . 137 . 141 . 146 . 155 . 162 . 168

iii

iv 3.9 Change of Variables . . . . . . . . . . . . . . . . . . . . . . . . . . 172 3.10 Surface Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 3.11 Green’s, Stokes’, and Gauss’ Theorems . . . . . . . . . . . . . . . . 179 A Answers and Hints 186 Answers and Hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 GNU Free Documentation License 1. APPLICABILITY AND DEFINITIONS . . . . . . 2. VERBATIM COPYING . . . . . . . . . . . . . . 3. COPYING IN QUANTITY . . . . . . . . . . . . 4. MODIFICATIONS . . . . . . . . . . . . . . . . 5. COMBINING DOCUMENTS . . . . . . . . . . . 6. COLLECTIONS OF DOCUMENTS . . . . . . . 7. AGGREGATION WITH INDEPENDENT WORKS 8. TRANSLATION . . . . . . . . . . . . . . . . . 9. TERMINATION . . . . . . . . . . . . . . . . . 10. FUTURE REVISIONS OF THIS LICENSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 . 250 . 250 . 250 . 251 . 251 . 252 . 252 . 252 . 252 . 252

Preface

These notes started during the Spring of 2003. They are meant to be a gentle introduction to multivariable and vector calculus. Throughout these notes I use Maple™ version 10 commands in order to illustrate some points of the theory. I would appreciate any comments, suggestions, corrections, etc., which can be addressed to the email below.

David A. SANTOS dsantos@ccp.edu

v

5). r′ ] and that r′ is its head.1 Vectors and Parametric Curves 1. Point. We say that r is the tail − of the bi-point [r. r′1 and is denoted by [r. 0) b = (2. 0). r ]. A vector → ∈ R2 is a codiﬁcation a ′ of movement of a bi-point: given the bi-point [r. r′ ] can be thus interpreted as an arrow starting at r and ﬁnishing. y) with x ∈ R and y ∈ R. Vector) A scalar α ∈ R is simply a real number. Some authors use the terminology “ﬁxed vector” instead of “bi-point. −4). 1 Deﬁnition (Scalar. 0) by O = (0. Those interested in more formal treatments can proﬁt by reading [BlRo] or [Lan]. 2 Example Consider the points a1 = (1.” 1 . −6). 1 b1 = (3.1. 0 Notice that inﬁnitely many different choices of departure and arrival points may give the same vector. the point (0. a2 = (3. r = (x. Here the ﬁrst coordinate x stipulates the location on the horizontal axis and the second coordinate y stipulates the location on the vertical axis. Bi-point. b2 = (5. Given two points r and r′ in R2 the directed line segment with departure point r and arrival point r′ is called the bi-point r. O = (0. See ﬁgure 1.1 Points and Vectors on the Plane W e start with a na¨ve introduction to some linear algebra necessary for the ı course. We will always denote the origin. that is. 2). −1). The zero 0 → − vector 0 = indicates no movement in either direction. A point r ∈ R2 is an ordered pair of real numbers. See ﬁgure 1. with the arrow tip. The bi-point [r. at r′ . y) in the horizon′ y −y tal axis and of y ′ − y units from the current position in the vertical axis. r′ ].2 for an example. we associate to it the vector −′ → x′ − x rr = stipulating a movement of x′ − x units from (x.

2. Though the bi-points [a1 . −5. .3: Example 2. have the same direction.1: A point in R2 . go two units right and six units down. 3Z+1 = {. According to their remainder upon division by 3. 3. b2 ] and [O. all bi-points that have the same length. 3Z+2 = {. . [a2 .}. . that is. 8. Analogously. [a1 . . See ﬁgure 1. 4. −3. −1. . consider the set of integers Z. −4. 0. . and the name of this equivalence is a vector. . they represent the same vector. that is.y axis r′ r = (x. 6. O b2 b1 b Figure 1. −6 −6 Free to photocopy and distribute 2 . 1. b1 ] ∈ . b] are in different locations on the plane. . and point in the same sense are equivalent. 5. 7. . a vector is an equivalence class of bi-points. . each integer belongs to one of the three sets 3Z = {. −18 ∈ 3Z.2: A bi-point in R2 . . . in example 2.}. . and for example. −2. the bi-point [a1 . b1 ] belongs to the 2 2 equivalence class . b1 ]. .}. a 1 Figure 1. .3. The equivalence class 3Z comprises the integers divisible by 3. In more technical language. −6. y) Points and Vectors on the Plane r x axis a2 Figure 1. as 3−1 = 5−3 = 2−0 = 2 −6 . . . As an simple example of an equivalence class. −4 − 2 −1 − 5 −6 − 0 The instructions given by the vector are all the same: start at the point. .

They have opposite sense if they have the same direction and if when translating one so as to its tail is over the other’s tail. − → − → Similarly we write. 6 Deﬁnition The Euclidean length or norm of bi-point [a. b]. ab = −ba. See ﬁgure 1. b] and [z. ← → 4 Deﬁnition Let a = b be points on the plane and let ab be the line passing through a and b.4. a] has the opposite sense of [a. say that the bi-points [a. a] = −[a. both their heads lie on the same half-plane made by the line perpendicular to then at their tails. The sense of a vector is the sense of any of its bi-point representatives.6: Bi-points with opposite sense. Bi-point [b. The direction of a vector − → = → is the direction of any of its bi-point representatives. b] is simply the distance between a and b and it is denoted by ||[a. Two vectors are parallel. Figure 1. if bi-point representatives of them are parallel. . w] if ab = ← We zw.6 .5: Bi-points with the same sense. The direction of the bi-point [a. Two bi-points are parallel if the lines containing them are parallel. b] and so we write [b.5 and 1. their heads lie on different half-planes made by the line perpendicular to them at their tails. π[ that the line ab makes with the positive x-axis (horizontal axis). − v 0 ← → → 5 Deﬁnition We say that [a. that is.4: Direction of a bi-point Figure 1. B A θ A B D C D A B C Figure 1. w] have the same sense if they have the same direction and if when translating one so as to its tail is over the other’s tail. 3 . See ﬁgures 1. b]|| = Free to photocopy and distribute (a1 − b1 )2 + (a2 − b2 )2 . the angle θ ∈ [0.Chapter 1 − → 3 Deﬁnition The vector Oa that corresponds to the point a ∈ R2 is called the position vector of the point a. when measured counterclockwise. b] has the same direction as [z. b] is the direction of the line ← → L.

It is clear that the norm of a vector satisﬁes the following properties: ¬¬− ¬¬ 1.Points and Vectors on the Plane A bi-point is said to have unit length if it has norm 1. See ﬁgures 1. that → − − the vector 0 acts as an additive identity. Free to photocopy and distribute 4 (1.10.2) . then we have Chasles’ Rule: − → − → − → AB + BC = AC. and that the additive inverse of → a →. − − − − 8 Deﬁnition If → and → are two vectors in R2 their vector sum → + → is deﬁned u v u v by the coordinatewise addition → + → = u1 + v1 = u1 + v1 . 1.9. − − u v u2 v2 u2 + v2 − −2→ u → − u B → − v C 1→ − u 2 → − u (1. proceed as follows. A vector is completely determined by three things: (i) its norm. The norm of a vector is the norm of any of its bi-point representatives. In particular. ¬¬− ¬¬ = 0 ⇐⇒ → = 0 . a ¬¬− ¬¬ 1 − 7 Example The vector → = √ has norm ¬¬→¬¬ = v v 2 √ √ 12 + ( 2)2 = 3.8: Scalar multiplication of vectors. Draw a bi-point − is − a − − representative of →. It is easy to see that vector addition is commutative and associative. and 1. and (iii) its sense. if → = − and → = − − − − whose tip is that of the bi-point for v u AB v BC. The sum → + → is the vector whose tail is that of the bi-point for → and − − − − of u u v u → → →.11.1) A →+→ − − u v Figure 1. Find a bi-point representative of → having its tail at the tip u v →. Figure 1. a ¬¬→¬¬ → − − a 2. To add two vectors geometrically. ¬¬→¬¬ ≥ 0.7: Addition of Vectors.7. 1. (ii) its direction.

3) It is easy to see that vector addition and scalar multiplication satisﬁes the following properties. The afﬁne u u u u1 − line with direction vector → = u and passing through a is the set of points on u2 the plane ´ − a + R→ = u x y ∈ R : x = a1 + tu1 . as x is constant. If u1 = 0.12.11: Difference → − − − − 10 Deﬁnition Let → = 0 . . the afﬁne line deﬁned above is vertical. → − → − − − α(→ + b ) = α→ + α b a a − − − (α + β)→ = α→ + β → a a a − − 1→ = → a a − − (αβ)→ = α(β →) a a − → v − → v → − w + → − → − v v + → − w → − w → − u + → − v → − u − → v − → − → +w − → + v) (u − →) w − → (v + − →+ u → − v → − v +→ − w − → w Figure 1. 2 y = a2 + tu2 . t∈R See ﬁgure 1. a2 αa2 (1. Put R→ = {λ→ : λ ∈ R} and let a ∈ R2 . then x − a1 (x − a1 ) u2 u2 = t =⇒ y = a2 + u2 = x + a2 − a1 .10: Associativity Figure 1. u1 u1 u1 u1 Free to photocopy and distribute 5 − → w − → − v µ − → w → − w .9: Commutativity Figure 1.Chapter 1 − 9 Deﬁnition If α ∈ R and → ∈ R2 we deﬁne scalar multiplication of a vector and a a scalar by the coordinatewise multiplication a1 αa1 − α→ = α a = . If u1 = 0.

if y = that is. and so the vector sought is v √ → − v 3 1 − 3 √ . v v a Figure 1. every Cartesian line is also an afﬁne line and one may take the vector 1 − − as its direction vector. Conversely. It also follows that two vectors → and → are parallel u v m − − − − if and only if the afﬁne lines R→ and R→ are parallel. then x y = 1 m t+ 0 k .12: Parametric equation of a line on the plane. the vector ¬¬→¬¬ has unit norm and v ¬¬− ¬¬ v − has the same direction and sense as →. → → if there u v u v → = λ→. −3 ¬¬→¬¬ = − √ √ = ¬¬− ¬¬ v 3 2 − 6 ◭ Free to photocopy and distribute 6 → − u − r = a + t→ u . parallel to → = √ but in the opposite v 2 sense. − − exists a scalar λ ∈ R such that u v → − → − − − Because 0 = 0→ for any vector →. the 0 is parallel to every vector.Points and Vectors on the Plane that is. Hence. u2 u1 . Solution: → − √ ¬¬− ¬¬ v ◮ Since ¬¬→¬¬ = 3. the afﬁne line is the Cartesian line with slope mx + k is the equation of a Cartesian line. 1 − 11 Example Find a vector of length 3.

Therefore. − → BC − → − → = BA + AC − → − → = −AB + AC −− −→ −− −→ = −2AMC + 2AMB −− −→ −− −→ = 2MC A + 2AMB −− −→ −− −→ = 2(MC A + AMB ) −−→ −− = 2MC MB . determine the vector sum AB + BC + − → − → − → CD + DE + EA. be MC and MB . − → − → 13 Example Given a pentagon ABCDE. . Demonstrate that the line segment joining the midpoints of two sides is parallel to the third side and it is in fact. Solution: ◮ Utilising Chasles’ Rule several times: − → − → − → − → − → − → → − 0 = AA = AB + BC + CD + DE + EA. as we were to shew.Chapter 1 1 −1 12 Example Find the parametric equation of the line passing through in the direction of the vector 2 −3 . let MC be the midpoint of [A. − → −−→ −− respectively. We have. t ∈ R. Here are some examples. and Solution: ◮ The desired equation is plainly x y ◭ Some plane geometry results can be easily proved by means of vectors. A]. = 1 −1 +t 2 −3 =⇒ x = 1 + 2t. Solution: ◮ Let the midpoints of [A. ◭ 15 Example In △ABC. CMC = 2 Free to photocopy and distribute 7 . Demonstrate that 1 − → − → −− −→ CA + CB . −− −→ − → −− −→ − → 2AMC = AB and 2AMB = AC. half its length. We will demonstrate that BC = 2MC MB . B] and [C.◭ 14 Example Consider a △ABC. y = −1 − 3t. B].

we have. −−→ −− −− −→ (2 − b)MA MB = (b − a)GMA . From example 14. Free to photocopy and distribute 8 . −−→ −− −− −→ Since △ABC is non-degenerate. from where the result follows. ◭ 16 Example If the medians [A. ←− −→ Solution: ◮ Since the triangle is non-degenerate. the lines AMA and ←→ − − → BMB are not parallel. − → − → CA + CB = = = −− −→ −− −→ −− −→ −− −→ CMC + MC A + CMC + MC B −− −→ −− −→ −− −→ 2CMC − AMC + MC B −− −→ 2CMC . Therefore. See ﬁgure 1. MA MB and GMA are not parallel. − → −− −→ In the same fashion. =⇒ a = b = 2.13. ◭ and thus − → −− −→ BG = 2GMB . See ﬁgure 1. MB ] of the non-degenerate △ABC intersect at the point G. −−→ −− − → 2MA MB = BA − → − → = BG + GA −− −→ −− −→ = bGMB − aGMA −− −→ −−→ −− −− −→ = bGMA + bMA MB − aGMA . b − a = 0.Points and Vectors on the Plane −− −→ −− −→ Solution: ◮ As AMC = MC B.13. AG and −− −→ − → −− −→ GMA are parallel and hence there is a scalar a such that AG = aGMA . demonstrate that − → −− −→ AG = 2GMA . whence 2 − b = 0. The point of concurrency G is called the barycentre or centroid of the triangle. MA ] and [B. and hence meet at a point G. there is a scalar b such that BG = bGMB . 17 Example The medians of a non-degenerate triangle △ABC are concurrent.

Problem 1. Let Q be a point on side [x.4 Let A. demonstrating the result. Problem 1.1. z be points on the plane with x = y and consider △xyz. and then demonstrate that for any arbitrary point M on the plane −→ − −→ − − → MA + 3MB = 4MI and −→ −→ − − − → 3MA + MB = 4MJ. the quadrilateral formed by the midpoints of the sides is a parallelogram. We must shew that the line ←→ − ←→ − ←→ − CMC also passes through G.1 Is there is any truth to the statement “a vector is that which has magnitude and direction”? such that − → − → IA = −3IB. Therefore −→ − −→ − −→ − −→ − → − → − GG′ = −2GG′ =⇒ 3GG′ = 0 =⇒ GG′ = 0 =⇒ G = G′ .6 Let x.1.3 (Varignon’s Theorem) Use vector algebra in order to prove that in any quadrilateral ABCD. Let the line CMC and BMB meet in G′ . − → → 1− JA = − JB. y = 2 − t. C] and F is the midpoint of [D. −→ − GG′ −→ − −− −→ BG′ = 2G′ MB . Construct two points I and J Problem 1. Problem 1. It follows that − → −− −→ BG = 2GMB .1. Solution: ◮ Let G be as in example 16.1. − → −− −→ AG = 2GMA . Prove that − → − → − → AC + BD = 2BC. y. B be two points on the plane.C MB G A MC MA Chapter 1 B Figure 1. 3 Problem 1. − − → −→ = GB + BG′ −− −→ −− −→ = −2GMB + 2G′ MB −→ −− −→ − − = 2(MB G + G′ MB ) −→ − = 2G′ G.13: Example 17.2 ABCD is a parallelogram. ◭ Homework Problem 1. By the aforementioned example. z] such that ||[x. E is the midpoint of [B. C].5 Find the Cartesian equation corresponding to the line with parametric equation x = −1 + t.1. whose sides do not intersect.1. Q]|| : Free to photocopy and distribute 9 . −→ − −− −→ CG′ = 2G′ MC .

8. y] if and only if there exists scalars α ≥ 0. The point a belongs to the interior of the triangle △xyz if and only if there exists scalars α > 0.17: 1.14: 1.8.1. [B]. 1. [E]. Problem Free to photocopy and distribute 10 .8.19).19: 1. 2.1. four equal. β ≥ 0 with α + β = 1 such that → = α→ + β → − − − za zx zy.Points and Vectors on the Plane ||[Q. 3.8.1. 2.8. Problem Figure 1. Problem 1. → − c → − d → − e [F].18: 1.1. z] such that ||[y. Find rational numbers α and β such − → − → − → that TQ = αTx + β Tz. as suggested in the ﬁgures. Find rational numbers l. [A]. The point a belongs to the line ← if xy and only if there exists scalars α. Q]|| = 7 : 2. m. Problem → − a → − c → − b Figure 1. y. [D]. The point a belongs to the line segment [x. β > 0 with α + β < 1 such that → = α→ + β → − − − za zx zy. z be points on the plane with x = y. Demonstrate that → 1. [C].16: 1. Assume that the divisions start or stop at the centre of the circle.8 A circle is divided into three. Let T be an arbitrary point on the plane.1. Problem → − b → − c → − c → − b Figure 1. or six equal parts (ﬁgures 1. Find the sum of the vectors. z]|| = 3 : 4 and let P be a point on [y.7 Let x. P]|| : ||[P.8.1. n such − → − → − → − → that TP = lTx + m Ty + nTz.1. β with α + β = 1 such that → = α→ + β → − − − za zx zy. Problem 1. Problem → − a → − d → − c → − b Figure 1. → − a → − a → − c → − a → − d → − b Figure 1.15: 1. Problem → − b → − d → − a → − f Figure 1.17 through 1.1.

Chapter 1 1. Their scalar product (dot product. α ∈ R.9) → → − b −− a → − b → − a Figure 1.8) (1. SP1 Bilinearity − − − − − − − y z (→ + →)•→ = →•→ + →•→. x y x y x y SP3 Commutativity SP4 SP5 SP6 → •→ = → •→ − − − − x y y x → •→ ≥ 0 − − x x − →•→ = 0 ⇐⇒ → = → − − − x x x 0 Ô ¬¬→¬¬ − − ¬¬− ¬¬ = →•→ x x x →•(→ + →) = →•→ + →•→ − − − − − − − x y z x y x z (1. inner x y product) is deﬁned and denoted by →•→ = x y + x y . a 19 Example If → = a 2 4 The following properties of the scalar product are easy to deduce from the deﬁnition.7) (1. − − x y 1 1 2 2 1 3 → − − − → − and b = . x y z x z SP2 Scalar Homogeneity − − − − − − (α→)•→ = →•(α→) = α(→•→).2 Scalar Product on the Plane We will now deﬁne an operation between two plane vectors that will provide a further tool to examine the geometry on the plane.20: Theorem 21. − − 18 Deﬁnition Let → ∈ R2 and → ∈ R2 .4) (1.6) (1. then →• b = 1 · 3 + 2 · 4 = 11. Free to photocopy and distribute 11 .5) (1.

→). → − It follows that the vector 0 is simultaneously parallel and perpendicular to any vector! − 23 Deﬁnition Two vectors are said to be orthogonal if they are perpendicular. b ) ∈ [0. − − − − − a b a b a b Proof: From ﬁgure 1.9) we have → − − → − → − − − − − → || b − →||2 = ||→||2 + || b ||2 − 2||→|||| b || cos (→. b ) a a a a a →→ − − − → − → − − − → − − − − − → ⇐⇒ b • b − 2→• b + →•→ = ||→||2 + || b ||2 − 2||→|||| b || cos (→.20. a − → − − → − denoted by (→. we write a b ¬¬→¬¬ ¬¬− ¬¬ → − → − ¬¬− ¬¬ − − 24 Deﬁnition If → ⊥ b and ¬¬→¬¬ = ¬¬ b ¬¬ = 1 we say that → and b are ora a a thonormal.4) through (1. as the angle between the afﬁne lines R→ and R b . b ) a a a a a a → − − → − → − − − → − − − − → ⇐⇒ || b ||2 − 2→• b + ||→||2 = ||→||2 + || b ||2 − 2||→|||| b || cos (→. a a a as we wanted to shew. Then a − →•→ = ||→||||→|| cos (→. 2 22 Corollary Two vectors in R2 are perpendicular if and only if their dot product is 0. we deﬁne the (convex) angle between them. a a → − − 21 Theorem Let → and b be vectors in R2 . If → a → − − → ⊥ →. b ) = a in Theorem 21 we obtain the following corollary. and (1. u π − − → Putting (→. b ). 12 . π].Scalar Product on the Plane → − − 20 Deﬁnition Given vectors → and b . − Equality occurs if and only if → a Free to photocopy and distribute → − b. Since | cos θ| ≤ 1 we also have 25 Corollary (Cauchy-Bunyakovsky-Schwarz Inequality) ¬ ¬ ¬¬→¬¬¬¬→¬¬ ¬→•→¬ − − − ¬¬− ¬¬ ¬ a b ¬ ≤ ¬¬ a ¬¬¬¬ b ¬¬. − is orthogonal to b . b ) a a a a → − − → − − → − → − − − − − → ⇐⇒ ( b − →)•( b − →) = ||→||2 + || b ||2 − 2||→|||| b || cos (→. b ) a a a a a − → − − − → − − → ⇐⇒ →• b = ||→|||| b || cos (→. using Al-Kashi’s Law of Cosines on the length of the vectors.

10) 26 Example Let a. 1 = |a · 1 + b · 1| ≤ (a2 + b2 )1/2 (12 + 12 )1/2 =⇒ a2 + b2 ≥ Equality occurs if and only if a b =λ 1 1 1 2 . In such case. ¬¬ ¬¬ ¬¬ ¬¬ ¬¬ − ¬¬→ − + → + →¬¬ = ¬¬ x + y + z − ¬¬ b c ¬¬ a ¬¬ ¬¬ x + y + z Also. u 28 Example Let x. and so equality is achieved for a = b = 27 Corollary (Triangle Inequality) 1 . Then y z x ¬¬ ¬¬ √ ¬¬ ¬¬ = 2(x + y + z).Chapter 1 a1 b1 → − − If → = a and b = . Proof: → − − ||→ + b ||2 a → − − → − − = (→ + b )•(→ + b ) a a − − − − − − → → → = →•→ + 2→• b + b • b a a a → − → − − − ≤ ||→||2 + 2||→|||| b || + || b ||2 a a → 2 − − = (||→|| + || b ||) .− = c . Solution: ◮ By the CBS Inequality. y. b = . − Solution: ◮ Put → = a y → z → − . ¬¬− ¬¬ ¬¬→¬¬ ¬¬− ¬¬ ¬¬− ¬¬ + ¬¬→¬¬ + ¬¬→¬¬ = a c ¬¬ b ¬¬ x2 + y 2 + y2 + z2 + Ô z 2 + x2 . 1 2 1 2 (1. . Minimise a2 + b2 subject to the constraint a + b = 1. the CBS Inequality takes the form a2 b2 |a1 b1 + a2 b2 | ≤ (a2 + a2 )1/2 (b2 + b2 )1/2 . 13 Free to photocopy and distribute . b be positive real numbers. a = b = λ. a from where the desired result follows. Prove that √ 2(x + y + z) ≤ x x2 + y 2 + y2 + z2 + Ô z 2 + x2 . z be positive real numbers.◭ 2 ¬¬ ¬¬→¬¬ ¬¬→¬¬ →¬¬ − ¬¬ ¬¬→ ¬¬− ¬¬ − − ¬¬ a + b ¬¬ ≤ ¬¬ a ¬¬ + ¬¬ b ¬¬.

AB AB D′ D C A B A′ B′ A B C A′ B′ C′ Figure 1. We will denote this distance by AB − . − or more routinely. by AB. − 29 Deﬁnition Let A and B be points on the plane and let → be a unit vector. u − → →. Then ← ′ − → AA D D′ ← ′ → AC A′ C ′ CC ⇐⇒ = ′ ′. ← → ← → B = B′ . ◭ We now use vectors to prove a classical theorem of Euclidean geometry. On the one hand. A = B. AB AC A′ B ′ AC On the other hand. ← − → Since AA′ Free to photocopy and distribute 14 . there is a scalar λ ∈ R such that −′ → − −→ − − → A B′ = AB + λAA′ . because they are unit vectors in the same direction. Let AA′ BB′ . u u → is patent. C be distinct points of D . and A′ . A = A′ . B. if the vector u ← ← → → 30 Theorem (Thales’ Theorem) Let D y D′ be two distinct lines on the plane. Let ← → ← → A.Scalar Product on the Plane and the assertion follows by the triangle inequality ¬¬ ¬¬→¬¬ ¬¬→¬¬ ¬¬→¬¬ → →¬¬ − ¬¬→ ¬¬− ¬¬ − − ¬¬ − − ¬¬ a + b + c ¬¬ ≤ ¬¬ a ¬¬ + ¬¬ b ¬¬ + ¬¬ c ¬¬. Proof: Refer to ﬁgure 1. B′ . = ′ ′. then λ is the directed distance or algebraic measure of the line − If AB = λ u − → segment [AB] with respect to the vector →. Figure 1. ← ′ → BB . A′ = B′ .2. C′ be distinct points of D′ . −→ − − − −→ − − −→ − − → −→ − → → BB′ = BA + AA′ + A′ B′ = (A′ B′ − AB) + AA′ . C = C′ . Observe that AB = −BA. −′ → − −′ → − − → − → AB AC A B′ A C′ = .22: Corollary to Thales’ Theorem.21: Thales’ Theorem. by Chasles’ Rule.

and a fortiori x → →.) ← → ← → 31 Corollary Let D and D′ are distinct lines. say. and A′ . B ′ . when can we ﬁnd scalars a.u From the preceding theorem. In the ﬁrst case. BB ⇐⇒ CA CA′ 1. −→ − − − − → −→ − → CC′ = CA + AA′ + A′ C′ − −→ − AC − A′ C ′ − → −→ → = − · AB + AA′ + ′ ′ AB + λAA′ AB AB − A′ C ′ AC − A′ C ′ −→ → = − AB + 1 + λ ′ ′ AA′ . since all vectors are parallel to the zero vector. → = 0 . − y 2 j 15 Free to photocopy and distribute . → and →. Under which condix y → on the plane as a linear combination − tions can we write an arbitrary vector v − − of → and →. and y1 y2 to x1 y1 − − =λ ⇐⇒ → →. say. AB AB proving the theorem. In the second case we − − x y have − → = x →. − x 2 j − → = y →. points on line D′ . B. then either x2 = 0 or y1 = 0. b such that x y C → = a→ + b→ ? − − − v x y The answer can be promptly obtained algebraically. But. → = a→ + b→ − − − v x y ⇐⇒ ⇐⇒ v1 = ax1 + by1 .3 Linear Independence − − onsider now two arbitrary vectors in R2 . ← → ← → Let A. Operating formally. v1 y2 − v2 y1 a= . Then ← ′ − → AA ← ′ → CB CB ′ = . x y x2 y2 → − − If x1 = 0. what x1 x2 does it mean x1 y2 = x2 y1 ? If none of these are zero then = = λ. intersecting in the unique point C. we immediately gather the following corollary.2. the equality above reveals that ← ′ ← ′ − → → AC A′ C ′ AA CC ⇐⇒ − ′ ′ = 0. that is. x 1 y2 − x 2 y1 The above expressions for a and b make sense only if x1 y2 = x2 y1 . x 1 y2 − x 2 y1 v2 = ax2 + by2 x1 v2 − x2 v1 b= . (See ﬁgure 1.Chapter 1 Assembling these results. A′ B ′ AB AB ← − → ← → As the line AA′ is not parallel to the line AB. be points on line D .

. → and →. 1]. coloured in 0 1 brown. . In essence 0 0 1 1 then. . The fundamental parallelo− − gram of the the vectors → and → is the set x y − − {a→ + b→ : a ∈ [0. − − − Consider now two linearly independent vectors → and →. Observe of ´ 0 1 1 2 that the vertices of this parallelogram are . an arbitrary vector → can be x y v written as the linear combination → = a→ + b→. − − 34 Deﬁnition Given two linearly independent vectors → and → consider bi-point x y representatives of them with the tails at the origin. for b ∈ [0.23 shews the fundamental parallelogram of Free to photocopy and distribute 16 1 1 µ . and the respective tiling of the plane by various translationsµ it. We have demonstrated − y − − − 32 Theorem Given two vectors in R2 . In this last case we say that → is linearly x y x →. →. If two vectors are not linearly independent.Linear Independence → − − and so both vectors are parallel to j and hence → x the following theorem. since one a b we can express is not a scalar multiple of the other. 1]. 1 0 1 1 33 Example The vectors and are clearly linearly independent. 1]. a→ x y x → and traverses the whole length of →: from its tip (when a = 1) − − is parallel to x x − − to its tail (when a = 0). 1]}. then we − independent from vector y say that they are linearly dependent. b∈R − − − if and only if → is not parallel to →. b→ is parallel to → y y →. In the same manner. x y ´ . The linear combination a→ + b→ is also a − − − and traverses the whole length of y x y vector on the plane. Given an arbitrary vector it as a linear combination of these vectors as follows: a b 1 0 1 1 = (a − b) +b . b ∈ [0. . linear independence of two vectors on the plane means that we may obtain every vector on the plane as a linear combination of these two vectors and hence Figure 1. − − − v x y a ∈ R. For a ∈ [0.

3 Consider the line with Cartesian equation L : ax + by = c.3. Find a formula for the distance from t to L. will be certain construct called matrices which we will study in the next section.4 Geometric Transformations in two dimensions e now are interested in the following fundamental functions of sets (ﬁgures) on the plane: translations.3. and draw their fundamental parallelogram. Homework Problem 1. Let t be a point not on L.23: Tiling and the fundamental parallelogram.1 Prove that æ é 1 æ and 1 1 are linearly independent. as a linear combination b æ é æ é 1 −1 of the vectors and . It will turn out that a handy tool for investigating all of these (with the exception of translations).2 Write an arbitrary vector æ é a on the plane.3. b are zero. and rotations about the origin.Chapter 1 cover the whole plane by all these linear combinations. scalings (stretching or shrinking) reﬂexions about the axes. Problem 1. where not both of a. Free to photocopy and distribute 17 W . é −1 Problem 1. 1 1 1.3.4 Prove that two non-zero perpendicular vectors in R2 must be linearly independent. Figure 1. Problem 1.

v v v v 1 v 2 v 1 v 2 v 2 1 2 1 and − − − − − − − − v v v (T→2 ◦ T→1 )(a) = T→2 (T→1 (a)) = T→1 (a) + →2 = a + →1 + →2 . v v v A translation simply shifts an object on the plane rigidly (that is. Observe that the identity transformation leaves a point untouched. 35 Example The function I : R2 → R2 . and the output is also a point on the plane. I(x) = x is called the identity transformation. that is. It is clear that the composition of any two translations commutes.5 Free to photocopy and distribute = .0. it does not distort it shape or re-orient it). We start with the simplest of these functions. 18 .5y Figure 1. See ﬁgure 1. 37 Deﬁnition A function Sa. Then v v v − − − − − − − − (T→ ◦ T→ )(a) = T→ (T→ (a)) = T→ (a) + → = a + → + → .Geometric Transformations in two dimensions First observe what is meant by a function F : R2 → R2 .24 for an example. − 36 Deﬁnition A function T→ : R2 → R2 is said to be a translation if it is of the v − − − form T→ (x) = x + →.25 shews the scaling S2.25: A scaling. → − v Figure 1. where → is a ﬁxed vector on the plane. This means that the input of the function is a point of the plane. to a copy of itself a given amount of units from where it was. v v v v v from where the commutativity claim is deduced. if − − − − − − T→1 . For let v v v v v v − − − T1 (a) = a + →1 and T→2 (a) = a + →2 .b : R2 → R2 is said to be a scaling if it is of the form ax Sa. b > 0 are real numbers. where a > 0. then T→1 ◦ T→2 = T→2 ◦ T→1 .24: A translation. T→2 : R2 → R2 are translations. but nevertheless an important one is the following. Figure 1.b (r) = . A rather uninteresting example. by x y 2x 0.

1 0 0 = 0 0 . = .1 ◦ T→ ) i −1 0 = S2. A function RV : R2 → R2 y Free to photocopy and distribute 19 . −1 0 − = T→ i S −1 0 − = T→ i −2 0 = −1 0 .b ◦ Sa′ .b ◦ Sa′ .1 − T→ i −1 0 = S2. For (Sa.b .b and (Sa′ .b (Sa′ .b′ = Sa′ . For consider 2a1 → − − the translation T→ (a) = a + i and the scaling S2. 38 Deﬁnition A function RH : R2 → R2 is said to be a reﬂexion about the y-axis or −x horizontal reﬂexion if it is of the form RH (r) = .Chapter 1 It is clear that the composition of any two scalings commutes. The original object (in the ﬁrst quadrant) is yellow. Then i a2 − (T→ ◦ S2.b′ ◦ Sa. then Sa. if Sa. that is. but − (S2.b′ )(r) = Sa. Translations and scalings do not necessarily commute. Its reﬂexion about the origin is blue (on the third quadrant).1 (a) = .26: Reﬂexions.b′ (Sa. y axis x axis Figure 1.b′ from where the commutativity claim is deduced. Its reﬂexion about the x-axis is cyan (on the fourth quadrant). Sa′ . Its reﬂexion about the y-axis is magenta (on the second quadrant).b )(r) = Sa′ .b′ : R2 → R2 are scalings. however.b′ (r)) = Sa.1 ) i a′ x b′ y ax by = a(a′ x) b(b′ y) a′ (ax) b′ (by) .b .b′ ◦ Sa.b (r)) = Sa′ .

y = ρ sin α as in ﬁgure 1. This deﬁnition will be somewhat harder than the others. Some reﬂexions appear in ﬁgure 1. 2π[. A few short computations establish various commutativity properties among reﬂexions. ρ cos α ρ sin α θ α x-axis Figure 1.26. Hence the point x y is mapped to the point y-axis ρ cos(α + θ) ρ sin(α + θ) x cos θ − y sin θ x sin θ + y cos θ . and scalings. If we rotate it. Free to photocopy and distribute 20 . so let us develop some ancillary results. A function RO : R2 → R2 is said to be a reﬂexion about origin if RV (r) = −y it is of the form RH (r) = −x −y .27. But ρ cos(α + θ) = ρ cos θ cos α − ρ sin θ sin α = x cos θ − y sin θ.4. We may now formulate the deﬁnition of a rotation. We now deﬁne rotations. in the levogyrate sense.27: Rotation by an angle θ in the levogyrate (counterclockwise) sense from the x-axis.4. by an angle θ. we land on the new point x′ with x′ = ρ cos(α + θ) and y ′ = ρ sin(α + θ). and ρ sin(α + θ) = ρ sin α cos θ + ρ sin θ cos α = y cos θ + x sin θ. Consider a point r with polar coordinates x = ρ cos α.Geometric Transformations in two dimensions is said to be a reﬂexion about the x-axis or vertical reﬂexion if it is of the form x . translations. Here ρ = x2 + y 2 and α ∈ [0. See problem 1.

the identity transformation I(x) = x.6. We now codify some properties shared by scalings.5 and 1.Chapter 1 39 Deﬁnition A function Rθ : R2 → R2 is said to be a levogyrate rotation about the x cos θ − y sin θ . 2 columns) array whose columns are (in this order) 1 0 L and L . and rotations are linear transformations from R2 to R2 .b is . and rotations. it is veriﬁed that L(a + b) = L(a) + L(b). b on the plane and every scalar λ. (2 rows. Various properties of the composition of rotations with other plane transformations are explored in problems 1. reﬂexions. The matrix of a 0 the scaling transformation Sa. 40 Deﬁnition A function L : R2 → R2 is said to be a linear transformation from R2 to R2 if for all points a. then → − → − → − → − L(r) = L(x i + y j ) = xL( i ) + yL( j ). and thus a linear transformation from R2 to R2 is solely determined by the values → − → − L( i ) and L( j ). reﬂexions and rotations are linear transformations. We will now introduce a way to codify these values. We have seen that scalings. origin by the angle θ measured from the positive x-axis if Rθ (r) = x sin θ + y cos θ Here ρ = x2 + y 2 . b > 0 be a real numbers. L(λa) = λL(a). 41 Deﬁnition A function A : R2 → R2 is said to be an afﬁne transformation from R2 to R2 if there exists a linear transformation L : R2 → R2 and a ﬁxed vector → ∈ R2 such that for all points x ∈ R2 it is veriﬁed that − v − A(x) = L(x) + →. v It is easy to see that translations are then afﬁne transformations.4.b 1 0 = a·1 b·0 = a 0 21 Free to photocopy and distribute . It is easy to prove that scalings. 0 1 43 Example (Scaling Matrices) Let a > 0. The matrix AL associated to L is the 2 × 2. If L : R2 → R2 is a linear transformation. where for the linear transformation L in the deﬁnition we may take I : R2 → R2 . but not so translations.4. For 0 b Sa. 42 Deﬁnition Let L : R2 → R2 be a linear transformation. reﬂexions.

Id(x) = x is I2 = . that the matrix for the transformation RV is 0 1 0 −1 . 45 Example (Rotating Matrices) It is easy to verify that the matrix for a rotation cos θ − sin θ .4. r t s u a c b d and .b 0 1 = 0·1 b·1 = 0 b . Rθ is sin θ cos θ 46 Example (Identity Matrix) The matrix for the identity linear transformation Id : 1 0 R2 → R2 . N(x) = O is 02 = .Geometric Transformations in two dimensions and Sa. Then the compo- Proof: We need to ﬁnd (L ◦ L′ ) and (L ◦ L′ ) 0 1 . We are now interested in how to codify the matrix of a composition of linear transformations L1 ◦ L1 in terms of their individual matrices. 0 1 47 Example (Zero Matrix) The matrix for the null linear transformation N : R2 → 0 0 R2 . and the matrix for the transformation RO is −1 0 0 −1 . 44 Example (Reﬂexion Matrices) It is easy to verify that the matrix for the trans−1 0 1 0 formation RH is .7 we know that the composition of two linear transformations is also linear. 48 Theorem Let L : R2 → R2 have the matrix representation AL = let L′ : R2 → R2 have the matrix representation AL′ = sition L ◦ L′ has matrix representation ar + bt as + bu cr + dt cs + du 1 0 . 22 Free to photocopy and distribute . 0 0 From problem 1.

We deﬁne scalar multiplication of a matrix as λA = λ a c b d = λa λc λb λd . 1 0 = L L′ 1 0 =L r t → − → − = rL( i )+tL( j ) = r a c +t b d = ar + bt cr + dt . Since the composition of functions is not necessarily commutative.Chapter 1 We have (L◦L′ ) and (L◦L′ ) 0 1 = L L′ 0 1 =L s u a b → − → − = sL( i )+uL( j ) = s +u c d ar + bt cr + dt as + bu cs + du . and λ ∈ R = a+r c+t b+s d+u . neither is matrix multiplication. so is matrix multiplication. MN = 3 −2 1 1 . 1 0 −1 1 . as = as + bu cs + du . We deﬁne matrix multiplication as AB = a c b d r t s u = ar + bt cr + dt as + bu cs + du . whence we conclude that the matrix of L ◦ L′ is we wanted to shew. a b r s u c d t be a scalar. 2M = 2 0 −2 2 . Free to photocopy and distribute 23 . 50 Example Let M = Then M +N = 2 −2 1 2 . Since the composition of functions is associative. N = 1 −2 2 1 . We deﬁne matrix addition as A+B = a c b d + r t s u 49 Deﬁnition Let A = and B = be two 2 × 2 matrices.u The above motivates the following deﬁnition.

a c b d 1 1 = 1 3 =⇒ a+b c+d = 1 3 =⇒ b = −1. Assume in each case that the vertices of the ﬁgures are lattice points.29. Figure 1. transformed to itself.Geometric Transformations in two dimensions 51 Example Find a 2 × 2 matrix that will transform the square in ﬁgure 1. that each vertex of the square is transformed in the same order. a c b d Solution: ◮ Let be the desired matrix. Then the point a c b d 1 0 = 2 2 =⇒ a c = 2 2 =⇒ a = c = 2.28 into the parallelogram in ﬁgure 1. Homework Free to photocopy and distribute 24 . is a fortiori.28: Example 51. 0 Using these values. that is. d = 1. Figure 1. coordinate points with integer coordinates. Then since a c b d 0 0 0 0 = . And so the desired matrix is 2 2 ◭ 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 −1 1 . We now assume. with0 out loss of generality. to each vertex of the rectangle. counterclockwise.29: Example 51.

1 If A = 1 2 æ a 1 b é and é −1 . and so the composition of two rotations on the plane is commutative.31: Problem 1. S1. Problem 1.3 Find the effects of the reﬂexions R π . Prove that their composition L ◦ L′ is also a linear transformation. R π .Chapter 1 æ Problem 1.10.4.8 Find the effects of the reﬂexions RH .10 Find the image of the ﬁgure below (consisting of two circles and a æ é 1 1 triangle) under the matrix .30.4. and S2. 5 4 −2 (A + B)2 = A2 + 2AB + B 2 .30. as in ﬁg2 −2 1 ure 1.2 . Problem 1.5 Determinants in two dimensions e now desire to deﬁne a way of determining areas of plane ﬁgures on the plane.4.1 . Problem 1.4. B 3 = Problem 1.4. and R− π on the 2 4 2 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 Figure 1.7 Let L : R2 → R2 and L′ : R2 → R2 be linear transformations. Prove that the composition of a rotation and a translation is not necessarily commutative. R− π . and RO on the triangle in ﬁgure 1. Prove that the composition of a reﬂexion and a scaling is commutative. Problem 1.4. 1. Prove that the composition of a reﬂexion and a rotation is not necessarily commutative.2 Consider △ABC with A = −1 0 2 .4.30. 1. Determine the the effects of the following scaling transformations on the triangle: S2.4. Prove that the composition of a reﬂexion and a translation is not necessarily commutative. −1 3 2 1 0 0 1 2 Figure 1. triangle in ﬁgure 1.4.6 Prove that the composition of a scaling and a rotation is not necessarily commutative. C = .4.30: Problems 1.2.8. Problem 1.9 Find all matrices A M2×2 (R) such that A2 = 02 ∈ Problem 1. ﬁnd a and b.4. Problem 1.5 Prove that the composition of two rotations on the plane Rθ and Rθ ′ satisﬁes Rθ ◦ Rθ ′ = Rθ+θ ′ = Rθ ′ ◦ Rθ .4.2 .4.4 Prove that the composition of two reﬂexions is commutative. and 1. Problem 1. RV .3.4. It seems reasonable to require that this area determination agrees 25 W Free to photocopy and distribute . B = .

as in ﬁgure 1.32: Area of a parallelogram. This motivates the following deﬁnition.33: (a + c)(b + d) − 2 · ab 2 −2 · c(2b+d) 2 = ad − bc. Consider now a simple quadrilateral with vertices r1 = (x1 . From ﬁgures (1. This quadrilateral is spanned by the vectors −→ = x2 − x1 . the area of a parallelogram should be as we learn in elementary geometry and the area of a unit square should be 1.34. r2 = (x2 .32) and (1.33). y3 ). listed in counterclockwise order. r4 = (x4 . = ad − bc. c d b d a b a c Figure 1. the area of a parallelogram spanned by and c d is a b c d a b . y4 ). D . y2 ). in particular. Figure 1. r− 2 1r y2 − y1 Free to photocopy and distribute −→ = x4 − x1 . r− 4 1r y4 − y1 26 . r3 = (x3 .Determinants in two dimensions with common formulæ of areas of plane ﬁgures. y1 ). a b c d 52 Deﬁnition The determinant of the 2 × 2 matrix a b c d is det = ad − bc.

y4 ). y2 ). r1 r2 r2 r r r3 r3 r1 4 This is equivalent to 1 − − − − − − ¡ 1 − − − − − − − − − − D(→. → − →). → − →) r2 r1 r4 r1 r4 r3 r2 r3 →. →) + D(→. →) − D(→. →) − D(→. →) + D(→.36. → − →). reﬂect it about − r3 3 3 one of its sides. → − →) + D(→ − →. →) + D(→. r− 2 3r y2 − y3 Using the properties derived in Theorem ??. →) + D(→. → = (x2 . which. →) + D(→. (x3 . →)¡ − − − − − − − − D( r r r r r r r r 2 4 2 →.11. →) − − − 1 + 2 D( r4 r2 2 4 2 1 1 4 1 1 − − − − − − ¡ D(→. →) − D(→. →) . r1 r2 r2 r3 r3 r4 r4 r1 1 1 4 2 4 2 3 2 4 3 We conclude that the area of a quadrilateral with vertices (x1 . y1 ). creating a parallelogram. r2 r1 r4 r1 Similarly. r1 r2 r2 r r r3 r3 r1 r2 r3 Free to photocopy and distribute 27 . listed in counterclockwise order. is 1 → → − − − − − − ¡ D(− . →) − D(→. − ) + D(→.35. →) − D(→. →) r3 r2 r4 r3 r3 r3 →.Chapter 1 and hence. its area is given by A = det x2 − x1 y2 − y1 x4 − x1 y4 − y1 − − − − = D(→ − →. noticing that the quadrilateral is also spanned by −→ = x4 − x3 . →) − D(→. y ). →) + D(→. →) + 2D(→. we see that A = = = = 1 2 1 2 1 2 1 2 − − − − − − − − ¡ D(→ − →. →) + D(→. →) − D(→. →) = 0. y2 ). →) − D(→. →) + 2D(→. (x2 .11) − − Similarly. →) − D(→. →) − D(→. r1 r2 → = (x . →) − D(→. by virtue of 1. →)¡ − − − − − − − − − − − − D( r r r r r r r r r r r r − − − − − − − − ¡ D(→. (1. r4 r3 r2 r3 −→ = x2 − x3 . y1 ). →) + D(→. r− 4 3r y4 − y3 its area is also given by A = det x4 − x3 y4 − y3 x2 − x3 y2 − y3 − − − − = D(→ − →. as in ﬁgure 1. →) + D(→. →))¡ + 1 D(→. to ﬁnd the area of a triangle of vertices → = (x1 . The area of the triangle is now half the area of the parallelogram. y3 ). →) − D(→. →) . (x4 . →) + D(→. →) + D(→. listed in counterclockwise order is 1 2 det x1 y1 x2 y2 + det x2 y2 x3 y3 + det x3 y3 x4 y4 + det x4 y4 x1 y1 . →) r1 r2 r2 r3 r3 r1 r1 r2 r2 r r r3 r3 r1 r2 r3 2 4 We will prove that − − − − − − − − − − D(→. as in ﬁgure 1.

→) + 2D(→. →) − D(→. . (1. If P is not convex. then there must be a vertex Free to photocopy and distribute 28 . 53 Theorem (Surveyor’s Theorem) Let (x1 . → − →) + D(→ + → − →. y1 ). Consider now a simple polygon P with n vertices. We have proved then that the area of a triangle. →) − D(→.12) and (1. (x2 . . We have already proved the cases n = 3 and n = 4 in (1.Determinants in two dimensions To do this. →) − D(→. y1 ). → − →) + 2D(→. − − − − − − = D(→ + →. yn ) be the vertices of a simple (non-crossing) polygon.12) In general. →) − − − − = D( r r r r 3 2 2 3 − − − − − − − = D(→. which r r3 r2 r1 → = → + → − →. →) − − − − − − −D( r r r r r r 3 2 1 3 3 1 2 3 = 0. → − → = → − →. (x2 . →) − − − − − − = D( r3 r2 r2 r3 r2 r3 →. listed in counterclockwise order. and observe. we have the following theorem. (xn . is 1 2 det x1 y1 x2 y2 + det x2 y2 x3 y3 + det x3 y3 x1 y1 .11). triangulating the polygon. →) + D(→. y2 ). →) r3 r2 r2 r3 r2 r3 →. Then its area is given by 1 2 det x1 y1 x2 y2 + det x2 y2 x3 y3 + · · · + det xn−1 yn−1 xn yn + det xn yn x1 y1 . y2 ). →) + 2D(→. (x3 . Proof: The proof is by induction on n. Thus − − − − means r r3 r2 r1 − − − − − − − − − − − − − − − − D(→. → r1 r2 r3 r3 r2 r1 r →. →) − D(→. → + → − →) + 2 r1 r2 r2 r r r3 r3 r1 r2 r3 r1 r2 r2 r3 r2 r1 → + → − →. →) = D(→. creating n − 2 triangles. →) − D(→. whose vertices (x1 . →) + D(→. we appeal once again to the bi-linearity properties derived in Theorem − − − − ??. →) − − +2D( r r as claimed. . that since we have a parallelogram. y3 ) are listed in counterclockwise order. If P is convex then we may take any vertex and draw a line to the other vertices. respectively. .

y3 ) r3 r2 Chapter 1 (x2 . r1 (x1 . ﬁnding another diagonal and creating a new sub-polygon.35: Area of a triangle. the terms det xi yi xj yj + det xj yj xi yi = 0. Eventually. disappear from the sum and we are simply left with 1 2 u We may use the software Maple™ in order to speed up computations with vectors. but having opposite orientations. y3 ) r3 r4 (x4 . → = a and b = and a matrix A := 2 1 Free to photocopy and distribute 29 det x1 y1 x2 y2 + det x2 y2 x3 y3 + · · · + det xn−1 yn−1 xn yn + det xn yn x1 y1 . y2 ) Figure 1. A ray produced from this vertex must hit another vertex. y1 ) r2 (x2 . the polygon can be triangulated in such a way that all triangles inherit the positive orientation of the original polygon but each neighbouring pair of triangles have opposite orientations. In the latter case. r1 (x1 .34: Area of a quadrilateral. that has a reﬂex angle. Moreover. y3 ) (x . y ) 1 1 r3 Figure 1. we repeat the argument. y) r2 (x2 . since the number of vertices is inﬁnite.12) we obtain that the area is det xi yi xj yj . .(x3 . These two sub-polygons are either both convex or at least one is not convex. For 1 2 → − − example. y1 ) Figure 1. Applying (1. y4 ) r (x. This diagonal divides the polygon into two sub-polygons. we end up triangulating the polygon. Since each diagonal occurs twice. let us deﬁne two vectors. Most of the commands we will need are in the linalg package. y2 ) r1 (x3 . creating a diagonal.36: Area of a triangle. where the sum is over each oriented edge. y2 ) (x3 . otherwise the polygon would have inﬁnite area.

5. Prove that − − u v → • → = 1 ||→ + →||2 − ||→ − →||2 ¡ . −2 Homework − Problem 1.4 Let → . > A:=matrix([[1. The result will not display if a colon is chosen.2]. . a a → − − then→ = b .7 Find the Cartesian equa−1 tion of all lines L′ passing through 2 π and making an angle of radians with the 6 Cartesian line L : x + y = 1. Using Corollary 22.) > > > > with(linalg): a:=vector([2.5. Prove that if − − − − − → ∀ → ∈ R2 .5. prove that − a b ¬¬− ¬¬2 ¬¬→¬¬2 → − ¬¬− ¬¬ − ||→ + b ||2 = ¬¬→¬¬ + ¬¬ b ¬¬ . 2] normalize(a).6 Consider two lines on the plane L1 and L2 with Cartesian equations L1 : y = m 1 x + b1 and L2 : y = m 2 x + b1 . (There must be either a colon or a semicolon at the end of each statement. v v a v Free to photocopy and distribute 30 . m 2 = 0.2]).4]]). √ √ 5 2 5 [ .b). 1] b:=vector([1. where m 1 = 0. ﬁnd a unit vector in the direction of →. → be vectors in R2 .Determinants in two dimensions 1 2 − .b).[3.5. a Problem 1. 2 Problem 1. → • → = → • b . a := [2. A := 1 3 > det(A).2 (Pythagorean Theorem) If − → ⊥ →.5 (Polarisation Identity) Let →. Prove that (a2 + b2 )2 ≤ 2(a4 + b4 ).3 Let a. b be arbitrary real numbers. − − − − − − u v u v u v 4 Problem 1. − − → Problem 1. prove that L1 ⊥ L2 ⇐⇒ m 1 m 2 = −1. a 3 4 and the angle between the vectors. angle(a.5. b := [1. Let us compute their dot product. Problem 1.5.5.1]). ] 5 5 4 arccos 4 5 2 4 > > dotprod(a. b be ﬁxed vectors a 2 in R . Problem 1.1 æ Findé all vectors → ∈ R2 a √ −3 − such that → ⊥ a and ||a|| = 13.

that is. b] → R2 . with w 0 → → − − → = → − v • w → is perpendicular to → . Prove that the vector − plane. but in obtaining suitable parametrisations of simple Cartesian curves. 1 + t2 1 − t2 . and a theory akin to the one studied for Cartesian equations in a ﬁrst Calculus course has been developed. then p is a multiple point of the curve. r(a) is the initial point of the curve and r(b) its terminal point.37: x = sin 2t.39: x = y= t − t3 . 1 + t2 Figure 1. A curve is closed if its initial point and its ﬁnal point coincide. Figure 1. b] ⊆ R.Chapter 1 − − Problem 1. y = cos 6t. The curve is simple if its has no multiple points. A parametric curve representation r of a curve Γ is a function r : [a. b]) = Γ. A closed curve whose only multiple points are its endpoints is called a Jordan curve. be vectors on the v w − → = → .38: x = 2 y = 2t/10 sin t. Graphing parametric equations is a difﬁcult art. Figure 1. Our interest is not in graphing curves. Figure 1. − − − − a v w ¬¬→¬¬2 w ¬¬− ¬¬ w 1. The trace of the curve r is the set of all images of r. We mention in passing however that Maple Free to photocopy and distribute 31 . →. Γ. y = (sin t)(1 + cos t)/2.5. and such that r([a.40: x = (1 + cos t)/2.8 Let →. with r(t) = .6 Parametric Curves on the Plane x(t) y(t) 54 Deﬁnition Let [a. t/10 cos t. If there exist t1 = t2 such that r(t1 ) = r(t2 ) = p.

π].y=-5. y : t → b(t) and an interval I such that the graphs of f (x.2*Pi].5. the cosine decreases from 1 to −1 and so the curve is traversed from right to left in this interval.2].. These parametrisations may differ in features. y = t.t=-20. x decreases.x=-5..5). > plot([(1-tˆ2)/(1+tˆ2). the curve is traversed from left to right.sin(t)*(1+cos(t))/2.. √ √ 2.. 55 Example Consider the parabola with Cartesian equation y = x2 . +∞[ gives the half of the curve for which x ≤ 0. √ √ 3. We will give various parametrisations for portions of this curve. For example. and so the curve is traversed from right to left.37 through 1. This works only for t ≥ 0. y = t.. > plot([2ˆ(t/10)*cos(t).2ˆ(t/10)*sin(t). the curve is traversed from left to right.. +∞[ gives the half of the curve for which x ≥ 0. the commands to graph the various curves in ﬁgures 1. y) = 0 and f (a(t).5. For t ∈ [0. according to the choice of functions and the choice of intervals. t ∈ I coincide. t ∈ [0. As t increases. then again y = t = (− t)2 = x2 . This works for every t ∈ R. we want to ﬁnd functions x : t → a(t). That is. 1. b(t)) = 0. Notice that as t increases.10].5. If x = t and y = t. and hence the parametrisation x = t. If x = t and y = t2 . 4. t ∈ [0. then clearly y = t2 = x2 .x=-5.y=-5.cos(6*t).x=-5. if x = − t and y = t.x=-5. Given a Cartesian curve with equation f (x.5). As t increases.40 follow.. Free to photocopy and distribute 32 .t=0. > with(plots): > plot([sin(2*t).Parametric Curves on the Plane has excellent capabilities for graphing parametric equations. If x = cos t and y = cos2 t. then again y = t = ( t)2 = x2 .5.. and hence the parametrisation √ x = t. and so this parametrisation only agrees with the curve y = x2 when −1 ≤ x ≤ 1..5).. Our main focus of attention will be the following. Both x and y are periodic with period 2π. we wish to ﬁnd suitable parametrisations for them.t=-2. and hence the parametrisation √ x = − t. This works only for t ≥ 0.. then again y = cos2 t = (cos t)2 = x2 .. > plot([(1+cos(t))/2. t∈R works for the whole curve. Similarly.y=-5.(t-tˆ3)/(1+tˆ2).5). y = t2 .t=0. y) = 0.2*Pi].y=-5.

y = t. Figure 1. +∞[. −2). 1]. 1].Chapter 1 Figure 1. The second parametrisation must satisfy that as t traverses the interval [0. notice that as t traverses the interval [0. ﬁnishing at (3. sin t) traverses the unit circle once. 3 x = 1 + 2 cos t. 0) and ending there. √ Figure 1. 2π]. √ t. For the second parametrisation. 2 and y+2 = sin t =⇒ y = −2 + 3 sin t.42: x = t ∈ [0. −2). y = t. Put x−1 = cos t =⇒ x = 1 + 2 cos t. a sort of “polar coordinates. clockwise. one starts at the point (3.44: x = cos t.41: x = t. −2). The ﬁrst parametrisation must satisfy that as t traverses the values in the interval [0. are often useful when parametrising quadratic curves. t ∈ R. +∞[. 2π] Then is the desired ﬁrst parametrisation. 2. cosh2 θ − sinh2 θ = 1. (x − 1)2 4 + (y + 2)2 9 = 1. (sin 4πt. y = t2 . traverses the ellipse once counterclockwise. 56 Example Give two distinct parametrisations of the ellipse 1. −2). the point (cos t. Figure 1. traverses the ellipse twice clockwise.” Observe that for t ∈ [0. t ∈ [0. Solution: ◮ What formula do we know where a sum of two squares equals 1? We use a trigonometric substitution. y = cos2 t. y = −2 + 3 sin t. cos 4πt) traverses the unit interval twice. The identities cos2 θ + sin2 θ = 1. and returns to (3. 2π]. π].43: x = − t. tan2 θ − sec2 θ = 1. starting at (1. t ∈ [0. t ∈ [0. one starts at the point (3. but Free to photocopy and distribute 33 .

Parametric Curves on the Plane begins and ends at the point (0, 1). To begin at the point (1, 0) we must π π make a shift: sin 4πt + , cos 4πt + will start at (1, 0) and 2 2 travel clockwise twice, as t traverses [0; 1]. Hence we may take x = 1 + 2 sin 4πt + π 2 , y = −2 + 3 cos 4πt + π 2 , t ∈ [0; 1]

as our parametrisation. ◭ Some classic curves can be described by mechanical means, as the curves drawn by a spirograph. We will consider one such curve.

O′ φ B P A

θ O

Figure 1.45: Construction of the hypocycloid.

Figure 1.46: Hypocycloid with R = 5, ρ = 1.

Figure 1.47: Hypocycloid with R = 3, ρ = 2.

57 Example A hypocycloid is a curve traced out by a ﬁxed point P on a circle C of radius ρ as C rolls on the inside of a circle with centre at O and radius R. If the R initial position of P is , and θ is the angle, measured counterclockwise, that 0 a ray starting at O and passing through the centre of C makes with the x-axis, shew that a parametrisation of the hypocycloid is x = (R − ρ) cos θ + ρ cos y = (R − ρ) sin θ − ρ sin Free to photocopy and distribute (R − ρ)θ ρ (R − ρ)θ ρ , . 34

Chapter 1 Solution: ◮ Suppose that starting from θ = 0, the centre O ′ of the small circle moves counterclockwise inside the larger circle by an angle θ, and the point P = (x, y) moves clockwise an angle φ. The arc length travelled by the centre of the small circle is (R − ρ)θ radians. At the same time the point P has rotated ρφ radians, and so (R − ρ)θ = ρφ. See ﬁgure 1.45, where O ′ B is parallel to the x-axis. π Let A be the projection of P on the x-axis. Then ∠OAP = ∠OP O′ = , 2 π ′ ∠OO P = π − φ − θ, ∠P OA = − φ, and OP = (R − ρ) sin(π − φ − θ). 2 Hence x = (OP ) cos ∠P OA = (R − ρ) sin(π − φ − θ) cos( y = (R − ρ) sin(π − φ − θ) sin( Now x = = = = = (R − ρ) sin(π − φ − θ) cos( π 2 − φ) π 2 − φ). π 2 − φ),

(R − ρ) sin(φ + θ) sin φ (R − ρ) (cos θ − cos(2φ + θ)) 2 (R − ρ) (cos θ + cos(2φ + θ)) (R − ρ) cos θ − 2 (R − ρ) cos θ − (R − ρ)(cos(θ + φ) cos φ). ρ OO′ =− ρ R−ρ and cos φ =

Also, cos(θ + φ) = − cos(π − θ − φ) = − cos (R − ρ)θ ρ and so

x = (R−ρ) cos θ−(R−ρ)(cos(θ+φ) cos φ = (R−ρ) cos θ+ρ cos

(R − ρ)θ ρ

,

as required. The identity for y is proved similarly. A particular example appears in ﬁgure 1.47. ◭

Free to photocopy and distribute

35

− d→ x

**Parametric Curves on the Plane → − x → + d→ − − x x
**

Figure 1.49: Area enclosed by a simple closed curve Figure 1.48: Length of a curve.

Given a curve Γ how can we ﬁnd its length? The idea, as seen in ﬁgure 1.48 is to consider the projections dx, dy at each point. The length of the vector dr = is ||dr|| = Hence the length of Γ is given by ||dr|| = (dx)2 + (dy)2 .

Γ

dx dy

(dx)2 + (dy)2 .

(1.13)

Γ

Similarly, suppose that Γ is a simple closed curve in R2 . How do we ﬁnd the (oriented) area of the region it encloses? The idea, borrowed from ﬁnding areas of polygons, is to split the region into triangles, each of area 1 2 det x x + dx y y + dy = 1 2 det x y dx dy = 1 2 (xdy − ydx),

**and to sum over the closed curve, obtaining a total oriented area of 1 2
**

Á Á

det

Γ

x y

dx dy

=

1 2

Á

Γ

(xdy − ydx).

(1.14)

Here

Γ

denotes integration around the closed curve.

Figure 1.50: Example 58.

Figure 1.51: Example 59.

Free to photocopy and distribute

36

Chapter 1 58 Example Let (A, B) ∈ R2 , A > 0, B > 0. Find a parametrisation of the ellipse Γ: (x, y) ∈ R2 : x2 A2 + y2 B2 =1 .

Furthermore, ﬁnd an integral expression for the perimeter of this ellipse and ﬁnd the area it encloses. Solution: ◮ Consider the parametrisation Γ : [0; 2π] → R2 , with x y = A cos t B sin t .

This is a parametrisation of the ellipse, for x2 A2 + y2 B2 = A2 cos2 t A2 + B 2 sin2 t B2 = cos2 t + sin2 t = 1.

Notice that this parametrisation goes around once the ellipse counterclockwise. The perimeter of the ellipse is given by

¬¬ →¬¬ ¬¬d− ¬¬ = r

2π 0

A2 sin2 t + B 2 cos2 t dt.

Γ

The above integral is an elliptic integral, and we do not have a closed form for it (in terms of the elementary functions studied in Calculus I). We will have better luck with the area of the ellipse, which is given by 1 2

Á

Γ

(xdy − ydx) = = =

1 2 1 2 1 2

Á

**(A cos t d(B sin t) − B sin t d(A cos t))
**

2π 0 2π

(AB cos2 t + AB sin2 t)) dt AB dt

.

0

= πAB. ◭ 59 Example Find a parametric representation for the astroid Γ : (x, y) ∈ R2 : x2/3 + y 2/3 = 1 , in ﬁgure 1.51. Find the perimeter of the astroid and the area it encloses. Solution: ◮ Take x y Free to photocopy and distribute = cos3 t sin3 t 37

**Parametric Curves on the Plane with t ∈ [0; 2π]. Then x2/3 + y 2/3 = cos2 t + sin2 t = 1. The perimeter of the astroid is
**

2π Γ

||dr|| = = =

9 cos4 t sin2 t + 9 sin4 t cos2 t dt 3| sin t cos t| dt

0 2π 0 2π 0 π/2 0

3 2

| sin 2t| dt sin 2t dt

= 6 = 6.

**The area of the astroid is given by 1 2
**

Á

Γ

(xdy − ydx) = = = = = =

1 2 1 2 3 2 3 8 3

Á

**(cos3 t d(sin3 t) − sin3 t d(cos3 t))
**

2π 0 2π 0 2π 0

(3 cos4 t sin2 t + 3 sin4 t cos2 t)) dt (sin t cos t)2 dt (sin 2t)2 dt (1 − cos 4t) dt

2π 0

16 3π 8

.

We can use Maple™ (at least version 10) to calculate the above integrals. For example, if (x, y) = (cos3 t, sin3 t), to compute the arc length we use the path integral command and to compute the area, we use the line −y/2 integral command with the vector ﬁeld . x/2 > with(Student[VectorCalculus]): > PathInt( 1, [x,y]=Path( <(cos(t))ˆ3,(sin(t))ˆ3>,0..2*Pi)); > LineInt( VectorField(<-y/2,x/2>), Path( <(cos(t))ˆ3,(sin(t))ˆ3>,0..2*Pi)); ◭ We include here for convenience, some Maple commands to compute various arc lengths and areas. 60 Example To obtain the arc length of the path in ﬁgure 1.52, we type > with(Student[VectorCalculus]): > PathInt(1, [x,y] = LineSegments( <0,0>, <1,1>, <1,2> ,<2,1>,<3,3>,<4,1>); To obtain the arc length of the path in ﬁgure 1.53, we type Free to photocopy and distribute 38

4 Parametrise the curve y = π log cos x for 0 ≤ x ≤ . 6 5 5 4 3 2 1 1 2 3 4 5 Figure 1.52: Line Path. y = 2 + (1 + cos t)(sin t). 1. [x. In each case give an implicit Cartesian equation for the parametric representation and describe the trace of the parametric curve. y = b sinh t 4.54 > with(Student[VectorCalculus]): > LineInt( VectorField(<-y/2. 4π]. angle ≤θ≤ . y(t) = 1 − t2 . and ﬁnd its length.5 Describe the trace of the parametric curve æ é x y æ = sin t é .6.(1+cos(t))*(sin(t))+2>. y= .. t ∈ [0.y] = Arc( Circle( <0. 3). 1 −1 −1 1 Figure 1.53: Arc of circle of π π radius 3. Problem 1.3 Let a. y(t) = t3 + 2t.55: Problem 1.7 Consider the graph given parametrically by x(t) = t3 + 1.6. d be strictly positive real constants.2 Give an implicit Cartesian equation for the parametric representation t2 t3 x= .6 Consider the plane curve √ √ deﬁned implicitly by x + y = 1.6. [ 2 2 Problem 1. The graph of the curve appears in ﬁgure 1. > Path( <(1+cos(t))*(cos(t))+1. y = b tan t.6. 5 4 3 2 1 1 2 3 4 5 Figure 1. Pi/5 ) . b.0>. Give a suitable parametrisation of this curve.1 A curve is represented parametrically by x(t) = t3 − 2t. x = cos t.6. x = a cosh t. To obtain the area inside the curve in 1. Pi/6.55. t ∈] − π π . Problem 1. over Free to photocopy and distribute 39 . y = ct + d 2.x/2>).2*Pi)). Find its Cartesian equation. x = at + b. Then ﬁnd its arc 3 length. Homework Problem 1.54: x = 1 + (1 + cos t)(cos t).0. > > 5 4 3 2 1 1 2 3 4 5 Figure 1. Problem 1. c. 1 + t5 1 + t5 Problem 1.6.6. y = 0 3.6. Find the area under the graph. 2 sin t + 1 Problem 1. x = a sec t.Chapter 1 with(Student[VectorCalculus]): PathInt(1.6.

Free to photocopy and distribute 40 .6. Problem 1. 2 2 P d θ C ρ B ρ.Parametric Curves on the Plane the x axis. The curve traced out by P as the circle rolls along a straight line. The curve appears in ﬁgure 1.6.6. is called a cycloid.57: Problem 1. known that the shell was ﬁred from a is 1 + t3 1 + t3 gun on the ground with a muzzle velocwhich you may see in ﬁgure 1.57.11 Let C be the curve in R deﬁned by 2 2 Figure 1. y = et sin t. 2 y(t) = (2t + 1)3/2 .6.6.16 The parabola y 2 = −4px rolls without slipping around the parabola y 2 = 4px.8 Find the arc length of the curve given parametrically by x(t) = 3t2 . Find the length of this curve.56: Problem 1.9 Let C be the curve in R2 deﬁned by x(t) = t2 .13 Find the arc length of the arc of the cycloid x = ρ(t − cos t). Problem 1.10 Find the area enclosed by the curve x(t) = sin3 t. 2π]. π]. Problem 1. known.58: Cycloid Problem 1.12 Let P be a point at a distance d from the centre of a circle of radius Problem 1. y(t) = (cos t)(1 + sin2 t). 3 1 1 t ∈ [− . Find the equation of the locus of the vertex of the rolling parabola.11. Figure 1. Find a parametrisation of the cycloid. + ].10. Problem 1. y(t) = 2t3 for 0 ≤ t ≤ 1. t ∈ [0. and between the lines x = 1 and x = 2. without slipping. Problem 1. but the position of the gun and its angle of elevation are both unarea enclosed by the loop of this curve.56. Problem 1.15 A shell strikes an airplane ﬂying at height h above the ground. Problem 1. y = ρ(1 − cos t).6.6.6.6.6. Deduce that the gun is situated within a circle whose centre lies directly below the airplane and whose radius is 1 Ô V V 2 − 2gh . y(t) = . g −1 −1 1 Figure 1.6. It 3t 3t x(t) = . t ∈ [0. Find the ity of magnitude V . t ∈ R\{−1}.14 Find the length of the parametric curve given by x = et cos t.

Free to photocopy and distribute 41 . Most of what we did in R2 transfers to R3 without major complications. Let us explain.60. z) = x i + y j + z k . 1 ¾ ¿ → − → − → − r = (x. y. z x y Figure 1. 61 Deﬁnition The 3-dimensional Cartesian Space is deﬁned and denoted by In ﬁgure 1. 0 and observe that 1 ¾ ¿ → − j = 1 .Chapter 1 1. In analogy to R2 we put ¾ ¿ → − i = 0 . we have a choice for the orientation of the the x and y-axis. We adopt a convention known as a right-handed coordinate system. 1.59 we have pictured the point (2.7 Vectors in Space R3 = {r = (x. y ∈ R. as in ﬁgure 1.59: A point in space. 0 0 0 → − k = 0 . 3). Having oriented the z axis upwards. y. z ∈ R} . z) : x ∈ R.

The Cauchy-Schwarz-Bunyakovsky Inequality takes the form ¬¬− ¬¬¬¬→¬¬ − ¬¬− ¬¬ − → |→• b | ≤ ¬¬→¬¬¬¬ b ¬¬ =⇒ |a1b1 +a2 b2 +a3 b3 | ≤ (a2 +a2 +a2 )1/2 (b2 +b2 +b2 )1/2 .61: Right Hand. z be positive real numbers such that x2 + 4y 2 + 9z 2 = 27. By 42 Free to photocopy and distribute . Solution: ◮ Since x. 63 Example Let x. Figure 1. z are positive. the dot product satisﬁes →• b = ¬¬→¬¬¬¬ b ¬¬ cos θ. y.62: system. − − a b 1 1 2 2 3 3 − The norm of a vector → in R3 is a Ô ¬¬→¬¬ − − ¬¬− ¬¬ = →•→ = a a a (a1 )2 + (a2 )2 + (a3 )2 . Maximise x + y + z. |x + y + z| = x + y + z. y. Left-handed ℓ1 Figure 1.→ − k → − k Vectors in Space → − j → − j → − i → − i Figure 1. ¬¬− ¬¬¬¬→¬¬ − ¬¬− ¬¬ − → a Just as in R2 . ℓ1 and → − − 62 Deﬁnition The dot product of two vectors → and b in R3 is a →•→ = a b + a b + a b .60: Right-handed ℓ3 system. ℓ2 .63: ℓ1 ℓ3 are skew. ℓ 2 Figure 1. π] a is the convex angle between the two vectors. where θ ∈ [0. a a 1 2 3 1 2 3 equality holding if an only if the vectors are parallel.

z= √ 2 3 7 . The desired equation is à à x y z ◭ Free to photocopy and distribute = 1 2 3 ¾ ¿ 3 3 +t 3 . v à 1 65 Example Find the parametric equation of the line passing through à 2 3 and −2 −1 0 .z = λ 9 =⇒ λ + 2 λ2 4 + λ2 9 = 27 =⇒ λ = ± √ 18 3 7 . 2 Equality occurs if and only if ¾ x 2y 3z ¿ ¾ 1 ¿ = λ 1/2 1/3 =⇒ x = λ. Solution: ◮ The line follows the direction ¾ 1 − (−2) 3−0 ¿ ¾ ¿ 3 3 2 − (−1) = 3 . y= √ 9 3 14 . The → is the set − v − r ∈ R3 : r = a + t→ . Therefore for a maximum we take x= ◭ − 64 Deﬁnition Let a be a point in R3 and let → = v parametric line passing through a in the direction of √ 18 3 7 . 43 . ¬ ¬ |x+y+z| = ¬x + 2y ¬ 1 2 + 3z 1 1 ¬ 1 ¬ ≤ (x2 +4y 2 +9z 2 )1/2 1 + + ¬ 3 4 9 ¬ 1/2 √ 7 = 27 6 √ 7 3 = . y = λ 4 .Chapter 1 Cauchy’s Inequality. → − 0 be a vector in R3 .

A plane is determined by three non-collinear points. t ∈ R} = {λ→ : λ ∈ R}.Vectors in Space Given two lines in space. b arbitrary scalars. In the case at hand the vectors are in space. one over the other. − − − u v w − → and →. is coplanar with both v w t → and → can be uniquely expressed in the form − − coplanar with both v w → − − − t = p→ + q → . →. they still span a plane.64. abusing language. one of the following three situations might arise: (i) the lines intersect at a point. a We saw in the preceding chapter that if the vectors were on the plane. Conversely. v w See ﬁgure 1. We will say. (iii) the lines are skew (non-parallel. If → b . Suppose à a. All the above gives the following result. − From the above theorem. → b } a a is a plane passing through the origin. → in R3 be non-parallel vectors. without intersecting. again abusing language. c . c − − → − 67 Theorem Three vectors →. lying on different planes). that a vector is parallel to a speciﬁc plane or that it lies on a speciﬁc plane if there exists a bi-point representative of the vector that lies on the particular plane. t ∈ R. (ii) the lines are parallel. This prompts the following theorem. passing through the origin. any vector → − − a. Thus → − − − − → {s→ + t b : s ∈ R. then linear combinations of these three vectors may lie outside the → − − plane containing b . b. − − − 66 Theorem Let →. and that x c are non-collinear points on the same plane and that r = y z Free to photocopy and distribute 44 is another . if a vector → is not a linear combination of two other a → → − − vectors b . a a → − − which is a line through the origin. they would span the whole plane R2 . → − − − − → Consider now two non-zero vectors → and b in R3 .63. → in R3 are said to be linearly independent if a c → − → − − − α→ + β b + γ → = 0 =⇒ α = β = γ = 0. Suppose now that → and b are not parallel. b . We will say. a c Any vector in R3 can be written as a linear combination of three linearly independent vectors in R3 . then the set a a → − − − {s→ + t b : s ∈ R. Then every vector → of the v w u form → = a→ + b→. See ﬁgure 1. that two vectors are coplanar if there exists bi-point representatives of the vector that lie on the same plane.

and parallel to the vectors → and → is u v given by → − → = p→ + q → . Since a. − → = → + p(→ − →) + q(→ − →). Free to photocopy and distribute 45 . (u3 v1 − u1 v3 )(y − a2 ) = (u3 v1 − u1 v3 )(pu2 + qv2 ). Since ax Lemma 66. Multiplying the ﬁrst equation by u2 v3 − u3 v2 . (u2 v3 − u3 v2 )(x − a1 ) = (u2 v3 − u3 v2 )(pu1 + qv1 ). the second by u3 v1 − u1 v3 . − − → − where →. and c. and the third by u1 v2 − u2 v1 . y − a2 = pu2 + qv2 . − − 68 Theorem Let → and → be linearly independent vectors. b.64: Theorem 66. and c are non-collinear. − − − − r a u v Componentwise this takes the form x − a1 = pu1 + qv1 . Thus we have the a c following theorem. The parametric equau v − − tion of a plane containing the point a. By Chasles’ Rule.Chapter 1 − → − arbitrary point on this plane. ab and → ac. − − − − − r a b a c a is the equation of a plane containing the three non-collinear points a. →= − n ¾ ¿ a b c → − v − − p→ + q → v w ax + by + cz = d → − x → − w Figure 1. we have by → which are coplanar. are non-parallel. q with → → = p− + q → − − ar ab ac. b. (u1 v2 − u2 v1 )(z − a3 ) = (u1 v2 − u2 v1 )(pu3 + qv3 ). we obtain.65: Theorem 69. z − a3 = pu3 + qv3 . Figure 1. that there exist real numbers p. − also lies on the plane. b . and → are the position vectors of these points.

c is non-zero. Then x− y z d a á ¾ =s − b¿ +t ¾ a 1 0 − c¿ a 0 1 is a parametric equation for the plane. (u2 v3 − u3 v2 )(x − a1 ) + (u3 v1 − u1 v3 )(y − a2 ) + (u1 v2 − u2 v1 )(z − a3 ) = 0.Vectors in Space Adding gives. c are zero. observe that if → and → are non-parallel vectors and → − u v r → = p→ + q → is the equation of the plane containing the point a and − − − a u v − − − parallel to the vectors → and →. Moreover. − − − − − − − − − ular to u v r a n u n v n since at least one of a. The argument is similar if one of the other letters is non-zero and a = 0.u Free to photocopy and distribute 46 b a + b (1) + c (0) = 0. d = a1 (u2 v3 − u3 v2 ) + a2 (u3 v1 − u1 v3 ) + a3 (u1 v2 − u2 v1 ). b = u3 v1 − u1 v3 . a a a Put y = s and z = t. Proof: We have already proved the ﬁrst statement. We have a − a b c ing the second statement. Put and a = u2 v3 − u3 v2 . at ¾ ¿ a →= − least one of the coefﬁcients is non-zero. we may assume a = 0. c = u1 v2 − u2 v1 . prov1 1 0 . For the second − − − statement. then if → is simultaneously perpendicu v n → and → then (→ − →)•→ = 0 for →•→ = 0 = →•→. a − c a + b (0) + c (1) = 0. This gives the − − Since v u following theorem. ¾ ¾ ¿ and so is simultaneously perpendicular to − ¾ c¿ b¿ − a a and 0 . In this case we can see that d b c x = − y − z. b. 69 Theorem The equation of the plane in space can be written in the form ax + by + cz = d. not all of a. the vector n b is normal c to the plane with Cartesian equation ax + by + cz = d. Now. that is. → is linearly independent from →. b. Here a2 + b2 + c2 = 0. which is the Cartesian equation of the plane.

◭ − − 72 Deﬁnition If → is perpendicular to plane Π1 and →′ is perpendicular to plane n n − Π2 . 71 Example Find both the parametric equation and the Cartesian equation of ¾ ¿ ¾ ¿ 1 1 the plane parallel to the vectors 1 and 1 and passing through the point à 1 . − n 73 Example Free to photocopy and distribute 47 .Chapter 1 à 1 70 Example The equation of the plane passing through the point ¾ −1 2 and normal to the vector −3 2 4 ¿ is −3(x − 1) + 2(y + 1) + 4(z − 2) = 0 =⇒ −3x + 2y + 4z = 3. 0 0 −1 2 Solution: ◮ The desired parametric equation is à x y+1 z−2 This gives s = z − 2. and ¾ ¿ 1 1 ¾ ¿ 1 0 =s 1 +t 1 . the angle between the two planes is the angle between the two vectors → and n →′ . t=y+1−s=y+1−z+2=y−z+3 x = s + t = z − 2 + y − z + 3 = y + 1. Hence the Cartesian equation is x − y = 1.

This appears in ﬁgure 1. √ 2 + 12 · 2 + 12 2 3 1 1 4. 1. Find the angle between the planes z = 1 − x and z = 1 − y. This appears in ﬁgure 1. and 1 hence the volume required is .67: The plane z = 1 − y. Draw the solid S which results from the intersection of the planes z = 1−x and z = 1 − y with the ﬁrst octant. If θ is the angle between these two vectors. 3 ◭ Homework Free to photocopy and distribute 48 . The vector 0 is normal to the plane x + z = 1. Solution: ◮ 1.66. Figure 1. Find the volume of the solid S . Draw the intersection of the plane z = 1 − x with the ﬁrst octant. Figure 1. 5. 2. ¾ ¿ 1 ¾ ¿ 0 3.z z z Vectors in Space x y x y x y Figure 1.68. 4. and the vector 1 1 1 is normal to the plane y + z = 1.66: The plane z = 1 − x.67. Recall that the volume of a pyramid is given by the Ah formula V = . then 1·0+0·1+1·1 1 π cos θ = √ =⇒ cos θ = =⇒ θ = . 3. This appears in ﬁgure 1. Now. the height of this pyramid is clearly 1.68: Solid bounded by the planes z = 1 − x and z = 1 − y in the ﬁrst octant. 5. Draw the intersection of the plane z = 1 − y with the ﬁrst octant. 2. The resulting solid is a pyramid with square base of area A = 1 · 1 = 1. where A is area of the base of the pyramid and 3 h is its height.

6 It is known that ¬¬→ ¬¬ = 3. Problem 1.Chapter 1 Problem 1. b. Problem 1. z = 1−t. Problem 1. a ¬¬ b ¬¬ ¬¬ ¬¬ → ¬¬ − ¬¬− Find ¬¬→ − b ¬¬.2 Find theǑ equation of the 1 line passing through 2 3 in the direction ¾ of ¿ −2 0 −1 .7 Find the equation of the line perpendicular to the plane ax + a2 y + a3 z = 0. Find the angle between them.7. 2. b.7. Prove that f (a. c > 0 be the lengths of the sides of △ABC. a = 0 and passing through the Free to photocopy and distribute 49 .7. (Vertex A is opposite to the side measuring a. a ¬¬ ¬¬ ¬¬→ ¬¬ − − ¬¬→¬¬ − ¬¬ = 5 and that → + → + − a b ¬¬ b ¬¬ = 4. → satisfy − b ¬¬ ¬a ¬ − ¬¬→ →¬¬ − ¬¬ a + b ¬¬ = 24. Problem 1.7. and ﬁnd this maximum. etc. c > 0.12 Let a > 0. Prove that (a2 + b2 + c2 )2 ≤ 3(a4 + b4 + c4 ).7. a − → .7. ¬¬→¬¬ = 19. b. c be arbitrary real numbers. c) ¬¬− ¬¬ Problem 1. A ⊆ B). intersect. c. y = −2t. Problem 1.9 Find the (shortest) distance from the point (1. ¬¬ c − → = →. b.7. where a > 0.7. 1 Ǒ Ǒ ¾ 2 1 ¿ + t −1 . Let B be the set of all points which are at distance at most 1 from some point of A (in particular. S(a. 3) to the plane x − y + z = 1.11 Let a. b. b > 0. Ǒ .7. Problem 1.) Recall that by Heron’s Formula. Find − c 0 − − → •→ + → •→ + → •→ . c) is maximised when △ABC is a 2 + b2 + c 2 equilateral.5 (Putnam Exam 1984) Let A be a solid a × b × c rectangular brick in three dimensions.7. Problem 1. −1 and contain- ¾ ¿ 2 +t 1 . 0 point 0 1 Problem 1. the area of this triangle is = s(s − a)(s − b)(s − c).8 Find the equation of the plane perpendicular to the line ax = by = cz. a+b+c where s = is the semiperime2 ter of the triangle.7. b > 0. Ǒ = 0 and passing through the point abc 1 1 1 Problem 1.3 Find the Ǒ equation of plane 1 containing the point 1 and perpendic1 ular to the line x = 1+t. − − − − a b b c c a Problem 1.4 Find the equation of plane Ǒ 1 containing the point −1 ing the line x = 2y = 3z.7.10 Determine lines Ǒ Ǒ x 1 L1 : y z x L2 : y z = 1 1 0 = 0 1 whether the in R3 .1 Vectors ¬¬ ¬¬ ¬¬→¬¬ − ¬¬− ¬¬ = 13. c) = S(a. Express the volume of B as a polynomial in a.

Problem 1.19 Given a polyhedron with n faces. y. Find the volume of the tetrahedron with vertices at Ǒ Ǒ Ǒ Ǒ 0 1 0 0 0 0 . y. and let α ∈ R be a scalar. Determine the number of vertices and the number of edges of S. Prove that √ √ √ √ x+y+ y+z+ z+x ≤ 6. each normal to a face of the polyhedron. 2x + 3z ≤ 24.18 (Putnam Exam 1980) Let S be the solid in three-dimensional space Ǒ x consisting of all points y satisfying the z following system of six conditions: x ≥ 0. Prove that x+y+z ≤ 2 x2 y2 z2 + + y+z z+x x+y . 0 0 . Draw this plane and its 0 1 intersection with the ﬁrst octant. then the distance from b to the plane is ¬ ¬ − − ¬ → → •→ ¬ − ¬( a − b ) n ¬ ¬¬→¬¬ .8 Cross Product We now deﬁne the standard cross product in R3 as a product satisfying the following properties. 1 0 and 0 1 . consider n vectors. − − − 74 Deﬁnition Let →. → be vectors in R3 . − − − Problem 1.13 Let x.17 Let (→ − →)•→ = 0 be a r a n plane passing through the point a and per− pendicular to vector →. z be strictly positive numbers.14 Let x.7.7. →. z be strictly positive numbers. Prove that the sum of → − these vectors is 0 . 2x + 4y + 3z ≤ 36. Ǒ 0 1 and 0 0 Ǒ . If b is not a point n on the plane.7.7. ¬¬− ¬¬ n Problem 1. The cross x y z product × is a closed binary operation satisfying − − − − Anti-commutativity: → × → = −(→ × →) x y y x Bilinearity: − − − − − − − (→ + →) × → = → × → + → × → and x z y x y z y → × (→ + →) = → × → + → × → − − − − − − − x z y x z x y → − − − →×→= 0 x x − − − − − − Scalar homogeneity: (α→) × → = → × (α→) = α(→ × →) x y x y x y Free to photocopy and distribute 50 . y ≥ 0. z ≥ 0. Problem 1. Problem 1. x + y + z ≤ 11. and length equal to the area of the face. √ x+y+z Problem 1.7.15 Find the Cartesian equaǑ 1 tion of the plane passing through 0 0 .7.16 Prove that there do not exist three unit vectors in R3 such that the 2π angle between any two of them be > .Cross Product Problem 1.7. 3 1.

It follows that the cross product is an operation that. j × k = i . allows us to “get out” of that plane. given two non-parallel vectors on a plane.Chapter 1 Right-hand Rule: → → − − → → → − − − → → → − − − → − i × j = k. Operating as in example 75 we obtain ¾ − 76 Theorem Let → = x2 x x3 x1 ¿ ¾ − and → = y2 y y3 y1 ¿ be vectors in R3 . Then − − − → × → = (x y − x y )→ + (x y − x y )→ + (x y − x y )→. Solution: ◮ We have → − → − → − → − ( i −3k)×( j +2k) = = = → → − − → → − − → → − − → → − − i × j +2 i × k −3k × j −6k × k → − → − → − → − k −2 j +3 i +60 → − → → − − 3 i −2 j + k. − − x y 2 3 3 2 i 3 1 1 3 j 1 2 2 1 k Free to photocopy and distribute 51 . since → − → → − − → → − − → − i ×( i × j )= i × k =− j → → − − → − − → → − → − ( i × i )× j = 0 × j = 0. 75 Example Find ¾ ¿ ¾ ¿ 1 0 −3 0 2 × 1 . k × i = j . ¾ ¿ Hence ¾ 1 0 −3 ¿ 0 2 ¾ 3 1 ¿ × 1 = −2 . ◭ but The cross product of vectors in R3 is not associative.

Proof: We will only check the ﬁrst assertion. = x 1 x 2 y3 − x 1 x 3 y2 + x 2 x 3 y1 − x 2 x 1 y3 + x 3 x 1 y2 − x 3 x 2 y1 completing the proof. 78 Theorem − → × (→ × →) = (→•→)→ − (→•→)→.69: Theorem 79. the following theorem. Figure 1. that is. obtaining. we have.70: Area of a parallelogram → → − − → → − − → → − − → − Proof: Since i × i = j × j = k × k = 0 . however. u Using the cross product. the second veriﬁcation is analogous. →•(→ × →) − − − x x y → − → − → − → − = (x1 i + x2 j + x3 k )•((x2 y3 − x3 y2 ) i → − → − +(x3 y1 − x1 y3 ) j + (x1 y2 − x2 y1 ) k ) = 0. u Although the cross product is not associative. →×→ − − x y → − → − → − → − → − → − = (x1 i + x2 j + x3 k ) × (y1 i + y2 j + y3 k ) → → − − → → − − → → − − → → − − = x 1 y2 i × j + x 1 y3 i × k + x 2 y1 j × i + x 2 y3 j × k → → − − → → − − +x3 y1 k × i + x3 y2 k × j → → − − → → − − → → − − = (x1 y2 − y1 x2 ) i × j + (x2 y3 − x3 y2 ) j × k + (x3 y1 − x1 y3 ) k × i → − → − → − = (x1 y2 − y1 x2 ) k + (x2 y3 − x3 y2 ) i + (x3 y1 − x1 y3 ) j . we only worry about the mixed products.×→ − y ¬ ¬¬¬→ ¬¬ − y ¬¬ → − x Cross Product θ − → x → − y ¬¬→¬¬ ¬¬− ¬¬ x Figure 1. − − − − − − 77 Theorem → ⊥ (→ × →) and → ⊥ (→ × →). − − − − − − − − a b c a c b a b c Free to photocopy and distribute 52 ¬¬→¬¬¬¬→¬¬ ¬¬− ¬¬¬¬− ¬¬ sin θ x y . proving the theorem. we may obtain a third vector simultaneously perpendicular to two other vectors in space. the cross product of two x x y y x y vectors is simultaneously perpendicular to both original vectors.

a c a completing the proof. C y − − − 79 Theorem Let (→. π] be the convex angle between two vectors → and x y x →. →).71: Theorem 82. →) ∈ [0.Chapter 1 Proof: − → × (→ × →) = − − a b c = = = → − → − → − → − (a1 i + a2 j + a3 k ) × ((b2 c3 − b3 c2 ) i + → − → − +(b3 c1 − b1 c3 ) j + (b1 c2 − b2 c1 ) k ) → − → − → − a1 (b3 c1 − b1 c3 ) k − a1 (b1 c2 − b2 c1 ) j − a2 (b2 c3 − b3 c2 ) k → − → − → − +a2 (b1 c2 − b2 c1 ) i + a3 (b2 c3 − b3 c2 ) j − a3 (b3 c1 − b1 c3 ) i → − → − → − (a1 c1 + a2 c2 + a3 c3 )(b1 i + b2 j + b3 i )+ → − → − → − (−a1 b1 − a2 b2 − a3 b3 )(c1 i + c2 j + c3 i ) → − → → − − − − − (→•→) b − (→• b ) c . Figure 1. Then − y − − − − − − ||→ × →|| = ||→||||→|| sin (→. x y x y x y Free to photocopy and distribute 53 . u z P D′ − →×→ − a b A′ → − c D x B′ C′ N θ −b → A → − a M B Figure 1.72: Example 83.

The area of the base of the parallelepiped is the → − − area of the¬ parallelogram determined by the vectors → and b . The volume of the paral− − where θ is the angle between c a b lelepiped is thus ¬¬ →¬¬¬¬→¬¬ − ¬¬ − − ¬¬→ − → → − → − ¬¬ a × b ¬¬¬¬ c ¬¬ cos θ = ( a × b ) • c . The signed a c → − − − volume of the parallelepiped spanned by them is (→ × b ) • →. it follows → − → − − → − − − − − − (→ × b ) • → = ( b × →) • → = (→ × → ) • b . x y x y The following result mixes the dot and the cross product. See ﬁgure 1. The following corollaries easily follow from Theorem 79. a c c a c a In particular. 81 Corollary (Lagrange’s Identity) − − − − ||→ × →||2 = ||x||2 ||y||2 − (→•→)2 . →) x y x y x y →||2||→||2 sin2 (→. b . a c Proof: See ﬁgure 1. → − − − − − 80 Corollary Two non-zero vectors →. u that Since we may have used any of the faces of the parallelepiped. it is possible to “exchange” the cross and dot products: − − → • (→ × → ) = (→ × →) • → − − − − a b c a b c Free to photocopy and distribute 54 .71. The altitude of the parallelepiped is ¬¬− ¬¬ cos θ a c − → and → × →. − − − − = || x y x y whence the theorem follows. − − → − 82 Theorem Let →. proving the theorem. → satisfy → × → = 0 if and only if they x y x y are parallel. →. →). u 2 2 2 = (x2 + y 2 + z 2 )(y1 + y2 + y3 ) − (x1 y1 + x2 y2 + x3 y3 )2 − − − − = ||→||2||→||2 − (→•→)2 x y x y ¬¬− − ¬¬ Theorem 79 has the following geometric signiﬁcance: ¬¬→ × →¬¬ is the area x y of the parallelogram formed when the tails of the vectors are joined.Cross Product Proof: We have − − ||→ × →||2 x y = (x2 y3 − x3 y2 )2 + (x3 y1 − x1 y3 )2 + (x1 y2 − x2 y1 )2 2 2 2 = y 2 y3 − 2x2 y3 x3 y2 + z 2 y2 + z 2 y1 − 2x3 y1 x1 y3 + 2 2 2 +x2 y3 + x2 y2 − 2x1 y2 x2 y1 + y 2 y1 − − − − − − = ||→||2||→||2 − ||→||2 ||→||2 cos2 (→. which a ¬ ¬¬ ¬¬→¬¬ →¬¬ − ¬¬− has area ¬¬→ × b ¬¬. be linearly independent vectors in R3 .70.

−6 −3 ¿ The equation of the plane is thus ¾ x−2 z−0 ¿ ¾ y − 0 • −1 = 0 =⇒ 3(x−2)+1(y)+6z = 0 =⇒ 3x+y+6z = 6. 0). The area of the triangle is ¬¬− → −→¬¬ − ¬¬ ¬¬ − ′ ¬¬AD × AM¬¬ 2 ¾ 1Ô 2 = 3 + 12 + 62 = 2 √ 46 . B(2. B ′ (2. ← → 3. y = 3t. 1). −6 −3 ¿ 2. and hence the line AC′ has parametric à x y z = 2 0 0 +t ¾ −2 3 1 ¿ =⇒ x = 2 − 2t. −→ − AM = −1 3 0 ¿ ¾ − → −→ − − =⇒ AD′ × AM = −1 . 3. C(0. and M. Find the parametric equation of MP. 2 −→ − 3. Let M be the midpoint of the line segment joining the vertices B and C. 1.Chapter 1 83 Example Consider the rectangular parallelepiped ABCDD ′C ′ B ′ A′ (ﬁgure 1. Find the area of △AD ′ M . 3. Form the following vectors and ﬁnd their cross product: ¾ −→ − AD′ = −2 0 1 ¿ ¾ . 0. Free to photocopy and distribute 55 . A′ (2. 4. 3. z = t. Suppose that a line through M is drawn cutting the line segment [AC ′ ] in ← − → ← → N and the line DD′ in P . 1). 3. Find the Cartesian equation of the plane containing the points A. 0). 0). 1). D ′ . 0). Find the parametric equation of the line AC′ . C ′ (0. We have AC′ = equation à −2 3 1 ¿ ← → . 0. D ′ (0. 2. 1). D(0. Solution: ◮ 1. 0. 0.72) with vertices A(2.

x y z = 1 3 0 +s −3 z′ −1 =⇒ x = 1−s y = 3−3s.5 Find the equation of the plane passing through the points (a. Find a vector of unit length simultaneously perpendicu- Free to photocopy and distribute 56 . 1. and (0.8. 0.8.t = 2 3 . t = sz ′ . (−a. Problem 1. z = sz ′ . a). 0). Putting this into à 0 the third equation we deduce z ′ = 2. P = 0 z′ ← → The parametric equation of the line MP is thus à à ¾ ¿ for some real number z ′ > 0. y = 3−3s. − − i − j and a b j k b c Problem 1. ﬁnd → × → .8. ← → ← → Since N is on both MP and AC′ we must have 2 − 2t = 1 − s. Problem 1.1 Prove that → − → − → − − − − (→ − b ) × (→ + b ) = 2→ × b . a a a → − − − Problem 1. z = 2s.8.2 Prove that → × → = 0 x x follows from the anti-commutativity of the cross product. −1. → − − − − Problem 1.8. 2a) in R3 . by ﬁnding a normal to the plane. 1.4 Redo example 71. Since P is on the z-axis. Solving the ﬁrst two equations gives s = 1 3 . that is. ﬁnd the Cartesian equation of the plane parallel to the vectors ¾ ¿ 1 1 and ¾ ¿ 1 1 and 1 0 passing through the point (0.3 If b − → and → − → are a c a − − parallel and it is known that → × → = c a → → − − − − − − → × → = → + →. Homework Problem 1. 3t = 3 − 3s.Cross Product à 0 4.8. Thus P = desired equation is à à 0 2 and the x y z ◭ = 1 3 0 ¾ +s −3 2 −1 ¿ =⇒ x = 1−s.6 Let a ∈ R. 2).

4. The plane Π intersects a 3 × 3 × 3 cube.12 The vectors →.8. → are a c constant vectors. 5.8. Find the coordinates of the points P . as in the ﬁgure. Find the parametric equation of the line LD E joining D and E. Find the intersection point between the lines LC A and LD E . x x x − − → Problem 1.8. x y a y x a c − − − → − → Problem 1. Find x → 2 − → − → − − − − ||→ × i || + ||→ × j ||2 + ||→ × k ||2 . This b . Find the CP QRS. − lar to → = v −a − and → = w a . 3).8 Let → ∈ R3 . β. b are cona stant vectors. d be veca c 3 tors in R . Find CA × CB. b . as shewn in ﬁga 0 ure 1. Find the parametric equation of the line LC A joining C and A. ||x|| = 1.15 Consider the plane Π passing through the points A(6.Chapter 1 Problem 1. − − − − − − − → − → − − → → − → → − c (→ × b )•(→ × d ) = (→ •→)( b • d )−(→ • d )( b •→).73 below. Find the equation of the plane Π.73: Problem 1. Prove the following vector identity. with a parameter t ∈ R.11 Assume → • ( b × →) = 0 a c and that − → = α→ + β → + γ →. 0). Solve the system of equations → − − − − − − − − 2→ + → × → = b . − a c a c a 0. − Problem 1. c be vectors in R3 . 3. 7.8. be veca c 3 tors in R . and S. 0.8.13 Let → . a c − → − → 1. Solve the equation − − → × (→ × →) = → × (→ × →). 0) and C(0. 8.. area of the pentagon − − → − Problem 1. → . 9.15. Q. d .8. with a parameter s ∈ R. R. Prove that → → → → → → → → − − − − → − → ×( b × c )+ b ×(− ×− )+− ×(− × b ) = intersection forms a pentagon CP QRS. the origin and that has three of its edges on a → → − − the coordinate axes. − − − x a b c → − − − Find α. 6. 0. 3→ + → × → = →. one of whose vertices is at − Problem 1.8. and γ in terms of → • ( b × →). B(0.10 If → + b + → = 0 . Figure 1.8.9 The vectors → . →. − − − − a b b c c a → − − − Problem 1. Prove that → − − → − xA − ( b × →) • (→ × d ) c a → → − − − − +(→ × → ) • ( b × d ) c a → − → − − − +(→ × b ) • (→ × d ) a c = S Q R B y 0.14 Let → . Find ¬¬CA × CB¬¬.8.7 (Jacobi’s Identity) Let → . Find the area of △ABC. 4. ¾ 0 ¿ ¾ ¿ 1 ¬¬ ¬¬ → − ¬¬ → ¬¬− 2. prove a c that − − → × → = → × → = → × →. b . a c a c a z C P − − − → − → Problem 1.8. − − − − a x b b x a − → − − → − Problem 1. b . Free to photocopy and distribute 57 .

with àà à x1 L x2 x3 = x1 − x2 − x3 x1 + x2 + x3 x3 .9 Matrices in three dimensions We will brieﬂy introduce 3 × 3 matrices. and T (k). 85 Example Consider L : R3 → R3 . b in R3 and all scalars λ. Then àà u1 + v1 L(u + v) = L u2 + v2 u3 + v3 = à (u1 + v1 ) − (u2 + v2 ) − (u3 + v3 ) (u1 + v1 ) + (u2 + v2 ) + (u3 + v3 ) 3 à u + v3 + v1 − v2 − v3 v1 + v2 + v3 3 àà à = u1 − u2 − u3 u1 + u2 + u3 u3 u1 àà v v1 +L v2 v3 = = L u2 u3 L(u) + L(v).Matrices in three dimensions 1. Prove that L is a linear transformation. v be points in R3 . for all points a. Most of the material will ﬂow like that for 2 × 2 matrices. T (λa) = λT (a). T (j). Such a linear transformation has a 3 × 3 matrix representation whose columns are the vectors T (i). Solution: ◮ Let α ∈ R and let u. Find the matrix corresponding to L under the standard basis. 58 Free to photocopy and distribute . 84 Deﬁnition A linear transformation T : R3 → R3 is a function such that T (a + b) = T (a) + T (b).

à à à 1 We have L à 1 = 1 0 . Free to photocopy and distribute 59 . whence the desired matrix is ¾ 1 1 0 −1 1 0 −1 1 1 ¿ . B be 3 × 3 matrices. scalar multiplication. Then we deﬁne 3 A + B = [aij + bij ]. ◭ Addition.Chapter 1 and also àà αu1 L(αu) = L αu2 αu3 = à α(u1 ) − α(u2 ) − α(u3 ) α(u1 ) + α(u2 ) + α(u3 ) αu3 à = α u1 − u2 − u3 u1 + u2 + u3 u3 u1 àà = = à αL αL(u). and L 0 1 = −1 1 1 .L 0 1 0 = 0 0 −1 1 0 0 . αA = [αaij ]. 86 Deﬁnition Let A. AB = k=1 aik bkj . à u2 u3 proving that L is a linear transformation. and matrix multiplication are deﬁned for 3 × 3 matrices in a manner analogous to those operations for 2 × 2 matrices.

b.Matrices in three dimensions ¾ 1 2 6 0 ¾ 3 0 ¿ ¾ a b b c 0 ¿ 87 Example If A = 4 5 0 . where a > 0. Easy to ﬁnd counterexamples should convince the reader that the product of two rotations in space does not necessarily commute. A rotation matrix about the x-axis by an angle θ in the counterclockwise sense is ¾ 1 0 0 cos θ sin θ 0 cos θ ¿ Rx (θ) = 0 − sin θ . 6a 6b 6c 88 Deﬁnition A scaling matrix is one of the form ¾ a 0 0 0 b 0 0 c ¿ Sa. and B = a ¿ 0 . c > 0. ¿ a + 4b + 6c 2a + 5b 3a a + 4b a 2a 3a 6a 3b c BA = 2a + 5b 3a .c = 0 . 1 ¿ A rotation matrix about the y-axis by an angle θ in the counterclockwise sense is cos θ 0 sin θ 0 − sin θ 0 0 cos θ Ry (θ) = 1 . ¿ 3A = 12 ¾ 15 0 . then ¾ ¿ a 0 1+a 2+b 3+c 0 0 6+a 0 3 18 6 0 9 0 A+B = 4+a 5+b ¾ . 89 Deﬁnition A rotation matrix about the z-axis by an angle θ in the counterclockwise sense is ¾ ¿ cos θ − sin θ 0 Rz (θ) = sin θ 0 ¾ cos θ 0 0 . Free to photocopy and distribute 60 . AB = 9a 9b 4c . b > 0. It is an easy exercise to prove that the product of two scaling matrices commutes.

. m(a)m(b) = m(a + b).9. a general formula for An .. m(a)m(−a) = I3 .Chapter 1 90 Deﬁnition A reﬂexion matrix about the x-axis is ¾ Rx = −1 0 0 0 1 0 0 1 ¿ 0 . Conjecture and..3 Consider the n × n matrix ¾ ¿ 1 1 1 1 . 0 1 1 . using induction. the 3 × 3 identity matrix. . prove by induction.. A reﬂexion matrix about the y-axis is ¾ 1 0 0 0 0 1 ¿ Ry = 0 −1 A reﬂexion matrix about the z-axis is ¾ 0 ... . . 1 . . .2 Let A ∈ M3×3 (R) be given by Describe A2 and A3 in terms of n. . b are real numbers. 1 1 0 A= 0 . Find 1 1 1 A2 . 0 . 0 1 1 . 0 1 0 ..9. prove that 1. 1 Demonstrate. ... Homework ¾ Problem 1. Problem 1. 1 0 0 1 0 0 0 −1 ¿ Rz = 0 . . Free to photocopy and distribute 61 .. Problem 1. A3 and A4 ..1 Let A = 1 1 0 1 0 ¿ 0 . 0 1 1 ... n ≥ 1. . 1 1 . and put ¾ ¿ 1 0 x 2 m(x) = −x 1 − x .. 2 0 0 1 If a..9. Problem 1. ¾ A= 1 1 1 1 1 1 ¿ 1 1 . 2.9.. that An = 3n−1 A for n ∈ N.4 Let x be a real number.

− − − linearly dependent. b . →) = det a2 b2 c2 = → • ( b × →). If D(→.15) D(→. then D( a b c → → → − − − 3. →) = → • ( b × →). if →. we deﬁne the determinant of A.15) satisﬁes the following properties: 1. and the 3 × 3 matrix A = a c ¾ a1 a2 a3 b1 b2 b3 c1 c3 ¿ a3 b3 c3 c2 . c ) + D(→′ . Proof: − → − − → − − − 1. the volume of the parallelepiped → − − − spanned by these vectors is → • ( b × →). 91 Theorem The determinant of a 3 × 3 matrix A as deﬁned by (1. For example. Since thanks to Theorem 82. − → − − → − − − − → − − − − → − D(λ→. b . b . are a c →. →) = 0. a c to be ¾ ¿ a1 b1 c1 → − − − − − → − a c (1. and accords with the right-hand rule. →) = a a c = = and if λ ∈ R. →. b .Determinants in three dimensions 1. →). a c a3 b3 c3 We now establish that the properties of the determinant of a 3 × 3 as deﬁned above are analogous to those of the determinant of 2 × 2 matrix deﬁned in the preceding chapter. c ). − − D( a a . →) = (λ→)•( b ×→) = λ((→)•( b ×→)) = λD(→. b . det A. → = c2 . b = b2 .10 Determinants in three dimensions We now deﬁne the notion of determinant of a 3 × 3 matrix. j . D is linear in each of its arguments. linearity of the ﬁrst component a c a c follows by the distributive law for the dot product: − − − → − D(→ + →′ . Consider now the ¾ ¿ ¾ ¿ ¾ ¿ a1 b1 c1 → − − − vectors → = a2 . for Free to photocopy and distribute 62 → − − − − (→ + →′ ) • ( b × →) a a c − − → • (→ × →) + →′ • (→ × →) − − − − a b c a b c → → − − → → − − →. b . D( i . k ) = 1. that is. in R3 . b . b . − − → − 2. a c a c a c a c The linearity on the second and third component can be established by using the distributive law of the cross product. →. If the parallelepiped is ﬂat then the volume is 0.

b .D(→.16) = = = − − → − → • (b c − b c )→ + (b c − b c )→ + (b c − b c (1. →). → → − − → − →→ − − 3. →) + D(→. (1. If →. b . hence. λ b . u Observe that ¾ ¿ a1 a3 b1 b2 b3 c1 c2 c3 det a2 = − → • (→ × →) − − a b c (1. j . b . − → − − − − − → − − − → − − → − D(→. →) = a c 0. Since j × k = i . − − − − = D( a b c a and if λ ∈ R. Solution: ◮ Using (1. we have det A = 1 det 5 8 6 9 − 4 det 2 3 8 9 + 7 det 2 5 3 6 = 1(45 − 48) − 4(18 − 24) + 7(12 − 15) = 0. c ). are linearly dependent. − − → → − − − → → − − − D(→.19).Chapter 1 the second component we have. → → → − − − → − → → − − →→ − − D( i . ◭ Free to photocopy and distribute 63 = −3 + 24 − 21 . b . then they lie on the same plane a c − − → − and the parallelepiped spanned by them is ﬂat. →) = → • (( b + b ′ ) × →) a c a c −′ → → → → − →•(b ×− + b ×−) − = a c c −′ − → • (→ × →) + → • (→ × →) − − − − = a b c a b c → → −′ − →.17) − a 2 3 3 2 i 3 1 1 3 j 1 2 2 1) k a1 (b2 c3 − b3 c2 ) + a2 (b3 c1 − b1 c3 ) + a3 (b1 c2 − b2 c1 ) (1. →) = →•((λ b )×→) = λ(→•( b ×→)) = λD(→. b + b ′ . k ) = i • ( j × k ) = i • i = 1. →.19) which reduces the computation of 3 × 3 determinants to 2 × 2 determinants. a c a c a c a c − − → − 2.18) b2 b3 c2 c3 − a2 det b1 b3 c1 c3 + a3 det b1 b2 c1 c2 a1 det . →. and i • i = 1. ¾ 1 7 2 5 8 3 9 ¿ 92 Example Find det 4 6 .

b := [−1.2 Prove that ¾ ¿ a b c det b c c a a b = 3abc − a3 − b3 − c3 . 93 Example Cube ABCDD′ C′ B′ A in ﬁgure 1. dotprod(a. we may use the Maple™ packages linalg.10. Prove that AD′ −→ − MN and ﬁnd the area of the quadrilateral MND′ A. D′ A′ B′ D M A B Figure 1. Problem 1.75: Example 93.b). 1] b:=vector([-1. An example follows with linalg. 0. angle(a. or Student[VectorCalculus] to perform many of the vector operations.74: Example 93. 1. 3. > > > > > > with(linalg): a:=vector([-2.74 has side of length a. M is the −→ − midpoint of edge [BB′ ] and N is the midpoint of edge [B′ C′ ].1]).b).Some Solid Geometry Again.3. LinearAlgebra.1 Prove that ¾ ¿ 1 1 1 det a a2 b b2 c c2 = (b − c)(c − a)(a − b).0.b). −1. arccos [−3. a := [−2. C′ N N M C D′ Q P A Figure 1.10. 0] crossprod(a.11 Some Solid Geometry In this section we examine some examples and prove some theorems of threedimensional geometry. −6] 2 √ 50 25 Homework Problem 1. Free to photocopy and distribute 64 .0]).

In consequence. AD′ 2 2 −→ − MN. Similarly the plane containing ← → points A. The plane containing points A. M and N are the midpoints of the sides [B ′ B] and [B ′ C ′ ] of −→ − −→ − △B ′ C ′ B. Since P 2 and P 3 Free to photocopy and distribute 65 . Because − → −→ − − they are diagonals that belong to parallel faces of the cube. we must prove AE EB = CF FD . √ à √ a 2 a 2+ 9a2 2 = . N are all on the same √ √ a 2 ′ plane. B. M. Hence MND A is a trapezoid with bases of length a 2 and 2 (ﬁgure 1. AD′ BC′ . Proof: that ← → ← → See ﬁgure 1.76. Given the lines AB and CD. ← → and D intersects plane P 2 in the line EG. Now. From the ﬁgure Solution: √ ¬¬−→¬¬ ¬¬ 1 ¬¬− →¬¬ ¬¬−→¬¬ a 2 →¬¬ ¬¬ − ¬¬ ¬¬− ¬¬ ¬¬ − ′ ¬¬ ¬¬ − ¬¬ ′ . and D intersects plane P 2 in the line GF. C. The aforementioned √ ¬¬ −→ − 1 ¬¬−→¬¬ a 2 − ¬¬ ¬¬−→¬¬ ¬¬ − ′ ¬¬ example also gives ¬¬MN¬¬ = ¬¬AD ¬¬ = . and hence MN BC′ by example 14. The height of this trapezoid is thus ¬¬− ¬¬ ¬¬ →¬¬ ¬¬NQ¬¬ = 5a2 4 − a2 8 = 3a √ . by the Pythagorean Theorem. 94 Theorem (Thales’ Theorem) Of two lines are cut by three parallel planes. ¬¬D Q¬¬ = ¬¬AP¬¬ = ¬¬AD ¬¬ − ¬¬MN¬¬ = 2 4 Also. 2 2 The area of the trapezoid is ﬁnally. D′ . ← → Draw line AD cutting plane P 2 in G. 2 8 3a √ · 2 2 ◭ Let us prove a three-dimensional version of Thales’ Theorem. ¬¬− →¬¬ ¬¬ − ¬¬ ¬¬D′ N¬¬ = Ö ¬¬− →¬¬2 ¬¬−→¬¬2 ¬¬ − ′ ¬¬ ¬¬ − ¬¬ ¬¬D′ C ¬¬ + ¬¬C′ N¬¬ = a2 + a2 4 = √ a 5 2 .Chapter 1 ¬¬− →¬¬ √ ¬¬ − ¬¬ ◮ By the Pythagorean Theorem. their corresponding segments are proportional. ¬¬AD′ ¬¬ = a 2. This means that the four points A.75).

76: Thales’ Theorem in 3D. ← → GF and P1 A C A′ D′ C′ B′ D M A N C B P2E G F P 3B D D′ C′ D ′′ B′ Figure 1.78: Example 95. EG (theorem 30).77: Example 95. u = CF FD . = AG GD . AC EB CF FD It follows that AE EB as needed to be shewn. ′′ C ′′ A D N M B ′′ A C B A′ Figure 1. Figure 1.Some Solid Geometry ← → are parallel planes. Free to photocopy and distribute 66 . since P 1 and P 2 are parallel. ← → BD. and so by Thales’ Theorem on the plane AE = AG . GD ← → Similarly.

Put ¬¬−→¬¬ ¬¬− →¬¬ ¬¬−→¬¬ ¬¬−→¬¬ ¬¬ − ¬¬ ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬ − ¬¬ ¬¬AM¬¬ ¬¬MB ¬¬ ¬¬AB ¬¬ − ¬¬AM¬¬ ¬¬−→¬¬ = x =⇒ ¬¬−→¬¬ = ¬¬−→¬¬ = 1 − x. P also contains N. Since MN is parallel to face ABCD . ¬¬− →¬¬ ¬¬−→¬¬ ¬¬ − ′′ ¬¬ ¬¬ − ¬¬ ¬¬BB ¬¬ ¬¬AM¬¬ ¬¬−→¬¬ = ¬¬−→¬¬ . ﬁnd the ratios ¬¬−→¬¬ and ¬¬−→¬¬ . 3 ¬¬ ¬¬− − ¬¬2 ¬¬− →¬¬2 5 2 − ¬¬ ¬¬−→¬¬2 ¬¬ − →¬¬ ¬¬ − ¬¬ ′′ ′′ a = (1−x)2 a2 +x2 a2 =⇒ x ∈ ¬¬MN¬¬ = ¬¬MB ¬¬ +¬¬B N¬¬ =⇒ 9 1 2 . − ′ ¬¬ ¬¬ ¬¬ − ′ ¬¬ 3 ¬¬AB ¬¬ ¬¬BC ¬¬ Solution: ◮ There is a unique plane parallel P to face ABCD and −→ − containing M. ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ 3 ¬¬AB ¬¬ ¬¬BC ¬¬ ¬¬ ¬¬ − ¬¬ →¬¬ ¬¬−→¬¬ ¬¬− ¬¬ ¬¬AM¬¬ ¬¬BN¬¬ 2 ¬¬−→¬¬ = ¬¬−→¬¬ = .78. ¬¬− → ¬¬ ¬¬− − ¬¬ ¬¬ − ′ ¬¬ ¬¬ − →¬¬ ¬¬MB ¬¬ ¬¬B′′ B′ ¬¬ ¬¬−→¬¬ = ¬¬−→¬¬ . by the Pythagorean Theorem. ¬¬− →¬¬ ¬¬ − ¬¬ ′′ ¬¬B N¬¬ = xa. The intersection of P with the cube produces a lamina A′′ B′′ C′′ D′′ . ¬¬−→¬¬ ¬¬ − ¬¬ First notice that ¬¬AB′ ¬¬ = ¬¬−→¬¬ √ ¬¬ − ′ ¬¬ ¬¬BC ¬¬ = a 2. 3 3 . ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬AB ¬¬ ¬¬BB ¬¬ ¬¬− →¬¬ ¬¬− − ¬¬ ¬¬ − ′ ¬¬ ¬¬ − →¬¬ ¬¬MB ¬¬ ¬¬MB′′ ¬¬ ¬¬−→¬¬ = ¬¬− ¬¬ ¬¬ − ′ ¬¬ ¬¬ →¬¬ ¬¬AB ¬¬ ¬¬AB¬¬ =⇒ ¬¬− − ¬¬ ¬¬ − →¬¬ ¬¬MB′′ ¬¬ = (1 − x)a. the points M and N are −→ − located on diagonals [AB′ ] and [BC′ ] such that MN¬ is parallel to¬ the¬face ABCD ¬ −→¬¬ ¬− ¬ ¬¬ − ¬¬ ¬¬ →¬¬ √ ¬¬ ¬¬−→¬¬ ¬¬ ¬¬AM¬¬ ¬¬BN¬¬ 5 ¬¬− ¬¬ → ¬¬ − ¬¬ of the cube. ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬AB ¬¬ ¬¬AB ¬¬ ¬¬AB ¬¬ Now. as in ﬁgure 1. If ¬¬MN¬¬ = ¬¬AB¬¬.Chapter 1 95 Example In cube ABCDD′ C′ B′ A′ of edge of length a. giving the solutions ¬¬ ¬¬ − ¬¬ →¬¬ ¬¬−→¬¬ ¬¬− ¬¬ ¬¬AM¬¬ ¬¬BN¬¬ 1 ¬¬−→¬¬ = ¬¬−→¬¬ = . ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ 3 ¬¬AB ¬¬ ¬¬BC ¬¬ ◭ Homework Free to photocopy and distribute 67 . as △B ′ AB ∼ △B ′ M B ′′ and △BC ′ B ′ ∼ △BN B ′′ . ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬BB ¬¬ ¬¬AB ¬¬ ¬¬− →¬¬ ¬¬− →¬¬ ¬¬ − ¬¬ ¬¬ − ′′ ¬¬ ¬¬B′′ N¬¬ ¬¬BB ¬¬ ¬¬− →¬¬ = ¬¬−→¬¬ ¬¬ − ′ ¬¬ ¬¬ − ′ ¬¬ ¬¬B′ C ¬¬ ¬¬BB ¬¬ =⇒ √ ¬¬ 5 − ¬¬ ¬¬−→¬¬ Since ¬¬MN¬¬ = a. There are two possible positions for the segment.

Prove that MN ⊥ AB and MN ⊥ − → CD.Cavalieri. b > 0.11.12 Cavalieri. a > 0. Proof: We only provide the proof for the second statement. Prove − → − → that AC ⊥ BD. x2 ]. By Cavalieri’s ﬁrst principle. y= bÔ a a 2 − x2 . where c > 0 is the constant of proportionality. and hence. u 97 Example Use Cavalieri’s Principle in order to deduce that the area enclosed x2 y2 by the ellipse with equation 2 + = 1. respectively.11. x2 x2 From elementary calculus. Then. as in ﬁgure 1.79.2 ¬ In¬ a tetrahedron ¬¬ ABCD. Find the angle between the lines [M N ] and [BC]. is πab. Cut any two such solids by horizontal planes that produce cross sections of area A(x) and cA(x). a b Solution: ◮ Consider the circle with equation x2 + y 2 = a2 . we know that x1 A(x)dx and x1 x2 cA(x)dx x2 give the volume of the portion of each solid cut by all horizontal planes as x runs over some interval [x1 . ¬¬ ¬¬ ¬¬ ¬ ¬ ¬¬ ¬¬ → → → → ¬¬− ¬¬ ¬¬− ¬¬ ¬¬− ¬¬ ¬¬− ¬¬ ¬¬AB¬¬ = ¬¬BC¬¬. Area of the ellipse = b Area of the circle a b = πa2 a = πab. B. The corresponding ordinate for the ellipse and the circle are proportional. at an arbitrary height x above a ﬁxed base. points M and N are the midpoints of the edges [AB] and [CD]. 2. C.1 In a regular tetrahedron ¬¬ ¬¬ → ¬¬− ¬¬ with vertices A. As x1 A(x)dx = c x1 A(x)dx the corresponding volumes must also be proportional. All solids with cross sections of proportional areas at the same height have their volume in the same proportion.3 In cube ABCDD′ C′ B′ A′ of edge of length a.11. Problem 1. ¬¬AD¬¬ = ¬¬DC¬¬. and the Pappus-Guldin Rules 96 Theorem (Cavalieri’s Principle) All planar regions with cross sections of proportional length at the same height have area in the same proportion. Problem 1. Find the length of the segment [M N ]. D and with ¬¬AB¬¬ = a. and the Pappus-Guldin Rules Problem 1. 68 Free to photocopy and distribute . −→ − − → −→ − 3. y= Ô a 2 − x2 . 1. ﬁnd the distance between the lines that contain the diagonals [A′ B] and [AC]. the corresponding chords for the ellipse and the circle will be proportional. for y > 0. 1. as the proof for the ﬁrst is similar.

which agrees with a horizontal slab for the sphere at the same height. By Cavalieri’s Principle. as in ﬁgure 1. x2 + r 2 = a2 . consider a punctured cylinder of base radius a and height a. and so this circular slab has area πr 2 = π(a2 − x2 ). as in ﬁgure 1. Volume of the hemisphere = Volume of the punctured cylinder πa3 = πa3 − 3 2πa3 = . We need to recall that the volume of a right circular cone πa2 h with base radius a and height h is . 3 Free to photocopy and distribute 69 . with a cone of height a and base radius a cut from it. 3 Solution: ◮ The following method is due to Archimedes.80.Chapter 1 ◭ x a x a r a x Figure 1. By the Pythagorean Theorem. who was so proud of it that he wanted a sphere inscribed in a cylinder on his tombstone.80: Punctured cylinder. A horizontal slab at height x is an annular region of area πa2 − πx2 .81: Hemisphere. Figure 1. producing a circle of radius r.81. 3 Consider a hemisphere of radius a.79: Ellipse and circle. a Figure 1. 98 Example Use Cavalieri’s Principle in order to deduce that the volume of a 4 sphere with radius a is πa3 . Cut a horizontal slice at height x. Now.

99 Theorem (Pappus-Guldin Rule) The area of the lateral surface of a solid of revolution is equal to the product of the length of the generating curve on the side of the axis of revolution and the length of the path described by the centre of gravity of the generating curve under a full revolution. the volume of the solid torus is (πr 2 )(2πR) = 2π 2 r 2 R. by the PappusGuldin Rule. See ﬁgure 1. Homework Problem 1. The volume of a solid of revolution is equal to the product of the area of the generating plane on one side of the revolution axis and the length of the path described by the centre of gravity of the area under a full revolution about the axis.13 Dihedral Angles and Platonic Solids 101 Deﬁnition When two half planes intersect in space they intersect on a line. 1.12. 100 Example Since the centre of gravity of a circle is at its centre.1 Use the Pappus-Guldin Rule to ﬁnd the lateral area and the volume of a right circular cone with base radius r and height h.82) is (2πr)(2πR) = 4π 2 rR. The intersecting line is called the edge of the dihedral angle and each of the two half planes of the dihedral angle is called a face.Dihedral Angles and Platonic Solids 2πa3 3 4πa3 3 It follows that the volume of the sphere is 2 = .◭ Essentially the same method of proof as Cavalieri’s Principle gives the next result.83. r R Figure 1.82: A torus. Also. and radius of gyration R (as in ﬁgure 1. Free to photocopy and distribute 70 . The portion of space bounded by the half planes and the line is called the dihedral angle. the surface area of the torus with the generating circle having radius r.

See ﬁgure 1. In the trihedral angle of ﬁgure 1. each on each of the faces. Free to photocopy and distribute 71 . Hence the measure of a dihedral angle is the measure of any one of its rectilinear angles. 103 Deﬁnition The opening of three or more planes that meet at a common point is called a polyhedral angle or solid angle. V is the vertex. 104 Theorem The sum of any two face angles of a trihedral angle is greater than the third face angle. Each of the intersecting lines of two consecutive planes is called an edge of the polyhedral angle. △V BC. any two adjacent faces form a dihedral angle. The angles formed by adjacent edges are called face angles. 102 Deﬁnition The rectilinear angle of a dihedral angle is the angle whose sides are perpendicular to the edge of the dihedral angle at the same point.84: Rectilinear of a Dihedral Angle.83: Dihedral Angles.84. In the particular case of three planes. notice that in any polyhedral angle. The common point is called the vertex of the polyhedral angle. A polyhedral angle is said to be convex if the section made by a plane cutting all its edges forms a convex polygon. All the rectilinear angles of a dihedral angle measure the same.85. Figure 1. The portion of the planes lying between consecutive edges are called the faces of the polyhedral angle. we use the term trihedral angle.Chapter 1 Figure 1. △V CA are faces. In analogy to dihedral angles we now deﬁne polyhedral angles. Also. △V AB.

An illustration can be seen in ﬁgure 1. so assume that. Hence AD = AB. D W V C Z Dihedral Angles and Platonic Solids A3 P A5 A2 A1 Figure 1. This implies that ∠BV C > ∠DV C.87: A.86. by the triangle inequality in △ABC.86: Polyhedral Angle. Observe that the polygon A1 A2 · · · An is convex and that the sum of its interior angles is π(n − 2). Observe that △AV D ∼ AV B. Through any point D of the segment [V W ]. If ∠ZV X is smaller or equal to in size than either ∠XV Y or Y V Z. we may draw. We must demonstrate that ∠XV Y + ∠Y V Z > ∠ZV X. Z. . where for convenience. as wanted. Consider now the plane containing the line segment [AC] and the point B. An . ∠ZV X > XV Y . we have depicted only ﬁve edges. X. A2 . Let the faces be cut by a plane. = Now.85. intersecting the edges at the points A1 . . Since we are assuming that ∠ZV X > XV Y . in ∠XV Y the line segment [V W ] such that ∠XV W = ∠XV Y . Take the point B ∈ [V Y ] so that V D = V B. Proof: Let the polyhedral angle have n faces and vertex V . draw △ADC on the plane P containing the points V . which proves that ∠XV Y + ∠Y V Z > ∠ZV X. u 105 Theorem The sum of the face angles of any convex polyhedral angle is less than 2π radians. Hence ∠AV B + ∠BV C = > = ∠AV D + ∠BV C ∠AV D + ∠DV C ∠AV C. Ak Ak+1 Ak−1 Figure 1. Ak+1 are three consecutive vertices. say.V V A4 A X B Y Figure 1. .85: Trihedral Angle. then we are done. Ak . Proof: Consider ﬁgure 1. Free to photocopy and distribute 72 . say. AB + BC > CA.

and ∠V Ak Ak+1 . . Now. . This notation means that Ak−1 Ak Ak+1 represents any of the n triplets A1 A2 A3 . But clearly V Ak Ak+1 = 1≤k≤n 1≤k≤n ∠V Ak+1 Ak . 1≤k≤n being the sum of the interior angles of the polygon A1 A2 · · · An . An+2 = A2 . Ak . u Free to photocopy and distribute 73 . V Ak Ak−1 = 1≤k≤n 1≤k≤n ∠V Ak Ak+1 . Consider the trihedral angle with vertex Ak and whose face angles at Ak are ∠Ak−1 Ak Ak+1 . By Theorem 104. Hence ∠Ak V Ak+1 1≤k≤n = πn − = πn − = 2π. An A1 A2 . let Ak−1 . ∠V Ak Ak−1 + ∠V Ak Ak+1 > ∠Ak−1 Ak Ak+1 . . Also. . as in ﬁgure 1. 1≤k≤n since this is summing the sum of the angles of the n triangles of the faces. that is. An+1 = A1 . An−1 An A1 . Thus V Ak Ak−1 + ∠V Ak Ak+1 > 1≤k≤n 1≤k≤n ∠Ak−1 Ak Ak+1 = π(n − 2).87.Chapter 1 We would like to prove that ∠A1 V A2 + ∠A2 V A3 + ∠A3 V A4 + · · · + ∠An−1 V An + ∠An V A1 < 2π. A3 A4 A5 . V Ak Ak+1 + ∠V Ak+1 Ak + ∠Ak V Ak+1 = πn. we let A0 = An . since one sum adds the angles in one direction and the other in the opposite direction. Observe that as k ranges from 1 through n. An−2 An−1 An . A2 A3 A4 . Ak+1 be three consecutive vertices of of the polygon A1 A2 · · · An . For the same reason. (V Ak Ak+1 + ∠V Ak+1 Ak ) 1≤k≤n (V Ak Ak+1 + ∠V Ak Ak−1 ) 1≤k≤n < πn − π(n − 2) as we needed to shew. the sum ∠Ak−1 Ak Ak+1 = π(n − 2). ∠V Ak Ak−1 . etc.

Dodecahedron. From O(0. z). Suppose a regular polygon with n ≥ 3 sides is a face of a platonic solid with m ≥ 3 faces meeting at a corner. where V is the number of vertices. Thus there are at most ﬁve Platonic solids. 0. F is the number of faces. We now project the line segment [OB] onto the xy-plane in order to ﬁnd the polar coordinates of x and y. Appealing to Euler’s Formula for polyhedrons. y. 0) we draw a straight line to B(x. m). which states that V + F = E + 2.Spherical Trigonometry 106 Deﬁnition A Platonic solid is a polyhedron having congruent regular polygon as faces and having the same number of edges meeting at each corner. π] radians. we obtain the values in the following table. Observe that z = ρ cos φ. z) in Cartesian coordinates.92 depict the Platonic solids. y. Since each interior angle of this polygon π(n − 2) measures . φ ∈ [0. Hexahedron or Cube. Figures 1. Since n ≥ 3 and m ≥ 3. Since OP = ρ sin φ we ﬁnd x = ρ cos θ sin φ. Free to photocopy and distribute 74 . Icosahedron. let us say it is an angle of φ. Tetrahedron or regular Pyramid. Let θ be angle that this projection makes with the positive x-axis. 1. and let its distance be ρ.14 Spherical Trigonometry Consider a point B(x. Octahedron. m 3 4 3 5 3 n 3 3 4 3 5 S 4 8 6 20 12 E 6 12 12 30 30 F 4 6 8 12 20 Name of regular Polyhedron. the above inequality only holds for ﬁve pairs (n.93. n m π(n − 2) n < 2π =⇒ m(n − 2) < 2n =⇒ (m − 2)(n − 2) < 4. we must have in view of Theorem 105. as in ﬁgure 1. We measure its inclination from the positive z-axis. y = ρ sin θ sin φ. and E is the number of edges of a polyhedron. That there are exactly ﬁve can be seen by explicit construction.88 through 1.

z = ρ cos φ. This plane will be unique if and only if the points are not diametrically opposite. 109 Deﬁnition Given the centre of the sphere. We call the smaller arc the geodesic joining the two points. 0 ≤ θ ≤ 2π and 0 ≤ φ ≤ π. the great circle formed is split into a larger and a smaller arc by the two points. A pole of a circle is equidistant from every point of the circumference of the circle. z) in Cartesian coordinates. y. The poles of a great circle are equally distant from the plane of the circle. Figure 1.92: Icosahedron. the intersection will be a circle.90: Octahedron.91: Dodecahedron Figure 1. measured from the positive z-axis. and θ is the azimuthal angle.89: Cube or hexahedron. If this circle contains the centre of the sphere.88: Tetrahedron. Here φ is the polar angle. Otherwise we talk of a small circle. we call it a great circle. By the pole of a small circle. a plane can be drawn.Chapter 1 Figure 1. 107 Deﬁnition Given a point (x. In the case where the two points are not diametrically opposite. 108 Deﬁnition If a plane intersects with a sphere. we mean the closest pole to the plane containing the circle. The radius of a great circle is the radius of the sphere. y = ρ sin θ sin φ. Figure 1. and any two points of the surface of the sphere. but this is not the case in a small circle. The endpoints of such a diameter are called the poles of the circle. Figure 1. its spherical coordinates are given by x = ρ cos θ sin φ. Spherical coordinates are extremely useful when considering regions which are symmetric about a point. measured from the positive x-axis. If the two points are diametrically opposite then every plane containing the line Free to photocopy and distribute 75 . The axis of any circle on a sphere is the diameter of the sphere which is normal to the plane containing the circle. By convention.

Observe that if O is the centre of the sphere then − − → → ∠a = ∠ OB. OC × OB . − − → → ∠b = ∠ OC. − − → → ∠c = ∠ OA. OA . − → − − → → − → ∠B = ∠ OB × OC. ρ cos φ Free to photocopy and distribute ρ si n φ P − → − − → → − → ∠C = ∠ OC × OA. ∠C. B φ ρ θ Figure 1. 76 . In this case we take any such arc as a geodesic. and the arcs formed are then of equal length. OB . and A spherical triangle has then six angles: three vertex angles ∠A. the angles formed by the arcs at the points where they meet are called the angles of the spherical triangle. The three arcs of great circles which form a spherical triangle are called the sides of the spherical triangle. B. OA × OC .93: Spherical Coordinates. ∠a. and three arc angles. OC . 110 Deﬁnition A spherical triangle is a triangle on the surface of a sphere all whose vertices are connected by geodesics. − → − − → → − → ∠A = ∠ OA × OB. ∠B. C are the vertices of a spherical triangle. ∠c. OB × OA . it is customary to label the opposite arcs with the same letter name. If A. ∠b.Spherical Trigonometry forms with the sphere a great circle. but in lowercase.

OC. Then cos π I n sin = π . y1 = ρ sin θ1 sin φ1 . and join CE and DE.94. x2 = ρ sin θ2 cos φ2 . say. bisect AB at E. u Figure 1. that is. e respectively. Let AB be the edge common to the two adjacent faces. C and D the centres of the faces. By a rotation we may assume that the z-axis passes through C. x2 + y2 + z2 = ρ2 . c. z2 = ρ cos θ2 . and the angle CED is the angle of inclination of the two adjacent faces. y1 . so that Free to photocopy and distribute 77 . Then the following quantities give the square of the distance of the line segment [AB]: (x1 − x2 )2 + (y1 − y2 )2 + (z1 − z2 )2 . In spherical coordinates this is. we shall denote it by I. and meeting at O. z1 = ρ cos θ1 . ρ2 + ρ2 − 2ρ2 cos ∠ (AOB) . OE at a. z2 ). In the plane containing CE and DE draw CO and DO at right angles to CE and DE respectively. B(x2 . 112 Theorem Let I be the dihedral angle of two adjacent faces of a regular polyhedron. we gather that 1 2 x1 x2 + y1 y2 + z1 z2 = ρ2 cos ∠ (AOB) .94: Theorem 112. z1 ). and let O be the centre and ρ be the radius of the sphere. about O as centre describe a sphere meeting OA. Therefore we obtain cos θ2 cos θ1 + sin θ2 sin θ1 cos(φ1 − φ2 ) = cos ∠ (AOB) . 2 2 2 2 Since x2 + y1 + z1 = ρ2 . y2 . x1 = ρ sin θ1 cos φ1 . Proof: Consider a spherical triangle ABC with A(x1 . cos a cos b + sin a sin b cos C = cos c. CE and DE will be perpendicular to AB. Then cos a cos b + sin a sin b cos C = cos c. y2 = ρ sin θ2 sin φ2 . 2 sin m Proof: See ﬁgure 1.Chapter 1 111 Theorem Let △ABC be a spherical triangle.

and 3 4 m Free to photocopy and distribute 78 . r = CE tan CEO = CE tan also therefore u r = R cos aOc = R cot eca cot eac = R cot R = r tan π m tan π n = a 2 tan I 2 tan π n . so that r is the radius of the inscribed sphere. Then r= a 2 cot π m tan I 2 . Then CE = AE cot ACE = a 2 cot π m . that is. Since AB is perpendicular to CE and DE. R= a 2 tan I 2 tan π n . π n . 2 sin m 113 Theorem Let r and R be. From the right-angled triangle cae 2n n cos cae = cos cOe sin ace. the radii of the inscribed and circumscribed spheres of a regular polyhedron. therefore u 2 π cos I n sin = π . Proof: Let the edge AB = a. n the number of the plane angles 2π π which form each solid angle. Here a is the length of any edge of the polyhedron. therefore the plane AOB which contains AB is perpendicular to the plane CED. I 2 = a 2 cot π m tan π m I 2 . it is perpendicular to the plane CED. 2m m and the angle cae is half one of the n equal angles formed on the sphere 2π π round a. cot From the above formulæ we now easily ﬁnd that the volume of the pyramid which r ma2 π has one face of the polyhedron for base and O for vertex is cot . cae = = . Let m be the number of sides in each face of the polyhedron.Spherical Trigonometry cae forms a spherical triangle. Then the angle ace = ACE = = . hence the angle cea of the spherical triangle is a right angle. let OC = r and OA = R. that is cos π n = cos π − I 2 sin π m . and R is the radius of the circumscribed sphere. respectively. and I is the dihedral angle of any two faces.

5 Consider a dodecahedron whose edge measures a.3 Consider a cube whose edge measures a. its surface area is a2 3. m ma2 π cot . V3 = 0 √ 2 What is the cosine of the dihedral angle between any pair of faces of the tetrahedron? Problem 1. 4 m therefore the volume of the polyhedron is cot Homework Problem 1. 12 1. V4 = 0 Ǒ . it will become important for us to be able to identify certain families of often-occurring surfaces.4 Consider an octahedron whose edge measures a.1 The four vertices of a regular tetrahedron are Ǒ Ǒ 1 −1/2 √ V1 = 0 . and that the raa dius of the inscribed sphere is .2 Consider a tetrahedron whose edge measures a. Shew that its vol√ a3 ume is 15 + 7 5 . Shew that its vol√ √ a3 2 . Shew that its volume is a3 . the area of one face of the polyhedron is 4 m mF a2 π the surface area of the polyhedron is cot . a circle.14.6 Consider an icosahedron whose edge measures a. 6 Problem 1. We shall explore both their Cartesian and their parametric form. V2 = 3/2 . We remark that in order to parametrise curves (“one-dimensional entities”) we needed one parameter.14. Just like in one-variable Calculus it is important to identify the equation and the shape of a line.14. and that the radius of the inscribed √ √ a sphere is 5 3 + 15 . and therefore Furthermore. its surface area is 4 √ 3a2 25 + 10 5. its surface area is 12 √ 5a2 3.14. Shew that its vol√ 5a3 ume is 3 + 5 . 12 Problem 1. Free to photocopy and distribute 79 . and that the radius of the inscribed sphere is a 4 10 + 22 1 . 2 Problem 1. and ume is 12 that the radius of the inscribed sphere is √ a 6 . its surface area is 6a2 .14. and that in order to parametrise surfaces we shall need to parameters.. Shew that its vol√ √ a3 2 .Chapter 1 mF ra2 12 π . its surface area is 2a2 3.15 Canonical Surfaces In this section we consider various surfaces that we shall periodically encounter in subsequent sections. 5 Ǒ . etc. a parabola. 0 −1/2 √ − 3/2 0 0 Problem 1.14. ume is 3 and √ that the radius of the inscribed sphere a 6 is .

z − z0 = pu3 + qv3 . y. the lines generating S will be parallel to the line of equation A = 0. qthe equation of the plane is x − x0 = pu1 + qv1 . u ∈ R. z = u. Figure 1. B = 0. c are real numbers. or z is missing. if one of the variables x.Canonical Surfaces Let us start with the plane. b.96: The parabolic cylinder z = y2 . One starts with the circle x2 + y 2 = 1 on the xy-plane and moves it up and down the z-axis. Under these conditions. z0 ) is a(x − x0 ) + b(y − y0 ) + c(z − z0 ) = 0. not all ¾ ¿ a zero. z z x y x y Figure 1. y − y0 = pu2 + qv2 . If it is of the form f (A.95: Circular cylinder x2 + y2 = 1. then the Cartesian equation of a plane with normal vector b and passing c through the point (x0 . B) = 0. 2π]. − − If we know that the vectors → and → are on the plane (parallel to the plane) then u v with the parameters p. Recall that if a. y = sin v. where A. y0 . B are secant planes. v ∈ [0. whose directrix will be the axis of the missing coordinate.95 shews the cylinder with Cartesian equation x2 +y 2 = 1. then the curve is a cylinder. In practice. 80 Free to photocopy and distribute . then the surface is a cylinder. A parametrisation for this cylinder is the following: x = cos v. 115 Example Figure 1. To recognise whether a given surface is a cylinder we look at its Cartesian equation. The line ∆ is called the directrix of the cylinder. 114 Deﬁnition A surface S consisting of all lines parallel to a given line ∆ and passing through a given curve Γ is called a cylinder.

z = u.1.z=-10. z = z.10. y = v. > plot3d([t. One starts with the parabola z = y 2 on the yz-plane and moves it up and down the x-axis.1. 0) parallel to the z-axis to (x.numpoints=5001). r θ x y Figure 1. > plot3d([cos(s).sin(s).numpoints=5001.axes=boxed)..96 shews the parabolic cylinder with Cartesian equation z = y 2 .10.10. A parametrisation for this parabolic cylinder is the following: x = u. and then lift (x. v ∈ R. y.. The method of parametrisation utilised above for the cylinder is quite useful when doing parametrisations in space. v ∈ R.x=-10.Chapter 1 The Maple™ commands to graph this surface are: > with(plots): > implicitplot3d(xˆ2+yˆ2=1. r sin θ.98 shews the hyperbolic cylinder with Cartesian equation x2 − y 2 = 1. A parametrisation for this parabolic cylinder is the following: x = ± cosh v.t]. y = sinh v. One starts with the hyperbola x2 − y 2 on the xy-plane and moves it up and down the z-axis..t=-10.sˆ2]. See ﬁgure 1. z (r cos θ..y=-10. we ﬁrst ﬁnd the polar coordinates of x. y in the xy-plane.y=-1.s=-10. We have used the fact that cosh2 v − sinh2 v = 1. u ∈ R.10.10. 117 Example Figure 1.10. y. 116 Example Figure 1... We need a choice of sign for each of the portions.10. The Maple™ commands to graph this surface are: Free to photocopy and distribute 81 . u ∈ R.10).. In general. The Maple™ commands to graph this surface are: > with(plots): > implicitplot3d(z=yˆ2.s. z) y = r sin θ. z = v2 ..t=-10.97..x=-1.97: Cylindrical Coordinates.z=-10.. We refer to it as the method of cylindrical coordinates. z): x = r cos θ.numpoints=5001).s=-10.

0. 0).10. ) = 0. if the Cartesian equation of a surface can be put into the form A B . The intersection of S with a half-plane bounded by ∆ is called a meridian. Σ) = 0. where A is a plane and Σ is a sphere. the apex is located at (0. y.t].. and its distance to the axis Ô of rotation would be |x| = 1 + 4z 2 . f( In practice.t]}.y=-10.Canonical Surfaces > > > > with(plots): implicitplot3d(xˆ2-yˆ2=1.sinh(s).10.t=-10. z) to Free to photocopy and distribute 82 . Considering z z the planes x = 0. The graph is shewn in ﬁgure 1. y = 0. where A. z) be a point on S. s=-2.numpoints=5001). We then say that ∆ is the axis of revolution.[cosh(s).100. then the surface is of revolution. the surface S obtained by drawing rays from Ω and passing through Γ is called a cone. Solution: ◮ Let (x. z) to the axis of rotation is the same as the distance of (x. If this point were on the xz plane.. then the C C surface is a cone. 118 Deﬁnition Given a point Ω ∈ R3 (called the apex) and a curve Γ (called the generating curve)..z=-10. 119 Example The surface in R3 implicitly given by z 2 = x2 + y 2 is a cone. The axis of S is the line passing through the centre of Σ and perpendicular to the plane A.10.2. C = 0. x 2 y 2 If the Cartesian equation of S can be put in the form f (A.axes=boxed). are planes secant at exactly one point. + 120 Deﬁnition A surface S obtained by making a curve Γ turn around a line ∆ is called a surface of revolution.. it would be on the hyperbola. B. and its apex is given by A = 0.sinh(s). B = 0.numpoints=5001.x=-10.. Anywhere else. y. plot3d({[-cosh(s). the distance of (x. z = 0.10. as its equation can be put in the form −1 = 0. y. C. 121 Example Find the equation of the surface of revolution generated by revolving the hyperbola x2 − 4z 2 = 1 about the z-axis.

a2 b c (0. Observe that when z = 0. Free to photocopy and distribute 83 . r are positive real numbers) is revolved around the z-axis. sgn(t) = 1 if t > 0. v ∈ [0. and sgn(0) = 0) is the sign of t. y= 1 + 4u2 sin v. y. x2 + y 2 . Solution: ◮ Let (x. A parametrisation for this hyperboloid is x= ◭ 122 Example The circle (y − a)2 + z 2 = r 2 . If this point were on the yz plane.104. which is to say x2 + y 2 − 4z 2 = 1. We must have Ô x2 +y 2 = (a+sgn(y−a) r 2 − z 2 )2 = a2 +2asgn(y−a) r 2 − z 2 +r 2 −z 2 . Find the equation of this torus. 0. Figure 1. z) be a point on T . y. u ∈ R. that is x2 + y 2 . This surface is called a hyperboloid of one sheet. y 2 − 4z 2 = 1 is a hyperbola on the yz plane. We must have x2 + y 2 = Ô 1 + 4z 2 . Figure 1. When y = 0. z) : Ô Ô Ô 1 + 4u2 cos v. z) to the z-axis is the distance of this point to the point (x. See ﬁgure 1.99: The torus. x2 − 4z 2 = 1 is a hyperbola on the xz plane.98: The hyperbolic cylinder x2 − y2 = 1. When x = 0. where sgn(t) (with sgn(t) = −1 if t < 0.Chapter 1 Figure 1. Anywhere else. the distance from (x. forming a torus T . it would be on the circle. z). on the yz plane (a. y.100: Cone x2 y2 z2 + 2 = 2. z = u. 2π]. and the of the distance to the axis Ô of rotation would be y = a + sgn(y − a) r 2 − z 2 . x2 + y 2 = 1 is a circle on the xy plane.

See ﬁgure 1. z = u. When y = 0. c > 0. The equation of the torus thus.101: Paraboloid z = x2 y2 + 2. or (x2 + y 2 + z 2 − (a2 + r 2 ))2 = 4a2 r 2 − 4a2 z 2 since (sgn(y − a))2 = 1. Free to photocopy and distribute 84 . z = r sin α. z = x2 is a parabola on the xz plane.99. Rearranging again. why?). The equation clearly requires that z ≥ 0.102: Figure 1. y = u sin v.103: Two-sheet hyx2 y2 z2 perboloid 2 = 2 + 2 +1. u ∈ [0. and its graph appears in ﬁgure 1. +∞[. z = y 2 is a parabola on the yz plane. x2 + y 2 = c is a circle. with (θ. a2 b Figure Hyperbolic y2 x2 paraboloid z = 2 − 2 a b 1. A parametrisation for the torus generated by revolving the circle (y − a)2 + z 2 = r 2 around the z-axis is x = a cos θ+r cos θ cos α. α) ∈ [−π.101. The following is a parametrisation of this paraboloid: √ √ x = u cos v. is of fourth degree. Ô ◭ z z z x y x y x y Figure 1. (it could not be 0. π]2 .Canonical Surfaces Rearranging x2 + y 2 + z 2 − a2 − r 2 = 2asgn(y − a) r 2 − z 2 . When x = 0. (x2 + y 2 + z 2 )2 − 2(a2 + r 2 )(x2 + y 2 ) + 2(a2 − r 2 )z 2 + (a2 − r 2 )2 = 0. For ﬁxed z = c. v ∈ [0. c a b 123 Example The surface z = x2 +y 2 is called an elliptic paraboloid. y = a sin θ+r sin θ cos α. 2π].

Chapter 1 124 Example The surface z = x2 − y 2 is called a hyperbolic paraboloid or saddle. When z = c is a constant c > 1. For x2 + y 2 < 1. then the x2 + y 2 = c2 − 1 are circles. v ∈ R. For z 2 − 1 < 0. x2 + y 2 < 0 is impossible. 126 Example The surface z 2 = x2 + y 2 − 1 is called an hyperboloid of one sheet. When x = 0. When z = c is a constant.104. x2 − y 2 = 0 is a pair of lines in the xy plane. z 2 − x2 = 1 is a hyperbola on the xz plane.102. y = u sin v. z 2 − y 2 = 1 is a hyperbola on the yz plane. 2π]. y = v. The following is a parametrisation for this hyperboloid of one sheet x= Ô u2 + 1 cos v. u ∈ R.103. z 2 − x2 = −1 is a hyperbola on the xz plane. When y = 0. v ∈ [0. v ∈ [0. 125 Example The surface z 2 = x2 + y 2 + 1 is called an hyperboloid of two sheets. u ∈ R. u ∈ R. 2 c a b One-sheet hyperboloid Figure 1. a2 b c Free to photocopy and distribute 85 . z = u. See ﬁgure 1. u ∈ R. y = u sin v. The following is a parametrisation of this hyperbolic paraboloid: x = u. When x = 0. Figure 1. z = −u2 − 1. and hence there is no graph when −1 < z < 1. z = x2 is a parabola on the xz plane. If z = 0. and hence there is no graph when x2 + y 2 < 1. y= Ô u2 + 1 sin v. When x = 0. z = u2 − v 2 . z 2 < 0 is impossible. z = −y 2 is a parabola on the yz plane.104: z2 x2 y2 = 2 + 2 − 1. When y = 0. z 2 − y 2 = −1 is a hyperbola on the yz plane. v ∈ [0. then the x2 + y 2 = c2 + 1 are circles See ﬁgure 1. When y = 0.105: Ellipsoid x2 y2 z2 + 2 + 2 = 1. See ﬁgure 1. x = u cos v. 2π]. 2π] and the following parametrises the bottom sheet. z = u2 + 1. The following is a parametrisation for the top sheet of this hyperboloid of two sheets x = u cos v.

z = c cos φ. au). 2π]. Problem 1. the radius of the largest circle which can lie on the ellipsoid x2 y2 z2 + 2 + 2 = 1. (u. +∞) × (0. Problem 1. b > 0. +y 2 +z 2 − (x + z)e−2xz = 0.9 Shew that the surface S in R3 implicitly deﬁned as xy + yz + zx + x + y + z + 1 = 0 is of revolution and ﬁnd its axis. is a cylinder. y. We may parametrise b2 the ellipsoid using spherical coordinates: x = a cos θ sin φ. 2π). The surface z 2 + + c2 z 2 = 1 is called an ellipsoid. and Cylindrical coordinates.15.1 Find the equation of the surface of revolution S generated by revolving the ellipse 4x2 + z 2 = 1 about the zaxis.When y = 0. See ﬁgure 1. b. = 1 is an ellipse on the yz plane.3 Describe the surface parametrised by ϕ(u. π].105. y = b sin θ sin φ. (u. bu sin v. 2 a b c a > b > c > 0. z) ∈ R3 : x2 +y 2 +z 2 = 1. 2π) × (0. Homework Problem 1.4 Describe the face parametrised by ϕ(u. x 2 x 2 plane.8 Shew that the surface S in R3 given implicitly by the equation 1 1 1 + + =1 x−y y−z z−x is a cylinder and ﬁnd the direction of its directrix. Problem 1. φ ∈ [0. For z = 0.15.15. the spherical Problem 1.Canonical Surfaces x2 a2 y2 b2 127 Example Let a. a.15. When x = 0. v) → (v cos u. and ﬁnd its axis of revolution. with proof.15.10 Demonstrate that √ S = {(x. z ≥ 1/ surface in R3 given implicitly by 2}. 1). Free to photocopy and distribute 86 .15. implicitly deﬁned. v) (1.15.15. Problem 1.5 Consider cap deﬁned by the sur= ∈ Problem 1. a > 0.6 Demonstrate that the surface in R3 S : ex 2 z 2 − xy = 2z − 1 is a cone Problem 1.7 Shew that the surface in R3 implicitly deﬁned by x4 + y 4 + z 4 − 4xyz(x + y + z) = 1 is a surface of revolution.15. θ ∈ [0.15.15. Parametrise S using Cartesian.2 Find the equation of the surface of revolution generated by revolving the line 3x + 4y = 1 about the y-axis .11 (Putnam Exam 1970) Determine. v) ∈ (0. Problem 1. Problem 1. Problem 1. Spherical. u2 ). c be strictly positive real numbers. v sin u. c2 + y 2 a2 + z 2 a2 + y 2 b2 1 is an ellipse on the xy c2 = 1 is an ellipse on the xz plane. v) (au cos v.

15. b] → R3 . Γ.106 has the property that if it is cut by planes at z = ±2. 4x2 + y 2 = 1 y z = −2. The length of the curve is ¬¬ →¬¬ ¬¬d− ¬¬.106: Problem 1.y2 z = 2. r Γ . A curve is closed if its initial point and its ﬁnal point coincide. with à x(t) r(t) = y(t) z(t) and such that r([a. b] ⊆ R. b]) = Γ.15.16 Parametric Curves in Space In analogy to curves on the plane. x2 + =1 4 z Chapter 1 Problem 1.12 The hyperboloid of one sheet in ﬁgure 1.107: Helix. its projection on the xy plane produces the ellipse y2 x2 + = 1. The trace of the curve r is the set of all images of r. x z = 0. Find its equation. r(a) is the initial point of the curve and r(b) its terminal point. that is.12. 128 Deﬁnition Let [a. and if it is cut by a plane at 4 z = 0. x2 + y2 =1 4 Figure 1. its projection on the xy plane produces the ellipse 4x2 + y 2 = 1. we now deﬁne curves in space. Free to photocopy and distribute 87 . A parametric curve representation r of a curve Γ is a function r : [a. 1. z x y Figure 1.

and acts as a source light projecting a shadow of R onto the xyplane.t>. y = sin t.axes=normal). 2π].2*Pi. This ellipse can be parametrised as √ 2 x = cos t. Solution: ◮ The projection of the intersection of the plane 3x+y+z = 1 and the cylinder is the ellipse x2 + 2y 2 = 1.z]=Path(<cos(t).0. The Maple™ commands to graph this curve and to ﬁnd its length are: > with(plots): > with(Student[VectorCalculus]): > spacecurve([cos(t).sin(t). 2 Thus we may take the parametrisation ¾ x(t) z(t) ¿ ¾ r(t) = y(t) = 1 − 3 cos t − cos t √ 2 sin t 2 √ ¿ . ﬁrst observe that √ ¬¬ →¬¬ ¬¬ ¬¬ ¬¬d− ¬¬ = ¬¬(sin t)2 + (− cos t)2 + 1¬¬dt = 2dt. c be strictly positive real numbers. b. z) ∈ R3 : |x| ≤ a.Parametric Curves in Space 129 Example The trace of → − → − → − r(t) = i cos t + j sin t + k t is known as a cylindrical helix. A point P moves along the ellipse = 1.. Find the area of this shadow. |y| ≤ b. 130 Example Find a parametric representation for the curve resulting by the intersection of the plane 3x + y + z = 1 and the cylinder x2 + 2y 2 = 1 in R3 .t=0. > PathInt(1.sin(t). z =c+1 a2 b2 once around. y. 2 From the equation of the plane. Consider the the region R = (x.2*Pi)).t].[x. √ 2 z = 1 − 3x − y = 1 − 3 cos t − sin t. 0 ≤ t ≤ 2π.y.. x and thus the length is 2π 0 √ √ 2dt = 2π 2. 2 sin t 2 ◭ 131 Example Let a. on the xy-plane. Free to photocopy and distribute 88 x2 + y2 . To ﬁnd the length of the helix as t traverses the interval [0. z = c .

4.4. Figure 1.109). and an ellipse = 1. z = c . being the centre of a (2c + 2) × (2c + 2) square. 0) (ﬁgure 1.4. z) ∈ R3 : |x| ≤ a. z = c . For ﬁxed P (u. Q moves along the circle with equation x2 +y 2 = c2 on the xy-plane (ﬁgure 1. y. Solution: ◮ First consider the same problem as P moves around the circle x2 + y 2 = 1.16. v. and the central square has side 2c + 2.108: Problem 1. As P moves along the circle.109: Problem 1. the image of R′ (a 2 × 2 square) on the xy plane is a (2c + 2) × (2c + 2) square with centre at the point Q(−cu. Resizing to a region R = (x.110: Problem 1. c + 1) on the circle. of area πc2 + 4(c + 1)2 + 8c(c + 1). a(c + 1) (parallel to the x-axis) and b(c + 1) (parallel to the y-axis). y. ◭ Free to photocopy and distribute 89 x2 + y2 . −cv. Figure 1. This creates a region as in ﬁgure 1.110.z Chapter 1 x y Figure 1. |y| ≤ 1.109). so that the area shadowed is πab(c + 1)2 + 4ab(c + 1)2 + 4abc(c + 1) = c2 ab(π + 12) + 16abc + 4ab.16. z = c + 1 a2 b2 we use instead of c + 1. where each quarter circle has radius c. z) ∈ R3 : |x| ≤ 1.16. z = c + 1 and the region is R′ = (x. |y| ≤ b.

1. 132 Deﬁnition Rn is the n-dimensional space.16. As a point Q moves along C . Problem 1.xy-plane.16. 1 + t2 2 t 1 + t2 Let tk . the collection x1 Rn = x2 . Let (x. P2 : x − ay + az = −1. let → z−x2 −y 2 −1 = 0. if they at all Problem 1. if they at all in. y = 4t3/2 . and r r r → (t ) are coplanar if and only if − r 4 1 1 1 1 + + + = 0. 32. Problem 1. xn ã : xk ∈ R . Find an equation relating x and y.2 Consider the surfaces in R3 implicitly deﬁned by which lies on top of the plane x + y + z = 0. R be the point of intersection of ← and PQ Describe.16. Graph all points R on the faces and their intersection. 1. and consider the planes Problem 1.5 Let P be the point (2. Calculate the distance along C from (1. →(t3 ). Free to photocopy and distribute 90 .1 Let C be the curve in R3 deﬁned by x = t2 . Problem 1. z = 9t.Multidimensional Vectors Homework Problem 1.3 Consider the space curve ¾ ¿ t4 1 + t2 t3 →:t→ − r . z+x2 +y 2 −3 = 0. 1 ≤ k ≤ 4 be non-zero real num− − − bers. and the surface S as c varies. Find the volume bounded by the two planes. 1) and consider the curve C : z = y 2 on the yz-plane. y. 2. +∞[. l describes a surface S in R3 . z) be the point of intersection of this surface and the plane z = c.17 Multidimensional Vectors We brieﬂy describe space in n-dimensions. tersect. Let l be their intersection line.16. The ideas expounded earlier about the plane and space carry almost without change.16.6 Let a be a real number paintersect. as vividly as possible these sur. and x = 1. Prove that → (t1 ). rameter. t1 t2 t3 t4 P1 : ax + y + z = −a. 4. As a varies through R.16. . 0. . →(t2 ). 9) to (16.4 Give a parametrisation for the part of the ellipsoid x2 + y2 z2 + =1 9 4 t ∈ [0. 36). Find a direction vector for l.the xy-plane. Find a parametric equation for the curve on which they intersect. x = 0. 3.

Free to photocopy and distribute 91 . αn − − → 136 Deﬁnition Given vectors →. . . k=1 We now establish one of the most useful inequalities in analysis.Chapter 1 → − − − − → 133 Deﬁnition If → and b are two vectors in Rn their vector sum → + b is deﬁned a a by the coordinatewise addition ¾ a1 + b1 a2 + b2 . . . 0 (a 1 in the k slot and 0’s everywhere else). Observe that ¾ n k=1 α1 ¿ − αk →k = e α2 . . αan 135 Deﬁnition The standard ordered basis for Rn is the collection of vectors − − − {→1 . If α ∈ R and → ∈ Rn a we deﬁne scalar multiplication of a vector and a scalar by the coordinatewise multiplication ¾ ¿ αa1 − α→ = a αa2 . b of Rn . (1. →2 .21) 0 . . . . . →n . → = − ek 1 . an + bn ¿ − →+→= − a b . . . . } e e e with ¾ ¿ (1. . . . their dot product is a →•→ = − − a b n ak bk . .20) − 134 Deﬁnition A real number α ∈ R will be called a scalar.

x y x y x y x y giving the theorem. Then the angle (→. that is. we have ¬¬→ − ¬¬ ¬¬− + t→¬¬ ≥ 0 x y ⇐⇒ ⇐⇒ ⇐⇒ − − − − (→ + t→)•(→ + t→) ≥ 0 x y x y →•→ + 2t→•→ + t2 →•→ ≥ 0 − − − − − − x x x y y y ¬¬→¬¬2 ¬¬− ¬¬ − − ¬¬− ¬¬ + 2t→•→ + t2 ¬¬→¬¬2 ≥ 0. The form of the Cauchy-Bunyakovsky-Schwarz most useful to us will be ¬ ¬ ¬ ¬ ¬ n k=1 ¬ ¬ ¬ x k yk ¬ ≤ ¬ n 1/2 n 2 yk k=1 1/2 x2 k k=1 .22) for real numbers xk . ¬¬− ¬¬2 ¬¬− ¬¬2 ¬¬− ¬¬¬¬− ¬¬ − − − − (2→•→)2 − 4(¬¬→¬¬ )(¬¬→¬¬ ) ≤ 0 ⇐⇒ |→•→| ≤ ¬¬→¬¬¬¬→¬¬. u Again. x y x y This last expression is a quadratic polynomial in t which is always nonnegative. the preceding proof is valid in any vector space that has a norm. (1. x y − ¬¬¬¬− ¬¬ ¬¬ x y Free to photocopy and distribute 92 . → − − 138 Corollary (Triangle Inequality) Let → and b be any two vectors in Rn . As such its discriminant must be non-positive. yk . →) between them is given by the relation x y → •→ − − x y − − cos (→. Proof: → − − ||→ + b ||2 a → − − → − − = (→ + b )•(→ + b ) a a − − − − − − → → → = →•→ + 2→• b + b • b a a a → − → − − − ≤ ||→||2 + 2||→|||| b || + || b ||2 a a → − − = (||→|| + || b ||)2 . a from where the desired result follows. Then a we have ¬¬ ¬¬→¬¬ ¬¬→¬¬ →¬¬ − ¬¬ ¬¬→ ¬¬− ¬¬ − − ¬¬ a + b ¬¬ ≤ ¬¬ a ¬¬ + ¬¬ b ¬¬.Multidimensional Vectors − − 137 Theorem (Cauchy-Bunyakovsky-Schwarz Inequality) Let → and → be any two x y n vectors in R . u The above proof works not just for Rn but for any vector space (cf. x y x y Proof: Since the norm of any vector is non-negative. − − 139 Deﬁnition Let → and → be two non-zero vectors in a vector space over the x y − − real numbers. →) = ¬¬→¬¬¬¬→¬¬ . Then we have ¬¬− ¬¬¬¬− ¬¬ − − |→•→| ≤ ¬¬→¬¬¬¬→¬¬. below) that has an inner product.

We need some preparatory work. . k = 1. Shew that n 4 n n n 2 ak bk ck k=1 ≤ n a4 k k=1 k=1 b4 k k=1 c2 k . 140 Example Assume that ak . Thus n x→0 1/x lim log k=1 qk a x k = = = = x x k=1 qk (ak − 1) lim x→0 x n (ax − 1) lim qk k x→0 x k=1 x→0 lim log Èn Èn k=1 qk a x k ¡ n qk log ak . ◭ We now use the CBS inequality to establish another important inequality. k=1 Free to photocopy and distribute 93 . Solution: ◮ Using CBS on k=1 n k=1 (ak bk )ck once we obtain n 1/2 n 1/2 ak bk ck ≤ n a2 b2 k k k=1 1/2 k=1 c2 k . .Chapter 1 This expression agrees with the geometry in the case of the dot product for R2 and R3 . Then 1/x n lim log k=1 qk a x k = k=1 qk log ak . . . with k=1 n x→0 qk = 1. Proof: Recall that log(1 + x) ∼ x as x → 0. n. n 141 Lemma Let ak > 0. ck . bk . qk > 0. are positive real numbers. Using CBS again on k=1 n a2 b2 k k n we obtain 1/2 n 1/2 ak bk ck k=1 ≤ ≤ a2 b2 k k k=1 n 1/4 c2 k k=1 n b4 k k=1 1/4 n 1/2 a4 k k=1 c2 k k=1 . which gives the required inequality.

.23) Successive applications of (1. Notice that since the factors are unequal we have strict inequality. which by Lemma 141 has limit exp giving a1 + a2 + · + an √ n a1 a2 · · · an ≤ . a2 > 0. .. n as wanted. As a corollary to AMGM we obtain Free to photocopy and distribute 94 . . Their harmonic mean is given by n 1 a1 + 1 a2 + ··· + 1 an . . = n2 n n = nn . then by CBS 1 n n k=1 bk ≥ 1 n n k=1 Ô 2 bk . 1 · 3 · 5 · · · · (2n − 1) < 1 + 3 + 5 + · · · + (2n − 1) n n 1 n n log ak k=1 = √ n a1 a2 · · · an . n Proof: If bk ≥ 0. an > 0.. For.Multidimensional Vectors u 142 Theorem (Arithmetic Mean-Geometric Mean Inequality) Let ak ≥ 0. u 143 Example For any positive integer n > 1 we have 1 · 3 · 5 · · · · (2n − 1) < nn . Then a1 + a2 + · + an √ n a1 a2 · · · an ≤ . 144 Deﬁnition Let a1 > 0.23) yield the monotone decreasing sequence 1 n n k=1 ak ≥ 1 n n k=1 √ 2 ak ≥ 1 n n k=1 √ 4 4 ak ≥ . (1. by AMGM..

b2 > 0. . Solution: ◮ Put bk = n s s − ak 1 bk = .Chapter 1 145 Corollary (Harmonic Mean-Geometric Mean Inequality) Let b1 > 0. For then 1 1 b1 b2 u ··· 1 bn 1/n 1 1 1 + + ··· + b b2 bn ≤ 1 . . b2 > 0. bn > 0. and s = a1 + a2 + · · · + an . we deduce 146 Corollary (Harmonic Mean-Arithmetic Mean Inequality) Let b1 > 0. Then n ≤ (b1 b2 · · · bn )1/n . 1 1 1 + + ··· + b1 b2 bn 1 bk Proof: This follows by putting ak = in Theorem 142 . Èn n n−1 k=1 ≤ s s − ak . . Prove that n k=1 s s − ak ak s − ak ≥ n2 n−1 n n−1 and n k=1 ≥ . . . Free to photocopy and distribute 95 . n from where the ﬁrst inequality is proved. Then n k=1 s − ak s k=1 =n−1 and from Corollary 146. Then b1 + b2 + · · · + bn n ≤ . n Combining Theorem 142 and Corollary 145. . bn > 0. 1 1 1 n + + ··· + b1 b2 bn 147 Example Let ak > 0. . .

Prove that ∀x ∈ R. k A1 + A2 + · · · + An = n. In this exercise you will follow the steps of a proof by George Polya. e be real numbers such that a+b+c+d+e = 8. = an . c. x ≤ ex−1 . Prove the AMGM inequality by assembling the results above.6 Let f (x) = (a + x)5 (a − x)3 .Multidimensional Vectors Since s s − ak −1= ak s − ak ak s − ak . .17. . . . 4.. .7 Prove that the sequence 1 n xn = 1 + . −1 −n = k=1 2 ≥ = ◭ n n−1 n n−1 Homework Problem 1. + xn ) + + .. Deduce that a1 + a2 + · · · + an Gn ≤ n n Problem 1.3 (USAMO 1978) Let a. creasing. . n = 1. . x2 . = (a1 + a2 + · · · + an )n ak = 96. is strictly in2 n ≥n . .4 Find all positive real numbers a1 ≤ a2 ≤ .17. n! < n+1 2 n . k k=1 a3 = 216. Prove that n Problem 1. . ≤ an such that n n n A1 A2 · · · An and that n Gn . Find the maximum value of de f using the AM-GM inequality.17. . 2. .5 Demonstrate that for integer n > 1 we have. Problem 1. k=1 k=1 a2 = 144. 3. . a]. . are strictly positive real numbers then 1 1 1 (x1 + x2 + . (a1 a2 · · · an )1/n ≤ n Equality occurs if and only if a1 = a2 = . xn .17. Maximise the value of e.17. Problem 1. we have n n k=1 = k=1 n s s − ak s s − ak −n . . ´ 1.2 Demonstrate that if x1 . a1 + a2 + · · · + an and Gn = a1 a2 · · · an .17. d.17. Put nak Ak = .1 The Arithmetic Mean Geometric Mean Inequality says that if ak ≥ 0 then a1 + a2 + · · · + an . Problem 1. x ∈ [−a. 2. + x1 x2 xn Problem 1. Free to photocopy and distribute 96 . a2 +b2 +c2 +d2 +e2 = 16. b.

1: Open ball in R2 . 97 b2 − a2 ε . bk are real numbers. where the ak .1 Some Topology An open box is a Cartesian product of open intervals 148 Deﬁnition Let a ∈ Rn and let ε > 0.2 Differentiation 2. b2 [× · · · ×]an−1. An open ball centred at a of radius ε is the set Bε (a) = {x ∈ Rn : ||x − a|| < ε}. Figure 2. (a1 . ]a1 . a2 ) b1 − a1 Figure 2. bn [. b1 [×]a2. bn−1 [×]an .2: Open rectangle in R2 .

an open ball in R2 is an open disk (see ﬁgure 2. An open box in R is an open interval. y) ∈ R2 : x2 ≤ y ≤ x} 7. however. G = {(x. closed. |y| < 1} 2. 1] is neither open nor closed in R.1 Determine whether the following subsets of R2 are open. The reader will recognise that open boxes. however. y) ∈ R2 : |x| < 1.4). R \ [−1. The interval ] − 1. open ellipsoids and their unions and ﬁnite intersections are open sets in Rn . 152 Example The region ] − 1. 1[×]0. C = {(x. 1] × [0. F = {(x. That is. x < y 2 } 3. y 149 Example An open ball in R is an open interval. +∞[ is open in R. or neither. D = {(x. Homework Problem 2.4: Open box in R3 . |y| ≤ 1} 5. y) ∈ R2 : |y| ≤ 9. as its complement. 1]. A = {(x. y) ∈ R2 : xy ≤ 1} 4. a3 ) Some Topology x Figure 2. an open box in R2 is a rectangle without its boundary (see ﬁgure 2. y) ∈ R2 : xy > 1} 6. 1[ is open in R.3: Open ball in R3 . 1] =] − ∞.z z ε (a1 .2) and an open box in R3 is a box without its boundary (see ﬁgure 2. a2 . ∀a ∈ O ∃ε > 0 such that Bε (a) ⊆ O. y) ∈ R2 : x2 + 4y 2 < 4 is open in R2 . +∞[ is open in R2 . 1. 155 Example The closed interval [−1. 156 Example The region [−1. B = {(x. +∞[×[0. 2] is closed in R3 . 150 Deﬁnition A set O ⊆ Rn is said to be open if for every point belonging to it we can surround the point by a sufﬁciently small open ball so that this balls lies completely within the set. E = {(x. 151 Example The open interval ] − 1.1) and an open ball in R3 is an open sphere (see ﬁgure 2.1. 153 Example The ellipsoidal region (x. y x Figure 2. y) ∈ R2 : |x| ≤ 1. in R2 . |y| ≤ 1} Free to photocopy and distribute 98 . 1] is closed in R. 1] is not open. as no interval centred at 1 is totally contained in ] − 1. The interval ] − 1.3). −1[∪]1. 154 Deﬁnition A set F ⊆ Rn is said to be closed in Rn if its complement Rn \ F is open. y) ∈ R2 : |x| < 1.

Shew that there is a pairwise disjoint subcollection Dk . y) → cos(x2 − y 2 ) Problem 2. 1. the introduction of more variables greatly complicates the analysis. z) → xyz 3.1. the level curve at z = c is the curve resulting from the intersection of the surface z = f (x.2.2. y) and the plane z = c. y. (x. y) → sin(x2 + y 2 ) 7.1 Sketch the level curves for the following maps. For example. and integrability of multivariable functions. we have a tri-dimensional surface. deﬁned over the entire plane R2 . our concern will be functions of the form If m = 1. j =1 2. continuity.2 Sketch the level surfaces for the following maps. c > 0. (x. Homework Problem 2. y) → x3 − x 5. (x. y) = x2 +y 2 (an elliptic paraboloid) are the concentric circles x2 + y 2 = c. y. (x. y) → x + y 2. Given c ∈ R. 158 Example The level curves of the surface f (x. differentiability. Needless to say. we would like to examine limits. is the set {(x. y) be a polynomial with real coefﬁcients in the real variables x and y. |y|) 4. (x. We will now brieﬂy examine this case. (x. y. We would like to develop a calculus analogous to the situation in R. For most of this course. (x. y) → x2 + 4y 2 6.1. Let A ⊆ Rn . y) → min(|x|. we say that f is a vector ﬁeld. (x. In particular. if there is such a curve. What are the possibilities for the image (range) of p(x. m = 1. 1. y)? Problem 2. we have an object of more than three-dimensions! In the case n = 2. |y|.Chapter 2 Problem 2.2 Multivariable Functions f : A → Rm . z) → x + y + z 2. |z|) Free to photocopy and distribute 99 . (x. we say that f is a scalar ﬁeld.3 (Putnam 1998) Let F be a ﬁnite collection of open disks in R2 whose union contains a set E ⊆ R2 . (x. If m + n > 3. k ≥ 1 in F such that n E⊆ 3Dj . y) → xy 3. If m ≥ 2. 157 Deﬁnition Let A ⊆ R2 and let f : A → R be a function.2 (Putnam Exam 1969) Let p(x. z) → min(|x|. recall that the graph of a function f : A → Rm . f (x)) : x ∈ A)} ⊆ Rn+m . A ⊆ Rn .

(x. > limit(xˆ2*y/(xˆ2+yˆ2). > with(plots): > plot3d(xˆ2*y/(xˆ2+yˆ2).y)→(0.y)→(0.0) ¬ x2 + y 2 ¬ (x. (x. limits at inﬁnity. ◭ 161 Example Find (x. Thus x + y2 ¬ ¬ ¬ x2 y ¬ ¬ ¬≤ lim 0≤ lim (x. Free to photocopy and distribute 100 . 159 Deﬁnition A function f : Rn → Rm is said to have a limit L ∈ Rm at a ∈ Rn if ∀ǫ > 0∃δ > 0 such that 0 < ||x − a|| < δ =⇒ ||f (x) − L|| < ǫ. and so 0 ≤ 2 ≤ 1. The notions of inﬁnite limits.y)→(0. z) → sin(z − x2 − y 2 ) 7.0) lim x5 y 3 x6 + y 4 .0) (x. y.x=-10. and continuity at a point.0) ◮ We use the sandwich theorem.y=-10. y.10.0) lim |y|. y. (x.axes=boxed. x→a lim f (x) = L. are analogously deﬁned. as one must approach the test point from inﬁnitely many directions. y.y)→(0. (x. z) → x2 + y 2 5.x=0.Limits 4. Observe that 0 ≤ x2 ≤ x2 x2 + y 2 . The Maple™ commands to graph this surface and ﬁnd this limits appear below. x2 + y 2 (x.3 Limits We will start with the notion of limit.y=0). z) → x2 + y 2 + z 2 2.. and hence (x. In such a case we write. Limits in more than one dimension are perhaps trickier to ﬁnd.y)→(0.. 160 Example Find Solution: lim x2 y .color=xˆ2+yˆ2).10.0) lim x2 y x2 + y 2 = 0.y)→(0. z) → x2 + 4y 2 6. Notice that Maple is unable to ﬁnd the limit and so returns unevaluated.

then ¬ ¬ ¬ x5 y 3 ¬ x8 2 ¬ ¬ ¬ x6 + y 4 ¬ ≤ x6 = x . 0). y) → (0. −y 2 as y → 0. Y = ρ sin θ. Solution: then ◮ Either |x| ≤ |y| or |x| ≥ |y|. x2 + y2 Figure 2. ¬ x6 + y 4 ¬ X2 + Y 2 Passing to polar coordinates X = ρ cos θ. we obtain ¬ ¬ ¬ x5 y 3 ¬ X 5/3 Y 3/2 ¬ ¬= = ρ5/3+3/2−2| cos θ|5/3| sin θ|3/2 ≤ ρ7/6 → 0. x ) ≤ y + x −→ 0.5: (x.0) lim 1+x+y x2 − y 2 .7: Example 162.Chapter 2 Figure 2. as (x. Solution: ◮ When y = 0. 0). y) → x5 y3 . Thus ¬ ¬ ¬ x5 y 3 ¬ 4 2 4 2 ¬ ¬ ¬ x6 + y 4 ¬ ≤ max(y . ¬ x6 + y 4 ¬ y4 If |y| ≤ |x|. y) → x2 y . Aliter: Let X = x3 .8: Example 163. x6 + y4 Figure 2.y)→(0. ¬ ¬ ¬ x5 y 3 ¬ y8 ¬ ¬≤ = y4 . ¬ ¬ ¬ x5 y 3 ¬ X 5/3 Y 3/2 ¬ ¬= . Figure 2. y) → (0. Y = y 2 . The limit does not exist.6: (x. 1+y → −∞. ◭ 162 Example Find (x. x2 as x → 0. ¬ x6 + y 4 ¬ X2 + Y 2 as (x. ◭ Free to photocopy and distribute 101 . 1+x → +∞. Observe that if |x| ≤ |y|. When x = 0.

0).0) lim ((x − 1)2 + y 2 ) loge ((x − 1)2 + y 2 ) |x| + |y| . Figure 2.10: Example 165. (x. y = t3 . But when y = 0.y)→(0. Solution: ◮ sin(x4 ) + sin(y 4 ) ≤ x4 + y 4 and so ¬ ¬ ¬ sin(x4 ) + sin(y 4 ) ¬ ¬ ¬≤ Ô ¬ ¬ x4 + y 4 x4 + y 4 → 0.y)→(0. as t → 0. Figure 2. ◭ 165 Example Find lim sin(x4 ) + sin(y 4 ) Ô (x. y) → (0.12: Example 163. as (x. 0). ◭ 166 Example Find lim sin x − y x − sin y . the function is 0. Solution: ◮ When y = 0 we have 2(x − 1)2 ln(|1 − x|) 2x ∼− .Limits xy 6 x6 + y 8 163 Example Find (x.0) x4 + y 4 . 164 Example Find (x. Figure 2. Solution: ◮ Putting x = t4 .y)→(0.0) lim .9: Example 164.y)→(0. we ﬁnd xy 6 x6 + y8 = 1 2t2 → +∞. Thus the limit does not exist.0) Free to photocopy and distribute 102 . |x| |x| and so the function does not have a limit at (0. ◭ Figure 2.11: Example 166.

It may occur that does not.6 For what c will the function Ô 1 − x2 − 4y 2 . lim f (x. as x → 0. lim f (x. y) → log(x + y). but that 3.y )→(0. y) exists. if x2 + 4y 2 ≤ 1. y) = c. both exist. It may occur that lim (x.3. y) → 2 x + y2 Problem 2. y) 2. y) → (x0 . If the iterated limits exist and lim then (x.1 Sketch the domain of deﬁÔ nition of (x.y)→(x0 . y) = lim lim f (x.7 Find (x.4 Find y 2 ) sin 1 .y)→(x0 . y) .y)→(x0 .y)→(x0 . y) . x→x0 lim y→y0 lim f (x.3. y0 ).3.y0 ) Homework Problem 2. then each of the iterated limits lim y→y0 x→x0 lim f (x. The following possibilities might occur. x2 + y 2 Free to photocopy and distribute 103 . y) exists. If (x. y) does not exist. f (x. (x.Chapter 2 Solution: ◮ When y = 0 we obtain sin x x → 1.y )→(0. y) and lim y→y0 lim f (x. Thus the limit does not exist.3.y0 ) y→y0 x→x0 x→x0 y→y0 lim f (x. y) .0) (x. y) → 4 − x2 − y 2 . When y = x the function is identically −1.y )→(0. Problem 2. if x2 + 4y 2 > 1 be continuous everywhere on the xy-plane? Problem 2.2 Sketch the domain of deﬁnition of (x. nition of (x.5 Find (x. but one of the iterated limits 4. x Problem 2.y0 ) x→x0 lim f (x. y) exists.3. it may be that the limits y→y0 lim x→x0 lim f (x.2) lim sin xy . xy lim (x2 + Problem 2. y) does not exist. y) = lim lim f (x.3. Problem 2.0) lim Ô x2 + y 2 sin 1 . ◭ If f : R2 → R. These are called the iterated limits of f as (x. 1. lim f (x.y0 ) y→y0 x→x0 x→x0 y→y0 lim f (x.3 Sketch the domain of deﬁ1 .3.

if f (h) = 0. x4 + y 4 Problem 2.13 Prove that x→0 lim Problem 2.+∞) 1 1 + y sin x y if x = 0.y )→(0. since lim h→0 h lim h3 + 2h2 h = lim h2 + 2h = 0.3. Let f : R → R be a function.3.0) lim but still not exist.9 Find lim 2x2 sin y 2 + y 4 e−|x| Ô . + y2 + z2 Problem 2.3. y) = x sin 0 Problem 2. y) do not exist. x2 + y 2 Prove that (x. We write f (h) = o (h) if f (h) goes faster to 0 than h. The limit condition above is equivalent to saying that lim g(x) − g(a) − g ′ (a)(x − a) = 0. x−a g(a + h) − g(a) − g ′ (a)(h) h = 0.0) x 2 lim x2 y 2 z 2 = 0.3. h3 + 2h2 = o (h). |y|) Ô . y = 0 otherwise lim f (x. h→0 lim Free to photocopy and distribute .Deﬁnition of the Derivative Problem 2.11 Prove that lim lim x−y x+y = 1 = − lim x−y exist?. Problem 2.3.0) x 2 y 2 lim x2 y 2 does + (x − y)2 2.12 Let f (x. let us introduce some necessary notation. x→0 y →0 y →0 x→0 lim x→0 x 2 y 2 lim = 0.4 Deﬁnition of the Derivative Before we begin. Does (x. y) exists.y . 104 x→a or equivalently. y →0 y →0 x 2 y 2 lim x2 y 2 + (x − y)2 x2 y 2 + (x − y)2 = 0.z)→(0. y) lim f (x. (x.8 Find lim max(|x|.y )→(0.y )→(+∞. h→0 h→0 We now deﬁne the derivative in the multidimensional space Rn .0 that the iterated limits lim and lim y →0 x→0 lim f (x.10 Demonstrate that (x.y )→(0. a function g : R → R is said to be differentiable at x = a if the limit x→a lim g(x) − g(a) = g ′ (a) x−a exists. that is. x+y lim x−y x+y and that . Recall that in one variable. but x→0 y →0 (x.0) (x.0.y )→(0. For example.3.

the corresponding deﬁnition in higher dimensions involves quotients of norms. The condition for differentiability at a is equivalent to f (x) − f (a) = Da (f )(x − a) + o (||x − a||) . i. as h → 0. x→a ||x − a|| Equivalently. and f (a + h) − f (a) = L(h) + o (||h||) . we have f (a + h) − f (a) = Da (f )(h) + o (||h||) . Da (f ) is uniquely determined. called the derivative of f at a. Since A is open. ||Da (f )(v) − L(v)|| ≤ ||Da (f )(h) − f (a + h) + f (a)|| = o (||h||) + o (||h||) = o (||h||) . L(v) = Da (f )(v). as x → a. as h → 0. Observe that since we may not divide by vectors. A function f : A → Rm is said to be differentiable at a ∈ A if there is a linear transformation.e.. f is differentiable at a if there is a linear transformation Da (f ) such that f (a + h) − f (a) = Da (f )(h) + o (||h||) . 168 Theorem If A is an open set in deﬁnition 167. L(v) = Da (f )(v). By the triangle inequality. a + h ∈ A for sufﬁciently small ||h||. observe that Da (f )(v) − L(v) = Da (f )(h) − f (a + h) + f (a) + f (a + h) − f (a) − L(h).Chapter 2 We may write this as g(a + h) − g(a) = g ′ (a)(h) + o (h) . We must prove that ∀v ∈ Rn . Proof: Let L : Rn → Rm be another linear transformation satisfying deﬁnition 167. The above analysis provides an analogue deﬁnition for the higher-dimensional case. Da (f ) : Rn → Rm such that ||f (x) − f (a) − Da (f )(x − a)|| lim = 0. u Free to photocopy and distribute 105 +||f (a + h) − f (a) − L(h)|| . 167 Deﬁnition Let A ⊆ Rn . as h → 0 This means that ||L(v) − Da (f )(v)|| → 0. Now. By deﬁnition. completing the proof.

| h k | ≤ ¬¬ h ¬¬¬¬ k ¬¬ = o ¬¬ h ¬¬ . we¬¬ ¬¬ by the¬ Cauchy-Buniakovskii-Schwarz ¬¬→¬¬ have ¬ →¬¬ →•→ − − − ¬¬− ¬¬¬¬→¬¬ ¬¬− ¬¬ inequality. 171 Theorem Suppose A ⊆ Rn is open and f : A → Rn is differentiable on A. − x = = → → − − → → − − As ( h . we must shew that x→a lim f (x) = f (a). implies continuity at that point. ◭ Just like in the one variable case. which proves the assertion. ◭ 170 Example Let f : R3 × R3 − − (→. → + k ) − f (→. for any a ∈ Rn . then Da (L) = L. 106 Free to photocopy and distribute .→) f ( h . Shew that f is differentiable and that − − → → − − − → →− − − y D(→. then Da (f ) is not uniquely determined. 0 ). k ) = →• k + h •→. k ) → ( 0 . Any linear transformation T will satisfy the deﬁnition. But by linearity ||L(x) − L(a) − L(x − a)|| = ||L(x) − L(a) − L(x) + L(a)|| = ||0|| = 0. whence the claim follows. For ||x − a|| < δ holds only for x = a. 169 Example If L : Rn → Rm is a linear transformation. Proof: Given a ∈ A. →) x y → → R →•→ − − x y be the usual dot product in R3 . and ||f (x) − f (a) − T (x − a)|| = ||0|| = 0. then the claim is established. →) = x y x y → − − → − − − − (→ + h )•(→ + k ) − →•→ x y x y − − − − − − →•→ + →•→ + →•→ + →•→ − →•→ − − − − x y h y h k x y x k → →•→ →•→ − − − − − →• k + h y + h k . we know that Da (L) uniquely determined. Thus if L satisﬁes deﬁnition 167. as T (x − a) = T (0) = 0. differentiability at a point. and so f (x) = f (a). x x y Solution: ◮ We have → − − → − − − − f (→ + h . Then f is continuous on A.Deﬁnition of the Derivative If A = {a}. Solution: ◮ Since Rn is an open set. a singleton. identically.

. f (→ ) = ¬¬→¬¬ be the usual norm in Rn .4. . and put ¾ f1 (x1 . x2 . . xj . . x2 . . x2 . . x2 . proving the theorem. . xn ) f2 (x1 . Free to photocopy and distribute 107 . v ¬¬− ¬¬ x → − − for → = 0 . f : A → Rm . The partial derivative ∂fi ∂xj (x) = lim ∂fi (x) is deﬁned as ∂xj . h→0 fi (x1 . fm (x1 . . xj + h. 172 Deﬁnition Let A ⊆ Rn . xn ) − fi (x1 . . . . we simply keep the other variables ﬁxed and differentiate with respect to the j-th variable. but that f is not differentiable x → − at 0 . Prove that with x x x → •→ − − x v − Dx (f )(→ ) = ¬¬→¬¬ . x2 . xn ) . n ≥ ¬¬− ¬ − 1. .4. xn ) h whenever this limit exists. . ¿ f (x) = .5 The Jacobi Matrix We now establish a way which simpliﬁes the process of ﬁnding the derivative of a function at a given point. .1 Let L : R3 → R3 be a linear transformation and R3 F : → − x → R3 → × L(→ ) .2 ¬Let f : Rn → R. x x ¬¬→ ¬¬2 − − ¬¬− ¬¬ = →•→. − − x x Problem 2. u Homework Problem 2. . . . xn ) Here fi : Rn → R. . . x x 2. .Chapter 2 Since f is differentiable at a we have f (x) − f (a) = Da (f )(x − a) + o (||x − a||) . . . To ﬁnd partial derivatives with respect to the j-th variable. . . . . → Shew that F is differentiable and that → − → − → − − − Dx (F )( h ) = → × L( h ) + h × L(→). . as x → a. . . . and so f (x) − f (a) → 0.

z). x2 . (x) (x) ··· ··· . The following theorem will allow us to determine the matrix representation for Da (f ) under the standard bases of Rn and Rm . f ′ (x) = ∂x1 . x2 . . y. (x.y. diff(f(x.z)->x+yˆ2+zˆ3+3*x*yˆ2*zˆ3. . be the standard basis for Rn . y. . . . fm (x1 . it can be represented by aid of matrices. xn ) Suppose A ⊆ Rn is an open set and f : A → Rm is differentiable. . . . z) = x + y 2 + z 3 + 3xy 2 z 3 then ∂f ∂x ∂f ∂y and ∂f ∂z (x.z). Since the derivative of a function f : Rn → Rm is a linear transformation. xn ) f2 (x1 . diff(f(x. Then each ∂fi partial derivative (x) exists. .z). (x. f (x) = . ∂f1 ∂xn ∂f2 ∂xn . . . . and the matrix representation of Dx (f ) with ∂xj respect to the standard bases of Rn and Rm is the Jacobi matrix ¾ ∂f1 ∂x1 ∂f2 (x) (x) ∂f1 ∂x2 ∂f2 ∂x2 . . z) = 2y + 6xyz 3 . . .z). > > > > f:=(x. .y. we must compute Dx (f )(→j ). z) = 1 + 3y 2 z 3 . To obtain the e − Jacobi matrix. 174 Theorem Let ¾ ¿ f1 (x1 .y. . ¿ (x) (x) . . ∂fn (x) ∂x1 ∂fn (x) ∂x2 ··· ∂fn (x) ∂xn − Proof: Let →j .y. and f (x. y. xn ) . diff(f(x. which will give us the j-th e Free to photocopy and distribute 108 . . .The Jacobi Matrix 173 Example If f : R3 → R.x). x2 . . The Maple™ commands to ﬁnd these follow.y). 1 ≤ j ≤ n. . y. z) = 3z 2 + 9xy 2 z 2 .

xn ) − f (x1 . . . . y) x ∂z ¿ x+z 1 y y 0 ¿ . . . y) ∂x ∂f2 (x. however. . . xj + ε. xn ) − εDx (f )(→j )|| e |ε| . xj + h. . ◭ Free to photocopy and distribute 109 . . . . Let f ′ (x) = (Jij ). x2 . . y) ∂x ∂f1 (x. . x2 . and refer to f ′ when we mean the Jacobi matrix of f . . . u Strictly speaking. . and observe that ¾ J1j J2j . . y) ∂y ∂f2 (x. . . − and put y = x + ε→j . . x2 .Chapter 2 column of the Jacobi matrix. . y) y ∂z = 1 ∂f2 (x. . . . . x2 . y) = ∂f1 (x. xn ) − fi (x1 . It is a matrix representation of the derivative in the standard basis of Rn . . . . . Solution: ◮ The Jacobi matrix is the 2 × 3 matrix ¾ f ′ (x. . Notice that e ||f (y) − f (x) − Dx (f )(y − x)|| = ||y − x|| − ||f (x1 . . . . . Compute the Jacobi matrix of f . xj . xj + h. . and so. . ε ∂xj → 0. ε→0 This ﬁnishes the proof. the Jacobi matrix is not the derivative of a function at a point. x2 . ε ∈ R. y) ∂y ¾ ∂f1 (x. xn ) − fi (x1 . 175 Example Let f : R3 → R2 be given by f (x. Since the sinistral side → 0 as ε → 0. . . . . . xj . . . loge xy). xn ) − εJij | |ε| This entails that Jij = lim fi (x1 . x2 . the so does the i-th component of the numerator. . . We will abuse language. xj . y) = (xy + yz. . . . . . Jnj ¿ − Dx (f )(→j ) = e . xn ) ∂fi = (x) . |fi (x1 .

θ) = (ρ cos θ sin φ. θ. 0) = (0. x if Observe that f is not continuous at (0. is not true. 0)). We have ¾ cos θ sin θ 0 f ′ (ρ. Free to photocopy and distribute 110 .The Jacobi Matrix 176 Example Let f (ρ. You should also notice that both partial derivatives are not continuous at (0. xy = 0. (0. y = 0. 0) = 0 but f (x. and let f : Rn → Rm . 1 177 Example Let f (ρ.. then f is differentiable on A. z) = (ρ cos θ. We have. 0) (f (0. it is not differentiable there. ρ sin θ sin φ. ρ cos φ) be the function which changes from spherical coordinates to Cartesian coordinates. 178 Example Let f : R2 → R be given by y f (x. Thus even if both partial derivatives exist at ∂x ∂y (0. y) = 1 if if x = 0. and hence. Thus differentiability at a point implies that the partial derivatives of the function exist at the point. θ. The Jacobi matrix provides a convenient computational tool to compute the derivative of a function at a point. The converse. however. however. fm If each of the partial derivatives Dj fi exists and is continuous on A. 0) = 1. the following. ρ sin θ. z) be the function which changes from cylindrical coordinates to Cartesian coordinates. We have ¾ cos θ sin φ sin θ sin φ cos φ f ′ (ρ. z) = −ρ sin θ ρ cos θ 0 0 ¿ 0 . ¿ 179 Theorem Let A ⊆ Rn be an open set. 0). φ. y) = 1 for values arbitrarily close to (0. Put f = . We have ∂f ∂f however. 0). θ) = ρ cos θ cos φ −ρ sin φ sin θ ρ sin θ cos φ −ρ sin φ ρ cos θ sin φ 0 ¿ . ¾ f1 f2 . 0) is no guarantee that the function will be differentiable at (0.. φ.

Chapter 2 The concept of repeated partial derivatives is akin to the concept of repeated differentiation. 1. ∂x ∂x Hence. 2 z xy . z ∂x ∂ ez log xy ∂x ∂z z = log xy + (xy)z . v. √ e 2 which is at the desired point. ∂z ∂y . Find and at (1. ∂ xy z ∂y = = ∂ ∂y (xy)z = = ∂ ∂y xy 2 z 3 = ∂ xy log z e ∂y x log z + ∂ xy ∂z z ∂y +1+1+3 ∂z ∂x = 0 =⇒ ∂z ∂x 1 =− . 180 Example Let f (u. −1. The following examples should be self-explanatory. Similarly with the concept of implicit partial differentiation. w) at (1. v. ◭ 2 181 Example The equation z xy + (xy)z + xy 2 z 3 = 3 deﬁnes z as an implicit ∂z ∂z function of x and y. ez log xy ∂y ∂z z log xy + ∂y y (xy)z . π 4 ). ∂2 ∂u∂v f (u. ∂x x ∂ ∂z xy 2 z 3 = y 2 z 3 + 3xy 2 z 2 . Determine Solution: ◮ We have ∂2 ∂u∂v (eu v cos w) = ∂ ∂u (eu cos w) = eu cos w. 1) we have ∂x ∂z ∂x Similarly. ∂x ∂y Solution: ◮ We have ∂ ∂x z xy = = ∂ (xy)z = ∂ xy log z e ∂x xy ∂z y log z + z xy . 1). w) = eu v cos w. 111 2xyz 3 + 3xy 2 z 2 Free to photocopy and distribute . at (1. 1.

v) = u + v . we have the following rules of differentiation.The Jacobi Matrix Hence. y) = Find (f ◦ h)′ (x. Then we have • Addition Rule: Dx ((f + αg)) = Dx (f ) + αDx (g). be differentiable on A. ey+z 1 y+z (x2 + y)ey+z 1 x +y 2 ¿ f ′ (h(x. h : B → Rl be differentiable on B. 4 182 Example Let ¾ uev uv ¿ f (u. (2. the Chain Rule takes the alternative form when applied to the Jacobi matrix. and h′ (x. • Chain Rule: Dx ((h ◦ f )) = Df (x) (h) ◦ (Dx (f )). h(x. y) = Observe also that ¾ 2x 0 1 0 1 1 . v) = . Let A ⊆ Rn . g : A → Rm . B ⊆ Rm be open sets f. y). α ∈ R. Free to photocopy and distribute 112 . at (1. 1. ev 1 v uev 1 u ¿ f ′ (u. (h ◦ f )′ = (h′ ◦ f )(f ′ ).1) +1+2+3 ∂z ∂y = 0 =⇒ ∂z ∂y 3 =− . Since composition of linear mappings expressed as matrices is matrix multiplication. Solution: ◮ We have ¾ x2 + y y+z . y)) = . 1) we have ∂z ∂y ◭ Just like in the one-variable case. and f (A) ⊆ B.

v(x. y) = y + z. y)) = 2xz Thus å è ä ey+z . Solution: ◮ Put g : R3 → R2 . f ′ (u. f (u. y)) å ¾ è z ¿ = å 2xz 2xz 2 ey+z 0 1 x 1 è . x2 + 2y + z x2 + y ◭ 183 Example Let f : R2 → R. y) ∂x ∂h (x. y. Observe 0 1 x 1 ç . v) = u2 + ev . g ′ (x. Put h(x. y) = z 0 ä u(x. v : R3 → R u(x. 0 ey+z = Free to photocopy and distribute 2x2 z + ey+z 113 . y) ∂y ∂h (x. y) = (f ′ (g(x. y) ∂z = h′ (x. z) . y) v(x. g(x. u. Find the partial derivatives of h. z) v(x. y) = that h = f ◦ g.Chapter 2 Hence (f ◦ h)′ (x. ç ∂h (x. y) ¾ ¿ ey+z = ¾ (x2 + y)ey+z ¾ 2x 1 0 1 1 1 x2 + y 2 ¿ 1 y+z 2xe y+z 0 ¿ y+z (1 + x + y)e 2 (x + y)e 1 2 y+z = 2x 2xy + 2xz . y))h′ (x. y. y) = f ′ (h(x. y) = xz. f ′ (h(x. Now. y)))(g ′ (x. y = xz y+z . y) = f u(x. v) = 2u ev .

y) dy exists as a continuously differentiable function of x on [a. f (x. y) = 2xz 2 . The above implies that (x2 − 1) · F ′ (x) = 2x = = = = = Free to photocopy and distribute π/2 0 π/2 0 dθ sin2 θ + x2 cos2 θ 2 x2 cos2 θ + sin θ − 1 sin2 θ + x2 cos2 θ π/2 dθ (x2 − 1) cos2 θ dθ π 2 π − 0 π/2 − 2 tan2 θ + ¬ 2 x 0 π 1 tan θ ¬π/2 − arctan ¬ 2 x x 0 π π − . Then f (x. b] × Y → R be a function. F ′ (x) = π/2 0 ∂ ∂x log(sin2 θ + x2 cos2 θ)dθ cos2 θ sin2 θ + x2 cos2 θ dθ π/2 0 = 2x . with derivative d dx 185 Example Prove that π/2 Y Y (x. Solution: ◮ Differentiating under the integral.The Jacobi Matrix Equating components. we obtain ∂h (x. y) are continuous in the vari∂x ables x and y jointly. y) dy = Y ∂ ∂x f (x. F (x) = 0 log(sin2 θ + x2 cos2 θ)dθ = π log x+1 2 . with [a. (x. y) and f (x. b]. 184 Theorem (Differentiation under the integral sign) Let f : [a. ∂x ∂h ∂y ∂h ∂z ◭ Under certain conditions we may differentiate under the integral sign. 2 2x sin2 θ + x2 cos2 θ sec2 θdθ 114 . y) = 2x2 z + ey+z . and Y being a closed and bounded ∂ subset of R. y) dy. b] being a closed interval. y) = ey+z . Suppose that both f (x.

with a ≥ 0. Finally x+1 2 . cos2 θdθ = π 2 For x = 1 one sees immediately that F ′ (1) = 2 agreeing with the formula. F ′ (x) = π/2 π/2 0 . 2 ◭ Homework Free to photocopy and distribute 115 . Solution: ◮ Put I(a) = 0 sin2 ax I ′ (a) = = +∞ 0 +∞ 0 +∞ 2x sin ax cos ax x2 sin 2ax dx x sin u du u dx = = Integrating each side gives I(a) = 0 π .Chapter 2 which in turn implies that for x > 0. F ′ (x) = 2x x2 − 1 π 2 − π 2x = π x+1 . π =⇒ F (x) = π log(x + 1) + C. +∞ 186 Example Given that 0 sin x x dx = π 2 +∞ . +∞ dx. the interval of integration in the above theorem need not be compact. x+1 Since F (1) = thus 0 log 1dθ = 0. Now. The desired integral is I(1) = . Differentiating x2 both sides with respect to a. we gather that C = 0. we gather that C = −π log 2. x = 1. π Since I(0) = 0. F (x) = π log(x + 1) − π log 2 = π log ◭ Under certain conditions. and making the substitution u = 2ax. compute 0 sin2 x x2 dx. 2 π 2 a + C.

. y) = xy é and ¾ g(x.5. If m = 1. +∞[×]0.4 Let f (x. 1). If undeﬁned. if at all deﬁned. ﬁnd . it is called a scalar ﬁeld.10 Let z be an implicitlydeﬁned function of x and y through the equation (x + z)2 + (y + z)2 = 8. f (x. 0). ∂r Problem 2. a ∂f (x. y) ∂f (x. Find ∂z at (1. 0.1 Let f : [0. 1). f (r. f : R2 → R. explain.5. determine .2 Let f : R2 → R. Find the partial derivatives with respect to x and y of x2 y r2 ∂f ∂r .6 Gradients and Directional Derivatives Rn x → Rm A function f : → f (x) is called a vector ﬁeld.8 Assuming that the equation xy 2 + 3z = cos z 2 deﬁnes z implicitly as a ∂z function of x and y.5.5.5. 2. where n is a constant. 1. Determine n such that ∂f 1 ∂ = 2 ∂t r ∂r Problem 2. y. evaluate a a b 0 Problem 2. Problem 2. explain. g(x. ∂x æ Problem 2.7 Prove that +∞ 0 é æ f (x. y) = g(t) dt. ∂x Problem 2. y) Find and .5.Gradients and Directional Derivatives Problem 2.5 Suppose g : R → R is continuous and a ∈ R is a constant. (x2 + a2 )2 b Problem 2.5.9 If w = euv and u = r + s.5.5. Problem 2.3 Let f : R2 → R2 and g : R3 → R2 be given by dx = x2 + a2 0 dx . f (x. 1). ∂x ∂y Problem 2. y 2 ). x 2 . +∞[→ 2 R. y) = xy 2 x2 y é æ arctan ax − arctan x π dx = log π. 187 Deﬁnition Let f : Free to photocopy and distribute Rn x → R → f (x) 116 . z) = x − y + 2z xy Compute (f ◦ g)′ (1. Compute (g ◦ f )′ (1. If undeﬁned. ∂w v = rs. t) = tn e−r /4t . y) = min(x. y) = x−y x2 y 2 x+y ¿ x+y Find (g ◦ f )′ (0.6 Given that 1 b arctan . if at all deﬁned.5.

Then (f ◦ c)(t0 ) = f (c(t)) = K. and using the chain rule f ′ (c(t0 ))c′ (t0 ) = 0. . ∂f ∂xn (x) The gradient operator is the operator ¾ ∂ ∂x1 ∂ ∂x2 . Let K ∈ R be a constant. Since c′ (t0 ) is tangent to the surface and its dot product with ∇f (x) is 0. Free to photocopy and distribute 117 . 1. . which translates to (∇f (x))•(c′ (t0 )) = 0. Proof: Let c: R t → Rn → c(t) be a curve lying on this surface. ∂ ∂xn ¿ ∇= . u Let θ be the angle between ∇f (x) and c′ (t0 ). . Choose t0 so that c(t0 ) = x. . The gradient of f is the vector deﬁned and denoted by ¾ ∂f ∂x1 ∂f ¿ (x) (x) . ∇f (x) = ∂x2 . Then ∇f (x) is orthogonal to the surface implicitly deﬁned by f (x) = K. we conclude that ∇f (x) is normal to the surface. 2). Since |(∇f (x))•(c′ (t0 ))| = ||∇f (x)||||c′ (t0 )|| cos θ.Chapter 2 be a scalar ﬁeld. 189 Example Find a unit vector normal to the surface x3 + y 3 + z = 4 at the point (1. and assume that f is differentiable in A. ∇f (x) is the direction in which f is changing the fastest. 188 Theorem Let A ⊆ Rn be open and let f : A → R be a scalar ﬁeld.

Free to photocopy and distribute 118 ¾ ¿ . y.Gradients and Directional Derivatives Solution: ◮ Here f (x. 1. Normalising this vector we obtain 2e2 1 1 √ 2 . 3). Here ¾ yez xez xyez ¿ ∇f (x. Normalising this vector we obtain 1 ¾ 3 ¿ √ 19 3 √ . 2. 2) is 3 . z) = xyez at the point (2. y. z) = x + y 2 − z 2 . 2). z) = x3 + y 3 + z − 4 has gradient ¾ 3x2 3y 2 1 ¿ ∇f (x. y. 1. 5 2 ◭ 191 Example Let f : R3 → R be given by f (x. Find the equation of the tangent plane to f at (1. Solution: ◮ The direction is that of the gradient vector. 19 1 √ 19 ◭ 190 Example Find the direction of the greatest rate of increase of f (x. y. 2) becomes 2e2 . z) = ¾ ¿ 3 which at (1. y. z) = ¾ e2 ¿ which at (2. 1.

193 Example If f (x. 3). y. ◭ 192 Deﬁnition Let f : be a vector ﬁeld with Rn x → Rn → f (x) ¾ ¿ f1 (x) f (x) = f2 (x) . 2. Now ¾ 1 2y −2z ¿ ∇f (x. or x + 4y − 6z = −9.Chapter 2 Solution: ◮ A vector normal to the plane is ∇f (1. 2 2 Free to photocopy and distribute 119 . y 2 . 3). fn (x) The divergence of f is deﬁned and denoted by divf (x) = ∇•f (x) = ∂f1 ∂x1 (x) + ∂f2 ∂x2 (x) + · · · + ∂fn ∂xn (x) . yez ) then divf (x) = 2x + 2y + 2yzez . The equation of the tangent plane is thus 1(x − 1) + 4(y − 2) − 6(z − 3) = 0. z) = (x2 . . 2. z) = which is ¾ 1 4 −6 ¿ at (1. . y. .

y. . . . Then the curl of (g1 . z) = x3 + y 3 − z 2 in the ¾ ¿ 1 direction of 2 . . y. . . 0. gi2 . gn−2 ) ¾ curl(g1 . . v 198 Theorem Let A ⊆ Rn be open and let f : A → R be a scalar ﬁeld. . y 2 . and − − assume that f is differentiable in A. y. . g)(x. z. Let → ∈ Rn \ { 0 } be such that → + t→ ∈ A for v x v − sufﬁciently small t ∈ R. w) = (0. g2 .. . . gin ). . 199 Example Find the directional derivative of f (x. . z. 0. w2 ). Let → ∈ Rn \ {0} be such that x + t→ ∈ A v v for sufﬁciently small t ∈ R.. z)) = ∇ × f (x. y. g(n−2)1 (x) e2 ∂ ∂x2 g12 (x) g22 (x) . z. y. . 196 Example If f (x. and assume → − − − − that f is differentiable in A. . y.. Then the directional derivative of f in the direction of → v → is given by − at the point x − v ∇f (x)•→. 1 ≤ k ≤ n − 2 be vector ﬁelds with gi = (gi1 . . yez ) then curlf ((x. z) = (ez )i.. . t − Some authors require that the vector → in deﬁnition 197 be a unit vector. w) = det ∂x1 exyz 0 ¾ ¿ 2 2 e2 ∂ ∂x2 0 0 e3 ∂ ∂x3 0 z e4 ∂ ∂x4 w2 0 = (xz 2 exyz )e4 . g2 . . . . 3 Free to photocopy and distribute 120 . 0. . 197 Deﬁnition Let A ⊆ Rn be open and let f : A → R be a scalar ﬁeld. Then the directional derivative of f in the direction of → at the point x is deﬁned and denoted by − v − D→ f (x) = lim v t→0 − f (x + t→) − f (x) v . g(n−2)2 (x) ··· ··· ··· . g(x. en ∂ ∂xn g1n (x) g2n (x) . 0) then e1 ∂ curl(f. g(n−2)n (x) ¿ . .Gradients and Directional Derivatives 194 Deﬁnition Let gk : Rn → Rn . y. w) = (exyz . z. 195 Example If f (x. z) = (x2 . gn−2 )(x) = det e1 ∂ ∂x1 g11 (x) g21 (x) .

and let U. Problem 2.2 Let f (x. 1) with the coordinate axes.6. ∇ × (∇φ) = 0 4. Problem 2. ∇ •φV = φ∇ •V + V•∇φ 2. Prove that 1. 2) = u → → − − 2 i − j . ∇ × φV = φ∇ × V + (∇φ) × V → − 3. 1).Chapter 2 Solution: ◮ We have ¾ 3x2 3y 2 −2z ¿ ∇f (x. If ∇f (1. y. 1).4 Find the point on the surface x2 + y 2 − 5xy + xz − yz = −3 for which the tangent plane is x − 7y = −6. y) in the di− rection of the unit vector → . y.6. 1. Problem 2. −1. z) = xey z .6. Problem 2.10 Let φ : R3 → R be a scalar ﬁeld. y. y) = ex cos 2y at (0.11 Find the angles made by √ the gradient of f (x. y. y) denote the u directional derivative of f at (x. 0.1 Let f (x.6. ∇ •(∇ × V) = 0 5. v ◭ Homework Problem 2. V : R3 → R3 be vector ﬁelds.6.6.9 Find the point on the surface 2x2 + xy + y 2 + 4x + 8y − z + 14 = 0 for which the tangent plane is 4x + y − z = 0.6.6. 0).2).6.6.6.6 Let D→ f (x. 5 5 Free to photocopy and distribute 121 . 1.3 Find the tangent plane to x2 the surface − y 2 − z 2 = 0 at the point 2 (2. Problem 2. 1). z) = 3xy − 9xz 2 + y increases most rapidly at the point (1.7 Use a linear approximation of the function f (x. 0) to estimate f (0.8 Prove that ∇ • (u × v) = v • (∇ × u) − u • (∇ × v). Problem 2. ∇(U•V) = (U•∇)V + (V•∇)U + U × (∇ × V) + +V × (∇ × U) Problem 2. Problem 2. z) Find (∇ × f )(2. z) = and so − ∇f (x. 2). − Problem 2. ¾ Problem 2. 4 ) f (1. = xz z ¿ exy . Find (∇f )(2. 1.1.5 Find a vector pointing in the direction in which f (x. y) = x 3 + y at the point (1. ﬁnd D( 3 . z)•→ = 3x2 + 6y 2 − 6z. y.

dl are the components of vectors in R3 . 200 Deﬁnition (Einstein’s Summation Convention) In any expression containing subscripted variables appearing twice (and only twice) in any term. we will enclose any terms under consideration with · . the ij-th entry (AB)ij of the product of two matrices A ∈ Mm×n (R) and B ∈ Mn×r (R) can be written as n (AB)ij = k=1 Aik Bkj = Ait Btj . ck . 205 Example Demonstrate. a c 203 Example Using Einstein’s Summation convention. d a c respectively. − − − → − → 202 Example Given that ai .Levi-Civitta and Einstein 2. →. In order to emphasise that we are using Einstein’s convention. what is the meaning of ai bi ck dk ? Solution: ◮ We have 3 ai bi ck dk ◭ = i=1 ai bi ck dk − − → = → • b ck dk a − − → = →• b a 3 k=1 − − − → − → ck dk = (→• b )(→• d ). we are dealing in the space R3 and so. via Einstein’s Summation convention. →. unless otherwise noted. that if A. subscripts are in the set {1. the subscripted variables are assumed to be summed over. 3}. then tr (AB) = tr (BA) . Free to photocopy and distribute 122 . 201 Example Using Einstein’s Summation convention.7 Levi-Civitta and Einstein In this section. the dot product of two − − vectors → ∈ Rn and → ∈ Rn can be written as x y →•→ = − − x y n x i yi = i=1 x t yt . B are two n × n matrices. 2. bj . 204 Example Using Einstein’s Summation convention. the trace tr (A) of a square n matrix A ∈ Mn×n (R) is tr (A) = Att = t=1 Att . b .

for any other permutation. the permutations 123 (identity). 2. a Recall that a permutation of distinct objects is a reordering of them. ◭ 206 Deﬁnition (Kroenecker’s Delta) The symbol δij is deﬁned as follows: ´ δij = 0 1 if i = j if i = j. Atk Bkt . 213 → 123. 231. 3} εjkl = −1 +1 if if 1 j 1 j 2 2 k 3 l 3 l k is an odd permutation is an even permutation Free to photocopy and distribute 123 . and 312 are even. 321 → 123. 3} can be classiﬁed into even or odd. 2. We start with the identity permutation 123 and say it is even. 321. since the indices are dummy variables and can be exchanged. Now. Since 132 → 123. and odd if it takes an odd number of transpositions to regain the identity permutation. the permutations 132. The 3! = 6 permutations of the index set {1. 208 Example We see that 3 3 δij ai bj = i=1 j=1 δij ai bj = k=1 − − → ak bk = →• b . 209 Deﬁnition (Levi-Civitta’s Alternating Tensor) The symbol εjkl is deﬁned as follows: 0 if {j.Chapter 2 Solution: ◮ We have tr (AB) = tr ((AB)ij ) = tr ( Aik Bkj ) = and tr (BA) = tr ((BA)ij ) = tr ( Bik Akj ) = Btk Akt . from where the assertion follows. k. and 213 are odd. we will say that it is even if it takes an even number of transpositions (switching only two elements in one move) to regain the identity permutation. 312 → 132 → 123. Since 231 → 132 → 123. 3 207 Example It is easy to see that δik δkj = k=1 δik δkj = δij . l} = {1.

this means that given x2 ∈ Rn . Again. − − − 212 Example Let →. Hence. − − − x y jkl k l ej 211 Example If A = [aij ] is a 3×3 matrix. det A = εijk a1i a2j a3k . Demonstrate that x i yi z j − − − = (→ •→)→ . if we differentiate T at x0 again. Rm ). x → Dx (f ) where L (Rn . →. ε132 = ε321 = ε213 = −1. if one subindex is repeated we have εrrs = εrsr = εsrr = 0.Extrema Also.7. This means that we have a function A → L (Rn . k = e1 e2 →. Rm ). Dx0 (f )(x1 ))(x2 ) ∈ m R . using the Levi-Civitta alternating tensor. x2 ) → Free to photocopy and distribute 124 . given x1 ∈ Rn . Let A ⊂ Rn and f : A → Rm be differentiable on A. 210 Example Using the Levi-Civitta alternating tensor and Einstein’s summation → − − − − → − → convention. We know that for ﬁxed x0 ∈ A. if i = →. Hence. then − e3 → × → = ε (a b )→ . let us consider the following. the derivative will provide us with information about the extrema. → . In particular. we obtain a linear transformation Dx0 (T ) = 2 Dx0 (Dx0 (f )) = Dx0 (f ) from Rn to L (Rn . Then x y z → • (→ × →) = − − − x y z − − xi (→ × →)i y z = xi εikl (yk zl ) . ε123 = ε231 = ε312 = 1. x y z 2. → be vectors in x y z R3 . j = →. Dx0 (f ) is a linear transformation from Rn to Rm . Homework − − − Problem 2. To deﬁne an analogue for the second derivative. and the “second derivative” will provide us with information about the nature of these extreme points. Thus the function Bx 0 : Rn × Rn → L (Rn . → be vectors in R3 . the cross product can also be expressed. then. As in the one-variable case. we have 2 2 Dx0 (f )(x1 ) ∈ L (Rn . Rm ) 2 Dx0 (f )(x1 .8 Extrema We now turn to the problem of ﬁnding maxima and minima for vector functions.1 Let →. Rm ) denotes the space of linear transformations from Rn to Rm . x2 ) (x1 . Rm ) T : .

216 Deﬁnition Let A ⊆ Rn be an open set. Then the matrix of Dx (f ) : Rn × Rn → R with respect to the standard basis is given by the Hessian matrix: ¾ ∂2f ∂x1 ∂x1 ∂2f (x) (x) ∂2f ∂x1 ∂x2 ∂2f ∂x2 ∂x2 . if there is some open ball Bx1 (r) on which Free to photocopy and distribute 125 . ∀(x1 . . . Similarly. x1 ). that is. are equal. ∂2f ∂xn ∂x2 (x) (x) ··· ··· . z) = xy 2 z 3 . we have the following analogue representation of the second derivative. and f : A → R be twice differentiable 2 on A. First. . ∂2f ∂xn ∂x1 (x) (x) ··· (x) 214 Example Let f : R3 → R be given by f (x. that is. . f (x0 ) ≥ f (x).z) f = From the preceding example. as the ∂2 ∂2 mixed partial derivatives f = f . . Just as the Jacobi matrix was a handy tool for ﬁnding a matrix representation of Dx (f ) in the natural bases. when f maps into R. . If Dx0 (f ) is continuous. 215 Theorem Let A ⊆ Rn be an open set. ∂2f ∂xn ∂xn ¿ (x) (x) Hx f = ∂x2 ∂x1 . 213 Theorem Let A ⊆ Rn be an open set.Chapter 2 is well deﬁned. as guaranteed by the following theorem. This is no coinci∂x∂y ∂y∂x dence. it is a bilinear function.y. Then ¾ 0 2yz 3 3y 2 z 2 2yz 3 2xz 3 6xyz 2 3y 2 z 2 6xyz 2 6xy 2 z ¿ H(x. and f : A → R be twice differentiable 2 2 on A. then Dx0 (f ) is symmetric. and f : A → R. we make some analogous deﬁnitions. we say that f (x0 ) is a local maximum of f . etc. The basic theorems resemble those of one-variable calculus. . x2 ) = Dx0 (f )(x2 . we notice that the Hessian is symmetric. and linear in each variable x1 and x2 .. . ∂2f ∂x1 ∂xn ∂2f ∂x2 ∂xn . If there is some open ball Bx0 (r) on which ∀x ∈ Bx0 (r). y. x2 ) ∈ n n R × R we have 2 2 Dx0 (f )(x1 . We are now ready to study extrema in several variables.

then f has a saddle point. and f : A → R be differentiable on A. then Dx0 (f ) = 0. and so to ﬁnd the critical points we solve 2x + y + 2 = 0. A critical point which is neither a maxima nor a minima is called a saddle point. If x0 is an extreme point. − 3 3 − ◭ Free to photocopy and distribute 7 1 4 = f − . 3 3 H(x. we say that f (x1 ) is a local maximum of f . 1 4 which yields x = − . If Hx0 f is indeﬁnite. Solution: ◮ We have (∇f )(x. since ∆1 = 2 > 0 and ∆2 = det 1 4 3 > 0. 218 Example Find the critical points of f : R2 → R (x. If Hx0 f is semi-deﬁnite (positive or negative). y) = x2 + xy + y 2 + 2x + 3y. y) → x2 + xy + y 2 + 2x + 3y and investigate their nature. then f has a local maximum at x0 . if f is twice-differentiable with continuous second derivative and x0 is a critical point such that Hx0 f is negative deﬁnite. Thus x0 = − . x0 is a critical point. that is. y = − .Extrema ∀x ∈ Bx0 (r′ ). the test is inconclusive. A point t is called a critical point if f is differentiable at t and Dt (f ) = 0. which is positive deﬁnite. f (x1 ) ≤ f (x). y) = 2x + y + 2 x + 2y + 3 . 126 .− 3 3 3 2 1 1 2 = is a relative minimum and we have ≤ f (x. 217 Theorem Let A ⊆ Rn be an open set.y) f = 2 1 1 2 . Now. A point is called an extreme point if it is either a local minimum or local maximum. Moreover. If Hx0 f is positive deﬁnite. then f has a local minimum at x0 . x + 2y + 3 = 0.

If a = 0. Solution: ◮ We have ¾ (∇f )(x. 3 127 Free to photocopy and distribute . y.Chapter 2 219 Example Find the extrema of f : R3 (x. which vanishes when x = y = z = 0. Solution: ◮ We have (∇f )(x. 0. ◭ 220 Example Let f (x. Now. y) = 3x2 + ay −3y + ax 2 = 0 0 =⇒ 3x2 = −ay. ∆2 = det ¾ 2 −1 −1 2 = 2 2 3 > 0. z) = x2 + y 2 + 3z 2 − xy + 2xz + yz. since ∆1 = 2 > 0. 0) ≤ f (x. with a ∈ R a parameter. 0) and −1 2 1 2 1 6 ¿ = 4 > 0. y) = x3 − y 3 + axy. Thus f has a relative 0 = f (0. z) → → R x2 + y 2 + 3z 2 − xy + 2xz + yz . y. 0. and ∆3 = det −1 minimum at (0. y. ¾ 2 2 Hr f = −1 −1 2 1 2 6 ¿ 1 . Determine the nature of the critical points of f . which is positive deﬁnite. z) = 2x − y + 2z 6z + 2x + y 2y − x + z ¿ . 0) is a critical point. then x = y = 0 and so (0. If a = 0 then 3 3 y2 a 2 = −ay =⇒ =⇒ =⇒ =⇒ 27y 4 = −a3 y y(27y 3 + a3 ) = 0 y(3y + a)(9y 2 − 3ay + a2 ) = 0 a y = 0 or y = − . 3y 2 = ax.

**Extrema If y = 0 then x = 0, so again (0, 0) is a critical point. If y = − a Now, x= 3 · − a 3
**

2

a 3

then

=

a 3

so

a 3

,−

a 3

is a critical point.

Hf (x,y) =

6x a

a −6y

=⇒ ∆1 = 6x,

∆2 = −36xy − a2 .

At (0, 0), ∆1 = 0, ∆2 = −a2 . If a = 0 then there is a saddle point. a a a a , − , ∆1 = 2a, ∆2 = 3a2 , whence ,− At will be a local 3 3 3 3 minimum if a > 0 and a local maximum if a < 0. ◭

Homework

Problem 2.8.1 Determine the nature of the critical points of f (x, y) = y 2 −2x2 y+4x3 + 20x2 . Problem 2.8.2 Determine the nature of the critical points of f (x, y) = (x − 2)2 + 2y 2 . Problem 2.8.3 Determine the nature of the critical points of f (x, y) = (x − 2)2 − 2y 2 . Problem 2.8.4 Determine the nature of the critical points of f (x, y) = x4 + 4xy − 2y 2 . Problem 2.8.5 Determine the nature of the critical points of f (x, y) = x4 + y 4 − 2x2 + 4xy − 2y 2 . Problem 2.8.6 Determine the nature of the critical points of f (x, y, z) = x2 + y 2 + z 2 − xy + x − 2z. Problem 2.8.7 Determine the nature of the critical points of f (x, y) = x4 + y 4 − 2(x − y)2 . Problem 2.8.8 Determine the nature of the critical points of f (x, y, z) = 4x2 z − 2xy − 4x2 − z 2 + y. Problem 2.8.9 Find the extrema of f (x, y, z) = x2 + y 2 + z 2 + xyz. Problem 2.8.10 Find the extrema f (x, y, z) = x2 y + y 2 z + 2x − z. of

Problem 2.8.11 Determine the nature of the critical points of f (x, y, z) = 4xyz − x4 − y 4 − z 4 . Problem 2.8.12 Determine the nature of the critical points of f (x, y, z) = xyz(4 − x − y − z). Problem 2.8.13 Determine the nature of the critical points of g(x, y, z) = xyze−x

2

−y 2 −z 2

.

x 2 +y

Problem 2.8.14 Let f (x, y) =

y 2 −x

g(t)dt,

where g is a continuously differentiable function deﬁned over all real numbers and g(0) = 0, g ′ (0) = 0. Prove that (0, 0) is a saddle point for f . Problem 2.8.15 Find the minimum of F (x, y) = (x−y)2 + √

Ô 144 − 16x2 − 4 − y2 3

2

,

for −3 ≤ x ≤ 3, −2 ≤ y ≤ 2.

Free to photocopy and distribute

128

Chapter 2

2.9

Lagrange Multipliers

In some situations we wish to optimise a function given a set of constraints. For such cases, we have the following. 221 Theorem Let A ⊆ Rn and let f : A → R, g : A → R be functions whose respective derivatives are continuous. Let g(x0 ) = c0 and let S = g −1 (c0 ) be the level set for g with value c0 , and assume ∇g(x0 ) = 0. If the restriction of f to S has an extreme point at x0 , then ∃λ ∈ R such that ∇f (x0 ) = λ∇g(x0 ).

The above theorem only locates extrema, it does not say anything concerning the nature of the critical points found.

222 Example Optimise f : R2 → R, f (x, y) = x2 − y 2 given that x2 + y 2 = 1. Solution: ◮ Let g(x, y) = x2 + y 2 − 1. We solve ∇f for x, y, λ. This requires 2x −2y = 2xλ 2yλ . x y = λ∇g x y

From 2x = 2xλ we get either x = 0 or λ = 1. If x = 0 then y = ±1 and λ = −1. If λ = 1, then y = 0, x = ±1. Thus the potential critical points are (±1, 0) and (0, ±1). If x2 + y 2 = 1 then f (x, y) = x2 − (1 − x2 ) = 2x2 − 1 ≥ −1, and f (x, y) = 1 − y 2 − y 2 = 1 − 2y 2 ≤ 1. Thus (±1, 0) are maximum points and (0, ±1) are minimum points. ◭ 223 Example Find the maximum and the minimum points of f (x, y) = 4x + 3y, subject to the constraint x2 + 4y 2 = 4, using Lagrange multipliers. Solution: ◮ Putting g(x, y) = x2 + 4y 2 − 4 we have ∇f (x, y) = λ∇g(x, y) =⇒ Free to photocopy and distribute 4 3 =λ 2x 8y . 129

Lagrange Multipliers Thus 4 = 2λx, 3 = 8λy. Clearly then λ = 0. Upon division we ﬁnd x 16 = . Hence y 3 x2 + 4y 2 = 4 =⇒ 256 16 3 y 2 + 4y 2 = 4 =⇒ y = ± √ , x = ± √ 9 73 73.

The maximum is clearly then 16 3 4 √ +3 √ 73 73 √ and the minimum is − 73. ◭ = √ 73,

224 Example Let a > 0, b > 0, c > 0. Determine the maximum and minimum x y z x2 y2 z2 values of f (x, y, z) = + + on the ellipsoid 2 + 2 + 2 = 1. a b c a b c Solution: ◮ We use Lagrange multipliers. Put g(x, y, z) = z

2

x2 a2

+

y2 b2

+

c2

− 1. Then

¾

1/a 1/b 1/c

¿

¾

2x/a2 2z/c2

¿

∇f (x, y, z) = λ∇g(x, y, z) ⇐⇒ a

= λ 2y/b2 .

It follows that λ = 0. Hence x =

b c x2 y2 ,y = ,z = . Since 2 + 2 + 2λ 2λ a b √ 2λ z2 3 3 = 1, we deduce = 1 or λ = ± . Since a, b, c are positive, f c2 4λ2 2 will have a maximum when all x, y, z are positive and a minimum when all x, y, z are negative. Thus the maximum is when a b c x = √ ,y = √ ,z = √ , 3 3 3 and √ 3 f (x, y, z) ≤ √ = 3 3

and the minimum is when a b c x = −√ , y = −√ , z = −√ , 3 3 3 and √ 3 f (x, y, z) ≥ − √ = − 3. 3 130

Free to photocopy and distribute

**Chapter 2 Aliter: Using the CBS Inequality,
**

¬ ¬ ¬x ¬ y z ¬ ·1+ · 1 + · 1¬ ≤ ¬a ¬ b c

x2 a2

+

y2 b2

+

z2 c2

1/2

12 + 12 + 12

1/2

√ √ x y z = (1) 3 =⇒ − 3 ≤ + + ≤ a b c

◭ 225 Example Let a > 0, b > 0, c > 0. Determine the maximum volume of the parallelepiped with sides parallel to the axes that can be enclosed inside the y2 z2 x2 ellipsoid 2 + 2 + 2 = 1. a b c ◮ Let 2x, 2y, 2z, be the dimensions of the box. We must x2 maximise f (x, y, z) = 8xyz subject to the constraint g(x, y, z) = 2 + a y2 z2 + 2 − 1. Using Lagrange multipliers, b2 c Solution:

¾

8yz 8xz 8xy

¿

¾

2x/a2 2z/c

2

¿

∇f (x, y, z) = λ∇g(x, y, z) ⇐⇒

= λ 2y/b2

=⇒ 4yz = λ

x a2

, 4xz = λ

y b2

, 4xy = λ

z c2

.

Multiplying the ﬁrst inequality by x, the second by y, the third by z, and adding, 4xyz = λ Hence x2 a , 4xyz = λ 2 y2 z2 , 4xyz = λ 2 , =⇒ 12xyz = λ b2 c =λ x2 =λ y2 =λ z2 x2 a2 + y2 b2 + z2 c2 = λ.

. 3 a2 b2 c2 If λ = 0, then 8xyz = 0, which minimises the volume. If λ = 0, then a x= √ , 3 and the maximum value is b y= √ , 3 abc 8xyz ≤ 8 √ . 3 3 Aliter: Using the AM-GM Inequality, (x2 y 2 z 2 )1/3 = (abc)2/3 ◭ x

2

λ

c z= √ , 3

a2

·

y

2

b2

·

z

2

1/3

x2 ≤ (abc)2/3 · a

2

+

y2 b2 3

+

z2 8 c2 = 1 =⇒ 8xyz ≤ √ (abc). 3 3 3

a2

Homework

Free to photocopy and distribute 131

Lagrange Multipliers

Problem 2.9.1 A closed box (with six outer faces), has ﬁxed surface area of S square units. Find its maximum volume using Lagrange multipliers. That is, subject to the constraint 2ab + 2bc + 2ca = S, you must maximise abc. Problem 2.9.2 Consider the problem of ﬁnding the closest point P ′ on the plane Π : ax + by + cz = d, a, b, c non-zero constants with a + b + c = d to the point P (1, 1, 1). In this problem, you will do this in three essentially different ways. 1. Do this by a geometric argument, arguing the the point P ′ closest to P on Π is on the perpendicular passing through P and P ′ . 2. Do this by means of Lagrange multipliers, by minimising a suitable function f (x, y, z) subject to the constraint g(x, y, z) = ax + by + cz − d. 3. Do this considering the unconstrained extrema of a suitable funcd − ax − by tion h x, y, . c Problem 2.9.3 Given that x, y are positive real numbers such that x4 +81y 4 = 36 ﬁnd the maximum of x + 3y. Problem 2.9.4 If x, y, z are positive real 1 numbers such that x2 y 3 z = 2 , what is 6 the minimum value of f (x, y, z) = 2x + 3y + z? Problem 2.9.5 Find the maximum and the minimum values of f (x, y) = x2 + y 2 subject to the constraint 5x2 + 6xy + 5y 2 = 8. Problem 2.9.6 Let a > 0, b > 0, p > 1. Maximise f (x, y) = ax + by subject to the constraint xp + y p = 1. Problem 2.9.7 Find the extrema of f (x, y, z) = x2 + y 2 + z 2 subject to the constraint (x − 1)2 + (y − 2)2 + (z − 3)2 = 4. abc3 ≤ 27 Problem 2.9.8 Find the axes of the ellipse 5x2 + 8xy + 5y 2 = 9. Problem 2.9.9 Optimise f (x, y, z) = x + y + z subject to x2 + y 2 = 2, and x + z = 1. Problem 2.9.10 Let x, y be strictly positive real numbers with x + y = 1. What is the √ maximum value of x + xy? Problem 2.9.11 Let a, b be positive real constants. Maximise f (x, y) = xa e−x y b e−y on the triangle in R2 bounded by the lines x ≥ 0, y ≥ 0, x + y ≤ 1. Problem 2.9.12 Determine the extrema of f (x, y) = cos2 x+cos2 y subject to the conπ straint x − y = . 4 Problem 2.9.13 Determine the extrema of f (x, y, z) = x − 2y + 2z subject to the constraint x2 + y 2 + z 2 = 1. Problem 2.9.14 Find the points on the curve determined by the equations x2 + xy + y 2 − z 2 = 1, x2 + y 2 = 1

which are closest to the origin. Problem 2.9.15 Does there exist a polynomial in two variables with real coefﬁcients p(x, y) such that p(x, y) > 0 for all x and y and that for all real numbers c > 0 there exists (x0 , y0 ) ∈ R2 such that p(x0 , y0 ) = c? Problem 2.9.16 Maximise f (x, y, z) = log x + log y + 3 log z on the portion of sphere x2 + y 2 + z 2 = 5r 2 which lies on the ﬁrst octant. Demonstrate using this that for any positive real numbers a, b and c, there follows the inequality a+b+c 5

5

.

Free to photocopy and distribute

132

.3 Integration 3. let {j1 . . . . x2 . . dxj1 ∧ dxj2 ∧ · · · ∧ dxjk (a1 . . the wedge product ∧ of differential forms has the following properties anti-commutativity: da ∧ db = −db ∧ da. j2 . . . . . . ak ) = det ··· ··· ··· ··· aj1 k ¿ aj2 k . . By virtue of being a determinant. 2. . . 226 Deﬁnition Consider n variables x1 . . associativity: (da ∧ db) ∧ dc = da ∧ (db ∧ dc). . . . 133 . . ajk k In other words. . ajk 2 dxj1 ∧ dxj2 ∧ · · · ∧ dxjk (a1 . . we need to consider the concept of differential forms. a2 . . n} be a collection of k sub-indices. ajk 1 aj1 2 aj2 2 . . . a2 .1 1 scalar homogeneity: if λ ∈ R. and let ¾ a1j a2j . . . . . An elementary k-differential form (k > 1) acting on the vectors aj . xjk component of the signed k-volume of a k-parallelotope in Rn spanned by a1 . ak ) is the xj1 xj2 . be k ≤ n vectors in Rn . anj ¿ aj = ∈ Rn . In order to smooth our discussion. 1 ≤ j ≤ k is deﬁned and denoted by ¾ aj1 1 aj2 1 . . x n in n-dimensional space (used as the names of the axes). . . 1 ≤ j ≤ k. linearity: d(a + b) = da + db. ak . . jk } ⊆ {1. then dλa = λda. Moreover. a2 . . . .1 Differential Forms We will now consider integration in several variables. Notice that associativity does not hold for the wedge product of vectors.

where the aj1 j2 .. In order to avoid redundancy we will make the convention that if a sum of two or more terms have the same differential form up to permutation of the variables.. are the (signed) 1-volumes (that is. 229 Example In R3 we have dx ∧ dz + 5dz ∧ dx + 4dx ∧ dy − dy ∧ dx + 12dx ∧ dx = −4dx ∧ dz + 5dx ∧ dy. w) = x + y 2 + z 3 + w4 is a 0-form in R4 .. dx(a) = det(1) = 1. dy(a) = det(0) = 0. y. since we have a repeated variable.jk dxj1 ∧ dxj2 ∧ · · · dxjk .Differential Forms Anti-commutativity yields da ∧ da = 0. the length) of the projections of a onto the coordinate axes. z. Free to photocopy and distribute 134 . dz(a) = det(−1) = −1. 233 Example An example of a 1-form ﬁeld in R3 is ω = xdx + y 2 dy + xyz 3 dz.jk are differentiable functions in Rn . 232 Example g(x. we will simplify the summands and express the other differential forms in terms of the one differential form whose indices appear in increasing order.. 230 Deﬁnition A 0-differential form in Rn is simply a differentiable function in Rn . 231 Deﬁnition A k-differential form ﬁeld in Rn is an expression of the form ω= 1≤j1 ≤j2 ≤···≤jk ≤n aj1 j2 . 228 Example In R3 we have dx ∧ dy ∧ dx = 0. 227 Example Consider a= Then ¾ 1 0 −1 ¿ ∈ R3 .

xn ) be a 0-form in Rn . is ω2 = −2xdx + 2ydy. x2 . x2 . 237 Deﬁnition Let f (x1 . . Then dx = du + dv. 236 Example The product of the 1-form ﬁelds in R3 ω1 = ydx + xdy. xn ) ∧ dxj1 ∧ dxj2 ∧ · · · ∧ dxjk . . then d(x3 y 4 ) = 3x2 y 4 dx + 4x3 y 3 dy. We shew now how to multiply differential forms. The exterior derivative df of f is n df = i=1 ∂f ∂xi dxi . . . . Free to photocopy and distribute 135 . ω = x3 y 4 . 238 Example If in R2 . x2 . the exterior derivative dω of ω is the (k + 1)-form dω = df (x1 .Chapter 3 234 Example An example of a 2-form ﬁeld in R3 is ω = x2 dx ∧ dy + y 2 dy ∧ dz + dz ∧ dx. ω1 ∧ ω2 = (2x2 + 2y 2 )dx ∧ dy. . y = uv. if is a k-form in Rn . . . . dy = vdu + udv. . whence dx ∧ dy = (u − v)du ∧ dv. . 235 Example An example of a 3-form ﬁeld in R3 is ω = (x + y + z)dx ∧ dy ∧ dz. xn )dxj1 ∧ dxj2 ∧ · · · ∧ dxjk 240 Example Consider the change of variables x = u + v. ω = x2 ydx + x3 y 4 dy then dω = = = = d(x2 ydx + x3 y 4 dy) (2xydx + x2 dy) ∧ dx + (3x2 y 4 dx + 4x3 y 3 dy) ∧ dy (3x2 y 4 − x2 )dx ∧ dy x2 dy ∧ dx + 3x2 y 4 dx ∧ dy ω = f (x1 . Furthermore. . 239 Example If in R2 .

−x has opposite orientation to +x and −x ω=− ω. 0). dv = − dw = − y+z (x + y + z)2 z (y + z)2 dy + x (x + y + z)2 y (y + z)2 dz. 0) is given negative orientation. 2. 0)}2 be an oriented 0-manifold.Zero-Manifolds 241 Example Consider the transformation of coordinates xyz into uvw coordinates given by z y+z u = x + y + z. 0. 0. (y + z)2 (x + y + z)2 − dx ∧ dy ∧ dz = 3. Free to photocopy and distribute 136 . w= . −{(1. 0. +x 244 Example Let M = −{(1. say. x (x + y + z)2 dz. 0)} ∪ +{(1.2 Zero-Manifolds 242 Deﬁnition A 0-dimensional oriented manifold of Rn is simply a point x ∈ Rn . −2. y+z x+y+z Then du = dx + dy + dz. The ﬁrst one means that the point (1. M 2 Do not confuse. A general oriented 0-manifold is a union of oriented points. the second means that (−1. and let ω be a 0-form. 0) = (−1. 3) = −(1) + (14) − (−4) = 17. 0. 0. Then M ω = ω(b) − ω(a). and let ω = x + 2y + z 2 . 243 Deﬁnition Let M = +{b} ∪ −{a} be an oriented 0-manifold. 2. 0) is the additive inverse of (1. v = . 0. 3)} ∪ −{(0. 0)) + ω(1. 3) − ω(0. 0. dx + dy + Multiplication gives du ∧ dv ∧ dw = zx y(y + z) − (y + z)2 (x + y + z)2 (y + z)2 (x + y + z)2 z(y + z) xy + − 2 (x + y + z)2 2 (x + y + z)2 (y + z) (y + z) z 2 − y 2 − zx − xy dx ∧ dy ∧ dz. 0. 0. Then ω = −ω((1. 0). 0)} with −(1. with a choice of the + or − sign.

with a choice of a + orientation if the curve traverses in the direction of increasing t.√These conditions are met with the parametrisation √ x = t − 1. 1) to (2.3 One-Manifolds 245 Deﬁnition A 1-dimensional oriented manifold of Rn is simply an oriented smooth curve Γ ∈ Rn . 246 Example Calculate xydx + (x + y)dy Γ where Γ is the parabola y = x2 . Solution: ◮ We parametrise the curve as x = t. t ∈ [0. The curve −Γ has opposite orientation to Γ and −Γ ω=− ω. 4 t3 + t4 −1 What would happen if we had given the curve above a different parametrisation? First observe that the curve travels from (−1. r Γ We now turn to the problem of integrating 1-forms.Chapter 3 3. y = ( t − 1)2 . 2 t 137 Free to photocopy and distribute . y = t2 . or with a choice of a − sign if the curve traverses in the direction of decreasing t. Then xydx + (x + y)dy = t3 dt + (t + t2 )dt2 = (3t3 + 2t2 )dt. whence ω Γ 2 = å −1 (3t3 + 2t2 )dt 3 4 è2 = = 2 3 69 . the classical way of writing this is dy → − − f • d→ . 2] oriented in the positive direction. 9]. Then xydx + (x + y)dy = = = √ √ √ √ √ ( t − 1)3 d( t − 1) + (( t − 1) + ( t − 1)2 )d( t − 1)2 √ √ √ (3( t − 1)3 + 2( t − 1)2 )d( t − 1) √ √ 1 √ (3( t − 1)3 + 2( t − 1)2 )dt. A general oriented 1-manifold is a union of oriented curves. x ∈ [−1. Γ æ é dx → − − If f : R2 → R2 and if d→ = r . 4) on the parabola y = x2 .

0) to (1. ◭ 248 Example Calculate the path integral x+y Γ x2 + y 2 dy + x2 + y 2 x−y dx around the closed square Γ = ABCD with A = (1. B = (−1. (sin t − cos t)dt To solve this problem using Maple you may use the code below. and D = (1. Line(<0.1>)). To solve this problem using Maple you may use the code below.One-Manifolds whence ω Γ √ √ 1 √ (3( t − 1)3 + 2( t − 1)2 )dt 2 æ0 2 t 3/2 é9 √ 3t 7t 5t = − + − t 4 3 2 0 69 = .0>.x+y> ). Γ where Γ is the line segment from (0. so we choose the parametrisation x = y = t. we only need to worry about ﬁnding a suitable parametrisation. 4 9 = as before. The integral is thus 1 y sin xdx + x cos ydy Γ = 0 (t sin t + t cos t)dt [t(sin x − cos t)]1 − 0 2 sin 1 − 1. ◭ It turns out that if two different parametrisations of the same curve have the same orientation. 1). C = (−1. t=-1.<1.tˆ2>. then their integrals are equal. 1) in the positive direction.x*cos(y)> ). > with(Student[VectorCalculus]): > LineInt( VectorField( <y*sin(x). 247 Example Calculate the line integral y sin xdx + x cos ydy. > with(Student[VectorCalculus]): > LineInt( VectorField( <x*y. Free to photocopy and distribute 138 . −1).2)). Solution: ◮ This line has equation y = x. 1). Hence. −1) in the direction ABCDA. Path( <t.. 1 0 = = upon integrating by parts.

Ellipse(9*xˆ2 + 4*yˆ2 -36)). The positive direction of integration is that sense that when traversing the path. x = −1. ◭ When the integral is along a Á closed path.-1>. To solve this problem using Maple you may use the code below. dy = 0. Solution: ◮ Parametrise the ellipse as x = 2 cos t. ◭ Free to photocopy and distribute 139 . dy = 0. Observe that when traversing this closed curve.Chapter 3 Solution: ◮ On AB. We have Á 2π 2 2 ω Γ = 0 2π ((4 cos2 t)(3 cos t) + (9 sin t)(−2 sin t))dt (12 cos3 t − 18 sin3 t)dt = 0 = 0.xˆ2> ). The integral is thus ω Γ = = AB −1 1 ω+ x2 1 −1 ω+ BC ω+ CD −1 1 ω DA 1 −1 x−1 = 4 +1 1 dx + dx y2 y−1 +1 dy + x+1 x2 +1 1 dx + −1 y+1 y2 + 1 dy x2 + 1 = 4 arctan x|1 −1 = 2π. t ∈ [0. > with(Student[VectorCalculus]): > LineInt( VectorField( <yˆ2. 249 Example Calculate the path integral Á x2 dy + y 2 dx.1>. 2π]. Γ where Γ is the ellipse 9x + 4y = 36 traversed once in the positive sense.<-1. y = 1. rather than Γ Γ it is customary to use the symbol . dx = 0. the area of the ellipse is on the left hand side of the path. dx = 0. x = 1. on CD. > LineSegments(<1. on BC.-1>.(x-y)/(xˆ2+yˆ2)> ).1>. like in the preceding example.<-1. so this parametrisation traverses the curve in the positive sense.<1. > with(Student[VectorCalculus]): > LineInt( VectorField( <(x+y)/(xˆ2+yˆ2). y = −1.<1. y = 3 sin t. the area enclosed by the curve is to the left of the curve.1>)). and on DA. To solve this problem using Maple you may use the code below.

3 3 and on LBC √ 2 x||dx|| = x (dx)2 + (dy)2 = x( 1 + (−4) )dx = x 17dx. 2).-1>. and LBC : y = −4x + 6. [x.<1.2> ) ). > with(Student[VectorCalculus]): > PathInt( x. On LAB 3 √ x 10dx 1 2 dx = x||dx|| = x (dx)2 + (dy)2 = x 1 + − . 251 Example Find Γ x||dx|| where Γ is the triangle starting at A : (−1. Solution: ◮ The lines passing through the given points have equations −x − 4 LAB : y = . The integral f (x)||dx|| Γ is called the path integral of f along Γ. −1) to B : (2. and ending in C : (1. 2 2 To solve this problem using Maple you may use the code below.-2>. −2). ◭ Homework Free to photocopy and distribute 140 . Hence x||dx|| Γ = LAB 2 x||dx|| + x||dx|| = = LBC √ 1 √ x 10dx + x 17dx 3√ −1 2 √ 10 3 17 − .y]=LineSegments( <-1.One-Manifolds 250 Deﬁnition Let Γ be a smooth curve. <2.

2 Find Γ xdx + ydy where Γ is the path shewn in ﬁgure 3. by parametrising this path. 0).2. xn ) then dω = k=1 n ∂f ∂xk dxk 141 Free to photocopy and distribute . xy||dx|| using this parametriProblem 3.3. Calculate also parametrisation. traversed in the positive direction. Evaluate C C A O xy||dx|| using this Figure 3.3.3. Calculate also sation. 0) going on a straight line to ¡ A 4 cos π . Evaluate C xdx + ydy where C is the straight line path that starts at (−1. 0). x2 . .3 Find Γ the semicircle that starts at (−1.1: Example 251. For p = 0 if ω = f (x1 . . Proof: We will prove this by induction on p.4 Closed and Exact Forms ´ 252 Lemma (Poincare Lemma) If ω is a p-differential form of continuously differentiable functions in Rn then d(dω) = 0. xdx + ydy where C is Problem 3.3. 4 sin π and continuing on an 6 6 ¡ arc of a circle to B 4 cos π . start- where Γ is the intersection of the sphere x2 + y 2 + z 2 = 1 and the plane x + y = 1.3. by parametrising this path. 1) and ends at (1.3 2 Chapter 3 1 -3 -2 -1 -1 1 2 3 -2 -3 Figure 3. 2.2 and 3.2. 1) and ends at (1. 5 5 B 1. 3. C x||dx|| where Γ is the path shewn in ﬁgure 3. Problem 3. 0) goes to (0.1 Consider C xdx + ydy and xy||dx||. 0) goes to (0. C ing at O(0. .2: Problems 3.3. .3.4 Find Á zdx + xdy + ydz Γ Problem 3. 4 sin π .

where the aj1 j2 . u 253 Deﬁnition A differential form ω is said to be exact if there is a continuously differentiable function F such that dF = ω. +da ∧ d dxj1 ∧ dxj2 ∧ · · · dxjp since dda = 0 from the case p = 0. since ω is continuously differentiable and so the mixed partial derivatives are equal.jp dxj1 ∧ dxj2 ∧ · · · dxjp .. p > 0.jp ∂xi dxi ∧ dxj1 ∧ dxj2 ∧ · · · dxjp .....Closed and Exact Forms and n d(dω) = k=1 n d ∂f ∂xk n j=1 ∧ dxk ∂2f ∧ dxj − ∧ dxk dxj ∧ dxk = k=1 n ∂xj ∂xk ∂2f ∂xj ∂xk = 1≤j≤k≤n ∂2f ∂xk ∂xj = 0. Consider now an arbitrary p-form.. Since such a form can be written as ω= 1≤j1 ≤j2 ≤···≤jp ≤n aj1 j2 ..jp are continuous differentiable functions in Rn . But an independent induction argument proves that d dxj1 ∧ dxj2 ∧ · · · dxjp = 0. completing the proof.jp ∧ dxj1 ∧ dxj2 ∧ · · · dxjp n i=1 = 1≤j1 ≤j2 ≤···≤jp ≤n ∂aj1 j2 . it is enough to prove that for each summand d da ∧ dxj1 ∧ dxj2 ∧ · · · dxjp = 0. we have dω = 1≤j1 ≤j2 ≤···≤jp ≤n daj1 j2 . Free to photocopy and distribute 142 . But d da ∧ dxj1 ∧ dxj2 ∧ · · · dxjp = = dda ∧ dxj1 ∧ dxj2 ∧ · · · dxjp da ∧ d dxj1 ∧ dxj2 ∧ · · · dxjp ..

∂F ∂y Free to photocopy and distribute 143 . if dω = 0 on a star-shaped domain then ω is exact. dx + y x2 + y 2 dy 1 2 (x2 + y 2 ) . Let ω = dF be an exact form. A result of Poincare says that for certain domains (called star´ shaped domains) the converse is also true. ´ dω = ddF = 0. Solution: ◮ Assume there is a function F such that dF = ω.Chapter 3 254 Example The differential form xdx + ydy is exact. 257 Example Determine whether the differential form ω= is exact. since ydx + xdy = d (xy) . since x x2 + y dx + 2 y x2 + y dy = d 2 1 2 loge (x2 + y 2 ) . = . since xdx + ydy = d 255 Example The differential form ydx + xdy is exact. 256 Example The differential form x x2 + y 2 is exact. that is. By the Poincare Lemma Theorem 252. 2x(1 − ey ) ey dx + dy (1 + x2 )2 1 + x2 2x(1 − ey ) (1 + x2 )2 ey 1 + x2 . By the Chain Rule dF = This demands that ∂F ∂x ∂F ∂x = dx + ∂F ∂y dy.

. We then take F (x. y) = ◭ 258 Example Is there a continuously differentiable function such that dF = ω = y 2 z 3 dx + 2xyz 3 dy + 3xy 2 z 2 dz ? Solution: ◮ We have dω = (2yz 3 dy + 3y 2 z 2 dz) ∧ dx 1 + x2 + c. So put dF = Then ∂F ∂x dx + ∂F ∂y dy + ∂F ∂z dz = y 2 z 3 dx + 2xyz 3 dy + 3xy 2 z 2 dz. ∂x (1 + x2 )2 Comparing this with our ﬁrst expression for φ′ (x) = that is φ(x) = − 2x (1 + x2 )2 ∂F ∂x . y) = 1 + x2 where φ(x) is a function depending only on x.Closed and Exact Forms We have a choice here of integrating either the ﬁrst. we ﬁnd ey + φ(x). we differentiate the obtained expression for F with respect to x and ﬁnd ∂F 2xey =− + φ′ (x). +(3y 2 z 2 dx + 6xyz 2 dy + 6xy 2 zdz) ∧ dz so this form is exact in a star-shaped domain. z). z). Since integrating the second expression (with respect to y) is easier. or the second expression. +(2yz 3 dx + 2xz 3 dy + 6xyz 2 dz) ∧ dy = 0. To ﬁnd it. we ﬁnd 1 + c. 1 + x2 ey − 1 where c is a constant. ∂F ∂x ∂F ∂y = y 2 z 3 =⇒ F = xy 2 z 3 + a(y. F (x. 144 Free to photocopy and distribute . = 2xyz 3 =⇒ F = xy 2 z 3 + b(x.

Chapter 3 ∂F ∂z = 3xy 2 z 2 =⇒ F = xy 2 z 3 + c(x. Γ 260 Example Evaluate the integral Á Γ 2y 2x dx + 2 dy x2 + y 2 x + y2 where Γ is the closed polygon with vertices at A = (0. 1). E = (1. ◭ 261 Example Calculate the path integral Á Γ (x2 − y)dx + (y 2 − x)dy. the integral is 0. then ω = 0. Solution: ◮ Observe that d 2x x2 + y 2 dx + 2y x2 + y 2 dy =− 4xy (x2 + y 2 )2 dy∧dx− 4xy (x2 + y 2 )2 dx∧dy = 0. 259 Theorem Let U ⊆ Rn be an open set. 0). where Γ is a loop of x3 + y 3 − 2xy = 0 traversed once in the positive sense. Comparing these three expressions for F . the form is exact. and so the form is exact in a start-shaped domain. y). 0). Assume ω = dF is an exact form. we obtain F (x. C = (7. D = (3. ◭ Free to photocopy and distribute 145 . traversed in the order ABCDEA. y. Then B ω= Γ A dF = F (B) − F (A). 2). 2). ◭ We have the following equivalent of the Fundamental Theorem of Calculus. and Γ a path in U with starting point A and endpoint B. Solution: ◮ Since ∂ ∂y (x2 − y) = −1 = ∂ ∂x (y 2 − x). the integral is 0. By virtue of Theorem 259. B = (5. and since this is a closed simple path. Á In particular. if Γ is a simple closed path. z) = xy 2 z 3 .

Two-Manifolds 3. In order to evaluate double integrals. b] × [c. D −D We will often write f (x.5 Two-Manifolds 262 Deﬁnition A 2-dimensional oriented manifold of R2 is simply an open set (region) D ∈ R2 . where the + orientation is counter-clockwise and the − orientation is clockwise. unless otherwise noticed. y)dA D where dA denotes the area element. d]. and let f : A → R be continuous. This corresponds to using the area form dxdy. Free to photocopy and distribute 146 . Then b d d b f dA = D a c f (x. we need the following. the integral f dA D is the sum of all the values of f restricted to D. A general oriented 2-manifold is a union of open sets. In particular. y)dy dx = c a f (x. Given a function f : D → R. In this section. The region −D has opposite orientation to D and ω=− ω. y)dx dy Fubini’s Theorem allows us to convert the double integral into iterated (single) integrals. we will choose the positive orientation for the regions considered. 263 Theorem (Fubini’s Theorem) Let D = [a. Let D ⊆ R2 . dA D is the area of D.

y]=Region(0. 5 4 Also.4.0. Free to photocopy and distribute 147 . 265 Example We have 4 3 1 4 1 0 (x + 2y)(2x + y) dxdy 0 = 3 4 (2x2 + 5xy + 2y 2 ) dxdy dy = = 409 12 3 2 5 + y + 2y 2 3 2 .3] xdx 1 2 0 ydy 2 = = 5 4 5 2 To solve this problem using Maple you may use the code below. this integral is “factorable into x and y pieces” meaning that 1 3 xydA = [0..1. To solve this problem using Maple you may use the code below. 4 Notice that if we had integrated ﬁrst with respect to x we would have obtained the same result: = 3 2 1 3 xydx dy 0 = 2 3 x2 y 2 y 2 3 2 1 dy 0 = 2 dx = = y2 4 . [x.1]×[2.y]=Region(3.3)).3] 0 1 2 xydy dx xy 2 2 9x 2 1 3 = 0 1 dx 2 = 0 2 − 2x dx 5x 4 0 5 = .1]×[2...Chapter 3 264 Example 1 3 xydA = [0..1)). > with(Student[VectorCalculus]): > int(x*y. > with(Student[VectorCalculus]): > int((x + 2*y)*(2*x + y). [x.2.

The lower point of the dashed line is (1. we must divide the region as in ﬁgure 3. The right point of the dashed line is (7/4. If we integrate ﬁrst with respect to x. 2 LBC : y = 2x − 12. B : (2. The integral is thus 1 (3x+1)/2 2 x −1 (−x−4)/3 y dy dx + 1 x −4x+6 (−x−4)/3 y dy dx = − 11 8 .-2>. P Solution: ◮ The lines joining the points have equations LAB : y = x .-1>. we must divide the region as in ﬁgure 3. 5). > with(Student[VectorCalculus]): > int(x*y. −1). ◭ 267 Example Consider the region inside the parallelogram P with vertices at A : (6. D : (7. 4). Solution: ◮ The lines passing through the given points have equations −x − 4 3x + 1 LAB : y = . If we integrate ﬁrst with respect to y. C : (9. The integral is thus −1 −2 (6−y)/4 2 (6−y)/4 y −4−3y x dx dy + −1 y (2y−1)/3 x dx dy = − 11 8 . To solve this problem using Maple you may use the code below.<1. 2). 3). we 3 2 draw the region carefully. one variable is kept constant. as in ﬁgure 3. −2). Now.Two-Manifolds In the cases when the domain of integration is not a rectangle. LBC : y = −4x + 6. −5/3). because there are two upper lines which the upper value of y might be.3. LCA : y = . Free to photocopy and distribute 148 . because there are two left-most lines which the left value of x might be. Find xy dxdy. [x. −1). we decompose so that. C : (1.2>). B : (8. 266 Example Find D xy dxdy in the triangle with vertices A : (−1.y]=Triangle(<-1. 6).5.4.<2.

Solution: ◮ The integral is 0. > with(Student[VectorCalculus]): > int(x*y.y]=Triangle(<8.5>)) > + int(x*y. y).y]=Triangle(<6. y) ∈ R2 |x ≥ 0.3>.<8. LCD : y = x 2 + 3 2 .4>.<9.<7.6>. Observe that if (x. Figure 3. −y) = −f (x.5>)). [x. Figure 3.3: Example 266. y) ∈ D then (x. f (x. x2 + y 2 ≤ 1}. −y) ∈ D. Notice that we have split the parallelogram into two triangles. ◭ 268 Example Find y D x2 + 1 dxdy where D = {(x.<7. To solve this problem using Maple you may use the code below. [x.4>.7 3 3 6 5 4 1 1 3 2 -3 -2 -1 -1 1 2 3 -3 -2 -1 -1 1 2 3 4 5 6 7 8 9 10 1 2 3 1 2 2 Chapter 3 -2 -2 -3 -3 Figure 3.4: Example 266. The integral is thus 4 3 2y 5 (y+12)/2 6 (y+12)/2 xy dxdy+ (y+9)/2 4 (y+9)/2 xy dxdy+ 5 2y−3 xy dxdy = 409 4 . Integration order dydx.5: Example 267. ◭ Free to photocopy and distribute 149 . LDA : y = 2x − 9. Also. Integration order dxdy.

y) ∈ R2 : 2e2 − (2e2 − e2 + 1) e2 − 1 Ô √ √ √ x + y ≥ 1. 0 0 1 2 Figure 3. We then have 4 0 √ y y 2 ≤x≤ √ y =⇒ 0 ≤ x ≤ 2.6: Example 269. Solution: ◮ We have 0 ≤ y ≤ 4. 269 Example Find 4 0 √ y y/2 ey/x dx dy.8: Example 271. Solution: ◮ The area is given by 1 1−(1− (1− √ 1−x)2 dA = D 0 1 0 √ 2 x) dy dx = = ◭ Free to photocopy and distribute 2 2 3 .5 4 3 2 2 1 0 1 0 1 2 3 4 5 1 Two-Manifolds 0 0 Figure 3. 1 Figure 3. ey/x dx dy 2 2x x2 = 0 2 ey/x dy dx ¬ y/2 = 0 2 xey/x ¬2x dx x2 (xe2 − xex ) dx = 0 = = ◭ 270 Example Find the area of the region R = {(x. x2 ≤ y ≤ 2x. 1 − x + 1 − y ≥ 1}.7: Example 270. √ √ ( 1 − x + x − 1)dx 150 .

x2 + y 2 dA. x ≥ y 2 }. y 2 )dxdy.5.y≥0 2≤x2 +y2 <3. we have 0 ≤ x2 +y 2 ≤ ( 2)2 +( 2)2 = 4. and (2. and so.5. Problem 3. where R is the rectangle [0. 3}. gral 1 0 x Problem 3.x≥0.2 Let S be the interior and boundary of the triangle with vertices (0.1 Evaluate the iterated inte3 x 1 dydx. Now the integrals can be computed by realising that they are areas of quarter annuli.y≥0 Hence x2 + y 2 dA = R π 4 (1 + 2 + 3) = 3π 2 .x≥0.6 Find 0 0 min(x2 . x2 + y 2 dA = R 1≤k≤3 Rk k ∈ {1.y≥0 = 2dA + 3≤x2 +y2 <4.5. π]2 . For (x.5. where D = [0. Thus we decompose R as the union of the Rk = {(x. 0). Free to photocopy and distribute 151 . y) ∈ R2 : x ≥ 0. 4 4 k≤x2 +y2 <k+1.5. y) ∈ R2 : x ≥ 0. 0). Find S x2 dA. 1 ≤ x2 +y 2 ≤ 4}.4 Find xydxdy D where D = {(x. 1 1 Problem 3. y) ∈ R2 |y ≥ x2 .x≥0.5.Chapter 3 271 Example Evaluate . y ≥ 0. 2. kdA = k · 1 πk · π(k + 1 − k) = . ◭ Homework Problem 3. x2 + y 2 dA 1dA + 1≤x2 +y2 <2.5 Find (x + y)(sin x)(sin y)dA D ydA.y≥0 3dA.3 Let S = {(x. (2. 1). y) ∈ R. k ≤ x2 +y 2 < k+1}.x≥0. R √ 2] × [0. y ≥ 0. Problem 3. y) → x2 + y 2 jumps √ every time x2 + y 2 √ is an integer. √ 2] Solution: ◮ The function (x. Find S Problem 3.

x + y ≤ 1}. 9 < x2 +y 2 < 16. y ≥ 0. B Problem 3.8 Evaluate R xdA where R where D = {(x. y) ∈ R2 : y ≥ − √ 3 x+4 . + y)dA log (1 + x D Problem 3.7 Find D xydxdy where D = {(x.8.2]2 is the (unoriented) circular segment in ﬁgure 3. Problem 3.9.5.5.5.10: Problem 3.11: Problem 3. is the (unoriented) △OAB in ﬁgure 3. 2].5. which is created by the intersection of regions {(x.12 Find Problem 3. y) ∈ R2 |x ≥ 0. 1 < x2 −y 2 e< 16}.1]2 min(x.11. A(3.5.9 Find 0 2ex dxdy 2 Problem 3. y 2 )dA.14 Evaluate R x + y dA. is the quarter annulus in ﬁgure 3.10 Evaluate [0. 1 1 y Problem 3.15 Evaluate R xdA where R Figure 3. 4). Problem 3.11.11 Find R xydA.5.5.5.10 with O(0. and B(4. 1] × [0 .13 Evaluate [0. y) ∈ R2 : x > 0. y) ∈ R2 : x2 + y 2 ≤ 16} and (x.12. y > 0.Two-Manifolds Problem 3. A Problem 3. where R is the rectangle [0 .5.5. which formed by the the area between the circles x2 + y 2 = 1 and x2 + y 2 = 4 in the ﬁrst quadrant.15.9: Problem 3.16 Evaluate R xdA where R is the E-shaped ﬁgure in ﬁgure 3. B Problem 3. 3 x + y 2 dA.5.5. 0). A O Figure 3. 1). where R Figure 3. Free to photocopy and distribute 152 .5.

5.5. xy+y+x ≤ A(−1. and C(1. B(2. y ≥ 0. as shewn in ﬁgure 3.5. |y| ≤ 1}. −1). y Problem 3.5.5. 1}.Chapter 3 Figure 3.13: Problem 3.26 Find D (2x + 3y + 1) dA.5. D where Problem 3. y where loge (1 + x2 + y)dA D Problem 3.12: Problem 3. Problem 3.5. where R is the region between the circles x2 +y 2 = 4 and x2 +y 2 = 2y. πx sin dy dx. y) ∈ R2 : x4 + y 4 + x2 − y 2 ≤ 1}.22 Find π /2 π /2 Problem 3. x2 +y ≤ 1}. √ 2y x 2 Problem 3.5.19 Find 2x(x2 + y 2 )dA D R xdA. where where D is the triangle with vertices at D = {(x.5.21 Find xydA D D = {(x. y) ∈ R2 : x ≥ 0. y ≥ 0. Free to photocopy and distribute 153 .17 Evaluate 0 x cos y dydx.24 Evaluate 0 ex/y dydx.5.5.25 Find |x − y|dA Figure 3. where D = {(x. 1 1 √ x Problem 3.5.13. −4).5.14. y) ∈ R2 : x ≥ 0.20 Find the area bounded by y2 x2 the ellipses x2 + = 1 and + y 2 = 1.14: Problem 3. 3). Problem 3.20. Problem 3. y) ∈ R2 : |x| ≤ 1. Figure 3.23.23 Evaluate Problem 3. 4 4 as in ﬁgure 3.18 Find 2 1 πx sin dy √ 2y x x 4 dx+ 2 D = {(x.16.

Free to photocopy and distribute 154 .32 Find 2 xy dA.5.36 Let I be the rectangle [1.5. g : [0. √ D Problem 3. 1] →]0. . Problem 3.5. where 2 D = {(x. It is known that at least one side of each of the rectangles Rk is an integer.29 Let f. Problem 3.5. with f increasing.38 Evaluate √ 4 4−y Ô 12x − x3 dxdy.5. y2 16 Problem 3. x−1 dx log x = Then demonstrate that log 2. 0 0 2 2 where S = {(x.5. 1] → [0.33 A rectangle R on the plane is the disjoint union R = ∪N Rk of rectank=1 gles Rk . f (x)dx Problem 3.27 Let f : [0.42 Prove that +∞ 0 +∞ 2x xe−y sin y 1 dydx = . y) ∈ R : |x| 2 1/2 + |y| 1/2 ≤ 1}.39 Evaluate 0 y y Ô 1 + x3 dxdy.28 Find (xy(x + y))dA D (g(y) − f (x)) dxdy ≥ 0. .Two-Manifolds Problem 3. Problem 3. 2] × [1. +∞] be a decreasing function. x + y ≤ 1}.35 Evaluate 1 0 0 1 1 xf 2 (x)dx 0 1 f 2 (x)dx ≤ 0 1 ··· 0 (x1 + x2 + · · · + xn ) dx1 dx2 .5.5. y) ∈ R2 |x ≥ 1.43 Prove that 1 y y2 ··· 0 (x1 x2 · · · xn )dx1 dx2 .5.5. y ≥ 1. (x + y) ≤ 2x}. Shew that at least one side of R is an integer. dxn . Problem 3. . Problem 3. dxn .5.34 Evaluate 1 0 0 1 1 where D = {(x. 1 + x4 where a and b are positive.37 Find 0 1 0 xy dxdy. g : [1.31 Evaluate a 0 0 b 1 2 1 y emax(b 2 x . y ≥ 0.40 Evaluate dydx. g be continuous functions f.41 Find 1 dA (x + y)4 D 0 √ xy dxdy. 1] be continuous.5.5. 1 1 where D = {(x.a y ) 2 2 Problem 3. 2] such that f (x) ≤ g(x). Prove that 1 1 Problem 3. Demonstrate that I xf (x)dx 0 0 Problem 3.30 Compute S (xy + y 2 )dA Problem 3. 0 (f ◦g)(x)dx ≤ f (x)dx+ 0 0 g(x)dx.5.5. y) ∈ R : y ≥ 0. . Problem 3. 2] and let f.5. x + y ≤ 4}. 0 x Ô y x2 + y 2 dxdy = log(1 + √ 2). y) ∈ R2 |x ≥ 0. Problem 3.5. . Problem 3. Prove that 1 0 1 1 Problem 3. 2] → [1.

xn )dx1 ∧ dx2 ∧ . dxn . b]. . Problem 3.6 Change of Variables We now perform a multidimensional analogue of the change of variables theorem in one variable. dx2 dx equals 1 n! a n f (x)dx 0 . bounded sets in Rn with volume and let g : ∆ → D be a continuously differentiable bijective mapping such that 1 det g ′ (u) = 0. . .46 Evaluate 1 n→+∞ 1 0 1 lim 0 ··· cos2 0 Problem 3. are bounded on ∆. Deduce Cauchy’s Inequality for integrals. 2n a x x x 1 n−2 n−1 Problem 3. Let f : [0. a] → R be continuous. . . . . n > 0. f (x1 . x+2y−4 ≤ 0}.44 Prove that 1 1 equals 1 1 b a b 2 x−y dxdy. D = {(x. Prove that æ é2 b b f (x) g(x) 1 det dx dy 2 a f (y) g(y) a 0 0 ··· 0 0 (f (x1 )f (x2 ) · · · f (xn ))dxn dxn−1 .45 Find xdA D x−y 1 dydx = =− (x + y)3 2 (g(x))2 dx − (f (x)g(x))dx a . u2 .5. ∧ dun . x−y+1 ≥ 0. a > 0.5. un ))| det g ′ (u)|du1 ∧ du2 ∧ . . .5.Chapter 3 Problem 3.48 Let a ∈ R.5. n ∈ N. . . This is an integral analogue of Lagrange’s Identity. y) ∈ R |y ≥ 0.47 Let f. ∆) ∈ (Rn )2 be open. 272 Theorem Let (D. . π Prove that (x1 + x2 + · · · + xn ) dx1 dx2 . x2 . .5. ∧ dxn = ··· ∆ f (g(u1 . f ◦ g| det g ′ (u)| is integrable on ∆ and ··· that is ··· D f = D ··· ∆ (f ◦ g)| det g ′ (u)|. . b 2 b b where 2 a (f (x)g(x))dx ≤ (f (x))2 dx a a (g(x))2 dx . . 155 Free to photocopy and distribute . b (x + y)3 0 0 0 0 (f (x))2 dx a Is this a contradiction to Fubini’s Theorem? Problem 3. For f : D → R | det g ′ (u)| bounded and integrable. . g be continuous functions in the interval [a. and both | det g ′ (u)|. 3.

Let us start with a rather trivial example. (u. and hence it maps quadrilaterals into quadrilaterals.Change of Variables One normally chooses changes of variables that map into rectangular regions. giving du ∧ dv = −3dx ∧ dy. 4). v = 2x + y =⇒ dv = 2dx + dy. 5). (1. 4). The form dx ∧ dy has opposite orientation to du ∧ dv so we use dv ∧ du = 3dx ∧ dy Free to photocopy and distribute 156 . 3). v) = 1 2 2 1 x y is a linear transformation. The corners of the rectangle in the area of integration in the xyplane are (0. xy-plane. or that simplify the integrand. 0 Solution: ◮ Observe that we have already computed this integral in example 265.16: Example 273. (traversed counter-clockwise) and they map into (6. 3). 3). 6). (7. (9. 273 Example Evaluate the integral 4 3 1 (x + 2y)(2x + y)dxdy. Put u = x + 2y =⇒ du = dx + 2dy. 7 6 5 4 3 2 1 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Figure 3. Now.16). 4). and (8. (1. respectively. in the uv-plane (see ﬁgure 3. uv-plane. Figure 3. and (0.15: Example 273.

0 ≤ v ≤ 2. y) : 0 ≤ y ≤ 1}. on DA. v = 0. 4 Free to photocopy and distribute 157 . The region ∆ is thus v2 the area in the uv plane enclosed by the parabolas u ≤ − 1. dv = 2ydx + 2xdy. v = 0. 0 ≤ v ≤ 2. 0) : 0 ≤ x ≤ 1}. 1 − y) : 0 ≤ y ≤ 1}. Put AB = {(x. We use the fact that boundaries will be mapped into boundaries. Since 0 ≤ y ≤ 1. 4 On CD we have u = (1 − x)2 − 1. CD = {(1 − x. 1]2 is mapped. On BC we have u = 1 − y 2 . v = 2y. 1) : 0 ≤ x ≤ 1}. BC = {(1. DA = {(0. 1 5 [0. 4 Finally. Solution: ◮ First we ﬁnd du = 2xdx − 2ydy. This means that v2 u= − 1. v = 0. ◭ 274 Example The integral (x4 − y 4 )dA = 1 0 P uv dvdu = 409 12 . v = 2xy. v = 0. AB is thus mapped into the line segment 0 ≤ u ≤ 1. v = 2(1 − x).1]2 − y 4 dy = 0. We now determine the region ∆ into which the square D = [0. DA is mapped into the line segment −1 ≤ u ≤ 0. we have u = −(1 − y)2 . Evaluate it using the change of variables u = x2 − y 2 . On AB we have u = x. 0 ≤ v ≤ 2. Since 0 ≤ x ≤ 1.Chapter 3 instead. Thus u = 1 − mapped to the portion of the parabola u = 1 − v 2 v2 4 . Hence BC is . The integral sought is 1 3 from example 267. u ≤ 4 2 v 1− with −1 ≤ u ≤ 1. and so du ∧ dv = (4x2 + 4y 2 )dx ∧ dy.

dx ∧ dy = 3u2 v 2 du ∧ dv. The integral becomes f (x. y = uv 2 . y 3 = v 2 u and verify that the same result is obtained. as before.1]2 (x4 − y 4 ) ∆ 1 4(x2 + y2 ) dudv 4 1 4 1 (x2 − y 2 )dudv ududv 1−v 2 /4 ∆ 2 0 4 = 0. ◭ 275 Example Find e(x D 3 udu dv v 2 /4−1 +y3 )/xy dA where D = {(x. dy = v 2 du + 2uvdv. +∞[ ﬁxed}. The region transforms into ∆ = {(u. p ∈]0. 0 ≤ v ≤ (2p)1/3 }. y)dxdy D = ∆ exp 3 u6 v 3 + u3 v 6 u3 v 3 3 (3u2 v 2 ) dudv = 3 ∆ eu ev u2 v 2 dudv (2p)1/3 2 2 u3 = = 3u e 3 0 1 2p (e − 1)2 . v) ∈ R2 |0 ≤ u ≤ (2p)1/3 . x2 − 2py ≤ 0. Solution: ◮ We have dx = 2uvdu + u2 dv. you may try the (more natural) substitution x3 = u2 v. y) ∈ R2 |y 2 − 2px ≤ 0. 3 1 du As an exercise. ◭ Free to photocopy and distribute 158 . using the change of variables x = u2 v.Change of Variables We deduce that (x4 − y 4 )dA = = = = 1 ∆ [0.

Prove that n=1 1 (2n − 1)2 = 0 dxdy 1 − x2 y 2 sin u cos v 3. 4 2. S. then S = 2 n 4 1 1 0 1 +∞ n=1 1 (2n − 1)2 . . Kock to prove that n2 6 n=1 +∞ 1. hence the sum of the odd terms must be three 3 quarters of the sum. 1 − x2 y 2 Solution: ◮ 1. Prove that if S = n=1 +∞ 3 . Observe that 1 2n − 1 1 = 0 x2n−2dx =⇒ 1 2n − 1 2 1 = 0 x2n−2 dx 1 0 y 2n−2 dy 159 1 1 0 = 0 (xy)2n−2 dx Free to photocopy and distribute . Figure 3.Chapter 3 π 1 0 0 1 2 π 2 Figure 3.17: Example 276. xy-plane.18: Example 276. uv-plane. 276 Example In this problem we will follow an argument of Calabi. a quarter of the sum. Use the change of variables x = 1 0 1 0 . and +∞ π2 1 = . Observe that the sum of the even terms is +∞ n=1 1 (2n)2 = 1 4 +∞ n=1 1 n2 = 1 4 S. Beukers. y = sin v cos u in order to evaluate dxdy . 2.

We will accept it for our purposes. This exchange of integral and sum needs justiﬁcation. 2 2 .Change of Variables Thus +∞ n=1 1 = (2n − 1)2 +∞ n=1 1 0 1 0 (xy)2n−2 dxdy = 1 0 1 +∞ 0 n=1 (xy)2n−2 dxdy = 1 0 1 0 dxdy . Free to photocopy and distribute 160 . 1 − x2 y 2 = 1 − This gives dxdy = dudv. 1 − x2 y 2 We now have to determine the region that the transformation x = sin u sin v . We deduce. 1 0 1 0 dxdy 1 − x2 y 2 π/2 π/2−v π/2 = 0 0 dudv = 0 (π/2 − v) dv = π 2 v− v2 2 ¬π/2 π2 π2 π ¬ = − = ¬ 0 4 8 Finally. This means that the square in the xy-plane in ﬁgure 3. Also. 4 8 6 ◭ Homework Problem 3. v) ∈ R2 : u ≤ 1.3 sin v sin u 3.17 is transformed into the triangle in the uv-plane in ﬁgure 3. 1 − x2 y 2 as claimed. Consider R2 Φ: 3 → → (u.18. If x = . from where dx∧dy = du∧dv−(tan2 u)(tan2 v)du∧dv = 1 − (tan2 u)(tan2 v) du∧dv. 1−x . 2 v = arctan y 1 − x2 1 − y2 .6. −u ≤ v ≤ u}. then cos v cos u dx = (cos u)(sec v)du+(sin u)(sec v)(tan v)dv. 3 π2 π2 S= =⇒ S = . Observe that cos v cos u u = arctan x 1 − y2 × dy = (sec u)(tan u)(sin v)du+(sec u)(cos v)d sin2 v cos2 v · sin2 v cos2 u = 1 − (tan2 u)(tan2 v).y= forms in the uv-plane.1 Let D ′ = {(u. v) R2 u+v u−v .y= .

y 2 − x2 ≤ 1. y) ∈ R2 |a ≤ xy ≤ b. 2).6. evaluate x−y dA.6. (1. R x + y Problem 3. v ≥ 0.Chapter 3 Find the image of Φ on D ′ . y)dA where D = {(x. that is. b) ∈ R2 . 2). v = x − y to shew that 1 0 0 1 dxdy =2 1 − xy 1 0 u −u dv 4 − u2 + v 2 2 2−u u−2 du + 2 1 dv 4 − u2 + v 2 du. where R is the square with vertices at (0. y ≥ x ≥ 0. (a. 1). Ô evaluate x2 + y 2 dA. y = 2uv. y) ∈ R2 : −1 ≤ x ≤ 1. 3). (2. 0 < a < b} and f (x. 2 n 6 Use the series expansion 1 = 1 + t + t2 + t3 + · · · 1−t in order to prove (formally) that 1 0 0 1 ∞ n=1 |t| < 1. n2 Use the change of variables u = x + y. ﬁnd D = Φ(D ′ ). u ≥ 0. Shew that the above integral reduces to 1 2 0 √ 2 u arctan √ du + 2 2 4−u 4 − u2 2 1 √ 2 2−u arctan √ du.6. 2 4−u 4 − u2 Finally. 0 ≤ y ≤ 2 1 − |x|} Problem 3. Problem 3.3 Using the change of variables u = x − y and v = x + y. prove that the above integral is π2 u by using the substitution θ = arcsin . where R R = {(x.2 Using the change of variables x = u2 − v 2 .4 Find D f (x.6. v = y 2 − x2 . dxdy = 1 − xy 1 .5 Use the following steps (due to Tom Apostol) in order to prove that ∞ n=1 1 π2 = . Find (x + y)2 ex D 2 −y 2 dA. 6 2 Free to photocopy and distribute 161 . y) = y 4 − x4 by using the change of variables u = xy. (1. Problem 3.

Since x2 + y 2 = r 2 . that is. y ≤ x. y = ρ sin θ =⇒ dy = sin θdρ + ρ cos θdθ. the radius sweeps from r 2 = 2r sin θ to r 2 = 4. Solution: ◮ We use polar coordinates.Change to Polar Coordinates 3. One of the most common changes of variable is the passage to polar coordinates where whence dx ∧ dy = (ρ cos2 θ + ρ sin2 θ)dρ ∧ dθ = ρdρ ∧ dθ.23. 277 Example Find xy D x2 + y 2 dA where D = {(x. where R is the region bounded by the circles x2 + y 2 = 4 and x2 + y 2 = 2y. 278 Example Evaluate R xdA. Free to photocopy and distribute 162 . x2 + y 2 ≤ 1}. y ≥ 0.5. Solution: ◮ Observe that this is problem 3. from 2 sin θ to 2. The region D transforms into the region π ∆ = [0. 1] × 0. 4 Therefore the integral becomes ρ4 cos θ sin θ dρdθ ∆ π/4 1 = 0 cos θ sin θ dθ 0 ρ4 dρ = ◭ 1 20 . y) ∈ R2 |x ≥ 0. .7 Change to Polar Coordinates x = ρ cos θ =⇒ dx = cos θdρ − ρ sin θdθ.

Figure 3. The angle clearly sweeps from 0 to π/2 π 2 .20: Example 278.19: Example 277. Solution: ◮ Completing squares x + xy + y = x + y 2 √ 3y 2 2 y 2 2 + √ 3y 2 2 .V = e−x 2 2 .Chapter 3 Figure 3. Figure 3. 3 . Figure 3.21: Example 279. Thus the integral becomes r 2 cos θdrdθ 2 sin θ xdA = R 3 = 2. Put U = x + . where D = {(x. ◭ 279 Example Find D = 1 0 2 π/2 0 (8 cos θ − 8 cos θ sin3 θ)dθ e−x 2 −xy−y2 dA. The integral becomes dxdy = 2 √ 3 e−(U {U 2 +V 2 ≤1} 2 −xy−y2 +V 2 ) dU dV. y) ∈ R2 : x2 + xy + y 2 ≤ 1}. the above equals 2 √ 3 ◭ Free to photocopy and distribute 163 2π 0 1 0 2π 2 ρe−ρ dρdθ = √ (1 − e−1 ).22: Example 280. {x2 +xy+y2 ≤1} Passing to polar coordinates.

y = br sin θ =⇒ dy = b sin θdr + br cos θdθ. 15 Free to photocopy and distribute 164 . y ≥ 1 (x2 + y 2 )3/2 1 280 Example Evaluate (ﬁgure 3.22). R Solution: ◮ The radius sweeps from r = integral is dA = (x2 + y 2 )3/2 = = ◭ 281 Example Evaluate R 1 sin θ 2 csc θ to r = 2. Then x = ar cos θ =⇒ dx = a cos θdr − ar sin θdθ.Change to Polar Coordinates dA over the region (x. Observe that on the ellipse x2 a2 + y2 b2 = 1 =⇒ a2 r 2 cos2 θ a2 + b2 r 2 sin2 θ b2 = 1 =⇒ r = 1. Thus the required integral is (x3 + y 3 )dA = R π/2 0 1 0 1 0 abr 4 (cos3 θ + sin3 θ)drdθ π/2 0 = = = ◭ ab ab 1 r 4 dr (a3 cos3 θ + b3 sin3 θ)dθ 2a3 + 2b3 5 3 2ab(a3 + b3 ) . The desired 1 5π/6 π/6 5π/6 1 r2 drdθ 1 2 dθ R √ π/6 π 3− . a > 0 and b > 0. y = br sin θ. a2 b Solution: ◮ Put x = ar cos θ. y) ∈ R2 : x2 + y 2 ≤ 4. whence dx ∧ dy = (abr cos2 θ + abr sin2 θ)dr ∧ dθ = abrdr ∧ dθ. 3 sin θ − (x3 + y 3 )dA where R is the region bounded by the ellipse x2 y2 + 2 = 1 and the ﬁrst quadrant.

4 Find and f (x. D = {(x. x ≥ 1.7. b) ∈]0. Problem 3. +∞[ ﬁxed} Problem 3. y) ∈ R2 |(x − 1)2 + y 2 ≤ 1}. (a. y) ∈ R2 : x2 + y 2 ≤ 16.7.7. y) ∈ R2 : b2 x2 + a2 y 2 = a2 b2 .1. Problem 3. 2 a b Free to photocopy and distribute 165 . Prove that Ö a 2 b2 − a 2 y 2 − b2 x 2 πab(π − 2) dA = . f dA where D Problem 3.7.Chapter 3 Homework Problem 3. y) ∈ R2 |(x2 + y 2 )2 ≤ 2xy}. xydA where R is the region Figure 3.7.2 Find (x2 − y 2 )dA D where D = {(x.5 Let a > 0 and b > 0. y ≥ 1 . as in the ﬁgure 3. a 2 b2 + a 2 y 2 + b2 x 2 8 R where R is the region bounded by the ellipse x2 y2 + 2 = 1 and the ﬁrst quadrant.23: Problem 3. Set up the integral in both Cartesian and polar coordinates.1 Evaluate R ¨ © R = (x.7. y) = x3 + y 3 .23.3 Find D √ xydA where D = {(x.

Compute D dA . y) = 1 . (1 + x2 + y 2 )2 Free to photocopy and distribute 166 . y) ∈ R2 : x ≥ 1. Problem 3. x2 + y 2 − 2x ≤ 0} and f (x. 2 +y where R = {(x.7. x2 + y 2 ≤ 4}.8 Find D Ô x2 + y 2 dA where D = {(x. y) ∈ R2 |x ≥ 1.7. y) ∈ R2 : x2 + y 2 − y ≤ 0. 5 Problem 3.7. x2 + y 2 − 2y ≥ 0}. y ≥ 0. Find xdA. f dA where D Problem 3.12 Let D = {(x.10 Let D = {(x. Find the integral (x + y)2 dA.7.7. x2 + y 2 − x ≤ 0}. y) ∈ R2 |x ≥ 0.9 Find D = {(x. x2 + y 2 − 2x ≤ 0} and f (x.6 Prove that Ô R x2 √ y dxdy = 2 − 1. y) ∈ R2 : 0 < x < 1. (x2 + y 2 )2 Problem 3.11 Find D f dA where D = {(x. x2 + y 2 ≤ 1. y) = x2 y.7. D Problem 3.7.13 Let D = {(x.7. o < y < x2 }.7 Prove that the ellipse (x − 2y + 3)2 + (3x + 4y − 1)2 = 4 bounds an area of 2π . Problem 3. D Problem 3. y) ∈ R2 |y ≥ 0.Change to Polar Coordinates Problem 3. y) ∈ R2 |y ≤ x2 + y 2 ≤ 1}.

7. Liouville was a mathematician. y 2 = ρ sin θ. y) be the distance from (x.x 4 +y 4 ≤1} Ô 1 − x4 − y 4 dA using x2 = ρ cos θ.y )∈R2 :x≥0.y ) dxdy = 2π + L + A.7. Prove that Ia ≤ Ja ≤ Ia √ 2 . Find Ia = Da e−(x 2 +y 2 ) dxdy. y) ∈ R2 |x2 + y 2 ≤ a2 }. 2 √ π 2 Let a > 0 be a real number and put Da = {(x.15 William Thompson (Lord Kelvin) is credited to have said: “A mathematician is someone to whom √ +∞ 2 π e−x dx = 2 0 is as obvious as twice two is four to you.Chapter 3 Problem 3.y ≥0. let D(x. Free to photocopy and distribute 167 .7. y) to the nearest point R.7.18 In the xy-plane. Deduce that 0 +∞ e−x dx = 2 √ π . y) ∈ R2 ||x| ≤ a. if R is the set of points inside and on a convex polygon. x2 + xy + y 2 Problem 3.” Prove that +∞ 0 e−x dx = by following these steps. Problem 3. 2 1 .7. Let a > 0 be a real number and put ∆a = {(x. y)dA. Problem 3.17 Prove that every closed convex region in the plane of area ≥ π has two points which are two units apart. y) ∈ R2 : 4 ≤ x2 + y 2 ≤ 16} and f (x. Show that +∞ −∞ +∞ e−D (x. |y| ≤ a}. Problem 3.16 Let D = {(x. Let Ja = ∆a e−(x 2 +y 2 ) dxdy. y) = Find D f (x.14 Evaluate x3 y 3 {(x. −∞ where L is the perimeter of R and A is the area of R.

8 Three-Manifolds 282 Deﬁnition A 3-dimensional oriented manifold of R3 is simply an open set (body) V ∈ R3 . Given a function f : V → R. dV V is the oriented volume of V . unless otherwise noticed. The region −M has opposite orientation to M and −M ω=− ω. 2 = e− ◭ Free to photocopy and distribute 168 . In this section. the integral f dV V is the sum of all the values of f restricted to V . In particular. 283 Example Find x2 yexyz dV. we will choose the positive orientation for the regions considered. where the + orientation is in the direction of the outward pointing normal to the body.Three-Manifolds 3. and the − orientation is in the direction of the inward pointing normal to the body. This corresponds to using the volume form dx ∧ dy ∧ dz.1]3 Solution: ◮ The integral is 1 0 1 0 1 0 x2 yexyz dz dy dx 1 1 0 = 0 1 x(exy − 1) dy dx = 0 (ex − x − 1)dx 5 . [0. A general oriented 3-manifold is a union of open sets. Let V ⊆ R3 . M We will often write f dV M where dV denotes the volume element.

z) ∈ R3 |x ≥ 0. Solution: ◮ The integral is 1 (1− √ 2 z) (1− 0 √ √ z− x)2 zdxdydz R = 0 1 z 0 √ (1− z)2 dy (1 − √ z)4 dz √ √ z − x)2 dx dx dz dz = 0 z 1 6 1 1 0 0 = = ◭ 285 Example Prove that z(1 − 840 . (0. y ≥ 0. 0). y ≥ 0. (3. y. z ≥ 0. 3.24. and (0.Chapter 3 284 Example Find R z dV if √ √ √ x + y + z ≤ 1}. xdxdydz V = 0 0 xdxdydz 1 2 1 ay az − b c 0 3 a2 (−z + c) b dx c3 a− b−bz/c 2 = = = ◭ dydz 6 0 a2 bc 24 286 Example Evaluate the integral S xdV where S is the (unoriented) tetrahe- dron with vertices (0. z ≥ 0. Solution: ◮ We have c b−bz/c 0 c 0 c a−ay/b−az/c Ò x a + y b + z c ≤1 . 0). 2. 0. z) ∈ R3 : x ≥ 0. y. xdV = V a2 bc . 24 Ó where V is the tetrahedron V = (x. 2). 0). See ﬁgure 3. R = {(x. Free to photocopy and distribute 169 . 0.

3. on x and y.z]=Tetrahedron(<0. 2) has equation 2x + 6y + 9z = 18.z C B Three-Manifolds O x A Figure 3.<3. The line 3 2 passing through AC has equation y = x. We must now ﬁgure out the xy limits of integra9 tion. 0. 3 We ﬁnd. 0). 18 − 2x − 6y 0≤z≤ .2>)).25 we draw the projection of the tetrahedron on the xy x plane.0. 1. With the order dzdxdy.y.25: xy-projection. Given an arbitrary point in the solid. 2. dxdydz Solution: ◮ We must ﬁnd the projections of the solid on the the coordinate planes. Use the following orders of integration: 1. y = x. where R is the solid formed by the intersection of the parabolic cylinder z = 4 − x2 . A C B y Figure 3. the limits of integration of z can only depend. and y = 0. its Free to photocopy and distribute 170 .<0. The line passing through AB has equation y = − + 3. the planes z = 0.0>.0>.<0. (0. Hence.3. Solution: ◮ A short computation shews that the plane passing through (3. ◭ 287 Example Evaluate R xyzdV .24: Problem 286. and (0. > with(Student[VectorCalculus]): > int(x.0.0>. dzdxdy 2. ﬁnally.2. 0). 3 3−x/3 2x/3 3−x/3 2x/3 (18−2x−6y)/9 xdV S = 0 3 0 xdzdydx dydx 9 18xy − 2x2 y − 3y 2 x ¬3−x/3 ¬ dx ¬ 2x/3 9 3 x − 2x2 + 3x dx 3 18x − 2x2 − 6yx = 0 3 = 0 3 = = 9 4 0 To solve this problem using Maple you may use the code below. if at all.[x. In ﬁgure 3.

1 Compute E zdV where E is the region in the ﬁrst octant bounded by the planes y + z = 1 and x + z = 1. the limits of integration of x can only depend. (1.8. x = 2. so the limits for z are from z = 0 to z = 4 − x2 . if at all. 0). x sweeps √ from the plane to x = 2. so the limits for x are from x = y to x = 4 − z. Problem 3. Free to photocopy and distribute 171 . and then by integrating in the order dydxdz. 1.8. y = x.Chapter 3 lowest z coordinate is 0 and its highest one is on the cylinder. bounded by the parabolic cylinder x2 2 z = 2− and the planes z = 0. and the z and y axes. Prove that xyz = ﬁrst by 2 3 S integrating in the order dzdydx. on y and z. 0. 0).4 Compute E xdV where E is the region in the ﬁrst octant bounded by the plane y = 3z and the cylinder x2 + y 2 = 9. (1. 1). With the order dxdydz.2 Consider the solid S in the ﬁrst octant. and y = 0. ◭ Homework Problem 3. 2 0 y 0 4−x2 xyzdzdxdy 0 = = = 1 2 1 2 2 2 0 2 0 y 0 y 0 xy(4 − x2 )2 dxdy y(16x − 8x3 + x5 )dxdy y7 12 dy 0 4y 3 − y 5 + = 8. The projection of the solid on the xy-plane is the area bounded by the lines y = x. The projection of the solid on the yz-plane is the area bounded by z = 4 − y 2 .8. where R is the tetrahedron with vertices at (0. 0. 2. 4 0 √ 0 4−z 2 xyzdxdydz y = = 1 2 4 4 0 √ 0 4−z 0 2z − z3 8 (4y − y 3 )zdydz dz = 8. 0. and the x and y axes. Problem 3. Given an arbitrary point in the solid. Problem 3.8. and (0.3 Evaluate the integrals R 1dV and R xdV . 1).

5 Find D = {(x. so we choose. 2x − z = 1 =⇒ u + v + 2w = 2. z) = (ρ cos θ. Consider a transformation to cylindrical coordinates (x. x + y − z = 0. Solution: ◮ We make the change of variables u = x + y + z =⇒ du = dx + dy + dz. R 288 Example Find where R is the tetrahedron bounded by the planes x + y + z = 0.9 Change of Variables (x + y + z)(x + y − z)(x − y − z)dV.8. du ∧ dw ∧ dv = 4dx ∧ dy ∧ dz which have the same orientation. This gives w = x − y − z =⇒ dw = dx − dy − dz. 3. The integral becomes 1 4 2 0 1−v/2 0 2−v−2w uvw dudwdv = 0 1 180 . ρ sin θ. dx ∧ dy ∧ dz = ρdρ ∧ dθ ∧ dz.Change of Variables dV where (1 + x2 z 2 )(1 + y 2 z 2 ) D Problem 3. From what we know about polar coordinates dx ∧ dy = ρdρ ∧ dθ. du ∧ dv ∧ dw = −4dx ∧ dy ∧ dz. say. x − y − z = 0. Also. and 2x − z = 1. z). y. v = 0. z) ∈ R3 : 0 ≤ x ≤ 1. z ≥ 0}. u + v + 2w = 1 in the uvw-coordinate frame. y. v = x + y − z =⇒ dv = dx + dy − dz. ◭ Free to photocopy and distribute 172 . Since the wedge product of forms is associative. These forms have opposite orientations. The tetrahedron in the xyz-coordinate frame is mapped into a tetrahedron bounded by u = 0. 0 ≤ y ≤ 1.

x2 +y 2 ≤ R2 .Chapter 3 289 Example Find R z 2 dxdydz if R = {(x. z) ∈ R3 |x2 + y 2 ≤ 1. y 2 +z 2 ≤ R2 . 16 ◭ 291 Example Three long cylinders of radius R intersect at right angles. z)dxdydz R = = π 3 0 dθ 0 ρ dρ 0 z 2 dz . y. 0 ≤ z. Using cylindrical coordinates ∆′ = (θ. Solution: ◮ The integral is 1 0 π/2 π/4 2 1 ρ3 dρdθdz = 15π . z = 1. x = y. 2π] × [0. Find the volume of their intersection. z) ∈ 0. D 2 (x2 + y 2 )dxdydz over the ﬁrst octant region bounded = 1 and x2 + y 2 = 4 and the planes z = 0. By symmetry. ρ. In this case it is easier to integrate with respect to z ﬁrst. V = 24 V ′ . 173 Ó Free to photocopy and distribute . +∞[. 1] through a cylindrical coordinate change. 1] × [0. z) ∈ R3 : 0 ≤ y ≤ x. R] × [0. The integral is therefore 2π 1 1 f (x. 0 ≤ z ≤ 1}. 0 ≤ z ≤ R2 − ρ2 cos2 θ . Solution: ◮ Let V be the desired volume. ◭ 290 Example Evaluate by the cylinders x2 + y x = 0. Ò π 4 × [0. Solution: ◮ The region of integration is mapped into ∆ = [0. z 2 +x2 ≤ R2 }. y. y. where V′ = D′ dxdydz. D ′ = {(x.

x ≥ 0. and so we choose dx ∧ dy ∧ dz = ρ2 sin φdρ ∧ dφ ∧ dθ instead. 174 Free to photocopy and distribute . = sin θ sin φdρ + ρ cos θ sin φdθ + ρ sin θ cos φdφ. z = ρ cos φ. ≤ 1. c) ∈]0. z ≥ 0 . ◭ Consider now a change to spherical coordinates x = ρ cos θ sin φ. y ≥ 0. = cos φdρ − ρ sin φdφ. √ 2 V = 16V ′ = 8(2 − ä çR 1 (R2 − ρ2 cos2 θ)3/2 dθ 2θ 0 3 cos π/4 1 − sin3 θ dθ cos2 θ √ 0 2 2 1 − u2 π/4 [tan θ]0 + du u2 1 ρ R2 − ρ2 cos2 θdρ dθ − √ 2)R3 . z) ∈ R3 : x2 a2 + y2 b2 + z2 c2 xyz dV if R = cos θ sin φdρ − ρ sin θ sin φdθ + ρ cos θ cos φdφ. b. We have dx dy dz This gives dx ∧ dy ∧ dz = −ρ2 sin φdρ ∧ dθ ∧ dφ. V′ π/4 R 0 R 0 √ 0 = 0 π/4 R2 −ρ2 cos2 θ dz ρdρ dθ = 0 π/4 = = u = cos θ = = Finally = R3 3 R3 0 3 √ ä ç 22 R3 1 − u−1 + u 1 3 √ 2−1 3 R . +∞[3 be ﬁxed. Find R= (x.Change of Variables Now. 292 Example Let (a. the form dρ ∧ dθ ∧ dφ is negatively oriented. y. y = ρ sin θ sin φ. From this derivation.

c > 0. The integration region is mapped into ∆ = [0. 2 2 cos θ sin θ dθ 0 0 ρ5 dρ π/2 0 cos3 φ sin φ dφ = (abc)2 . Also. y. z) ∈ R3 : x2 + y 2 + z 2 ≤ 9. Also. z) = (aρ cos θ sin φ. cρ cos φ). 3 Free to photocopy and distribute 175 .4 Prove that the volume enclosed by the ellipsoid x2 y2 z2 + 2 + 2 =1 2 a b c is 4πabc . ﬁnd this volume. ]. ﬁnd this volume. Set up integrals for the volume of this region in Cartesian. b > 0. Set up integrals for the volume of this region in Cartesian. where R is the region above the paraboloid z = x2 + y 2 and under the sphere x2 + y 2 + z 2 = 4. Then 2π π/2−arcsin 1/3 π/2−arcsin 2/3 2/ cos φ 1/ cos φ π/2 1 π π ] × [0. Problem 3. The integral becomes (abc)2 ◭ 293 Example Let V = {(x. 1 ≤ z ≤ 2}. Set up integrals for the volume of this region in Cartesian. y. Homework Ô Problem 3. cylindrical and spherical coordinates. where (x. Also.9.9.Chapter 3 Solution: ◮ We use spherical coordinates. Problem 3.2 Consider the integral R xdV . 1] × [0. cylindrical and spherical coordinates. Here a > 0. 48 dxdydz V = 0 ρ2 sin φ dρdφdθ = 63π 4 .1 Consider the region R below the cone z = x2 + y 2 and above the 2 2 paraboloid z = x + y for 0 ≤ z ≤ 1.9. ﬁnd this volume. bρ sin θ sin φ. Problem 3. cylindrical and spherical coordinates.3 Consider the region R bounded by the sphere x2 + y 2 + z 2 = 4 and the plane z = 1. We have dx ∧ dy ∧ dz = abcρ2 sin φdρ ∧ dφ ∧ dρ.9.

9. unless otherwise noticed.10 (Putnam Exam 1984) Find x1 y 9 z 8 (1 − x − y − z)4 dxdydz. x + y + z ≤ 1}.6 Prove that √ 2 2 2 e− x +y +z dV = π 2 − 2e−R − 2Re−R − R2 e−R . A general oriented 2-manifold in R3 is a union of surfaces.9. The surface −Σ has opposite orientation to Σ and −Σ ω=− ω. Σ In this section. dz ∧ dx. R where R = {(x. Problem 3. z) ∈ R3 : x ≥ 0.9 Compute the 4-dimensional integral ex x 2 +y 2 +u 2 +v 2 ≤1 2 +y 2 +u 2 +v 2 dxdydudv.10 Surface Integrals 294 Deﬁnition A 2-dimensional oriented manifold of R3 is simply a smooth surface D ∈ R3 .9.5 Compute and x2 + y 2 = 4. where the + orientation is in the direction of the outward normal pointing away from the origin and the − orientation is in the direction of the inward normal pointing towards the origin.9.8 Compute E z Ô x2 + y 2 + z 2 dV where E is is the upper solid hemisphere bounded by the xy-plane and the sphere of radius 1 about the origin. we will choose the positive orientation for the regions considered. z ≥ 0. Problem 3.Surface Integrals ydV where E is the region between the cylinders x2 +y 2 = 1 E Problem 3.9.y≥0 x 2 +y 2 +z 2 ≤R 2 Problem 3. dx ∧ dy}. Problem 3. x≥0.9. Free to photocopy and distribute 176 . E y 2 z 2 dV where E is bounded by the paraboloid x = 1 − y 2 − z 2 Problem 3. 3. This corresponds to using the ordered basis {dy ∧ dz. y. y ≥ 0.7 Compute and the plane x = 0. below the plane x − z = −2 and above the xy-plane.

y = v.Chapter 3 295 Deﬁnition Let f : R3 → R. Σ Ô ¬¬ 2 ¬¬ ¬¬d x¬¬ = 1 + 4u2 + 4v 2 du ∧ dv. dz = dv. and z = v. 296 Example Evaluate Σ z ¬¬d2 x¬¬ where Σ is the outer surface of the section of the paraboloid z = x2 + y 2 . y − 1 = sin u. The integral of f over the smooth surface Σ (oriented in the positive sense) is given by the expression f ¬¬d2 x¬¬. Observe that the domain D of Σ is the unit disk u2 +v 2 ≤ 1. 0 ≤ z ≤ 1. dz ∧ dx = −2vdu ∧ dv. Ô To evaluate this last integral we use polar coordinates. dy ∧ dz = −2udu ∧ dv. Solution: ◮ We have x2 + y 2 = 2y ⇐⇒ x2 + (y − 1)2 = 1. Solution: ◮ We parametrise the paraboloid as follows. and so (u2 + v 2 ) 1 + 4u2 + 4v 2 dudv D Ô 2π 1 0 = 0 ρ3 1 + 4ρ2 dρdθ = ◭ √ 1 (5 5 + ). We see that dx ∧ dy = du ∧ dv. Let x = u. 12 5 π 297 Example Find the area of that part of the cylinder x2 + y 2 = 2y lying inside the sphere x2 + y 2 + z 2 = 4. We parametrise the cylinder by putting x = cos u. Free to photocopy and distribute 177 . and so Now. dy = cos udu. Hence dx = − sin udu. z = u2 +v 2 . z ¬¬d2 x¬¬ = D ¬¬ ¬¬ (u2 + v 2 ) 1 + 4u2 + 4v 2 dudv. Σ ¬¬ ¬¬ Here ¬¬ 2 ¬¬ ¬¬d x¬¬ = (dx ∧ dy)2 + (dz ∧ dx)2 + (dy ∧ dz)2 ¬¬ ¬¬ is the surface area element.

(z 2 − zx)dzdx = Σ (z 2 − zx)dzdx = 8. 0). Also 0 ≤ u ≤ π. i. The integral is thus ¬¬ 2 ¬¬ ¬¬d x¬¬ Σ π √ 2−2 sin u π = 0 = 2 2 1 − sin u du √ 0√ = 2 2 4 2−4 . Σ ◭ Free to photocopy and distribute 178 . 2. v 2 = 4−2(1+sin u) = 2−2 sin u. Σ √ √ − 2−2 sin u πÔ dvdu = 0 2 2 − 2 sin udu Ô where Σ is the top side of the triangle with vertices at (2. 0). when z 2 = 4−2y.e. 3 2−y 0 − Σ xydxdy = − 0 and hence xdydz + (z 2 − zx)dzdx − xydxdy = 10. We project this plane onto the coordinate axes obtaining 4 2−z/2 0 2 0 2 xdydz = Σ 0 (2 − y − z/2)dydz = 4−2x 0 8 3 .Surface Integrals whence dx ∧ dy = 0. (0. dy ∧ dz = cos udu ∧ dv. 4). and so ¬¬ 2 ¬¬ ¬¬d x¬¬ = = (dx ∧ dy)2 + (dz ∧ dx)2 + (dy ∧ dz)2 cos2 u + sin2 u du ∧ dv = du ∧ dv. that is. dz ∧ dx = sin udu ∧ dv. (0. 0. Solution: ◮ Observe that the plane passing through the three given points has equation 2x + 2y + z = 4. ◭ 298 Example Evaluate xdydz + (z 2 − zx)dzdx − xydxdy. The cylinder and the sphere intersect when x2 + y 2 = 2y and x2 + y 2 + z 2 = 4. 2 xydxdy = − . 0.

Problem 3.10.11 Green’s.10.6 Evaluate xydydz − x2 dzdx + (x + z)dxdy. S ¬¬ ¬¬ Ô z ¬¬d2 x¬¬ over the conical surface z = x2 + y 2 between Problem 3. Your argument must use surface integrals. In R2 . Σ where Σ is the top of the triangular region of the plane 2x + 2y + z = 6 bounded by the ﬁrst octant.1 Evaluate Σ ¬¬ ¬¬ y ¬¬d2 x¬¬ where Σ is the surface z = x + y 2 . If ω is a differential form. Find the surface area of the part of the cone which lies between the planes z = 1 and z = 2.10.10. 3.2 Consider the cone z = x2 + y 2 .4 Evaluate z = 0 and z = 1. 0 ≤ y ≤ 2.10. this takes the name of Green’s Theorem. 0 ≤ x ≤ 1. Á 300 Example Evaluate C (x − y 3 )dx + x3 dy where C is the circle x2 + y 2 = 1. then ω= ∂M M dω. 299 Theorem (General Stoke’s Theorem) Let M be a smooth oriented manifold.10. Stokes’. if ω is a 1-form. having boundary ∂M . Problem 3.Chapter 3 Homework Problem 3.3 Evaluate z 2 = 1.5 You put a perfectly spherical egg through an egg slicer. ¬¬ ¬¬ x2 ¬¬d2 x¬¬ where Σ is the surface of the unit sphere x2 + y 2 + Σ Problem 3. but you forgot to peel it ﬁrst! Shew that the amount of egg shell in any of the slices is the same. 179 Free to photocopy and distribute . Ô Problem 3. resulting in n slices of identical height. and Gauss’ Theorems We are now in position to state the general Stoke’s Theorem.

y) ∧ dx + dg(x. and using polar coordinates. again resorting to polar coordinates. y)dx + g(x. y) − f (x. let ω = f (x. Thus by Green’s Theorem. Then dω = df (x. Free to photocopy and distribute 180 . y) − ∂ ∂y f (x. observe that 1 = (sin2 θ + cos2 θ)2 = sin4 θ + 2 sin2 θ cos2 θ + cos4 θ. Á C (x − y 3 )dx + x3 dy 2π = 0 2π (cos θ − sin3 θ)(− sin θ)dθ + (cos3 θ)(cos θ)dθ (sin4 θ + cos4 θ − sin θ cos θ)dθ. y) dx ∧ dy ∂x ∂y ∂ M 2 ∂ ∂ g(x. y)dy ∧ dy ∂x ∂y which gives the classical Green’s Theorem f (x. y)dx + g(x. y) ∧ dy ∂ ∂ = f (x. y) dxdy. y)dy = ∂M 3 ∂x g(x. Á The region M is the area enclosed by the circle x2 + y 2 = 1. whence the integral equals 2π 0 (sin4 θ + cos4 θ − sin θ cos θ)dθ 2π = = 3π . Stokes’. y)dy be a 1-form in R . y)dx + f (x. and Gauss’ Theorems Solution: ◮ We will ﬁrst use Green’s Theorem and then evaluate the integral directly. y)dy ∧ dx + ∂x ∂y ∂ ∂ = g(x. C (x − y 3 )dx + x3 dy = = = 3π 0 (3y 2 + 3x2 )dxdy M 2π 1 0 3ρ2 ρdρdθ . 2 Aliter: We can evaluate this integral directly. In R . the above theorem takes the name of Gauß’ or the Divergence Theorem. We have dω = (dx − 3y 2 dy) ∧ dx + (3x2 dx) ∧ dy = d(x − y 3 ) ∧ dx + d(x3 ) ∧ dy = (3y 2 + 3x2 )dx ∧ dy. 2 0 (1 − 2 sin2 θ cos2 θ − sin θ cos θ)dθ ◭ In general. y)dx + g(x. if ω is a 2-form. = 0 To evaluate the last integral.Green’s.

z)dz ∧ dx + h(x. 3 2π 0 xdydz Σ = = 36π. The integral becomes dxdydz M = 4π 3 (27) = 36π. z)dx ∧ dy Free to photocopy and distribute 181 . Σ Σ since (x. ◭ In general. y. z) → z is an odd function of z and the domain of integration is symmetric with respect to z. z) → −y is an odd function of y and the domain of integration is symmetric with respect to y. Now. Σ since (x. Now. y. let ω = f (x. y. We have (x − y)dydz = xdydz. Similarly −ydxdy = 0. Aliter: We could evaluate this integral directly. y. dω = = (dx − dy) ∧ dy ∧ dz + dz ∧ dz ∧ dx − dy ∧ dx ∧ dy dx ∧ dy ∧ dz. Solution: ◮ The region M is the interior of the sphere x2 +y 2 +z 2 = 9. y. Σ since (x. y.Chapter 3 301 Example Evaluate of the sphere S (x − y)dydz + zdzdx − ydxdy where S is the surface x2 + y 2 + z 2 = 9 and the positive direction is the outward normal. z)dy ∧ dz + g(x. z) → −y is an odd function of y and the domain of integration is symmetric with respect to y. −3 |ρ| 9 − ρ2 dρdθ Also zdzdx = 0.

y. y. ∂x ∂y ∂z ∂ M which gives the classical Gauss’s Theorem f (x. y. z) h(x. Á 302 Example Evaluate ydx+(2x−z)dy+(z−x)dz where C is the intersection C 2 of the sphere x2 + y 2 + z = 4 and the plane z = 1. z) dx ∧ dy ∧ dz. y.. y. d S = dzdx . y. z) dx Using classical notation. z)dx ∧ dy ∂ ∂ ∂ = f (x. y. the surface Σ on which we are integrating is the inside of the circle x2 + y 2 + 1 = 4. y. z)dy + f (x. Σ Free to photocopy and distribute 182 . z)dz ∧ dz ∧ dx ∂x ∂y ∂z ∂ ∂ ∂ + h(x. z)dy + g(x. Σ Σ Since this is just the area of the circular region x2 + y 2 ≤ 3. i. y. Stokes’.e. y. z)dz ∧ dx ∧ dy ∂x ∂y ∂z ∂ ∂ ∂ = f (x. if ¾ f (x. y. Also. y. y. z)dy + h(x. z) + g(x. z = 1 implies dz = 0 and so dω = dxdy. y. z) ¿ ¾ →= − a then M → − g(x. z) . z)dz ∧ dx + dh(x. z = 1. y. y. z)dzdx+h(x. y. dxdy − → • d→. z)dxdy = ∂M ∂x ¿ f (x.Green’s. and Gauss’ Theorems be a 2-form in R3 . Solution: ◮ We have dω = = = (dy) ∧ dx + (2dx − dz) ∧ dy + (dz − dx) ∧ dz dx ∧ dy + dy ∧ dz + dz ∧ dx. y. the integral evaluates to dxdy = 3π. y. z)dx + g(x. y. z) + ∂ ∂z h(x. z)dydz+g(x. z)dx + f (x. z)dx + h(x. −dx ∧ dy + 2dx ∧ dy + dy ∧ dz + dz ∧ dx Since on C. z) + ∂ ∂y g(x. z) + h(x. − a S ∂M dydz − (∇ • →)dV = a The classical Stokes’ Theorem occurs when ω is a 1-form in R3 . x2 + y 2 = 3. y. y. z)dy ∧ dz + dg(x. y. y. z)dz ∧ dy ∧ dz ∂x ∂y ∂z ∂ ∂ ∂ + g(x. Then dω = df (x.

y. z)dx + g(x. y. y. z) dx dz →= − a g(x.1 Evaluate (2. z) dy ∧ dz ∂y ∂z ∂ ∂ f (x. − − a r Homework Á Problem 3. y. y. y. y. Then dω = = df (x. y. y. y. ∂x ∂y Using classical notation. y. z) dxdy. z)dy + g(x. y. y. y. y. (2. y. y. y. let ω = f (x. y. z)dz ∧ dz ∂x ∂y ∂z ∂ ∂ h(x. y. z) . z) dydz f (x. Free to photocopy and distribute 183 . z)dz ∂M = M ∂ ∂y + h(x. z) − ∂ ∂z g(x. z)dy + +h(x. z) dx ∧ dy ∂x ∂y = which gives the classical Stokes’ Theorem f (x. y. y. z) dz ∧ dx + ∂z ∂x ∂ ∂ g(x. then M → − − (∇ × →) • d S = a ∂M → • d→. C x3 ydx + xydy where C is the square with vertices at (0.Chapter 3 ◭ In general. z) − f (x. z)dz ∧ dy ∂x ∂y ∂z ∂ ∂ ∂ + h(x. z) − h(x. y. y. z)dz be a 1-form in R3 . z) ∧ dz ∂ ∂ ∂ f (x. z)dy + h(x. if ¾ ¿ ¾ ¿ ¾ ¿ ∂ g(x. z)dz ∧ dx ∂x ∂y ∂z ∂ ∂ ∂ + g(x. y. y. z) − ∂ f (x. z)dx + f (x. r → − dS = dydz dxdy dzdx . z) h(x. z) − f (x. z)dy + f (x. y. z) ∧ dx + dg(x. 0). z) dxdy ∂z ∂x ∂ ∂ + h(x. 2) and (0. y. y. 0). z)dx + h(x. y. y.11. y. z) ∧ dy + dh(x. z) − g(x. z)dy + h(x. z)dx + g(x. 2). z)dx + g(x. y. − d→ = dy .

and LB C . and Gauss’ Theorems Problem 3. 1. Prove that in Cartesian coordinates. y. 0 ≤ y ≤ 1.2 Consider the triangle △ with vertices A : (0. Evaluate If LP Q denotes the equation of the line joining P and Q ﬁnd LAB . 2. Here the boundary of the box is positively oriented considering outward normals. △ Find (1 − 2y)dx ∧ dy D where D is the interior of △. Prove that 2 0 0 1 0 1 ∂M (arctan y − x2 )dydz + (cos x sin z − y 3 )dzdx + (2zx + 6zy 2 )dxdy = 3y 2 dxdydz = 2. Problem 3.11. The surface ∂M of the solid is positively oriented upon considering outward normals. the boundary of the box without the upper top S = ∂M \ U . Prove that in cylindrical coordinates.4 Problems 1 through 4 refer to the box M = {(x. Prove that ∂M xdydz + ydzdx + 2zdxdy = 2π. Prove that dω = 3y 2 dx ∧ dy ∧ dz. the upper face of the box U = {(x. Problem 3. 4. 1 √ − 1−x 2 1−x 2 0 2. Stokes’. and the solid M whose boundaries are the paraboloid z = 1 − x2 − y 2 . 0 Free to photocopy and distribute 184 . 3.11. 0 ≤ z ≤ 1 and the disc x2 + y 2 ≤ 1. 0). and the differential form ω = (arctan y − x2 )dy ∧ dz + (cos x sin z − y 3 )dz ∧ dx + (2zx + 6zy 2 )dx ∧ dy. 1 0 0 1−r 2 3. 0 ≤ z ≤ 2}.3 Problems 1 through 4 refer to the differential form ω = xdy ∧ dz + ydz ∧ dx + 2zdx ∧ dy. 1). 0 ≤ y ≤ 1.11. 2). M dω = 0 4rdzdrdθ. Á y 2 dx + xdy. 1. z) ∈ R3 : 0 ≤ x ≤ 1. Prove that the integral on the upper face of the box is 1 1 0 U (arctan y − x2 )dydz + (cos x sin z − y 3 )dzdx + (2zx + 6zy 2 )dxdy = 4x + 12y 2 dxdy = 6. y. Prove that dω = 4dx ∧ dy ∧ dz. ∂M ω= −1 2π 1−x 2 −y 2 √ 4dzdydx. z) ∈ R3 : 0 ≤ x ≤ 1.Green’s. B : (1. z = 0. LAC . C : (−2. z = 2}.

Free to photocopy and distribute 185 . and C(0. Also. Prove that the integral on the open box is y 3 )dzdx + (2zx + 6zy 2 )dxdy = −4. following to C. Use Stoke’s Theorem to prove that ydx + zdy + xdz = −2π Γ √ 2. as viewed from a point far above the triangle. Γ where Γ is the circle (x − 2)2 + y 2 = 4. 3. continuing to B. oriented clockwise when viewed from the origin.11.5 Problems 1 through 3 refer to a triangular surface T in R3 and a differential form ω. The differential form is ω = (2xz + arctan ex ) dx + (xz + (y + 1)y ) dy + xy + y2 + log(1 + z 2 ) dz. The vertices of T are at A(6. 12. Prove this directly by parametrising the boundary of the surface and evaluating the path integral.7 Let Γ denote the curve of intersection of the plane x + y = 2 and the sphere x2 − 2x + y 2 − 2y + z 2 = 0. prove this directly by using a path integral. Prove that 3 0 0 12−4z ∂T (2xz + arctan ex ) dx+(xz + (y + 1)y ) dy+ xy + 6 0 0 3−x/2 y2 + log(1 + z 2 ) dz = 2 ydydz + 2xdzdx=108. B(0. and ending again at A.6 Use Green’s Theorem to prove that (x2 + 2y 3 )dy = 16π. 3).11. Prove that dω = ydy ∧ dz + (2x − y) dz ∧ dx + zdx ∧ dy. 0). 2. Prove that the equation of the plane that contains the triangle T is 2x + y + 4z = 12. Problem 3.8 Use Green’s Theorem to evaluate Á (x3 − y 3 )dx + (x3 + y 3 )dy. 0.11. Problem 3. Problem 3. ∂ M \U Problem 3. 0.11. 2 1. 0). The boundary of of the triangle ∂T is oriented positively by starting at A. C where C is the positively oriented boundary of the region between the circles x2 + y 2 = 2 and x2 + y 2 = 4. The surface T is oriented positively by considering the top of the triangle.Chapter 3 (arctan y − x2 )dydz +(cos x sin z − 4.

1. and BD = BE + ED.5 x + 1 = t = 2 − y =⇒ y = −x + 1. The zero vector 0 .2 We have 2BC = BE + EC.β = .A Answers and Hints → − 1. − → − → − → − → − → − → − → − → − → 1.m = . 1. [C].1.4 We have IA = −3IB ⇐⇒ IA = −3(IA + AB) = −3IA − 3AB. AD = BC.1.1. l = . ⇐⇒ − → − → 4IA = −3AB − → − → 4AI = 3AB − → → 3− AI = AB. [B]. We deduce that − → − → − → − → − → − → − → − → AC + BD = AE + EC + BE + ED = AD + BC.n = .1 No. 0 . [E]. has magnitude but no direction. → − → − → − → − → − − 1. − → − → − → − → − → − → − → 1. 0 . − → → − → → 3− 1− Thus we take I such that AI = AB and J such that AJ = AB.1. 4 − → − → 3JA = −JB − → − → − → 3JA = −JA − AB − → − → 4JA = −AB − → → 1− AJ = AB 4 − → − → − → − → 3MJ + 3JA + MJ + JB − → − → − → 4MJ + 3JA + JB − → 4MJ. 0 .1. 7 7 9 9 3 − → − → − → MI + IA + 3IB − → − → − → 4MI + IA + 3IB − → 4MI. [D]. Hence − → − → − → − → − → AC + BD = AD + BC = 2BC.6 α = 4 3 4 2 1 . 4 4 Now −→ − −→ − MA + 3MB = = = and −→ −→ − − 3MA + MB = = = 1. − → − → But since ABCD is a parallelogram. 0 .8 [A]. 2→ (= 2 d ) c 186 . By Chasles’ Rule AC = AE + EC. Thus we deduce − → − → − → IA + 3IA = −3AB ⇐⇒ ⇐⇒ Similarly → − → 1− JA = − JB 3 ⇐⇒ ⇐⇒ ⇐⇒ ⇐⇒ .

8 The transformations are shewn in ﬁgures A.3.3 The desired transformations are shewn in ﬁgures through A. æ 1.4. a2 ) = (−2. 1. −3).8 through A. 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 Figure A. bc = −a2 1.2 The desired transformations are in ﬁgures A. Figure A.2: Vertical Stretch. 2 1.4 A. a2 ) = (2.2 Plainly. Horizontal Figure A.1: Stretch.1 through A.4.Appendix A 1. 2 2 1 1 1 1.1 Upon solving the equations −3a1 + 2a2 = 0.10. b = − . æ é a b æ é æ é b − a −1 a+b 1 = + .9 a c b −a é .5.4.3.1 a = −3. a2 + a2 = 13 1 2 Free to photocopy and distribute 187 .3: Horizontal and Vertical Stretch.4.7.4. 1. we ﬁnd (a1 . 3) or (a1 .

2 4 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 Figure A. − − → 1. a a a → − → − − But the norm of a vector is 0 if and only if the vector is the 0 vector. we v v a v a gather → − − → − → − − − (→ − b )•(→ − b ) = ||→ − b ||2 = 0. 2 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 Figure A. i.. a → − − → − − (→ + b )•(→ + b ) a a − − →•→ + 2→ •→ + → •→ − − − − a b b b a a → •→ − − → •→ + 0 + b b − − a a → − − ||→||2 + || b ||2 .5: Levogyrate rotaπ tion radians. we have a → − − ||→ + b ||2 a = = = = from where the desired result follows. whence the assertion follows.3 By the CBS Inequality.8: Reﬂexion about the x-axis.10: Reﬂexion about the origin. → = b .6: Dextrogyrate roπ tation radians.5 4 3 2 1 0 -1 -2 -3 5 -4 4 3 2 1 0 -1 -2 -3 -4 -4 5 -3 4 -2 -1 0 1 2 3 4 5 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 5 4 3 2 1 0 -1 -2 -3 -4 -4 -3 -2 -1 0 1 2 3 4 5 Answers and Hints Figure A. → − → − − − − − − 1. Figure A. 1.2 Since →• b = 0. 2 Figure A.5. Figure A. → • (→ − b ) = 0. (a2 · 1 + b2 · 1) ≤ (a4 + b4 )1/2 (12 + 12 )1/2 .4 We have ∀→ ∈ R2 . a Free to photocopy and distribute 188 . 4 Figure A.e. choosing → = → − b . In particular.5.4: Levogyrate rotaπ tion radians.5. Therefore → − b = a → − → − − 0 .9: Reﬂexion about the y-axis .7: Dextrogyrate ro3 π tation radians.

6 A parametric equation for L1 is x y A parametric equation for L2 is x y = 0 b2 +t = 0 b1 +t æ 1 m1 é .5.8 We must prove that → •→ = 0. since L′ must pass through 3)x + b =⇒ 2 = (−2 ± 3)(−1) + b =⇒ b = ± 3 √ √ √ √ and the lines are y = (−2 + 3)x + 3 and y = (−2 − 3)x − 3. Using the distributive law for the dot product. a w − •→ w = = = = Free to photocopy and distribute 189 .7 The line L has a parametric equation x y Let L′ have parametric equation x y = 0 b +t = 0 1 +t æ 1 é . − − 4u v giving the result. 6 and −1 2 æ 1 é This gives two possible values for the slope of L′ . m1 m2 1.5. y = (−2 ± √ √ √ − − 1. −1 æ 1 é . The lines are perpendicular if and only if. æ é æ é 1 1 • = 0 ⇐⇒ 1 + m 1 m 2 = 0 ⇐⇒ m 1 m 2 = −1. 6 −1 √ √ π 1 + m 2 2 cos =⇒ m = −2 ± 3. respectively.5. m æ We need the angle between 1−m = 1 é m Ô π to be and so by Theorem 21. 1.Appendix A 1.5. according to Corollary 22.5 We have − − − − ||→ + → ||2 − ||→ − → ||2 u v u v = = = − − − − − − − − (→ + → )•(→ + →) − (→ − →)•(→ − → ) u v u v u v u v → •→ + 2→ •→ + → •→ − (→ •→ − 2→ •→ + →•→ ) − − − − − − − − − − − − u u u v v v u u u v v v → •→ . Now. æ 1 m2 é . → → − − v w − → − ¬¬ • ¬¬ → − v w − ¬¬→ ¬¬2 w → → − − v w − − →•→ − ¬¬ • ¬¬ → •→ − − v w w w − ¬¬→¬¬2 w → → − − →•→ − v • w ¬¬→¬¬2 − − v w w ¬¬→¬¬2 ¬¬− ¬¬ ¬¬− ¬¬ w → •→ − → •→ − − − − v w v w 0.

6. 1. What happens as t → −1? 1. 1) (at t = 0). y = log cos t. 2 − 2 = 1. 3) at t = . = tan θ =⇒ dt = 1 sec2 θdθ. −1 ≤ x ≤ 1. Since in the interval [0. 1) at t = . 1. 3) again at t = . 1) when t = 4π.1 We have y − x = 4t =⇒ t = x= is the Cartesian equation sought. x2 y2 − 2 = 1. this is a straight line with positive slope. −1}.3 1. 2. π /3 π .4 We may simply give the trivial parametrisation: x = t. 0).6 First observe that x + y = 1 demands x ∈ [0. so the trace is part of this line. + y5 y−x and so 4 y−x 4 3 −2 y−x 4 5 x5 If t = 0. + 1 dt. This is one branch of a hyperbola.Answers and Hints 1. This gives (dx)2 + (dy)2 = To integrate 1 2 4 t− √ 2 2 0 we now use the trigonometric substitution 2 t− 1 2 1 2 Ô 2 4t2 + (2 − 2t)2 dt = 2 2t2 − 2t + 1 dt = √ 2 4 t− 1 2 2 + 1 dt. we want the portion of the line y = 2x + 1 with −1 ≤ x ≤ 1 (and. 1].2 Observe that for t = {0. 4π]. √ √ 1. a b 3.6. The curve starts at the middle point (0.6. t ∈ [0. 2 Free to photocopy and distribute 190 . −1 ≤ sin t ≤ 1. 3 Hence the arc length is sec tdt = log(2 + 0 √ 3). 1]. y = 0. and ﬁnishes in the 2 2 middle point (0. one can give many parametrisations. ay − cx = ad − bc. then x = 0. 1] and [0. reaches its low point (−1. it goes to its low point (−1.6.6. One is x = t2 . reaches π 3π the high point (1. y 2 y x = t =⇒ x = y x 1+ x =⇒ x = y 2 x3 =⇒ x5 + y 5 = x2 y 2 . 2 a b y2 x2 4. 1) at t = .6. y = (1 − t)2 . 0 ≤ t ≤ Then (dx)2 + (dy)2 = (1 + tan2 t)(dt)2 = sec2 t(dt)2 . y = 0 and our Cartesian equation agrees. 0) to (1. this is the line segment on the plane joining (−1. 1. reaches its high 2 2 5π 7π point (1. x > 0. This is a hyperbola. thus −1 ≤ y ≤ 3).5 Observe that y = 2x + 1. Again.

10 Observe that the parametrisation traverses the curve once clockwise if t ∈ [0. dy = √ 2t + 1dt =⇒ (dx)2 + (dy)2 = Ô √ t2 + ( 2t + 1)2 dt = t2 + 2t + 1dt = (t+1)dt. and it appears in most Calculus texts: √ π /4 2 0 sec3 θdθ = = = √ 1 1 sec θ tan θ − log | sec θ + tan θ| 2 2 √ √ 1 √ 1 2 · 2 + log( 2 + 1) 2 2 √ 1 √ log( 2 + 1) + 1. 2π]. Hence.8 Observe that dx = 6tdt. The area under the graph is t=1 t=1 t=1 ydx = t=0 t=0 (1 − t2 )d(t3 + 1) = t=0 3t2 (1 − t2 )dt = 2 .6. The area is given by æ é Á Á x dx 1 1 det = xdy − ydx 2 Γ 2 y dy 0 4 = (sin3 t(− sin t(1 + sin2 t) + 2 sin t cos2 t) 2 π /2 − cos t(1 + sin2 t)(3 sin2 t cos t))dt 0 = = = 2 π /2 0 (−3 sin2 t + sin4 t)dt − 9 1 + cos 2t + cos 4t dt 8 8 2 9π .9 Observe that dx = tdt. 8 π /2 Free to photocopy and distribute 191 .Appendix A The integral thus becomes √ π /4 2 0 sec3 θdθ. −1/2 1.6. (You write sec θd tan θ = tan θ sec θ − tan θd sec θ. etc.7 First notice that x = 1 =⇒ t3 + 1 = 1 =⇒ t = 0 and x = 2 =⇒ t3 + 1 = 2 =⇒ t = 1. 1. 2 2 π /4 0 1. 5 1.) I will simply quote it. as I assume most of you have seen it.6. (dx)2 + (dy)2 = t=0 t=0 6t Ô √ 1 + t2 dt = 4 2 − 2. and so the arc length is t=1 t=1 dy = 6t2 dt =⇒ (dx)2 + (dy)2 = 6t Ô 1 + t2 dt.6. the famous secant cube integral. the arc length is 1/2 (1 + t)dt = −1/2 t+ t2 2 1/2 = 1. which is a standard example of integration by parts where you “solve” for the integral.

making the point P to rotate an angle of θ radians. Hence to obtain the loop Using integration by substitution (u = 1 + t3 and du = 3t2 dt) . the y-axis is vertical.6.11 Using the quotient rule. dx = and dy = Hence xdy−ydx = 9t2 (1 + t3 ) 9t2 − 18t5 9t2 + 9t5 9t2 18t2 − 9t5 ·dt− ·dt = ·dt = ·dt = ·dt. ρ).12 See ﬁgure 1. ρ).Answers and Hints 1. ρ − d). This is the desired parametrisation. 1. h) with u ≥ 0. Let θ be the angle (in radians) of rotation of the circle. Since we know that the shell strikes the plane. 2 where a is the angle of elevation.6. h = V t sin a − gt2 . The point P has moved x = ρθ − d sin θ horizontal units and y = ρ − d cos θ units. Then the centre of the circle has displaced rθ units horizontally. If the gun were ﬁred at t = 0. At θ = 0 the centre of the circle is at (0. then x = V t cos a.6. dy = et (sin t+cos t)dt =⇒ The arc length is thus π (dx)2 + (dy)2 = et (cos t − sin t)2 + (sin t + cos t)2 dt = (dx)2 + (dy)2 = 0 √ π 2 0 et dt = √ 2(eπ − 1). and P = (0. and the airplane is on a point with coordinates (u. and let C be the centre of the circle.15 Choose coordinates so that the origin is at the position of the gun. d cos θ). then x → 0 and y → 0. we must have u = V t cos a.14 We have dx = et (cos t−sin t)dt. and so is now located at (ρθ. y = V t sin a − gt2 .58. (1 + t3 )3 (1 + t3 )3 (1 + t3 )3 (1 + t3 )3 (1 + t3 )2 6t(1 + t3 ) − 3t2 (3t2 ) 6t − 3t4 18t2 − 9t5 · dt = · dt =⇒ xdy = · dt 3 )2 3 )2 (1 + t (1 + t (1 + t3 )3 3(1 + t3 ) − 3t2 (3t) 3 − 6t3 9t2 − 18t5 · dt = · dt =⇒ ydx = · dt (1 + t3 )2 (1 + t3 )2 (1 + t3 )3 Observe that when t = 0 then x = y = 0. u2 2 A shorter way of obtaining xdy − ydx would have been to argue that y 9t2 xdy − ydx = x2 d = dt. As t → +∞. the area is given by 1 2 +∞ 0 9t2 3 · dt = (1 + t3 )2 2 +∞ 0 3t2 3 dt = (1 + t3 )2 2 +∞ 1 du 3 = . in relation to the centre of the circle (notice that the circle moves clockwise). The polar coordinates of the point P are (d sin θ. x (1 + t3 )2 1.6. Suppose the circle is displaced towards the right. t is the time and g is the acceleration due to gravity. 1. 2 Free to photocopy and distribute 192 .

g 2 t4 + (gh − V 2 )2 t2 + h2 + u2 = 0.7. ¬¬ ¬¬ ¬¬ ¬¬ − − ¬¬→ →¬¬2 ¬¬→ →¬¬2 − − = ¬¬ a − b ¬¬ + ¬¬ a + b ¬¬ = whence ¬¬ ¬¬2 ¬¬ ¬¬ ¬¬ ¬¬2 ¬¬→¬¬2 − − − − − − − ¬¬− ¬¬ − 2→ •→ + ¬¬→ ¬¬ + ¬¬→ ¬¬2 + 2→•→ + ¬¬→ ¬¬ a a b a a b ¬¬ b ¬¬ ¬¬ b ¬¬ ¬¬ ¬¬2 ¬¬− ¬¬2 − ¬¬→¬¬ 2¬¬→¬¬ + 2¬¬ b ¬¬ . 1. y(t)). Ǒ = 2 3 + t −1 . 8px2 16px3 0 0 . 4 The quadratic equation in t2 has a real root if (gh − V 2 )2 ≥ g 2 (h2 + u2 ) =⇒ g 2 u2 ≤ V 2 (V 2 − 2gh). 1. 0 the vertex V of the rolling parabola is the reﬂexion of the origin about the line tangent to 1 their point of contact. in general.1 Observe that. from where the equation of 2x0 the tangent is x + 2px0 .Appendix A whence u2 + h + and thus 1 2 gt 2 2 = V 2 t2 . 2 1 + 4x0 1 + 4x2 0 := (x(t). and so the lines intersect at − from where V = − 4px2 8px3 0 0 . As −2x0 x(t) = y(t). a ¬¬ ¬¬ − ¬¬→ →¬¬ − ¬¬ a − b ¬¬ = x 1.6.7. giving the equation of the locus. from where the assertion follows.16 Suppose the parabolas have a point of contact P = (4px2 . 4px0 ). y= 2x0 The line normal to this line and passing through the origin is hence y = −2xx0 . Free to photocopy and distribute 193 . By symmetry. 1 + 4x2 1 + 4x2 0 0 . eliminating t yields x=− or 2p y ¡2 y ¡2 x x 1+ (x2 + y 2 )x + 2py 2 = 0.2 y z ¬¬ ¬¬2 ¬¬ ¬¬ ¬¬− ¬¬2 − → ¬¬ − ¬¬→¬¬ ¬¬− 2¬¬→¬¬ + 2¬¬ b ¬¬ − ¬¬→ + b ¬¬ = a a Ǒ 1 ¾ ¿ −2 0 2(13)2 + 2(19)2 − (24)2 = 22. The slope of the tangent at P is .

−1 and thus the equation of the desired plane is −2 . The desired equation is thus 1 4 3 x − y + z = 0. 6 3 2 1.4 Observe that Thus the vector 0 0 (x − 1) − 2(y − 1) − (z − 1) = 0. Two a × b × 1 bricks.5 The set B can be decomposed into the following subsets: The set A itself. and four quarter-cylinders of length c and radius 1. z = t/3. this means that −2 −1 Ǒ 0 1. two b × c × 1 bricks. and hence on the plane. and thus ¾ 1 ¿ × ¾ 1 ¿ = ¾ 1/6 3/2 ¿ −1 −1 1/2 1/3 −4/3 is normal to the plane. Free to photocopy and distribute 194 .7. if x = 2y = 3z = t.7. Now. then x = t. y = t/2. of volume abc. four quarter-cylinders of length b and radius 1. Hence. (as 0 = 2(0) = 3(0)) is on the line.3 The vectorial form of the equation of the line is ¾ ¿ ¾ ¿ 1 1 →= − r 0 + t −2 . and two c × a × 1 bricks. 0 1/3 1/3 ¾ This means that 1 1/2 1/3 ¿ also lies on the plane. the vectorial form of the equation of the line is ¾ ¿ ¾ ¿ ¾ ¿ 0 1 1 →= − r 0 + t 1/2 = t 1/2 . Eight eighth-of-spheres of radius 1.7.Answers and Hints 1. ¾ −1 − 0 −1 − 0 1−0 ¿ = ¾ 1 ¿ −1 −1 lies on the plane. 1 ¾ Since the line follows the direction of 1 ¿ −1 ¾ 1 ¿ is normal to the plane. Four quarter-cylinders of length a and radius 1.

the equation of the plane ¾ 1/a 1/b 1/c ¿ ¾ • x−1 y−1 z−1 ¿ = ¾ ¿ 0 0 . the shortest distance from Free to photocopy and distribute 195 . − − − − a b b c c a 2 ¾ ¿ a 1. so x = t/a.7 A vector normal to the plane is a2 . Therefore. z t ∈ R. ¾ Thus the vector is 1/a 1/b 1/c ¿ is perpendicular to the plane.7. 3 1.8 Put ax = by = cz = t. The parametric equation of the line is ¾ ¿ ¾ ¿ x 1/a y =t 1/b 1/c .7. ¾ 1. 1. 3) is obtained by the perpendicular line to the plane that pierces the plane.6 We have. The line sought has the same direction as a2 this vector. z = 3 + t. z = t/c.9 The vector 1 ¿ 1 (1. 0 or We may also write this as x y z 1 1 1 + + = + + . thus the equation of the line is ¾ ¿ ¾ ¿ x 0 y z = 0 1 ¾ a a 2 ¿ t ∈ R. ¬¬ ¬¬ ¬¬→ ¬¬2 ¬¬→¬¬2 ¬¬→¬¬2 − ¬¬→ → →¬¬2 ¬¬− ¬¬ → → → − − − → → → − − − → → → → → → − − − − − − − − − − ¬¬ a + b + c ¬¬ = ( a + b + c )•( a + b + c ) = ¬¬ a ¬¬ +¬¬ b ¬¬ +¬¬ c ¬¬ +2( a • b + b • c + c • a ). + t a2 . from where we deduce that 2 2 2 − − →•→ + →•→ + → •→ = −3 − 4 − 5 = −25. 2. −1 is perpendicular to the plane.7.Appendix A Thus the required formula for the volume is abc + 2(ab + bc + ca) + π(a + b + c) + 4π . y = 2 − t. Hence. a b c a b c bcx + cay + abz = ab + bc + ca. y = t/b. this perpendicular line to the plane has equation ¾ ¿ Ǒ Ǒ x 1 1 y z = 2 3 + t −1 1 =⇒ x = 1 + t.7.

3 12 4 3 Free to photocopy and distribute 196 .11 Observe the CBS Inequality in R3 given the vectors → = (x1 . (a2 + b2 + c2 )2 3 S(a. and the distance sought 3 3 3 √ 7 8 2 3 . b. 3 − − 1. x2 . which is clearly impossible. c) 1 ≤ a 2 + b2 + c 2 4 the desired maximum. which proves the claim at once. b. 3) is therefore is 2 7 8 . and so the lines are skew.7. x2 = b2 . c) = = = Hence S(a. (a2 + b2 + c2 )2 a 4 + b4 + c 4 . y3 ) let θ be the angle between them. we don’t need the absolute values) |x1 y1 +x2 y2 +x3 y3 | ≤ x2 + x2 + x2 1 2 3 2 2 2 y1 + y2 + y3 =⇒ (a2 +b2 +c2 )2 ≤ (a4 +b4 +c4 )(3). Then → •→ = ¬¬→ ¬¬¬¬→ ¬¬ cos θ =⇒ |x y + x y + x y | ≤ − − ¬¬− ¬¬¬¬− ¬¬ x y x y 1 1 2 2 3 3 x2 + x2 + x2 1 2 3 2 2 2 y1 + y2 + y3 . there would be a value t′ for which ¾ ¿ ¾ ¿ ¾ ¿ ¾ ¿ Ǒ Ǒ 1 2 0 2 1−0 2−2 1 1 + t′ 1 1 = 0 1 + t′ −1 1 =⇒ 1−0 1−1 = t′ −1 − 1 1−1 ¾ ¿ 1 =⇒ 1 0 = t′ ¾ 0 0 ¿ −2 . From problem (a2 + b2 + c2 )2 and thus maximising f is equivalent to minimising 2 1. 1. 1− which in turn gives √ 2 1 3 = √ = .10 If the lines intersected.12 First observe that S(a. 3 The closest point on the plane to (1. . a 4 + b4 + c 4 1 ≥ . .Answers and Hints The intersection of the line and the plane occurs when 1 1 + t − (2 − t) + (3 + t) = 1 =⇒ t = − . y2 . Then cos θ = Ô 2 · 2 + 1 · (−1) + 1 · 1 4 Ô = √ √ =⇒ θ = arccos (2)2 + (1)2 + (1)2 (2)2 + (−1)2 + (1)2 6 6 2 . x3 = c2 and y1 = y2 = y3 = 1. (1 − )2 + (2 − )2 + (3 − )2 = 3 3 3 3 1. Now take x1 = a2 . 2.11.7. This gives (since squares are positive.7. Let θ be the angle between them. x3 ). c) 1 = a 2 + b2 + c 2 4 1 4 1 4 s(s − a)(s − b)(s − c) (a + b + c)(b + c − a)(c + a − b)(a + b − c) (a2 + b2 + c2 )2 − 2(a4 + b4 + c4 ) Ö 1−2 a 4 + b4 + c 4 . b.7. → = x y (y1 .

b •→ < − . V1 V6 . 3. 0. V0 V2 . ||projn || = ¬¬ ¬¬→ ¬¬2 ¬¬− ¬¬ ¬¬− ¬¬ ¬¬ ¬¬ n n − − − − Since R0 is on the plane.18 There are 7 vertices (V0 = (0. and so r0 n a n r ||projn0 − b || = → − → − → − − → − − → − − − − − − − | → •→ − b •→ | r0 n n |(→ − b )•→ | a n |→•→ − b •→ | a n n ¬¬→ ¬¬ ¬¬→¬¬ ¬¬→¬¬ = . 0.8. V2 V4 . Then bx0 = x0 b distance we seek is ¬¬ ¬¬ → − − − ¬¬ → → •→ ¬¬ − → −→ − − |(→ − b )•→| r0 n r0 n − ¬¬ (− − b ) − → ¬¬ r0 b ¬¬→¬¬ n ¬¬ = . 0)) and 11 edges (V0 V1 . V6 = (4. V4 V5 . 2). V5 = (9.18. Then →• b < − .7.Appendix A 1. V3 V5 . since a norm of vectors is always ≥ 0. x x Free to photocopy and distribute 197 . 0).16 Assume contrariwise that → . 0. y → − − − − − − − − − 1. 8).7. V2 V6 .7.7. → are three unit vectors in R3 such that the a c − − 2π 1 → − 1 − → − − angle between any two of them is > . = − ¬¬| − ¬¬ − ¬¬ ¬¬ n ¬¬ n ¬¬ n as we wanted to shew. V1 V5 .11: Problem 1. Thus 2→ × → = 0 and x x x x y x x x → − − − hence → × → = 0 . 8). z x Figure A. Let x be the point − s t s 0 0 (||s||)2 − → − → − → is orthogonal to the plane. V3 = (0. V0 V3 .17 Let projt = s → •→ − − t s → − − − be the projection of → over → = →. 6 − − → − 1.7. 0). and V4 V6 ). V2 = (0. Thus 2 ¬¬ ¬¬ ¬¬− ¬¬2 ¬¬→¬¬2 ¬¬− ¬¬2 − ¬¬→ → → ¬¬2 ¬¬− ¬¬ − − = ¬¬→ ¬¬ + ¬¬ b ¬¬ + ¬¬→¬¬ a c ¬¬ a + b + c ¬¬ − − → •→ + 2→ •→ + 2→ •→ − − − − +2 a b b c c a < = 0. V1 = (11. 0). V3 V4 . and the − on the plane that is nearest to b. 1+1+1−1−1−1 which is impossible. 1. 1 . 7. 0. b . and → •→ < a c c a 3 2 2 1 − . V4 = (0. 1. → •→ = →•→ .15 x + y + z = 1. 9.2 We have → × → = −→ × → by letting → = → in ??.

4 The vector is normal to the plane.8. The equation of the plane is thus given by ¾ ¿ ¾ ¿ −2a x−a −3a2 a • z−a y−0 = 0.8.6 Either of Free to photocopy and distribute 198 . →×→ − − →×→ − − v w v w → × →|| or − ||→ × → || will do. 1.8. A vector normal to the plane is ¾ ¿ ¾ ¿ ¾ ¿ 2a a −2a −1 a × 0 = 2a a −3a2 .Answers and Hints 1.3 One has → − − → → − − − → − → − − − − − − ( b − → ) × (→ − →) = 0 =⇒ → × b + b × → + → × → = 0 a c a a c c a This gives → → − → − − → → → → − − − − → → − − − − b × − = −(→ × b + → × →) = −( j + k + i − j ) = − i − k .8. c a c a ¾ ¿ ¾ ¿ ¾ ¿ 1 1 −1 1 × 1 1 = 1 0 0 ¿ ¾ ¿ −1 • 1. Now − − − − || v w v w − − − − → × → = (−a→ + a→) × (→ + a→) − − v w j k i j → → − − → → − − → → − − → → − − 2 = −a( j × i ) − a ( j × j ) + a( k × i ) + a2 ( k × j ) → − → − → − = a k + a j − a2 i Ǒ −a2 = a a . that is. 1. as obtained before. The plane has thus equation ¾ x z−2 y+1 1 0 = 0 =⇒ −x + y + 1 = 0 =⇒ x − y = 1. 2ax + 3a2 y − az = a2 .5 The vectors ¾ a − (−a) a−0 0−1 ¿ = ¾ 2a a ¿ −1 and ¾ 0 − (−a) 2a − 0 1−1 ¿ = ¾ a 0 2a ¿ lie on the plane.

c a a c c a → − → → − − → × (→ × b ) = (→ • b ) a − (→ •→ )→ . →.8. x x x x ¬¬− ¬¬2 − − − − → − → − → x and since (→• i )2 +(→• j )2 +(→• k )2 = ¬¬→ ¬¬ = 1.7 From Theorem ?? we have − → × (→ × → ) = (→ •→ )→ − (→ •→ )→ . a c Free to photocopy and distribute 199 . x x x x ¬¬− ¬¬2 ¬¬→¬¬2 → − − − ¬¬− ¬¬ − − → − → ||→ × j ||2 = ¬¬→¬¬ ¬¬ k ¬¬ − (→ • k )2 = 1 − (→• k )2 . x x x x ¬¬− ¬¬2 ¬¬→¬¬2 → − − − ¬¬− ¬¬ − − → − → ||→ × k ||2 = ¬¬→¬¬ ¬¬ j ¬¬ − (→ • j )2 = 1 − (→• j )2 . ¬¬− ¬¬2 ¬¬→¬¬2 → − − − ¬¬− ¬¬ − − → − → ||→ × i ||2 = ¬¬→¬¬ ¬¬ i ¬¬ − (→ • i )2 = 1 − (→• i )2 . the desired sum equals 3− 1 = 2. 1. Hence we may take either v w Ǒ −a2 1 √ 2 a 2a + a4 a or −√ 1 2a2 + a4 −a2 a a Ǒ .8 By Lagrange’s Identity. y = → and → = d we gather that − − − − Now.8.13 First observe that → • (→ × →) = (→ × →) • →. x x a 1.8.12 → − → → → − − − → → − − → = ( a • b ) a + 6 ¬b +¬2 a × c − x ¬→¬ 2 12 + 2¬¬− ¬¬ a − → → → − − − → − → → − → = (a•c)a +6c +3a × b − y ¬¬→¬¬2 18 + 3¬¬− ¬¬ a 1. x x x 1.8. − − − − − − x y z x y z − − − This is so because both sides give the volume of the parallelogram spanned by →. → .Appendix A Ô Ô − − and ||→ × →|| = a4 + a2 + a2 = 2a2 + a4 . − − − − − − c a c c a b and adding yields the result. − − − − − − − − − − − − − a x b b x a a b x a x b b a x b x a a x b x → − − − − The answer is thus {→ : → ∈ R→ × b }. putting x a c z → − → − − − (→ × b ) • (→ × d ) a c = → − → − − − ((→ × b ) × → ) • d .8. x y z → → − − → − → = → × b .9 − − − − − − − − − → ×(→ ×→) = →×(→ ×→ ) ⇐⇒ (→•→ )→−(→ •→)→ = (→ •→)→ −(→•→ )→ ⇐⇒ →•→ = → •→ = 0. 1. − − − − − − − − a b c a c b a b c → − → − − − → − − − − − b × (→ × → ) = ( b •→ )→ − ( b •→ )→ .

**Answers and Hints
**

Now, again, → − − − − − − − − − − → − → − − − → − − → − → − (→ × b ) × → = −→ × (→ × b ) = −((→ • b )→ − (→•→ ) b ) = (→ •→) b − (→ • b )→ . a c c a c a c a c a c a This gives → − → − − − → − − − − − − − − − → − → − − − → → − → − → ((→ × b ) × →) • d = ((→ •→) b − (→ • b )→ ) • d = (→•→ )( b • d ) − (→ • b )(→ • d ), a c c a c a c a c a proving the identity. 1.8.14 By problem 1.8.13, − − − − − − − − − − − − y u (→ × →) • (→ × →) = (→•→ ) • (→•→ ) − (→ •→) • (→ •→). x y u v x u y v x v Using this three times: → − − → − − ( b × →) • (→ × d ) c a → → − − − − (→ × →) • ( b × d ) c a → − → − − − (→ × b ) • (→ × d ) a c = = = → − − − → → − − − − → − → ( b •→) • (→ • d ) − ( b • d ) • (→ • d ) a c c − − − − − → − → − → − → (→ • b ) • (→ • d ) − (→ • d ) • (→ • b ) c a c a → •→ − − → − → •→ − − − − − (→ •→) • ( b d ) − (→ • d ) • ( b c ) a c a

Adding these three equalities, and using the fact that the dot product is commutative, we see that all the terms on the dextral side cancel out and we obtain 0, as required. 1.8.15 1. We have ¾ − → CA =

6 0

¿

, − → CB =

¾

0 4

¿

− → − → → − → − → − → − → − → − → − =⇒ CA×CB = (6 i −3 k )×(4 j −3 k ) = 24 k +18 j +12 i =

¾

12 24

¿

18 .

−3 2. We have

−3

¬¬ ¬¬ Ô √ √ → − ¬¬ → ¬¬− ¬¬CA × CB¬¬ = 122 + 182 + 242 = 1044 = 6 29.

3. The desired line has equation ¾ ¿ Ǒ Ǒ x 0 6 y z = 0 3 +t 0 −3 =⇒ x = 6t, y = 0, z = 3 − 3t.

4. The desired line has equation Ǒ Ǒ x 3 y z = 0 0 +s

¾

0 0

¿

=⇒ x = 3, y = 0, z = −3s.

−3

5. From the preceding items, the line LC A is x = 6t, y = 0, z = 3 − 3t and the line LD E is x = 3, y = 0, z = −3s. If the line intersect then 6t = 3, 0 = 0, 3 − 3t = 1 1 3 −3s gives t = and s = − . The point of intersection is thus 3, 0, . item The 2 2 2 area is ¬¬ ¬¬ √ √ → − ¬¬ → 1 ¬¬− 1 ¬¬CA × CB¬¬ = · 6 29 = 3 29. 2 2

Free to photocopy and distribute

200

Appendix A

6. Observe that 3 P = 0 z

Ǒ

, Q=

3 y 0

Ǒ

, R=

x 3 0

Ǒ

, S =

0 3 z

Ǒ

.

Since all this points lie on the plane 2x + 3y + 4z = 12, we ﬁnd 3 2(3) + 3(0) + 4z = 12 =⇒ P = 0 3 2

Ǒ

,

Ǒ 3

2 0 3Ǒ 2 , 3 0 ,

2(3) + 3y + 4(0) = 12 =⇒ Q =

2x + 3(3) + 4(0) = 12 =⇒ R =

0 2(0) + 3(3) + 4z = 12 =⇒ S =

Ǒ

3 . 3 4 7. A possible way is to decompose the pentagon into three triangles, say △CP Q, △CQR and △CRS and ﬁnd their areas. Another way would be to subtract from the area of △ABC the areas of △AP Q and △RSB. I will follow the second approach. Let [△AP Q], [△RSB] denote the areas of △AP Q and △RSB respectively. Then

¬¬¾ ¿ ¾ ¬¬¾ ¿¬¬ ¿¬¬ ¬¬ 3 ¬¬ −3 ¬¬ 0 ¬¬ ¬¬ ¬¬ ¬¬ ¬¬ ¬¬ ¬¬ → − ¬¬ → 1 ¬¬− 1 ¬¬ 1 2 ¬¬ = ¬¬ 9 ¬¬ = [△AP Q] = ¬¬PA × PQ¬¬ = ¬¬ 0 × ¬¬ ¬¬ ¬¬ 2 2 ¬¬ 3 2 ¬¬ 2 ¬¬ 3 ¬¬ ¬¬ ¬¬ ¬¬ 6 ¬¬ − 2 2 ¬¬¾ ¿¬¬ ¬¬¾ ¿ ¾ ¬¬ 3 ¬¬ ¿¬¬ ¬¬ 3 ¬¬ ¬¬ ¬¬ 0 ¬¬ ¬¬ ¬¬ ¬¬ ¬¬ 4 ¬¬ 2 ¬¬ → − ¬¬ → 1 ¬¬ 1 ¬¬ 9 ¬¬ 1 ¬¬− 1 ¬¬ = ¬¬ [△RSB] = ¬¬SR × SB¬¬ = ¬¬ 0 × ¬¬ = ¬¬ 2 2 ¬¬ 2 ¬¬ 8 ¬¬ 3 ¬¬ ¬¬ 3 ¬¬ 3 ¬¬ − ¬¬ − ¬¬ ¬¬ ¬¬ 4 4 2

Hence the area of the pentagon is √ 3√ 3 √ 33 √ 3 29 − 29 − 29 = 29. 4 16 16

1 2

9+

81 3√ + 36 = 29, 4 4

1 2

9 81 9 3 √ + + = 29. 16 64 4 16

1.9.2 The assertion is trivial for n = 1. Assume its truth for n − 1, that is, assume An−1 = 3n−2 A. Observe that ¾ ¿¾ ¿ ¾ ¿ 1 1 1 1 1 1 3 3 3 A2 = 1 1 1 1 1 1 1 1 1 1 1 1 = 3 3 3 3 3 3 = 3A.

Free to photocopy and distribute

201

**Answers and Hints
**

Now An = AAn−1 = A(3n−2 A) = 3n−2 A2 = 3n−2 3A = 3n−1 A, and so the assertion is proved by induction. 1.9.3 First, we will prove that ¾

1 0 0

2 1 0 ... 0

3 2 1 . . . 0

4 3 2 . . . 0

... ... ... . . . ...

n−1 n−2 n−3 ... 0

n

¿

A =

2

n−1 ... 1

n−2 .

... 0

Observe that A = [aij ], where aij = 1 for i ≤ j and aij = 0 for i > j. Put A2 = [bij ]. Assume ﬁrst that i ≤ j. Then

n j

bij =

k=1

aik akj =

k=i

1 = j − i + 1.

**Assume now that i > j. Then
**

n n

bij =

k=1

aik akj =

k=1

0 = 0,

proving the ﬁrst statement. Now, we will prove that ¾ (n − 1)n 1 3 6 10 . . . 2 (n − 2)(n − 1) 0 1 3 6 ... 2 (n − 3)(n − 2) A3 = 0 0 1 3 ... 2 . . . . . . ... ... . . . ... 0 0 0 0 ... 0

n(n + 1) 2 (n − 1)n 2 (n − 2)(n − 1) 2 ... 1

¿

.

For the second part, you need to know how to sum arithmetic progressions. In our case, we need to know how to sum (assume i ≤ j),

j j

S1 =

k=i

a,

S2 =

k=i

k.

**The ﬁrst sum is trivial: there are j − i + 1 integers in the interval [i; j], and hence
**

j j

S1 =

k=i

a = S1 = a

k=i

1 = a(j − i + 1).

For the second sum, we use Gauß trick: summing the sum forwards is the same as summing the sum backwards, and so, adding the ﬁrst two rows below, S2 S2 2S2 2S2 = = = = i j (i + j) (i + j)(j − i + 1) + + + i+1 (i + j) j−1 + + + i+2 (i + j) j−2 + + + ··· ··· ··· + + + j−1 + + + j i (i + j) ,

···

(i + j)

i−1

Free to photocopy and distribute

202

Appendix A

(i + j)(j − i + 1) . 2

which gives S2 =

**Put now A3 = [cij ]. Assume ﬁrst that i ≤ j. Since A3 = A2 A,
**

n

cij

=

k=1 j

bik akj (k − i + 1)

j j

=

k=i j

=

k=i

k−

i+

k=i k=i

1

= =

(j + i)(j − i + 1) − i(j − i + 1) + (j − i + 1) 2 (j − i + 1)(j − i + 2) . 2

**Assume now that i > j. Then
**

n n

cij =

k=1

bik akj =

k=1

0 = 0.

This ﬁnishes the proof. 1.9.4 For the ﬁrst part, observe that ¾ m(a)m(b) =

1 −a

0 1 0 1

¾

=

0

a a2 − 2 1 0 1 0 0 1 0

¿¾

1 −b 0

0 1 0

−a − b 0

a+b a2 b2 − − + ab 2 2 1 a+b (a + b)2 − 2 1

b b2 − 2 1 ¿

¿

¾

=

1 −(a + b) 0

¿

=

m(a + b)

For the second part, observe that using the preceding part of the problem, ¾ ¿ 1 0 0 2 m(a)m(−a) = m(a − a) = m(0) = −0 1 − 0 = I3 , 2 0 0 1 giving the result. a π 1.11.1 √ , . 2 4

Free to photocopy and distribute

203

**Answers and Hints
**

a 1.11.3 √ 3 1.12.1 Consider a right triangle △ABC rectangle at A with legs of length CA = h and AB = r, as in ﬁgure A.12. The cone is generated when the triangle rotates about CA. The gyrating curve is theÔ hypotenuse, whose centroid is its centre. The length of the generating curve is thus r 2 + h2 and the length of curve described by the centre of Ô r gravity is 2π = πr. The lateral area is thus πr r 2 + h2 . 2

C

r/2

′ h/3 G

G B

A

Figure A.12: Generating a cone.

rh . We need 2 to ﬁnd the centroid of the triangle. But from example 17, we know that the centroid G of the triangle is is two thirds of the way from A to the midpoint of BC. If G′ is the perpendicular projection of G onto [CA], then this means that G′ is at a vertical height h 2 h GG′ h/3 r of · = . By similar triangles = =⇒ GG′ = . Hence, the length of the 2 3 3 r h 3 2 curve described by the centre of gravity of the triangle is πr. The volume of the cone is 3 2 rh π thus πr · = r 2 h. 3 2 3 −− −→ −− −→ − 1.14.1 Find a vector → mutually perpendicular to V1 V2 and V1 V3 and another vector a → − −− −→ −− −→ 1 and a vector b mutually perpendicular to V1 V3 and V1 V4 . Then shew that cos θ = , 3 → − − where θ is the angle between → and b . a Ǒ x To ﬁnd the volume, we gyrate the whole right triangle, whose area is 1.15.1 Let y z ellipse, and its distance to the axis of rotation would be |x| = be a point on S. If this point were on the xz plane, it would be on the 1Ô 1 − z 2 . Anywhere else, 2

Free to photocopy and distribute

204

1. Σ : x2 + y 2 + z 2 = 0. (x2 + y 2 + z 2 )2 − 1 ((x + y + z)2 − (x2 + y 2 + z 2 )) − 1 = 0. Its axis is the line in the direction i + j + k .15. 2 and so we may take A : x + y + z = 0. y z = t 1 . x2 + z 2 = 1 |1 − 4y|. To¾ ¿ its directrix.15. A B A+B thus the equation represents a cylinder. 1. We must have to the axis of rotation is the same as the distance of y z to y 0 .15.15. 1. We thus have Ô 1Ô x2 + y 2 = 1 − z 2. the equation takes the form f (A. B) = 1 1 1 + − − 1 = 0. Ǒ x 1. The directrix has direction → → − − i − k. it would be on the Ô Ô Ô which is to say 1. 1. the 3 Ǒ Ǒ Ǒ x x 0 distance of that is y z x2 + z2. This distance is the same as the length of the segment on the xz-plane going from the z-axis. Anywhere else. be a point on S. and it is thus a cylinder. B) = eA +B − A = 0. and its distance to the axis of rotation would be |x| = |1 − 4y|.3 A spiral staircase.4 A spiral staircase.8 Considering the planes A : x − y = 0. The direction vector is thus 1 Free to photocopy and distribute 205 .6 The planes A : x + z = 0 and B : y = 0 are secant.2 Let y z 1 line. shewing that the surface is → → → − − − of revolution. B : y − z = 0.15.7 Rearranging. If this point were on the xy plane. This gives → → → − − − i + j + k. 3 9x2 + 9z 2 − 16y 2 + 8y − 1 = 0. The surface has equation of 2 2 the form f (A.15. we ﬁnd the intersection of ﬁnd ¾ ¿ x 1 the planes x = y and y = z. 2 or 4x2 + 4y 2 + z 2 = 1.Appendix A x the distance from y z Ǒ to the z-axis is the distance of this point to the point 0 0 z Ǒ : x2 + y 2 .

B : y = 0. we obtain (z − 1)2 − xy = 0.15. 1 1 Free to photocopy and distribute 206 .15. c2 a b When z = 0 we must have 4x2 + y 2 = 1 =⇒ a = Thus Hence. with apex at (0. To shew that circular cross section of radius b actually exist. 1. 4 1. Every plane through the origin which makes a circular cross section must intersect the yz-plane. 1). 2 c c 4 4 4 4 since at z = ±2. 0. one may verify that the two planes given by a2 (b2 − c2 )z 2 = c2 (a2 − b2 )x2 give circular cross sections of radius b. 1. The desired length is ¬4 4 ¬ (2t + 9) dt = (t2 + 9t)¬ = (16 + 36) − (1 + 9) = 42. b = 1. the radius of the circle is at most b. so we may take A : x + y + z = 0.11 The largest circle has radius b. hence we may increase the size of these by moving towards the centre of the ellipse. c2 Ô 4t2 + 36t + 81 dt = (2t + 9) dt. letting z = ±2.1 The arc length element is (dx)2 + (dy)2 + (dz)2 = 1 . z−1z−1 A : x = 0. Therefore. The → → → − − − axis of revolution is then in the direction of i + j + k . C : z = 1.16. Σ : x2 + y 2 + z 2 = 0 as our plane and sphere. (x + y + z)2 − (x2 + y 2 + z 2 ) + 2(x + y + z) + 2 = 0.9 Rearranging.15. The equation is thus 4 3z 2 = 4x2 + y 2 − 1. 4 1 y2 1 1 3 = 4x2 + y 2 − 1 =⇒ 2 = x2 + − = 1− = . 2 + 2 = 1. 2 z2 = 4x2 + y 2 − 1.15.12 Any hyperboloid oriented like the one on the ﬁgure has an equation of the form z2 x2 y2 = 2 + 2 − 1.10 After rearranging. or − Considering the planes x y + 1 = 0. We need t = 1 to t = 4.Answers and Hints 1. 1. Parallel cross sections of the ellipsoid are similar ellipses. we see that our surface is a cone. Arguing b c similarly on the xy-plane shews that the radius of the circle is at least b. and the diameter of any such cross section must be a diameter of z2 y2 the ellipse x = 0. x2 + y2 = 1.

Hence. Then we must have r a t4 t3 t2 +b +c = d =⇒ (at4 +bt3 +ct2 ) = d(1+t2 ) =⇒ at4 +bt3 +(c−d)t2 −d = 0.16. we may parametrise this circle as x = cos t. − 1. t1 t2 t3 + t1 t2 t4 + t1 t3 t4 + t2 t3 t4 1 1 1 1 = 0 =⇒ + + + = 0. t1 t2 t3 t4 t1 t2 t3 t4 The projection of the plane x + y + z = 0 onto the xy-plane is the line y = −x. t1 t2 t3 +t1 t2 t4 +t1 t3 t4 +t2 t3 t4 = 0 =⇒ as required. Hence to be above the plane we need 0 ≤ φ ≤ . 1 + t2 1 + t2 1 + t2 − which means that if →(t) is on the plane ax + by + cz = d. the plane makes 4 4 1 π an angle of with the z-axis. then t must satisfy the quartic r polynomial p(t) = at4 + bt3 + (c − d)t2 − d = 0. so they intersect at the circle x2 +y 2 = 1. On the other hand. 0. then the sum of the roots of p(t) taken three at a time is 0. a2 − 1 a2 + 1 1 −2a ¿ 2. 0. 4 1. Adding the equations we obtain 2z = z + z = (1 + x2 + y 2 ) + (3 − x2 − y 2 ) = 4 =⇒ 2z = 4 =⇒ z = 2. Since 1 is normal to the plane x + y + z = 0. 1). z = 2.Appendix A 1. 2π]. the angle¾ ¿measured from the positive x-axis needs to be in the θ. Recall that φ is measured from φ = 0 (positive z-axis) to 4 3π φ = π (negative z-axis). the tk are coplanar if and only if they are roots of p(t). so they intersect at the plane z = 2.17. z = cos φ.2 Observe that z = 1 + x2 + y 2 is a paraboloid opening up. Since they meet at the circle x2 +y 2 = 1. Subtracting the equations.4 Observe that in this problem you are only parametrising the ellipsoid! The tricky part is to ﬁgure out the bounds in your parameters so that only the part above the plane x + y + z = 0 is described.16.16. y = sin t.6 ¾ 1. z = 3 − x2 − y 2 is another paraboloid opening down. with highest point at (0.3 Let →(t) lie on the plane ax + by + cz = d. with vertex (lowest point) at (0.1 Free to photocopy and distribute 207 . To be “above” this line. 1 π 3π interval − ≤ θ ≤ . 3). 1. A common parametrisation found was: x = cos θ sin φ. π 0 (c2 + 1)dc = 4π 3 1. t ∈ [0. Since the coefﬁcient of t in this polynomial is 0. that is. y = 3 sin θ sin φ. x2 + y 2 = c2 + 1 3. z = 2. (1 + x2 + y 2 ) − (3 − x2 − y 2 ) = z − z = 0 =⇒ 2x2 + 2y 2 − 2 = 0 =⇒ x2 + y 2 = 1.16.

If f ′ (x) = 0 then ex−1 = 1 implying that x = 1. for equality to occur. By the preceding results. . This occurs. Easy Algebra! 3. .17. in view of the preceding lemma. A2 ≤ exp(A2 − 1). which translates into a1 = a2 = . f ′′ (x) = ex−1 . if and only if Ak = 1. 1. This completes the proof. Since all the quantities involved are non-negative. . In view of the observations above. = an . we may multiply all these inequalities together. ¡ (a + b + c + d)2 ≤ (1 + 1 + 1 + 1) a2 + b2 + c2 + d2 Hence. we need each of the inequalities Ak ≤ exp(Ak − 1) to hold. we have A1 ≤ exp(A1 − 1).. to obtain. Now. 5 16 6 The maximum value e = is reached when a = b = c = d = . Easy Algebra! 4. . 1. + xn ) 1 1 1 + +. (a1 + a2 + · · · + an )n We deduce that Gn ≤ which is equivalent to (a1 a2 · · · an )1/n ≤ a1 + a2 + · · · + an . 5 5 Free to photocopy and distribute 208 . Put f : R → R.2 By CBS.. f ′ (x) = ex−1 − 1. = ¡ 4 a 2 + b2 + c 2 + d 2 . . A1 A2 · · · An ≤ exp(A1 + A2 + · · · + An − n). . n a1 + a2 + · · · + an n n An ≤ exp(An − 1). Thus f has a single minimum point at x = 1. Clearly f (1) = e0 − 1 = 0.17. + x1 x2 xn n ≥ = √ i=1 1 xi √ xi 2 n2 . which gives the desired result. . Now. the preceding inequality is equivalent to nn Gn ≤ exp(n − n) = e0 = 1. ∀k. . (x1 + x2 + . ¡ 16 (8 − e)2 ≤ 4 16 − e2 ⇐⇒ e (5e − 16) ≤ 0 ⇐⇒ 0 ≤ e ≤ . Thus for all real numbers x 0 = f (1) ≤ f (x) = ex−1 − x. 2. f (x) = ex−1 − x.Answers and Hints 1.3 By CBS.

. .. a1 = a2 = · · · = a5 = 5 3 a+x 5 from where f (x) ≤ with equality if and only if a+x a−x = . 1+ that is 1+ which means xn+1 > xn . a3 ) 1 2 n for some real number t. 1 n n/(n+1) .17. and thus we may use AM-GM with a−x a+x and a6 = a7 = a8 = . . Hence a1 = a2 = .4 Observe that 96 · 216 = 1442 and by CBS. an ) = t(a3 . 2. a]. . 1. As there is equality. then a + x ≥ 0 and a − x ≥ 0.. 1 + has arithmetic mean 1+ and geometric mean 1+ Therefore. 1 n n > 1+ . 2 1. from where na = 96. . Free to photocopy and distribute 209 .7 Applying AM-GM to the set of n + 1 numbers 1.1 + . n n n 1 n+1 n/(n+1) 1 n . .. 5 3 5 1 +2 +··· + n n+1 = . 1 n+1 1 > n+1 1+ n+1 1 1 1 .17. . a2 . giving the assertion. n n n k=1 a2 ≤ k a3 k k=1 k=1 ak . . a3 . 48 1. = an = a. We deduce that n = 8. .17.5 Applying the AM-GM inequality. .1 + . . n 2 a−x 3 3 5 ≤ a+x 5 +3 8 a−x 3 8 = a 4 8 . . na2 = 144 3 gives a = y n = 32. n: n!1/n = (1 · 2 · · · n)1/n < with strict inequality for n > 1.6 If x ∈ [−a.Appendix A 1. for 1.17. . .. (a1 . 55 33 a8 .

3. kx) would be bounded for every constant k.2 Since polynomials are continuous functions and the image of a connected set is connected for a continuous function.5 2 2. 2. p(x. a semi-inﬁnite interval with no endpoint (take for example p(x. and so the image would just be the point f (0. 2.1 We have → − − − F (→ + h ) − F (→ ) x x = = = → − → − − − − − (→ + h ) × L(→ + h ) − → × L(→ ) x x x x − − → + → ) × (L(→ ) + L(→ )) − → × L(→ ) − − − − (x h x h x x − − − − → × L(→ ) + → × L(→) + → × L(→ ) − − x h h x h h ¬¬ ¬¬ − → − → − → − → − ¬¬→ ¬¬ Now. y) = (xy − 1)2 + x2 whose image is ]0.10 By AM-GM. and by the Cauchy-Bunyakovsky-Schwarz inequality. a semi-inﬁnite interval with an endpoint (take for example p(x. Then e → − L( h ) = n − hk L(→ k ). +∞[). x2 (x2 + y 2 + z 2 )3 (x2 + y 2 + z 2 )2 x2 y 2 z 2 ≤ = →0 2 + z2 2 + y2 + z 2) +y 27(x 27 as (x. e k=1 and hence by the triangle inequality.6 c = 0. the image must be an interval of some sort. 2. The possibilities are thus 1. all real numbers (take for example p(x. 0. a single point (take for example. y) = x2 whose image is [0.7 0 2. y) = 0). e Free to photocopy and distribute 210 . If the image were a ﬁnite interval. 0). For let → − h = n − hk → k .3.1. 4. 3. we will prove that || h × L( h )|| = o ¬¬ h ¬¬ as h → 0 . +∞[).3. 0). e k=1 − where the → k are the standard basis for Rn .3. 2. y) = x).Answers and Hints 2. z) → (0. then f (x.3. y.4.4 0 2. n → − − ||L( h )|| ≤ |hk |||L(→k )|| e n k=1 1/2 n 1/2 ≤ = k=1 |hk |2 ¬¬ ¬¬ − ¬¬→ ¬¬ ¬¬ h ¬¬( n k=1 − ||L(→k )||2 e k=1 − ||L(→k )||2 )1/2 .

¬¬ ¬¬ ¬¬− ¬¬ ¬¬→¬¬2 ¬¬→ ¬¬ → ¬¬→ ¬¬2 − − − ¬¬ h ¬¬ ¬¬− ¬¬ + 2→ • h + ¬¬ h ¬¬ + ¬¬− ¬¬ x x x proving the ﬁrst assertion. L( 0 ) = 0 . and the dextral side → 0 as h → 0 . and so. x x h x x ¬¬ h ¬¬ ¬¬ ¬¬2 − − − → ¬¬→¬¬ ¬¬ ¬¬ 2→ • h + ¬¬ h ¬¬ x → •→ − − − x h ¬¬→ ¬¬ → ¬¬→ ¬¬ + o ¬¬ h ¬¬ . L : Rn → R such that ¬¬ ¬¬ − → − − → → − → − ¬¬→¬¬ ||f ( 0 + h ) − f ( 0 ) − L( h )|| = o ¬¬ h ¬¬ . − ¬¬→ ¬¬ ¬¬ ¬¬ h ¬¬ ¬¬ or → − → − → − → − But the sinistral side → 1 as h → 0 . → − → − Recall that by theorem ??. → − t − 2. → − ¬¬ ¬¬ − L( h ) → − → − → − → − → − ¬¬→ ¬¬ D0 (L)( 0 ) = L( 0 ) = 0 . e as it was to be shewn. This is a → − contradiction. This implies that ¬¬→¬¬ → D0 (L)( 0 ) = 0 . → − → − → − → − → − → − − || h × L( h )|| ≤ || h ||||L( h )| ≤ || h ||2 |||L(→k )||2 )1/2 = o || h || .Appendix A whence.4. We use the fact that (1 + t)1/2 = 1 + + o (t) as t → 0. To prove the second assertion.2 Assume that → = 0 . we have ¬¬ ¬¬2 ¬¬ ¬¬ ¬¬→¬¬2 ¬¬− ¬¬ − − − − ¬¬− ¬¬ + 2→ •→ + ¬¬→¬¬ + ¬¬→ ¬¬ → 2¬¬→ ¬¬. as ¬¬ h ¬¬ → 0. such linear transformation L does not exist at the point 0 . and so by example 169. f ( h ) = ¬¬ h ¬¬ this would imply that ¬¬ ¬¬ − ¬¬→¬¬ as ¬¬ h ¬¬ → 0. We x 2 have ¬¬− ¬¬ → − → − − − − x f (→ + h ) − f (→ ) = ||→ + h || − ¬¬→ ¬¬ x x x ¬¬− ¬¬ → • → → − − − = (→ + h ) (− + h ) − ¬¬→¬¬ x x x ¬¬ ¬¬2 ¬¬ ¬¬ ¬¬→ ¬¬2 − − − − ¬¬− ¬¬ + 2→•→ + ¬¬→ ¬¬ − ¬¬→¬¬ x h = x x ¬¬ h ¬¬ ¬¬ ¬¬2 − − − → ¬¬→ ¬¬ 2→ • h + ¬¬ h ¬¬ x = ¬¬ ¬¬2 ¬¬ ¬¬ . ¬¬− ¬¬ ¬¬ h ¬¬ ¬¬ ¬¬ → − →¬¬ − → − ¬¬ Since f ( 0 ) = ||0|| = 0. Free to photocopy and distribute 211 . assume that there is a linear transformation D0 (f ) = L. ¬¬ ¬¬ ¬¬ ¬¬1 − ¬¬ ¬¬ ¬¬ → ¬¬ − L( h ) ¬¬ ¬¬ ¬¬ ¬¬ = o (1) . ¬¬→ ¬¬2 − − − − ¬¬− ¬¬ + 2→•→ + ¬¬→¬¬ + ¬¬→¬¬ x x h x ¬¬ h ¬¬ → − → − As h → 0 . ¬¬ ¬¬ ¬¬ ¬¬2 − − → − → − ¬¬→ ¬¬ ¬¬→ ¬¬ Since ¬¬ h ¬¬ = o ¬¬ h ¬¬ as h → 0 . again by the Cauchy-Bunyakovsky-Schwarz Inequality. ¬¬¬¬ ¬¬ ¬¬ ¬¬ ¬¬ → ¬¬ − − − ¬¬¬¬→¬¬ ¬¬→ ¬¬ ¬¬¬¬ h ¬¬ − L( h )¬¬ = o ¬¬ h ¬¬ .

1)) = f (3. 1)) = (g ′ (0. 0. g (x. r2 1 ∂ r 2 ∂r 2 2 ∂f 2r 1 = r 2 − rtn e−r /4t = − r 3 tn−1 e−r /4t . 1) = f ′ (g(1.Answers and Hints 2. 0) = ′ ′ æ 0 0 é 0 9 . the output of f is R2 . via the Chain-Rule.5. 0. 1) = (g ′ (f (0.2 Observe that f (x.5. æ 1 0 −1 1 é 2 0 . 1) = æ é 0 1 . + r t 2 4 We conclude that ntn−1 + 2. 0 and hence g (1. 0. y) = ′ æ 1 y −1 x 2 0 é . 0.1 We have 2 ∂f = ntn−1 e−r /4t + tn ∂t r 2 −r 2 /4t e 4t2 = ntn−1 + r 2 tn−2 4 e−r 2 /4t . 1))(f ′ (0. 1))g ′ (1. 2. y) = ′ y2 2xy 2xy x2 é . 1) = æ 0 0 éæ 0 1 9 0 −1 1 2 0 é = æ 0 0 0 9 é 0 0 . 0. f (g(1.4 Since f (0. 1) = ′ r 2 tn−2 3 1 3 = − tn−1 + r 2 tn−2 =⇒ n = − . y) = ∂x and ∂ f (x. ∂r 4t 2 2 ∂f 1 ∂ 1 r2 = − r 3 tn−1 e−r /4t ∂r r 2 ∂r 2 2 1 4 n−1 −r 2 /4t 1 3 r t e = − r 2 tn−1 e−r /4t + r2 2 4t 3 n−1 1 2 n−2 −r 2 /4t = − t e .5. 1)))(f ′ (0. 1)) = 1 0 1 ¿ −1 æ 1 0 1 1 é 0 1 ¾ = 0 0 2 ¿ −1 0 1 Free to photocopy and distribute 212 . the Chain Rule gives ¾ (g◦f )′ (0. 4 2 4 2 x y 2 if x ≤ y 2 if x > y 2 1 0 if x ≤ y 2 if x > y 2 0 2y if x > y 2 if x ≤ y 2 æ f (x.5. 0. The composition g ◦ f is undeﬁned. For. but the input of g is in R3 . (f ◦ g)′ (1. This gives. y) = Hence ∂ f (x.3 Observe that æ é 3 g(1. y) = ∂y 2. 1) = .

∂y and 1 b 2. y) = 2xyg(x2 y).6.1 ∇f (x.6.10 We have ∂ ∂ ∂ ∂z ∂z (x + z)2 + (y + z)2 = 8 =⇒ 2(1 + )(x + z) + 2 (y + z) = 0. the integral is arctan + 2a3 a b 2a2 (a2 + b2 ) 2. 1) = 2e . 1. Observe that ∇f (x.6 Differentiating both sides with respect to the parameter a. z) = x2 + y 2 − 5xy + −7 ¾ xz − yz + 3.5 We have ∂f (x. 1) = ¾ 0 e2 ¿ 2 . z = 4a. ∂x ∂x ∂x 2 ¿ ¾ ¿ e =⇒ (∇f )(2.4 The vector 1 0 ¿ is perpendicular to the plane. Then (∇f )(x.5. 1) the last equation becomes 4(1 + ∂z ∂z 1 ∂z )+4 = 0 =⇒ =− . ∂x ∂x ∂x ∂x ∂x At (1.Appendix A 2.2 (∇ × f )(x. y.6. 2e 0 x yexy ¾ 2. 1.5. ∂x ∂f (x. z) = ey z xzey z yz xye ¾ 2. y. Free to photocopy and distribute 213 . z) = 2x − 5y + z x−y ¿ 2y − 5x − z . y.5. = a −7 Since the point is on the plane x − 7y = −6 =⇒ a − 7a = −6 =⇒ a = 1. Thus x = y = 1 and z = 4. z) is parallel ¾ to the vector 1 0 ¿ −7 . y. y = a. 1. Put f (x. ¾ 2. y) = x2 g(x2 y). and hence there exists a constant a such that ¾ 2x − 5y + z 2y − 5x − z x−y ¿ ¾ 1 0 ¿ =⇒ x = a. y. z) = ¿ =⇒ (∇ × f )(2.

y. y) = −2x sin 2yex cos 2y =⇒ fy (0. y) = (cos 2y)ex cos 2y =⇒ fx (0. fx (x.Answers and Hints 2. ∆2 = det 2 −1 é −1 2 ¾ = 3. 0) = 1. ﬁrst keeping → constant and then keeping → v u u v constant we then see that → − − − − − ∇→ • (→ × → ) = ( ∇ × →) • → . from where we gather that the critical point is a local minimum. u − v v u v u v ¾ ∇f (x. y = − . v u v u 2.6 We have π π with the x-axis and with the y-axis.8. z) = ¿ 2x − y + 1 2y − x 2z − 2 ¾ ¿ 0 = 0 0 1 2 =⇒ x = − .2) ≈ 1 + 0. 2. 2 0 The principal minors are æ ∆1 = 2. 3 3 2 1 whence − . y) ≈ 1 + x. we saw that a c − − → • (→ × →) = (→ × →) • →. z) = −1 0 2 0 . u − v u v u Thus → − − − → − − − − − − − − − − − u v u v v u u v ∇•(u×v) = (∇→ +∇→ )•(u×v) = ∇→ •(→ ×→ )+∇→ •(→ ×→) = ( ∇ ×→)•→ −( ∇ ×→)•→ . fy (x. → . −0. b .6. 0) + fx (0. Hence f (x. y.6.6. whence the hessian is positive deﬁnite at the critical point. 0)(x − 0) + fy (0. Recall that when we deﬁned the volume of a parallelepiped − − → − spanned by the vectors →. where ∇ acts the differential operator and × is the product. 0) = 1.11 An angle of 2. ∆3 = 2 −1 0 −1 2 0 ¿ 0 0 2 = 2 det æ 2 −1 é −1 2 = 6. Free to photocopy and distribute 214 .1. − . 6 3 → − − − − − − − ∇→ • (→ × →) = −∇ • (→ × → ) = −( ∇ × →) • → .1. 0)(y − 0) =⇒ f (x. 0) = 0. − − − − a b c a b c − − − − Treating ∇ = ∇→ + ∇→ as a vector. z = 1.1 = 1. 1 is the only critical point.7 Observe that f (0. This gives f (0. y) ≈ f (0. The hessian is 3 3 ¾ ¿ 2 −1 0 Hf (x.8 This is essentially the product rule: duv = udv +vdu.

0)? We have f (x. 4x3 − 4(x − y) = 0 =⇒ 4x3 − 8x = 0 =⇒ 4x(x2 − 2) = 0 =⇒ x ∈ {− Since x = −y. 4 12y 2 − 4 √ √ √ 2. y) = . so the matrix is negative semideﬁnite and further testing is needed. f (x. the critical points are thus (− 2. If (x. The Hessian is and its principal minors are ∆1 = 12x2 − 4 and ∆2 = (12x2 − 4)(12y 2 − 4) − 16. meaning that there is a saddle point there.8. and 8. √ 2}. ¾ H = 8z − 8 −2 8x −2 0 0 8x 0 ¿ . z) = The hessian is 8xz − 2y − 8x 4x − 2z −2x + 1 2 ¿ = ¾ ¿ 0 0 0 =⇒ x = 1/2. 2.8. −x) = 2x4 − 4x2 = 2x2 (x2 − 1). the matrix is negative deﬁnite and the critical point is thus a saddle point. y) = ( 2. 2. What happens in a neighbourhood of (0. − 2). √ √ 2). z = 1/2. 0) then ∆1 = −4 < 0 and ∆2 = 0.Appendix A 2. and Hr f = ¾ 2 z y z 2 x y 2 ¿ x .9 We have (∇f )(x. 0). y) = (− 2. now. y = −1. −4.7 We have æ (∇f )(x. æ é 12x2 − 4 4 Hf (x. = 2x2 (x2 −1) < 0. y. so the matrix is positive deﬁnite and we have a local minimum at each of these points. −2 The principal minors are 8z − 8. ( 2. 0. which means that the function both increases and decreases to 0 in a neighbourhood of (0. Free to photocopy and distribute 215 . (0. − 2). y.8.8 We have ¾ ∇f (x. then ∆1 = 20 > 0 and ∆2 = 384 > 0. If x is close enough to 0. At z = 1/2. 2) or (x. z) = ¾ 2x + yz 2z + xy ¿ 2y + xz . √ √ √ √ If (x. 0). x) = 2x4 > 0. y) = Hence 4x3 − 4(x − y) 4y 3 + 4(x − y) é = æ é 0 0 =⇒ 4x3 = 4(x − y) = −4y 3 =⇒ x = −y. y) = (0.

∆2 (0. If xyz = −8. 0) is a critical point. and solving for x. −2).10 We have (∇f )(x. y. z) = 8xyz = −x2 y 2 z 2 . z) = 2x ¾ ¿ 0 0 0 xy = −1. −2). 0. 0) = 8 > 0. y. ∆2 (x. all equal 0. y. 0) is a minimum point. Clearly. y = ±1. It is easy to see then that either exactly one of the variables is negative. y. If (∇f )(x. so these points are saddle points. then we must have 2 z z 2 é = 4 − z 2 and ∆3 (x. 2y = −xz. The other critical points are therefore (−2. 0) = 2 > 0. or zy = −4. z) = and Hr f = ¾ 2xy + 2 x2 + 2yz y −1 2 ¿ . z) = 0 0 2x = −yz. say. in which case x = 2. 2. 0). then we must have at least one of the variables equalling 0. (2. 0. 2). Thus (0.Answers and Hints æ We see that ∆1 (x. and substituting this into 2x + yz = 0 we gather (yz)2 = 16. by virtue of the original three equations. then ∆2 (x. If x2 = y 2 = z 2 = 4. we must have 8 x = − . ∆1 (0. and upon multiplication of the three equations. 2. z) = 0. meaning that yz either yz = 4. 0. z) = det det Hr f = −8y 3 . −2. (2. y. ∆1 (0. or all three are negative. 2). 2y 2 0 If (∇f )(x. in which case. z) = 2y. z) = det det Hr f = 8 − 2x2 − ¾ ¿ − 2z 2 + 2xyz. that is. xyz(xyz + 8) = 0. y. z) = 2y 2x 2z é = 4yz − 4x2 and ∆3 (x. y. then we must have Free to photocopy and distribute 216 . and (−2. y. 2. y. 0) = 4 > 0. 2z = −xy. −2. y. We see that ∆1 (x. 0. At (0. x2 = −2yz. 0. ∆2 (x. ∆3 = −32. then none of the variables is 0. if xyz = 0. xyz = −8.8. 0. ¾ 2y 2x 0 2x 2z 2y æ 0 0 ¿ 2y . in which case x = −2. z) = 2. and thus (0.

0. Thus (1. and ∆3 (1. 4xy = 4z 3 =⇒ xyz = x4 = y 4 = z 4 . 0. ¾ Hf (x. 0). 1). 2 2. −1. (−1. (−1. −1). z) = ¿ y(4 − 2x − y − 2z) x(4 − x − 2y − 2z) Free to photocopy and distribute 217 . −1. ) is a saddle point. 1) = −7. Now f (0. 1. (1. z) = 0. 1).8. y. 1. Then 4yz = 4x3 . y. 0) is the only critical point with at least one of the variables 0. z) = xyz(4 − x − y − z) = 4xyz − x2 yz − xy 2 z − xyz 2 . y. x ¾ 4yz − 4x3 4xy − 4z 3 ¿ 4xz − 4y 3 . y = ±1. 1. − ) = 2. and ∆3 = = = < −1728x2 y 2 z 2 + 192x4 + 192z 4 + 128zyx + 192y 4 −1728 + 192 + 192 + 128 + 192 −1024 0. and ∆3 (−1. −1). Since xyz = 1. (−1. (1. (−1. Observe that at these critical points f = 1. −1.Appendix A 1 1 1 and hence (1. 1. −1. −1). Thus (0. −1. Similarly. −1). z) = Assume ∇f (x.8. ¾ ¿ 4yz − 2xyz − y 2 z − yz 2 (∇f )(x. 0. This means that for xyz = 0 the Hessian is negative deﬁnite and the function has a local maximum at each of the four points (1. ∆2 = 144x2 y 2 − 16z 2 = 144 − 16 = 128 > 0. 2. y. Assume now that xyz = 0. The Hessian is Hx f = ¾ −12x2 4z 4y 4z −12y 2 4x 4y 4x −12z 2 ¿ . 1. ) is also a saddle point. 1. Thus we ﬁnd the following critical points: (0. ) = −6. ∆1 (1.12 Rewrite: f (x. −1. we have ∆1 = −12x2 = −12 < 0. − ) are the critical points. Similarly. and (−1. Thus xyz ≥ 0. z = ±1. ) = −2. (1. − ) = −8. 2 2 2 1 1 1 ∆1 (−1.11 We ﬁnd ∇f (x. 1). 4xz = 4y 3 . 1. exactly two of the variables can be negative. −1. Now. 0) = 0 and f (−1. 2 2 2 1 1 1 ∆2 (1. Then. 1). −1. y. shewing that 2 2 2 1 (1. and if one of the variables is 0 so are the other two. − ) = −6. ) = 8. −1. z) = 4xy − x2 y − xy 2 − 2xyz z(4 − 2x − 2y − z) −2yz z(4 − 2x − 2y − z) x(4 − x − 2y − 2z) −2xz y(4 − 2x − y − 2z) −2xy 4xz − x2 z − 2xyz − xz 2 . −1. 1. ∆2 (−1. If 1 = xyz = x2 = y 2 = z 2 . Then (xyz)3 = x4 y 4 z 4 = (xyz)4 =⇒ xyz = 1 =⇒ yz = 1 =⇒ x4 = 1 =⇒ x = ±1. 1. ).

1. which means that in some neighbourhood of (0. z) = (e−x )(e−y )(e−z ). x) > 0 and f (x. 0. 4 − x − y − 2z = 0 =⇒ x = y = z = 1. we will get ∆3 = 0. x. If either of x. y. y = ± √ z = ± √ . x. If xyz = 0 then we must have 4 − 2x − y − z = 0. z) (1 − 2y 2 )(1 − 2x2 )z 2 2 2 2 2 (1 − 2z )(1 − 2x )y (1 − 2z )(1 − 2y )x (4y 3 − 6y)(xz) 2 2 (1 − 2y 2 )(1 − 2z 2 )x . ¾ ¿ −1 −1 −2 and the principal minors are ∆1 = −2 < 0. In this case Hf (1. −x. y. Free to photocopy and distribute 218 . z) = (1 − 2y 2 )(xz)(e−x )(e−y )(e−z ) . −2 −1 −1 −1 −2 −1 xz(4 − x − 2y − z) = 0.Answers and Hints Equating the gradient to zero. 1. which means that (0. z) = 0. we can discard any point with a 0. 0) the function is both decreasing towards 0 and increasing towards 0. (4z 3 − 6z)(xy) with t(x. and ∆3 = −4 < 0. so the matrix is negative deﬁnite and we have a local maximum at (1. Thus as x → 0+ then f (x. We ﬁnd 2 2 2 ¾ ¿ (4x3 − 6x)(yz) (1 − 2x2 )(1 − 2y 2 )z (1 − 2x2 )(1 − 2z 2 )y Hx g = t(x. yz(4 − 2x − y − z) = 0. 1). then x = 1 1 1 ± √ . −x.13 To facilitate differentiation observe that g(x. 1) = 4 − x − 2y − z = 0. (1 − 2z 2 )(xy)(e−x )(e−y )(e−z ) 2 2 2 2 2 2 2 2 2 The function is 0 if any of the variables is 0. f (x. x) < 0. z) = (xe−x )(ye−y )(ze−z ). Since at the critical points we have 1 − 2x2 = 1 − 2y 2 = 1 − 2z 2 = 0. Now. x) = x3 (4 − 3x). y. we obtain. Now ¾ ¿ 2 2 2 (1 − 2x2 )(yz)(e−x )(e−y )(e−z ) ∇g(x. so further testing is needed. ∆2 = 3 > 0. If ∇(x. the Hessian reduces to ¾ ¿ (4x3 − 6x)(yz) 0 0 Hx g = (e−3/2 ) 0 0 We have ∆2 = (4x3 − 6x)(4y 3 − 6y)(xyz 2 ) ∆1 = (4x3 − 6x)(yz) (4y 3 − 6y)(xz) 0 0 . y. f (x. y. (4z − 6z)(xy) 3 ∆3 = (4x3 − 6x)(4y 3 − 6y)(4z 3 − 6z)(x2 y 2 z 2 ). or z is 0. y. Since the function clearly assumes positive and negative values.8. 2. x) = x3 (−4 + x). 0) is a saddle point. xy(4 − x − y − 2z) = 0. 0.

( √ . y) = and so 2g(x2 + y) + 4x2 g ′ (x2 + y) − g ′ (y 2 − x) 2xg ′ (x2 − y) + 2yg ′ (y 2 − x) g ′ (x2 + y) − 2g(y 2 − x) − 4y 2 g ′ (y 2 − x) 2xg ′ (x2 + y) + 2yg ′ (y 2 − x) é . √ Free to photocopy and distribute 219 . − √ . − √ .14 By the Fundamental Theorem of Calculus. − √ .15 Since the coordinates (x. √ . Now. then we the Hessian is negative deﬁnite. ∂y This gives ∂f ∂f (0. 0) is a critical point. 0) is a saddle point. the Hessian of f is æ Hf (x. √ . Geometrically this is easily seen to be 1. (− √ . This means that if an even number of the variables is negative (0 or 2). −2 ≤ y ≤ 2 describe a circle centred at the origin with radius 2. (− √ . − √ ). 4 − y 2 ). the determinant of of this last matrix is −(g ′ (0))2 < 0. − √ ). and if an odd numbers of the variables is positive (1 or 3). Regardless of the sign of g ′ (0). 2. √ .8. We conclude that we have local maxima at 1 1 1 1 1 1 1 1 1 1 1 1 ( √ . y) = G′ (x2 + y) − 2yG′ (y 2 − x) = g(x2 + y) − 2yg(y 2 − x). and the coordinates (y. 0) = æ ′ −g (0) 0 0 g ′ (0) é . the Hessian is positive deﬁnite. 4 1 √ 2 3 −6 1 √ 2 √ = −2 2 < 0. √ ). ( √ . − √ ) 2 2 2 2 2 2 2 2 2 2 2 2 and local minima at 1 1 1 1 1 1 1 1 1 1 1 1 (− √ . Hf (0. 2 2 2 2 2 2 2 2 2 2 2 2 2. ∂x ∂f (x. y) = y 2 −x g(t)dt = G(x2 + y) − G(y 2 − x). Hence ∂f (x. y) = 2xG′ (x2 + y) + G′ (y 2 − x) = 2xg(x2 + y) + g(y 2 − x). the problem reduces to ﬁnding the minimum between the boundaries of the circle and the ellipse. ( √ . and so (0. 144 − 16x2 ). √ ). 0) = 0. √ ). −3 ≤ x ≤ 3 describe an ellipse centred 3 Ô at the origin and semi-axes 3 and 4. 1 4 −√ 2 3 1 − 6 −√ 2 √ = 2 2 > 0. there exists a continuously differentiable function G such that x 2 +y f (x. ∂x ∂y so (0. − √ ). √ .8. (− √ . √ ). 0) = g(0) = (0.Appendix A Also. − √ .

a 2 + b2 + c 2 1+a· Free to photocopy and distribute 220 .2 1.9. the equation of the perpendicular c passing through P is Ǒ Ǒ x 1 y z = 1 1 +t ¾ ¿ a b c =⇒ x = 1 + ta. Hence. y = 1 + tb. Hence. The intersection of the line and the plane happens when a(1 + at) + b(1 + tb) + c(1 + tc) = d =⇒ t = Hence P′ = d − a − b − cá a 2 + b2 + c 2 d−a−b−c 1+b· 2 a + b2 + c 2 d−a−b−c 1+c· 2 a + b2 + c 2 d−a−b−c .Answers and Hints 2. a c+a Therefore c a+c = =⇒ b = c. The vector √ ¾ ¿ a b is perpendicular to Π. abc = 0 and so λ = 0. b+c b = =⇒ a = b.1 We have ¾ =⇒ bc ca ab ¿ ¾ 2b + 2c 2b + 2a ¿ ∇(abc) = λ∇(2ab + 2bc + 2ca − S) = λ 2a + 2c = = = 2λ(b + c) 2λ(a + c) 2λ(b + a) bc =⇒ ca ab By physical considerations. 6 and the maximum volume is √ 3 ( S) abc = √ . c b+c √ S 2a2 + 2a2 + 2a2 = S =⇒ a = √ . z = 1 + tc. 3 3 6 Equality happens if S 2ab = 2bc = 2ca =⇒ a = b = c = √ .9. ( 6)3 The above result can be simply obtained by using the AM-GM inequality: S 2ab + 2bc + 2ca S 3/2 = ≥ ((2ab)(2bc)(2ca))1/3 = 2(abc)2/3 =⇒ abc ≤ 3/2 . b b+a a b+a = =⇒ a = c. by successively dividing the equations. 6 2.

2. 1. Substituting this in the equation of the plane gives the same coordinate of z. y. we gather that λ = 0. 2 z =1+ λc . z) = (x − 1)2 + (y − 1)2 + (z − 1)2 be the square of the distance from P to a point on the plane and let g(x. 1). 2 2 2 2.9. 2 a + b2 + c 2 z = 1+ λc d−a−b−c = 1+c· 2 . Now.9. y.3 Using CBS. 1 61/3 = Ô 6 x2 y 3 z ≤ 2x + 3y + z =⇒ 2x + 3y + z ≥ 62/3 . x =1+ λa . Since (1. z) =⇒ 2(y − 1) 2(z − 1) =λ b c =⇒ 2(x−1) = λa. which is the square of the distance from a point (x. y. Using Lagrange multipliers. 2(z−1) = λc. a 2 + b2 + c 2 Then the coordinates of P ′ are x = 1+ λa d−a−b−c = 1+a· 2 . z) = ax + by + cz − d. as before. 2(y−1) = λb. 2 Putting these into the equation of the plane.4 Using AM-GM. Consider the function t(x. 2 a + b2 + c 2 as before. 2 y =1+ λb .Appendix A The distance is then Ö d−a−b−c a· 2 a + b2 + c 2 2 + b· d−a−b−c a 2 + b2 + c 2 2 + c· d−a−b−c a 2 + b2 + c 2 2 2. 2 a + b2 + c 2 y = 1+ λb d−a−b−c = 1+b· 2 . y) = (x − 1)2 + (y − 1)2 + d − ax − by −1 c 2 . Now. 1. ¾ ¿ æ é a d − ax − by 2(x − 1) − 2 −1 0 c c ∇t(x. y. Let f (x. y) = = b d − ax − by 0 2(y − 1) − 2 −1 c c which implies x= d−a−b−c −b2 − c2 + ab − ad + ac = 1+a· 2 . y. z) = λ∇g(x. 3. 6 Free to photocopy and distribute 221 . a 1+ λa 2 +b 1+ λb 2 +c 1+ λc 2 = d =⇒ λ = 2 · d−a−b−c . √ √ x + 3y x4 + 81y 4 1/4 361/4 ≤( ) = 1/4 =⇒ x + 3y ≤ 23/4 6 = 25/4 3. z) on the plane to the point (1. a 2 + b2 + c 2 a + b2 + c 2 y= c2 + a2 − ab + bd − bc d−a−b−c = 1+b· 2 a 2 + b2 + c 2 a + b2 + c 2 as before. 1) is not on the plane and abc = 0. ¾ ¿ ¾ ¿ 2(x − 1) a ∇f (x.

(1/ 2. 0 = 3x + 5(λ − 1)y. Substituting this into the third equation we gather that √ 1 10x2 ± 6x2 = 8.9. 5 5 5 5 2 8 5 2.9. (p−1)/p + b p p/(p−1) . We then have 1/(p−1) 1/(p−1) a b x= . but this solution does not lie on the third equation. (−1/ 2.7 Let g(x. λp λp Thus 1 = xp + y p = which gives λ= This gives x= (a1/(p−1) a1/(p−1) b1/(p−1) . y) = 5x2 + 6xy + 5y 2 − 8 and argue using Lagrange multipliers. 2). y. these points deﬁne a maximum for f and so ax + by ≤ ap/(p−1) bp/(p−1) + . 0) as long as 25(λ−1)2 −9 = 0. Since f is non-negative. y) =⇒ =λ . 5x2 + 6xy + 5y 2 = 8. (a1/(p−1) + b1/(p−1) )1/p (a1/(p−1) + b1/(p−1) )1/p 2. Aliter: Observe that. We have æ é æ é 2x 10x + 6y ∇f (x. Taking into account the third equa2 √ √ √ √ √ √ √ √ tion. If 25(λ−1)2 −9 = 0. − 2) = 4 and the minimum is f (1/ √ 2.6 Put g(x. z) = (x − 1)2 + (y − 2)2 + (z − 3)2 − 4. the feasible values are ( 2. (− 2. The linear system (the ﬁrst two equations) will have the unique solution (0. resulting in x = ± 2 or x = ± √ . using AM-GM. y= .9. 1/ √ 2) = f (−1/ √ 2. −1/ √ 2) = 1. 2) = f ( 2. 5x2 +6xy +5y 2 = 8 =⇒ x2 +y 2 = 8 6 8 6 x2 + y 2 5 8 − xy ≥ − · =⇒ x2 +y 2 ≥ · = 1. We need a = pλxp−1 and b = pλy p−1 . y) = λ∇g(x. λ = 0. 2y 6x + 10y This gives the three equations 0 = 5(λ − 1)x + 3y. 1/ 2).Answers and Hints 2. We solve ¾ ¿ ¾ ¿ x x ∇f y z = λ∇g y z Free to photocopy and distribute 222 . y) = xp + y p − 1. 1/(p−1) )1/p 1/(p−1) + b1/(p−1) )1/p +b (a a p p/(p−1) a λp p/(p−1) + b λp p/(p−1) . Clearly then . −1/ 2) The desired maximum is thus √ √ √ √ f (− 2.5 We put g(x. then we deduce that x = ±y. y= . − 2).

14 14 14 √ 12 14 The ﬁrst point gives an absolute maximum of 18 + and the second an absolute 7 √ 12 14 . This requires ¾ 2x 2y 2z ¿ = ¾ 2(x − 1)λ 2(z − 3)λ ¿ 2(y − 2)λ . z) = x + z − 1. We may shew that (0. z). y. 2. z) = and 2 4 6 1− √ . λ. 1) yields a minimum.2 + √ .10 One can use Lagrange multipliers here. minimum of 18 − 7 (x. We must ﬁnd λ. z = 1. which translates into 1 = 2λx + δ. y. y.Appendix A for x. z) = 2. y.9 Put g(x.9. y. z) = λ∇g(x. 1) yields a maximum 2. 1 = δ. you will ﬁnd that 2a = 2 and 2b = 6 2. If a and b are the semi-axes. Free to photocopy and distribute 223 . z) = x2 + y 2 − 2. This gives x = −λ −2λ −3λ . y. Substituting into 1−λ 1−λ 1−λ 2 2 2 (x − 1) + (y − 2) + (z − 3) = 4. z) + δ∇h(x. we gather that −λ −1 1−λ from where λ =1± 2 + −2λ −2 1−λ 2 + −3λ −3 1−λ 2 = 4. and x2 + y 2 = 1. √ x + z = 1. y = . Now maximise and minimise the distance between a point on the ellipse and the origin. y = ± 2. √ 2 4 6 1 + √ . λ = 1. √ and that (0. 1 = 2λy. √ 14 . y. But perhaps the easiest approach is to put y = 1 − x and maximise f (x) = x + x(1 − x).2 − √ . 2 This gives the two points (x. h(x. and z = .9. δ such that ∇f (x.8 Observe that the ellipse is symmetric about the origin.9.3 − √ .3 + √ 14 14 14 We deduce that x = 0. − 2. Clearly. y.

so f is doubly-periodic of period π in each argument. y). −2 sin y cos y = −λ − sin 2x = sin 2y = sin 2 x − π 4 = (sin 2x)(cos a a+b a b . y) =⇒ =λ . This means that we may restrict ä ç ä ç ourselves to x ∈ 0. Using Lagrange multipliers. y) = λ∇g(x. Then ∇f (x. If cos 2x = 0 then x = this case f π π 1 . y) = λ(∇g)(x. Since x + y = 1.Answers and Hints For this we have √ 1 1 − 2x 2 f (x) = 0 =⇒ 1 + Ô = 0 =⇒ x = + .9. For this we compute the gradient and set it equal to the zero vector: æ é æ é −axa−1 y b e−(x+y ) 0 (∇f )(x. Hence we maximise f subject to the constraint x + y = 1. Put g(x. æ a−1 b é æ é ax y 1 (∇h)(x.11 Claim: the function achieves its maximum on the boundary of the triangle. π and y ∈ 0. y) = xa y b subject to the constraint x + y = 1. 4(x(1 − x))3/2 x(1 − x) 2. y) = x − y. f = + 2 4 2 2 (1 − 2x)2 1 − Ô < 0. y) =⇒ =⇒ =⇒ −2 sin x cos x = λ.12 Observe that f (x + π. a b−1 bx y 1 which in turn =⇒ axa−1 y b = λ = bxa y b−1 =⇒ ay = bx =⇒ ay = b(1−y) =⇒ y = Finally. y) = xa y b e−(x+y ) ≤ xa y b e−1 ≤ 2. 2 sin x cos x = −2 sin y cos y π π ) − (sin )(cos 2x) = − cos 2 2 2 A similar analysis gives sin 2y = − cos 2y. Since the function is identically 0 for x = 0 or y = 0. 2 4 2 x(1 − x) ′ Since f ′′ (x) = − the value sought is a maximum. In 4 Free to photocopy and distribute 224 . a+b x= a . a+b b a+b b e−1 . 4 4 2 π and so y = 0. y + π) = cos2 (x + π) + cos2 (y + π) = cos2 x + cos2 y = f (x. so the problem reduces to maximising h(x. we can see that f (x. To prove this claim we have to prove that there are no critical points strictly inside the triangle. 0 −bxa y b−1 e−(x+y ) which means that the critical points occur on the boundary. f (x.9. This maximum is thus √ √ 1 2 1 2 + . 0 = cos2 = . we only need to look on the line x + y = 1 for the maxima. y) = xa y b e−(x+y ) = xa y b e−1 on the line. y) = = =⇒ x = 0 or y = 0. π .

y) = (y 2 + 1)x2 + 2xy + 1. Then 2 2 π /2 xdx + ydy C = = 0.Appendix A If cos 2y = 0 then y = π π and so x = . Then xdx + ydy C = L1 1 xdx + ydy + L2 xdx + ydy 1 0 = = 0. L1 xy||dx|| + 1 −1 L2 2 x(x + 1)dx − √ 1 2 0 x(−x + 1)dx ä π πç 2. −π /2 (sin t)(cos t)dt − (cos t)(sin t)dt Also. t ∈ − . thus xy||dx|| = = = √ 0. however.3. both on L1 and on L2 we have ||dx|| = xy||dx|| C √ 2dx. . ||dx|| = (cos t)2 + (− sin t)2 dt = dt. 8 8 = 2 cos2 3π ≈ . and thus π /2 xy||dx|| C = = = (sin t)(cos t)dt ¬ (sin t)2 ¬π /2 ¬ 2 −π /2 0. 3. 2 4 = 1 .1 1. In this case 4 2 f If.2928932190. y = cos t. 8 8 π ≈ 1. 2 π π =⇒ y = − . however.15 Try p(x. L2 : −x + 1.− 8 8 If. We put x = sin t. 8 8 sin 2y = − cos 2y =⇒ tan 2y = −1 =⇒ y = In this case we have f 5π 3π . − cos 2x = 0 then − sin 2x = − cos 2x =⇒ tan 2x = 1 =⇒ x = In this case we have π π . 8 3π 5π =⇒ x = . cos 2y = 0 then f = 2 cos2 π π . −π /2 Free to photocopy and distribute 225 . Let L1 : y = x + 1. −1 xdx(x + 1)dx + xdx − (−x + 1)dx Also.707106781.9. 8 2.

To solve this problem using Maple you may use the code below. and since the integrand is 0.2 Let Γ1 denote the straight line segment path from O to A = (2 3. we ﬁnd on Γ1 that x||dx|| = x whence 2 (dx)2 + (dy)2 = x √ 1+ 1 2 dx = √ xdx. 5 Free to photocopy and distribute 226 . Then on Γ1 3 x x 4 xdx + ydy = xdx + √ d √ = xdx. <2*sqrt(3). Arc(Circle( <0. Pi/5) ). 3. 3 3 3 x||dx|| = Γ1 0 √ 2 √ xdx = 4 3.3 Using the parametrisations from the solution of problem 3.3. 5 6 5 Assembling these we gather that x||dx|| = Γ Γ1 x||dx|| + Γ2 √ π x||dx|| = 4 3 − 8 + 16 sin . 2) and Γ2 denote π the arc of the circle centred at (0.3. 3 3 3 Hence 2 √ 3 xdx + ydy = Γ1 0 4 xdx = 8. 0) and radius 4 going counterclockwise from θ = to 6 π θ= . Pi/6. 3 π On the arc of the circle we may put x = 4 cos θ. 3 On Γ2 that x||dx|| = x whence π /5 (dx)2 + (dy)2 = 16 cos θ sin2 θ + cos2 θdθ = 16 cos θdθ.y> ).0>. Assembling these two pieces.3. xdx + ydy = Γ Γ1 xdx + ydy + Γ2 xdx + ydy = 8 + 0 = 8. x||dx|| = Γ2 π /6 16 cos θdθ = 16 sin π π π − 16 sin = 4 sin − 8.y> ).0>. Line( <0. the integral will be zero. Observe that there 5 xdx + ydy = (cos θ)d cos θ + (sin θ)d sin θ = − sin θ cos θdθ + sin θ cos θdθ = 0.Answers and Hints √ 3. 5 ← → x Observe that the Cartesian equation of the line OA is y = √ . y = 4 sin θ and integrate from θ = 6 π to θ = .2> )) > +LineInt( VectorField( <x. > with(Student[VectorCalculus]): > LineInt( VectorField( <x.3. 4).

0) the positive x-axis is to your left. [x. and not your right). 1. Pi/6.y]=Arc(Circle( <0. we dispose of one of the variables.5. 4 x− 2 We must integrate on the ¿ ¾ side of the plane that can be viewed from the point (1.2> )) +PathInt(x. 2 0 dt 3.y]= Line( <0. To check that these two are indeed 10 5 the same. 4 x − +2z 2 = 2 0 1 is an ellipse. 1. To obtain a positive parametrisation we must integrate from t = 2π to t = 0 (this is because when you look at the ellipse from the point (1. 2 y =1−x = 1 cos t − . use the code > > > is(16*cos(3*Pi/10)=16*sin(Pi/5)). [x. 0) 1 1 2 (observe that the vector 1 is normal to the plane).Appendix A To solve this problem using Maple you may use the code below. Pi/5) ).3 Free to photocopy and distribute 227 . > 3. <2*sqrt(3). and to parametrise this curve. from where y = 1 − x and x2 + y 2 + z 2 = 1 give x2 + (1 − x)2 + z 2 = 1 =⇒ =⇒ =⇒ So we now put x= 1 cos t + . 4). 2 2 sin t z= √ . Thus Á zdx + xdy + ydz Γ = = = 3.5.2 1 3 15π 16 sin t 1 cos t + √ d 2 2 2 2π 0 cos t cos t 1 1 + − + d 2 2 2 2 2π 0 1 cos t sin t + − d √ 2 2 2 2π 0 sin t cos t cos t sin t 1 + √ + − √ 4 4 2 2 2 2 2π π √ .3. y say. which returns true. with(Student[VectorCalculus]): PathInt(x. 3π π Maple gives 16 cos rather than our 16 sin .0>.1 2 3.5.0>. 2 2 2x2 − 2x + z 2 = 0 1 2 1 2 x− + z2 = 2 2 1 2 + 2z 2 = 1.4 The curve lies on the sphere. On the zx-plane.

y = − The integral is 2 0 √ 3 √ 3 x + 4 =⇒ 16 − x2 = 3 − √ 3 x+4 3 2 √ =⇒ x = 0.8 Observe that x2 + y 2 = 16. 2 3.5. √ 16−x 2 − 33 x+4 √ xdydx = = = √ Ô 3 2 + x 16 − x x − 4 dx 3 0 √ ¬2√ 3 1 3 3 ¬ − (16 − x2 )3/2 + x − 2x2 ¬ 3 9 0 8 .Answers and Hints 3. 3.5.5 The integral equals 1 .6 The integral is 1 y 1 1 x2 dxdy + x≤y y ≤x y 2 dxdy = 0 1 0 x2 dxdy + 0 y y 2 dxdy = = = = 21 8 1 2 ¡ y3 dy + y − y 3 dy 3 0 ¬ 0 ¬ y 4 ¬1 y 4 ¬1 y3 − ¬ + ¬ 12 0 3 4 0 1 1 1 + − 12 3 4 1 . 12 0 1 x(x − x4 ) dx 2 π π x sin x sin ydxdy + D D y sin x sin ydxdy = = 2 0 y sin y dy 0 sin x dx 4π.7 3. 6 3.5.5.5.9 e − 1 Free to photocopy and distribute 228 .5. 3 2 √ 3 3.4 The integral equals 1 √ x xydxdy D = 0 1 x x2 y dy dx = = 3.

1]2 xdA + [0. Free to photocopy and distribute 229 . since Maple is quite whimsical about which order of integration to choose. [x. y goes from the line OA to the ← → ← → ← → line OB. 30 xdxdy + 0 ¬y 2 1 ¬ x2 ¬ dy + 1 0 y 2 dxdy ¬1 ¬ y 2 x¬ dy y2 4 y2 (y − y )dy ← → x 3.5. > > with(Student[VectorCalculus]): int(x*y.1>.<3.0>.11 Begin by ﬁnding the Cartesian equations of the various lines: for OA is y = 3 ← → ← → (0 ≤ x ≤ 1).<4. one notices that it does not make much of a difference. ¬x ¬ xy 2 ¬ To solve this problem using Maple you may use the code below.10 We have min(x.y]=Triangle(<0. I will integrate with ← → respect to y ﬁrst. and for 3 ≤ x ≤ 4.5.4>).1]2 y 2 dA y 2 <x x≤y 2 1 y2 1 0 1 0 2 1 = 0 = = = = 1 2 0 0 1 1 4 y dy + 2 0 2 1 + 10 15 7 . It also evaluates expressions that it deems below its dignity to return unevaluated. > int(x*y.0>.<4. y goes from the line AB to the line OB 3 x 4 x xydA R = 0 x/3 3 xydydx + xydydx 3 3x−8 ¬x 4 2¬ = = = = = 1 1 dx + xy ¬ dx 2 0 2 3 x/3 3x−8 3 4 2 ¡ 1 x 1 x x2 − dx + x x2 − (3x − 8)2 dx 2 0 9 2 3 3 4 ¡ 4 1 x3 dx + −8x3 + 48x2 − 64x dx 9 0 2 3 9+9 18. for AB is y = 3x − 8 (3 ≤ x ≤ 4).4>)). Maple can also provide the limits of integration. We have a choice of whether integrating with respect to x or y ﬁrst.Appendix A 3. In such a case notice that for 0 ≤ x ≤ 3.1>. Upon examining the region.1]2 = [0. y 2 )dA [0. [x. and for BO is y = x (0 ≤ x ≤ 4).y]=Triangle(<0.<3. but this command is limited. ’inert’).

**Answers and Hints
**

3.5.12 Integrating by parts,

1 1−x

loge (1 + x + y)dxdy

D

=

0 1 0

loge (1 + x + y) dy

dx

=

0 1

[(1 + x + y) loge (1 + x + y) − (1 + x + y)]1−x dx 0 (2 loge (2) − 1 − loge (1 + x) − x loge (1 + x) + x) dx

= = 1 . 4

0

3.5.13 First observe that on [0; 2]2 , 0 ≤ x + y 2 ≤ 6, so we decompose the region of integration according to where x + y 2 jumps across integer values. We have x + y 2 dA

[0;2]2

[0;2]2 [0;2]2 [0;2]2 [0;2]2 [0;2]2

=

x+y 2 =1

1dA +

x+y 2 =2

2dA +

x+y 2 =3

3dA +

x+y 2 =4

4dA +

x+y 2 =5

5dA

=

[0;2]2 1≤x+y 2 <2

dA + 2

[0;2]2 2≤x+y 2 <3

dA + 3

[0;2]2 3≤x+y 2 <4

dA + 4

[0;2]2 4≤x+y 2 <5

dA + 5

[0;2]2 5≤x+y 2 <6

dA

By looking at the regions (as in ﬁgures A.13 through A.17 below) (I am omitting the details of the integrations, relying on Maple for the evaluations), we obtain

1

√ √

2−x

2

√

0

2−x

dA =

0

[0;2]2 1≤x+y 2 <2

dydx +

1−x 1

dydx = −

4 4√ 2 2 4√ + 2+ =− + 2. 3 3 3 3 3

2

√ √

3−x

2

[0;2]2

dA = 2

0

2−x

√ 8√ 4 dy dx = 4 3 − 2− . 3 3

2≤x+y 2 <3 2

√ √

4−x

3

[0;2]2 3≤x+y 2 <4

dA = 3

0

3−x

√ √ dy dx = 18 − 6 3 − 4 2.

1

2

2

√ √

5

5−x

4

[0;2]2

dA = 4

0

√

dy dx+4

4−x 1

4−x

dy dx = −

√ √ 40 64 16 √ +8 3+ −16 3+ 2. 3 3 3

4≤x+y 2 <5 2

5

[0;2]2

dA = 5

1

√

5−x

dy dx = −

√ 50 + 10 3. 3

4≤x+y 2 <5

**Adding all the above, we obtain x + y 2 dA =
**

[0;2]2

22 4√ 4√ + 3− 2 ≈ 7.7571. 3 3 3

Free to photocopy and distribute

230

2 1 0 2 0 1 0 0 1 2

2 1 0 1 2 0 1 2

2 1 0 0 1 2

2 1Appendix A 0 0 1 2

Figure A.13: 1 ≤ x + y2 < 2.

Figure A.14: 2 ≤ x + y2 < 3.

Figure A.15: 3 ≤ x + y2 < 4.

Figure A.16: 4 ≤ x + y2 < 5.

Figure A.17: 5 ≤ x + y2 < 6.

**3.5.14 Observe that in the rectangle [0 ; 1] × [0 ; 2] we have 0 ≤ x + y ≤ 3. Hence x + y dA
**

R

= =

1

R

1dA + 1dydx +

R

2dA 2dydx

1≤x+y <2 2 2−x 1−x

2≤x+y <3 2 2 1 2−x

=

1

4. √

0 2 1 4−x 2

√ √

4−x 2

2

3.5.15

0 1

xdydx +

1−x 2 2 1 −1 −2

xdydx =

7 . 3

2 3

3

3.5.16

0 −2

xdxdy +

1

xdxdy +

1 0

xdxdy +

1 2

xdxdy = 7.

**3.5.17 Exchanging the order of integration,
**

π /2 0 0 y

cos y dxdy = y

π /2

cos ydy = 1.

0

**3.5.18 Upon splitting the domain of integration, we ﬁnd that the integral equals
**

2 1 y y2

sin

πx dx 2y

2

å

−

2 1

dy

=

1

2y πx cos π 2y

èy 2

dy

y

= = upon integrating by parts.

−

2y πy cos dy π 2 4(π + 2) , π3 −

**3.5.19 The integral is 0. Observe that if (x, y) ∈ D then (−x, y) ∈ D. Also, f (−x, y) = −f (x, y).
**

2/ √ 5 5

1 2

√

4−y 2

3.5.20

−2/

√

1 −2

√

dxdy =

4−y 2

8 + 4 arcsin 5

√

5 5

.

Free to photocopy and distribute

231

**Answers and Hints
**

3.5.21 The integral equals

1

xydA

D

=

0 1 0

1−x 1 + x xy dy

dx

= = = 3.5.22 Using integration by parts,

1 1−x 2 x dx 2 1+x 0 2 2 (t − 1)(t − 2) dt t2 1 11 4 loge 2 − . 4

1

1−x 2

loge (1 + x2 + y)dA

D

=

0 1 0

loge (1 + x2 + y) dy

dx

=

0

(2 loge (2) − 1 − loge (1 + x2 ))dx

1

+

0

(−x2 loge (1 + x2 ) + x2 ) dx

=

2

2 8 π loge 2 + − . 3 9 3

√ √

4−y 2

3.5.23

0

xdxdy = 2.

2y −y 2

**3.5.25 Let D1 = {(x, y) ∈ R2 | − 1 ≤ x ≤ 1, x ≤ y}, D2 = {(x, y) ∈ R2 | − 1 ≤ x ≤ 1, x > y}. Then D = D1 ∪ D2 , D1 ∩ D2 = ∅ and so f (x, y)dxdy =
**

D D1

f (x, y)dxdy +

D2

f (x, y)dxdy.

**By symmetry, f (x, y)dxdy =
**

D1 D2

f (x, y)dxdy,

and so f (x, y)dxdy

D

= = = =

2

D1 1

f (x, y)dxdy

1 x

2

−1 1 −1

(y − x) dy

dx

(1 − 2x + x2 ) dx

8 . 3

3.5.26 The line joining A, and B has equation y = −x − 2, line joining B, and C has equation y = −7x + 10, and line joining A, and C has equation y = 2x + 1. We split

Free to photocopy and distribute

232

Appendix A

the triangle along the vertical line x = 1, and integrate ﬁrst with respect to y. The desired integral is then

1 2x+1

(2x + 3y + 1) dxdy

D

=

−1

(2x + 3y + 1)dy +

1 1 −x−2 2 −7x+10 −x−2

dx dx

(2x + 3y + 1)dy

=

**21 2 3 x + 9x − dx 2 2 −1 2 ¡ 60x2 − 198x + 156 dx +
**

1

= =

4−1 3.

**3.5.27 Since f is positive and decreasing,
**

1 0 0 1

f (x)f (y)(y − x)(f (x) − f (y))dxdy ≥ 0,

**from where the desired inequality follows. 3.5.28 The domain of integration is a triangle. The integral equals
**

1 1−x

xy(x + y)dxdy

D

=

0 0 1

xy(x + y) dy

dx

= = =

å 2 è1−x y y3 x x + dx 2 3 0 0 1 2 x(1 − x) (1 − x)3 x + 2 3 0 1 . 30

1

dx

**3.5.29 For t ∈ [0; 1], ﬁrst argue that
**

1 0 1 1 0 1

0

f (x)dx ≥ (1 − t)f (t) ≥ f (t) − t. Hence

1 1

(f (x)dx) dy ≥

1

0

(f ◦ g)(y)dy −

g(y)dy.

0

Since

0 0

f (x)dxdy =

0

f (x)dx, the desired inequality is established.

**3.5.30 Put f (x, y) = xy + y 2 . If I, II, III, IV stand for the intersection of the region with each quadrant, then f (x, y)dxdy =
**

II II

f (−x, y)d(−x)dy = −

I

f (x, y)dxdy,

f (x, y)dxdy =

IV IV

f (x, −y)dxd(−y) = −

I

f (x, y)dxdy,

Free to photocopy and distribute

233

**Answers and Hints
**

and f (x, y)dxdy =

III III

f (−x, −y)d(−x)d(−y) = +

I

f (x, y)dxdy.

Thus (xy + y 2 )dxdy

S

=

I

f (x, y)dxdy +

II

f (x, y)dxdy +

III

f (x, y)dxdy +

IV

f (x, y)dxdy f (x, y)dxdy

I

=

I

f (x, y)dxdy −

I

f (x, y)dxdy +

I

f (x, y)dxdy −

=

0.

**3.5.31 We split the rectangle [0; a] × [0; b] into two triangles, depending on whether bx < ay or bx ≥ ay. Hence
**

a 0 0 b

emax(b

2

x 2 ,a 2 y 2 )

dydx

=

bx<ay

emax(b e

a2 y 2

2

x 2 ,a 2 y 2 )

dydx +

bx≥ay

emax(b dydx ea

2

2

x 2 ,a 2 y 2 )

dydx

=

dydx +

bx≥ay

2

e

a

b2 x2

bx<ay b ay /b

=

0 b 0

ea

a2y 2

y2

bx/a

dxdy +

a 0 0 a2y 2

y2

dydx

= = =

aye bxe dy + b a 0 0 2 2 2 2 ea b − 1 ea b − 1 + 2ab 2ab a2b2 e −1 . ab

dx

1 3.5.32 Observe that x ≥ (x + y)2 ≥ 0. Hence we may take the positive square root 2 √ √ giving y ≤ 2x − x. Since y ≥ 0, we must have 2x − x ≥ 0 which means that x ≤ 2. The integral equals

2 0 0 √ 2x−x

√

xydy

dx

= = = = =

2 3

2 0

√

2 √ 4 u2 (u 2 − u2 )3/2 du 3 0 1 1 (1 − v 2 )3/2 (1 + v)2 dv 6 −1 π /2 1 cos4 θ(1 + sin2 θ)dθ 6 −π /2 7π . 96

√

√ x( 2x − x)3/2 dx

3.5.33 Observe that

a

´

sin 2πx dx =

0

0 1 (1 − cos 2πa) 2π

if a is an integer if a is not an integer

Free to photocopy and distribute

234

Then 1 1 0 1 I = 0 ··· cos2 0 π (x1 + x2 + · · · + xn ) dx1 dx2 . .5. 48 4 xdxdy D = −1 2/3 0 dy x dx + 2/3 4 0 2− x 2 dy x dx = = 275 .5. dxn = xk dxk k=1 0 = k=1 1 1 = n. we deduce that at least one of the sides of R is an integer.41 The integral equals 1 dxdy (x + y)4 D 3 n . dxn 0 = k=1 1 2 = 3. dxn = k=1 n ··· xk dx1 dx2 . 54 −1 x(x + 1) dx + 2/3 x 2− x 2 dx 3.5.34 We have 1 0 0 1 1 n 1 n ··· 0 (x1 x2 · · · xn )dx1 dx2 . . dxn 2n Free to photocopy and distribute 235 . . . 2 4−x 1 = 1 3 dy dy (x + y)4 dx = = = 3.45 The integral equals 2/3 x+1 4−x 1 − (x + y)−3 dx 3 1 1 3 1 1 1 − dx 3 1 (1 + x)3 64 1 . . . 3.5. ﬁnishing the proof.Appendix A Thus 0 a sin 2πx dx = 0 ⇐⇒ a is an integer.35 This is 1 0 0 1 1 n n 1 0 0 1 1 ··· xk 0 k=1 dx1 dx2 .46 Make the change of variables xk = 1 − yk . Since N sin 2πx sin 2πy dxdy = R k=1 Rk sin 2πx sin 2πy dxdy. .5. . N Now sin 2πx sin 2πy dxdy = 0 k=1 Rk since at least one of the sides of each Rk is an integer. 2 2 3.

du ∧ dv = 2(y 2 + x2 )dx ∧ dy. u = 1. Observe that D ′ 2 2 is the triangle in the uv plane bounded by the lines u = 0. 1]. we have 2I = 1. Clearly also 1 dx ∧ dy = du ∧ dv.6. dyn . v = u. y)dx ∧ dy D (y 4 − x4 ) 1 2 (y − x2 )du ∧ dv 2 v du ∧ dv 2 1 du ∧ dv 2(y 2 + x2 ) = ∆ v du ∧ dv 1 du 2 a b−a .4 Here we argue that du = ydx + xdy.1 u+v u−v and y = . and so I = 3. The integral becomes f (x. Taking the wedge product of differential forms. y)dx ∧ dy = = = The region transforms into ∆ = [a. . 2 Since sin t + cos t = 1. dv = −2xdx + 2ydy. b] × [0.6. v = −u. 4 3. 2n 1 . 4 b 1 = = v dv 0 Free to photocopy and distribute 236 .Answers and Hints equals 1 0 2 0 2 1 1 ··· sin2 0 π (y1 + y2 + · · · + yn ) dy1 dy2 . x + y = 1. Hence f (x. 2 Put x = From the above (x + y)2 ex D 2 −y 2 dA = = = = 1 u2 euv dudv 2 D′ 1 u 1 u2 euv dudv 2 0 −u 1 2 2 1 u(eu − e−u )du 2 0 1 (e + e−1 − 2). . y = 0. Then x + y = u and x − y. Its image under Φ is the triangle bounded by the equations x = 0.

2 √ 1 15 15y − y 3 dy Free to photocopy and distribute 237 .6. (2. (1. 1 0 0 1 1 1 0 dxdy 1 − xy = 0 1 (1 + xy + x2 y 2 + x3 y 3 + · · · )dxdy 1 0 = = = x2 y 3 x3 y 4 xy 2 + + +··· 2 3 4 0 1 2 3 x x x (1 + + + + · · · )dx 2 3 4 0 1 1 1 1 + 2 + 2 + 2 + ··· 2 3 4 y+ dx This change of variables transforms the square [0. 1). The integral becomes 1 0 0 1 dxdy 1 − xy = = 1 2 T 1 ∪T 2 1 1− u −u du ∧ dv u 2 −v 2 4 2 2−u u−2 2 0 dv 4 − u2 + v 2 du + 2 1 dv 4 − u2 + v 2 du.Appendix A 3. (1. −1). 0).5 Formally. 1). 0). We will split this region of integration into two disjoint triangles: T1 with vertices at (0. as desired. Observe that dx ∧ dy = 1 du ∧ dv. 0). 0). (2. u − v = 2y and so 4xy = u2 − v 2 . 1] × [0. This follows by using the identity t 0 dω = arctan t. (1.7.1 The integral in Cartesian coordinates is √ 1 15 1 √ 16−y 2 xydxdy = = 1 2 49 . 1] in the xy plane into the square with vertices at (0. −1) in the uv plane. 2 and that u + v = 2x. 1 + Ω2 This is straightforward but tedious! 3. and T2 with vertices at (1. 1). (1. and (1. −1).

7. √ D cos4 θ(cos2 θ − sin2 θ) dθ π. π /6 1 f (x. π /2 2 cos θ 1 4 arccos 1 4 arccos 1 4 arcsin π 4 π 4 1 cos4 θ sin θ sin θ cos θdθ (cot θ)(csc2 θ)dθ (tan θ)(sec2 θ)dθ 1 4 1 4 π 4 arcsin 1 4 arccos 1 4 π 4 7 7 28 − − 4 4 49 2 x2 − y 2 dxdy D = −π /2 π /2 0 ρ3 dρ 8 0 (cos2 θ − sin2 θ)dθ = = 3. π /4 √ 0 sin 2θ xydxdy = = = = 4 0 ρ π /4 0 Ô ρ2 cos θ sin θ dρ dθ 4 3 √ √ ( sin 2θ)3 cos θ sin θ dθ π /4 4 √ 3 √2 π 2 . 18 9 = = Free to photocopy and distribute 238 .Answers and Hints The integral in polar coordinates is π 4 4 r 3 sin θ cos θdrdθ + 1/ sin θ arccos 1 4 π 4 4 r 3 sin θ cos θdrdθ 1/ cos θ = arcsin 1 4 1 4 π 4 arcsin 1 4 44 − 1 sin4 θ 44 − sin θ cos θ + = 44 4 − − = = 3.4 Using x = aρ cos θ.7.7. y = bρ sin θ.3 Using polar coordinates.2 Using polar coordinates. the integral becomes 2π 1 (ab) 0 a3 cos3 θ + b3 sin3 θdθ 0 ρ4 dρ = 2 (ab)(a3 + b3 ). 12 sin2 2θ dθ 0 3.8 Using polar coordinates. 15 3. y)dA D = 0 π /6 2 sin θ ρ2 dρ dθ 1 (1 − 8 sin3 θ) dθ 3 0 √ π 16 − + 3.7.

7. y) ∈ R2 : y ≥ x. Free to photocopy and distribute 239 . 2ydy = sin θdρ + ρ cos θdθ. x2 + y 2 − y ≤ 0. x2 + y 2 − x ≤ 0}.12 Put D ′ = {(x. whence 4xydx ∧ dy = ρdρ ∧ dθ.11 Using polar coordinates the integral becomes π /4 −π /4 2 cos θ 1/ cos θ 1 dρ ρ3 π /4 dθ = 0 cos2 θ − sec2 θ 4 dθ = π .7. (1 + x2 + y 2 )2 D D2 D1 Using polar coordinates we have dxdy = (1 + x2 + y 2 )2 2π 0 0 1 D2 ρ π dρdθ = (1 + ρ2 )2 2 and dxdy (1 + x2 + y 2 )2 π /2 = = 2 0 +∞ 0 D1 ρ sin2 θdθ dρdθ = 2 )2 (1 + ρ 1 + sin2 θ 0 0 √ dt dt π π 2 − 2 = − . t2 + 1 2t + 1 2 4 sin θ π /2 (We evaluated this last integral using t = tan θ) Finally. Using polar coordinates the integral equals π /2 cos θ 2 π /4 (cos θ + sin θ)2 0 ρ3 dρ dθ = = 1 cos4 θ(1 + 2 sin θ cos θ)dθ 2 π /4 3π 5 − . Then the integral equals 2 D′ (x + y)2 dxdy.9 Using polar coordinates the integral becomes π /2 0 0 2 cos θ ρ4 dρ cos2 θ sin θdθ = 4 .7.13 Observe that D = D2 \D1 where D2 is the disk limited by the equation x2 +y 2 = 1 and D1 is the disk limited by the equation x2 + y 2 = y. 2 2 4 4 3.7. 8 3. the integral equals √ √ π π π 2 π 2 − − = . 64 48 π /2 3.7.14 We have 2xdx = cos θdρ − ρ sin θdθ.Appendix A 3. 5 3. Hence dxdy = (1 + x2 + y 2 )2 dxdy − (1 + x2 + y 2 )2 dxdy .

7. Free to photocopy and distribute 240 . we deduce that Ja → π. First observe that Ja = −a a e−x dx 2 2 . 2 2 The domain of integration of Ja is a square of side 2a centred at the origin.7. This gives the result. To ﬁnd this integral. 60 π /2 1 ρ3 0 Ô 1 − ρ2 dρ 3. 3. whose Calculus book I recommend greatly.15 Using polar coordinates 2π a 0 Ia = 0 ρe−ρ dρ dθ = π(1 − e−a ).16 x2 4≤x 2 +y 2 ≤16 1 dA + xy + y 2 2π 4 2 = 0 2π 4 r drdθ r 2 + r 2 sin θ cos θ = = = = = = so the problem reduces to evaluate I = 1 drdθ r(1 + sin θ cos θ) 2π 4 dθ dr 1 + sin θ cos θ r 0 2 2π dθ log 2 1 + sin θ cos θ 0 2π dθ 2 log 2 2 + sin 2θ 0π dθ 4 log 2 2 + sin 2θ 0 4I log 2. we now 2 + sin 2θ 0 use what has been dubbed as “the world’s sneakiest substitution”1 : we put tan θ = t.Answers and Hints It follows that Ô x3 y 3 1 − x4 − y 4 dx ∧ dy = = 1 2 2 Ô (x y )( 1 − x4 − y 4 )(4xy dx ∧ dy) 4 Ô 1 3 (ρ cos θ sin θ 1 − ρ2 )dρ ∧ dθ 4 Observe that x4 + y 4 ≤ 1 =⇒ ρ2 cos2 θ + ρ2 sin2 θ ≤ 1 =⇒ ρ ≤ 1. Since both Ia and Ia √ 2 tend to π as a → +∞. The respective domains of integration of Ia and Ia √ 2 are the inscribed and the exscribed circles to the square. 1 by Michael Spivak. Hence the integral becomes π /2 0 0 1 Ô 1 3 (ρ cos θ sin θ 1 − ρ2 )dρdθ 4 = = = 1 cos θ sin θdθ 4 0 1 1 2 · · 4 2 15 1 . 0 2 π dθ . Since the integration takes place on the ﬁrst quadrant. we have 0 ≤ θ ≤ π/2.

Parametrise the curve enclosing the region by polar coordinates so that the region is tangent to the polar axis at the origin.Appendix A In so doing we have to pay attention to the fact that θ → tan θ is not continuous on π π [0. π]. Since D(x. π]. The area of the region is then given by 1 2 π ρ2 dθ = 0 1 2 π (f (θ))2 dθ = 0 1 2 π /2 ((f (θ))2 + (f (θ + π/2))2 )dθ.17 Recall from formula 1. π] = [0. the integral above is the integral of the square of the chord in question.7. Let L be a side of R with length l and let S(L ) be the half strip consisting of Free to photocopy and distribute 241 . and so the area π /2 1 < 4dθ = π. Hence 1 + t2 1 + t2 1 + t2 π 0 dθ 2 + sin 2θ π /2 = 0 +∞ dθ + 2 + sin 2θ 2 dt 1+t 2 2t + 1+t 2 π π /2 0 dθ 2 + sin 2θ = 0 +∞ + = = = = = We conclude that dt dt + 2 2(t2 + t + 1) −∞ 2(t + t + 1) +∞ 0 2 dt 2 dt + 2t 1 2 2t 1 3 0 3 −∞ ( √ + √ )2 + 1 ( √3 + √3 ) + 1 3 √ ¬+∞ √ √ √ ¬3 √ √ 0 3¬ 2t 3 3 3¬ 2t 3 3 arctan + + arctan + ¬ ¬ 3 0 3 3 3 −∞ 3 3 √ √ 3 π π 3 π π − + − − 3 6 3 6 2 √ 2 π 3 . 3.7. 3 0 dt 1+t 2 2t −∞ 2 + 1+t 2 0 4π 1 dA = x2 + xy + y 2 4≤x 2 +y 2 ≤16 √ 3 log 2 . y) = 0 inside R. Then 2 2 2t 1 − t2 dt sin 2θ = . dθ = .18 Let I(S) denote the integral sought over a region S. I(R) = A. If no two points are farther than 2 units. their squares are no farther than 4 units. 3 3. [0.14 that the area enclosed by a simple closed curve Γ is given by 1 xdy − ydx. so we split the interval of integration into two pieces. 0 By the Pythagorean Theorem. 2 Γ Using polar coordinates xdy − ydx = = (ρ cos θ)(sin θdρ + ρ cos θdθ) − (ρ sin θ)(cos θdρ − ρ sin θdθ) ρ2 dθ. ]∪] . cos 2θ = . 2 0 a contradiction. Let the equation of the curve be ρ = f (θ).

B. Set up coordinates uv so that u is measured parallel to L and v is measured perpendicular to L. O B. C. 0). The projection of the solid on the xy plane produces the region bounded by the lines x = 0.8. The sum of these integrals over all the sides of R is L. Assembling all these integrals we deduce the result.3 Let A = (1. B = (1. Using the order of integration dzdxdy. The sum of these integrals over all the vertices of R is 2π. x = 1 and x = y on the ﬁrst quadrant of the xy-plane. C = (0. B. 1). O A. so the limits are z = y to z = 1 − x + y. 0). 6 Free to photocopy and distribute 242 .Answers and Hints the points of the plane having a point on L as nearest point of R. Then l +∞ I(S(L )) = 0 0 e−v dudv = l. z sweeps from P2 to P1 . Thus 1 0 0 x y 1−x+y 1 x 0 dzdydx = 0 1 (1 − x)dydx = = = ¡ x − x2 dx 0 ¬ x2 x 3 ¬1 − ¬ 2 3 0 1 . 0. C. 3. 0. 12 3. then using polar coordinates α +∞ I(S(V )) = 0 0 ρe−ρ dρdθ = α. If α is the measure of that angle.8. O x−y+z =1 z=y x=y y = 0.1 We have 1 1−y 0 0 1−z zdV E = 0 1 1−y zdxdzdy z − z 2 dzdy = 0 1 0 = = = (1 − y)2 (1 − y)3 − dy 2 3 0 4 3 ¬1 (1 − y) (1 − y) ¬ − ¬ 12 6 0 1 = . and O = (0. A. the points that have V as nearest from R lie inside an angle S(V ) bounded by the rays from V perpendicular to the edges meeting at V . If V is a vertex of R. We have four planes passing through each triplet of points: P1 : P2 : P3 : P4 : A. 0. 1). 1. C.

z) is on the intersection of the surfaces then √ −1 ± 17 2 z + z = 4 =⇒ z = . We have 1 0 0 x y 1−x+y 1 x 0 xdzdydx = 0 1 (x − x2 )dydx = = = 3. Observe that if (x.8. 0 r2 Spherical: π /2 π /4 0 2π 0 (cos φ)/(sin φ)2 r 2 sin φdrdθdφ. 12 y /3 √ 0 9−x 2 0 xdV = E 0 xdzdydx = 27 . √ √ − 1−y 2 1−y 2 2π 0 x 2 +y 2 √ 1 dzdxdy.Appendix A We use the same limits of integration as in the previous integral.1 Cartesian: 1 −1 π 2 0 1 0 π(x − y) dxdy 2(x2 − y 2 ) π dxdy 2(x + y) ¬1 ¬ π ¬ · ((y + 1) log(y + 1) − (y + 1) − y log y + y)¬ 2 ¬ 0 log(y + 1) − log ydy π log 2. 8 3. 2 Free to photocopy and distribute 243 . we start our integration with the z-variable. y. x 2 +y 2 r Cylindrical: rdzdθdr.8.5 The desired integral is 1 0 0 1 0 ∞ dxdydz (1 + x2 z 2 )(1 + y 2 z 2 ) 1 1 0 0 = 0 1 1 0 1 1 0 = 0 dxdydz ¬z=∞ ¬ 1 ¬ (x arctan(xz) − y arctan(yz)) ¬ dxdy x2 − y 2 ¬ z=0 ∞ 1 x2 − y 2 x2 y2 − 1 + x2 z 2 1 + y2z2 = 0 1 1 = 0 = = = 3. 3 3.9. π The volume is . Since x2 + y 2 ≤ z ≤ 4 − x2 − y 2 .9.2 Ô 1.4 We have 3 ¡ x2 − x3 dx 0 ¬ x3 x4 ¬1 − ¬ 3 4 0 1 .

Free to photocopy and distribute 244 . √ 17 − 1 √ 17 − 1 2 z = r cos φ =⇒ cos φ = =⇒ φ = arccos 2 4 The desired integral √ 2π 0 0 arccos 17 − 1 4 2 r 3 cos θ sin2 φdrdφdθ. It is clear that the limits of the angle θ are from θ = 0 to θ = 2π. 3. Observe that z = x2 +y 2 =⇒ r cos φ = r 2 (cos θ)2 (sin φ)2 +r 2 (sin θ)2 (sin φ)2 =⇒ r ∈ {0. 3. x 2 +y 2 2. we must have z = √ √ 17−1 2 √ 17−1 −x 2 2 √ 4−x 2 −y 2 − √ 17−1 2 − √ 17−1 −x 2 2 xdzdydx. The desired integral 2 2π 0 0 √ 17−1 2 √ r2 4−r 2 r 2 cos θdzdrdθ. The z-limits remain the same as in the Cartesian coordinates. since the integrand is an odd function of x and the domain of integration is symmetric about the origin in x. The projection of the interseccylindrical coordinates. a conclusion that is easily reached. 2 The projection of the circle of intersection of the paraboloid and the sphere onto the xy-plane satisﬁes the equation √ 17 − 1 z 2 + z = 4 =⇒ x2 + y 2 + (x2 + y 2 )2 = 4 =⇒ x2 + y 2 = . The angle φ starts at φ = 0. (csc φ)(cot φ)}.Answers and Hints 17 − 1 only. (csc φ)(cot φ) Perhaps it is easiest to evaluate the integral using cylindrical coordinates.3 Cartesian: √ − 3 3 √ − 3−y 2 3−y 2 1 √ 4−x 2 −y 2 √ √ dzdxdy. Now. 2 Ö√ 17 − 1 a circle of radius . We obtain √ 2π 0 0 17 − 1 2 √ r2 4−r 2 r 2 cos θdzdrdθ = 0. The desired integral is thus 2 Since x + y + z 2 2 2 = 4 =⇒ −2 ≤ z ≤ 2. but translated into Ô 4 − r 2 . and so r 2 ≤ z ≤ tion circle onto the xy-plane is again a circle with centre at the origin and radius Ö√ 17 − 1 .9.

Therefore ex x 2 +y 2 +u 2 +v 2 ≤1 2 +y 2 +u 2 +v 2 π 2π 0 1 0 2 π 0 1 dxdydudv = 0 r 3 er sin φ sin2 tdrdφdθdt 2π π π 2 = 0 r 3 er dr 1 π (2π)(2) 2 2 π 2. 3 2π 2 1 0 2+r cos θ ydV = E 0 r 2 sin θdzdrdθ = 0. 0 dθ 0 sin φdφ 0 sin2 tdt = = Free to photocopy and distribute 245 . u = ρ cos φ sin t.9 We put x = ρ cos θ sin φ sin t.5 We have 5π . dx ∧ dy ∧ du ∧ dv = r 3 sin φ sin2 tdr ∧ dφ ∧ dθ ∧ dt.9.9. v = ρ cos t. y = ρ sin θ sin φ sin t.Appendix A Cylindrical: √ 0 3 0 2π 0 2π 1 2 √ 4−r 2 rdzdθdr.9.8 π 96 π 14 3. r 2 cos2 θ sin2 φ sin2 t + r 2 sin2 θ sin2 φ sin2 t + r 2 cos2 φ sin2 t + r 2 cos2 t r 2 cos2 θ sin2 φ + r 2 sin2 θ sin2 φ + r 2 cos2 φ r 2 cos2 θ + r 2 sin2 θ r2. Spherical: π /3 0 r 2 sin φdrdθdφ. Upon using sin2 a + cos2 a = 1 three times. 3. dx dy du dv = = = = cos θ sin φ sin tdr − ρ sin θ sin φ sin tdθ + ρ cos θ cos φ sin tdφ + ρ cos θ sin φ cos tdt . x2 + y 2 + u2 + v 2 = = = = Now. sin θ sin φ sin tdr + ρ cos θ sin φ sin tdθ + ρ sin θ cos φ sin tdφ + ρ sin θ sin φ cos tdt cos φ sin tdr − ρ sin φ sin tdφ + ρ cos φ cos tdt cos tdr − ρ sin tdt After some calculation. 1/ cos φ The volume is 3.9.7 3.

0 ≤ uv ≤ 1. dz ∧ dx = −2vdu ∧ dv. dy ∧ dz = −du ∧ dv. 0 ≤ uv − uvw ≤ 1 − uvw. 0 1 0 1 0 1 u20 v 18 w 8 (1 − u)4 (1 − v)(1 − w)9 dudvdw. To ﬁnd the limits of integration we observe that the limits of integration using dx ∧ dy ∧ dz are 0 ≤ z ≤ 1. Free to photocopy and distribute 246 . which in turn is 1 0 1 1 u20 (1 − u)4 du 0 v 18 (1 − v)dv 0 w 8 (1 − w)9 dw = 1 1 1 · · 265650 380 437580 which is = 1 . 0 ≤ v ≤ 1. using the fact that for positive integers m. which ﬁnally give 0 ≤ u ≤ 1. Thus 0 ≤ uvw ≤ 1. z = uvw.1 We parametrise the surface by letting x = u. and so Ô ¬¬ 2 ¬¬ ¬¬d x¬¬ = 2 + 4v 2 du ∧ dv. This gives x = u(1 − v). 0 ≤ y ≤ 1 − z. This translates into 0 ≤ uvw ≤ 1. n one has 1 0 2 xm (1 − x)n dx = m!n! . The integral sought is then.9. 0 ≤ x ≤ 1 − y − z. y = uv(1 − w). uv = y + z =⇒ udv + vdu = dy + dz. 0 ≤ u ≤ 1. u v du ∧ dv ∧ dw = dx ∧ dy ∧ dz. z = u + v 2 . uvw = z =⇒ uvdw + uwdv + vwdu = dz. 1] × [0. 0 ≤ u − uv ≤ 1 − uv + uvw − uvw. 44172388260000 3. (m + n + 1)! we deduce.10.10 We make the change of variables u = x + y + z =⇒ du = dx + dy + dz. 0 ≤ w ≤ 1. Observe that dx ∧ dy = du ∧ dv. Observe that the domain D of Σ is the square [0. 2]. y = v.Answers and Hints 3.

cos φ).10. whence u v dx ∧ dy = du ∧ dv. 3. z) = (cos θ sin φ.2 Using x = r cos θ. π] is the longitude. z z and so ¬¬ 2 ¬¬ ¬¬d x¬¬ = = = u2 + v 2 du ∧ dv z2 √ 2 du ∧ dv.4 Put x = u.10. 3 = 3.3 We use spherical coordinates. and so The integral becomes ¬¬ 2 ¬¬ ¬¬d x¬¬ = sin φdφ ∧ dθ. 1+ (dx ∧ dy)2 + (dz ∧ dx)2 + (dy ∧ dz)2 Hence ¬¬ ¬¬ z ¬¬d2 x¬¬ = Σ u 2 +v 2 ≤1 Ô u2 + v 2 √ 2 dudv = √ 2π 1 2 0 0 ρ2 dρdθ = √ 2π 2 . ¬¬ ¬¬ x2 ¬¬d2 x¬¬ Σ 2π π = 0 0 cos2 θ sin3 φdφdθ 4π . 3 0 √ Ô 2 + 4v 2 dudv 2 = = y 0 Ô 2 + 4v 2 dv 3. 3 Free to photocopy and distribute 247 . the surface area is √ 2π 2 2 0 1 rdrdθ = 3π √ 2. 2π] is the latitude and φ ∈ [0. Here θ ∈ [0. dy ∧ dz = cos θ sin2 φdφ ∧ dθ. y = v. zdz = udu + vdv. dy = dv. Then dx = du. sin θ sin φ. (x.10. Observe that dx ∧ dy = sin φ cos φdφ ∧ dθ. 0 ≤ θ ≤ 2π. dz ∧ dx = − sin θ sin2 φdφ ∧ dθ. dz ∧ dx = − du ∧ dv. dy ∧ dz = − du ∧ dv.Appendix A The integral becomes ¬¬ ¬¬ y ¬¬d2 x¬¬ Σ 2 1 = 0 1 0 v du 13 2 . 1 ≤ r ≤ 2. y = r sin θ. y. z 2 = u2 + v 2 .

Free to photocopy and distribute 248 . The area of the part of the surface of the sphere in slice is 2π 0 φ2 φ1 R2 sin φdφdθ = 2πR2 (cos φ1 − cos φ2 ) = 4πR2 /n.10. φ1 ≤ φ ≤ φ2 . and LB C is clearly y = − x + . 2 27 . one for each side of the square. 4 − Σ x2 dzdx = − 3 0 x2 dzdx = − 27 . with R cos φ1 − R cos φ2 = 2R/n. LAC is y = −x.2 We have −4. 3. 2 (x + z)dxdy = Σ 0 (6 − x − 2y)dxdy = 27 . 1 4 LAB is y = x. The region M is the area enclosed by the square.1 Evaluating this directly would result in evaluating four path integrals.10.Answers and Hints 3. We have dω = = = (3x2 ydx + x3 dy) ∧ dx + (ydx + xdy) ∧ dy d(x3 y) ∧ dx + d(xy) ∧ dy (y − x3 )dx ∧ dy. We will use Green’s Theorem. 4 and hence xydydz − x2 dzdx + (x + z)dxdy = Σ 3. 3 3 We have 1 y 2 dx + xdy AB = 0 (x2 + x)dx −2 0 = 2 5 6 − 15 2 14 3 y 2 dx + xdy BC = 1 1 4 − x+ 3 3 − 1 x dx 3 = = y 2 dx + xdy CA = −2 (x2 − x)dx Adding these integrals we ﬁnd Á △ y 2 dx + xdy = −2.6 We project this plane onto the coordinate axes obtaining 6 3−z/2 0 3 xydydz = Σ 0 (3 − y − z/2)ydydz = 6−2x 0 3−y 0 27 . A slice can be parametrised by 0 ≤ θ ≤ 2π.11. This means that each of the n slices has identical area 4πR2 /n. The integral equals Á 2 2 x3 ydx + xydy = (y − x3 )dxdy C 0 0 = 3.11. each slice will have height 2R/n.5 If the egg has radius R.

y = 2 sin t. 0 ≤ t ≤ 2π. Since dx ∧ dy = 0.11.6 Observe that 10 44 − 27 27 −2. Then the integral becomes 2π 2π ((2 + 2 cos t)2 + 16 sin3 t)d2 sin t 0 = 0 (8 cos t + 16 cos2 t +8 cos3 t + 32 cos t sin3 t)dt = 3. which describes an ellipse on the yz-plane. dydz − 2(x−1)2 +z 2 ≤2 √ dzdx = −2π( 2).7 At the intersection path 16π. The integral becomes 2π 0 (1 + cos θ)d(1 − cos θ) + 2π √ √ 2 sin θd(1 + cos θ) + (1 − cos θ)d( 2 sin θ) √ √ √ which in turn = 0 sin θ + sin θ cos θ − 2+ 2 cos θdθ = −2π 2. z = 2 sin θ. 2xdx ∧ dy = {(x−2)2 +y 2 ≤4} −π /2 2ρ2 cos θdρ ∧ dθ = 16π. To do it directly.Appendix A We have 0 (1 − 2y)dx ∧ dy D = −2 1 −x/3+4/3 −x (1 − 2y)dy dx dx + = = 3. (To evaluate the integrals you may resort to the fact that the area of the elliptical region (x − x0 )2 (y − y0 )2 + ≤ 1 is πab).11. We have d (ydx + zdy + xdz) = dy ∧ dx + dz ∧ dy + dx ∧ dz = −dx ∧ dy − dy ∧ dz − dz ∧ dx. 2 a b2 If we were to evaluate this integral directly. we would set √ y = 1 + cos θ. Free to photocopy and distribute 249 . put x − 2 = 2 cos t. Similarly we get 2(x − 1)2 + z 2 = 2 on the xz-plane. by Stokes’ Theorem the integral sought is − 2(y −1)2 +z 2 ≤2 0 = x2 + y 2 + z 2 − 2(x + y) = (2 − y)2 + y 2 + z 2 − 4 = 2y 2 − 4y + z 2 = 2(y − 1)2 + z 2 − 2. x = 2 − y = 1 − cos θ. − 0 x −x/3+4/3 (1 − 2y)dy Hence by the generalised Stokes’ Theorem the integral equals π /2 4 cos θ 0 d(x2 + 2y 3 ) ∧ dy = 2xdx ∧ dy.

A copy made in an otherwise Transparent ﬁle format whose markup. You accept the license if you copy. to use that work under the conditions stated herein. Examples of suitable formats for Transparent copies include plain ASCII without markup. refers to any such manual or work. while not being considered responsible for modiﬁcations made by others. which is a copyleft license designed for free software. the title page itself.2. Such a notice grants a world-wide. Preamble The purpose of this License is to make a manual. SGML or XML for which the DTD and/or processing tools are not generally available. a Secondary Section may not explain any mathematics. philosophical. provided that this License. or of legal.) To “Preserve the Title” of such a section when you modify the Document means that it remains a section “Entitled XYZ” according to this deﬁnition. legibly. and the license notice saying this License applies to the Document are reproduced in all copies. below. which means that derivative works of the document must themselves be free in the same sense. as being those of Invariant Sections. We recommend this License principally for works whose purpose is instruction or reference. Fifth Floor. A “Transparent” copy of the Document means a machine-readable copy. in the notice that says that the Document is released under this License. PostScript or PDF designed for human modiﬁcation. as Front-Cover Texts or Back-Cover Texts. Texinfo input format. in the notice that says that the Document is released under this License. The “Document”. “Endorsements”. The Document may include Warranty Disclaimers next to the notice which states that this License applies to the Document. plus such following pages as are needed to hold. LaTeX input format. It complements the GNU General Public License. November 2002 Copyright c 2000. because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does. The “Title Page” means. unlimited in duration. Inc. Boston. 51 Franklin St. SGML or XML using a publicly available DTD. but only as regards disclaiming warranties: any other implication that these Warranty Disclaimers may have is void and has no effect on the meaning of this License. A copy that is not “Transparent” is called “Opaque”. But this License is not limited to software manuals.2002 Free Software Foundation. textbook. or with modiﬁcations and/or translated into another language. you may accept compensation in exchange for copies. for a printed book. Any member of the public is a licensee. ethical or political position regarding them. but changing it is not allowed. XCF and JPG. and you may publicly display copies. preceding the beginning of the body of the text. and standard-conforming simple HTML.2001. and a Back-Cover Text may be at most 25 words. such as “Acknowledgements”. modify or distribute the work in a way requiring permission under copyright law. under the same conditions stated above. You may also lend copies. If a section does not ﬁt the above deﬁnition of Secondary then it is not allowed to be designated as Invariant. either commercially or noncommercially. A Front-Cover Text may be at most 5 words. “Title Page” means the text near the most prominent appearance of the work’s title. commercial. and that is suitable for input to text formatters or for automatic translation to a variety of formats suitable for input to text formatters. the material this License requires to appear in the title page. However.) The relationship could be a matter of historical connection with the subject or with related matters. that is suitable for revising the document straightforwardly with generic text editors or (for images composed of pixels) generic paint programs or (for drawings) some widely available drawing editor. (Here XYZ stands for a speciﬁc section name mentioned below. and the machine-generated HTML. represented in a format whose speciﬁcation is available to the general public. royalty-free license. 2. If the Document does not identify any Invariant Sections then there are none. The “Invariant Sections” are certain Secondary Sections whose titles are designated. MA 02110-1301 USA Everyone is permitted to copy and distribute verbatim copies of this license document. For works in formats which do not have any title page as such. if the Document is in part a textbook of mathematics. or other functional and useful document “free” in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it. These Warranty Disclaimers are considered to be included by reference in this License. Opaque formats include proprietary formats that can be read and edited only by proprietary word processors. You may not use technical measures to obstruct or control the reading or further copying of the copies you make or distribute. regardless of subject matter or whether it is published as a printed book. in any medium. Secondarily. this License preserves for the author and publisher a way to get credit for their work. 3. (Thus. APPLICABILITY AND DEFINITIONS This License applies to any manual or other work. that contains a notice placed by the copyright holder saying it can be distributed under the terms of this License. with or without modifying it. COPYING IN QUANTITY 250 . The “Cover Texts” are certain short passages of text that are listed. or “History”. it can be used for any textual work. A “Secondary Section” is a named appendix or a front-matter section of the Document that deals exclusively with the relationship of the publishers or authors of the Document to the Document’s overall subject (or to related matters) and contains nothing that could fall directly within that overall subject. An image format is not Transparent if used for any substantial amount of text. has been arranged to thwart or discourage subsequent modiﬁcation by readers is not Transparent. PostScript or PDF produced by some word processors for output purposes only. This License is a kind of “copyleft”. and that you add no other conditions whatsoever to those of this License. A “Modiﬁed Version” of the Document means any work containing the Document or a portion of it. “Dedications”. either commercially or noncommercially. The Document may contain zero Invariant Sections. We have designed this License in order to use it for manuals for free software. and is addressed as “you”. Examples of transparent image formats include PNG.GNU Free Documentation License Version 1. either copied verbatim. 1. A section “Entitled XYZ” means a named subunit of the Document whose title either is precisely XYZ or contains XYZ in parentheses following text that translates XYZ in another language. If you distribute a large enough number of copies you must also follow the conditions in section 3. the copyright notices. or absence of markup. VERBATIM COPYING You may copy and distribute the Document in any medium.

be listed in the History section of the Document). to give them a chance to provide you with an updated version of the Document. if any) a title distinct from that of the Document. If you use the latter option. in parentheses. add their titles to the list of Invariant Sections in the Modiﬁed Version’s license notice. if any. Preserve the network location. J. The author(s) and publisher(s) of the Document do not by this License give permission to use their names for publicity for or to assert or imply endorsement of any Modiﬁed Version. Make the same adjustment to the section titles in the list of Invariant Sections in the license notice of the combined work. If there is no section Entitled “History” in the Document. thus licensing distribution and modiﬁcation of the Modiﬁed Version to whoever possesses a copy of it. together with at least ﬁve of the principal authors of the Document (all of its principal authors. Preserve the section Entitled “History”. Such a section may not be included in the Modiﬁed Version. If you publish or distribute Opaque copies of the Document numbering more than 100. forming one section Entitled “History”. you must do these things in the Modiﬁed Version: A. You must delete all sections Entitled “Endorsements”. It is requested. given in the Document for public access to a Transparent copy of the Document. as authors. create one stating the title. If the required texts for either cover are too voluminous to ﬁt legibly. and from those of previous versions (which should. and Back-Cover Texts on the back cover. In the combination. Preserve any Warranty Disclaimers. if it has fewer than ﬁve). statements of peer review or that the text has been approved by an organization as the authoritative deﬁnition of a standard. The front cover must present the full title with all words of the title equally prominent and visible. and any sections Entitled “Dedications”. new authors. can be treated as verbatim copying in other respects. provided that you include in the combination all of the Invariant Sections of all of the original documents. If the Modiﬁed Version includes new front-matter sections or appendices that qualify as Secondary Sections and contain no material copied from the Document. one or more persons or entities responsible for authorship of the modiﬁcations in the Modiﬁed Version. and that you preserve all their Warranty Disclaimers. If there are multiple Invariant Sections with the same name but different contents. If the Document already includes a cover text for the same cover. and continue the rest onto adjacent pages. M. but not required. immediately after the copyright notices. 4. you must either include a machine-readable Transparent copy along with each Opaque copy. O. likewise combine any sections Entitled “Acknowledgements”. These may be placed in the “History” section. F. Free to photocopy and distribute 251 . and publisher of the Modiﬁed Version as given on the Title Page. in the form shown in the Addendum below. You may add a section Entitled “Endorsements”. with the Modiﬁed Version ﬁlling the role of the Document. Preserve its Title. on explicit permission from the previous publisher that added the old one.Appendix A If you publish printed copies (or copies in media that commonly have printed covers) of the Document. unaltered in their text and in their titles. as long as they preserve the title of the Document and satisfy these conditions. as the publisher. and a passage of up to 25 words as a Back-Cover Text. you must enclose the copies in covers that carry. Include. but you may replace the old one. You may omit a network location for a work that was published at least four years before the Document itself. D. year. numbering more than 100. List on the Title Page. to the end of the list of Cover Texts in the Modiﬁed Version. Delete any section Entitled “Endorsements”. K. Copying with changes limited to the covers. The combined work need only contain one copy of this License. you may at your option designate some or all of these sections as invariant. C. provided that you release the Modiﬁed Version under precisely this License. under the terms deﬁned in section 4 above for modiﬁed versions. when you begin distribution of Opaque copies in quantity. you should put the ﬁrst ones listed (as many as ﬁt reasonably) on the actual cover. you may not add another. I. and multiple identical Invariant Sections may be replaced with a single copy. Preserve all the copyright notices of the Document. the name of the original author or publisher of that section if known. that you contact the authors of the Document well before redistributing any large number of copies. H. if there were any. 5. and preserve in the section all the substance and tone of each of the contributor acknowledgements and/or dedications given therein. Section numbers or the equivalent are not considered part of the section titles. Preserve the Title of the section. Add an appropriate copyright notice for your modiﬁcations adjacent to the other copyright notices. to ensure that this Transparent copy will remain thus accessible at the stated location until at least one year after the last time you distribute an Opaque copy (directly or through your agents or retailers) of that edition to the public. E. State on the Title page the name of the publisher of the Modiﬁed Version. and publisher of the Document as given on its Title Page. Preserve in that license notice the full lists of Invariant Sections and required Cover Texts given in the Document’s license notice. Include an unaltered copy of this License. COMBINING DOCUMENTS You may combine the Document with other documents released under this License. G. make the title of each such section unique by adding at the end of it. MODIFICATIONS You may copy and distribute a Modiﬁed Version of the Document under the conditions of sections 2 and 3 above. authors. For any section Entitled “Acknowledgements” or “Dedications”. In addition. provided it contains nothing but endorsements of your Modiﬁed Version by various parties–for example. To do this. and likewise the network locations given in the Document for previous versions it was based on. then add an item describing the Modiﬁed Version as stated in the previous sentence. or else a unique number. clearly and legibly. Preserve all the Invariant Sections of the Document. and the Document’s license notice requires Cover Texts. Both covers must also clearly and legibly identify you as the publisher of these copies. Use in the Title Page (and on the covers. free of added material. These titles must be distinct from any other section titles. a license notice giving the public permission to use the Modiﬁed Version under the terms of this License. You may add a passage of up to ﬁve words as a Front-Cover Text. N. previously added by you or by arrangement made by the same entity you are acting on behalf of. or if the original publisher of the version it refers to gives permission. Do not retitle any existing section to be Entitled “Endorsements” or to conﬂict in title with any Invariant Section. and list them all as Invariant Sections of your combined work in its license notice. L. or state in or with each Opaque copy a computer-network location from which the general network-using public has access to download using public-standard network protocols a complete Transparent copy of the Document. unless they release you from this requirement. You may add other material on the covers in addition. all these Cover Texts: Front-Cover Texts on the front cover. you must combine any sections Entitled “History” in the various original documents. B. and add to it an item stating at least the title. year. unmodiﬁed. Only one passage of Front-Cover Text and one of Back-Cover Text may be added by (or through arrangements made by) any one entity. You may use the same title as a previous version if the original publisher of that version gives permission. you must take reasonably prudent steps.

10. If the Document speciﬁes that a particular numbered version of this License “or any later version” applies to it. and will automatically terminate your rights under this License. sublicense or distribute the Document is void. provided that you follow the rules of this License for verbatim copying of each of the documents in all other respects. 9. so you may distribute translations of the Document under the terms of section 4. modify. Replacing Invariant Sections with translations requires special permission from their copyright holders. parties who have received copies. is called an “aggregate” if the copyright resulting from the compilation is not used to limit the legal rights of the compilation’s users beyond what the individual works permit. or “History”. provided you insert a copy of this License into the extracted document. in or on a volume of a storage or distribution medium. revised versions of the GNU Free Documentation License from time to time. or distribute the Document except as expressly provided for under this License. but may differ in detail to address new problems or concerns. Que ´ languidezca de dolor gritando por piedad. Such new versions will be similar in spirit to the present version. y que no haya coto a su agon´ a hasta la ultima disolucion. the Document’s Cover Texts may be placed on covers that bracket the Document within the aggregate. “Dedications”. but you may include translations of some or all Invariant Sections in addition to the original versions of these Invariant Sections. que arda en las n llamas del Inﬁerno para siempre. and all the license notices in the Document. If the Document does not specify a version number of this License.gnu. Free to photocopy and distribute 252 . and replace the individual copies of this License in the various documents with a single copy that is included in the collection. you may choose any version ever published (not as a draft) by the Free Software Foundation. TERMINATION You may not copy. se le convierta en una serpiente en las manos y lo venza. cuando llegue al ﬁnal de su castigo. sublicense. and any Warranty Disclaimers. Any other attempt to copy. Que a quien robe este libro. FUTURE REVISIONS OF THIS LICENSE The Free Software Foundation may publish new. When the Document is included in an aggregate. Otherwise they must appear on printed covers that bracket the whole aggregate. Each version of the License is given a distinguishing version number. or the electronic equivalent of covers if the Document is in electronic form. or rights. then if the Document is less than one half of the entire aggregate. and follow this License in all other respects regarding verbatim copying of that document. the original version will prevail. o lo tome prestado y no lo devuelva. However. You may extract a single document from such a collection. If a section in the Document is Entitled “Acknowledgements”. and distribute it individually under this License. Barcelona. See http://www. TRANSLATION Translation is considered a kind of modiﬁcation. from you under this License will not have their licenses terminated so long as such parties remain in full compliance. this License does not apply to the other works in the aggregate which are not themselves derivative works of the Document. You may include a translation of this License.Answers and Hints 6. -Maldicion anonima contra los ladrones de libros en el monasterio de ´ ´ San Pedro. 8. Que sea golpeado por la paralisis y todos sus miembros arruinados. AGGREGATION WITH INDEPENDENT WORKS A compilation of the Document or its derivatives with other separate and independent documents or works. COLLECTIONS OF DOCUMENTS You may make a collection consisting of the Document and other documents released under this License. 7. provided that you also include the original English version of this License and the original versions of those notices and disclaimers. modify. Que las polillas roan sus ı ´ ´ entra˜ as y. If the Cover Text requirement of section 3 is applicable to these copies of the Document.org/copyleft/. you have the option of following the terms and conditions either of that speciﬁed version or of any later version that has been published (not as a draft) by the Free Software Foundation. the requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual title. In case of a disagreement between the translation and the original version of this License or a notice or disclaimer.

Edmund ROBERTSON. MARSDEN and Anthony TROMBA. Sean DINEEN Multivariate Calculus and Geometry ´ [MarTro] Jerrold E. Analysis on Manifolds. Isaac TODHUNTER. Advanced Calculus. [Lan] [Spik] [Edw] [Mun] [Tod] Serge LANG. Linear Algebra.Bibliography [BlRo] [Mun] [Buck] [Din] Thomas BLYTH. Introduction to Linear Algebra. MUNKRES. James R. Analysis on Manifolds. James R. EDWARDS. A Uniﬁed Approach. MUNKRES. Michael SPIVAK. Vector Calculus [HubHub] John H. HUBBARD and Barbara Burke HUBBARD. Spherical Trigonometry.C. Advanced Calculus: A Differential Forms Approach. and Differential Forms. BUCK. Vector Calculus. R. Basic Linear Algebra. Calculus on Manifolds: A Modern Approach to Classical Theorems of Advanced Calculus. Harold M. 253 .

- Hard calculus problems in university_ do viet hung
- Vector calculus
- 05_ismchap5
- calculus
- Mathematics Mar 2009 Eng
- Calculus
- XVII – Trigonometric Functions of Real Numbers
- C++ Programing Fundamentals
- aime 2015 2
- Calculus
- Maths Problems
- Vector Calculus Supplement
- Latex Cheatsheet
- Latex Quickguide
- 67TRIG~1_4
- 38970239-formulaes
- Special Paper 1992-95
- 21.03.09
- crmosol-12-2 Previous year Question Papers of Regional Mathematical Olympiad with solutions
- M0IITU01 - Quadratic Equations Qns
- h6
- 2012
- Lecture 08
- ABFormulas
- Problems 1
- Mathematics Test
- MC0063(a) Unit1 Final
- Mouse Catia
- Tut02 Mouse
- Tut02 e Mouse

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd