Math 1 All Answers

Examination papers and
Examiners’ reports
Mathematics 1
279005a, 990005a, 996D05a
2003, 2004
Undergraduate study in
Economics, Management,
Finance and the Social Sciences
05a Mathematics 1
Examiner’s report 2004

Zone B
Exam technique: general remarks
We start by emphasising that candidates should always include their working. This
means two things. First, they should not simply write down the answer in the exam
script, but explain the method by which it is obtained. Secondly, they should include
rough working. The examiners want you to get the right answers, of course, but it is
more important that you prove you know what you are doing: that is what is really
being examined.
We also stress that if a student has not completely solved a problem, they may still be
awarded marks for a partial, incomplete, or slightly wrong, solution; but, if they have
written down a wrong answer and nothing else, no marks can be awarded.
Candidates should ensure that they have covered the bulk of the course in order to
perform well in the examination: it is bad practice to concentrate only on a small
range of major topics in the expectation that there will be lots of marks obtainable
for questions on these topics. There are no formal options in this course: all students
should study the full extent of the topics described in the syllabus and subject guide. In
particular, since the whole syllabus is examinable, any topic could potentially appear
in Section A.
Students are reminded that calculators are not permitted in the examination for this
subject, under any circumstances. It is a good idea to prepare for this by attempting
not to use your calculator as you study and revise this subject.
Specific comments on questions
1. It should be realised that the curves are parabolas, and their intersections with the
axes should be indicated. Finding the intersection point involves solving the equation
x2 + 4x + 1 = −x2 − x + 6 which, rewritten as a quadratic equation in standard form,
becomes 2x2 + 5x √ − 5 = 0, and this can be solved in the usual ways. The positive
solution is x = ( 65 − 5)/4.
2. We find f (x) = e−x/2 − (1/2)(1 + x)e−x/2 = (1/2)e−x/2 (1 − x). So f (x) = 0 means

x = 1. The nature of the critical point can be checked by examining the behaviour
of the derivative around each point or by computing the second derivative at each
point. It is found that x = 1 gives a maximum and therefore the maximum value is
f (1) = 2e−1/2 . (Here, we are also using the fact that as x gets larger and larger, f (x)
2
Examination papers and Examiners’ reports 2004
approaches 0. Many forgot to calculate the maximum value. The question asks for the
maximum value, not simply the value of x that gives the maximum value.
3. This integral can be solved using the partial fractions method. We find that
x 5/7 2/7
= +
x2 − 3x − 10 x−5 x+2
and so the integral is
5 2
ln |x − 5| + ln |x + 2| + c.
7 7
4. We have total revenue given, as a function of x, y, by
T R = xpX + ypY = x(100 − x/2) + y(100 − y),
and the profit is

3
Π(x, y) = T R − T C = − x2 − 2y 2 + 100x + 100y − 24y − 20.
2
Solving ∂Π/∂x = 0 and ∂Π/∂y = 0 simultaneously gives x = 25 and y = 25/2. The
second order conditions are easily checked to show that these quantities do indeed
maximise profit. Now, for the second part of the question, there is no need to use the
Lagrange multiplier method. If twice as much Y as X must be produced, then y = 2x
and the profit, as a function of x is
27 2
Π(x, 2x) = − x + 300x − 20.
2
This single-variable function is easily seen (in the usual way, by setting its derivative
equal to 0 and checking the second order condition) to be maximised when x = 100/9.
So the answer to the second part is x = 100/9, y = 2x = 200/9.
√
5. The Lagrangean is L = xy 2 − λ(x + y − 100) and the first-order conditions are
∂L 1
= x−1/2 y 2 − λ = 0
∂x 2
∂L √
= 2 xy − λ = 0
∂y
x + y = 100.
From the first two equations,

√
λ = (1/2)x−1/2 y 2 = 2 xy,
which on simplification becomes y = 4x. Then the third equation shows that x =
20, y = 80. There is no need to do a second-order test for the nature of the point: that
is not part of the syllabus and there is no credit for it.
3
05a Mathematics 1
6. The system expressed in matrix form (a step some forgot) is

    
2 1 −1 x 4
    
 1 3 −1   y  =  4  .
1 5 1 z 8
(The system ‘expressed in matrix form’ is not the augmented matrix.) The standard
matrix method approach is now to reduce the augmented matrix to reduced form. Here
is one way. (There are others, equally valid.)
     
2 1 −1 4 1 3 −1 4 1 3 −1 4
     
 1 3 −1 4  →  2 1 −1 4  →  0 −5 1 −4 
1 5 1 8 1 5 1 8 0 2 2 4
   
1 3 −1 4 1 3 −1 4

→ 0 1 1 2  
→ 0 1 1 2 

0 −5 1 −4 0 0 6 6
It follows, from this last matrix, that
x + 3y − z = 4
y+z = 2
6z = 6.
So, z = 1/2, y = 2 − z = 1 and x = 4 + z − 3y = 2. Simply manipulating equations

does not constitute a matrix method, which the question specifically asks for, and gets
no credit. Some candidates used Cramer’s rule. This is a technique not discussed in
the Subject Guide, and the computations required are more difficult. We recommend
the reduction technique, but Cramer’s rule, carefully applied, gives the correct answer.
7. For some a and d, the nth term of the sequence takes the form a + (n − 1)d. The
sum of the first n terms is (n/2)(2a + (n − 1)d). So we must have
13
a + 4d = 4, (2a + 12d) = 65.
2
So we have the simultaneous equations
a + 4d = 4, a + 6d = 5.
The solutions are d = 1/2 and a = 2.
8. (a) We have T C = V C + F C = q(AV C) + F C = q 3 + q 2 + eq + 10 and

d
MC = T C = 3q 2 + 2q + eq .
dq
(b) The first integral ismost easily done using a substitution. If we set u = x2 + 2 the
integral becomes (1/2) (u − 2)u1/2 du, which is
1 3/2 1 2 1 2
u − 2u1/2 du = u5/2 − u3/2 + c = (x2 + 2)5/2 − (x2 + 2)3/2 + c.
2 5 3 5 3
4
√
The substitution u = x2 + 2 also works. So does integration by parts,
√ though
this is
harder. If using parts, one has to think of the integrand as x2 × x x2 + 2 . (Some
candidates attempted to use integration by parts, but only succeeded in ending up with
a more complicated integral. No credit was given for this. Integration by parts is only
useful if the resulting integral is easier than the one we start with.) For the second
integral, we can use integration by parts, twice, as follows:

x−2 e−x dx = −x2 e−x + 2xe−x dx = −x2 e−x − 2xe−x − 2e−x + c.
(c) This was not well done. It is useful to take logarithms and then differentiate. We
2
have ln f = ln(2x y ) = x2 y ln 2 and so, differentiating each side with respect to x,
1 ∂f
= 2xy ln 2,
f ∂x
so
∂f 2
= (2xy ln 2)f = (2xy ln 2)2x y .
∂x
Similarly, we find that
∂f 2
= (x2 ln 2)f = (x2 ln 2)2x y .
∂y
9. (a) The total cost is T C = q(AC) = 9q+(3/10)q 2 +30. Since the firm is a monopoly,
we can find the price in terms of q from the demand equation. Hence the total revenue
is T R = pq = (30 − (3/4)q)q and profit is
21 2
Π(q) = T R − T C = 21q − q − 30.
20
To maximise profit, we solve Π (q) = 0, which gives q = 10. We can see also that
Π = −21/10 < 0, so q = 10 maximises profit.
(b) Writing
g(x, y) = x2 − 2xy + 2y 2 − 2y + 2,
we have f (x, y) = ln(g(x, y)).Thus
∂f 1 ∂g
= ,
∂x g ∂x
∂f 1 ∂g
= .
∂y g ∂y
The stationary point of g(x, y) is found to be given by x = y = 1 by solving
∂g
= 2x − 2y = 0,
∂x
∂g
= −2x + 4y − 2 = 0.
∂y
5
05a Mathematics 1
Now we need to show that this point is a local minimum. One possible way to do this
is to consider g. We have (writing gxx for ∂ 2 g/∂x2 and so on), gxx = 2, gxy = −2
2
and gyy = 4, so, since gxx > 0 and gxx gyy − gxy > 0, g(x, y) has a local minimum at
(1, 1). That is to say, g(1, 1) ≤ g(x, y) holds locally; from this it then follows that
f (1, 1) ≤ f (x, y) holds locally and hence (1, 1) is a local minimum of f . Alternatively,
we may compute directly the second derivatives of f . We find that
2 (2x − 2y)2 −2 (2x − 2y)(−2x + 4y − 2)
fxx = − , fxy = − ,
g g2 g g2
4 (−2x + 4y − 2)2
−
fyy = .
g g2
Now—and this is very important—we calculate these derivatives at the critical point.
Putting x = 1, y = 1, the derivatives are
fxx (1, 1) = 2, fxy (1, 1) = −2, fyy (1, 1) = 4,
so, since fxx (1, 1) > 0 and
fxx (1, 1)fyy (1, 1) − (fxy (1, 1))2 > 0,
we see that (1, 1) is a local minimum.
10. (a) Carefully equating the supplies and demands, we have the following system of
linear equations:
4p1 − 2p2 + 2p3 = 50

−2p1 + 3p2 − 2p3 = 20
p1 − 2p2 + 2p3 = 35.
Solving these using elementary row operations (but not simply by manipulating the
equations, since the question explicitly asks that a matrix method be used) we find
p1 = 5, p2 = 60 and p3 = 75.

(b) We have T C = M C dq = q + q 2 /2 + c and so the required change in the total
cost function is T C(4) − T C(2) = 8.
11. (a) This is a constrained optimisation problem. We want to find the values of
k and l that maximise q(k, l) subject to the constraint k + 3l = 80. The Lagrange
multiplier method can be used, with

−4
L = 10 3l−1/4 + k −1/4 − λ(k + 3l − 80).
The equations to be solved are

∂L
−5
= −40 3l−1/4 + k −1/4 (−1/4)k −5/4 − λ = 0
∂k
∂L
−5
= −40 3l−1/4 + k −1/4 (−3/4)l−5/4 − 3λ = 0
∂l
k + 3l = 80.
6

10k −5/4 = 10l−5/4 ,
so l = k. Then the third equation tells us that k = 20 and l = 20.
(b) Let yN be the required number. Then we have
y0 = 100,
y1 = (0.95)100 + 3,
y2 = (0.95) ((0.95)100 + 3) + 3 = (0.95)2 100 + (0.95)3 + 3,

y3 = (0.95) (0.95)2 100 + (0.95)3 + 3 + 3 = (0.95)3 100 + (0.95)2 3 + (0.95)3 + 3,

and, in general,
yN = (0.95)N 100 + (0.95)N −1 3 + (0.95)N −2 3 + · · · + (0.95)3 + 3

= (0.95)N 100 + 3 + 3(0.95) + 3(0.95)2 + · · · + 3(0.95)N −1

3(1 − (0.95)N )
= (0.95)N 100 +
1 − 0.95
= 100(0.95) + 60(1 − (0.95)N )
N
= 60 + 40(0.95)N .
This decreases with N , converging to 60.
Examination paper for 2005
The format of the 2005 examination will be the same as that of the 2004 examination;
namely seven compulsory questions in Section A and two questions to be chosen from
four in Section B.
7

Zone B
General remarks
19
05a Mathematics 1
20
21
05a Mathematics 1
22
23
05a Mathematics 1
24
25
05a Mathematics 1
26
27
Examiners’ commentaries 2010

05a Mathematics 1
Important note
This commentary reflects the examination and assessment arrangements for this unit in the
academic year 2009–10. The format and structure of the examination may change in future years,
and any such changes will be publicised on the virtual learning environment (VLE).
Specific comments on questions – Zone A
Section A
Answer all six questions from this section (60 marks in total).
Question 1
Show that the curve with equation y = x2 − x − 2 does not intersect the line with
equation y = x − 4. Sketch the two curves on the same diagram.
For which values of the number a will the curve with equation y = x2 − x − 2
intersect the line with equation y = x − a? For which particular value of a will there
be precisely one point at which the curve and the line intersect?
This question concerns curve sketching, discussed in Chapter 2 of the subject guide.
The equation x2 − x − 2 = x − 4 is equivalent to x2 − 2x + 2 = 0. This has no solution. We can

see this either by observing that its discriminant is negative, or by noting that it’s equivalent to
(x − 1)2 = −1. So the two curves do not intersect, because an intersection of y = f (x) and
y = g(x) corresponds to a solution of f (x) = g(x).
To find where the quadratic crosses the x-axis, note that x2 − x − 2 = (x − 2)(x + 1) so it crosses
the x-axis at 2, −1. The line crosses the x-axis at 4. Clearly, the quadratic crosses the y-axis at
−2 and the line at −4. The minimum of the quadratic is when 2x − 1 = 0, so x = 1/2 and
y = −9/4. The curves are illustrated in Figure 1.
The graphs of f (x) = x2 − x − 2 and g(x) = x − a intersect if and only if there’s a solution to
x2 − x − 2 = x − a.
This equation is equivalent to x2 − 2x + (a − 2) = 0.
This has discriminant

∆ = 4 − 4(a − 2) = 12 − 4a.
For solutions (and hence intersection) we require ∆ ≥ 0, meaning a ≤ 3.
There will be exactly one solution when ∆ = 0; that is, when a = 3.
1
05a Mathematics 1
25
20
15
10
–4 –2 2 4
x
–5
Figure 1: Sketch of the curves y = x − 4 and y = x2 − x − 2.
Question 2
Express the following system of equations in matrix form, and solve it using a
matrix method.
4x − y − z = 4,
2x − 3y − z = 4,
2x − 5y + z = 8.
The recommended method for solving linear equations can be found in Chapter 6 of the subject
guide. It is known by several names: the row operation method, the Gauss-Jordan method, the
row-reduction method, and so on. Many candidates like to use a different method, not covered in
the subject guide, especially ‘Cramer’s rule’. It is acceptable to do so, but our view is that the
row operations method is easier and less prone to error. But, whatever method you use, it has to
be a matrix method, as the question makes explicit: manipulation of the equations will not
suffice.
The system expressed in matrix form is

    
4 −1 −1 x 4
2 −3 −1 y  = 4 .
2 −5 1 z 8
The standard matrix method approach is now to reduce the augmented matrix to reduced form.
The augmented matrix is  
4 −1 −1 4
 2 −3 −1 4  .
2 −5 1 8
Here is one way. (There are others, equally valid.)

     
4 −1 −1 4 4 −1 −1 4 4 −1 −1 4
 2 −3 −1 4  →  4 −6 −2 8 → 0 −5 −1 4 
2 −5 1 8 4 −10 2 16 0 −9 3 12
     
4 −1 −1 4 4 −1 −1 4 4 −1 −1 4
→  0 45 9 −36  →  0 45 9 −36  →  0 5 1 −4  .
0 −45 15 60 0 0 24 24 0 0 1 1
2

4x + y − z = 4
5y − z = 4
z = 1.
So, z = 1, y = (1/5)(−4 − z) = −1 and x = (1/4)(4 + z + y) = 1.
This is probably a good point at which to make some general comments about how questions are
marked. Clearly, in a question like this, it is easy to get the wrong answer. (Though it should be
noted that in this particular question, you can always substitute the values that you have found
into the original equations, and this will show whether these are correct or not. So you can tell if
you have the wrong answer and, if you have time, you can then re-work the calculation.)
Examiners understand that arithmetical errors can be made, especially in the stressful
circumstances of an examination. Quite probably, the Examiners themselves would make some
mistakes if they sat the paper. So, although there are certainly some marks for correct
calculation, there are many marks for using the right method (even if you make a mistake). So,
here, for instance, the Examiners will award marks if you can indicate that you know how to
start to solve the equations (by writing down an augmented matrix); that you know what row
operations are; that you know what it is you want to achieve with row operations (the reduced
matrix, that is); and that you then know how to work from that reduced matrix to determine the
required solutions. There are marks for all these things.
Be sure to understand that only certain types of operations qualify as valid row operations. In
particular, a number of candidates make the mistake of thinking that subtracting a fixed
constant from each entry of a row is valid. It is not. (And, if you don’t know what we mean by
that, then you’re probably not doing it, which is good!)
Question 3
Determine the integrals

x x
Z Z
√ dx, dx.
ex x2 + 5x + 4
Integration is discussed in Chapter 4 of the subject guide. It can be difficult because it is not
always clear which technique will work. The three main techniques are: substitution, parts and
partial fractions.
x
Z
For √ dx, we can use integration by parts.
ex
We have
x
Z Z
√ dx = xe−x/2 dx
ex
so taking u = x, dv = e−x/2 yields du = 1, v = −e−x/2 , and
Z
= −2xe−x/2 + 2 e−x/2 dx
= −2xe−x/2 − 4e−x/2 + c.
x
Z
For dx, we use partial fractions. We have
x2 + 5x + 4
x x 4/3 1/3
= = − .
x2 + 5x + 4 (x + 4)(x + 1) x+4 x+1
3
05a Mathematics 1
So,
x 4 1
Z
dx = ln |x + 4| − ln |x + 1| + c.
x2 + 5x + 4 3 3
Question 4
The function f is defined for positive y and all x by
f (x, y) = x2 ln y − y ln y.
Find the critical (or stationary) points of f and determine, for each, whether it is a
local maximum, a local minimum, or a saddle point.
This is a fairly standard type of question, using the material in Chapter 5 of the subject guide.
With f (x, y) = x2 ln y − y ln y, we have
x2
fx = 2x ln y, fy = − ln y − 1.
y
So we solve fx = fy = 0.
From 2x ln y = 0 we have x = 0 or y = 1.
If x = 0, then fy = 0 becomes ln y = −1, so y = e−1 = 1/e.
If y = 1 then fx = 0 becomes x2 − 1 = 0, so x = ±1.
So the critical points are

(0, 1/e), (1, 1), (−1, 1).
To determine the nature of the critical points, we use the second-derivative test. We have
x2 1 2x
fxx = 2 ln y, fyy = − − , fxy = .
y2 y y
2
Consider first (0, 1/e). Here, fxx = −2, fyy = −e, fxy = 0. So, fxx fyy − fxy = 2e > 0 and
fxx < 0, so the point is a local maximum.
2
Consider (1, 1). Here, fxx = 0, fyy = −2, fxy = 2. So, fxx fyy − fxy = −4 < 0, so the point is a
saddle point.
2
Consider (−1, 1). Here, fxx = 0, fyy = −2, fxy = −2. So, fxx fyy − fxy = −4 < 0, so the point is
a saddle point.
Question 5
The production of an amount q of a very specialized high technology good requires

the input of capital, k, and labour, l, where
q = k1/4 l1/4 .
If the price of capital is $100 per unit per week and the cost of labour is $400 per
unit per week, find the maximum possible weekly output, Q(M ), if an amount $M
is spent each week on capital and labour.
4
In determining its production schedule, the manufacturer spends an amount M that

maximises 160Q(M ) − M . What is this optimal value of M ?
This is a constrained optimisation problem and the appropriate method to solve it is the
Lagrange multiplier method, discussed in Chapter 5 of the subject guide.
For this, we need to maximise k 1/4 l1/4 subject to 100k + 400l = M .

1, CAO, B, for objective and constraint. Must be exact. No marks for getting these the wrong
way round. This can be implicit in what follows, in the writing down of the correct Lagrangian.
The Lagrangian is
L = k 1/4 l1/4 − λ(100k + 400l − M ).
Here, one can also use +λ in place of −λ, or M − 100k − 400l in place of 100k + 400l − M .
We have
1 −3/4 1/4
Lk = k l − 100λ = 0
4
1 1/4 −3/4
Ll = k l − 400λ = 0
4
100k + 400l = M.
(Here, Lk denotes ∂L/∂k and Ll denotes ∂L/∂l.)
Eliminating λ,
4k −3/4 l1/4 = k 1/4 l−3/4 ,
so
k = 4l.
Then, 100k + 400l = M implies 800l = M.
So l = M/800 and k = 4l = 4M/800.
You are not expected to check second order conditions for constrained optimisation problems.
This is not part of the syllabus and there is no credit for doing it.
Then, the maximum output is

√ √
1/4 1/4 1/4 M M
Q(M ) = k l =4 √ = .
800 20
For the next part, we need to maximise

√
f (M ) = 160Q(M ) − M = 8 M − M.
We have
4
f ′ (M ) = √ − 1,
M
and f ′ = 0 ⇐⇒ M = 16.
We also note f ′′ (M ) < 0, so M = 16 is a maximum.
5
05a Mathematics 1
Question 6
An investor saves money in a bank account paying interest at a fixed rate of 100r%,
where the interest is paid once per year, at the end of the year. She deposits an
amount D at the beginning of each of the next N years. Show that she will then
have saved an amount equal to
D
((1 + r)N − 1),
r
just after the last of these deposits.
Chapter 7 of the subject guide gives the required background material.
Let yn be the amount of money after the nth deposit. Then:
y1 = D,
y2 = D(1 + r) + D,
y3 = y2 (1 + r) + D = D(1 + r)2 + D(1 + r) + D.
Spotting the pattern,
yN = D(1 + r)N −1 + D(1 + r)N −2 + · · · + D(1 + r) + D (∗)

(1 + r)N − 1
= D
(1 + r) − 1
D
= ((1 + r)N − 1).
r
Section B
Answer both questions from this section (20 marks each).
Question 7
(a) A firm is a monopoly and it has fixed costs of 10 and marginal cost function
5q 4 − 2q − 15q 2 . The demand equation for the good it produces is p + q = 20, where
p is the selling price. Find an expression for the firm’s profit, as a function of q, and
determine the production level that maximises the firm’s profit.
This is essentially a one-variable optimisation problem, and the underlying material can be found
in Chapter 3 of the subject guide.
A common error in a question of this type is to misunderstand what is meant by a ‘marginal’.
For example, a number of candidates thought that T C = q(M C) + F C. This confuses marginal
cost with average cost. Recall that since the marginal cost is the derivative of the total cost, we
must integrate the marginal cost to find the total cost.
R
We have T C = M C dq = q 5 − q 2 − 5q 3 + c where, since T C(0) = F C = 10, we must have
c = 10. (You should explain, as here, why c = 10.)
The total revenue is T R = pq = (20 − q)q. Here, we’ve used the fact that the firm is a monopoly,
so that we can use the demand equation to obtain an expression for the price in terms of the
firm’s production level, q.
So Π = T R − T C = 20q − q 5 + 5q 3 − 10.
We solve Π′ = 0. This is 20 − 5q 4 + 15q 2 = 0, which we note is a quadratic in q 2 and simplifies as
−5(q 2 − 4)(q 2 + 1) = 0, to which the only economically meaningful solution is q = 2.
6
Further, since Π′′ = −20q 3 + 30q, we have Π′′ (2) < 0 and therefore we maximise profit.
You might think that an alternative method here is to solve the equation M R = M C in order to
maximise the profit. But this does not fully answer the question because, although it will indeed
lead to the value q = 4, it will not result in an explicit expression for the profit function in terms
of q, something the question clearly asks for. Nor, without extra work, will it show that the value
q = 4 does indeed maximise profit. If you took this approach, you would still get some marks,
but not all of them. So, this is a case where a careful reading of the question is important.
(b) The function f is given, for x, y > 0, by

x
f (x, y) = (y − x) ln .
y
Determine the first and second-order partial derivatives of f and show that
∂ 2f ∂ 2f ∂ 2f
x2 + 2xy + y2 = 0.
∂x2 ∂x∂y ∂y 2
Partial differentiation is discussed in Chapter 5 of the subject guide.

It would be useful to note that

x
f (x, y) = (y − x) ln = (y − x)(ln x − ln y).
y
But suppose we just proceed directly. Then:

x 1 y x y−x x y
fx = − ln + (y − x) = − ln + = − ln −1+ ,
y y x y x y x

x x y x x
fy = ln + (y − x) − 2 = ln −1+ ,
y y x y y
1y y 1 y
fxx = − − = − − 2,
y x x2 x x
x y 1 1 1
fxy = fyx = + = + ,
y2 x x y x
x y x 1 x
fyy = − − = − − 2.
y2 x y2 y y
So,
∂2f ∂2f ∂2f

1 y 1 1 1 x
x2 2
+ 2xy + y2 2 = x2 − − 2 + 2xy + + y2 − − 2
∂x ∂x∂y ∂y x x y x y y
= −x − y + 2x + 2y − y − x
= 0.
Question 8
A consumer has utility function

xy
u(x, y) = ,
x+y
for two goods, X and Y . Here, x > 0 denotes the amount of X consumed and y > 0
the amount of Y consumed. Each unit of x costs p dollars, each unit of Y costs q
dollars, and the consumer has a budget for X and Y of M dollars.
7
05a Mathematics 1
Use the Lagrange multiplier method to find the quantities x∗ of X and y ∗ of Y the
consumer will consume in order to maximise his utility subject to the budget
constraint. (Your answers will depend on p, q and M .)
Show that
M
x∗ + y ∗ = √ √ .
p q
Hence find the optimal value, λ∗ , of the Lagrange multiplier, simplifying your
expression as much as possible.
Suppose that V = u(x∗ , y ∗ ) is the maximum achievable utility. Find an explicit

expression for V in terms of p, q and M , simplifying your answer as much as
possible.
∂V
Find and verify that it is equal to λ∗ .
∂M
∂V
Find . Use this to find an approximate expression for the change in V if the
∂p
price of X increases by one dollar.
This question primarily concerns the Lagrange multiplier method discussed in Chapter 5 of the
subject guide.
To determine optimal affordable consumption, the consumer maximises u(x, y) subject to the
budget constraint px + qy = M .
The Lagrangian is
xy
L= − λ(px + qy − M ).
x+y
We have
y(x + y) − xy y2
Lx = − pλ = − pλ = 0
(x + y)2 (x + y)2
x(x + y) − xy x2
Ly = 2
− qλ = − qλ = 0
(x + y) (x + y)2
px + qy = M.
Eliminating λ from the first two of these equations, we see that
y2 x2
= ,
p q
so that
p
r
y= x.
q
Then, px + qy = M implies
√
px + pqx = M,
so x∗ , the solution to this, is given by
M
x∗ = √ .
p + pq
8
Then,
p ∗ p M M
r r
∗
y = x = √ = √ .
q q p + pq q + pq
Now,

1 1
x∗ + y ∗ = M √ + √
p + pq q + pq
√
M (p + q + 2 pq)
= √ √
(p + pq)(q + pq)
√ √
M ( p + q)2
= √ √ √ √
p q( p + q)2
M
= √ √ .
p q
The corresponding value of λ∗ is

(y ∗ )2
λ∗ = .
p(x∗ + y ∗ )2
So,
(y ∗ )2 pq
λ∗ =
pM 2
M 2q
= √
M 2 (q + pq)2
1
= √ √ .
( q + p)2
We also have
x∗ y ∗
V =
x∗ + y∗
√ √
M M p q
= √ √ √ √ √ √
p( p + q) q( q + p) M
M
= √ √ .
( p + q)2
So
∂V 1
= √ √ = λ∗ .
∂M ( p + q)2
We have
∂V M
= −√ √ √ .
∂p p( q + p)3
√ √ √
So if p increases by 1 dollar, V decreases by approximately M/( p( q + p)3 ). This last part
relies on us understanding that the derivative of V with respect to p means the rate of change of
V when p is changed.
9
Mathematics 1
279005a, 990005a, 996D05a
2003, 2004
05a Mathematics 1

Zone A
means two things. First, they should not simply write down the answer in the exam
script, but explain the method by which it is obtained. Secondly, they should include
rough working. The examiners want you to get the right answers, of course, but it is
more important that you prove you know what you are doing: that is what is really
being examined.
We also stress that if a candidate has not completely solved a problem, they may still
be awarded marks for a partial, incomplete, or slightly wrong, solution; but, if they
have written down a wrong answer and nothing else, no marks can be awarded.
in Section A.
Specific comments on questions
1. It should be realised that the curves are parabolas, and their intersections with the
axes should be indicated. Finding the intersection point involves solving the equation
4x2 − 8x − 1 = −4x2 − 2x + 4 which, rewritten as a quadratic equation in standard
form, becomes 8x2 − 6x − 5 = 0, and this can be solved in the usual ways. The positive
solution is x = 5/4.
2 2 2
2. We find f (x) = 2e−x −2x(1+2x)e−x = 2e−x (1−x−2x2 ). Solving f (x) = 0 means
solving 1 − x − 2x2 = 0, which has solutions 1/2 and −1 (as can be seen by factorising
or by using the formula for the solutions of a quadratic equation). The nature of each
critical point can be checked by examining the behaviour of the derivative around each
point or (more difficult in this particular problem) by computing the second derivative
2
at each point. It is found that x = 1/2 gives a maximum.
3. This integral can be solved using the partial fractions method. We find that
x 2/3 1/3
= +
x2 −x−2 x−2 x+1
and so the integral is
2 1
ln |x − 2| + ln |x + 1| + c.
3 3
4. We have
∂f
= 2x − 2y,
∂x
∂f
= −2x + 4y − 2.
∂y
The stationary point of f is found to be given by x = y = 1 by setting these both to
equal 0. We have (writing fxx for ∂ 2 f /∂x2 and so on), fxx = 2, fxy = −2 and fyy = 4,
2
so, since fxx > 0 and fxx fyy − fxy > 0, f (x, y) has a local minimum at (1, 1).
5. The Lagrangean is L = xy 3/2 − λ(x + 2y − 100) and the first-order conditions are
∂L
= y 3/2 − λ = 0
∂x
∂L 3
= xy 1/2 − 2λ = 0
∂y 2
x + 2y = 100.
3
λ = y 3/2 = xy 1/2 ,
4
which on simplification becomes y = (3/4)x. Then the third equation shows that
x = 40, y = 30. There is no need to do a second-order test for the nature of the point:
that is not part of the syllabus and there is no credit for it.
6. The system expressed in matrix form (a step some forgot) is

    
4 1 −2 x 4
    
 2 3 −2   y  =  4  .
2 5 2 z 8
(The system ‘expressed in matrix form’ is not the augmented matrix.) The standard
matrix method approach is now to reduce the augmented matrix to reduced form. Here
is one way. (There are others, equally valid. But note here how we have in several
places been able to avoid introducing fractions by multiplying rows by appropriate
constants before cancelling entries.)
     
4 1 −2 4 4 1 −2 4 4 1 −2 4
     
 2 3 −2 4  →  4 6 −4 8  →  0 5 −2 4 
2 5 2 8 4 10 4 16 0 9 6 12
3
05a Mathematics 1
     
4 1 −2 4 4 1 −2 4 4 1 −2 4
     
→  0 45 −18 36  →  0 45 −18 36  →  0 5 −2 4  .
0 45 30 60 0 0 48 24 0 0 2 1
4x + y − 2z = 4
5y − 2z = 4
2z = 1.
So, z = 1/2, y = (1/5)(4+2z) = 1 and x = (1/4)(4+2z −y) = 1. Simply manipulating
equations does not constitute a matrix method, which the question specifically asks
for, and gets no credit. Some candidates used Cramer’s rule. This is a technique not
discussed in the Subject Guide, and the computations required are more difficult. We
recommend the reduction technique, but Cramer’s rule, carefully applied, gives the
correct answer.
7. For some a and r, the nth term of the sequence takes the form arn−1 . The sum to
infinity is a/(1 − r). So we must have
a
ar = 2, = 9.
1−r
These can now be solved simultaneously either by eliminating r or by eliminating a.
Since the question asks for r, it is easiest to eliminate a. We see that, since a = 9(1−r),
we have 9(1 − r)r = 2, so 9r2 − 9r + 2 = 0, with solutions r = 2/3, 1/3. (Of course,
alternatively, r may be eliminated, a found, and then r determined from a.)
8. (a) We have T C = V C + F C = q(AV C) + F C = q 3 + 2q 2 + ln(q 2 + 1) + 9 and

d 2q
MC = T C = 3q 2 + 4q + 2 .
dq q +1
(b) The first integral

is most easily done using a substitution. If we set u = x + 3 the
integral becomes (u − 3)2 u1/2 du, which is

2 12 2 12
u5/2 −6u3/2 +9u1/2 du = u7/2 − u5/2 +6u3/2 +c = (x+3)7/2 − (x+3)5/2 +6(x+3)3/2 +c
7 5 7 5
√
The substitution u = x + 3 also works. So does integration by parts, though this is
harder. (Some candidates attempted to use integration by parts, but only succeeded in
ending up with a more complicated integral. No credit was given for this. Integration
by parts is only useful if the resulting integral is easier than the one we start with.)
For the second integral, we can use integration by parts:

−2 1 11 1 1
x ln x dx = − ln x + dx = − ln x − + c.
x xx x x
(c) This was not well done. Although the partial derivative with respect to x can be
computed immediately as
∂f
= −yx−y−1 ,
∂x
4
the derivative with respect to y is more difficult. For this, it is useful either to realise
that f = exp(−y ln x) or to take logarithms and differentiate. We have
ln f = ln(x−y ) = −y ln x
and so, differentiating each side with respect to y,

1 ∂f
= − ln x,
f ∂y
so
∂f
= −(ln x)f = −(ln x)x−y .
∂y
9. (a) The total cost is T C = q(AC) = 40 + 20q − 3q 2 + q 3 /2. Since the firm is a
monopoly, we can find the price in terms of q from the demand equation. Hence the
total revenue is T R = pq = (40 − (5/2)q)q = 40q − (5/2)q 2 and profit is
3
Π(q) = T R − T C = 20q + q − q 2 .
2
To maximise profit, we solve Π (q) = 0, which means 3q 2 − 2q − 40 = 0. This has the
solutions q = −3/10 and q = 4, and it is the positive one that has economic meaning.
We can see also that Π (4) < 0, so q = 4 maximises profit.
(b) This is a constrained optimisation problem. We want to find the minimum cost of
producing an amount q and, given the information in the question, this means we want
to minimise k + 16l subject to the constraint k 1/4 l1/4 = q. The Lagrange multiplier
method can then be used, with

L = k + 16l − λ k 1/4 l1/4 − q .
The equations to be solved are

∂L 1 −3/4 1/4
=1− λk l = 0
∂k 4
∂L 1 1/4 −3/4
= 16 − λk l = 0
∂l 4
k 1/4 l1/4 = q.
From the first two equations, λ = 4k 3/4 l−1/4 = 64k −1/4 l3/4 , so k = 16l and then the
third tells us that (16)l1/4 l1/4 = q, so l1/2 = q/2 and hence l = q 2 /4, k = 16l = 4q 2 and
the minimum cost is 4q 2 + 16(q 2 /4) = 8q 2 .
10. (a) Carefully equating the supplies and demands, we obtain the following system
of linear equations:
4p1 − 2p2 + 4p3 = 50

−2p1 + 3p2 − 4p3 = 20
p1 − 2p2 + 4p3 = 35.
5
05a Mathematics 1
Solving these using elementary row operations (but not simply by manipulating the
equations, since the question explicitly asks that a matrix method be used) we find
p1 = 5, p2 = 60 and p3 = 75/2.

(b) We have T C = M C dq = q + 2e0.5q + 13 q 3 + c and, given that T C(0) = F C = 10,
we have 0 + 2 + 0 + c = 10, so c = 8 (and not 10!).
11. (a) We have total revenue given, as a function of q1 , q2 , by
T R = p1 q1 + p2 q2 = (1050 − 25q1 )q1 + (500 − 2q2 )q2 ,
and the profit is
Π(q1 , q2 ) = T R − T C = 1000q1 − 25q12 + 480q2 − q22 − 10.
Solving ∂Π/∂q1 = 0 and ∂Π/∂q2 = 0 gives q1 = 20 and q2 = 120. The second order
conditions are easily checked to show that these quantities do indeed maximise profit.
(b) Let yN be the required number. Then we have
y0 = 1000,
y1 = (0.98)1000 + 30,
y2 = (0.98) ((0.98)1000 + 30) + 30 = (0.98)2 1000 + (0.98)30 + 30,

y3 = (0.98) (0.98)2 1000 + (0.98)30 + 30 +30 = (0.98)3 1000+(0.98)2 30+(0.98)30+30,

and, in general,
yN = (0.98)N 1000 + (0.98)N −1 30 + (0.98)N −2 30 + · · · + (0.98)30 + 30

= (0.98)N 1000 + 30 + 30(0.98) + 30(0.98)2 + · · · + 30(0.98)N −1

30(1 − (0.98)N )
= (0.98)N 1000 +
1 − 0.98
= 1000(0.98) + 1500(1 − (0.98)N )
N
= 1500 − 500(0.98)N .
This increases with N , converging to 1500.
The format of the 2005 examination will be the same as that of the 2004 examination;
namely seven compulsory questions in Section A and two questions to be chosen from
four in Section B.
6
Examiners’ commentary 2008

05A Mathematics 1
Specific comments on questions Zone A
Question 1
The demand equation for a good is q(p + 3) = 4 and the supply equation is 2q − p + 4 = 0 where
p is the price and q is the quantity. Sketch the supply and demand functions for p ≥ 0.
Determine the equilibrium price and quantity.
The sketch of the two functions for p ≥ 0 is:
2q − p + 4 = 0
4
3
1 q(p + 3) = 4
2
O
4 5 p
−2
(It is quite acceptable to sketch p versus q instead.) A completely correct answer should
indicate:
• the shape of the demand curve (particularly, that it curves the right way and that it does
not intersect the p-axis);
• the shape of the supply curve (a straight line);
• the intercepts of both curves with the q-axis;
• the intercept of the supply curve with the p-axis.
(Partial credit would be given if some of the above had not been indicated.) You should also
remember, as indicated in the subject guide, that simply ‘plotting’ points and joining them
up is not an acceptable way to sketch curves.
1
05A Mathematics 1
For the equilibrium price, we solve the equation

4 1
= (p − 4).
p+3 2
This is equivalent to
(p + 3)(p − 4) = 8 =⇒ p2 − p − 20 = 0,
and so, factorising, we have
(p − 5)(p + 4) = 0 =⇒ p = 5, −4.
Thus, the equilibrium price is p = 5 (as the negative solution is not economically meaningful).
Of course, it is perfectly acceptable to use the formula for the solutions of a quadratic
equation instead of factorising.
The equilibrium quantity is then q = (5 − 4)/2 = 1/2. (You should not round this up or
down, since nowhere in the question does it indicate that the quantities must be integers.)
Question 2
A firm has fixed costs of 10 and its marginal revenue and marginal cost functions are given,
respectively, by
M R = 11 − 2q, M C = q 2 − 4q + 3,
where q is the level of production. Find an expression, in terms of q, for the firm’s profit. Find
also the value of q which maximises the profit (and show that this value of q does indeed
maximise profit).
This question combines differentiation, integration, and optimisation of one-variable functions

and the relevant material can be found in Chapters 3 and 4 of the subject guide and in the
readings indicated there.
As the marginal revenue is the derivative of the total revenue, we integrate to find that
Z
T R = M R dq = 11q − q 2 + c.
Now, since there can be no revenue if no goods are sold we have T R(0) = 0. But, from above,
T R(0) = c and so we have c = 0.
Similarly, as the marginal cost is the derivative of total cost, we have

Z
1
T C = M C dq = q 3 − 2q 2 + 3q + c.
3
To find c, note that T C(0) = F C = 10, so c = 10.
The profit, Π, is then given by

1
Π = T R − T C = 8q + q 2 − q 3 − 10.
3
To maximise the profit, we find that
Π0 (q) = 8 + 2q − q 2 ,
and solve the equation Π0 (q) = 0, which is
8 + 2q − q 2 = 0 =⇒ q 2 − 2q − 8 = 0,
2
so that, factorising, we get

(q + 2)(q − 4) = 0 =⇒ q = −2, 4.
Thus, we take q = 4 (as the negative solution is not economically meaningful). Of course, it is
perfectly acceptable to use the formula for the solutions of a quadratic equation instead of
factorising. Now,
Π00 (q) = 2 − 2q
and, since Π00 (4) < 0, q = 4 does indeed maximise the profit.
You might think that an alternative method here is to solve the equation M R = M C in order
to maximise the profit. But this does not fully answer the question because, although it will
indeed lead to the value q = 4, it will not result in an explicit expression for the profit
function in terms of q, something the question clearly asks for. Nor, without extra work, will
it show that the value q = 4 does indeed maximise profit. If you took this approach, you
would still get some marks, but not all of them. So, this is a case where a careful reading of
the question is important.

For example, a number of students thought that T C = q(M C) + F C. This confuses marginal
cost with average cost. Recall that since the marginal cost is the derivative of the total cost,
we must integrate the marginal costs to find the total cost. Furthermore, you should explain
why the constant of integration is what it is, and not just assume that it equals the fixed cost.
In finding the total revenue, you have to realise that when the quantity is 0, the revenue is
also 0: only with this observation can you see that the constant of integration in this case is 0.
Question 3
A function f is given, for x > 0, by

A
f (x) = + B + Cx,
x2
for some numbers A, B, C. If f (1) = 37, f (2) = 9 and f (3) = 3, show that
A + B + C = 37, A + 4B + 8C = 36, A + 9B + 27C = 27.
Express this system of equations in matrix form. Solve it, using a matrix method, to determine
A, B and C.
At first, this might look like an unusual sort of question, but it is in fact nothing more than a
system of linear equations. Now, the equations are given in the question, so even if you could
not do the first part of the question (showing that the equations hold), you could certainly
continue with the second part (solving the equations).
The recommended method for solving linear equations can be found in Chapter 6 of the
subject guide. It is known by several names: the row operation method, the Gauss-Jordan
method, the row-reduction method, and so on. Many students like to use a different method,
not covered in the subject guide, especially ‘Cramer’s rule’. It is acceptable to do so, but our
view is that the row operations method is easier and less prone to error. But, whatever
method you use, it has to be a matrix method, as the question makes explicit: manipulation
of the equations will not suffice.
Substituting in the given information leads to the equations:

A A
A + B + C = 37, + B + 2C = 9 and + B + 3C = 3.
4 9
To see this, note, for example, that f (1) = A + B + C and we know that f (1) = 37, so the
first equation follows. The other two equations are similarly obtained. Then, multiplying the
second equation by 4 and the third by 9 gives the required equations.
3
05A Mathematics 1
We solve these equations using row operations which, as noted above, is the method
recommended for Mathematics 1. The augmented matrix is
 
1 1 1 37
 1 4 8 36  ,
1 9 27 27
and using row operations to reduce this, we have

     
1 1 1 37 1 1 1 37 1 1 1 37
 1 4 8 36  →  0 3 7 −1  →  0 3 7 −1  .
1 9 27 27 0 8 26 −10 0 0 22 −22
So, this gives us
A + B + C = 37, 3B + 7C = −1, and 22C = −22,
and then, working backwards, we get

1
C = −1, B= (−1 − 7C) = 2 and A = 37 − B − C = 36.
3
This is probably a good point at which to make some general comments about how questions
are marked. Clearly, in a question like this, it is easy to get the wrong answer. (Though it
should be noted that in this particular question, you can always substitute the values of A, B
and C that you have found into the original equations, and this will show whether these are
correct or not. So you can tell if you have the wrong answer and, if you have time, you can
re-work the calculation.) Examiners understand that arithmetical errors can be made,
especially in the stressful circumstances of an examination. Quite probably, the examiners
themselves would make some mistakes if they sat the paper. So, although there are certainly
some marks for correct calculation, there are many marks for using the right method (even if
you make a mistake). So, here, for instance, examiners will award marks if you can indicate
that you know how to start to solve the equations (by writing down an augmented matrix);
that you know what row operations are; that you know what it is you want to achieve with
row operations (the reduced matrix, that is); and that you then know how to work from that
reduced matrix to determine the required values of A, B and C. There are marks for all these
things.
Be sure to understand that only certain types of operations qualify as valid row operations.
In particular, a number of students make the mistake of thinking that subtracting a fixed
constant from each entry of a row is valid. It is not. (And, if you don’t know what we mean
by that, then you’re probably not doing it, which is good!)
Question 4
Show that the function

f (x, y) = 19 − x2 + 6xy − 10y 2 − 2y
has one critical (or stationary) point and that the point is a local maximum.
This is a very standard type of question, using the material in Chapter 5 of the subject guide.
The partial derivatives are
fx = −2x + 6y and fy = 6x − 20y − 2.
To find the critical points, we solve the equations fx = 0 and fy = 0. The first of these
equations gives us x = 3y and substituting this into the second equation gives −2y − 2 = 0, so
y = −1. Using x = 3y again then gives us x = −3. Thus, the function has one critical point,
namely (−3, −1).
4
The second derivatives are
fxx = −2, fxy = 6 and fyy = −20.
Since these values assure us that

2
fxx fyy − fxy = (−2)(−20) − 62 = 4 > 0 and fxx < 0,
this point is indeed a local maximum. (Note that it is not correct to simply observe that
fxx < 0 and fyy < 0: you have to use the test correctly!)
Question 5
Determine the integral Z

√
x2 x − 1 dx.
always clear which technique will work. But you should of course realise that one of the
techniques will work! The three main techniques are: substitution, parts, and partial
fractions. The last of these applies only to very special types of integrals, and it turns out
that, for this problem, you can use either substitution or parts.
Substitution is probably the easiest way. If we let u = x − 1, then du = dx and so the integral
becomes
Z Z
√ √
2
x x − 1 dx = (u + 1)2 u du
Z
= (u2 + 2u + 1)u1/2 du
Z
= (u5/2 + 2u3/2 + u1/2 ) du
2 7/2 4 5/2 2 3/2
u + u + u +c
=
Z 7 5 3
√ 2 4 2
∴ x x − 1 dx = (x − 1) + (x − 1)5/2 + (x − 1)3/2 + c.
2 7/2
7 5 3
Although it is probably more difficult, the integral can also be done using integration by
parts, as follows.
Z Z
2
√ 2 2 3/2 4
x x − 1 dx = x (x − 1) − x(x − 1)3/2 dx
3 3
Z
2 8 8
= x2 (x − 1)3/2 − x(x − 1)5/2 + (x − 1)5/2 dx
3 15 15
Z
√ 2 8 16
∴ x2 x − 1 dx = x2 (x − 1)3/2 − x(x − 1)5/2 + (x − 1)7/2 + c.
3 15 105
This might look like a different answer to the one obtained using substitution, but it is, in
fact, the same. (A fact that you can verify by showing that these two answers only differ by a
constant.)
5
05A Mathematics 1
Question 6
Use
p the Lagrange multiplier method to find the values of x and y that minimise the function
2x2 + y 2 subject to the constraint x + y = 1.
The Lagrangean is p
L= 2x2 + y 2 − λ(x + y − 1),
and we need to solve the equations
2x
Lx = p − λ = 0,
2x2 + y 2
y
Ly = p − λ = 0.
2x2 + y 2
Eliminating λ, this means that
2x y
λ= p =p ,
2x2 + y2 2x2 + y2
so that y = 2x.
Then, substituting this into the constraint equation, x + y = 1, gives us 3x = 1 and so

x = 1/3. Then, using y = 2x again, we get y = 2/3.
Of course, you might realise that an equivalent problem is to minimise 2x2 + y 2 subject to the
constraint, and that is acceptable.
In questions like this, as the subject guide emphasises, you are not expected to verify that a
minimum has been found. Techniques for doing so are not part of the syllabus (and they are
rather more complicated than you might think).
It is possible to obtain the answer to this question without using Lagrange’s method, but that
does not answer the question since it explicitly asks that the Lagrange multiplier method be
used. So, if you did this, you would only get some of the marks. For example, you could use
the constraint equation to express y in terms of x as y = 1 − x. Then, we have to minimise
p p
f = 2x2 + (1 − x)2 = 3x2 − 2x + 1.
This gives
3x − 1
f0 = √ ,
3x2 − 2x + 1
and so, solving f 0 (x) = 0 gives x = 1/3 and, as y = 1 − x, we also therefore have y = 2/3.
To see that this gives a minimum, we note that f 0 changes sign from negative to positive at
1/3. (Or, although this is messier, we could have used the second derivative test to see this.)
That’s all correct, but look at the question: it says ‘Use the Lagrange multiplier method’ and
you need to do this in order to answer the question properly.
6
Question 7
An investor saves money in a bank account paying interest at a fixed rate of 4%, where the
interest is paid once each year, at the end of the year. She deposits an amount P at the
beginning of each year for n years. Determine the amount of money in the bank account
immediately after n such deposits have been made. Your answer should be an expression
involving P and n. (Show your working carefully.) Show also that if this amount is at least M
immediately after n deposits, then
ln ((M/25P ) + 1)
n≥ .
ln(1.04)
These questions always seem to cause more difficulty than they ought to. Chapter 7 of the
subject guide gives the background and some examples. Essentially, what you have to do is:
identify the sequence that you are trying to find; determine the first few values; spot a general
expression for the sequence; and simplify that expression. It is also possible to attack such
problems using general techniques for solving difference equations, but that is not expected
(as such techniques are not part of the 05A syllabus, though they are in 05B).
It is, more than ever, important to read questions such as this carefully: they might well be
subtly different from any ‘standard’ ones that you are used to.
Let yn be the required balance, i.e. the amount in the account immediately after n deposits
have been made. This means that
y1 = P,
y2 = (1.04)y1 + P = (1.04)P + P,
y3 = (1.04)y2 + P = (1.04)2 P + (1.04)P + P,
y4 = (1.04)y3 + P = (1.04)3 P + (1.04)2 P + (1.04)P + P,
and, in general,
yn = (1.04)n−1 P + (1.04)n−2 P + · · · + (1.04)P + P.
(At this point, we have spotted the pattern!)
Noting that we have a geometric series, this can then be simplified to get
µ ¶
1 − (1.04)n
yn = P = 25P [(1.04)n − 1].
1 − 1.04
So far, so good. But what about the final part of the question? The balance needs to be at
least M after n deposits, i.e. we need yn ≥ M , and so using our formula for the balance, this
must mean that we need
M
25P [(1.04)n − 1] ≥ M =⇒ (1.04)n ≥ + 1.
25P
Then taking natural logarithms of both sides this gives we have
µ ¶ µ ¶Á
M M
n ln(1.04) ≥ ln +1 =⇒ n ≥ ln +1 ln(1.04),
25P 25P
as required.
7
05A Mathematics 1
Question 8
(a) The function f is defined, for x > 0, by

f (x) = x − x(ln x)2 .
Show that its critical (or stationary) points satisfy the equation
(ln x)2 + 2 ln x − 1 = 0.
By solving the quadratic equation w2 + 2w − 1 = 0, or otherwise, find the critical points.
Determine the nature of each critical point.
(b) A firm is the only producer of two goods, X and Y . The demand quantities x and y for X
and Y (respectively) and the corresponding prices pX and pY are related as follows:
−pX + pY + 13 − 3x = 0, pX − 4pY + 26 − 3y = 0.
The firm’s joint total cost function (that is, the cost of producing x of X and y of Y ) is
14x + 7y. Find an expression in terms of x and y for the profit function. Determine the
quantities x and y that maximise the profit.
(a) This is a one-variable optimisation problem, which uses the techniques in Chapter 3 of the
subject guide. But it’s a little unusual, isn’t it? The examiners certainly thought so, which
is why they asked it. But — and you must always remember this — you can do it using the
methods of this unit.
We differentiate to find that
2 ln x
f 0 (x) = 1 − (ln x)2 − x = −(ln x)2 − 2 ln x + 1,
x
and solve the equation f 0 (x) = 0 to find the critical points. This means that, to find these
points, we need to solve the equation
(ln x)2 + 2 ln x − 1 = 0,
as required.
Using the formula for the solutions of a quadratic equation, we see that
w2 + 2w − 1 = 0
has solutions √
w = −1 ± 2.
So, letting w = ln x, we see that the critical points are given by
√ √ √
ln x = −1 ± 2 =⇒ x = e−1+ 2 , e−1− 2 .
Let’s call these r, s respectively.
To determine the nature of these critical points, we differentiate again to see that
2 ln x 2 2
f 00 (x) = − − = − (ln x + 1) .
x x x
This means that, as r > 0,
2 √
f 00 (r) = − ( 2) < 0,
r
i.e. r is a local maximum and, as s > 0,
2 √
f 00 (s) = − (− 2) > 0,
s
i.e. s is a local minimum.
The natures of r and s can also be found by considering the behaviour of the sign of f 0 (x)
around the two points remembering, of course, that f 0 (x) = −(ln x)2 − 2 ln x + 1 and not
(ln x)2 + 2 ln x − 1.
8
(b) This is a very standard application of unconstrained optimisation: see Chapter 5 of the
subject guide and the related readings.
The profit is given by
Π = xpX + ypY − T C
and we want this to be a function of x and y (because the question says so, and because we
need to reduce the problem to a two variable optimisation problem). This means that we
need to find pX and pY in terms of x and y.
One way of doing this is to add the two equations to get
−3pY + 39 − 3x − 3y = 0,
which gives us
pY = 13 − x − y.
Then, substituting this into the first equation we find that
pX = pY + 13 − 3x = 26 − 4x − y.
Putting these into our expression for the profit and simplifying, we find that the profit is
Π = 12x + 6y − 4x2 − y 2 − 2xy,
as a function of x and y.
To find the values of x and y that maximise the profit, we solve the equations Πx = 0 and
Πy = 0, i.e.
Πx = 12 − 8x − 2y = 0 and Πy = 6 − 2y − 2x = 0.
This is easily done and we find that x = 1 and y = 2.
Πxx = −8, Πxy = −2 and Πyy = −2.
Since these assure us that
Πxx Πyy − Π2xy = (−8)(−2) − (−2)2 = 12 > 0 and Πxx < 0,
these values of x and y do indeed maximise the profit. (Note that it is not correct to simply
observe that Πxx < 0 and Πyy < 0: you have to use the test correctly!)
An alternative approach would be to express the profit in terms of pX and pY . This would
be messy in this case and, furthermore, it is not what the question asks for and so it would
not get full credit. But, if you want the gory details, here they are.
We have
Π = xpX + ypY − (2x + y),
and we will write the quantities in terms of the prices. Using the two equations we get
1 1
x= (−pX + pY + 13) and y= (pX − 4pY + 26),
3 3
which means that, putting these into our expression for the profit and simplifying we get
1¡ 2 ¢
Π= −pX + 2pX pY + 20pX − 4p2Y + 40pY − 364 .
3
We now maximise f = 3Π (an equivalent problem to that of maximising Π and better
because we don’t have to deal with awkward fractions) by solving the equations fpX = 0 and
fpY = 0, i.e.
fpX = −2pX + 2pY + 20 = 0 and fpY = 2pX − 8pY + 40 = 0.
This is easily done and we find that pY = 10 and pX = 20 which means, in turn, that x = 1
and y = 2 as before.
The second-order test can then be applied to verify that we have indeed found a maximum.
9
05A Mathematics 1
Now, although this approach gives the right answer, it does not answer the question, which
said, quite explicitly, ‘Find an expression in terms of x and y for the profit function’. So if
used, this method will only get some of the marks!
Some students seem to think that there is a way in which they can deal separately with the
x-profit and the y-profit. This is clumsy and, again, does not answer the question. Indeed,
although this method works in some examples, it won’t work here. What we mean by this is
that they try to find the maximising values of x and y by solving equations like
Marginal x-revenue = Marginal x-cost,
and so, as pX = 26 − 4x − y, we’d need to solve the equation we’d get from
∂ ∂
(x(26 − 4x − y)) = (14x + 7y),
∂x ∂x
and so on. This just does not work. It gives the wrong answer and it is mathematically
incorrect.
Question 9
(a) Find the critical (or stationary) points of the function f (x, y) = x3 − 8xy + 8y 2 − 1 and
determine, for each, whether it is a local maximum, a local minimum, or a saddle point.
(b) Determine the integral Z
3
x5 ex dx.
(c) By simplifying the expression on the right-hand side of the equation, or otherwise, show that
x+2 2 2 3
= − + .
x(x − 1)2 x x − 1 (x − 1)2
Hence find Z
x+2
dx.
x(x − 1)2
(a) This uses the same techniques as Question 4, described in Chapter 5 of the subject guide. It
is, however, a little more complicated.
fx = 3x2 − 8y and fy = −8x + 16y.
To find the critical points, we solve the equations fx = 0 and fy = 0. The second of these
gives x = 2y and substituting this into the first equation gives
2
12y 2 − 8y = 0 =⇒ y(3y − 2) = 0 =⇒ y = 0, .
3
Using x = 2y again, these give us x = 0 and x = 4/3 respectively and so the critical points
are (0, 0) and (4/3, 2/3).
fxx = 6x, fxy = −8 and fyy = 16.
At the critical point (0, 0), this means that we have
2
fxx fyy − fxy = (0)(16) − (−8)2 = −64 < 0,
and so this is a saddle point. Whereas, at the critical point (4/3, 2/3), we have
2
fxx fyy − fxy = (8)(16) − (−8)2 = 64 > 0 and fxx = 8 > 0,
and so this is a local minimum.
Note that it is important to substitute the coordinates of each critical point, in turn, into
the second derivative expressions before applying the second derivative test.
10
(b) This question involves the substitution and integration by parts techniques discussed in
Chapter 4 of the subject guide.
It is sensible to start with a substitution in order to simplify the integral. The obvious
substitution (since it appears as the exponent in the exponential part of the integrand) is
u = x3 . This means that du = 3x2 dx and so
Z Z
3 1
x5 ex dx = ueu du.
3
Now using integration by parts, we find that
Z Z
ueu du = ueu − eu du = ueu − eu + c,
and, putting this all together, we get

Z
3 1 3 1 3
x5 ex dx = x3 ex − ex + c.
3 3
(c) Again, this is an integration problem: see Chapter 4 of the subject guide. It really helps to
follow the lead given in the question, i.e. start with the right-hand side and simplify it to
obtain the left-hand side, not the other way round.
The obvious way to simplify the expression on the right-hand side is to take a common
denominator. We have
2 2 3 2(x − 1)2 − 2x(x − 1) + 3x
− + =
x x − 1 (x − 1)2 x(x − 1)2
2(x − 2x + 1) − 2(x2 − x) + 3x
2
=
x(x − 1)2
x+2
= .
x(x − 1)2
So, using this, we have
Z Z
x+2 2 2 3 3
dx = − + dx = 2 ln |x| − 2 ln |x − 1| − + c.
x(x − 1)2 x x − 1 (x − 1)2 x−1
Question 10
(a) Using a matrix method, solve the following system of equations to find x, y and z.
50 − x − y + 2z = x − 2y + 7z − 10
40 + x + y + z = 3x − 2y − 100
240 + x − y − z = x − 2y − 40.
(b) The function f (x, y) is given by
f (x, y) = cos(nx) sin(y 2 ),
where n is a positive integer. Find
∂f ∂f ∂ 2 f ∂ 2 f
, , , .
∂x ∂y ∂x2 ∂y 2
Find the value of n if f satisfies
∂2f ∂2f ∂f
y3 2
− y 2
+ = 0.
∂x ∂y ∂y
11
05A Mathematics 1
(a) This involves the same techniques as Question 3, described in Chapter 6 of the subject
guide. But, it is important to realise that the given equations are not in the right form for
the standard method of row operations (or other methods, such as Cramer’s rule) to be
applied. Explicitly, the standard methods for solving linear systems assume that the
equations take the form Ax = b where A is a matrix, x is the vector of unknowns (i.e.
variables) and b is a vector whose entries are constants (that is, numbers). The system of
equations in the question has to be rewritten in this standard form before solving.
Having done this, the equations are
2x − y + 5z = 60
2x − 3y − z = 140
−y + z = 280.
and we can now solve them using row operations. The augmented matrix is
 
2 −1 5 60
 2 −3 −1 140 
0 −1 1 280

     
2 −1 5 60 2 −1 5 60 2 −1 5 60
 2 −3 −1 140  →  0 −2 −6 80  →  0 1 3 −40  .
0 −1 1 280 0 −1 1 280 0 0 4 240
So, this gives us
2x − y + 5z = 60, y + 3z = −40 and 4z = 240,

1
z = 60, y = −40 − 3z = −220 and x= (60 + y − 5z) = −230.
2
(b) This uses nothing more than the techniques for partial differentiation discussed in Chapter 5
of the subject guide and, although it may look daunting, it is, in fact, quite straightforward.
The first derivatives are given by
fx = −n sin(nx) sin(y 2 ) and fy = 2y cos(nx) cos(y 2 ).
This means that fxx is given by
fxx = −n2 cos(nx) sin(y 2 ),
and fyy is
fyy = 2 cos(nx) cos(y 2 ) − 4y 2 cos(nx) sin(y 2 ),
where, of course, we have had to use the product rule here.
To find the value of n that makes f satisfy the equation
y 3 fxx − yfyy + fy = 0,
we note that, substituting for the derivatives we have found on the left-hand side, we get the
equation
−n2 y 3 cos(nx) sin(y 2 ) − 2y cos(nx) cos(y 2 ) + 4y 3 cos(nx) sin(y 2 ) + 2y cos(nx) cos(y 2 ) = 0,
and this simplifies to

(4 − n2 )y 3 cos(nx) sin(y 2 ) = 0.
Now, as far as n is concerned, this equation can only be satisfied if 4 − n2 = 0, i.e. if
n = ±2. But, we are also told that n is positive, and so n = 2 is the value that makes f
satisfy this equation.
12
Question 11
(a) Suppose that 0 < a < 1 and that x = x∗ and y = y ∗ minimise the function
(x1/a + y 1/a )a
subject to the constraint px + qy = M . By using the Lagrange multiplier method, show that
µ ¶a/(1−a)
∗ q
y = x∗ .
p
Suppose now that a = 1/2. Show that
Mp Mq
x∗ = , y∗ = 2 .
p2
+q 2 p + q2
(b) An arithmetic progression has the following properties:

• its second term is 7;
• its thirteenth term is ten times its first term.
Determine the first term and the common difference.
(a) This uses the Lagrange multiplier method which we saw in Question 6, described in Chapter
5 of the subject guide. The question might look a bit frightening, but it contains a lot of
useful information, even providing you with the relationship between the optimal values of x
and y so that you can see whether you are on the right track.
The Lagrangean is
L = (x1/a + y 1/a )a − λ(px + qy − M ),
Lx = x(1/a)−1 (x1/a + y 1/a )a−1 − λp = 0,
Ly = y (1/a)−1 (x1/a + y 1/a )a−1 − λq = 0.
x(1/a)−1 (x1/a + y 1/a )a−1 y (1/a)−1 (x1/a + y 1/a )a−1

λ= = ,
p q
so that
y (1/a)−1 q ³ y ´(1−a)/a q
= =⇒ = .
x(1/a)−1 p x p
Raising both sides to the power a/(1 − a) we then get
µ ¶a/(1−a) µ ¶a/(1−a)
y q q
= =⇒ y= x,
x p p
and so the solutions to this constrained optimisation problem, i.e. x∗ and y ∗ , are related in
the required way.
When a = 1/2 we have a/(1 − a) = 1 and so this expression becomes
µ ¶
q
y= x.
p
Then, substituting this into the constraint equation, px + qy = M , gives us

µ ¶
q
px + q x =M =⇒ (p2 + q 2 )x = M p,
p
13
05A Mathematics 1
and so x∗ is
Mp
x= ,
p2 + q2
as required. Then using our expression again we get
µ ¶
q Mp Mq
y= = 2 ,
p p2 + q 2 p + q2
as required, for y ∗ .
(b) For this question, you need to know about arithmetic progressions and these are discussed in
The nth term of an arithmetic progression takes the form a + (n − 1)d and so the fact that
the second term is 7 means that we have
a + d = 7,
whereas the fact that the thirteenth term is ten times the first term means that we have
a + 12d = 10a.
Thus, to determine the first term, a, and the common difference, d, we need to solve the
simultaneous equations
a+d=7 and − 9a + 12d = 0.
This is easily done by noting that a = 4d/3 from the second equation and this means that
the first equation gives
4d 7
+d=7 =⇒ d=7 =⇒ d = 3.
3 3
Then, using a = 4d/3 again, we get a = 4.
14

MT105a Mathematics 1
Important note
This commentary reflects the examination and assessment arrangements for this course in the
Change from next year: no graph paper

There is one minor change to take effect next year. From the 2014 examination, candidates will
not be issued with graph paper. Any sketching of graphs should be done in the answer book.
Graph paper is not needed. Indeed, as we have mentioned often in Examiners’ commentaries,
the plotting of points in order to graph a function is not the correct approach. A sketch of
the graph of a function should indicate its shape, its position with respect to the axes, and its
intercepts on those axes: it need not be drawn to scale. Graph paper is not necessary for this.
Information about the subject guide
Unless otherwise stated, all cross-references will be to the latest version of the subject guide (2011).
You should always attempt to use the most recent edition of any Essential reading textbook, even if
the commentary and/or online reading list and/or subject guide refers to an earlier edition. If
different editions of Essential reading are listed, please check the VLE for reading supplements – if
none are available, please use the contents list and index of the new edition to find the relevant
section.
General remarks
Learning outcomes
By the end of this half course and having completed the Essential reading and activities you should
be able to demonstrate to the Examiners that you should have:
• used the concepts, terminology, methods and conventions covered in the half course to solve
mathematical problems in this subject
• the ability to solve unseen mathematical problems involving understanding of these concepts
and application of these methods
• seen how mathematical techniques can be used to solve problems in economics and related
subjects.
1
Showing your working
We start by emphasising that you should always include your working. This means two things.
First, you should not simply write down the answer in the examination script, but you should
explain the method by which it is obtained. Secondly, you should include rough working (even if it is
messy!). The Examiners want you to get the right answers, of course, but it is more important that
you prove you know what you are doing: that is what is really being examined. We also stress that
if you have not completely solved a problem, you may still be awarded marks for a partial,
incomplete, or slightly wrong, solution; but, if you have written down a wrong answer and nothing
else, no marks can be awarded. So it is certainly in your interests to include all your workings.
Covering the syllabus and choosing questions
You should ensure that you have covered the syllabus in order to perform well in the examination: it
is bad practice to concentrate only on a small range of major topics in the expectation that there
will be lots of marks obtainable for questions on these topics. There are no formal options in this
course: you should study the full extent of the topics described in the syllabus and subject guide. In
particular, since the whole syllabus is examinable, any topic could appear in the examination
questions.
Expectations of the examination paper
Every examination paper is different. You should not assume that your examination will be almost
identical to the previous year’s: for instance, just because there was a question, or a part of a
question, on a certain topic last year, you should not assume there will be one on the same topic this
year. Each year, the Examiners want to test that candidates know and understand a number of
mathematical methods and, in setting an examination paper, they try to test whether the candidate
does indeed know the methods, understands them, and is able to use them, and not merely whether
he or she vaguely remembers them. Because of this, every year there are some questions which are
likely to seem unfamiliar, or different, from previous years’ questions. You should expect to be
surprised by some of the questions. Of course, you will only be examined on material in the syllabus,
so all questions can be answered using the material of the course. There will be enough, routine,
familiar content in the examination so that a candidate who has achieved competence in the course
will pass, but, of course, for a high mark, more is expected: you will have to demonstrate an ability
to solve new and unfamiliar problems.
Answer the question
Please do read the questions carefully. You might be asked to use specific methods, even when other
methods could be used. The purpose of the examination is to test that you know certain methods,
so the Examiners might occasionally ask you to use a specific technique. In such circumstances, only
limited partial credit can be given if you do not use the specified technique. It is also worth reading
the question carefully so that you do not do more than is required (because it is unlikely that you
would get extra marks for doing so). For instance, if a question asked you only to find the critical
points of a function, but not their natures, then you should not determine their natures. Be careful
to read all questions carefully because, although they may look like previous examination questions
on first glance, there can be subtle differences.
Calculators
You are reminded that calculators are not permitted in the examination for this course, under any
circumstances. The Examiners know this, and so they set questions that do not require a calculator.
2
It is a good idea to prepare for this by attempting not to use your calculator as you study and revise
this course.
Question spotting
Many candidates are disappointed to find that their examination performance is poorer
than they expected. This can be due to a number of different reasons and the Examiners’
commentaries suggest ways of addressing common problems and improving your performance.
We want to draw your attention to one particular failing – ‘question spotting’, that is,
confining your examination preparation to a few question topics which have come up in past
papers for the course. This can have very serious consequences.
We recognise that candidates may not cover all topics in the syllabus in the same depth, but
you need to be aware that Examiners are free to set questions on any aspect of the syllabus.
This means that you need to study enough of the syllabus to enable you to answer the required
number of examination questions.
The syllabus can be found in the ‘Course information sheet’ in the section of the VLE dedicated
to this course. You should read the syllabus very carefully and ensure that you cover sufficient
material in preparation for the examination.
Examiners will vary the topics and questions from year to year and may well set questions that
have not appeared in past papers – every topic on the syllabus is a legitimate examination
target. So although past papers can be helpful in revision, you cannot assume that topics or
specific questions that have come up in past examinations will occur again.
If you rely on a question spotting strategy, it is likely you will find yourself in
difficulties when you sit the examination paper. We strongly advise you not to
adopt this strategy.
3

Important note

section.
Comments on specific questions – Zone A
Candidates should answer all EIGHT questions: all SIX questions of Section A (60 marks in total)
and BOTH questions from Section B (20 marks each).
Section A
Question 1
Let f and g be functions with f (x) = 2x2 + 4x + 2 and g(x) = x2 + 2x − 3.
(a) Show that the graphs of the functions f and g do not intersect.
(b) Sketch the graphs of both functions on the same diagram.
(c) For a constant c, let h be the function with h(x) = x2 + 2x + c. Find the values of
c for which the graphs of f and h intersect.
4
Approaching the question
For reading related to this question, see Chapter 2 of the subject guide.
(a) The x-coordinate of a point of intersection would satisfy

2x2 + 4x + 2 = x2 + 2x − 3.
x2 + 2x + 5 = 0,
and this has no solution because the discriminant (‘b2 − 4ac’ in the standard notation),
22 − 4(1)(5), is negative.
(b) The sketch should be as follows (with f being the upper of the two curves):
A fully correct solution should indicate

• intercept of f on x-axis, i.e., x = −1, obtained either by quadratic formula or by
factorisation.
• intercepts of g on x-axis, i.e., x = −3 and x = 1, obtained either by quadratic formula or
by factorisation.
• the U -shape of two non-intersecting quadratic curves.
• their y-intercepts (y = 2 for f and y = −3 for g). (These are not indicated in the above
diagram, for typesetting reasons, but you should mark them on your sketch.)
Note: there is absolutely no need to use graph paper for sketching curves. A sketch is
supposed to indicate the shape and position of curves: it is not meant to be a ‘plot’ of
points on the curve.
(c) There will be an intersection if and only if 2x2 + 4x + 2 = x2 + 2x + c has a solution.
Now,
2x2 + 4x + 2 = x2 + 2x + c ⇐⇒ x2 + 2x + (2 − c) = 0,
and this has a solution if and only if
22 ≥ 4(2 − c).
So we need 4c ≥ 4. That is, c ≥ 1.
5
Question 2
Solve the following system of linear equations, using a matrix method:
x+y+z = 2,
2y + z = 0,
−x + y − z = −4.
the subject guide, especially ‘Cramer’s rule’. It is acceptable to do so in this question, since it
only requires the use of a matrix method. (But, it has to be a matrix method, as the question
makes explicit: manipulation of the equations will not suffice.) However, our view is that the row
operations method is easier and less prone to error, and it is the method that we intend
candidates to learn. It is possible that in future years we might insist explicitly that a
row-operations method be used.
1 1 1 2
(A|b) =  0 2 1 0 .
−1 1 −1 −4
A valid reduction to row-echelon form is as follows:

   
1 1 1 2 1 1 1 2
R3 +R1
(A|b) =  0 2 1 0 −→ 0 2 1 0
−1 1 −1 −4 0 2 0 −2
 
1
R3
1 1 1 2
2
−→ 0 2 1 0
0 1 0 −1
 
1 1 1 2
R3
R2
−→ 0 1 0 −1
0 2 1 0
 
1 1 1 2
R3 −2R2
−→ 0 1 0 −1 .
0 0 1 2
We can stop here, or continue to row-reduced echelon form (RREF):

   
1 1 0 0 1 0 0 1
R1 −R3 R1 −R2
−→ 0 1 0 −1 −→ 0 1 0 −1 .
0 0 1 2 0 0 1 2
(Note that R3
R2 means interchange the second and third rows.)
So, from the row-echelon form, we have
z = 2, y = −1 and x = 2 − (−1) − 2 = 1.
6
mistakes if they sat the paper. So, although there are marks for correct calculation, there are
also marks for using the right method (even if you make a mistake). So, here, for instance, the
Examiners will award marks if you can indicate that you know how to start to solve the
equations (by writing down an augmented matrix); that you know what row operations are; that
you know what it is you want to achieve with row operations (the reduced matrix, that is); and
that you then know how to work from that reduced matrix to determine the required solutions.
There are marks for all these things.
Question 3
Find the critical (or stationary) points of the function f (x) = x6 e−x . Determine the
nature of each critical point.
See Chapter 3 of the subject guide for related reading. We have
f 0 (x) = 6x5 e−x − x6 e−x = x5 e−x (6 − x).
For a critical point, f 0 (x) = 0. Now, f 0 (x) = 0 if and only if x = 0 or 6.
We can determine the nature of the critical point by examining the behaviour of the derivative
around the critical point.
At 0, f 0 (x) changes sign from − to +, so this is a minimum.
At 6, f 0 (x) changes sign from + to −, so this is a maximum.
You might be tempted instead to use the second-derivative test (which, for many problems, is
fairly straightforward, though is less so here). We have
f 00 = 5x4 e−x (6 − x) − x5 e−x (6 − x) − x5 e−x .
(This equals x4 e−x (x2 − 12x + 30), but there is no need to simplify, and indeed no real point.)
We have f 00 (6) < 0 so this is a maximum.
However, we have f 00 (0) = 0 so the test fails. It is not correct to infer from the fact that
f 00 (0) = 0 that 0 is an inflexion. So we must use the sign of the derivative (the method above)
instead. At 0, f 0 (x) changes sign from − to +, so this is a minimum.
Alternatively, it could be observed that f (x) > 0 = f (0) if x < 0 and f (x) > 0 = f (0) if x > 0,
which is enough to show that the point is a minimum.
7
Question 4
Determine the following integrals:

Z
x3 ln x dx,
√
Z
x2 x + 3 dx,
Z
1
dx.
1 + e2x
always clear which technique will work. The three main techniques are: substitution, parts, and
partial fractions. More than one method might work (as we shall see), and some integrals require
a combination of methods.
First integral
x3 ln x dx. This can be solved by a straightforward integration by

R
Consider the first integral,
parts:
x4
Z Z 4
x 1
x3 ln x dx = ln x − dx
4 4 x
x4 1
= ln x − x4 + c.
4 16
Second integral
√
x2 x + 3 dx, is most easily done by substitution, but can also be done by
R
The next integral,
parts.
Method A: by substitution u = x + 3
Let u = x + 3. The integral becomes (u − 3)2 u1/2 du This is

R
Z
2 7/2 12 5/2
(u5/2 − 6u3/2 + 9u1/2 ) du = u − u + 6u3/2 + c
7 5
2 12
= (x + 3)7/2 − (x + 3)5/2 + 6(x + 3)3/2 + c.
7 5
√
Method B: by substitution u = x+3
√
Let u = x + 3. We have
du 1 1
= √ = ,
dx 2 x+3 2u
so dx = 2u du. The integral then becomes (u2 − 3)2 u(2u) du.
R
So the integral is
Z
2 7 12 5
2 (u6 − 6u4 + 9u2 ) du = u − u + 6u3 + c
7 5
2 12
= (x + 3)7/2 − (x + 3)5/2 + 6(x + 3)3/2 + c.
7 5
8
Method C: by parts
√
Z Z
2 4
x2 x + 3 dx = x2 (x + 3)3/2 − x(x + 3)3/2 dx.
3 3
And,
Z Z
2 2
x(x + 3)3/2 dx = x(x + 3)5/2 − (x + 3)5/2 dx
5 5
2 22
= x(x + 3)5/2 − (x + 3)7/2 + c.
5 57
So the integral is
2 2 8 16
x (x + 3)3/2 − x(x + 3)5/2 + (x + 3)7/2 + c.
3 15 105
Third integral
Z
1
The third integral, dx, can be done using a substitution, perhaps after rewriting the
1 + e2x
integral.
Method A: by substituting u = 1 + e2x and using partial fractions
Let u = 1 + e2x .
Then du = 2e2x dx = 2(u − 1) dx and the integral is

Z
1 du
.
2 (u − 1)u
Partial fractions means there are A, B such that the integrand takes the form
A B
+ .
u−1 u
We find (by a number of possible methods) that
A = 1/2, B = −1/2
and hence the integral is

Z
1 1 1 1
− du = ln |u − 1| − ln |u| + c
2(u − 1) 2u 2 2
1 1
= ln(e2x ) − ln(1 + e2x ) + c.
2 2
(Of course, the first term can be simplified to x.)
Method B: multiplying numerator and denominator by e−x first
The integral is
e−x
Z
dx.
e−x + ex
Now we use a substitution. Let u = e−x . Then du = −e−x dx and the integral is
Z Z
u 1 u
− du = − du
u + 1/u u 1 + u2
1
= − ln(1 + u2 ) + c
2
1
= − ln(1 + e−2x ) + c.
2
9
Method C: by multiplying numerator and denominator by e−2x first
The integral is
e−2x
Z
dx.
e−2x +1
Note now that the numerator is −1/2 times the derivative of the denominator, or perform the
substitution u = e−2x + 1, to deduce that the integral is
1
− ln(e−2x + 1) + c.
2
An explicit substitution argument is as follows. We let u = 1 + e−2x . Then du = −2e−2x dx, so

the integral is
Z
1 du 1
− = − ln |u| + c
2 u 2
1
= − − ln(e−2x + 1) + c.
2
Question 5
The function f (x, y) is given by

y
f (x, y) = xy ln ,
x
∂f ∂f
for x, y > 0. Find the partial derivatives and .
∂x ∂y
Verify that
∂f ∂f
x +y = 2f (x, y).
∂x ∂y
Partial differentiation is discussed in Chapter 5 of the subject guide. We have

∂f
= y ln(y/x) + xy(x/y)(−y/x2 ) = y ln(y/x) − y
∂x
and
∂f
= x ln(y/x) + xy(x/y)(1/x) = x ln(y/x) + x.
∂y
Then,
∂f ∂f
x +y = (xy ln(y/x) − xy) + (yx ln(y/x) + yx)
∂x ∂y
= 2xy ln(y/x) = 2f.
Alternatively, you could start by noting that
f (x, y) = xy (ln y − ln x) .
So
∂f
= y(ln y − ln x) + xy (−1/x)
∂x
and
∂f
= x (ln y − ln x) + xy (1/y) .
∂y
10
Question 6
A consumer can buy quantities of two√goods, X and Y . If she buys x units of X and
y units of Y , then her utility is u = y 2 x. Each unit of X and each unit of Y costs 1
dollar. Given that the consumer has a budget of 100 dollars, determine the number
of units of each good she should buy in order to maximize u.
The √
problem is to maximise u subject to x + y = 100. The Lagrangean is
L = xy 2 − λ(x + y − 100).
The first-order conditions are

∂L 1
= x−1/2 y 2 − λ = 0
∂x 2
∂L √
= 2 xy − λ = 0
∂y
∂L
= −(x + y − 100) = 0.
∂λ
√
We next eliminate λ. From the first two equations, λ = (1/2)x−1/2 y 2 = 2 xy. On simplification
(elimination of λ), this becomes y = 4x. Then the third equation shows that x = 20, y = 80.
Section B
Question 7
(a) A graduate starts work with an organisation for a salary in year 1 of $S. At
the beginning of each subsequent year, his salary is increased by 5% plus an
additional amount of $500. (So, for instance, if his salary in year 1 was $20000
then his salary in year 2 will be $21500.) Determine an expression, in as simple a
form as possible, for his salary in year N + 1 of his employment.

In year 1, the salary is S.
In year 2, it is (1.05)S + 500.
In year 3, it is
(1.05)((1.05)S + 500) + 500 = (1.05)2 S + (1.05)500 + 500.
In year 4, it is
(1.05)3 S + (1.05)2 500 + (1.05)500 + 500.
In general, the salary in year N + 1 is
(1.05)N S + (1.05)N −1 500 + (1.05)N −2 500 + · · · + 500.
11
This simplifies (noting the geometric progression) to
1 − (1.05)N
(1.05)N S + 500 = (1.05)N + 10000((1.05)N − 1)
1 − 1.05
or, equivalently, (10000 + S)(1.05)N − 10000.
(b) The function f is given, for x, y 6= 0, by

4 2x
f (x, y) = − + y.
x y
Show that f has one critical (or stationary) point and determine what type of
critical point this is.

4 2
fx = − −
x2 y
2x
fy = + 1.
y2
We solve fx = fy = 0.
fx = 0 means y = −(1/2)x2 . Then, fy = 0 implies 2x/(x2 /2)2 = −1, so x3 = −8 and
x = −2. Then, y = −x2 /2 = −2. Of course, there are other valid ways of solving these
equations.
So we have one critical point, (−2, −2). The second derivatives are
fxx = 8/x3 , fyy = −4x/y 3 , fxy = 2/y 2 .

2
At (−2, −2) we have fxx = −1, fyy = −1, fxy = 1/2. So fxx < 0 and fxx fyy − fxy > 0, so
this is a local maximum point.
Question 8
(a) If a company
√ uses k units of capital and l units of labour, then its output is
Q = k 1/4 l. Each unit of capital costs the company 20 dollars and each unit of
labour costs the company 10 dollars. Use the method of Lagrange multipliers to
find, in terms of Q, the company’s minimum total expenditure on capital and
labour.

This is a constrained optimisation problem, as discussed in Chapter 5 of the subject guide.
We have to minimise 20k + 10l subject to the constraint k 1/4 l1/2 = Q. The Lagrangean is
L = 20k + 10l − λ(k 1/4 l1/2 − Q). The optimal k and l satisfy the equations
∂L 1
= 20 − λk −3/4 l1/2 = 0,
∂k 4
∂L 1
= 10 − λk 1/4 l−1/2 = 0.
∂l 2
These show, on elimination of λ, that 80k 3/4 l−1/2 = 20k −1/4 l1/2 , so that l = 4k. Then,
given that k 1/4 l1/2 = Q, we have
1/2
k 1/4 l1/2 = k 1/4 (4k) = Q.
12
4/3
So 2k 3/4 = Q and so k = (Q/2) . Therefore l = 4k = 4(Q/2)4/3 and the minimum cost is
4/3 4/3 4/3
Q Q Q
20k + 10l = 20 + 40 = 60 .
2 2 2
(b) Show that the only critical point of the function
f (x, y) = x4 − x2 y 2 − y 4
is (0, 0). By considering f (x, 0) and f (0, y), or otherwise, determine the nature of
this critical point.

This question uses the material in Chapter 5 of the subject guide.
4x3 − 2xy 2 = 2x(2x2 − y 2 ) = 0

2 3 2 2
−2x y − 4y = −2y(x + 2y ) = 0.
From the second equation, we have y = 0 or x2 + 2y 2 = 0. (We could, of course, equally well
start with the first equation.) Now, x2 + 2y 2 = 0 implies (x, y) = (0, 0). (And this satisfies
the first equation.) And, if y = 0, the first equation shows x = 0. So the only critical point
is (0, 0).
Note: it is not enough merely to observe that (0, 0) is a critical point because it makes the
derivatives 0: you need to manipulate the equations to show it is the only critical point.
We have f (x, 0) = x4 > 0 if x 6= 0 and f (0, y) = −y 4 < 0 if y 6= 0. Since f (0, 0) = 0, the
point (0, 0) is therefore neither a maximum nor a minimum and is hence a saddle point.
13

Important note

section.
Comments on specific questions – Zone B
Section A
Question 1
Let f and g be functions with f (x) = 2x2 + 4x + 2 and g(x) = x2 + 2x − 3.
(a) Show that the graphs of the functions f and g do not intersect.
(b) Sketch the graphs of both functions on the same diagram.
(c) For a constant c, let h be the function with h(x) = x2 + 2x + c. Find the values of
c for which the graphs of f and h intersect.
14
For reading related to this question, see Chapter 2 of the subject guide.
(a) The x-coordinate of a point of intersection would satisfy

2x2 + 4x + 2 = x2 + 2x − 3.
x2 + 2x + 5 = 0,
and this has no solution because the discriminant (‘b2 − 4ac’ in the standard notation),
22 − 4(1)(5), is negative.
(b) The sketch should be as follows (with f being the upper of the two curves):
A fully correct solution should indicate

• intercept of f on x-axis, i.e., x = −1, obtained either by quadratic formula or by
factorisation.
• intercepts of g on x-axis, i.e., x = −3 and x = 1, obtained either by quadratic formula or
by factorisation.
• the U -shape of two non-intersecting quadratic curves.
• their y-intercepts (y = 2 for f and y = −3 for g). (These are not indicated in the above
diagram, for typesetting reasons, but you should mark them on your sketch.)
Note: there is absolutely no need to use graph paper for sketching curves. A sketch is
supposed to indicate the shape and position of curves: it is not meant to be a ‘plot’ of
points on the curve.
(c) There will be an intersection if and only if 2x2 + 4x + 2 = x2 + 2x + c has a solution.
Now,
2x2 + 4x + 2 = x2 + 2x + c ⇐⇒ x2 + 2x + (2 − c) = 0,
and this has a solution if and only if
22 ≥ 4(2 − c).
So we need 4c ≥ 4. That is, c ≥ 1.
15
Question 2
Solve the following system of linear equations, using a matrix method:
x−y+z = −3,
−3x + 4y − z = 2,
x − 3y − 2z = 7.
operations method is easier and less prone to error, and it is the method that we intend
candidates to learn. It is possible that in future years we might insist explicitly that a
row-operations method be used.
1 −1 1 3
(A|b) = −3 4 −1 2.
1 −3 −2 7
   
1 −1 1 3 1 −1 1 −3
R2 +3R1
(A|b) = −3 4 −1 2 −→ 0 1 2 −7
R3 −R1
1 −3 −2 7 0 −2 −3 10
 
1 −1 1 −3
R3 +2R2
−→ 0 1 2 −7 .
0 0 1 −4
We can stop here, or continue to row-reduced echelon form (RREF):

   
1 −1 0 1 1 0 0 2
R1 −R3 R1 +R2
−→ 0 1 0 1  −→ 0 1 0 1 .
R2 −2R3
0 0 1 −4 0 0 1 −4
From the row-echelon form, we have
z = −4, y = −7 − 2(−4) = 1 and x = −3 − (−4) + 1 = 2.
16
Question 3
Find the critical (or stationary) points of the function f (x) = x4 e−x . Determine the
nature of each critical point.
See Chapter 3 of the subject guide for related reading. We have
f 0 (x) = 4x3 e−x − x4 e−x = x3 e−x (4 − x).
For a critical point, f 0 (x) = 0. Now, f 0 (x) = 0 if and only if x = 0 or 4.
We can determine the nature of the critical point by examining the behaviour of the derivative
around the critical point.
At 0, f 0 (x) changes sign from − to +, so this is a minimum.
At 4, f 0 (x) changes sign from + to −, so this is a maximum.
You might be tempted instead to use the second-derivative test (which, for many problems, is
fairly straightforward, though is less so here). We have
f 00 = 3x2 e−x (4 − x) − x3 e−x (4 − x) − x3 e−x .
(This equals x2 e−x (x2 − 8x + 12), but there is no need to simplify, and indeed no real point.) We
have f 00 (4) < 0 so this is a maximum.
However, we have f 00 (0) = 0 so the test fails. It is not correct to infer from the fact that
f 00 (0) = 0 that 0 is an inflexion. So we must use the sign of the derivative (the method above)
instead. At 0, f 0 (x) changes sign from − to +, so this is a minimum.
Alternatively, it could be observed that f (x) > 0 = f (0) if x < 0 and f (x) > 0 = f (0) if x > 0,
which is enough to show that the point is a minimum.
Question 4

Z
1
dx,
x2 − 3x + 2
Z p √
1+ x
√ dx,
x
Z
x3 ln(x2 − 1) dx.
17
always clear which technique will work. The three main techniques are: substitution, parts, and
partial fractions. More than one method might work (as we shall see), and some integrals require
a combination of methods.
First integral
Z
1
Consider the first integral, dx. This is a straightforward partial fractions integral:
x2 − 3x + 2
1 1
= .
x2 − 3x + 2 (x − 2)(x − 1)
So, there are constants A, B such that
1 A B
= + .
x2 − 3x + 2 x−2 x−1
We find (by a number of possible methods) that
A = 1, B = −1
and hence the integral is

ln |x − 2| − ln |x − 1| + c.
Second integral
Z p √
1+ x
The next integral, √ dx, is most easily done by substitution, and there are various
x
approaches.
√
Method A: by substitution u = 1 + x
√
let u = 1 + x. Then
1
du = √ dx
2 x
R√
and the integral becomes 2 u du. This is
4 3/2 4 √
u + c = (1 + x)3/2 + c.
3 3
p √
Method B: by substitution u = 1+ x
p √
Let u = 1 + x. Then
1
du = √ p √ dx
4 x 1+ x
u2 du. This is
R
and the integral becomes 4
4 3 4 √
u + c = (1 + x)3/2 + c.
3 3
√
Method C: by substitution u = x
√
Let u = x. Then
1
du = √ dx
2 x
R√
and the integral becomes 2 1 + u du. This is
4 4 √
(1 + u)3/2 + c = (1 + x)3/2 + c.
3 3
18
Third integral
The third integral, x3 ln(x2 − 1) dx, can be done using a substitution, followed by integration
R
by parts. (It can also be done directly by integration by parts, but we won’t present that
approach here.)
Method A: substitute u = x2 − 1
Let u = x2 − 1. Then du = 2xdx and the integral is

Z
1
(u + 1) ln u du.
2
Now we can use parts.
u2
Z Z 2
1 1 1 u 1
(u + 1) ln u du = + u ln u − +u du
2 2 2 2 2 u
u2

1 2 1 u
= u + u ln u − − + c.
4 2 8 2
So the answer is
(x2 − 1)2 x2 − 1

1 2 1 2
(x − 1) + (x − 1) ln(x2 − 1) −
2
− + c.
4 2 8 2
Method B: substitute u = x2
Let u = x2 . Then du = 2xdx and the integral is

Z
1
u ln(u − 1) du.
2
Now we can use parts.
u2
Z Z
1 1 2 1
u ln(u − 1) du = u ln(u − 1) − du.
2 4 4 u−1
Now we substitute v = u − 1 to see that
u2 (v + 1)2
Z Z Z 2 Z
v + 2v + 1 1
du = dv = dv = v + 2 + dv.
u−1 v v v
This is
v2 (u − 1)2
+ 2v + ln |v| + c = + 2(u − 1) + ln |u − 1|.
2 2
So the answer is
1 2 (u − 1)2 1 1
u ln(u − 1) − + (u − 1) + ln |u − 1| + c
4 8 2 4
1 4 (x2 − 1)2 1 1
= x ln(x2 − 1) − + (x2 − 1) + ln |x2 − 1| + c.
4 8 2 4
Question 5
The function f (x, y) is given by

x
f (x, y) = xy ln ,
y
∂f ∂f
for x, y > 0. Find the partial derivatives and .
∂x ∂y
19
Verify that
∂f ∂f
x +y = 2f (x, y).
∂x ∂y
Partial differentiation is discussed in Chapter 5 of the subject guide. We have
∂f
= y ln(x/y) + xy(y/x)(1/y) = x ln(x/y) + y
∂x
and
∂f
= x ln(x/y) + xy(y/x)(−x/y 2 ) = x ln(x/y) − x.
∂y
Then,
∂f ∂f
x +y = (xy ln(x/y) + xy) + (yx ln(x/y) − yx)
∂x ∂y
= 2xy ln(y/x) = 2f.
Alternatively, you could start by noting that
f (x, y) = xy (ln x − ln y) .
So
∂f
= y(ln y − ln x) + xy (−1/x)
∂x
and
∂f
= x (ln y − ln x) + xy (1/y) .
∂y
Question 6
A consumer can buy quantities of two goods, X and Y . If she buys x units of X and
y units of Y , then her utility is u = xy 3/2 . Each unit of X costs 1 dollar and each unit
of Y costs 2 dollars. Given that the consumer has a budget of 100 dollars, determine
the number of units of each good she should buy in order to maximize u.
The problem is to maximise u subject to x + 2y = 100. The Lagrangean is

L = xy 3/2 − λ(x + 2y − 100).

∂L
= y 3/2 − λ = 0
∂x
∂L 3
= xy 1/2 − 2λ = 0
∂y 2
∂L
= −(x + 2y − 100) = 0.
∂λ
We next eliminate λ. From the first two equations, λ = y 3/2 = 43 xy 1/2 . On simplification
(elimination of λ), this becomes y = (3/4)x. Then the third equation shows that x = 40, y = 30.
20
Section B
Question 7
(a) A graduate starts work with an organisation for a salary in year 1 of $S. At
the beginning of each subsequent year, his salary is increased by 5% plus an
additional amount of $500. (So, for instance, if his salary in year 1 was $20000
then his salary in year 2 will be $21500.) Determine an expression, in as simple a
form as possible, for his salary in year N + 1 of his employment.

In year 1, the salary is S.
In year 2, it is (1.05)S + 500.
In year 3, it is
(1.05)((1.05)S + 500) + 500 = (1.05)2 S + (1.05)500 + 500.
In year 4, it is
(1.05)3 S + (1.05)2 500 + (1.05)500 + 500.
In general, the salary in year N + 1 is
(1.05)N S + (1.05)N −1 500 + (1.05)N −2 500 + · · · + 500.
1 − (1.05)N
(1.05)N S + 500 = (1.05)N + 10000((1.05)N − 1)
1 − 1.05
or, equivalently, (10000 + S)(1.05)N − 10000.
(b) The function f is given, for x, y 6= 0, by
4 2y
f (x, y) = x + − .
y x
Show that f has one critical (or stationary) point and determine what type of
critical point this is.

2y
fx = 1 +
x2
4 2
fy = − 2 − .
y x
We solve fx = fy = 0.
fy = 0 means x = −(1/2)y 2 . Then, fx = 0 implies 2y/(y 2 /2)2 = −1, so y 3 = −8 and y = −2.
Then, x = −y 2 /2 = −2. Of course, there are other valid ways of solving these equations.
21
So we have one critical point, (−2, −2). The second derivatives are
fxx = −4y/x3 , fyy = 8/y 3 , fxy = 2/x2 .
2
At (−2, −2) we have fxx = −1, fyy = −1, fxy = 1/2. So fxx < 0 and fxx fyy − fxy > 0, so
this is a local maximum point.
Question 8
(a) If a company
√ uses k units of capital and l units of labour, then its output is
Q = k 1/3 l. Each unit of capital costs the company 20 dollars and each unit of
labour costs the company 10 dollars. Use the method of Lagrange multipliers to
find, in terms of Q, the company’s minimum total expenditure on capital and
labour.

This is a constrained optimisation problem, as discussed in Chapter 5 of the subject guide.
√
We have to minimise 20k√ + 10l subject to the constraint k 1/3 l = Q. The Lagrangean is
L = 20k + 10l − λ(k 1/3 l − Q). The optimal k and l satisfy the equations
∂L 1
= 20 − λk −2/3 l1/2 = 0,
∂k 3
∂L 1
= 10 − λk 1/3 l−1/2 = 0.
∂l 2
These show, on elimination of λ, that 60k 2/3 l−1/2 = 20k −1/3 l1/2 , so that l = 3k. Then,
given that k 1/3 l1/2 = Q, we have
1/2
k 1/3 l1/2 = k 1/3 (3k) = Q.
So 31/2 k 5/6 = Q and so k = 3−3/5 Q6/5 . Therefore l = 3k = 3 3−3/5 Q6/5 and the minimum
cost is
20k + 10l = 20 3−3/5 Q6/5 + 30 3−3/5 Q6/5 = 50 3−3/5 Q6/5 .
(b) Show that the only critical point of the function

f (x, y) = −x4 − x2 y 2 + y 4
is (0, 0). By considering f (x, 0) and f (0, y), or otherwise, determine the nature of
this critical point.

−4x3 − 2xy 2 = −2x(2x2 + y 2 ) = 0
2 3 2 2
−2x y + 4y = 2y(2y − x ) = 0.
2 2
From the first equation, we have x = 0 or 2x + y = 0. (We could, of course, equally well
start with the second equation.) Now, 2x2 + 2y 2 = 0 implies (x, y) = (0, 0). (And this
satisfies the second equation.) And, if x = 0, the second equation shows y = 0. So the only
critical point is (0, 0).
Note: it is not enough merely to observe that (0, 0) is a critical point because it makes the
derivatives 0: you need to manipulate the equations to show it is the only critical point.
We have f (x, 0) = −x4 < 0 if x 6= 0 and f (0, y) = y 4 > 0 if y 6= 0. Since f (0, 0) = 0, the
point (0, 0) is therefore neither a maximum nor a minimum and is hence a saddle point.
22

Important note
section.
General remarks
Learning outcomes
subjects.
1
questions.
Answer the question
Graph sketching
Some examinations in this subject ask you to sketch the graph of a function (though this was not
the case in 2014). Any sketching of graphs should be done in the answer book. Graph paper is not
needed. Indeed, as we have mentioned often in Examiners’ commentaries, the plotting of points in
order to graph a function is not the correct approach. A sketch of the graph of a function should
indicate its shape, its position with respect to the axes, and its intercepts on those axes: it need not
be drawn to scale. Graph paper is not necessary for this.
Calculators
this course.
2
Question spotting
Many candidates are disappointed to find that their examination performance is poorer
than they expected. This can be due to a number of different reasons and the Examiners’
commentaries suggest ways of addressing common problems and improving your performance.
We want to draw your attention to one particular failing – ‘question spotting’, that is,
confining your examination preparation to a few question topics which have come up in past
papers for the course. This can have very serious consequences.
We recognise that candidates may not cover all topics in the syllabus in the same depth, but
you need to be aware that Examiners are free to set questions on any aspect of the syllabus.
This means that you need to study enough of the syllabus to enable you to answer the required
number of examination questions.
The syllabus can be found in the ‘Course information sheet’ in the section of the VLE dedicated
to this course. You should read the syllabus very carefully and ensure that you cover sufficient
material in preparation for the examination.
Examiners will vary the topics and questions from year to year and may well set questions that
have not appeared in past papers – every topic on the syllabus is a legitimate examination
target. So although past papers can be helpful in revision, you cannot assume that topics or
specific questions that have come up in past examinations will occur again.
If you rely on a question spotting strategy, it is likely you will find yourself in
difficulties when you sit the examination paper. We strongly advise you not to
adopt this strategy.
3

Important note
section.
Section A
Question 1
Find the critical points of the function f given by
f (x) = 3x4 + 4x3 − 12x2 + 10.
For each critical point of f , determine whether it is a local maximum, local

minimum or inflexion point.
See Chapter 3 of the subject guide for related reading. We have:
f 0 (x) = 12x3 + 12x2 − 24x.
We solve f 0 = 0; that is:

f 0 = 12x(x2 + x − 2) = 0.
This is 12x(x − 1)(x + 2) = 0, so x = 0, 1 or −2.
There are two standard methods by which we can determine the nature of the critical points.
Either one is fine.
4
Method A We can use the second derivative. We have:
f 00 (x) = 12(3x2 + 2x − 2).
This is negative at 0, so this is a local maximum, and positive at −2 and 1, so these are local
minima.
Method B By examining the behaviour of the sign of f 0 around the points, we see that f 0
changes from positive to negative at 0, changes from negative to positive at −2, and from
negative to positive at 1. So we draw the same conclusions as to the natures of the points.
Question 2
√
Z
(a) Determine x x + 2 dx.

Integration is discussed in Chapter 4 of the subject guide. It can be difficult because it is
not always clear which technique will work. The three main techniques are: substitution,
parts, and partial fractions. More than one method might work (as we shall see), and some
integrals require a combination of methods.
There are at least three ways to perform this integration.
Method A: by substitution u = x + 2.
R √ R √
Let u = x + 2. Then the integral becomes I = x x + 2 dx = (u − 2) u du. This is:
Z
2 4
u3/2 − 2u1/2 du = u5/2 − u3/2 + c
5 3
which, on recalling that u = x + 2, gives the answer:
2 4
(x + 2)5/2 − (x + 2)3/2 + c.
5 3
√
Method B: by substitution u = x + 2.
√
Let u = x + 2. Then we have:
du 1 1
= √ = ,
dx 2 x+2 2u
so dx = 2u du. The integral then becomes (u2 − 2)u(2u) du.

R
So the integral is 2 (u4 − 2u2 ) du. This is:

R
2 5 4 3 2 4
u − u + c = (x + 2)5/2 − (x + 2)3/2 + c.
5 3 5 3
Method C: by parts.
Since the integrand is a product, we could try parts:
√
Z Z
2 2
x x + 2 dx, = x(x + 2)3/2 − (x + 2)3/2 dx.
3 3
And: Z
2
(x + 2)3/2 dx = (x + 2)5/2 + c.
5
So the answer is:
2 4
x(x + 2)3/2 − (x + 2)5/2 + c.
3 15
5
(b) Show that

2x x3 − x2
x+ = 2 .
(x + 1)(x − 2) x −x−2
Use this to determine
x3 − x2
Z
dx.
x2 − x − 2
We have:
2x 2x x(x2 − x − 2) + 2x x3 − x2
x+ =x+ 2 = 2
= 2 .
(x + 1)(x − 2) x −x−2 x −x−2 x −x−2
Then:
x3 − x2 x2
Z Z
2x
dx = x+ dx = +J
x2 − x − 2 (x + 1)(x − 2) 2
where, by partial fractions: Z
A B
J= + dx
x+1 x−2
for some A, B. We find (by one of the usual methods) that A = 2/3, B = 4/3, so:
Z
2/3 4/3 2 4
J= + dx = ln |x + 1| + ln |x − 2| + c.
x+1 x−2 3 3
So the answer is:
x2 2 4
+ ln |x + 1| + ln |x − 2| + c.
2 3 3
Question 3
Use the method of Lagrange multipliers to find the values of x and y that maximise
√
(x + 2) y and satisfy the equation x + 2y = 10.
The Lagrangian is:

√
L = (x + 2) y − λ(x + 2y − 10).
The first order conditions are:
√
Lx = y−λ=0
x+2
Ly = √ − 2λ = 0
2 y
Lλ = −(x + 2y − 10) = 0.
Eliminating λ, we have:
√ x+2
λ= y= √
4 y
showing that x = 4y − 2. (Or, you could express y in terms of x.)
Then x + 2y = 10 becomes 6y = 12 and so y = 2 and x = 6.
6
Question 4
For some number a, let f be the function of two variables given by
f (x, y) = 2xy + xa y 2a .
Find
∂f ∂f ∂ 2 f ∂ 2 f
, , ,
∂x ∂y ∂x2 ∂y 2
in terms of x, y and a. Given that f satisfies
∂2f 2
2∂ f
2x2 − y + 18f = 36xy,
∂x2 ∂y 2
determine the possible values of a.
It is important in answering questions like this to use a correct notation for partial derivatives.
∂2f
We can write fx instead of ∂f
∂x and fxx instead of ∂x2 , and so on, but it is a bad idea to invent
∂2f
your own notation! In particular, writing f (x, x) or f 0 (x, x) for ∂x2 is just plain wrong.
The derivatives are:
fx = 2y + axa−1 y 2a
fy = 2x + 2axa y 2a−1
fxx = a(a − 1)xa−2 y 2a
fyy = 2a(2a − 1)xa y 2a−2 .
So:
∂2f ∂2f
2x2 2
− y 2 2 + 18f = 2a(a − 1)xa y 2a − 2a(2a − 1)xa y 2a + 36xy + 18xa y 2a .
∂x ∂y
This equals 36xy if and only if:
xa y 2a [2a(a − 1) − 2a(2a − 1) + 18] = 0.
This is equivalent to:

2a(a − 1) − 2a(2a − 1) + 18 = 0
or, on simplifying:
9 − a2 = 0
which has solutions a = ±3.
Question 5
Find the critical points of the function f of two variables given by

f (x, y) = (x2 − x + 1)(y 3 − 3y). Determine the nature of each critical point.
7
f (x, y) = (x2 − x + 1)(y 3 − 3y), so we solve:
fx = (2x − 1)(y 3 − 3y) = 0. fy = (x2 − x + 1)(3y 2 − 3) = 0.
x2 − x + 1 = 0 has no solution (because it has a negative discriminant). (Recall that the

discriminant of the quadratic equation ax2 + bx + c = 0 is the quantity b2 − 4ac, and if this is
negative, the equation has no solutions.) So the second equation shows y = ±1. The first then
gives x = 1/2. There are therefore two stationary points: (1/2, 1) and (1/2, −1).
To determine the nature of each critical point, we use the Hessian. We have:
fxx = 2(y 3 − 3y) = 2y(y 2 − 3), fyy = 6y(x2 − x + 1), fxy = (2x − 1)(3y 2 − 3).
The Hessian is:
H = fxx fyy − (fxy )2 = 12y 2 (y 2 − 3)(x2 − x + 1) − (2x − 1)2 (3y 2 − 3)2 ,
which, at the critical points, equals:
12y 2 (y 2 − 3)(x2 − x + 1).
At (1/2, ±1), H = 12(−2)(3/4) < 0 and so both points are saddle points.
Alternatively, for this last part, we could just note that x2 − x + 1 > 0, and that fxy = 0 at the
stationary points, so we can see that both are saddle points because fxx and fyy have opposite
signs and hence H = fxx fyy − (fxy )2 will be negative.
Question 6
A sequence of numbers x0 , x1 , x2 , . . . is such that x0 = 4 and, for each n ≥ 1,

1
xn = xn−1 + 2.
3
Find an explicit expression for xn in terms of n. Hence find the behaviour of xn as
n → ∞.
Chapter 7 of the subject guide gives the required background material. This question was
intended to be extremely straightforward, but was, in fact, extremely badly done. The required
method is exactly that demonstrated in the Example in Section 7.8 of the subject guide and uses
the key method of that chapter. Rather than have the core mathematical problem embedded in a
complicated financial or other application, the underlying mathematical problem was explicitly
given. This should therefore have been easier to answer than questions involving complicated
financial applications. Perhaps the difficulty was that this was not the type of question that
candidates anticipated. See the advice on question spotting in the General remarks section of the
commentary.
We have:
x0 = 4

1
x1 = 4+2
3
2
1 1
x2 = 4+ 2+2
3 3
3 2
1 1 1
x3 = 4+ 2+ 2 + 2.
3 3 3
8
n n−1 !
1 1 1
In general, xn = 4 +2 + ··· + + 1 . This simplifies to:
3 3 3
n n n n
1 − (1/3)n

1 1 1 1
4 +2 =4 +3 1− =3+ .
3 1 − 1/3 3 3 3
As n → ∞, because (1/3)n tends to 0 and decreases, the sequence decreases to 3.
Section B
Question 7
(a) For some numbers a, b and c, let f be the function of one variable given by
b
f (x) = ax3 + + c(1 + ln x),
x
for x > 0. Given that
f (1) = 0, f 0 (1) = 20, f 00 (1) = 40,
find a system of linear equations for a, b and c. Find a, b and c by solving the
system using a matrix method.

We have:
b
f (x) = ax3 + + c(1 + ln x)
x
so:
b c
f 0 (x) = 3ax2 − 2
+
x x
2b c
f 00 (x) = 6ax + 3 − 2 .
x x
The given information means that:
a+b+c = 0
3a − b + c = 20
6a + 2b − c = 40.
method, the row-reduction method, and so on. Many candidates like to use a different
method, not covered in the subject guide, especially ‘Cramer’s rule’. It is acceptable to do
so in this question, since it only requires the use of a matrix method. (But, it has to be a
matrix method, as the question makes explicit: manipulation of the equations will not
suffice.) However, our view is that the row operations method is easier and less prone to
error, and it is the method that we intend candidates to learn. It is possible that in
future years we might sometimes insist explicitly that a row operations method
be used, in which case using Cramer’s rule or some other method would not be
answering the question. If we require the use of the row operation method we shall use
the phrase ‘Use row operations’.
The standard matrix method approach is now to reduce the augmented matrix to reduced
form.
9
The augmented matrix is:  

1 1 1 0
(A | b) = 3 −1 1 20 .
6 2 −1 40

   
1 1 1 0 1 1 1 0
R2 −3R1 ,R3 −6R1
(A | b) = 3 −1 1 20 −→ 0 −4 −2 20
6 2 −1 40 0 −4 −7 40
 
1 1 1 0
R3 /(−2)
−→ 0 2 1 −10
0 −4 −7 40
 
1 1 1 0
(R3 +2R2 )/(−5)
−→ 0 2 1 −10 .
0 0 1 −4
So, from the row-echelon form, we have:
c = −4, b = (−10 − c)/2 = −3 and a = −b − c = 7.
This is probably a good point at which to make some general comments about how
questions are marked. Clearly, in a question like this, it is easy to get the wrong answer.
(Though it should be noted that in this particular question, you can always substitute the
values that you have found into the original equations, and this will show whether these are
then re-work the calculation.) Examiners understand that arithmetical errors can be made,
especially in the stressful circumstances of an examination. Quite probably, the Examiners
themselves would make some mistakes if they sat the paper. So, although there are marks
for correct calculation, there are also marks for using the right method (even if you make a
mistake). So, here, for instance, the Examiners will award marks if you can indicate that
you know how to start to solve the equations (by writing down an augmented matrix); that
you know what row operations are; that you know what it is you want to achieve with row
operations (the reduced matrix, that is); and that you then know how to work from that
reduced matrix to determine the required solutions. There are marks for all these things.
In particular, a number of candidates make the mistake of thinking that subtracting a fixed
(b) A firm is a monopoly and the demand equation for the good it produces is
p + bq = a, where p is price, q is the quantity and a and b are positive numbers.
The firm’s fixed costs are 0 and the marginal cost function takes the positive
constant value k, where k < a. Determine the firm’s profit function, in terms of
q. Also find the maximum profit the firm can make, in terms of a, b and k,
simplifying your answer as much as possible.

The total cost is: Z
TC = M C dq = kq + c
where c is a constant. Given that T C(0) = F C = 0, we must have c = 0. The firm is a

monopoly, so its total revenue is T R = qpD (q) = q(a − bq). It follows that the profit is:
Π(q) = q(a − bq) − kq = (a − k)q − bq 2 .
We solve Π0 = 0, which is:

(a − k) − 2bq = 0
10
so q = (a − k)/(2b). This indeed gives a maximum since Π00 (q) = −2b < 0. Thus the
maximum profit is:
2
(a − k)2 (a − k)2

a−k a−k
Π = −b = .
2b 2b 2b 4b
Question 8
A firm has weekly production function q given by
q(k, l) = k 1/3 l1/3 ,

where k denotes the amount of capital and l the amount of labour. Each unit of
capital costs V dollars and each unit of labour costs W dollars. Show that the
minimum amount (in dollars) that the firm can spend in a week on capital and
labour if it manufactures a total weekly amount Q is
√
2 V W Q3/2 .
Now suppose that to produce each unit of the good, it costs the firm R dollars in
raw material costs. Suppose the good sells for a fixed price of P dollars per unit,
where P > R. Assuming that the only production costs are capital, labour and raw
materials, and that the firm minimizes its weekly expenditure on capital and labour,
show that the firm’s weekly profit Π(Q) is given by
√
Π(Q) = (P − R)Q − 2 V W Q3/2 .
Show that the maximum weekly profit is
(P − R)3
.
27 V W
The problem is to minimise V k + W l subject to the constraint k 1/3 l1/3 = Q.
The Lagrangian is:

L = V k + W l − λ(k 1/3 l1/3 − Q).
The first-order conditions are:
1
Lk = V − λk −2/3 l1/3 = 0
3
1
Ll = W − λk 1/3 l−2/3 = 0
3
1/3 1/3
Lλ = −(k l − Q) = 0.
From the first two equations, eliminating λ, we have:
k −2/3 l1/3 k 1/3 l−2/3

= .
V W
So l = (V /W )k.
11
Hence, by the third equation (the constraint):

1/3
V
k 2/3 = Q,
W
and so: r
W 3/2
k= Q
V
and: r
V W 3/2
l= Q .
W V
The minimum cost is therefore:
√
r r
W 3/2 V W 3/2
V k + Wl = V Q +W Q = 2 V W Q3/2 .
V W V
Using the given information, the profit function is:

√
Π(Q) = P Q − RQ + 2 V W Q3/2 .
Setting Π0 = 0 gives: √
P − R − 3 V W Q1/2 = 0,
which gives:
(P − R)2
Q= .
9V W
We have:
3√
Π00 = − V W Q−1/2
2
which is negative, so this is indeed going to give a maximum. The maximum value is:
(P − R)2

Π
9V W
This is:
(P − R)3 √ (P − R)3 (P − R)3 (P − R)3 (P − R)3
−2 VW √ = −2 = .
9V W 93/2 V W V W 9V W 27V W 27V W
12

Important note
section.
Section A
Question 1
Find the critical points of the function f given by
f (x) = 3x4 + 4x3 − 12x2 + 20.
For each critical point of f , determine whether it is a local maximum, local

minimum or inflexion point.
See Chapter 3 of the subject guide for related reading. We have:
f 0 (x) = 12x3 + 12x2 − 24x.
We solve f 0 = 0; that is:

f 0 = 12x(x2 + x − 2) = 0.
This is 12x(x − 1)(x + 2) = 0, so x = 0, 1 or −2.
There are two standard methods by which we can determine the nature of the critical points.
Either one is fine.
13
Method A We can use the second derivative. We have:
f 00 (x) = 12(3x2 + 2x − 2).
This is negative at 0, so this is a local maximum, and positive at −2 and 1, so these are local
minima.
Method B By examining the behaviour of the sign of f 0 around the points, we see that f 0
changes from positive to negative at 0, changes from negative to positive at −2, and from
negative to positive at 1. So we draw the same conclusions as to the natures of the points.
Question 2
√
Z
(a) Determine x x + 1 dx.

Integration is discussed in Chapter 4 of the subject guide. It can be difficult because it is
not always clear which technique will work. The three main techniques are: substitution,
parts, and partial fractions. More than one method might work (as we shall see), and some
integrals require a combination of methods.
There are at least three ways to perform this integration.
Method A: by substitution u = x + 1.
R √ R √
Let u = x + 2. Then the integral becomes I = x x + 1 dx = (u − 1) u du. This is:
Z
2 2
u3/2 − u1/2 du = u5/2 − u3/2 + c
5 3
which, on recalling that u = x + 1, gives the answer:
2 2
(x + 1)5/2 − (x + 1)3/2 + c.
5 3
√
Method B: by substitution u = x + 1.
√
Let u = x + 1. Then we have:
du 1 1
= √ =
dx 2 x+1 2u
R 2
so dx = 2u du. The integral then becomes (u − 1)u(2u) du.
So the integral is 2 (u4 − u2 ) du This is:
R
2 5 2 3 2 2
u − u + c = (x + 2)5/2 − (x + 2)3/2 + c.
5 3 5 3
Method C: by parts.
Since the integrand is a product, we could try parts:
√
Z Z
2 2
x x + 2 dx = x(x + 1)3/2 − (x + 1)3/2 dx.
3 3
Also: Z
2
(x + 1)3/2 dx = (x + 1)5/2 + c.
5
So the answer is:
2 4
x(x + 1)3/2 − (x + 1)5/2 + c.
3 15
14
(b) Show that

2x x3 + x2
x+ = 2 .
(x − 1)(x + 2) x +x−2
Use this to determine
x3 + x2
Z
dx.
x2 +x−2

We have:
2x 2x x(x2 + x − 2) + 2x x3 + x2
x+ =x+ 2 = = .
(x − 1)(x + 2) x +x−2 x2 − x − 2 x2 + x − 2
Then:
x3 + x2 x2
Z
2x
dx = x + = +J
x2 + x − 2 (x − 1)(x + 2) 2
where, by partial fractions: Z
A B
J =J = + dx
x−1 x+2
for some A, B. We find (by one of the usual methods) that A = 2/3, B = 4/3, so:
Z
2/3 4/3 2 4
J= + dx = ln |x − 1| + ln |x + 2| + c.
x−1 x+2 3 3
So the answer is:
x2 2 4
+ ln |x − 1| + ln |x + 2| + c.
2 3 3
Question 3
Use the method of Lagrange multipliers to find the values of x and y that maximise
√
(x + 1) y and satisfy the equation x + 2y = 11.
The Lagrangian is:

√
L = (x + 1) y − λ(x + 2y − 11).
The first order conditions are:
√
Lx = y−λ=0
x+1
Ly = √ − 2λ = 0
2 y
Lλ = −(x + 2y − 11) = 0.
√ x+1
λ= y= √
4 y
showing that x = 4y − 1. (Or, you could express y in terms of x.)
Then x + 2y = 11 becomes 6y = 12 and so y = 2 and x = 7.
15
Question 4
For some number a, let f be the function of two variables given by
f (x, y) = 2xy + x2a y a .
Find
∂f ∂f ∂ 2 f ∂ 2 f
, , ,
∂x ∂y ∂x2 ∂y 2
in terms of x, y and a. Given that f satisfies
∂2f 2
2∂ f
x2 − 2y − 18f + 36xy = 0,
∂x2 ∂y 2
determine the possible values of a.
∂2f
∂2f
The derivatives are:
fx = 2y + 2ax2a−1 y a
fy = 2x + ax2a y a−1
fxx = 2a(2a − 1)x2a−2 y a
fyy = a(a − 1)x2a y a−2 .
So:
∂2f ∂2f
x2 2
− 2y 2 2 − 18f + 36xy
∂x ∂y
= 2a(2a − 1)x2a y a − 2a(a − 1)x2a y a − 36xy − 18x2a y a + 36xy.
This equals 0 if and only if:
x2a y a [2a(2a − 1) − 2a(a − 1) − 18] = 0.

2a(2a − 1) − 2a(a − 1) − 18 = 0
or, on simplifying:
a2 − 9 = 0
which has solutions a = ±3.
Question 5
Find the critical points of the function f of two variables given by

f (x, y) = (x2 − x + 1)(y 3 − 12y). Determine the nature of each critical point.
16
f (x, y) = (x2 − x + 1)(y 3 − 12y), so we solve:

fx = (2x − 1)(y 3 − 12y) = 0. fy = (x2 − x + 1)(3y 2 − 12) = 0.
x2 − x + 1 = 0 has no solution (because it has a negative discriminant). (Recall that the
discriminant of the quadratic equation ax2 + bx + c = 0 is the quantity b2 − 4ac, and if this is
negative, the equation has no solutions.) So the second equation shows y = ±2. The first then
gives x = 1/2. There are therefore two stationary points: (1/2, 2) and (1/2, −2).
To determine the nature of each critical point, we use the Hessian. We have:
fxx = 2(y 3 − 12y) = 2y(y 2 − 12), fyy = 6y(x2 − x + 1), fxy = (2x − 1)(3y 2 − 12).
The Hessian is:
H = fxx fyy − (fxy )2 = 12y 2 (y 2 − 12)(x2 − x + 1) − (2x − 1)2 (3y 2 − 12)2
which, at the critical points, equals:
12y 2 (y 2 − 12)(x2 − x + 1).
At (1/2, ±2), H = 12(4)(−8)(3/4) < 0 and so both points are saddle points.
Alternatively, for this last part, we could just note that x2 − x + 1 > 0, and that fxy = 0 at the
stationary points, so we can see that both are saddle points because fxx and fyy have opposite
signs and hence H = fxx fyy − (fxy )2 will be negative.
Question 6

1
xn = xn−1 + 2.
2
Find an explicit expression for xn in terms of n. Hence find the behaviour of xn as
n → ∞.
Chapter 7 of the subject guide gives the required background material. This question was
intended to be extremely straightforward, but was, in fact, extremely badly done. The required
method is exactly that demonstrated in the Example in Section 7.8 of the subject guide and uses
the key method of that chapter. Rather than have the core mathematical problem embedded in a
complicated financial or other application, the underlying mathematical problem was explicitly
given. This should therefore have been easier to answer than questions involving complicated
financial applications. Perhaps the difficulty was that this was not the type of question that
candidates anticipated. See the advice on question spotting in the General remarks section of the
commentary.
We have:
x0 = 3

1
x1 = 3+2
2
2
1 1
x2 = 3+ 2+2
2 2
3 2
1 1 1
x3 = 3+ 2+ 2 + 2.
2 2 2
17
In general: !
n n−1
1 1 1
xn = 3 +2 + ··· + +1 .
2 2 2
This simplifies to:

n n n n
1 − (1/2)n

1 1 1 1
3 +2 =3 +4 1− =4− .
3 1 − 1/2 3 2 2
As n → ∞, because (1/2)n tends to 0 and decreases, the sequence increases to 3.
Section B
Question 7
(a) For some numbers a, b and c, let f be the function of one variable given by
b
f (x) = ax3 + + c(1 + ln x),
x
for x > 0. Given that
f (1) = 0, f 0 (1) = 20, f 00 (1) = 40,
find a system of linear equations for a, b and c. Find a, b and c by solving the
system using a matrix method.

We have: √
f (x) = ax2 + bx + c x
so:
c
f 0 (x) = 2ax + b + √ (.1)
2 x
Z 1 3 1
x x2 2 a b 2c
f (x) dx = a + b + cx3/2 = + + . (.2)
0 3 2 3 0 3 2 3
a+b+c = 2
c
2a + b + = 1
2
a b 2c
+ + = 3.
3 2 3
So:
a+b+c = 2
4a + 2b + c = 2
2a + 3b + 4c = 18.
(It is not essential to multiply the equations in this way to eliminate fractions, but it helps.)
method, the row-reduction method, and so on. Many candidates like to use a different
18
method, not covered in the subject guide, especially ‘Cramer’s rule’. It is acceptable to do
so in this question, since it only requires the use of a matrix method. (But, it has to be a
matrix method, as the question makes explicit: manipulation of the equations will not
suffice.) However, our view is that the row operations method is easier and less prone to
error, and it is the method that we intend candidates to learn. It is possible that in
future years we might sometimes insist explicitly that a row operations method
be used, in which case using Cramer’s rule or some other method would not be
answering the question. If we require the use of the row operation method we shall use
the phrase ‘Use row operations’.
The standard matrix method approach is now to reduce the augmented matrix to reduced
form.
The augmented matrix is:
 
1 1 1 2
(A | b) = 4 2 1 2 (1 F)
2 3 4 18

   
1 1 1 2 1 1 1 2
(A | b) = 4 2 1 2  → 0 −2 −3 −6
2 3 4 18 0 1 2 14
   
1 1 1 2 1 1 1 2
→ 0 1 2 14 → 0 1 2 14 .
0 2 3 6 0 0 1 22
c = 22, b = 14 − 2c = −30 and a = 2 − b − c = 10.
especially in the stressful circumstances of an examination. Quite probably, the Examiners
mistake). So, here, for instance, the Examiners will award marks if you can indicate that
you know how to start to solve the equations (by writing down an augmented matrix); that
you know what row operations are; that you know what it is you want to achieve with row
(b) A firm is a monopoly and the demand equation for the good it produces is
p + bq = a, where p is price, q is the quantity and a and b are positive numbers.
The firm’s fixed costs are 0 and the marginal cost function takes the positive
constant value k, where k < a. Determine the firm’s profit function, in terms of
q. Also find the maximum profit the firm can make, in terms of a, b and k,
simplifying your answer as much as possible.
19

The total cost is: Z
TC = M C dq = kq + c
where c is a constant. Given that T C(0) = F C = 0, we must have c = 0. The firm is a

monopoly, so its total revenue is T R = qpD (q) = q(a − bq). It follows that the profit is:
Π(q) = q(a − bq) − kq = (a − k)q − bq 2 .
We solve Π0 = 0, which is:

(a − k) − 2bq = 0
so q = (a − k)/(2b). This indeed gives a maximum since Π00 (q) = −2b < 0. Thus the
maximum profit is:
2
(a − k)2 (a − k)2

a−k a−k
Π = −b = .
2b 2b 2b 4b
Question 8
A firm has weekly production function q given by
q(k, l) = k 1/3 l1/3 ,
where k denotes the amount of capital and l the amount of labour. Each unit of
capital costs V dollars and each unit of labour costs W dollars. Show that the
minimum amount (in dollars) that the firm can spend in a week on capital and
labour if it manufactures a total weekly amount Q is
√
2 V W Q3/2 .
Now suppose that to produce each unit of the good, it costs the firm R dollars in
raw material costs. Suppose the good sells for a fixed price of P dollars per unit,
where P > R. Assuming that the only production costs are capital, labour and raw
materials, and that the firm minimizes its weekly expenditure on capital and labour,
show that the firm’s weekly profit Π(Q) is given by
√
Π(Q) = (P − R)Q − 2 V W Q3/2 .
Show that the maximum weekly profit is
(P − R)3
.
27 V W
The problem is to minimise V k + W l subject to the constraint k 1/3 l1/3 = Q.
The Lagrangian is:

L = V k + W l − λ(k 1/3 l1/3 − Q).
20

1
Lk = V − λk −2/3 l1/3 = 0
3
1
Ll = W − λk 1/3 l−2/3 = 0
3
Lλ = −(k 1/3 l1/3 − Q) = 0.
From the first two equations, eliminating λ, we have:
k −2/3 l1/3 k 1/3 l−2/3

= .
V W
So l = (V /W )k.
Hence, by the third equation (the constraint):

1/3
V
k 2/3 = Q
W
and so: r
W 3/2
k= Q
V
and: r
V W 3/2
l= Q .
W V
The minimum cost is therefore:
√
r r
W 3/2 V W 3/2
V k + Wl = V Q +W Q = 2 V W Q3/2 .
V W V
Using the given information, the profit function is:

√
Π(Q) = P Q − RQ + 2 V W Q3/2 .
Setting Π0 = 0 gives: √
P − R − 3 V W Q1/2 = 0,
which gives:
(P − R)2
Q= .
9V W
We have:
3√
Π00 = − V W Q−1/2
2
which is negative, so this is indeed going to give a maximum. The maximum value is:
(P − R)2

Π .
9V W
This is:
(P − R)3 √ (P − R)3 (P − R)3 (P − R)3 (P − R)3
−2 VW √ = −2 = .
9V W 93/2 V W V W 9V W 27V W 27V W
21

Important note
Information about the subject guide and the Essential reading

references
the commentary and/or online reading list and/or subject guide refer to an earlier edition. If
section.
General remarks
Learning outcomes
subjects.
1
questions.
Answer the question
Graph sketching
Some examinations in this subject ask you to sketch the graph of a function. Any sketching of
graphs should be done in the answer book. Graph paper is not needed. Indeed, as we have
mentioned often in Examiners’ commentaries, the plotting of points in order to graph a function is
not the correct approach. A sketch of the graph of a function should indicate its shape, its position
with respect to the axes, and its intercepts on those axes: it need not be drawn to scale. Graph
paper is not necessary for this.
Calculators
this course.
2
Examination revision strategy
Many candidates are disappointed to find that their examination performance is poorer than they
expected. This may be due to a number of reasons. The Examiners’ commentaries suggest ways of
addressing common problems and improving your performance. One particular failing is ‘question
spotting’, that is, confining your examination preparation to a few questions and/or topics which
have come up in past papers for the course. This can have serious consequences.
We recognise that candidates may not cover all topics in the syllabus in the same depth, but you
need to be aware that Examiners are free to set questions on any aspect of the syllabus. This
means that you need to study enough of the syllabus to enable you to answer the required number of
examination questions.
The syllabus can be found in the Course information sheet in the section of the VLE dedicated to
each course. You should read the syllabus carefully and ensure that you cover sufficient material in
preparation for the examination. Examiners will vary the topics and questions from year to year and
may well set questions that have not appeared in past papers. Examination papers may legitimately
include questions on any topic in the syllabus. So, although past papers can be helpful during your
revision, you cannot assume that topics or specific questions that have come up in past examinations
will occur again.
If you rely on a question-spotting strategy, it is likely you will find yourself in difficulties
when you sit the examination. We strongly advise you not to adopt this strategy.
3

Important note
Please note carefully the comment in relation to Question 3 concerning possible future requirement
to use row operations to solve systems of linear equations.

references
section.
Section A
Question 1
A monopolist’s marginal cost function is given by
M C = 20 + 4q,
where q is the quantity of good produced. Her fixed costs are 20, and the demand
equation for the good she produces is
p + 4q = 40,
where p and q are price and quantity, respectively. Find expressions for the total
revenue and for the profit, as functions of q. Determine the value of q which
maximises the profit.
4
Reading for this question
See Chapter 3 of the subject guide for related reading.
The total cost is T C = M C dq, which is 2q 2 + 20q + c. Since F C = 20 = T C(0), we have

R
c = 20. You should explain, as here, why the constant is 20 and not simply assume that it must
be. The monopolist’s demand equation is p + 4q = 40, so when the quantity produced is q, the
selling price will be p = 40 − 4q and hence the total revenue will be
T R = qp = q(40 − 4q) = 40q − 4q 2 . So the profit function is:
Π(q) = T R(q) − T C(q) = 40q − 4q 2 − (2q 2 + 20q + 20) = −6q 2 + 20q − 20.
We solve Π0 (q) = 0. We have Π0 (q) = 20 − 12q, and this is 0 when q = 5/3. This gives a
maximum because Π00 (q) = −12 < 0 (or because Π0 changes sign from positive to negative). (We
could simply note that Π is a quadratic with negative squared term and hence the critical point
must be a maximum.)
Question 2
Use the Lagrange multiplier method to determine the values of x and y that
minimise x + 2y subject to the constraint x3 y 2 = 27.
The Lagrange multiplier method for constrained optimisation is discussed in Chapter 5 of the
subject guide.
The Lagrangian is L = x + 2y − λ(x3 y 2 − 27). The first order conditions are:
1 − 3λx2 y 2 = 0,
2 − 2λx3 y = 0,
x3 y 2 = 27.
1 1
The first two equations imply that 2 2
= 3 , and so x = 3y. (It’s clear that the constraint is
3x y x y
not satisfied when x or y is 0, so we can assume they are non-zero.)
Then, (3y)3 y 2 = 27, which is y 5 = 1, so:
x = 3, y = 1.
Question 3
Suppose the numbers x, y, z satisfy the following equations, where a is some number:
x+y+z = 5−a
2x + y = 7−a
x − y − 2z = 0.
Use a matrix method to determine the values of x, y and z, in terms of a.
For what values of a will x, y, z all be positive?
5
guide.
It is known by several names: the row operation method, the Gauss–Jordan method, the
operations method is easier and less prone to error, and it is the method that we intend students
to learn. It is possible that in future years we might insist explicitly that a row
operations method be used, in which case using Cramer’s rule or some other
method would not be answering the question. If we require the use of the row operation
method we shall use the phrase ‘Use row operations’.

1 1 1 5−a
(A | b) = 2 1 0 7 − a
1 −1 −2 0

     
1 1 1 5−a 1 1 1 5−a 1 1 1 5−a
2 1 0 7 − a → 0 −1 −2 −3 + a → 0 1 2 3 − a .
1 −1 −2 0 0 −2 −3 a−5 0 0 1 1−a
(You can stop here, or continue reducing further, as some do, until the first three columns form
the identity matrix.)
z = 1 − a,
y = 3 − a − 2(1 − a) = 1 + a,
x = 5 − a − (1 + a) − (1 − a) − 1 = 3 − a.
For x, y, z all to be positive, we must have 1 − a, 1 + a, 3 − a > 0 which means a < 1, a > −1 and
a < 3. So we need −1 < a < 1.
6
Question 4
The function f is defined by

1
f (x, y) = p .
x2 + y2
Show that
1/3
∂2f ∂2f

+ = f (x, y).
∂x2 ∂y 2
∂2f
∂2f
We have:
−x
fx = ,
(x2 + y 2 )3/2
−y
fy = ,
(x2 + y 2 )3/2
−(x2 + y 2 )3/2 + 3x2 (x2 + y 2 )1/2 −(x2 + y 2 ) + 3x2 2x2 − y 2

fxx = = = ,
(x2 + y 2 )3 (x2 + y 2 )5/2 (x2 + y 2 )5/2
2y 2 − x2
fyy = .
(x2 + y 2 )5/2
Then:
2x2 − y 2 + 2y 2 − x2 x2 + y 2 1
fxx + fyy = 2 2 5/2
= 2 = 2
(x + y ) (x + y 2 )5/2 (x + y 2 )3/2
= (f (x, y))3 .
So:
1/3
∂2f ∂2f

+ = f (x, y).
∂x2 ∂y 2
Question 5
Show that the function f given by
f (x, y) = x4 + 2x2 y + 2y 2 + y
has three critical points. For each critical point of f , determine whether it is a local
minimum, local maximum, or saddle point.
7
In order to find the critical points of the function:
f (x, y) = x4 + 2x2 y + 2y 2 + y
we work out the partial derivatives of f :

∂f ∂f
= 4x3 + 4xy, = 2x2 + 4y + 1.
∂x ∂y
The critical points are given by the first-order conditions:
4x3 + 4xy = 0, 2x2 + 4y + 1 = 0.
The first of these equations says that x(x2 + y) = 0, and so either (i) x = 0 or (ii) y = −x2 .
We now analyse each of the two cases separately.
(i) When x = 0, the second equation is 4y + 1 = 0, giving y = −1/4. Thus one critical point is
(0, −1/4).
√
(ii) When
√ y = −x2 the second equation becomes 2x2 − 4x2 + 1 = 0, and so x is either 1/ 2 or
−1/ 2. The corresponding values of y are determined by y = −x2 , and are both −1/2.
√ √
Thus, there are three critical points: (0, −1/4), (1/ 2, −1/2) and (−1/ 2, −1/2).
To classify them we need the second partial derivatives of f :
fxx (x, y) = 12x2 + 4y, fxy (x, y) = 4x, fyy (x, y) = 4.
Note that, as is generally the case, these are not constant, so we have to look at each critical
point individually.
At (0, −1/4): we have fxx = −1, fxy = 0, fyy = 4, so D = (−1)(4) − 0 < 0. Therefore (0, −1/4)
is a saddle point.
√ √
At (1/ 2, −1/2):
√ we have fxx = 4, fxy = 2 2, fyy = 4, so fxx > 0 and D = (4)(4) − 8 > 0.
Therefore (1/ 2, −1/2) is a local minimum.
√ √
At (−1/ 2, −1/2):
√ we have fxx = 4, fxy = −2 2, fyy = 4, so fxx > 0 and D = (4)(4) − 8 > 0.
Therefore (−1/ 2, −1/2) is also a local minimum.
There are other ways of using the equations to derive the critical points. Here’s an alternative:
We have 4y = −1 − 2x2 from the second equation. Substituting into the first equation then
√ gives
4x3 − x − 2x3 = 0 so that 2x3 − x = 0. So √
x(2x2 − 1) = 0 and therefore x = 0 or x = ±1/ 2.
When x = 0, y = −1/4 and when x = ±1/ 2, y = −1/2.
Question 6
Sarah saves money in a bank account paying interest at a fixed rate of 5%, where
the interest is paid once per year, at the end of the year. She deposits an amount A
at the beginning of each of the next N years. Find an expression, in terms of A and
N , and in as simple a form as possible, for the final amount saved (the amount just
after the last deposit).
8
There are two steps in questions like this. The first is to model the situation that is being
described using mathematics and the second is to use that mathematics to find the answer. In
this case, we are asked to find out something about the final amount saved, i.e. the amount in
the account just after the last deposit, and so it makes sense to model this situation by seeing
what the balance is after each deposit. This means that we may want to let yn (say) represent
the amount of money after the nth deposit and then, using the information in the question, we
can figure out what the first few terms of this sequence (say y1 , y2 and y3 ) are. Having done this,
we should be able to spot a pattern and, by generalising what we are seeing, we should be able to
find what we want, i.e. yN , the balance after her N th (or last) deposit.
All this means is that a good solution to this question would run as follows if we bear in mind
how compound interest works and what the question tells us about Sarah’s savings.
Let yn be the amount saved after the nth deposit.
This means that we have:
(Amount after 1st deposit) y1 = A

(Amount after 2nd deposit) y2 = 1.05y1 + A = 1.05A + A
(Amount after 3rd deposit) y3 = 1.05y2 + A = 1.052 A + 1.05A + A
(Note that it is absolutely crucial that you specify what each of the items in this list represents
either in words (as we have done in the brackets) or by specifying exactly what yn is supposed to
represent. If you don’t do this, then no-one knows what you are trying to work out!)
We now look carefully at this list to try and spot a pattern and, if we can, we should see that it
generalises to give us:
(Amount after N th deposit) yN = 1.05N A + 1.05N −1 A + · · · + 1.05A + A,
and this is what we were after, i.e. Sarah’s balance after the N th (or last) deposit. However, this
is not the answer as it can be simplified. In particular, this is a geometric series with first term
A, commmon ratio 1.05 and N terms so it can be summed, using the formula in the subject
guide, to get:
1.05N − 1
(Amount after N th deposit) yN = A .
1.05 − 1
(Note that we can only apply the formula if our series for yN is clearly specified. That is, we
need to know how it starts, how it changes and, to get the number of terms, how it finishes!)
So, tidying this up, we can see that our final answer is:
(Amount after N th deposit) yN = 20A(1.05N − 1).
As always, it’s important to simplify the answer as much as possible!
9
Section B
Question 7
(a) The market value of a painting, if it is sold at time t ≥ 0, is assumed to be

V (t) = (t + 2)2 . The present value P (t) of money raised if it is sold at time t is
V (t)e−rt where r is a constant such that 0 < r < 1. Determine the value of t
that maximises P (t) and determine also the maximum value of P (t).

We have:
P 0 (t) = 2(t + 2)e−rt + (t + 2)2 e−rt (−r) = e−rt (t + 2)(2 − 2r − rt).
We solve P 0 (t) = 0 (for t ≥ 0). For t ≥ 0, P 0 (t) = 0 if and only if t = T = (2 − 2r)/r. (We
reject t = −2 because it’s negative.) Here, P 0 changes sign from positive to negative, so this
maximises P . So the maximum value is the value at T , which is:
2 2
2 −rT 2 − 2r −r(2−2r)/r 2
(T + 2) e = +2 e = e2r−2 .
r r
It is also possible (though not as easy) to use the second-derivative test to establish that
t = T gives a maximum value. You should find that:
P 00 (t) = −re−rt (t + 2)(2 − 2r − rt) + e−rt (2 − 2r − rt) − re−rt (t + 2).
So, P 00 (T ) = −re−rT 2−2r + 2 = −2e−rT < 0, so t = T gives a maximum.

r
(b) Determine the integrals Z √

x
√ dx
1+ x
Z e
x2 ln x dx.
1

Integration is discussed in Chapter 4 of the subject guide.
It can be difficult because it is not always clear which technique will work. The three main
techniques are: substitution, parts, and partial fractions. More than one method might
work, and some integrals require a combination of methods.
√
We can approach
√ the first integral
√ by making the substitution u = 1 + x. We then have
du = 1/(2 x) dx, so that dx = 2 x du = 2(u − 1) du and the integral becomes:
2(u − 1)2
Z
du.
u
This is: Z
1
2 u−2+ du = u2 − 4u + 2 ln u + c
u
√ √ √
= (1 + x)2 − 4(1 + x) + 2 ln(1 + x) + c.
√ √
The substitution u = x also works. With this choice, we have du = 1/(2 x) dx, so that
dx = 2u du and the integral becomes:
2u2
Z
du.
1+u
10
This is:
Z Z Z Z Z
2u(1 + u) 2u 2(1 + u) 1
du − du = 2u du − du + 2 du.
1+u 1+u 1+u 1+u
This is: √ √
u2 − 2u + 2 ln(1 + u) + c = x − 2 x + 2 ln(1 + x) + c.
For the second integral, we have:
Z e e Z e 2
x3

x
1 3 1 1 3
3 e
e3 x
= −
3 9 1
e3
3
e 1
= = − −
3 9 9
2e3 1
= + .
9 9
A correct answer must use the facts that ln 1 = 0 and ln e = 1.
Question 8
(a) A firm has production function q given by q(k, l) = k1/2 l1/4 where k denotes the
amount of capital and l the amount of labour. Each unit of capital costs 10
dollars and each unit of labour costs 20 dollars. Use the Lagrange multiplier
method to determine, in terms of q, the capital and labour that will minimise
the cost of producing an amount q. Determine this minimum cost, simplifying
your answer as much as possible.

We have to minimise 10k + 20l subject to the constraint k 1/2 l1/4 = q. The Lagrangian is:
L = 10k + 20l − λ(k 1/2 l1/4 − q),
and the optimal k and l satisfy the three equations:
1
10 − λk −1/2 l1/4 = 0,
2
1
20 − λk 1/2 l−3/4 = 0,
4
k 1/2 l1/4 = q.
The first two equations, on elimination of λ, show that:
80k −1/2 l3/4 = 20k 1/2 l−1/4 ,
so that k = 4l.
Then:
k 1/2 l1/4 = (4l)1/2 l1/4 = q,
and: q 4/3
l= ,
2
q 4/3

so k = 4l = 4 2 . The minimum cost is:
q 4/3 q 4/3 q 4/3
10k + 20l = 40 + 20 = 60 .
2 2 2
11
(b) Let n be any integer greater than 1. The function f is defined for positive x by
f (x) = nx1/n − x.
Find the maximum value of f (x) over all x > 0. Hence show that, for all x > 0,

x 1
x1/n ≤ + 1− .
n n

1
We have f 0 (x) = x n −1 − 1. We solve f 0 (x) = 0. The solution is x = 1. At x = 1, f 0 changes
1
sign from positive to negative. This is because n1 − 1 < 0 and therefore if x < 1, x n −1 > 1
1
and if x > 1, x n −1 < 1. The point is therefore a maximum. Alternatively,
1
f 00 (x) = ( n1 − 1)x n −2 so f 00 (1) = ( n1 − 1), and this is negative because n > 1. So the point is
a maximum.
We therefore have that the maximum value among positive x is f (1) = n − 1. What that
implies is that, for all x > 0:
f (x) ≤ f (1) = n − 1
and hence:
nx1/n − x ≤ n − 1
and so:
1/n x 1
x ≤ + 1− .
n n
12

Important note
Please note carefully the comment in relation to Question 3 concerning possible future requirement
to use row operations to solve systems of linear equations.

references
section.
Section A
Question 1
A monopolist’s marginal cost function is given by
M C = 10 + 2q,
where q is the quantity of good produced. Her fixed costs are 20, and the demand
equation for the good she produces is
p + 2q = 20,
where p and q are price and quantity, respectively. Find expressions for the total
revenue and for the profit, as functions of q. Determine the value of q which
maximises the profit.
13
The total cost is T C = M C dq, which is q 2 + 10q + c. Since F C = 20 = T C(0), we have c = 20.
R
You should explain, as here, why the constant is 20 and not simply assume that it must be. The
monopolist’s demand equation is p + 2q = 20, so when the quantity produced is q, the selling
price will be p = 20 − 2q and hence the total revenue will be T R = qp = q(20 − 2q) = 20q − 2q 2 .
So the profit function is:
Π(q) = T R(q) − T C(q)
= 20q − 2q 2 − (q 2 + 10q + 20) = −3q 2 + 10q − 20.
We solve Π0 (q) = 0. We have Π0 (q) = 10 − 6q, and this is 0 when q = 5/3. This gives a maximum
because Π00 (q) = −6 < 0 (or because Π0 changes sign from positive to negative). (We could
simply note that Π is a quadratic with negative squared term and hence the critical point must
be a maximum.)
Question 2
Use the Lagrange multiplier method to determine the values of x and y that
minimise 2x + y subject to the constraint x2 y 3 = 27.
subject guide.
The Lagrangian is L = x + 2y − λ(x3 y 2 − 27). The first order conditions are:

2 − 2λxy 3 = 0,
1 − 3λx2 y 2 = 0,
x2 y 3 = 27.
1 1
The first two equations imply that 3
= 2 2 , and so y = 3x. (It’s clear that the constraint is
xy 3x y
not satisfied when x or y is 0, so we can assume they are non-zero.)
Then, x2 (3x)3 = 27, which is x5 = 1, so:

x = 1, y = 3.
Question 3
Suppose the numbers x, y, z satisfy the following equations, where a is some number:
x+y+z = 4+a
2x + y = 3+a
x − y − 2z = −4.
Use a matrix method to determine the values of x, y and z, in terms of a.
For what values of a will x, y, z all be positive?
guide.
14
It is known by several names: the row operation method, the Gauss–Jordan method, the
operations method is easier and less prone to error, and it is the method that we intend students
to learn. It is possible that in future years we might insist explicitly that a row
operations method be used, in which case using Cramer’s rule or some other
method would not be answering the question. If we require the use of the row operation
method we shall use the phrase ‘Use row operations’.

 
1 1 1 4+a
(A | b) = 2 1 0 3 + a (1 C)
1 −1 −2 −4

     
1 1 1 4+a 1 1 1 4+a 1 1 1 4+a
2 1 0 3 + a → 0 −1 −2 −5 − a → 0 1 2 5 + a .
1 −1 −2 −4 0 −2 −3 −8 − a 0 0 1 2+a
z = 2 + a,
y = 5 + a − 2(2 + a) = 1 − a,
x = 4 + a − (1 − a) − (2 + a) − 1 = 1 + a.
For x, y, z all to be be positive, we must have 2 + a, 1 − a, 1 + a > 0 a > −2, a < 1 and a > −1.
So we need −1 < a < 1.
15
Question 4
The function f is defined for x, y > 0 by
f (x, y) = ln(x2 + y 2 ).
Find the partial derivatives
∂f ∂f ∂2f ∂2f ∂2f

, , , , .
∂x ∂y ∂x2 ∂x∂y ∂y 2
Show that
∂2f ∂2f
+ = 0.
∂x2 ∂y 2
∂2f
∂2f
We have:
2x
fx = ,
x2 + y 2
2y
fy = ,
x2 + y 2
2(x2 + y 2 ) − 2x(2x) 2y 2 − 2x2
fxx = = ,
(x2 + y 2 )2 (x2 + y 2 )2
2x2 − 2y 2
fyy = .
(x2 + y 2 )2
−4xy
fxy =
(x2 + y 2 )2
Then:
2y 2 − 2x2 2x2 − 2y 2
fxx + fyy = + = 0.
(x2 + y 2 )2 (x2 + y 2 )2
Question 5
Show that the function f given by
f (x, y) = x4 + 2x2 y + 2y 2 + y
has three critical points. For each critical point of f , determine whether it is a local
minimum, local maximum, or saddle point.
16
In order to find the critical points of the function:
f (x, y) = x4 + 2x2 y + 2y 2 + y
we work out the partial derivatives of f :

∂f ∂f
= 4x3 + 4xy, = 2x2 + 4y + 1.
∂x ∂y
The critical points are given by the first-order conditions:
4x3 + 4xy = 0, 2x2 + 4y + 1 = 0.
The first of these equations says that x(x2 + y) = 0, and so either (i) x = 0 or (ii) y = −x2 .
We now analyse each of the two cases separately.
(i) When x = 0, the second equation is 4y + 1 = 0, giving y = −1/4. Thus one critical point is
(0, −1/4).
√
(ii) When
√ y = −x2 the second equation becomes 2x2 − 4x2 + 1 = 0, and so x is either 1/ 2 or
−1/ 2. The corresponding values of y are determined by y = −x2 , and are both −1/2.
√ √
Thus, there are three critical points: (0, −1/4), (1/ 2, −1/2) and (−1/ 2, −1/2).
To classify them we need the second partial derivatives of f :
fxx (x, y) = 12x2 + 4y, fxy (x, y) = 4x, fyy (x, y) = 4.
Note that, as is generally the case, these are not constant, so we have to look at each critical
point individually.
At (0, −1/4): we have fxx = −1, fxy = 0, fyy = 4, so D = (−1)(4) − 0 < 0. Therefore (0, −1/4)
is a saddle point.
√ √
At (1/ 2, −1/2):
√ we have fxx = 4, fxy = 2 2, fyy = 4, so fxx > 0 and D = (4)(4) − 8 > 0.
Therefore (1/ 2, −1/2) is a local minimum.
√ √
At (−1/ 2, −1/2):
√ we have fxx = 4, fxy = −2 2, fyy = 4, so fxx > 0 and D = (4)(4) − 8 > 0.
Therefore (−1/ 2, −1/2) is also a local minimum.
There are other ways of using the equations to derive the critical points. Here’s an alternative:
We have 4y = −1 − 2x2 from the second equation. Substituting into the first equation then
√ gives
4x3 − x − 2x3 = 0 so that 2x3 − x = 0 So x(2x
√
2
− 1) = 0 and therefore x = 0 or x = ±1/ 2.
When x = 0, y = −1/4 and when x = ±1/ 2, y = −1/2.
Question 6
Sarah saves money in a bank account paying interest at a fixed rate of 4%, where
the interest is paid once per year, at the end of the year. She deposits an amount A
at the beginning of each of the next N years. Find an expression, in terms of A and
N , and in as simple a form as possible, for the final amount saved (the amount just
after the last deposit).
17
There are two steps in questions like this. The first is to model the situation that is being
described using mathematics and the second is to use that mathematics to find the answer. In
this case, we are asked to find out something about the final amount saved, i.e. the amount in
the account just after the last deposit, and so it makes sense to model this situation by seeing
what the balance is after each deposit. This means that we may want to let yn (say) represent
the amount of money after the nth deposit and then, using the information in the question, we
can figure out what the first few terms of this sequence (say y1 , y2 and y3 ) are. Having done this,
we should be able to spot a pattern and, by generalising what we are seeing, we should be able to
find what we want, i.e. yN , the balance after her N th (or last) deposit.
All this means is that a good solution to this question would run as follows if we bear in mind
how compound interest works and what the question tells us about Sarah’s savings.
Let yn be the amount saved after the nth deposit.
This means that we have:
(Amount after 1st deposit) y1 = A

(Amount after 2nd deposit) y2 = 1.04y1 + A = 1.04A + A
(Amount after 3rd deposit) y3 = 1.04y2 + A = 1.042 A + 1.04A + A
(Note that it is absolutely crucial that you specify what each of the items in this list represents
either in words (as we have done in the brackets) or by specifying exactly what yn is supposed to
represent. If you don’t do this, then no-one knows what you are trying to work out!)
We now look carefully at this list to try and spot a pattern and, if we can, we should see that it
generalises to give us:
(Amount after N th deposit) yN = 1.04N A + 1.04N −1 A + · · · + 1.04A + A,
and this is what we were after, i.e. Sarah’s balance after the N th (or last) deposit. However, this
is not the answer as it can be simplified. In particular, this is a geometric series with first term
A, commmon ratio 1.04 and N terms so it can be summed, using the formula in the subject
guide, to get:
1.04N − 1
(Amount after N th deposit) yN = A .
1.04 − 1
(Note that we can only apply the formula if our series for yN is clearly specified. That is, we
need to know how it starts, how it changes and, to get the number of terms, how it finishes!)
So, tidying this up, we can see that our final answer is:
(Amount after N th deposit) yN = 25A(1.04N − 1).
As always, it’s important to simplify the answer as much as possible!
18
Section B
Question 7
(a) The market value of a painting, if it is sold at time t ≥ 0, is assumed to be

V (t) = (t + 1)2 . The present value P (t) of money raised if it is sold at time t is
V (t)e−rt where r is a constant such that 0 < r < 1. Determine the value of t
that maximises P (t) and determine this maximum value, simplifying your
answer as much as possible.

We have:
P 0 (t) = 2(t + 1)e−rt + (t + 1)2 e−rt (−r) = e−rt (t + 1)(2 − r − rt).
We solve P 0 (t) = 0 (for t ≥ 0). For t ≥ 0, P 0 (t) = 0 if and only if t = T = (2 − r)/r. (We
reject t = −1 because it’s negative.) Here, P 0 changes sign from positive to negative, so this
maximises P . So the maximum value is the value at T , which is:
2 2
2 −rT 2−r −r(2−r)/r 2
(T + 1) e = +1 e = er−2 .
r r
It is also possible (though not as easy) to use the second-derivative test to establish that
t = T gives a maximum value. You should find that:
P 00 (t) = −re−rt (t + 1)(2 − r − rt) + e−rt (2 − r − rt) − re−rt (t + 1).
So, P 00 (T ) = −re−rT 2−r −rT

r + 1 = −2e < 0, so t = T gives a maximum.
(b) Determine the integrals √

x
Z
√ dx
1+ x
Z e
x3 ln x dx.
1

It can be difficult because it is not always clear which technique will work. The three main
techniques are: substitution, parts, and partial fractions. More than one method might
work, and some integrals require a combination of methods.
√
We can approach
√ the first integral
√ by making the substitution u = 1 + x. We then have
du = 1/(2 x) dx, so that dx = 2 x du = 2(u − 1) du and the integral becomes:
2(u − 1)2
Z
du.
u
This is: Z
1
2 u−2+ du = u2 − 4u + 2 ln u + c
u
√ √ √
= (1 + x)2 − 4(1 + x) + 2 ln(1 + x) + c.
√ √
The substitution u = x also works. With this choice, we have du = 1/(2 x) dx, so that
dx = 2u du and the integral becomes
2u2
Z
du.
1+u
19
This is:
Z Z Z Z Z
2u(1 + u) 2u 2(1 + u) 1
du − du = 2u du − du + 2 du.
1+u 1+u 1+u 1+u
This is: √ √
u2 − 2u + 2 ln(1 + u) + c = x − 2 x + 2 ln(1 + x) + c.
For the second integral, we have:

Z e e Z e 3
x4

x
1 4 1 1 4
4 e
e4 x
= −
4 16 1
e4 e4 3e4

1 1
= − − = + .
4 16 16 16 16
A correct answer must use the facts that ln 1 = 0 and ln e = 1.
Question 8
(a) A firm has production function q given by q(k, l) = k1/4 l1/2 where k denotes the
amount of capital and l the amount of labour. Each unit of capital costs 20
dollars and each unit of labour costs 10 dollars. Use the Lagrange multiplier
method to determine, in terms of q, the capital and labour that will minimise
the cost of producing an amount q. Determine this minimum cost, simplifying
your answer as much as possible.

We have to minimise 20k + 10l subject to the constraint k 1/4 l1/2 = q. The Lagrangian is:
L = 20k + 10l − λ(k 1/4 l1/2 − q),
and the optimal k and l satisfy the three equations:

1
20 − λk −3/4 l1/2 = 0,
4
1
10 − λk 1/4 l−1/2 = 0,
2
k 1/4 l1/2 = q.
80k 3/4 l−1/2 = 20k −1/4 l1/2 ,
so that l = 4k.
Then:
1/2
k 1/4 l1/2 = k 1/4 (4k) = q,
and: q 4/3
k= ,
2
q 4/3

so l = 4k = 4 2 . The minimum cost is:
q 4/3 q 4/3 q 4/3
20k + 10l = 20 + 40 = 60 .
2 2 2
20
(b) Let n be any integer greater than 1. The function f is defined for positive x by
f (x) = nx1/n − x.
Find the maximum value of f (x) over all x > 0. Hence show that, for all x > 0,

x 1
x1/n ≤ + 1− .
n n

1
We have f 0 (x) = x n −1 − 1. We solve f 0 (x) = 0. The solution is x = 1. At x = 1, f 0 changes
1
sign from positive to negative. This is because n1 − 1 < 0 and therefore if x < 1, x n −1 > 1
1
and if x > 1, x n −1 < 1. The point is therefore a maximum. Alternatively,
1
f 00 (x) = ( n1 − 1)x n −2 so f 00 (1) = ( n1 − 1), and this is negative because n > 1. So the point is
a maximum.
We therefore have that the maximum value among positive x is f (1) = n − 1. What that
implies is that, for all x > 0:
f (x) ≤ f (1) = n − 1
and hence:
nx1/n − x ≤ n − 1
and so:
1/n x 1
x ≤ + 1− .
n n
21
Mathematics 1
279005a, 990005a, 996D05a
2003, 2004, 2005

Zone A
5
05a Mathematics 1
6
7
05a Mathematics 1
8
9
05a Mathematics 1
10
11
Mathematics 1
279005a, 990005a, 996D05a
2003, 2004, 2005

Zone B
17
05a Mathematics 1
18
19
05a Mathematics 1
20
21
05a Mathematics 1
22
23

05a Mathematics 1
Specific comments on questions – Zone A
Section A
Answer all seven questions from this section (60 marks in total).
Question 1
A firm is the only producer of a particular good, and the demand equation for the
good is 2p + q = 18, where p denotes the selling price and q is the quantity
produced by the firm. The firm’s fixed costs are 12 and its marginal cost function is
M C = 1 + q. Find an expression for the profit function, Π(q). Sketch the graph of
Π for q between 0 and 8. Determine the value of the production, q, which
maximises the firm’s profit.
This question concerns curve sketching and optimisation, discussed in Chapters 2 and 3 of the
subject guide. From the demand equation:
q
p=9− .
2
So the total revenue is:
1
T R = pq = 9q − q 2 .
2
To obtain the total cost from the marginal cost, we must integrate. We have:
Z Z
1
T C = M C dq = (1 + q) dq = q + q 2 + c.
2
To find c, note that T C(0) = F C = 12, so c = 12. (You should explain, like this, why the
constant of integration is 12.)
So, the profit, as a function of q, is
Π = T R − T C = −q 2 + 8q − 12.
We now solve Π0 = 0 to maximise the profit. We have
8 − 2q = 0, so q = 4,
and
Π00 (q) = −2 < 0,
so q = 4 maximises the profit. (It is also acceptable, rather than checking the second derivative,
to comment that the critical point must be a maximum because the function is a negative
quadratic.)
1
05a Mathematics 1

cost with average cost. Recall that, since the marginal cost is the derivative of the total cost, we
must integrate the marginal cost to find the total cost.
Now,
Π(q) = 0 ⇐⇒ q 2 − 8q + 12 = 0 ⇐⇒ (q − 6)(q − 2) = 0 ⇐⇒ q = 2, 6.
The sketch is:
2
q
1 2 3 4 5 6 7 8
0
–2
–4
–6
–8
–10
–12
You should show the correct shape of the curve, and you should (as above) calculate the
intercepts with the q-axis.
Question 2
Functions f and g are as follows:
f (x) = x4 + 2x3 + 2x2 + 2, g(x) = −x4 + 2x3 + 18x2 + 20.
Show that the curves y = f (x) and y = g(x) intersect for exactly two values of x.
Find these values of x. (Do not attempt to sketch the curves.)
The intersections are the solutions of f (x) = g(x). Now,
f (x) = g(x) ⇐⇒ x4 + 2x3 + 2x2 + 2 = −x4 + 2x3 + 18x2 + 20 ⇐⇒ 2x4 − 16x2 − 18 = 0,
so we have the equation x4 − 8x2 − 9 = 0. We don’t have a general method for solving
polynomial equations of degree 4. However, this equation can be viewed as a quadratic in x2 , for
it is (x2 )2 − 8(x2 ) − 9 = 0.
Now we can use the formula for the solutions of a quadratic, or we can use factorisation to find
x2 :
(x2 − 9)(x2 + 1) = 0,
so x2 = 9. (Note that we cannot have x2 = −1.) So x = 3, −3 are the two required values.
2
Question 3
matrix method.
2x − 3y − z = 0
2x + 3y + z = 8
x − 2y + 3z = 3.
suffice.
In matrix form, this is     

2 −3 −1 x 0
 2 3 1  y  =  8 .
1 −2 3 z 3
2 −3 −1 0
 2 3 1 8 .
1 −2 3 3
Using row operations to reduce (and there are many ways this can be done), we have:
     
2 −3 −1 0 2 −3 −1 0 1 0 0 2
 2 3 1 8 → 4 0 0 8 → 2 −3 −1 0 
1 −2 3 3 1 −2 3 3 1 −2 3 3
   
1 0 0 2 1 0 0 2
→ 0 −3 −1 −4  →  0 3 1 4 .
0 −2 3 1 0 0 11 11
So, using back-substitution, we have
z = 1,
y = (4 − 1)/3 = 1,
x = 2.
here, for instance, Examiners will award marks if you can indicate that you know how to start to
solve the equations (by writing down an augmented matrix); that you know what row operations
are; that you know what it is you want to achieve with row operations (the reduced matrix, that
is); and that you then know how to work from that reduced matrix to determine the required
solutions. There are marks for all these things.
3
05a Mathematics 1
Question 4
Determine the integral

ln(x + 1)
Z
dx.
x2
partial fractions. The integral here is
Z
I = x−2 ln(x + 1) dx.
We start by using integration by parts:

Z
1
I = −x−1 ln(x + 1) + x−1 dx
x+1
Z
1
= −x−1 ln(x + 1) + dx
x(x + 1)
Now we use partial fractions:

Z
1 1
I = −x−1 ln(x + 1) + − dx = −x−1 ln |x + 1| + ln |x| − ln |x + 1| + c.
x x+1
Question 5
A travel company is the only provider of holidays (of one week’s duration) to two
private island resorts, X and Y . The demand equations for such holidays are given
by
x = 200 − pX ,
y = 100 − pY ,
where x and y are the numbers of week-long holidays at X and Y demanded

(respectively) and pX , pY are (respectively) the prices of these holidays. The
company’s joint total cost function (that is, the cost of providing x holidays in X
and y holidays in Y ) is
2x2 + y 2 + 2xy.
Find an expression in terms of x and y for the profit the company obtains from
selling these holidays. Determine the numbers x and y of holidays in resorts X and
Y that will maximise the company’s profit.

and we want to get this as a function of x and y, so we need to find pX and pY in terms of x and
y. We have
pX = 200 − x, pY = 100 − y.
So,
Π = xpX + ypY − T C = 200x + 100y − 3x2 − 2y 2 − 2xy.
4
Πx = 200 − 6x − 2y, Πy = 100 − 4y − 2x.
We now solve Πx = 0, Πy = 0, obtaining x = 30 and y = 10. We have Πxx = −6 < 0 and

Πxx Πyy − Π2xy = (−6)(−4) − (−2)2 > 0, so this does indeed maximise the profit.
Question 6
The weekly output (in units) of a factory depends on the amount of capital and
labour it employs as √follows:
√ if it uses k units of capital and l of labour then its
output is Q(k, l) = k l units. The cost to the firm of each unit of capital is 1
dollar, and the cost of each unit of labour is 4 dollars. Use the method of Lagrange
multipliers to find the minimum weekly cost of producing a quantity of 200 units.
The Lagrange multiplier method is discussed in Chapter 5 of the subject guide.

√ √
The problem here is to minimise k + 4l subject to k l = 200.
The Lagrangian is √ √
L = k + 4l − λ( k l − 200).
(It is fine to use +λ here rather than −λ.) We now set the partial derivatives of L to zero:
√
l
Lk = 1 − λ √ = 0,
2 k
√
k
Ll = 4 − λ √ = 0,
2 l
√ √
Lλ = −( k l − 200) = 0.
Eliminating λ from the first two equations,
√ √
l k
4 √ = √ , so k = 4l.
k l
√ √
Then,
√ the
√ constraint equation (equivalently, the third of the above equations) k l = 200 tells
us 4l l = 200, which is 2l = 200, and so l = 100, k = 400. So the minimum cost is
k + 4l = 400 + 4(100) = 800.
This is not part of the syllabus and no credit is given for doing it.
Question 7
Find the geometric series that has second term equal to 5 and sum to infinity equal
to 20.
With the usual notation, the sum to infinity is a/(1 − r) and the second term is ar. So
a
= 20, ar = 5.
1−r
So a = 20(1 − r) and
20(1 − r)r = 5,
5
05a Mathematics 1
giving
20r2 − 20r + 5 = 0, 4r2 − 4r + 1 = 0
(or an equivalent equation involving a if r has been eliminated instead). It follows that
(2r − 1)2 = 0,
and hence r = 1/2 The corresponding value of a is a = 10.
Section B
Question 8
i. The function f is given, for 0 < x < 1, by
f (x) = xx (1 − x)(1−x) .
Show that f has one critical (or stationary) point between 0 and 1, and that this
point is a local minimum.
We solve f 0 = 0. Calculating f 0 is quite tricky. Let g = xx . Then ln g = x ln x, so
g0
= 1 + ln x
g
and
g 0 = (1 + ln x)xx .
So, by the product rule and chain rule,
f 0 = (1 + ln x)xx (1 − x)(1−x) − (1 + ln(1 − x))xx (1 − x)(1−x) = xx (1 − x)(1−x) (ln x − ln(1 − x)) .
An alternative approach could be to start by taking the log of f as a step towards calculating
f 0 .We have
ln f = ln(xx (1 − x)(1−x) ) = x ln x + (1 − x) ln(1 − x),
so
1 0
f (x) = ln x − ln(1 − x).
f
So
f 0 (x) = xx (1 − x)(1−x) (ln x − ln(1 − x)) .
Now,
f 0 = 0 ⇐⇒ ln x = ln(1 − x) ⇐⇒ x = 1 − x ⇐⇒ x = 1/2.
As x passes through 1/2, f 0 changes sign from negative to positive, because for x < 1/2,
x < 1 − x and for x > 1/2, x > 1 − x. So it’s a local minimum. Alternatively, we can use the
second derivative test:

00 1 1 2
f = + xx (1 − x)(1−x) + (ln x − ln(1 − x)) xx (1 − x)(1−x)
x 1−x
and
f 00 (1/2) = 4(1/2)1/2 (1/2)1/2 > 0,
so it is a minimum.
6
ii. A firm’s marginal cost function is given by

q
M C = qeq + ,
q 2 + 4q + 3
where q denotes the quantity produced. If its production is increased from q = 1 to
q = 2, find the increase in total cost.
What we want is Z 2
I= M C dq.
1
(The answer is not simply the value of M C(1). Calculating M C(1) and giving this as the
solution gets no credit.)
I = J + K,
where Z 2 Z 2
q q
J= qe dq, K = dq.
1 1 q 2 + 4q + 3
By parts, Z 2
2
J = [qeq ]1 − eq dq
1
2
= [qeq − eq ]1 = 2e2 − e2 − (e − e) = e2
For K, we use partial fractions. Noting that

q q
=
q 2 + 4q + 3 (q + 1)(q + 3)
we must have
q A B
= +
q2 + 4q + 3 q+1 q+3
for some A, B. We find that A = −1/2, B = 3/2. Then
2 3 1
K = [(3/2) ln |q + 3| − (1/2) ln |q + 1|]1 = ln(5/4) − ln(3/2).
2 2
So,
3 1
I = e2 + ln(5/4) − ln(3/2).
2 2
Question 9
i. For some numbers a, b, c, the function f is given by
f (x) = ax2 + bx3 + cx4 .
Given that Z 1
f (1) = 15, f 0 (1) = 44, and f (x) dx = 4,
0
find the numbers a, b and c.
This turns out to be a linear equations question. We have, first,
a + b + c = 15.
Since
f 0 (x) = 2ax + 3bx2 + 4cx3 ,
we also have
2a + 3b + 4c = 44.
7
05a Mathematics 1
Now,
1 1
ax3 bx4 cx5
Z
a b c
f (x) dx = + + = + + .
0 3 4 5 0 3 4 5
So,
a b c
+ + = 4.
3 4 5
So we have the following system of linear equations:
a+b+c = 15
2a + 3b + 4c = 44
a b c
+ + = 4
3 4 5
or, equivalently,
a+b+c = 15
2a + 3b + 4c = 44
20a + 15b + 12c = 240.
Using row operations, we have

     
1 1 1 15 1 1 1 15 1 1 1 15
 2 3 4 44  →  0 1 2 14  →  0 1 2 14  .
20 15 12 240 0 −5 −8 −60 0 0 2 10
So,
c = 5,
b = 14 − 10 = 4,
a = 15 − 4 − 5 = 6.
ii. Determine each of the following integrals.

Z √
e x dx,
cos x
Z
dx.
(1 − sin x)(sin x + 2)
√ 1
For the first integral, we start by using substitution. Let u = x. Then du = √
2 x
dx and so
Z
I=2 ueu du.
Now we can use integration by parts:

√ √ √
Z
I = 2ueu − 2 eu du = 2ueu − 2eu + c = 2 xe x − 2e x + c.
For the second integral, let u = sin x. Then

Z
du
I= .
(1 − u)(u + 2)
Now we use partial fractions.
Z
1 1 1 1 1
I= + du = (− ln |1−u|+ln |u+2|)+c = (− ln |1−sin x|+ln | sin x+2|)+c.
3 1−u u+2 3 3
8
Question 10
i. A consumer has a utility function

1/3
u(x, y) = (x + 1)3 + y 3
for two goods, X and Y . (Here, x denotes the amount of X and y the amount of
Y .) Each unit of x costs p dollars and each unit of Y costs q dollars, and the
consumer has a budget for X and Y of M dollars. Use the Lagrange multiplier
method to find the quantities of X and Y the consumer will consume in order to
maximise his utility subject to the budget constraint. (Your answers should be
expressions depending on p, q and M .)
1/3
We want to maximise (x + 1)3 + y 3 subject to px + qy = M . The Lagrangian is
1/3
L = (x + 1)3 + y 3 − λ(px + qy − M ).
1/3
(L = (x + 1)3 + y 3 + λ(px + qy − M ) is perfectly acceptable too.) We have
−2/3
Lx = (x + 1)2 (x + 1)3 + y 3 − λp = 0
and −2/3
Ly = y 2 (x + 1)3 + y 3 − λq = 0.
These imply that
y2 (x + 1)2
= ,
q p
so r
q
y= (x + 1).
p
The constraint is px + qy = M , which means
r
q
px + q (x + 1) = M
p
which is r r
q q
px + q x+q = M,
p p
and hence, solving for x and simplifying,
√ √
pM − q q
x= √ √ .
p p+q q
Then, r √ √ √ √
pM − q q qM + p q
r
q q
y= (x + 1) = √ √ +1 = √ √ .
p p p p+q q p p+q q
ii. The function f (x, y) is given by
f (x, y) = 10 − x2 + 2y 2 − 2y 4 + 2xy 2 .
Find the critical (or stationary) points of f . Determine whether each critical point
is a local maximum, local minimum, or saddle point.
This involves techniques from Chapter 5 of the subject guide. The partial derivatives are
fx = −2x + 2y 2 , fy = 4y − 8y 3 + 4xy.
We solve fx = fy = 0. We have (from fx = 0) x = y 2 . Substituting into fy = 0 gives

4y − 8y 3 + 4y 3 = 0, which is 4y(1 − y 2 ) = 0. So y = 0, 1, −1. Then x = 0, 1, 1, respectively. So
there are three critical points: (0, 0), (1, 1), (1, −1).
9
05a Mathematics 1

fxx = −2, fxy = 4y, fyy = 4 − 24y 2 + 4x.
2
At (0, 0): fxx fyy − fxy = −8 < 0, so this is a saddle point.
2
At (1, 1), fxx fyy − fxy = (−2)(−16) − (4)2 > 0 and fxx = −2 < 0, so it’s a maximum.
2
At (1, −1), fxx fyy − fxy = (−2)(−16) − (−4)2 > 0 and fxx = −2 < 0, so it’s a maximum.
Question 11
i. James opens a savings account with a bank. The interest rate is guaranteed to be
fixed at a rate of 5% per annum for K years after he opens the account. (Here, K is
an integer greater than 1.) He opens the savings account with a payment of $200 on
1 January 2009, and makes further deposits of $200 yearly, on 1 January each year,
starting 1 January 2010. Find an expression (involving K, and in as simple a form
as possible) for the value of his savings K years after he opens the account (that is,
K years from 1 January 2009, just after he has made a new deposit of $200).
Suppose now that he continues to make the same deposits after this time, but that
the interest rate falls to the fixed rate of 4% for the next N years (where N is an
integer greater than 1). Find an expression (involving K and N , and in as simple a
form as possible) for the value of his savings K + N years after he opened the
account (that is, K + N years from 1 January 2009, just after he has made a new
deposit of $200).
subject guide gives the background and some examples. Essentially, what you have to do is
problems using general techniques for solving difference equations, but that is not expected (as
such techniques are not part of the 05A syllabus, though they are in 05B).
Let yt be the balance t years after opening the account (just after the (t + 1)st deposit). Then
y0 = 200
y1 = 200(1.05) + 200
y2 = (1.05)y1 + 200 = 200(1.05)2 + 200(1.05) + 200.
So, spotting the general pattern,
yK = 200(1.05)K + 200(1.05)K−1 + · · · + 200(1.05) + 200.
Now we simplify using the formula for a geometric series:
1 − (1.05)K+1
yK = 200 = 4000((1.05)K+1 − 1).
1 − 1.05
What happens next is that 1.05 is replaced in subsequent compounding by 1.04. So
yK+1 = (1.04)yK + 200
yK+2 = (1.04)yK+1 + 200 = (1.04)2 yK + (1.04)200 + 200
yK+3 = (1.04)3 yK + (1.04)2 200 + (1.04)200 + 200.
Then, spotting the general pattern,
yK+N = (1.04)N yK + 200((1.04)N −1 + (1.04)N −2 + · · · + (1.04) + 1)
1 − (1.04)N
= (1.04)N yK + 200
1 − 1.04
= 4000(1.04) ((1.05)K+1 − 1) + 5000((1.04)N − 1).
N
10
ii. The function f is given, for some number a, by
f (x, y) = 2xy + x2a y a .
Find, in terms of x, y and a, the partial derivatives
∂f ∂f ∂2f ∂2f
, , , .
∂x ∂y ∂x2 ∂y 2
Now suppose that we know that f satisfies
∂2f ∂2f
x2 − 2y 2 − 18f + 36xy = 0.
∂x2 ∂y 2
Determine the possible values of a.
The derivatives are
fx = 2y + 2ax2a−1 y a ,
fy = 2x + ax2a y a−1 ,
fxx = 2a(2a − 1)x2a−2 y a ,
fyy = a(a − 1)x2a y a−2 .
So,
∂2f ∂2f
x2 2
− 2y 2 2 − 18f + 36xy = 2a(2a − 1)x2a y a − 2a(a − 1)x2a y a − 36xy − 18x2a y a + 36xy.
∂x ∂y
Therefore,
∂2f 2
2∂ f
x2 − 2y − 18f + 36xy = 0
∂x2 ∂y 2
⇐⇒ x2a y a (2a(2a − 1) − 2a(a − 1) − 18) = 0
⇐⇒ 2a(2a − 1) − 2a(a − 1) − 18 = 0
⇐⇒ a2 − 9 = 0
⇐⇒ a = ±3.
New format of the examination in 2010
From the 2010 examination, the format of the examination will change. There will be
no optional questions. Instead, there will be six compulsory questions in Section A
rather than the present seven. These will be worth 60 marks in total. There will be
two compulsory questions in Section B, each worth 20 marks. (There will not be a
choice of questions in Section B.)
11

Important note

references
section.
General remarks
Learning outcomes
At the end of this half course and having completed the Essential reading and activities you should
have:
subjects.
explain the method by which it is obtained. Second, you should include rough working (even if it is
messy!). The examiners want you to get the right answers, of course, but it is more important that
1
questions.
year. Each year, the examiners want to test that candidates know and understand a number of
they vaguely remember them. Because of this, every year there are some questions which are likely
to seem unfamiliar, or different, from previous years’ questions. You should expect to be surprised
by some of the questions. Of course, you will only be examined on material in the syllabus, so all
questions can be answered using the material of the course. There will be enough, routine, familiar
content in the examination so that a candidate who has achieved competence in the course will pass,
but, of course, for a high mark, more is expected: you will have to demonstrate an ability to solve
new and unfamiliar problems.
Answer the question
Please do read the questions carefully. You might be asked to use specific methods, even when
others could be used. The purpose of the examination is to test that you know certain methods, so
the examiners might occasionally ask you to use a specific technique. In such circumstances, only
Graph sketching
mentioned often in the Examiners’ commentaries, the plotting of points in order to graph a function
is not the correct approach. A sketch of the graph of a function should indicate its shape, its
position with respect to the axes, and its intercepts on those axes: it need not be drawn to scale.
Graph paper is not necessary for this.
Calculators
circumstances. The examiners know this, and so they set questions that do not require a calculator.
this course.
2
expected. This may be due to a number of reasons, but one particular failing is ‘question
We recognise that candidates might not cover all topics in the syllabus in the same depth, but you
need to be aware that examiners are free to set questions on any aspect of the syllabus. This
The syllabus can be found in the Course information sheet available on the VLE. You should read
the syllabus carefully and ensure that you cover sufficient material in preparation for the
examination. Examiners will vary the topics and questions from year to year and may well set
questions that have not appeared in past papers. Examination papers may legitimately include
questions on any topic in the syllabus. So, although past papers can be helpful during your revision,
you cannot assume that topics or specific questions that have come up in past examinations will
occur again.
3

Important note

references
section.
Section A
Question 1
The demand equation for a good is q(p + 2) = 6 and the supply equation is
q − p + 3 = 0, where p is the price and q is the quantity demanded.
The demand and supply quantities are:
6
q= and q = p − 3.
p+2
4
We therefore solve:
6
= p − 3.
p+2
This gives:
(p + 2)(p − 3) = 6
which is:
p2 − p − 12 = 0.
Factorising:
(p − 4)(p + 3) = 0
and so the equilibrium price is p∗ = 4. The corresponding equilibrium quantity is q ∗ = p∗ − 3 = 1.
We could, alternatively, express p in terms of q for each of supply and demand and then equate
these two expressions.
Question 2
Z
x2 ex dx,
√
Z
x2 x − 2 dx.
partial fractions. More than one method might work, and some integrals require a combination
of methods.
For the first integral, we can use integration by parts.
Call the integral I. We have:

Z
I = x2 ex dx
Z
= x2 e x − 2 xex dx
Z
= x2 ex − 2 xex − ex dx
= x2 ex − 2xex + 2ex + c.
The second integral can be done in at least two ways: using substitution and using integration by
parts.
5
√
There is more than one substitution that will work. We could use u = x − 2 or u = x − 2. Here
is how it works with u = x − 2. Call the integral I. We then have du = dx and:
√
Z
I = (u + 2)2 u du
Z
= (u2 + 4u + 4)u1/2 du
Z
= u5/2 + 4u3/2 + 4u1/2 du
2 7/2 8 5/2 8 3/2

= u + u + u +c
7 5 3
2 8 8
= (x − 2)7/2 + (x − 2)5/2 + (x − 2)3/2 + c.
7 5 3
Here is how integration by parts can be used. We have:
Z
2 2 3/2 4
I = x (x − 2) − x(x − 2)3/2 dx
3 3
Z
2 2 8 8
= x (x − 2)3/2 − x(x − 2)5/2 + (x − 2)5/2 dx
3 15 15
2 2 8 16
= x (x − 2)3/2 − x(x − 2)5/2 + (x − 2)7/2 + c.
3 15 105
Question 3
The function f is defined, for x > 0, by

f (x) = a ln x − b(ln x)2 ,
where a and b are positive numbers. Show that f has one critical point. Determine
whether it is a local maximum, a local minimum or a point of inflexion.
We have:
a ln x
f 0 (x) =
− 2b .
x x
We solve f 0 (x) = 0. This gives 2b ln x = a, so ln x = a/(2b). Hence x = ea/2b . We have:
a 2b ln x 2b 2b ln x − a 2b
f 00 (x) = − + − 2 = − 2.
x2 x2 x x2 x
At the critical value of x, this equals −2be−a/b , which is negative (since b > 0). Therefore, the
critical point is a local maximum.
Alternatively, since f 0 (x) = (a − 2b ln x) /x, we can see that f 0 changes from positive to negative
as x increases through the critical value. The point is therefore a local maximum.
Question 4
√
Use the Lagrange multiplier method to find the values of x, y which maximise x2 y
subject to the constraint x + y = 100.
6
subject guide.

√
The Lagrangean is L = x2 y − λ(x + y − 100). The first-order conditions are:
∂L √
= 2x y − λ = 0
∂x
∂L 1 x2
= √ −λ=0
∂y 2 y
∂L
= −(x + y − 100) = 0.
∂λ
From the first two equations:
√ 1 x2
λ = 2x y = √ .
2 y
On simplification (elimination of λ), this becomes x = 4y. Hence the third equation shows that
x = 80 and y = 20.
Question 5
Find the critical points of the function
f (x, y) = y 3 − x3 − 2xy + 5.
For each critical point of f , determine whether it is a local minimum, local

maximum, or a saddle point.
This question uses the material in Chapter 5 of the subject guide. It is important in answering
questions like this to use a correct notation for partial derivatives. We can write fx instead of
∂f /∂x and fxx instead of ∂ 2 f /∂x2 , and so on, but it is a very bad idea to invent your own
notation!
We set the partial derivatives to 0. We have
fx = −3x2 − 2y = 0 and fy = 3y 2 − 2x = 0.
From the first equation, y = −3x2 /2. The second equation then gives:
4
9x 27 3
3 − 2x = 0 so x x − 2 = 0.
4 4
This has solutions x = 0 and x = 2/3 and the corresponding values of y are 0 and −2/3. So there
are two critical points: (0, 0) and (2/3, −2/3).
To test the nature of a critical point, we use the second derivative test. We have:
fxx = −6x, fxy = −2 and fyy = 6y.

2
At (0, 0), we have H = fxx fyy − fxy = −4 < 0, so this is a saddle point.
7
At (2/3, −2/3), we have H = 36(2/3)2 − 4 = 12 > 0 and fxx < 0. So this is a local maximum
point.
Question 6

1
xn = xn−1 + 2.
2
Show that xn can be written in the following form, for some numbers a and r
(which you should find):
xn = ar n + 2 1 + r + r 2 + · · · + r n−1 .

Hence, or otherwise, find an explicit expression for xn in terms of n, simplifying

your answer as far as possible. Hence describe the behaviour of xn as n → ∞.
We have:

1
x1 = 3+2
2
2
1 1
x2 = 3+ 2+2
2 2
3 2
1 1 1
x3 = 3+ 2+ 2 + 2.
2 2 2
In general: !
n n−1
1 1 1
xn = 3 +2 + ··· + +1 .
2 2 2
So a = 3 and r = 1/2.
This simplifies to:

n n n n
1 − (1/2)n

1 1 1 1
3 +2 =3 +4 1− =4− .
2 1 − 1/2 2 2 2
As n → ∞, the sequence increases and tends to 4 because (1/2)n decreases and tends to 0.
Section B
Question 7
(a) The function f is given, for x > 0, by

a
f (x) = + bx + cx2
x
8
for some constants a, b and c. The following facts about f and its derivative f 0
are known:
f (1) = 4, f (2) = −3, f 0 (1) = −9.
Show that a, b and c satisfy the following equations:
a+b+c = 4
a + 4b + 8c = −6
a − b − 2c = 9.
By solving this system of equations using row operations, find the values of a, b
and c.

The recommended method for solving linear equations using row operations can be found in
Chapter 6 of the subject guide. It is known by several names: the row operation method,
the Gauss–Jordan method, the row-reduction method, and so on.
We have:
a
f 0 (x) = − + b + 2cx.
x2
The given information therefore implies that:
a+b+c=4
a
+ 2b + 4c = −3
2
and:
−a + b + 2c = −9.
Therefore:
a+b+c = 4
a + 4b + 8c = −6
a − b − 2c = 9.
The system expressed in matrix form is:

    
1 1 1 a 4
1 4 8   b  = −6 .
1 −1 −2 c 9
We perform row operations on the augmented matrix.

Using row operations to reduce the augmented matrix (and there are many ways this can be
done), we have:
     
1 1 1 4 1 1 1 4 1 1 1 4
 1 4 8 −6  →  0 3 7 −10  →  0 3 7 −10  .
1 −1 −2 9 0 −2 −3 5 0 0 5 −5
So, by back-substitution, c = −1; 3b + 7c = −10, meaning b = −1; and a + b + c = 4,

meaning a = 6.
9

mistake). So, here, for instance, examiners will award marks if you can indicate that you
know how to start to solve the equations (by writing down an augmented matrix); that you
know what row operations are; that you know what it is you want to achieve with row
constant from each entry of a row is valid. It is not (and, if you don’t know what we mean
by that, then you’re probably not doing it, which is good!).
1
Z
(b) Determine the integral dx.
1 + e2x

There are several ways to solve this problem. Here are some.
Method A: by substituting u = 1 + e2x and using partial fractions.
Let u = 1 + e2x , then du = 2e2x dx = 2(u − 1) dx and the integral is:
Z
1 du
.
2 (u − 1)u
Partial fractions means there are A, B such that the integrand takes the form:
A B
+ .
u−1 u
We find (by a number of possible methods) that:
A = 1/2 and B = −1/2
and hence the integral is:

Z
1 1 1 1
− du = ln |u − 1| − ln |u| + c
2(u − 1) 2u 2 2
1 1
= ln(e2x ) − ln(1 + e2x ) + c
2 2
1
= x − ln(1 + e2x ) + c.
2
Method B: multiplying numerator and denominator by e−x first.

The integral is:
e−x
Z
dx.
e−x + ex
Now we use a substitution. Let u = e−x , then du = −e−x dx and the integral is:
Z Z
u 1 u
− du = − du
u + 1/u u 1 + u2
1
= − ln(1 + u2 ) + c
2
1
= − ln(1 + e−2x ) + c.
2
10
(This might look like a different answer to the first one, but it is not: can you see why?)
Method C: by multiplying numerator and denominator by e−2x first.
The integral is:
e−2x
Z
dx.
e−2x+1
Note now that the numerator is −1/2 times the derivative of the denominator, or perform
the substitution u = e−2x + 1, to deduce that the integral is:
1
− ln(e−2x + 1) + c.
2
An explicit substitution argument is as follows. We let u = 1 + e−2x , then du = −2e−2x dx.
So the integral is:
Z
1 du 1 1
− = − ln |u| + c = − ln(e−2x + 1) + c.
2 u 2 2
Question 8
(a) The function f is given, for x, y > 0, by

3 y
f (x, y) = y ln 1 + .
x
∂f ∂f
Find the partial derivatives and and verify that
∂x ∂y
∂f ∂f
x +y = 3f (x, y).
∂x ∂y


We have:
1 y y4
fx = y 3 − 2 =− 2
1 + y/x x x + xy
and:
y 1 1 y y3
fy = 3y 2 ln 1 + + y3 = 3y 2 ln 1 + + .
x 1 + y/x x x x+y
So:
∂f ∂f y4 y y4
x +y = − + 3y 3 ln 1 + +
∂x ∂y x+y x x+y
y
= 3y 3 ln 1 + = 3f (x, y).
x
(b) Consider the problem of maximising the function

−1
1 2
q(k, l) = +
k l
11
subject to the constraint vk + wl = M , where v, w, M are positive constants.

By using the Lagrange multiplier method, show that the optimal values of k and
l satisfy
1/2
2v
l= k.
w
Suppose now that v = 2w. Find the optimal values of k and l in terms of M and
v. Then find the maximum value of q(k, l), also in terms of M and v.

The Lagrange multiplier method for constrained optimisation is discussed in Chapter 5 of
the subject guide.
The Lagrangean is:
−1
1 2
L= + − λ(vk + wl − M ).
k l
−2
∂L 1 1 2
= + − λv = 0
∂k k2 k l
−2
∂L 2 1 2
= + − λw = 0
∂l l2 k l
∂L
= −(vk + wl − M ) = 0.
∂λ
−2 −2
1 1 2 2 1 2
λ= + = +
vk 2 k l wl2 k l
so:
l2 2v
=
k2 w
and hence: 1/2
2v
l= k.
w
It is sensible now to use the fact that v = 2w. It would be possible to continue without this
and solve generally, and then use this fact, but this is more difficult than using the fact now.
Using v = 2w, we have:
1/2
2v
l= k = 41/2 k = 2k.
w
Now we use the constraint equation (equivalently, the third of the first-order conditions).
We have vk + wl = M , so vk + w(2k) = M and therefore (since v = 2w) 2vk = M and hence
k = M/(2v). It follows that l = M/v.
The optimal value of q is:
−1 −1
1 2 2v 2v M
q∗ = + = + = .
(M/(2v)) (M/v) M M 4v
12

Important note

references
section.
Section A
Question 1
The demand equation for a good is q(p + 3) = 8 and the supply equation is
q − p + 4 = 0, where p is the price and q is the quantity demanded.
The demand and supply quantities are:

8
q= and q = p − 4.
p+3
We therefore solve:
8
= p − 4.
p+3
13
This gives:
(p + 3)(p − 4) = 8
which is:
p2 − p − 20 = 0.
Factorising:
(p − 5)(p + 4) = 0
and so the equilibrium price is p∗ = 5. The corresponding equilibrium quantity is q ∗ = p∗ − 4 = 1.
We could, alternatively, express p in terms of q for each of supply and demand and then equate
these two expressions.
Question 2
Z
x2 ex dx,
√
Z
x2 x − 1 dx.
partial fractions. More than one method might work, and some integrals require a combination
of methods.
For the first integral, we can use integration by parts.
Call the integral I. We have:

Z
I = x2 ex dx
Z
2 x
= x e −2 xex dx
Z
2 x x x
= x e − 2 xe − e dx
= x2 ex − 2xex + 2ex + c.
The second integral can be done in at least two ways: using substitution and using integration by
parts.
√
There is more than one substitution that will work. We could use u = x − 1 or u = x − 1. Here
14
is how it works with u = x − 2. Call the integral I. We then have du = dx and:

√
Z
I = (u + 1)2 u du
Z
= (u2 + 2u + 1)u1/2 du
Z
= u5/2 + 2u3/2 + u1/2 du
2 7/2 4 5/2 2 3/2

= u + u + u +c
7 5 3
2 4 2
= (x − 1)7/2 + (x − 1)5/2 + (x − 1)3/2 + c.
7 5 3
Here is how integration by parts can be used. We have:
Z
2 2 4
I = x (x − 1)3/2 − x(x − 1)3/2 dx
3 3
Z
2 2 8 8
= x (x − 1)3/2 − x(x − 1)5/2 + (x − 1)5/2 dx
3 15 15
2 2 8 16
= x (x − 1)3/2 − x(x − 1)5/2 + (x − 1)7/2 + c.
3 15 105
Question 3
The function f is defined, for x > 0, by
f (x) = a ln x − b(ln x)2 ,
where a and b are positive numbers. Show that f has one critical point. Determine
whether it is a local maximum, a local minimum or a point of inflexion.
We have:
a ln x
f 0 (x) =
− 2b .
x x
We solve f 0 (x) = 0. This gives 2b ln x = a, so ln x = a/(2b). Hence x = ea/2b . We have:
a 2b ln x 2b 2b ln x − a 2b
f 00 (x) = − 2
+ 2
− 2 = 2
− 2.
x x x x x
At the critical value of x, this equals −2be−a/b , which is negative (since b > 0). Therefore, the
critical point is a local maximum.
Alternatively, since f 0 (x) = (a − 2b ln x) /x, we can see that f 0 changes from positive to negative
as x increases through the critical value. The point is therefore a local maximum.
Question 4
√
Use the Lagrange multiplier method to find the values of x, y which maximise y 2 x
subject to the constraint x + y = 100.
15
subject guide.

√
The Lagrangean is L = x2 y − λ(x + y − 100). The first-order conditions are:
∂L 1 y2
= √ −λ=0
∂x 2 x
∂L √
= 2y x − λ = 0
∂y
∂L
= −(x + y − 100) = 0.
∂λ
√
λ = (1/2)x−1/2 y 2 = 2y x.
On simplification (elimination of λ), this becomes y = 4x. Then the third equation shows that
x = 20 and y = 80.
Question 5
Find the critical points of the function
f (x, y) = x3 − y 3 − 2xy + 3.
For each critical point of f , determine whether it is a local minimum, local

maximum, or a saddle point.
This question uses the material in Chapter 5 of the subject guide. It is important in answering
questions like this to use a correct notation for partial derivatives. We can write fx instead of
∂f /∂x and fxx instead of ∂ 2 f /∂x2 , and so on, but it is a very bad idea to invent your own
notation!
We set the partial derivatives to 0. We have:
fx = 3x2 − 2y = 0 and fy = −3y 2 − 2x = 0.
From the first equation, y = 3x2 /2. The second equation then gives:
4
9x 27 3
−3 − 2x = 0 so −x x + 2 = 0.
4 4
This has solutions x = 0, −2/3 and the corresponding values of y are 0, 2/3. So there are two
critical points: (0, 0) and (−2/3, 2/3).
To test the nature of a critical point, we use the second derivative test. We have:
fxx = 6x, fxy = −2 and fyy = −6y.

2
At (0, 0), we have H = fxx fyy − fxy = −4 < 0, so this is a saddle point.
16
At (−2/3, 2/3), we have H = 36(2/3)2 − 4 = 12 > 0 and fxx < 0. So this is a local maximum
point.
Question 6

1
xn = xn−1 + 2.
3
Show that xn can be written in the following form, for some numbers a and r
(which you should find):
xn = ar n + 2 1 + r + r 2 + · · · + r n−1 .

Hence, or otherwise, find an explicit expression for xn in terms of n, simplifying

your answer as far as possible. Hence describe the behaviour of xn as n → ∞.
We have:

1
x1 = 4+2
3
2
1 1
x2 = 4+ 2+2
3 3
3 2
1 1 1
x3 = 4+ 2+ 2 + 2.
3 3 3
In general: !
n n−1
1 1 1
xn = 4 +2 + ··· + +1 .
3 3 3
So a = 4 and r = 1/3.
This simplifies to:

n n n n
1 − (1/3)n

1 1 1 1
4 +2 =4 +3 1− =3+ .
3 1 − 1/3 3 3 3
As n → ∞, the sequence decreases and tends to 3 because (1/3)n decreases and tends to 0.
Section B
Question 7
(a) The function f is given, for x > 0, by

a
f (x) = + bx + cx2
x
17
for some constants a, b and c. The following facts about f and its derivative f 0
are known:
f (1) = 8, f (2) = 2, f 0 (1) = −9.
Show that a, b and c satisfy the following equations:
a+b+c = 8
a + 4b + 8c = 4
a − b − 2c = 9.
By solving this system of equations using row operations, find the values of a, b
and c.

We have:
a
f 0 (x) = − + b + 2cx.
x2
The given information therefore implies that:
a+b+c=8
a
+ 2b + 4c = 2
2
and:
−a + b + 2c = −9.
Therefore:
a+b+c = 8
a + 4b + 8c = 4
a − b − 2c = 9.
The system expressed in matrix form is:

    
1 1 1 a 8
1 4 8   b  = 4 .
1 −1 −2 c 9
We perform row operations on the augmented matrix.

Using row operations to reduce the augmented matrix (and there are many ways this can be
done), we have:
     
1 1 1 8 1 1 1 8 1 1 1 8
 1 4 8 4 → 0 3 7 −4  →  0 3 7 −4  .
1 −1 −2 9 0 −2 −3 1 0 0 5 −5
So, by back-substitution, c = −1; 3b + 7c = −4, meaning b = 1; and a + b + c = 8, meaning

a = 8.
18

mistake). So, here, for instance, examiners will award marks if you can indicate that you
know how to start to solve the equations (by writing down an augmented matrix); that you
know what row operations are; that you know what it is you want to achieve with row
constant from each entry of a row is valid. It is not (and, if you don’t know what we mean
by that, then you’re probably not doing it, which is good!).
1
Z
(b) Determine the integral dx.
1 + e2x

There are several ways to solve this problem. Here are some.
Method A: by substituting u = 1 + e2x and using partial fractions.
Let u = 1 + e2x , then du = 2e2x dx = 2(u − 1) dx and the integral is:
Z
1 du
.
2 (u − 1)u
A B
+ .
u−1 u
A = 1/2 and B = −1/2

Z
1 1 1 1
− du = ln |u − 1| − ln |u| + c
2(u − 1) 2u 2 2
1 1
= ln(e2x ) − ln(1 + e2x ) + c
2 2
1
= x − ln(1 + e2x ) + c.
2
Method B: multiplying numerator and denominator by e−x first.

The integral is:
e−x
Z
dx.
e−x + ex
Now we use a substitution. Let u = e−x , then du = −e−x dx and the integral is:
Z Z
u 1 u
− du = − du
u + 1/u u 1 + u2
1
= − ln(1 + u2 ) + c
2
1
= − ln(1 + e−2x ) + c.
2
19
(This might look like a different answer to the first one, but it is not: can you see why?)
Method C: by multiplying numerator and denominator by e−2x first.
The integral is:
e−2x
Z
dx.
e−2x +1
Note now that the numerator is −1/2 times the derivative of the denominator, or perform
the substitution u = e−2x + 1, to deduce that the integral is:
1
− ln(e−2x + 1) + c.
2
An explicit substitution argument is as follows. We let u = 1 + e−2x , then du = −2e−2x dx.
So the integral is:
Z
1 du 1 1
− = − ln |u| + c = − ln(e−2x + 1) + c.
2 u 2 2
Question 8
(a) The function f is defined, for x, y > 0, by

y x
f (x, y) = ln .
x y
∂f ∂f
Find the first-order partial derivatives . and
∂x ∂y
Use your result to find the first-order partial derivatives of the function g,
defined for x, y > 0 by
y/x
x
g(x, y) = .
y

We have:
y x y1y y x y
fx = − 2 ln + = − 2 ln + 2
x y xyx x y x
and:
1 x y x y 1 x 1
fy = ln − 2
= ln − .
x y xy x x y x
We have ln g = f . Therefore:
1
gx = fx .
g
So: y/x
y x y x
gx = gfx = − 2 ln + 2 .
x y x y
Similarly:
y/x
1 x 1 x
gy = gfy = ln − .
x y x y
20
(b) Consider the problem of maximising the function

−1
1 2
q(k, l) = +
k l
subject to the constraint vk + wl = M , where v, w, M are positive constants.
By using the Lagrange multiplier method, show that the optimal values of k and
l satisfy
1/2
2v
l= k.
w
Suppose now that v = 2w. Find the optimal values of k and l in terms of M and
v. Then find the maximum value of q(k, l), also in terms of M and v.

The Lagrange multiplier method for constrained optimisation is discussed in Chapter 5 of
the subject guide.
The Lagrangean is:
−1
1 2
L= + − λ(vk + wl − M ).
k l
−2
∂L 1 1 2
= + − λv = 0
∂k k2 k l
−2
∂L 2 1 2
= + − λw = 0
∂l l2 k l
∂L
= −(vk + wl − M ) = 0.
∂λ
−2 −2
1 1 2 2 1 2
λ= + = +
vk 2 k l wl2 k l
so:
l2 2v
2
=
k w
and hence: 1/2
2v
l= k.
w
It is sensible now to use the fact that v = 2w. It would be possible to continue without this
and solve generally, and then use this fact, but this is more difficult than using the fact now.
Using v = 2w, we have:
1/2
2v
l= k = 41/2 k = 2k.
w
Now we use the constraint equation (equivalently, the third of the first-order conditions).
We have vk + wl = M , so vk + w(2k) = M and therefore (since v = 2w) 2vk = M and hence
k = M/(2v). It follows that l = M/v.
The optimal value of q is:
−1 −1
1 2 2v 2v M
q∗ = + = + = .
(M/(2v)) (M/v) M M 4v
21

Important note

references
section.
General remarks
Learning outcomes
have:
subjects.
1
questions.
Answer the question
Graph sketching
Calculators
this course.
2
occur again.
3

Important note

references
section.
Section A
Question 1
Find the points of intersection of the curves
y = 3x − 1 and y = x2 + 2x − 3.
Sketch these curves on the same set of axes.
Curve sketching and related topics are discussed in Chapter 2 of the subject guide.
The x-coordinate of a point of intersection would satisfy:
3x − 1 = x2 + 2x − 3.

x2 − x − 2 = 0
4
which is:
(x − 2)(x + 1) = 0
and this has solutions x = 2, −1. The corresponding y-values are 5, −4 and the intersection
points are therefore (2, 5) and (−1, −4).
The sketch (or the calculations) should show:
• the U -shape of y = x2 + 2x − 3
• the position of the parabola y = x2 + 2x − 3
• the shape and position of the line
• the intercepts of y = 3x − 1 on axes: (0, −1) and (1/3, 0)
• that y = x2 + 2x − 3 intercepts the x-axis when x2 + 2x + 3 = 0, so at x = −3, 1
• that y = x2 + 2x − 3 has a minimum when y 0 = 0, so x = −1
• the intercept of y = x2 + 2x − 3 on the y-axis is (0, −3) and the minimum value is y = −4.
A sketch is as follows:
There are no marks merely for plotting points on graph paper. There is, in fact, no need to use
graph paper.
Question 2
A monopoly has a marginal cost function given by
M C(q) = eq −2q,
and fixed costs of 10. Find its total cost function.
5
The demand for its product is given by the equation
p + q = 20.
Find the profit function for this monopoly as a function of q.
What value of q will maximise this profit?
See Chapter 3 (and, for integration, Chapter 4) of the subject guide.
We have: Z Z
TC = M C dq = (eq − 2q) dq = eq − q 2 + c.
We have T C(0) = F C = 10. So, we have 10 = e0 − 0 + c, so c = 9. It is not correct, as many

thought, simply to assume that the ‘c’ must be the fixed cost of 10.
So T C = eq − q 2 + 9. The firm is a monopoly, so the selling price in terms of its production is,
from the demand equation, p = 20 − q. So T R = q(20 − q) and:
Π = T R − T C = q(20 − q) − (eq − q 2 + 9) = 20q − eq − 9.
To maximise Π, we set Π0 = 0. So 20 − eq = 0 and q = ln 20. This is a (local) maximum since

Π00 (q) = −eq is negative.
Question 3
Use row operations to solve the system of equations
x + 2y + 2z = 3,
5x − 2y + 3z = 18,
3x + 3y − z = −6.
Chapter 6 of the subject guide. It is known by several names: the row operation method, the
Gauss–Jordan method, the row-reduction method, and so on.
The standard matrix method approach is to reduce the augmented matrix to reduced form. Here
     
1 2 2 3 1 2 2 3 1 2 2 3
 5 −2 3 18  →  0 −12 −7 3 → 0 −12 −7 3 
3 3 −1 −6 0 −3 −7 −15 0 12 28 60
 
1 2 2 3
→ 0 −12 −7 3 .
0 0 21 63
6
It follows, from this last matrix, that:
x + 2y + 2z = 3
−12y − 7z = 3
21z = 63.
So, z = 3, y = (1/12)(−3 − 7z) = −2 and x = 3 − 2y − 2z = 1.
The examiners understand that arithmetical errors can be made, especially in the stressful
circumstances of an examination. Quite probably, the examiners themselves would make some
mistakes if they sat the paper. So, although there are marks for correct calculations, there are
examiners will award marks if you can indicate that you know how to start to solve the equations
(by writing down an augmented matrix); that you know what row operations are; that you know
what it is you want to achieve with row operations (the reduced matrix, that is); and that you
then know how to work from that reduced matrix to determine the required solutions. There are
marks for all of these things.
constant from each entry of a row is valid. It is not. (And, if you do not know what we mean by
that, then you are probably not doing it, which is good!)
It is also important to answer the question. The question explicitly asks you to use row
operations. Given this, other approaches to solving the system of equations are not acceptable
because to take such an approach is not to answer the question.
Question 4
Find the integrals
ex dx
Z Z
(i) (x + 1)2 ex dx and (ii) .
(1 − ex )(1 + ex )
See Chapter 4 of the subject guide.
(i) We use integration by parts. We have:

Z Z
I = (x + 1) e dx = (x + 1) e −2 (x + 1) ex dx
2 x 2 x
Z
= (x + 1)2 ex −2 (x + 1) ex − ex dx
= (x + 1)2 ex −2(x + 1) ex +2 ex +c.
7
(ii) We use a substitution. Let u = ex . (Other substitutions, such as u = 1 − ex or u = 1 + ex

will also work). Then du = ex dx and the integral is:
Z
du
.
(1 − u)(1 + u)
A B
+ .
1−u 1+u
A = 1/2 and B = 1/2
Z
1/2 1/2 1 1
+ du = − ln |1 − u| + ln |1 + u| + c
1−u 1+u 2 2
1 1
= − ln |1 − ex | + ln |1 + ex | + c.
2 2
Some candidates asserted at the outset that, for some constants A and B, we have:
ex A B
= + .
(1 − ex )(1 + ex ) 1 − ex 1 + ex
Now, actually, it turns out that such A and B do exist, but it is incorrect to say that this is
‘by partial fractions’. (It is not partial fractions, which is a method that simplifies
expressions that involve one polynomial divided by another.) A correct answer would have
to determine A and B and verify that they work. It turns out that, indeed:
ex 1/2 1/2
= − .
(1 − ex )(1 + ex ) 1 − ex 1 + ex
This can be verified by taking a common denominator and simplifying the right-hand side.
However, does this help integrate the function? To integrate 1/(1 − ex ), for example, you
will need to do something like a substitution, since it cannot be integrated directly. Taking
u = ex , for example: Z Z
dx du
=
1 − ex u(1 − u)
and we need ultimately to use (proper) partial fractions.
Question 5
Use the method of Lagrange multipliers to find the values of x and y that minimise
2x + 3y subject to the constraint x1/2 y 1/3 = 48 with x, y > 0.
subject guide.
The Lagrangian is L = 2x + 3y − λ(x1/2 y 1/3 − 48). The first-order conditions are:

∂L λ y 1/3
=2− =0
∂x 2 x1/2
∂L λ x1/2
=3− =0
∂y 3 y 2/3
∂L
= −(x1/2 y 1/3 − 48) = 0.
∂λ
8

4x1/2 9y 2/3
λ= = .
y 1/3 x1/2
On simplification, this becomes x = 9y/4. Then, substituting in the third equation shows that:
1/2
9
y y 1/3 = 48.
4
So y 5/6 = (2/3)48 = 32 and, therefore, y = (32)6/5 = 64. (It is acceptable to leave this as
(32)6/5 .) Then, x = (9/4)64 = 144. (It is acceptable to write x = 9((32)6/5 )/4.)
Question 6
At the beginning of 2018, Martin takes out a loan of $1,000 at an interest rate of 2%
per month. He has to make monthly repayments of $D at the beginning of each
subsequent month for the next 25 months.
Explain why he owes 1000(1.02) − D after his first repayment.
How much does he owe after n repayments? (Simplify your answer as far as
possible.)
Given that 1.0225 is approximately 1.64, show that D is approximately 51.25.
The amount owed after the first repayment is the initial loan 1000, plus interest (which is
0.02 × 1000), minus the payment D; so it is:
1000 + (0.02)(1000) − D = 1000(1.02) − D.
After the second repayment, the amount owing is:
1000(1.02)2 − D(1.02) − D.
After the third, it is:

1000(1.02)3 − D(1.02)2 − D(1.02) − D.
In general, after n repayments, it is:
1000(1.02)n − D(1.02)n−1 − D(1.02)n−2 − · · · − (1.02)D − D.
This is:
(1.02)n − 1

n
1000(1.02) − D = 1000(1.02)n − 50D((1.02)n − 1).
1.02 − 1
Here, we have used the formula for the sum of a geometric progression.
After 25 repayments, the balance is zero and hence:
0 = 1000((1.02)25 − 50D((1.02)25 − 1).
So:
20(1.02)25 20(1.64) 20(164)
D= 25
= = = 51.25.
(1.02) − 1 1.64 − 1 64
9
Section B
Question 7
(a) A company is the only producer of two goods, X and Y. When the prices for X
and Y are pX and pY per unit (respectively), consumers will purchase
quantities x and y (respectively) according to the demand equations
pY + pX = 45 − 3x − 3y and pY − pX = 11 + x − y.
If the company’s joint total cost function (that is, the cost of producing
quantities x of X and y of Y) is
TC(x, y) = 15 + 3x2 − 3xy + 3y 2 ,
find their profit function, π(x, y).

Hence determine the quantities x and y that maximise this profit.

The question asks us to find the profit as a function of x and y. That means we will need
the revenue, and hence the prices, in terms of x and y. Adding the two equations, we have
2pY = 56 − 2x − 4y, so pY = 28 − x − 2y. Then, pX = pY − 11 − x + y = 17 − 2x − y. The
profit is therefore:
Π = T R − T C = xpX + ypY − (15 + 3x2 − 3xy + 3y 2 ) = 17x − 5x2 + 28y − 5y 2 + xy − 15.
Solving for the critical points, we need:
Πx = 17 − 10x + y = 0, Πy = 28 − 10y + x = 0
10x − y = 17, x − 10y = −28.
So y = 10x − 17 and x − 10(10x − 17) = −28, which means −99x = −198 and x = 2. Then,
y = 20 − 17 = 3. The solution is (x, y) = (2, 3). We have:
Πxx = −10, Πyy = −10 and Πxy = 1.
So, since Πxx = −10 < 0 and:
Πxx Πyy − (Πxy )2 = (−10)(−10) − (1)2 > 0
this point maximises profit.
(b) Show that for all values of the number k, the function
f (x, y) = x2 + kxy + y 2 ,
has a critical (or stationary) point at (0, 0).

For each value of k, determine whether this critical point is a local minimum, a
local maximum or a saddle point.

10

We have:
fx = 2x + ky and fy = kx + 2y.
At (0, 0), fx = 0 and fy = 0, so it is a critical point.
Note: the question only asks us to show that there is a critical point at (0, 0), so it is enough
to substitute these coordinates into the partial derivatives and show that they are zero.
There is no need to show that the only critical point is (0, 0) by completely solving the
partial derivatives equal to zero.
We have:
fxx = 2, fxy = k and fyy = 2
so the Hessian is:
2
H = fxx fyy − fxy = 4 − k2 .
If |k| < 2, H > 0 and fxx > 0, so we have a local minimum. If |k| > 2, H < 0 and we have a
saddle point.
If |k| = 2, H = 0 and the test fails. However, in this case, when k = ±2 we have:
f (x, y) = x2 ± 2xy + y 2 = (x ± y)2 ≥ 0 = f (0, 0)
so (0, 0) is a minimum. (Why? Well, what it means for a point to be a minimum is that all
the values of the function nearby are at least as large. This argument shows that all other
values are at least as large as the value 0, at (0, 0), so the point (0, 0) is a (global) minimum.)
Question 8
A firm has a production function given by
q(k, l) = k1/4 l1/12
where k and l are the amounts of capital and labour used.
(a) If the prices of capital and labour are u and v per unit respectively, use the
method of Lagrange multipliers to find the values of k and l that will minimise
the firm’s costs when they produce a quantity Q.
The firm sells its product at a fixed price of w per unit.
(b) What is the firm’s profit function?

(c) Determine the firm’s breakeven quantity.
(d) Find the quantity that maximises the firm’s profit.
See Chapter 5 of the subject guide. The breakeven point is defined in Section 3.8 of the subject
guide. It is the production level at which the profit changes from negative to positive.
(a) The cost is uk + vl. We have to minimise this subject to the constraint k 1/4 l1/12 = Q, so the
Lagrangian is:
L = uk + vl − λ(k 1/4 l1/12 − Q).
11
The optimal k and l satisfy the three equations:

λ l1/12
Lk = u − =0
4 k 3/4
λ k 1/4
Ll = v − =0
12 l11/12
k 1/4 l1/12 = Q.
λ = 4uk 3/4 l−1/12 = 12vl11/12 k −1/4
so that k = 3(v/u)l. Then:
1/4
3v
k 1/4 l1/12 = l l1/12 = Q
u
and so: u 1/4
l1/3 = Q
3v
so: u 3/4
l= Q3 .
3v
Then: 1/4
3v 3v u 3/4 3 3v
k= l= Q = Q3 .
u u 3v u
(b) The profit is:
Π(Q) = wQ − C
where: 1/4
3v u 3/4
C = Cmin = u Q3 + v Q3 = 4(3−3/4 )u3/4 v 1/4 Q3 .
u 3v
So:
Π = wQ − 4(3−3/4 )u3/4 v 1/4 Q3 .
(c) Breakeven is when Π = 0. So, we need:

Q w − 4(3−3/4 )u3/4 v 1/4 Q2 = 0.
So (rejecting Q = 0):
w 33/4
Q2 =
4 u3/4 v 1/4
and so: √ √ 1/8
w 33/8 w 27
Q= = .
2 u3/8 v 1/8 2 u3 v
(d) We have:
12 3/4 1/4 2
Π0 = w − u v Q .
33/4
This is 0 when:
33/4 w 1
Q2 = .
12 u v 1/4
3/4
So: √ √ 1/8
w 33/8 w 27
Q= = .
12 u3/8 v 1/8 12 u3 v
Also:
24 3/4 1/4
Π00 = − u v Q<0
33/4
so it does give a maximum.
12

Important note

references
section.
Section A
Question 1
Find the points of intersection of the curves
y = 1 − 3x and y = 3 − 2x − x2 .
Sketch these curves on the same set of axes.
Curve sketching and related topics are discussed in Chapter 2 of the subject guide.
The x-coordinate of a point of intersection would satisfy:
1 − 3x = 3 − 2x − x2 .

x2 − x − 2 = 0
13
which is:
(x − 2)(x + 1) = 0
and this has solutions x = 2, −1. The corresponding y-values are −5, 4 and the intersection
points are therefore (2, −5) and (−1, 4).
The sketch (or the calculations) should show:
• the inverted U -shape of y = 3 − 2x − x2

• the position of the parabola y = 3 − 2x − x2
• the shape and position of the line
• the intercepts of y = 1 − 3x on axes: (0, 1) and (1/3, 0)
• that y = 3 − 2x − x2 intercepts the x-axis when 3 − 2x − x2 = 0, so at x = −3, 1
• that y = 3 − 2x − x2 has a maximum when y 0 = 0, so x = −1
• the intercept of y = 3 − 2x − x2 on the y-axis is (0, 3) and the maximum value is y = 4.
A sketch is as follows:
There are no marks merely for plotting points on graph paper. There is, in fact, no need to use
graph paper.
Question 2
A monopoly has a marginal cost function given by

M C(q) = eq −4q,
and fixed costs of 12. Find its total cost function.
The demand for its product is given by the equation

p + 2q = 14.
14
Find the profit function for this monopoly as a function of q.
What value of q will maximise this profit?
We have: Z Z
TC = M C dq = (eq −4q) dq = eq −2q 2 + c.
We have T C(0) = F C = 12. So, we have 10 = e0 −0 + c, so c = 11. It is not correct, as many

thought, simply to assume that the ‘c’ must be the fixed cost of 10.
So T C = eq −2q 2 + 11. The firm is a monopoly, so the selling price in terms of its production is,
from the demand equation, p = 14 − 2q. So T R = q(14 − 2q) and:
Π = T R − T C = q(14 − 2q) − (eq −2q 2 + 11) = 14q − eq −11.
To maximise Π, we set Π0 = 0. So 14 − eq = 0 and q = ln 14. This is a (local) maximum since
Π00 (q) = − eq is negative.
Question 3
Use row operations to solve the system of equations

x + 3y + z = 4,
4x + 2y − z = 1,
2x − 4y + 3z = 23.
Chapter 6 of the subject guide. It is known by several names: the row operation method, the
Gauss–Jordan method, the row-reduction method, and so on.
The standard matrix method approach is to reduce the augmented matrix to reduced form. Here
     
1 3 1 4 1 3 1 4 1 3 1 4
 4 2 −1 1  →  0 −10 −5 −15  →  0 2 1 3 
2 −4 3 23 0 −10 1 15 0 0 6 30
 
1 3 1 4
→  0 2 1 3 .
0 0 1 5
It follows, from this last matrix, that:

x + 3y + z = 4
2y + z = 3
z = 5.
15
So, z = 5, y = (1/2)(3 − z) = −1 and x = 4 − 3y − z = 2.
The examiners understand that arithmetical errors can be made, especially in the stressful
circumstances of an examination. Quite probably, the examiners themselves would make some
mistakes if they sat the paper. So, although there are marks for correct calculations, there are
examiners will award marks if you can indicate that you know how to start to solve the equations
(by writing down an augmented matrix); that you know what row operations are; that you know
what it is you want to achieve with row operations (the reduced matrix, that is); and that you
then know how to work from that reduced matrix to determine the required solutions. There are
marks for all of these things.
constant from each entry of a row is valid. It is not. (And, if you do not know what we mean by
that, then you are probably not doing it, which is good!)
operations. Given this, other approaches to solving the system of equations are not acceptable
because to take such an approach is not to answer the question.
Question 4
Find the integrals

ex dx
Z Z
(i) (x + 2)2 ex dx and (ii) .
(3 − ex )(3 + ex )
(i) We use integration by parts. We have:

Z Z
I = (x + 2)2 ex dx = (x + 2)2 ex −2 (x + 2) ex dx
Z
2 x x x
= (x + 2) e −2 (x + 2) e − e dx
= (x + 2)2 ex −2(x + 2) ex +2 ex +c.
(ii) We use a substitution. Let u = ex . (Other substitutions, such as u = 3 − ex or u = 3 + ex

will also work). Then du = ex dx and the integral is:
Z
du
.
(3 − u)(3 + u)
A B
+ .
3−u 3+u
16
A = 1/6 and B = 1/6

Z
1/6 1/6 1 1
+ du = − ln |3 − u| + ln |3 + u| + c
3−u 3+u 6 6
1 1
= − ln |3 − ex | + ln |3 + ex | + c.
6 6
Some candidates asserted at the outset that, for some constants A and B, we have:
ex A B
= + .
(3 − ex )(3 + ex ) 3 − ex 3 + ex
Now, actually, it turns out that such A and B do exist, but it is incorrect to say that this is
‘by partial fractions’. (It is not partial fractions, which is a method that simplifies
expressions that involve one polynomial divided by another.) A correct answer would have
to determine A and B and verify that they work. It turns out that, indeed:
ex 1/6 1/6
= − .
(3 − ex )(3 + ex ) 3 − ex 3 + ex
This can be verified by taking a common denominator and simplifying the right-hand side.
However, does this help integrate the function? To integrate 1/(3 − ex ), for example, you
will need to do something like a substitution, since it cannot be integrated directly. Taking
u = ex , for example: Z Z
dx du
=
3 − ex u(3 − u)
and we need ultimately to use (proper) partial fractions.
Question 5
Use the method of Lagrange multipliers to find the values of x and y that minimise
3x + 2y subject to the constraint x1/3 y 1/2 = 48 with x, y > 0.
subject guide.
The Lagrangian is L = 3x + 2y − λ(x1/3 y 1/2 − 48). The first-order conditions are:
∂L λ y 1/2
=3− =0
∂x 3 x2/3
∂L λ x1/3
=2− =0
∂y 2 y 1/2
∂L
= −(x1/3 y 1/2 − 48) = 0.
∂λ
9x2/3 4y 1/2
λ= = .
y 1/2 x1/3
17
On simplification, this becomes y = 9x/4. Then, substituting in the third equation shows that:
1/2
9
x1/3 x = 48.
4
So x5/6 = (2/3)48 = 32 and, therefore, x = (32)6/5 = 64. (It is acceptable to leave this as
(32)6/5 .) Then, y = (9/4)64 = 144. (It is acceptable to write y = 9((32)6/5 )/4.)
Question 6
At the beginning of 2018, Martin takes out a loan of $1,000 at an interest rate of 2%
per month. He has to make monthly repayments of $D at the beginning of each
subsequent month for the next 25 months.
Explain why he owes 1000(1.02) − D after his first repayment.
How much does he owe after n repayments? (Simplify your answer as far as
possible.)
Given that 1.0225 is approximately 1.64, show that D is approximately 51.25.
The amount owed after the first repayment is the initial loan 1000, plus interest (which is
0.02 × 1000), minus the payment D; so it is:
1000 + (0.02)(1000) − D = 1000(1.02) − D.
After the second repayment, the amount owing is:
1000(1.02)2 − D(1.02) − D.
After the third, it is:

1000(1.02)3 − D(1.02)2 − D(1.02) − D.
In general, after n repayments, it is:
1000(1.02)n − D(1.02)n−1 − D(1.02)n−2 − · · · − (1.02)D − D.
This is:
(1.02)n − 1

1000(1.02)n − D = 1000(1.02)n − 50D((1.02)n − 1).
1.02 − 1
Here, we have used the formula for the sum of a geometric progression.
After 25 repayments, the balance is zero and hence:
0 = 1000((1.02)25 − 50D((1.02)25 − 1).
So:
20(1.02)25 20(1.64) 20(164)
D= = = = 51.25.
(1.02)25 − 1 1.64 − 1 64
18
Section B
Question 7
(a) A company is the only producer of two goods, X and Y. When the prices for X
and Y are pX and pY per unit (respectively), consumers will purchase
quantities x and y (respectively) according to the demand equations
pX + pY = 54 − 4x − 4y and pX − pY = 22 − 2x + 2y.
If the company’s joint total cost function (that is, the cost of producing
quantities x of X and y of Y) is
TC(x, y) = 10 + 2x2 − 3xy + 2y 2 ,
find their profit function, π(x, y).

Hence determine the quantities x and y that maximise this profit.

The question asks us to find the profit as a function of x and y. That means we will need
the revenue, and hence the prices, in terms of x and y. Adding the two equations, we have
2pX = 76 − 6x − 2y, so pX = 38 − 3x − y. Then, pY = pX − 22 + 2x − 2y = 16 − x − 3y. The
profit is therefore:
Π = T R − T C = xpX + ypY − (10 + 2x2 − 3xy + 2y 2 ) = 38x − 5x2 + 16y − 5y 2 + xy − 10.
Solving for the critical points, we need:
Πx = 38 − 10x + y = 0, Πy = 16 + x − 10y = 0.
10x − y = 38, x − 10y = −16.
So y = 10x − 38 and x − 10(10x − 38) = −16, which means −99x = −396 and x = 4. Then,
y = 40 − 38 = 2. The solution is (x, y) = (4, 2). We have:
Πxx = −10, Πyy = −10 and Πxy = 1.
So, since Πxx = −10 < 0 and:
Πxx Πyy − (Πxy )2 = (−10)(−10) − (1)2 > 0
this point maximises profit.
(b) Show that for all values of the number k, the function
f (x, y) = x2 + kxy + y 2 ,
has a critical (or stationary) point at (0, 0).

For each value of k, determine whether this critical point is a local minimum, a
local maximum or a saddle point.

19

We have:
fx = 2x + ky and fy = kx + 2y.
At (0, 0), fx = 0 and fy = 0, so it is a critical point.

Note: the question only asks us to show that there is a critical point at (0, 0), so it is enough
to substitute these coordinates into the partial derivatives and show that they are zero.
There is no need to show that the only critical point is (0, 0) by completely solving the
partial derivatives equal to zero.
We have:
fxx = 2, fxy = k and fyy = 2
so the Hessian is:

2
H = fxx fyy − fxy = 4 − k2 .
If |k| < 2, H > 0 and fxx > 0, so we have a local minimum. If |k| > 2, H < 0 and we have a
saddle point.
If |k| = 2, H = 0 and the test fails. However, in this case, when k = ±2, we have:
f (x, y) = x2 ± 2xy + y 2 = (x ± y)2 ≥ 0 = f (0, 0)
so (0, 0) is a minimum. (Why? Well, what it means for a point to be a minimum is that all
the values of the function nearby are at least as large. This argument shows that all other
values are at least as large as the value 0, at (0, 0), so the point (0, 0) is a (global) minimum.)
Question 8
A firm has a production function given by
q(k, l) = k1/4 l1/12
where k and l are the amounts of capital and labour used.
(a) If the prices of capital and labour are u and v per unit respectively, use the
method of Lagrange multipliers to find the values of k and l that will minimise
the firm’s costs when they produce a quantity Q.
The firm sells its product at a fixed price of w per unit.
(b) What is the firm’s profit function?

(c) Determine the firm’s breakeven quantity.
(d) Find the quantity that maximises the firm’s profit.
See Chapter 5 of the subject guide. The breakeven point is defined in Section 3.8 of the subject
guide. It is the production level at which the profit changes from negative to positive.
(a) The cost is uk + vl. We have to minimise this subject to the constraint k 1/4 l1/12 = Q, so the
Lagrangian is:
L = uk + vl − λ(k 1/4 l1/12 − Q).
20
The optimal k and l satisfy the three equations:

λ l1/12
Lk = u − =0
4 k 3/4
λ k 1/4
Ll = v − =0
12 l11/12
k 1/4 l1/12 = Q.
λ = 4uk 3/4 l−1/12 = 12vl11/12 k −1/4
so that k = 3(v/u)l. Then:
1/4
3v
k 1/4 l1/12 = l l1/12 = Q
u
and so: u 1/4
l1/3 = Q
3v
so: u 3/4
l= Q3 .
3v
Then: 1/4
3v 3v u 3/4 3 3v
k= l= Q = Q3 .
u u 3v u
(b) The profit is:
Π(Q) = wQ − C
where: 1/4
3v u 3/4
C = Cmin = u Q3 + v Q3 = 4(3−3/4 )u3/4 v 1/4 Q3 .
u 3v
So:
Π = wQ − 4(3−3/4 )u3/4 v 1/4 Q3 .
(c) Breakeven is when Π = 0. So, we need:

Q w − 4(3−3/4 )u3/4 v 1/4 Q2 = 0
so (rejecting Q = 0):
w 33/4
Q2 =
4 u3/4 v 1/4
and so: √ √ 1/8
w 33/8 w 27
Q= = .
2 u3/8 v 1/8 2 u3 v
(d) We have:
12 3/4 1/4 2
Π0 = w − u v Q .
33/4
This is 0 when:
33/4 w 1
Q2 = .
12 u v 1/4
3/4
So: √ √ 1/8
w 33/8 w 27
Q= = .
12 u3/8 v 1/8 12 u3 v
Also:
24 3/4 1/4
Π00 = − u v Q<0
33/4
so it does give a maximum.
21
Examiners’ commentary 2008

05A Mathematics 1
Specific comments on questions Zone B
Question 1
The demand equation for a good is q(2p + 3) = 8 and the supply equation is q − 2p + 4 = 0 where
p is the price and q is the quantity. Sketch the supply and demand functions for p ≥ 0.
The sketch of the two functions for p ≥ 0 is:
q − 2p + 4 = 0
8
3
1 q(2p + 3) = 8
O
2 5 p
2
−4
(It is quite acceptable to sketch p versus q instead.) A completely correct answer should
indicate:
• the shape of the demand curve (particularly, that it curves the right way and that it does
not intersect the p-axis);
• the shape of the supply curve (a straight line);
• the intercepts of both curves with the q-axis;
• the intercept of the supply curve with the p-axis.
(Partial credit would be given if some of the above had not been indicated.) You should also
remember, as indicated in the subject guide, that simply ‘plotting’ points and joining them
up is not an acceptable way to sketch curves.
1
05A Mathematics 1
For the equilibrium price, we solve the equation

8
= 2p − 4.
2p + 3
(2p + 3)(2p − 4) = 8 =⇒ 4p2 − 2p − 20 = 0,
and so, factorising, we have

5
(2p − 5)(p + 2) = 0 =⇒ p= , −2.
2
Thus, the equilibrium price is p = 5/2 (as the negative solution is not economically
meaningful). Of course, it is perfectly acceptable to use the formula for the solutions of a
quadratic equation instead of factorising.
The equilibrium quantity is then q = 2(5/2) − 4 = 1. (You should not round this up or down,
since nowhere in the question does it indicate that the quantities must be integers.)
Question 2
A firm has fixed costs of 20 and its marginal revenue and marginal cost functions are given,
respectively, by
M R = 12 − q, M C = q 2 − 3q + 4,
where q is the level of production. Find an expression, in terms of q, for the firm’s profit. Find
also the value of q which maximises the profit (and show that this value of q does indeed
maximise profit).
This question combines differentiation, integration, and optimisation of one-variable functions

and the relevant material can be found in Chapters 3 and 4 of the subject guide and in the
readings indicated there.
As the marginal revenue is the derivative of the total revenue, we integrate to find that
Z
1
T R = M R dq = 12q − q 2 + c.
2
Now, since there can be no revenue if no goods are sold we have T R(0) = 0. But, from above,
T R(0) = c and so we have c = 0.
Similarly, as the marginal cost is the derivative of total cost, we have

Z
1 3
T C = M C dq = q 3 − q 2 + 4q + c.
3 2
To find c, note that T C(0) = F C = 20, so c = 20.
The profit, Π, is then given by

1
Π = T R − T C = 8q + q 2 − q 3 − 20.
3
To maximise the profit, we find that
Π0 (q) = 8 + 2q − q 2 ,
and solve the equation Π0 (q) = 0, which is
8 + 2q − q 2 = 0 =⇒ q 2 − 2q − 8 = 0,
2
so that, factorising, we get

(q + 2)(q − 4) = 0 =⇒ q = −2, 4.
Thus, we take q = 4 (as the negative solution is not economically meaningful). Of course, it is
perfectly acceptable to use the formula for the solutions of a quadratic equation instead of
factorising. Now,
Π00 (q) = 2 − 2q
and, since Π00 (4) < 0, q = 4 does indeed maximise the profit.
You might think that an alternative method here is to solve the equation M R = M C in order
to maximise the profit. But this does not fully answer the question because, although it will
indeed lead to the value q = 4, it will not result in an explicit expression for the profit
function in terms of q, something the question clearly asks for. Nor, without extra work, will
it show that the value q = 4 does indeed maximise profit. If you took this approach, you
would still get some marks, but not all of them. So, this is a case where a careful reading of
the question is important.

For example, a number of students thought that T C = q(M C) + F C. This confuses marginal
cost with average cost. Recall that since the marginal cost is the derivative of the total cost,
we must integrate the marginal costs to find the total cost. Furthermore, you should explain
why the constant of integration is what it is, and not just assume that it equals the fixed cost.
In finding the total revenue, you have to realise that when the quantity is 0, the revenue is
also 0: only with this observation can you see that the constant of integration in this case is 0.
Question 3
A function f is given, for x > 0, by

A
f (x) = + B + Cx,
x
for some numbers A, B, C. If f (1) = 12, f (2) = 3 and f (3) = −4, show that
A + B + C = 12, A + 2B + 4C = 6, A + 3B + 9C = −12.
Express this system of equations in matrix form. Solve it, using a matrix method, to determine
A, B and C.
At first, this might look like an unusual sort of question, but it is in fact nothing more than a
system of linear equations. Now, the equations are given in the question, so even if you could
not do the first part of the question (showing that the equations hold), you could certainly
continue with the second part (solving the equations).
method, the row-reduction method, and so on. Many students like to use a different method,
not covered in the subject guide, especially ‘Cramer’s rule’. It is acceptable to do so, but our
view is that the row operations method is easier and less prone to error. But, whatever
method you use, it has to be a matrix method, as the question makes explicit: manipulation
of the equations will not suffice.
Substituting in the given information leads to the equations:

A A
A + B + C = 12, + B + 2C = 3 and + B + 3C = −4.
2 3
To see this, note, for example, that f (1) = A + B + C and we know that f (1) = 12, so the
first equation follows. The other two equations are similarly obtained. Then, multiplying the
second equation by 2 and the third by 3 gives the required equations.
3
05A Mathematics 1
We solve these equations using row operations which, as noted above, is the method
recommended for Mathematics 1. The augmented matrix is
 
1 1 1 12
 1 2 4 6 ,
1 3 9 −12

     
1 1 1 12 1 1 1 12 1 1 1 12
 1 2 4 6  →  0 1 3 −6  →  0 1 3 −6  .
1 3 9 −12 0 2 8 −24 0 0 2 −12
So, this gives us
A + B + C = 12, B + 3C = −6, and 2C = −12,
C = −6, B = −6 − 3C = 12 and A = 12 − B − C = 6.
This is probably a good point at which to make some general comments about how questions
are marked. Clearly, in a question like this, it is easy to get the wrong answer. (Though it
should be noted that in this particular question, you can always substitute the values of A, B
and C that you have found into the original equations, and this will show whether these are
re-work the calculation.) Examiners understand that arithmetical errors can be made,
themselves would make some mistakes if they sat the paper. So, although there are certainly
some marks for correct calculation, there are many marks for using the right method (even if
reduced matrix to determine the required values of A, B and C. There are marks for all these
things.
In particular, a number of students make the mistake of thinking that subtracting a fixed
Question 4
Show that the function

f (x, y) = x2 − xy + y 2 − 3x + 3
has one critical (or stationary) point and that the point is a local minimum.
fx = 2x − y − 3 and fy = −x + 2y.
To find the critical points, we solve the equations fx = 0 and fy = 0. The first of these
equations gives us x = 2y and substituting this into the second equation gives 3y − 3 = 0, so
y = 1. Using x = 2y again then gives us x = 2. Thus, the function has one critical point,
namely (2, 1).
4
fxx = 2, fxy = −1 and fyy = 2.
Since these values assure us that

2
fxx fyy − fxy = (2)(2) − (−1)2 = 3 > 0 and fxx > 0,
this point is indeed a local minimum. (Note that it is not correct to simply observe that
fxx > 0 and fyy > 0: you have to use the test correctly!)
Question 5
Determine the integral Z

√
x2 x − 2 dx.
always clear which technique will work. But you should of course realise that one of the
techniques will work! The three main techniques are: substitution, parts, and partial
fractions. The last of these applies only to very special types of integrals, and it turns out
that, for this problem, you can use either substitution or parts.
Substitution is probably the easiest way. If we let u = x − 2, then du = dx and so the integral
becomes
Z Z
√ √
2
x x − 2 dx = (u + 2)2 u du
Z
= (u2 + 4u + 4)u1/2 du
Z
= (u5/2 + 4u3/2 + 4u1/2 ) du
2 7/2 8 5/2 8 3/2
u + u + u +c
=
Z 7 5 3
√ 2 8 8
∴ x x − 2 dx = (x − 2) + (x − 2)5/2 + (x − 2)3/2 + c.
2 7/2
7 5 3
Although it is probably more difficult, the integral can also be done using integration by
parts, as follows.
Z Z
2
√ 2 2 3/2 4
x x − 2 dx = x (x − 2) − x(x − 2)3/2 dx
3 3
Z
2 8 8
= x2 (x − 2)3/2 − x(x − 2)5/2 + (x − 2)5/2 dx
3 15 15
Z
√ 2 8 16
∴ x2 x − 2 dx = x2 (x − 2)3/2 − x(x − 2)5/2 + (x − 2)7/2 + c.
3 15 105
This might look like a different answer to the one obtained using substitution, but it is, in
fact, the same. (A fact that you can verify by showing that these two answers only differ by a
constant.)
5
05A Mathematics 1
Question 6
Use
p the Lagrange multiplier method to find the values of x and y that minimise the function
x2 + 2y 2 subject to the constraint x + y = 2.
The Lagrangean is p
L= x2 + 2y 2 − λ(x + y − 2),
x
Lx = p − λ = 0,
x2 + 2y 2
2y
Ly = p − λ = 0.
x2 + 2y 2
x 2y
λ= p =p ,
x2 + 2y 2 x2 + 2y 2
so that x = 2y.
Then, substituting this into the constraint equation, x + y = 2, gives us 3y = 2 and so

y = 2/3. Then, using x = 2y again, we get x = 4/3.
Of course, you might realise that an equivalent problem is to minimise x2 + 2y 2 subject to the
constraint, and that is acceptable.
In questions like this, as the subject guide emphasises, you are not expected to verify that a
minimum has been found. Techniques for doing so are not part of the syllabus (and they are
rather more complicated than you might think).
It is possible to obtain the answer to this question without using Lagrange’s method, but that
does not answer the question since it explicitly asks that the Lagrange multiplier method be
used. So, if you did this, you would only get some of the marks. For example, you could use
the constraint equation to express y in terms of x as y = 2 − x. Then, we have to minimise
p p
f = x2 + 2(2 − x)2 = 3x2 − 8x + 8.
This gives
3x − 4
f0 = √ ,
3x2 − 8x + 8
and so, solving f 0 (x) = 0 gives x = 4/3 and, as y = 2 − x, we also therefore have y = 2/3.
To see that this gives a minimum, we note that f 0 changes sign from negative to positive at
4/3. (Or, although this is messier, we could have used the second derivative test to see this.)
That’s all correct, but look at the question: it says ‘Use the Lagrange multiplier method’ and
you need to do this in order to answer the question properly.
6
Question 7
Jim has opened an account with a bank, and they pay a fixed interest rate of 4% per annum,
with the interest paid once a year, at the end of the year. He opened the account with a payment
of $200 on 1 January 2008, and will be making additional deposits of $100 yearly, on the 1st of
January each year from 1 January 2009 onwards. What will the amount in the account be
immediately after he has made N of these additional deposits? (Your answer will be an
expression involving N , and you should simplify it as much as possible.)
subject guide gives the background and some examples. Essentially, what you have to do is:
problems using general techniques for solving difference equations, but that is not expected
(as such techniques are not part of the 05A syllabus, though they are in 05B).
Let yn be the required balance, i.e. the amount in the account immediately after n deposits
have been made. This means that
y0 = 200,
y1 = (1.04)y1 + 100 = 200(1.04) + 100,
y2 = (1.04)y2 + 100 = 200(1.04)2 + 100(1.04) + 100,
y3 = (1.04)y3 + 100 = 200(1.04)3 + 100(1.04)2 + 100(1.04) + 100,
and, in general,
yN = 200(1.04)N + 100(1.04)N −1 + · · · + 100(1.04) + 100.
(At this point, we have spotted the pattern!)
Noting that we have a geometric series, this can then be simplified to get
µ ¶
1 − (1.04)N
yN = 200(1.04)N + 100
1 − 1.04
= 200(1.04)N + 2, 500[(1.04)n − 1]
∴ yN = 2, 700(1.04)N − 2, 500.
7
05A Mathematics 1
Question 8
(a) The function f is defined, for x > 0, by

f (x) = x − x(ln x)2 .
Show that its critical (or stationary) points satisfy the equation
(ln x)2 + 2 ln x − 1 = 0.
By solving the quadratic equation w2 + 2w − 1 = 0, or otherwise, find the critical points.
Determine the nature of each critical point.
(b) A firm is the only producer of two goods, X and Y . The demand quantities x and y for X
and Y (respectively) and the corresponding prices pX and pY are related as follows:
−pX + pY + 13 − 6x = 0, pX − 4pY + 26 − 6y = 0.
The firm’s joint total cost function (that is, the cost of producing x of X and y of Y ) is
2x + y. Find an expression in terms of x and y for the profit function. Determine the
quantities x and y that maximise the profit.
(a) This is a one-variable optimisation problem, which uses the techniques in Chapter 3 of the
subject guide. But it’s a little unusual, isn’t it? The examiners certainly thought so, which
is why they asked it. But — and you must always remember this — you can do it using the
methods of this unit.
We differentiate to find that
2 ln x
f 0 (x) = 1 − (ln x)2 − x = −(ln x)2 − 2 ln x + 1,
x
and solve the equation f 0 (x) = 0 to find the critical points. This means that, to find these
points, we need to solve the equation
(ln x)2 + 2 ln x − 1 = 0,
as required.
Using the formula for the solutions of a quadratic equation, we see that
w2 + 2w − 1 = 0
has solutions √
w = −1 ± 2.
So, letting w = ln x, we see that the critical points are given by
√ √ √
ln x = −1 ± 2 =⇒ x = e−1+ 2 , e−1− 2 .
Let’s call these r, s respectively.
To determine the nature of these critical points, we differentiate again to see that
2 ln x 2 2
f 00 (x) = − − = − (ln x + 1) .
x x x
This means that, as r > 0,
2 √
f 00 (r) = − ( 2) < 0,
r
i.e. r is a local maximum and, as s > 0,
2 √
f 00 (s) = − (− 2) > 0,
s
i.e. s is a local minimum.
The natures of r and s can also be found by considering the behaviour of the sign of f 0 (x)
around the two points remembering, of course, that f 0 (x) = −(ln x)2 − 2 ln x + 1 and not
(ln x)2 + 2 ln x − 1.
8
(b) This is a very standard application of unconstrained optimisation: see Chapter 5 of the
subject guide and the related readings.
and we want this to be a function of x and y (because the question says so, and because we
need to reduce the problem to a two variable optimisation problem). This means that we
need to find pX and pY in terms of x and y.
One way of doing this is to add the two equations to get
−3pY + 39 − 6x − 6y = 0,
which gives us
pY = 13 − 2x − 2y.
Then, substituting this into the first equation we find that
pX = pY + 13 − 6x = 26 − 8x − 2y.
Putting these into our expression for the profit and simplifying, we find that the profit is
Π = 24x + 12y − 8x2 − 2y 2 − 4xy,
as a function of x and y.
To find the values of x and y that maximise the profit, we solve the equations Πx = 0 and
Πy = 0, i.e.
Πx = 24 − 16x − 4y = 0 and Πy = 12 − 4y − 4x = 0.
This is easily done and we find that x = 1 and y = 2.
Πxx = −16 Πxy = −4 and Πyy = −4.
Since these assure us that
Πxx Πyy − Π2xy = (−16)(−4) − (−4)2 = 48 > 0 and Πxx < 0,
these values of x and y do indeed maximise the profit. (Note that it is not correct to simply
observe that Πxx < 0 and Πyy < 0: you have to use the test correctly!)
An alternative approach would be to express the profit in terms of pX and pY . This would
be messy in this case and, furthermore, it is not what the question asks for and so it would
not get full credit. But, if you want the gory details, here they are.
We have
Π = xpX + ypY − (2x + y),
and we will write the quantities in terms of the prices. Using the two equations we get
1 1
x= (−pX + pY + 13) and y= (pX − 4pY + 26),
6 6
which means that, putting these into our expression for the profit and simplifying we get
1¡ 2 ¢
Π= −pX + 2pX pY + 14pX − 4p2Y + 28pY − 52 .
6
We now maximise f = 6Π (an equivalent problem to that of maximising Π and better
because we don’t have to deal with awkward fractions) by solving the equations fpX = 0 and
fpY = 0, i.e.
fpX = −2pX + 2pY + 14 = 0 and fpY = 2pX − 8pY + 28 = 0.
This is easily done and we find that pY = 7 and pX = 14 which means, in turn, that x = 1
and y = 2 as before.
The second-order test can then be applied to verify that we have indeed found a maximum.
9
05A Mathematics 1
Now, although this approach gives the right answer, it does not answer the question, which
said, quite explicitly, ‘Find an expression in terms of x and y for the profit function’. So if
used, this method will only get some of the marks!
Some students seem to think that there is a way in which they can deal separately with the
x-profit and the y-profit. This is clumsy and, again, does not answer the question. Indeed,
although this method works in some examples, it won’t work here. What we mean by this is
that they try to find the maximising values of x and y by solving equations like
Marginal x-revenue = Marginal x-cost,
and so, as pX = 26 − 8x − 2y, we’d need to solve the equation we’d get from
∂ ∂
(x(26 − 8x − 2y)) = (2x + y),
∂x ∂x
and so on. This just does not work. It gives the wrong answer and it is mathematically
incorrect.
Question 9
(a) Find the critical (or stationary) points of the function f (x, y) = x3 − 4xy + 2y 2 + 3 and
determine, for each, whether it is a local maximum, a local minimum, or a saddle point.
(b) Determine the integral Z
2
x3 ex dx
(c) By simplifying the expression on the right-hand side of the equation, or otherwise, show that
x−1 1 1 2
= − + .
x(x + 1)2 x + 1 x (x + 1)2
Hence find Z
x−1
dx.
x(x + 1)2
(a) This uses the same techniques as Question 4, described in Chapter 5 of the subject guide. It
is, however, a little more complicated.
fx = 3x2 − 4y and fy = −4x + 4y.
To find the critical points, we solve the equations fx = 0 and fy = 0. The second of these
gives x = y and substituting this into the first equation gives
4
3y 2 − 4y = 0 =⇒ y(3y − 4) = 0 =⇒ y = 0, .
3
Using x = 2y again, these give us x = 0 and x = 4/3 respectively and so the critical points
are (0, 0) and (4/3, 4/3).
fxx = 6x, fxy = −4 and fyy = 4.
At the critical point (0, 0), this means that we have
2
fxx fyy − fxy = (0)(4) − (−4)2 = −16 < 0,
and so this is a saddle point. Whereas, at the critical point (4/3, 4/3), we have
2
fxx fyy − fxy = (8)(4) − (−4)2 = 16 > 0 and fxx = 8 > 0,
and so this is a local minimum.
Note that it is important to substitute the coordinates of each critical point, in turn, into
the second derivative expressions before applying the second derivative test.
10
(b) This question involves the substitution and integration by parts techniques discussed in
It is sensible to start with a substitution in order to simplify the integral. The obvious
substitution (since it appears as the exponent in the exponential part of the integrand) is
u = x2 . This means that du = 2x dx and so
Z Z
2 1
x3 ex dx = ueu du.
2
Now using integration by parts, we find that
Z Z
ueu du = ueu − eu du = ueu − eu + c,
and, putting this all together, we get

Z
2 1 2 1 2
x3 ex dx = x2 ex − ex + c.
2 2
(c) Again, this is an integration problem: see Chapter 4 of the subject guide. It really helps to
follow the lead given in the question, i.e. start with the right-hand side and simplify it to
obtain the left-hand side, not the other way round.
The obvious way to simplify the expression on the right-hand side is to take a common
denominator. We have
1 1 2 x(x + 1) − (x + 1)2 + 2x
− + =
x + 1 x (x + 1)2 x(x + 1)2
(x + x) − (x2 + 2x + 1) + 2x
2
=
x(x + 1)2
x−1
= .
x(x + 1)2
So, using this, we have
Z Z
x−1 1 1 2 2
dx = − + dx = ln |x + 1| − ln |x| − + c.
x(x + 1)2 x + 1 x (x + 1)2 x+1
Question 10
(a) Using a matrix method, solve the following system of equations to find x, y and z.
40 − x + y + 2z = x + 2y + 7z − 20
20 + x − y + z = 3x + 2y − 120
250 + x + y − z = x + 2y − 30.
(b) The function f (x, y) is given by
f (x, y) = sin(nx) cos(y 2 ),
where n is a positive integer. Find
∂f ∂f ∂ 2 f ∂ 2 f
, , , .
∂x ∂y ∂x2 ∂y 2
Find the value of n if f satisfies
∂2f ∂2f ∂f
y3 2
− y 2
+ = 0.
∂x ∂y ∂y
11
05A Mathematics 1
(a) This involves the same techniques as Question 3, described in Chapter 6 of the subject
guide. But, it is important to realise that the given equations are not in the right form for
the standard method of row operations (or other methods, such as Cramer’s rule) to be
applied. Explicitly, the standard methods for solving linear systems assume that the
equations take the form Ax = b where A is a matrix, x is the vector of unknowns (i.e.
variables) and b is a vector whose entries are constants (that is, numbers). The system of
equations in the question has to be rewritten in this standard form before solving.
Having done this, the equations are
2x + y + 5z = 60
2x + 3y − z = 140
y + z = 280
and we can now solve them using row operations. The augmented matrix is
 
2 1 5 60
 2 3 −1 140 
0 1 1 280

     
2 1 5 60 2 1 5 60 2 1 5 60
 2 3 −1 140  →  0 2 −6 80  →  0 1 −3 40  .
0 1 1 280 0 1 1 280 0 0 4 240
So, this gives us
2x + y + 5z = 60, y − 3z = 40 and 4z = 240,

1
z = 60, y = 40 + 3z = 220 and x= (60 − y − 5z) = −230.
2
(b) This uses nothing more than the techniques for partial differentiation discussed in Chapter 5
of the subject guide and, although it may look daunting, it is, in fact, quite straightforward.
The first derivatives are given by
fx = n cos(nx) cos(y 2 ) and fy = 2y sin(nx) sin(y 2 ).
This means that fxx is given by
fxx = −n2 sin(nx) cos(y 2 ),
and fyy is
fyy = −2 sin(nx) sin(y 2 ) − 4y 2 sin(nx) cos(y 2 ),
where, of course, we have had to use the product rule here.
To find the value of n that makes f satisfy the equation
y 3 fxx − yfyy + fy = 0,
we note that, substituting for the derivatives we have found on the left-hand side, we get the
equation
−n2 y 3 sin(nx) cos(y 2 ) + 2y sin(nx) sin(y 2 ) + 4y 3 sin(nx) cos(y 2 ) − 2y sin(nx) sin(y 2 ) = 0,
and this simplifies to

(4 − n2 )y 3 sin(nx) cos(y 2 ) = 0.
Now, as far as n is concerned, this equation can only be satisfied if 4 − n2 = 0 or n = 0 (as
this makes sin(nx) = 0), i.e. if n = 0, ±2. But, we are also told that n is positive, and so
n = 2 is the value that makes f satisfy this equation.
12
Question 11
(a) Suppose that 0 < a < 1 and that x = x∗ and y = y ∗ maximise the function
(xa + y a )1/a
subject to the constraint px + qy = M . By using the Lagrange multiplier method, show that
µ ¶1/(a−1)
∗ q
y = x∗ .
p
Suppose now that a = 1/2. Show that

Mq Mp
x∗ = , y∗ = .
p(p + q) q(p + q)
(b) An arithmetic progression has the following properties:

• its second term is 7;
• its seventh term is nine times its first term.
Determine the first term and the common difference.
(a) This uses the Lagrange multiplier method which we saw in Question 6, described in Chapter
5 of the subject guide. The question might look a bit frightening, but it contains a lot of
useful information, even providing you with the relationship between the optimal values of x
and y so that you can see whether you are on the right track.
The Lagrangean is
1
L = (xa + y a ) a − λ(px + qy − M ),
Lx = xa−1 (xa + y a )(1/a)−1 − λp = 0,
Ly = y a−1 (xa + y a )(1/a)−1 − λq = 0.
xa−1 (xa + y a )(1/a)−1 y a−1 (xa + y a )(1/a)−1

λ= = ,
p q
so that ³ y á−1
y a−1 q q
a−1
= =⇒ = .
x p x p
Raising both sides to the power 1/(a − 1) we then get
µ ¶1/(a−1) µ ¶1/(a−1)
y q q
= =⇒ y= x,
x p p
and so the solutions to this constrained optimisation problem, i.e. x∗ and y ∗ , are related in
the required way.
When a = 1/2 we have 1/(a − 1) = −2 and so this expression becomes
µ ¶−2
q p2
y= x = 2 x.
p q
Then, substituting this into the constraint equation, px + qy = M , gives us

µ 2 ¶
p
px + q x =M =⇒ p(q + p)x = M q,
q2
13
05A Mathematics 1
and so x∗ is
Mq
x= ,
p(q + p)
as required. Then using our expression again we get
µ 2¶
p Mq Mp
y= 2
= ,
q p(q + p) q(q + p)
as required, for y ∗ .
(b) For this question, you need to know about arithmetic progressions and these are discussed in
The nth term of an arithmetic progression takes the form a + (n − 1)d and so the fact that
the second term is 7 means that we have
a + d = 7,
whereas the fact that the seventh term is nine times the first term means that we have
a + 6d = 9a.
Thus, to determine the first term, a, and the common difference, d, we need to solve the
simultaneous equations
a+d=7 and − 8a + 6d = 0.
This is easily done by noting that a = 3d/4 from the second equation and this means that
the first equation gives
3d 7
+d=7 =⇒ d=7 =⇒ d = 4.
4 4
Then, using a = 3d/4 again, we get a = 3.
14
05a Mathematics 1
Examiners’ Report 2006
Zone A
General remarks
means two things. First, they should not simply write down the answer in the exam-
ination script, but explain the method by which it is obtained. Secondly, they should
include rough working. The examiners want you to get the right answers, of course,
but it is more important that you prove you know what you are doing: that is what is
really being examined.
We also stress that if a student has not completely solved a problem, they may still be
awarded marks for a partial, incomplete, or slightly wrong, solution; but, if they have
written down a wrong answer and nothing else, no marks can be awarded.
Covering the syllabus
in Section A.
Every examination paper is different. You should not assume that your examination
will be almost identical to the previous year’s: for instance, just because there was a
question, or a part of a question, on a certain topic last year, you should not assume
there will be one on the same topic this year. Each year, the examiners want to test
that candidates know and understand a number of mathematical methods and, in
setting an examination paper, they try to test whether the student does indeed know
the methods, understands them, and is able to use them, and not merely whether they
vaguely remember them. Because of this, every year there are some questions which are
likely to seem unfamiliar, or different from previous years’ questions. You should expect
to be surprised by some of the questions. Of course, candidates will only be examined
on material in the syllabus, so all questions can be answered using the material of the
subject. There will be enough, routine, familiar content in the examination so that
a student who has achieved competence in the subject will pass, but, of course, for a
1
high mark, more is expected: students will have to demonstrate an ability to solve new
and unfamiliar problems.
Specific comments on questions
1. This was a straightforward and standard question, very well done in general. In
sketching the demand curve, p = q 2 + 4q + 20, it is useful to obtain some additional
information. We know that it is a parabola, because the equation is quadratic. We
can see that q 2 + 4q + 20 = (q + 2)2 + 16, which shows that it always lies above the
q-axis and that its minimum value is at q = −2. (The same conclusions can be drawn
by noting that the equation has no roots and by finding that the derivative is 0 at −2.)
The graph should also indicate the value if p at which the graph crosses the p-axis
(which is p = 20). The graph is strange (but not impossible) for a demand curve, in
that it slopes upwards. The graph is shown below.
For the equilibrium, we need to solve q 2 + 4q + 20 = −q 2 − 10q + 176. This can be

written as a quadratic equation in standard form as 2q 2 + 14q − 156 = 0. This can
be solved by factorising or using the formula for the solutions of a quadratic. The
solutions are q = −13 and q = 6, so the solution we seek (the economically meaningful
one) is q = 6. The corresponding value of p is p = 80.
The curve should not be ‘sketched’ by plotting points: that is ’plotting’ not ’sketching’.
40
30
20
10
0 1 2 3 4
q
2. This question was quite straightforward and well done on the whole, though a few
candidates misunderstood the information given and, as a result, produced the wrong
2
√
profit function. From the information given, we can find that p = 4/ q, so the total
√ √ √
revenue is T R = (4/ q)q = 4 q and the profit is Π = T R − T C = 4 q − q 2 , noting
that the total cost function is q 2 . We then solve Π′ (q) = 0, which is 2√4 q −2q = 0, leading
√
to q q = 1, and hence q = 1. Corresponding to this, we have p = 4. Furthermore, we
should check that this does indeed maximise the profit. We have Π′′ (q) = −1/q 3/2 − 2
so Π′′ (1) < 0 and hence it’s a maximum.
3. This was a straightforward and standard question. The augmented matrix is

 
1 −2 10 5
 2 1 −2 4  .
1 3 4 7
Using row operations to reduce, we have

     
1 −2 10 5 1 −2 10 5 1 −2 10 5
 2 1 −2 4  →  0 5 −22 −6  →  0 5 −22 −6  ,
1 3 4 7 0 5 −6 2 0 0 16 8
so we have z = 1/2, 5y = −6 + 11 = 5, so y = 1 and x = 5 − 5z + 2y = 2.
Some candidates used invalid row operations (eg, multiplying two rows together or
subtracting a fixed number from each entry of a row, a surprisingly common error).
Be sure that you understand what row operations are allowable.
Some used Cramer’s rule. That is perfectly acceptable, but possibly more prone to
arithmetical error.
4. This is an easy integral to do√once the right substitution is chosen, and that is the key
to solving it. If we let u = 1 + e−x , then du = − 21 e−x/2 dx and the integral reduces to
R√ √
the very easy I = −2 u du. This√gives = − 43 u3/2 +c, which is = − 34 (1+ e−x )3/2 +c.
(Other substitutions, such as u = x will also work.)
5. This was badly done. This is an unusual question, but every examination contains
unusual questions, and this is a deliberate strategy on the examiners’ part to ensure
the standard of the examination. (See the discussion in the final section of this report.)
However, this is not fundamentally a hard question. All it needs is the chain rule and
the product rule.
Let g = −ax − by. Then, by the product and chain rules,

Vx = Ux eg + gx U eg = Ux eg − aU eg .
(Here, we use the Vx notation for ∂V /∂x, and so on.) Also,
Vy = Uy eg + gy U eg = Uy eg − bU eg .
Differentiating Vx again with respect to x, and using the product and chain rules, we
have
Vxx = Uxx eg − aUx eg − aUx eg + a2 U eg = Uxx eg − 2aUx eg + a2 U eg .
3
With a = 1, b = 4, we have (with g = −x − 4y), Vx = Ux eg − U eg Vy = Uy eg − 4U eg ,
Vx = Ux eg − 12 U eg , and Vx x = Uxx eg − 2Ux eg + U eg .
We have Vy = Vxx + 2Vx − 3V , so
Uy eg − 4U eg = (Uxx − 2Ux + U + 2Ux − 2U − 3U ) eg
and so, simplifying, Uy = Uxx , as required.
6. Two-variable optimisation problems are very standard. Finding the critical point in
this particular problem required solving some tricky equations, but the usual technique
of using the fact that the partial derivatives are 0 to substitute for one variable in terms
of the other leads to the answer.

2y
fx = 1 +
x2
and
4 2
fy = − 2
− .
y x
We solve fx = fy = 0. Now, fx = 0 means y = −(1/2)x2 . Then, fy = 0 implies
2/(−x2 /2)2 = −1/x, so x3 = −8 and x = −2. Then, y = −x2 /2 = −2.
fxx = −4y/x3 , fyy = 8/y 3 , fxy = 2/x2 .

2
At (−2, −2) we have fxx = −1, fyy = −1, fxy = 1/2, so fxx < 0 and fxx fyy − fxy > 0,
and hence this is a local maximum point.
7. Constrained optimisation problems such as this are standard and routine, and the
Lagrange method is the obvious way to tackle them. (Note, though, that the question
did not require the use of Lagrange’s method, so a method based on substituting for x
or y in terms of the other variable using the constraint was acceptable.) This problem
is similar to, but substantially easier than, Question 10 (a) of the 2005 ZB and ZA
examination papers.
The Lagrangian is L = (x + 1)3 (y + 1)2 − λ(x + y − 13). (Or, if you like, equally
acceptably, L = (x + 1)3 (y + 1)2 + λ(x + y − 13). We need to solve
Lx = 3(x + 1)2 (y + 1)2 − λ = 0
and
Ly = 2(x + 1)3 (y + 1) − λ = 0
simultaneously, alongside the constraint equation x + y = 13. These first two equations
imply that
3(x + 1)2 (y + 1)2 = 2(x + 1)3 (y + 1),
4
so 3(y + 1) = 2(x + 1) and y = 23 x − 31 (or, alternatively, we could express x in terms
of y). The constraint is x + y = 13, which means
5 1
x − = 13
3 3
and hence x = 8 and y = 5.
8(a) This is a completely standard type of problem. A very similar problem appeared
in the sample examination paper in the old edition of the subject guide (which students
taking the paper in 2006 will have used) and in previous examinations (Question 10(a),
2004 ZB and ZA papers). Those who attempted this did well.
The equations are equivalent to
2x1 + 5x2 + 7x3 = 294

2x1 + x2 = 70
12x1 − 2x2 + 4x3 = 44.

2 5 7 294
 2 1 0 70 
12 −2 4 44
Using row operations to reduce, we have

   
2 5 7 294 2 1 0 70
 2 1 0 70  →  2 5 7 294 
12 −2 4 44 6 −1 2 22
   
2 1 0 70 2 1 0 70
→  0 4 7 224  →  0 4 7 224  ,
0 −4 2 −188 0 0 9 36
so equilibrium prices are x3 = 4, x2 = 49 and x1 = 21/2.
8(b) Some candidates had difficulty with this first integral because they did not know
how to work with trigonometrical functions. But such functions are an important part
of this subject and are discussed in the subject guide. The easiest way is to make
1
the substitution u = tan x. Then du = dx. (This can be calculated using the
cos2 x
quotient rule, the fact that tan x = sin x/ cos x and the fact that sin2 x + cos2 x = 1.)
So the integral is
1
du
Z
√ 1
√ = 2 u 0 = 2.
0 u
(Note that we have changed the limits: if you don’t do this, you will need to revert back
to x once you have integrated: that’s perfectly acceptable.) Note that we have used the
fact that tan π/4 = 1. This is stated explicitly in a table in the subject guide. Students
5
should be comfortable with radians (as explained in the guide) and should know the
values of the trigonometrical functions at the key values of 0, π/6, π/4, π/3, π/2: see
the table in the subject guide.
For the second integral, the easiest (and most obvious) approach is integration by parts:
we have
1 1 1
Z Z
2 3
(x + 2) ln x dx = (x + 2) ln x − (x + 2)3 dx
3 3 x
1 1 1
Z
= (x + 2)3 ln x − (x3 + 6x2 + 12x + 8) dx
3 3 x
Z
1 1 8
= (x + 1)3 ln x − x2 + 6x + 12 + dx
3 3 x
1 1 8
= (x + 2)3 ln x − x3 − x2 − 4x − ln x + c.
3 9 3
9(a) This is a straightforward question. (It is similar to, but easier than, sample
question 5 of Chapter 4 in the subject guide: see the solution to that problem and
the discussion given there.) All that is required is a recognition that marginal cost is
the derivative of total cost, and an ability to use partial fractions. What we need is
T C(2) − T C(1), where T C is the total cost function. This is
Z 2
q+1
2
dq.
1 q + 6q + 8
Now, using partial fractions,

q+1 q+1 1/2 3/2
= =− + .
q2 + 6q + 8 (q + 2)(q + 4) q+2 q+4
So
T C(2) − T C(1) = [−(1/2) ln(q + 2) + (3/2) ln(q + 4)]21 = (1/2) ln(3/4) + (3/2) ln(6/5).
It is incorrect to simply calculate M C(1) and say that the extra cost is equal to this
value. Remember that the marginal cost is the derivative of the total cost and it
gives an approximation to the increased cost of producing one more unit when many
are being produced. But here the increase in production, from 1 to 2, is so large as
to make this approximation invalid. We need, to answer the question correctly, to
integrate to determine the total cost function.
√
9(b) We have P (t) = V e−0.1t = 10e2 t e−0.1t , so

1
P (t) = 10 √ − 0.1 e−0.1t .
′
t
√
Therefore, P ′ (t) = 0 ⇐⇒ t = 10 ⇐⇒ t = 100. At t = 100, P ′ changes sign from
positive to negative, so it is a maximum. (Alternatively, we could calculate P ′′ (t) and
note that P ′′ (100) < 0.)
6
9(c) This problem is hard if you do not follow the very explicit hint given in the
question. But if you do what it suggests, it is much easier. You should always follow
a hint if one is given. We have
fx = 4x(4xy + 5) + (2x2 + 2y 2 − 3)4y
and
fy = 4y(4xy + 5) + (2x2 + 2y 2 − 3)4x.
Now, fx = fy = 0 =⇒ fx + fy = 0, so, following the hint, and adding these two
equations,
fx + fy = (4x + 4y)(4xy + 5) + (4x + 4y)(2x2 + 2y 2 − 6) = 0,
so
4(x + y)(2x2 + 4xy + 2y 2 + 2) = 0.
This simplifies to
4(x + y)(2(x + y)2 + 2) = 0
so we must have y = −x. The equation fx = 0 then becomes 4x(−4x2 + 5) − 4x(4x2 −
3) = 00, which is 32x − 32x3 = 0, so x(1 − x2 ) = 0 and x = 0, 1 or −1 and y = 0, −1, 1,
correspondingly. So the critical points are (0, 0), (1, −1) and (−1, 1).
10(a). Problems such as this, although regularly featuring in these examinations and
not hard, seem to be problematic for students. This year, there was no question on
sequences and series in Section A, but no inference should be drawn from that fact for
future years. (See the comments on future examinations.) To answer this question,
we have to be clear about what we want to find, and how we will find it. Let yt be
balance in account after t full years. (This is the key thing to monitor.) So y0 = L.
Then y1 = (1.05)L − D,
y2 = (1.05)y1 − D = (1.05)2 L − (1.05)D − D,
y3 = (1.05)y2 − D = (1.05)3 L − (1.05)2 L − (1.05)L − L,

and, in general
yt = (1.05)t L − (1.05)t−1 D − (1.05)t−2 D − · · · − D.
Now, we must have yN ≥ 0 in order to make the required number of withdrawals. So
(1.05)N L − (1.05)N −1 D − (1.05)N −2 D − · · · − D ≥ 0.
1 − (1.05)N
(1.05)N L − D ≥ 0.
1 − 1.05
So
1
L ≥ 20D 1 − .
(1.05)N
7
10(b) This is a very standard utility-maximisation problem, with a budget constraint.
These occur frequently in these examination papers (e.g., Question 11 of the 2003 ZA
paper, Question 5 of the 2003 ZB paper, Question 5 of the 2004 ZA paper, Question
5 of the 2004 ZB paper, Question 10(a) of the 2005 ZA paper, and Question 10(a) of
the 2005 ZB paper) though, of course, the precise utility function is different from ones
previously used.
The problem, therefore, is the familiar one: maximise (xβ + 3y β )1/β subject to x +
y = M . As usual, the Lagrange multiplier approach is best. The Lagrangian is
L = (xβ + 3y β )1/β − λ(x + y − M ). We solve
Lx = xβ−1 (xβ + 3y β )(1/β−1) − λ = 0
Ly = 3y β−1 (xβ + 3y β )(1/β−1) − λ = 0

together with the constraint x + y = M . We solve Lx = Ly = 0. From the first two
equations, we have xβ−1 = 3y β−1 , from which we have x = 31/(β−1) y. The constraint,
x + y = M , becomes y(1 + 31/(β−1) ) = M so y = M/(1 + 31/(β−1) ) and
31/(β−1) M
x = 31/(β−1) y = .
1 + 31/(β−1)
The maximum u is (xβ + 3y β )1/β with these values of x and y. This is
M
(3β/(β−1) + 3)1/β .
1 + 31/(β−1)
11(a) This was an exercise in partial differentiation and the manipulation of expres-
sions, a standard type of problem (see Question 9(b), 2003 ZB examination, and Ques-
tion 9(b) in the mock examination in the old edition of the subject guide). Care is
required, so these problems should not be rushed. We have
!
1 4x
fx = q 2+ p ,
4x2 + y 2
p
2 2x + y + 4x2 + y 2
!
1 y
fy = q 1+ p .
4x2 + y 2
p
2 2x + y + 4x2 + y 2
So
!
1 4x2 y2
xfx + yfy = q 2x + p +y+ p
4x2 + y 2 4x2 + y 2
p
2 2x + y + 4x2 + y 2
!
2 2
1 4x + y
= q 2x + y + p
4x2 + y 2
p
2 2x + y + 4x2 + y 2
1 p
2 2

= q p 2x + y + 4x + y
2 2x + y + 4x2 + y 2
8
1
q p
= 2x + y + 4x2 + y 2
2
1
= f.
2
11(b) Two-commodity maximisation problems are standard problems (2003 ZB, Ques-
tion 8(b); 2003 ZA, Question 8(b); 2004 ZB, Question 4; 2005 ZB, Question 10(b); 2005
ZA, Question 10(b); and Chapter 5 of the subject guide). What is different here is the
presence of the constant c in the expression for the total cost function. But, in all other
respects, this is a completely standard question and the standard methods apply.
The profit function is
Π = xpX +ypY −T C = x(4−2x)+y(2−2y)−(x2 +cxy +y 2 ) = 4x+2y −3x2 −3y 2 −cxy.
We solve Πx = Πy = 0, which is 4 − 6x − cy = 0, 2 − 6y − cx = 0. These are equivalent

to 36x + 6cy = 24, c2 x + 6cy = 2c. Subtracting the second of these equations from
the first eliminates y and we find that
24 − 2c
x= .
36 − c2
(There are other ways of solving the equations.) Because 0 < c < 3, this value of x is
positive. Then,
1 12 − 4c
y = (4 − 6x) = .
c 36 − c2
Again, note that, because 0 < c < 3, this value of y is positive. We have
Πxx = −6, Πyy = −6, Πxy = −c,
so Πxx < 0 and Πxx Πyy − Π2xy = 36 − c2 > 0, since 0 < c < 3, so it is a maximum.
There will be no change to the format, style or number of questions in the examination
paper for 2007.
9

05a Mathematics 1
Specific comments on questions – Zone B
Section A
Answer all seven questions from this section (60 marks in total).
Question 1
A firm is the only producer of a particular good, and the demand equation for the
good is 2p + q = 20, where p denotes the selling price and q is the quantity
produced by the firm. The firm’s fixed costs are 12 and its marginal cost function is
M C = 2 + q. Find an expression for the profit function, Π(q). Sketch the graph of
Π for q between 0 and 8. Determine the value of the production, q, which
maximises the firm’s profit.
This question concerns curve sketching and optimisation, discussed in Chapters 2 and 3 of the
subject guide. From the demand equation,
q
p = 10 − .
2
So the total revenue is
1
T R = pq = 10q − q 2 .
2
To obtain the total cost from the marginal cost, we must integrate. We have
Z Z
1
T C = M C dq = (2 + q) dq = 2q + q 2 + c.
2
To find c, note that T C(0) = F C = 12, so c = 12. (You should explain, like this, why the
constant of integration is 12.)
So, the profit, as a function of q, is
Π = T R − T C = −q 2 + 8q − 12.
We now solve Π0 = 0 to maximise the profit. We have
8 − 2q = 0, so q = 4
and
Π00 (q) = −2 < 0
so q = 4 maximises the profit. (It is also acceptable, rather than checking the second derivative,
to comment that the critical point must be a maximum because the function is a negative
quadratic.)
1
05a Mathematics 1

cost with average cost. Recall that, since the marginal cost is the derivative of the total cost, we
must integrate the marginal costs to find the total cost.)
Now,
Π(q) = 0 ⇐⇒ q 2 − 8q + 12 = 0 ⇐⇒ (q − 6)(q − 2) = 0 ⇐⇒ q = 2, 6.
The sketch is:
2
q
1 2 3 4 5 6 7 8
0
–2
–4
–6
–8
–10
–12
You should show the correct shape of the curve, and you should (as above) calculate the
intercepts with the q-axis.
Question 2
Functions f and g are as follows:
f (x) = x4 + x3 + 2x2 + 2, g(x) = −x4 + x3 + 8x2 + 10.
Show that the curves y = f (x) and y = g(x) intersect for exactly two values of x.
Find these values of x. (Do not attempt to sketch the curves.)
The intersections are the solutions of f (x) = g(x). Now,
f (x) = g(x) ⇐⇒ x4 + x3 + 2x2 + 2 = −x4 + x3 + 8x2 + 10 ⇐⇒ x4 − 3x2 − 4 = 0.
We don’t have a general method for solving polynomial equations of degree 4. However, this
equation can be viewed is a quadratic in x2 , for it is (x2 )2 − 3(x2 ) − 4 = 0.
Now we can use the formula for the solutions of a quadratic, or we can use factorisation to find
x2 :
(x2 − 4)(x2 + 1) = 0,
so x2 = 4. (Note that we cannot have x2 = −1.) So x = 2, −2 are the two required values.
2
Question 3
matrix method.
2x + 3y − z = 4
2x − 3y + z = 0
x + 2y + 3z = 6.
suffice.
In matrix form, this is     

2 3 −1 x 4
 2 −3 1  y  =  0 .
1 2 3 z 6

2 3 −1 4
 2 −3 1 0 .
1 2 3 6
Using row operations to reduce (and there are many ways this can be done), we have
     
2 3 −1 4 1 2 3 6 1 2 3 6
 2 −3 1 0  →  2 −3 1 0  →  0 −7 −5 −12 
1 2 3 6 2 3 −1 4 0 −1 −7 −8
   
1 2 3 6 1 2 3 6
→ 0 1 7 8 → 0 1 7 8 .
0 7 5 12 0 0 −44 −44
So, using back-substitution, we have
z = 1,
y = 8 − 7z = 1,
x = 6 − 3z − 2y = 1.
3
05a Mathematics 1
Question 4
Determine the integral

dx
Z
.
ex − e−x
partial fractions. It turns out that, for this problem, substitution and then partial fractions is the
best approach.
One clever approach is to note that the integral is

ex dx
Z
(ex )2 − 1
and then use the substitution u = ex to reduce to a partial fractions integral. But if you didn’t
notice this, you might sensibly attempt a substitution u = ex or u = e−x . Both work in very
much the same way. Here’s how the substitution u = ex works:
Let u = ex . Then du = ex dx = u dx, so

Z Z
du du
I= −1
= 2
.
u(u − u ) u −1
Now we can use partial fractions, which tells us that for some numbers A and B,
du 1 A B
= = + .
u2 − 1 (u − 1)(u + 1) u−1 u+1
There are various ways of finding A, B. We determine that A = 1/2, B = −1/2. Then,
1 1 1 1
I= ln |u − 1| − ln |u + 1| + c = ln |ex − 1| − ln |ex + 1| + c.
2 2 2 2
Question 5
A firm is the only producer of two goods, X and Y . The demand equations for X
and Y are given by
x = 120 − pX ,
y = 200 − 2pY ,
where x and y are the quantities of X and Y demanded (respectively) and pX , pY

are (respectively) the prices of X and Y . The firm’s joint total cost function (that
is, the cost of producing x of X and y of Y ) is
1
x2 + y 2 + 2xy.
2
Find an expression in terms of x and y for the profit function. Determine the
quantities x and y that maximise the profit, and find the corresponding prices
pX , pY .
4

and we want to get this as a function of x and y, so we need to find pX and pY in terms of x and
y. We have
pX = 120 − x, pY = 100 − y/2.
So,
= 120x + 100y − 2x2 − y 2 − 2xy.
Πx = 120 − 4x − 2y, Πy = 100 − 2y − 2x.
We now solve Πx = 0, Πy = 0, obtaining x = 10 and y = 40. We have Πxx = −4 < 0 and

Πxx Πyy − Π2xy = (−4)(−2) − (−2)2 > 0, so this does indeed maximise the profit. The
corresponding prices are
pX = 120 − x = 110, pY = 100 − y/2 = 90.
Question 6
The weekly output (in units) of a factory depends on the amount of capital and
labour it employs as √follows:
√ if it uses k units of capital and l of labour then its
output is Q(k, l) = k l units. The cost to the firm of each unit of capital is 4
dollars, and the cost of each unit of labour is 1 dollar. Use the method of Lagrange
multipliers to find the minimum weekly cost of producing a quantity of 200 units.
The Lagrange multiplier method is discussed in Chapter 5 of the subject guide.

√ √
The problem here is to minimise 4k + l subject to k l = 200.
The Lagrangian is √ √
L = 4k + l − λ( k l − 200).
(It is fine to use +λ here rather than −λ.) We now set the partial derivatives of L to zero:
√
l
Lk = 4 − λ √ = 0,
2 k
√
k
Ll = 1 − λ √ = 0,
2 l
√ √
Lλ = −( k l − 200) = 0.
Eliminating λ from the first two equations,
√ √
l k
√ = 4 √ , so l = 4k.
k l
√ √
Then,
√ √ the constraint equation (equivalently, the third of the above equations) k l = 200 tells
us k 4k = 200, which is 2k = 200, and so k = 100, l = 400. So the minimum cost is
4k + l = 4(100) + 400 = 800.
5
05a Mathematics 1
Question 7
Find the geometric series that has second term equal to 3 and sum to infinity equal
to 12.
With the usual notation, the sum to infinity is a/(1 − r) and the second term is ar. So
a
= 12, ar = 3.
1−r
So a = 12(1 − r) and
12(1 − r)r = 3,
giving
12r2 − 12r + 3 = 0, 4r2 − 4r + 1 = 0
(or an equivalent equation involving a if r has been eliminated instead). It follows that
(2r − 1)2 = 0,
and hence r = 1/2 The corresponding value of a is a = 6.
Section B
Question 8
i. The function f is given, for 0 < x < 1, by
f (x) = xx (1 − x)(1−x) .
Show that f has one critical (or stationary) point between 0 and 1, and that this
point is a local minimum.
We solve f 0 = 0. Calculating f 0 is quite tricky. Let g = xx . Then ln g = x ln x, so
g0
= 1 + ln x
g
and
g 0 = (1 + ln x)xx .
So, by the product rule and chain rule,
f 0 = (1 + ln x)xx (1 − x)(1−x) − (1 + ln(1 − x))xx (1 − x)(1−x) = xx (1 − x)(1−x) (ln x − ln(1 − x)) .
An alternative approach could be to start by taking the log of f as a step towards calculating
f 0 .We have
ln f = ln(xx (1 − x)(1−x) ) = x ln x + (1 − x) ln(1 − x),
so
1 0
f (x) = ln x − ln(1 − x).
f
So
f 0 (x) = xx (1 − x)(1−x) (ln x − ln(1 − x)) .
Now,
f 0 = 0 ⇐⇒ ln x = ln(1 − x) ⇐⇒ x = 1 − x ⇐⇒ x = 1/2.
6
As x passes through 1/2, f 0 changes sign from negative to positive, because for x < 1/2,
x < 1 − x and for x > 1/2, x > 1 − x. So it’s a local minimum. Alternatively, we can use the
second derivative test:

00 1 1 2
f = + xx (1 − x)(1−x) + (ln x − ln(1 − x)) xx (1 − x)(1−x)
x 1−x
and
f 00 (1/2) = 4(1/2)1/2 (1/2)1/2 > 0,
so it is a minimum.
ii. A firm’s marginal cost function is given by

q
M C = qeq + ,
q2 + 3q + 2
where q denotes the quantity produced. If its production is increased from q = 1 to
q = 2, find the increase in total cost.
What we want is Z 2
I= M C dq.
1
(The answer is not simply the value of M C(1). Calculating M C(1) and giving this as the
solution gets no credit.)
I = J + K,
where Z 2 Z 2
q
J= qeq dq, K = dq.
1 1 q2 + 3q + 2
By parts, Z 2
2
J = [qeq ]1 − eq dq
1
q 2
= [qe − eq ]1 = 2e − e2 − (e − e) = e2
2
For K, we use partial fractions. Noting that

q q
=
q2 + 3q + 2 (q + 1)(q + 2)
we must have
q A B
= +
q 2 + 3q + 2 q+1 q+2
for some A, B. We find that A = −1, B = 2. Then
2
K = [2 ln |q + 2| − ln |q + 1|]1 = 2 ln(4/3) − ln(3/2).
So,
I = e2 + 2 ln(4/3) − ln(3/2).
Question 9
i. For some numbers a, b, c, the function f is given by
f (x) = ax + bx2 + cx3 .
Given that Z 1
f (1) = 16, f 0 (1) = 38, and f (x) dx = 5,
0
7
05a Mathematics 1
find the numbers a, b and c.

This turns out to be a linear equations question. We have, first,
a + b + c = 16.
Since
f 0 (x) = a + 2bx + 3cx2 ,
we also have
a + 2b + 3c = 38.
Now,
1 1
ax2 bx3 cx4
Z
a b c
f (x) dx = + + = + + .
0 2 3 4 0 2 3 4
So,
a b c
+ + = 5.
2 3 4
So we have the following system of linear equations:
a+b+c = 16
a + 2b + 3c = 38
a b c
+ + = 5
2 3 4
or, equivalently,
a+b+c = 16
a + 2b + 3c = 38
6a + 4b + 3c = 60.
Using row operations, we have

     
1 1 1 16 1 1 1 16 1 1 1 16
 1 2 3 38  →  0 1 2 22  →  0 1 2 22  .
6 4 3 60 0 −2 −3 −36 0 0 1 8
So,
c = 8,
b = 22 − 2c = 6,
a = 16 − b − c = 2.
ii. Determine each of the following integrals.
ln(ln x)
Z
dx,
x
e2x
Z
dx.
(e2x − 1)(e2x + 2)
1
For the first integral, we use substitution. Let u = ln ln x. Then du = x ln x dx and so
Z
I = ueu du.
We now use parts: Z

I = ueu − eu du = ueu − eu + c
= ln ln xeln ln x − eln ln x = ln x(ln ln x) − ln x + c.
8
For the second integral, let u = e2x . Then

Z
1 du
I= .
2 (u − 1)(u + 2)
Now we use partial fractions. Z

1 1 1
I= − du
6 u−1 u+2
1 1
= (ln |u − 1| − ln |u + 2|) + c = (ln |e2x − 1| − ln |e2x + 2|) + c.
6 6
Question 10
i. A consumer has a utility function
u(x, y) = (x + 1)2 (y + 3)
for two goods, X and Y . (Here, x denotes the amount of X and y the amount of
Y .) Each unit of x costs p dollars and each unit of Y costs q dollars, and the
consumer has a budget for X and Y of M dollars. Use the Lagrange multiplier
method to find the quantities of X and Y the consumer will consume in order to
maximise his utility subject to the budget constraint. (Your answers should be
expressions depending on p, q and M .)
We want to maximise (x + 1)2 (y + 3) subject to px + qy = M . The Lagrangian is
L = (x + 1)2 (y + 3) − λ(px + qy − M ).
(L = (x + 1)2 (y + 3) + λ(px + qy − M ) is perfectly acceptable too.) We have
Lx = 2(x + 1)(y + 3) − λp = 0
and
Ly = (x + 1)2 − λq = 0.
These imply that
2(x + 1)(y + 3) (x + 1)2
= ,
p q
so
2(y + 3) (x + 1)
=
p q
and
2q 6q
x= y+ −1
p p
(or we could alternatively obtain y in terms of x). The constraint is px + qy = M , which means
2qy + 6q − p + qy = M
and hence, solving for y,

M + p − 6q
y=
3q
and
2q (M + p − 6q) 6q 2M − 6q − p
x= + −1= .
p 3q p 3p
9
05a Mathematics 1
ii. The function f is given, for some number a, by
f (x, y) = 2xy + xa y 2a .
Find, in terms of x, y and a, the partial derivatives
∂f ∂f ∂2f ∂2f
, , , .
∂x ∂y ∂x2 ∂y 2
Now suppose that we know that f satisfies
∂2f ∂2f
2x2 − y2 + 18f = 36xy.
∂x2 ∂y 2
Determine the possible values of a.
The derivatives are
fx = 2y + axa−1 y 2a ,
fy = 2x + 2axa y 2a−1 ,
fxx = a(a − 1)xa−2 y 2a ,
fyy = 2a(2a − 1)xa y 2a−2 .
So,
∂2f 2
2∂ f
2x2 − y + 18f = 2a(a − 1)xa y 2a − 2a(2a − 1)xa y 2a + 36xy + 18xa y 2a .
∂x2 ∂y 2
So,
∂2f 2
2∂ f
2x2 − y + 18f = 36xy
∂x2 ∂y 2
⇐⇒ xa y 2a (2a(a − 1) − 2a(2a − 1) + 18) = 0
⇐⇒ 2a(a − 1) − 2a(2a − 1) + 18 = 0
⇐⇒ 9 − a2 = 0
⇐⇒ a = ±3.
Question 11
i. James opens a savings account with a bank. The interest rate is guaranteed to be
fixed at a rate of 5% per annum for K years after he opens the account. (Here, K is
an integer greater than 1.) He opens the savings account with a payment of $100 on
1 January 2009, and makes further deposits of $100 yearly, on 1 January each year,
starting 1 January 2010. Find an expression (involving K, and in as simple a form
as possible) for the value of his savings K years after he opens the account (that is,
K years from 1 January 2009, just after he has made a new deposit of $100).
Suppose now that he continues to make the same deposits after this time, but that
the interest rate falls to the fixed rate of 4% for the next N years (where N is an
integer greater than 1). Find an expression (involving K and N , and in as simple a
form as possible) for the value of his savings K + N years after he opened the
account (that is, K + N years from 1 January 2009, just after he has made a new
deposit of $100).
subject guide gives the background and some examples. Essentially, what you have to do is
problems using general techniques for solving difference equations, but that is not expected (as
such techniques are not part of the 05A syllabus, though they are in 05B).
10
Let yt be the balance t years after opening the account (just after the (t + 1)st deposit). Then
y0 = 100
y1 = 100(1.05) + 100
y2 = (1.05)y1 + 100 = 100(1.05)2 + 100(1.05) + 100.
So, spotting the general pattern,
yK = 100(1.05)K + 100(1.05)K−1 + · · · + 100(1.05) + 100.
Now we simplify using the formula for a geometric series:
1 − (1.05)K+1
yK = 100 = 2000((1.05)K+1 − 1).
1 − 1.05
What happens next is that 1.05 is replaced in subsequent compounding by 1.04. So
yK+1 = (1.04)yK + 100
yK+2 = (1.04)yK+1 + 100 = (1.04)2 yK + (1.04)100 + 100
yK+3 = (1.04)3 yK + (1.04)2 100 + (1.04)100 + 100.
Then, spotting the general pattern,
yK+N = (1.04)N yK + 100((1.04)N −1 + (1.04)N −2 + · · · + (1.04) + 1)
1 − (1.04)N
= (1.04)N yK + 100
1 − 1.04
= 2000(1.04) ((1.05)K+1 − 1) + 2500((1.04)N − 1).
N
ii. The function f (x, y) is given by

f (x, y) = 2x4 − 2x2 y + y 2 − 2x2 + 1.
Find the critical (or stationary) points of f . Determine whether each critical point
is a local maximum, local minimum, or saddle point.
This involves techniques from Chapter 5 of the subject guide. The partial derivatives are
fx = 8x3 − 4xy − 4x, fy = −2x2 + 2y.
We solve fx = fy = 0. We have (from fy = 0) y = x2 . Substituting into fx = 0 gives
8x3 − 4x3 − 4x = 0 = 0, which is 4x(x2 − 1) = 0. So x = 0, 1, −1. Then y = 0, 1, 1, respectively.
So there are three critical points: (0, 0), (1, 1), (−1, 1).
fxx = 24x2 − 4y − 4y, fxy = −4x, fyy = 2.
2
At (0, 0): fxx fyy − fxy = −8 < 0, so this is a saddle point.
2
At (1, 1), fxx fyy − fxy = 16(2) − (−4)2 > 0 and fxx = 16 > 0, so it’s a minimum.
2
At (−1, 1), fxx fyy − fxy = 16(2) − (4)2 > 0 and fxx = 16 > 0, so it’s a minimum.
New format of the examination in 2010
From the 2010 examination, the format of the examination will change. There will be
no optional questions. Instead, there will be six compulsory questions in Section A
rather than the present seven. These will be worth 60 marks in total. There will be
two compulsory questions in Section B, each worth 20 marks. (There will not be a
choice of questions in Section B.)
11

Important note

references
section.
General remarks
Learning outcomes
have:
subjects.
1
questions.
Answer the question
Graph sketching
Calculators
this course.
2
occur again.
3

Important note

references
section.
and BOTH questions from Section B (20 marks each). Candidates are strongly advised to divide
their time accordingly.
Section A
Question 1
The function f is given by:
f (x) = (x2 + 10x − 10)e−x .
Find the critical (or stationary) points of f and, for each, determine whether it is a
local maximum, local minimum or inflexion point.
The relevant background material is discussed in Chapter 3 of the subject guide.
We have:
f 0 (x) = 2x + 10 − (x2 + 10x − 10) e−x = (20 − 8x − x2 )e−x .

Critical points are given by f 0 = 0. This is x2 + 8x − 20 = 0 (noting that e−x 6= 0), which is
(x + 10)(x − 2) = 0. So the critical points are x = −10 and x = 2.
4
We can check their nature by examining the behaviour of the sign of the derivative around each
point, or we can use the second derivative test.
With the second-derivative approach, we have:
f 00 (x) = −e−x (20 − 8x − x2 ) − (2x + 8)e−x .
When x = −10, we have:

f 00 (−10) = e10 (0) + 12e10 > 0
so this is a local minimum.
When x = 2, we have:
f 00 (2) = e−2 (0) − 12e−2 < 0
so this is a local maximum.
If using the sign of the second derivative, we examine how the sign of the derivative changes
around the critical points.
At −10, this changes from − to +. So this point is a local minimum.
At 2, this changes from + to −. So this point is a local maximum.
Question 2
A firm has fixed costs of 10 and its marginal revenue and marginal cost functions
are given (respectively) by:
M R = 11 − 2q and M C = q 2 − 4q + 3,
where q is the level of production. Find an expression, in terms of q, for the firm’s
profit. Also find the value of q which maximises this profit.
Integrating, we have: Z
TR = M R dq = 11q − q 2 + c.
This is because marginal revenue is the derivative of total revenue.
Now, because T R(0) = 0 = c, since there can be no revenue if no goods are sold, we have c = 0.
Because marginal cost is the derivative of total cost, we have:

Z
1
T C = M C dq = q 3 − 2q 2 + 3q + c.
3
To find c, note that T C(0) = F C = 10, so c = 10. (You need to explain this for a fully correct
answer.)
The profit, Π, is given by:

1
Π = T R − T C = 8q + q 2 − q 3 − 10.
3
To maximise the profit, we now solve Π0 = 0. This is:
8 + 2q − q 2 = 0.
5
Factoring:
(2 + q)(4 − q) = 0.
So q = 4, as quantities are non-negative. (Alternatively, we can use the formula for the solutions
of a quadratic equation.)
Now:
Π00 (q) = 2 − 2q
and, since Π00 (4) < 0, q = 4 maximises profit.
It is possible to determine the maximising value q = 4 by simply solving M R = M C, but this is

not answering the question because it does not, in particular, determine the profit function,
which is specifically requested.
Question 3
Determine the integrals:

1
Z Z
dx and x3 ln(x2 − 1) dx.
x2 − 3x + 2
Consider the first integral. This is a straightforward partial fractions integral. We have:
1 1
= .
x2 − 3x + 2 (x − 2)(x − 1)
So, there are constants A and B such that:
1 A B
= + .
x2 − 3x + 2 x−2 x−1
A=1 and B = −1.
Hence the integral is:

ln |x − 2| − ln |x − 1| + c.
The second integral can be done using a substitution, followed by integration by parts. (It can
also be done directly by integration by parts, but we will not present that approach here.)
Method A: substitute u = x2 − 1.
Let u = x2 − 1. Hence du = 2x dx. The integral is:

Z
1
(u + 1) ln u du.
2
1 u2
Z Z 2
1 1 u 1
(u + 1) ln u du = + u ln u − +u du
2 2 2 2 2 u
u2

1 2 1 u
= u + u ln u − − + c.
4 2 8 2
6
So the answer is:
(x2 − 1)2 x2 − 1

1 2 1
(x − 1)2 + (x2 − 1) ln(x2 − 1) − − + c.
4 2 8 2
Method B: substitute u = x2 .
Let u = x2 . (Other substitutions work.) Hence du = 2x dx. The integral is:

Z
1
u ln(u − 1) du.
2
u2
Z Z
1 1 2 1
u ln(u − 1) du = u ln(u − 1) − du.
2 4 4 u−1
Now we substitute v = u − 1 to see that:
u2 (v + 1)2
Z Z Z 2 Z
v + 2v + 1 1
du = dv = dv = v + 2 + dv.
u−1 v v v
This is:
v2 (u − 1)2
+ 2v + ln |v| + c = + 2(u − 1) + ln |u − 1|.
2 2
So the answer is:
1 2 (u − 1)2 1 1
u ln(u − 1) − − (u − 1) − ln |u − 1| + c
4 8 2 4
1 4 (x2 − 1)2 1 1
= x ln(x2 − 1) − − (x2 − 1) − ln |x2 − 1| + c.
4 8 2 4
Question 4
and Y are given by:
x = 200 − 2pX and y = 100 − pY ,
where x and y are the quantities (in tonnes) of X and Y demanded (respectively)
and the prices of X and Y are pX and pY (respectively). The firm’s joint total cost
function (that is, the total cost of producing x of X and y of Y ) is:
x2 + 2xy + y 2 + 20.
quantities x and y that maximise the firm’s profit.
The question asks us to find the profit as a function of x and y. That means we will need the
revenue, and hence the prices, in terms of x and y.
7
The profit function is:

x
= x 100 − + y(100 − y) − (x2 + 2xy + y 2 + 20)
2
3
= 100x + 100y − x2 − 2y 2 − 2xy − 20.
2
We solve:
Πx = 100 − 3x − 2y = 0 and Πy = 100 − 4y − 2x = 0.
The second equation tells us 2y = 50 − x and then the first becomes 50 − 2x = 0, so x = 25.
Therefore, y = 25 − x/2 = 25/2.
The second derivatives are:

Πxx = −3, Πxy = −2 and Πyy = −4.
Since Πxx < 0 and Πxx Πyy − Π2xy > 0, the profit is maximised at these values of x and y.
Question 5
Use the Lagrange multiplier method to find the values of x and y which maximise
the function f (x, y) = x1/2 y 1/2 subject to the constraint x + 2y = 40.
subject guide.
The Lagrangian is:

L = x1/2 y 1/2 − λ(x + 2y − 40).
The resulting equations (first-order conditions) are:
∂L 1
= x−1/2 y 1/2 − λ = 0
∂x 2
∂L 1
= x1/2 y −1/2 − 2λ = 0
∂y 2
and:
x + 2y = 40.
The first two equations show that:
1 −1/2 1/2 1
λ= x y = x1/2 y −1/2
2 4
so that y = x/2.
The third equation implies that x = 20. So, the optimal quantities are x∗ = 20 and y ∗ = 10.
Question 6

1
xn =
xn−1 + 2.
2
Find an explicit expression for xn in terms of n. Hence describe the behaviour of xn
as n → ∞.
8
Note: In all question papers, due to a printing error, this question was unreadable.
The examiners took this fully into account in marking the papers. We present here
the question as it was intended to be asked.
We have:
x0 = 3

1
x1 = 3+2
2
2
1 1
x2 = 3+ 2+2
2 2
3 2
1 1 1
x3 = 3+ 2+ 2 + 2.
2 2 2
In general:
n n−1 !
1 1 1
xn = 3 +2 + ··· + +1 .
2 2 2
This simplifies to:

n n n n
1 − (1/2)n

1 1 1 1
3 +2 =3 +4 1− =4− .
2 1 − 1/2 2 2 2
As n → ∞, because (1/2)n tends to 0 and decreases, the sequence increases to 4.
Section B
Question 7
(a) For some numbers a, b and c, let f be the function of one variable given by:
b
f (x) = ax3 + + c(1 + ln x),
x
for x > 0. Given that:
f (1) = 0, f 0 (1) = 20, f 00 (1) = 40,
show that:
a+b+c=0
3a − b + c = 20
6a + 2b − c = 40.
Use row operations to solve this system (to find a, b and c).
9

We have:
b
f (x) = ax3 + + c(1 + ln x)
x
so:
b c
f 0 (x) = 3ax2 − +
x2 x
and:
2b c
f 00 (x) = 6ax + − 2.
x3 x
a+b+c=0
3a − b + c = 20
6a + 2b − c = 40.
The standard matrix method approach is to reduce the augmented matrix to reduced form.
1 1 1 0
(A | b) = 3 −1 1 20
6 2 −1 40
   
1 1 1 0 1 1 1 0
R2 −3R1 ,R3 −6R1
(A | b) = 3 −1 1 20 −→ 0 −4 −2 20
6 2 −1 40 0 −4 −7 40
 
1 1 1 0
R3 /(−2)
−→ 0 2 1 −10
0 −4 −7 40
 
1 1 1 0
(R3 +2R2 )/(−5)
−→ 0 2 1 −10 .
0 0 1 −4

−10 − c
c = −4, b= = −3 and a = −b − c = 7.
2
values which you have found into the original equations, and this will show whether these
are correct or not. So you can tell if you have the wrong answer and, if you have time, you
can then re-work the calculation.) Examiners understand that arithmetical errors can be
made, especially in the stressful circumstances of an examination. Quite probably, the
examiners themselves would make some mistakes if they sat the paper. So, although there
are marks for correct calculation, there are also marks for using the right method (even if
10
constant from each entry of a row is valid. It is not. (And, if you do not know what we
mean by that, then you are probably not doing it, which is good!)
operations. Given this, other approaches to solving the system of equations are not
acceptable because to take such an approach is not to answer the question.
(b) Show that, if I is the integral:

Z
I= sin2 x dx,
then:
I = − sin x cos x + x − I.
Hence find an expression for I.
Hence or otherwise, evaluate the integral:
Z π
sin2 x dx.
0

See Chapter 4 of the subject guide for integration.
This question was not well done. The key is to use integration by parts at the outset.
Using integration by parts, we have:
Z Z
I = sin x sin x dx = sin x(− cos x) − cos x(− cos x) dx
Z
= − sin x cos x + cos2 x dx.
Now, to relate this back to the original integral, it is useful to use the fact that
sin2 x + cos2 x = 1, so that cos2 x = 1 − sin2 x. We then have:
Z
I = − sin x cos x + cos2 x dx
Z
= − sin x cos x + (1 − sin2 x) dx
Z
= − sin x cos x + x − sin2 x dx
= − sin x cos x + x − I.
It follows that:
2I = − sin x cos x + x
and so:
1 x
I = − sin x cos x + + c.
2 2
Therefore, the definite integral equals:
π
1 x π
− sin x cos x + = .
2 2 0 2
11
Question 8
(a) Use the Lagrange multiplier method to find the minimum value of:
s
1 1
+ 2
x2 y
√
among all positive x, y satisfying x + y = 2.

We have: 1/2
√

1 1
L= + 2 − λ(x + y − 2).
x2 y
−1/2
−3 1 1
Lx = −x + 2 −λ=0
x2 y
−1/2
1 1
Ly = −y −3 + 2 −λ=0
x2 y
and: √
x+y = 2.
−1/2 −1/2
−3 1 1 −3 1 1
−x + 2 = −y + 2
x2 y x2 y
and so x−3 = y −3 . Therefore, x = y.

√ √
Next, the constraint equation x + y = 2 tells us 2x = 2. So:
√
2
x=y= .
2
Therefore, the minimum value is:
1/2
4 4
+ = 41/2 = 2.
2 2
Do not forget this last bit: the question does not simply ask for the values of x and y which
minimise the value of the function, but for this minimum value itself.
(b) The function f is defined for x, y > 0 by:
f (x, y) = ln(x2 + y 2 ).
Find the partial derivatives:
∂f ∂f ∂2f ∂2f ∂2f

, , , , .
∂x ∂y ∂x2 ∂x∂y ∂y 2
Show that:
∂2f ∂2f
+ = 0.
∂x2 ∂y 2
12

We have:
2x
fx =
x2 + y2
2y
fy =
x2 + y 2
2(x2 + y 2 ) − 2x(2x) 2y 2 − 2x2

fxx = =
(x2 + y 2 )2 (x2 + y 2 )2
2x2 − 2y 2
fyy =
(x2 + y 2 )2
and:
−4xy
fxy = .
(x2 + y 2 )2
Hence:
2y 2 − 2x2 2x2 − 2y 2
fxx + fyy = + = 0.
(x2 + y 2 )2 (x2 + y 2 )2
13

Important note

references
section.
and BOTH questions from Section B (20 marks each). Candidates are strongly advised to divide
their time accordingly.
Section A
Question 1
The function f is given by:
f (x) = (x2 + 10x + 1)e−x .
Find the critical (or stationary) points of f and, for each, determine whether it is a
local maximum, local minimum or inflexion point.
The relevant background material is discussed in Chapter 3 of the subject guide.
We have:
f 0 (x) = 2x + 10 − (x2 + 10x + 1) e−x = (9 − 8x − x2 )e−x .

Critical points are given by f 0 = 0. This is x2 + 8x − 9 = 0 (noting that e−x 6= 0), which is
(x + 9)(x − 1) = 0. So the critical points are x = −9 and x = 1.
14
We can check their nature by examining the behaviour of the sign of the derivative around each
point, or we can use the second derivative test.
With the second-derivative approach, we have:
f 00 (x) = −e−x (9 − 8x − x2 ) − (2x + 8)e−x .
When x = −9, we have:

f 00 (−9) = e9 (0) + 10e9 > 0
so this is a local minimum.
When x = 1, we have:
f 00 (1) = e−1 (0) − 10e−1 < 0
so this is a local maximum.
If using the sign of the second derivative, we examine how the sign of the derivative changes
around the critical points.
At −9, this changes from − to +. So this point is a local minimum.
At 1, this changes from + to −. So this point is a local maximum.
Question 2
A firm has fixed costs of 10 and its marginal revenue and marginal cost functions
are given (respectively) by:
M R = 11 − 2q and M C = q 2 − 4q + 3,
where q is the level of production. Find an expression, in terms of q, for the firm’s
profit. Also find the value of q which maximises this profit.
Integrating, we have: Z
TR = M R dq = 11q − q 2 + c.
This is because marginal revenue is the derivative of total revenue.
Now, because T R(0) = 0 = c, since there can be no revenue if no goods are sold, we have c = 0.
Because marginal cost is the derivative of total cost, we have:

Z
1
T C = M C dq = q 3 − 2q 2 + 3q + c.
3
To find c, note that T C(0) = F C = 10, so c = 10. (You need to explain this for a fully correct
answer.)
The profit, Π, is given by:

1
Π = T R − T C = 8q + q 2 − q 3 − 10.
3
To maximise the profit, we now solve Π0 = 0. This is:
8 + 2q − q 2 = 0.
15
Factoring:
(2 + q)(4 − q) = 0.
So q = 4, as quantities are non-negative. (Alternatively, we can use the formula for the solutions
of a quadratic equation.)
Now:
Π00 (q) = 2 − 2q
and, since Π00 (4) < 0, q = 4 maximises profit.
It is possible to determine the maximising value q = 4 by simply solving M R = M C, but this is

not answering the question because it does not, in particular, determine the profit function,
which is specifically requested.
Question 3

1
Z Z
x3 ln x dx and dx.
1 + e2x
Consider the first integral. This can be solved by a straightforward integration by parts. We have:
x4
Z Z 4
x 1
4 4 x
x4 1
= ln x − x4 + c.
4 16
The second integral can be done using a substitution, followed by partial fractions.
Let u = 1 + e2x . Hence du = 2e2x dx and so du = 2(u − 1) dx and the integral is:
Z
1 du
.
2 (u − 1) u
Partial fractions can now be applied. This means there are constants A and B such that the
integrand takes the form:
A B
+ .
u−1 u
1 1
A= and B = −
2 2
Z
1 1 1 1
− du = ln |u − 1| − ln |u| + c
2(u − 1) 2u 2 2
1 1
= ln(e2x ) − ln(1 + e2x ) + c.
2 2
(Of course, the first term can be simplified to x.)
16
The substitution u = e2x leads to a very similar calculation.
There are other, less obvious, ways, too. A particularly quick and clever way is to note that the
integral can be rewritten (on multiplying numerator and denominator of the integrand by e−2x )
as:
e−2x
Z
dx.
e−2x + 1
Noting that the derivative of the denominator is exactly −2 times the numerator, it follows that
this equals:
1
− ln(e−2x + 1) + c.
2
(This is the same as the previous answer, of course, just written differently.)
Question 4
and Y are given by:
x = 200 − 2pX and y = 100 − pY ,
where x and y are the quantities (in tonnes) of X and Y demanded (respectively)
and the prices of X and Y are pX and pY (respectively). The firm’s joint total cost
function (that is, the total cost of producing x of X and y of Y ) is:
x2 + 2xy + y 2 + 20.
quantities x and y that maximise the firm’s profit.
The question asks us to find the profit as a function of x and y. That means we will need the
revenue, and hence the prices, in terms of x and y.
The profit function is:
x
= x 100 − + y(100 − y) − (x2 + 2xy + y 2 + 20)
2
3
= 100x + 100y − x2 − 2y 2 − 2xy − 20.
2
We solve:
Πx = 100 − 3x − 2y = 0 and Πy = 100 − 4y − 2x = 0.
The second equation tells us 2y = 50 − x and then the first becomes 50 − 2x = 0, so x = 25.
Therefore, y = 25 − x/2 = 25/2.
The second derivatives are:
Πxx = −3, Πxy = −2 and Πyy = −4.
Since Πxx < 0 and Πxx Πyy − Π2xy > 0, the profit is maximised at these values of x and y.
17
Question 5
Use the Lagrange multiplier method to find the values of x and y which maximise
the function f (x, y) = x1/2 y 1/2 subject to the constraint x + 2y = 40.
subject guide.
The Lagrangian is:

L = x1/2 y 1/2 − λ(2x + y − 20).
The resulting equations (first-order conditions) are:
∂L 1
= x−1/2 y 1/2 − 2λ = 0
∂x 2
∂L 1
= x1/2 y −1/2 − λ = 0
∂y 2
and:
2x + y = 20.
The first two equations show that:
1 −1/2 1/2 1
λ= x y = x1/2 y −1/2
4 2
so that y = 2x.
The third equation implies that 4x = 20. So, the optimal quantities are x∗ = 5 and y ∗ = 10.
Question 6
1
xn = xn−1 + 2.
3
Find an explicit expression for xn in terms of n. Hence describe the behaviour of xn

as n → ∞.
Note: In all question papers, due to a printing error, this question was unreadable.
The examiners took this fully into account in marking the papers. We present here
the question as it was intended to be asked.
18
We have:
x0 = 4

1
x1 = 4+2
3
2
1 1
x2 = 4+ 2+2
3 3
3 2
1 1 1
x3 = 4+ 2+ 2 + 2.
3 3 3
In general,: !
n n−1
1 1 1
xn = 4 +2 + ··· + +1 .
3 3 3
This simplifies to:
n n n n
1 − (1/3)n

1 1 1 1
4 +2 =4 +3 1− =3+ .
3 1 − 1/3 3 3 3
As n → ∞, because (1/3)n tends to 0 and decreases, the sequence decreases to 3.
Section B
Question 7
(a) For some numbers a, b and c, let f be the function of one variable given by:
b
f (x) = ax3 + + c(1 + ln x),
x
for x > 0. Given that:
f (1) = 0, f 0 (1) = 20, f 00 (1) = 40,
show that:
a+b+c=0
3a − b + c = 20
6a + 2b − c = 40.
Use row operations to solve this system (to find a, b and c).

We have:
b
f (x) = ax3 + + c(1 + ln x)
x
so:
b c
f 0 (x) = 3ax2 − +
x2 x
19
and:
2b c
f 00 (x) = 6ax + − 2.
x3 x
a+b+c=0
3a − b + c = 20
6a + 2b − c = 40.
The standard matrix method approach is to reduce the augmented matrix to reduced form.
 
1 1 1 0
(A | b) = 3 −1 1 20
6 2 −1 40
   
1 1 1 0 1 1 1 0
R2 −3R1 ,R3 −6R1
(A | b) = 3 −1 1 20 −→ 0 −4 −2 20
6 2 −1 40 0 −4 −7 40
 
1 1 1 0
R3 /(−2)
−→ 0 2 1 −10
0 −4 −7 40
 
1 1 1 0
(R3 +2R2 )/(−5)
−→ 0 2 1 −10 .
0 0 1 −4
−10 − c
c = −4, b= = −3 and a = −b − c = 7.
2
values which you have found into the original equations, and this will show whether these
are correct or not. So you can tell if you have the wrong answer and, if you have time, you
can then re-work the calculation.) Examiners understand that arithmetical errors can be
made, especially in the stressful circumstances of an examination. Quite probably, the
examiners themselves would make some mistakes if they sat the paper. So, although there
are marks for correct calculation, there are also marks for using the right method (even if
constant from each entry of a row is valid. It is not. (And, if you do not know what we
mean by that, then you are probably not doing it, which is good!)
operations. Given this, other approaches to solving the system of equations are not
acceptable because to take such an approach is not to answer the question.
20
(b) Show that, if I is the integral:

Z
I= cos2 x dx,
then:
I = sin x cos x + x − I.
Hence find an expression for I.
Hence or otherwise, evaluate the integral:
Z π
cos2 x dx.
0

See Chapter 4 of the subject guide for integration.
This question was not well done. The key is to use integration by parts at the outset.
Using integration by parts, we have:
Z Z
I = cos x cos x dx = sin x cos x − sin x(− sin x) dx
Z
= sin x cos x + sin2 x dx.
Now, to relate this back to the original integral, it is useful to use the fact that
sin2 x + cos2 x = 1, so that sin2 x = 1 − cos2 x. We then have:
Z Z
I = sin x cos x + sin2 x dx = sin x cos x + (1 − cos2 x) dx
Z
= sin x cos x + x − cos2 x dx
= sin x cos x + x − I.
It follows that:
2I = sin x cos x + x
and so:
1 x
I= sin x cos x + + c.
2 2
Therefore, the definite integral equals:
π
1 x π
sin x cos x + = .
2 2 0 2
Question 8
(a) Use the Lagrange multiplier method to find the minimum value of:
s
1 1
2
+ 2
x y
√
among all positive x, y satisfying x + y = 2.
21


We have:
1/2
√

1 1
L= 2
+ 2 − λ(x + y − 2).
x y
−1/2
1 1
Lx = −x−3 2
+ 2 −λ=0
x y
−1/2
−3 1 1
Ly = −y 2
+ 2 −λ=0
x y
and:
√
x+y = 2.

−1/2 −1/2
1 1 1 1
−x−3 2
+ 2 = −y −3 2
+ 2
x y x y
and so x−3 = y −3 . Therefore, x = y.

√ √
Next, the constraint equation x + y = 2 tells us 2x = 2. So:
√
2
x=y= .
2
Therefore, the minimum value is:
1/2
4 4
+ = 41/2 = 2.
2 2
Do not forget this last bit: the question does not simply ask for the values of x and y which
minimise the value of the function, but for this minimum value itself.
(b) The function f is defined by:
1
f (x, y) = p .
x2 + y2
Find the partial derivatives:
∂f ∂f ∂2f ∂2f ∂2f

, , , , .
∂x ∂y ∂x2 ∂x∂y ∂y 2
Show that:
∂2f ∂2f 3
+ = (f (x, y)) .
∂x2 ∂y 2

22

We have:
−x
fx =
(x2 + y 2 )3/2
−y
fy =
(x2 + y 2 )3/2
−(x2 + y 2 )3/2 + 3x2 (x2 + y 2 )1/2 −(x2 + y 2 ) + 3x2 2x2 − y 2

fxx = 2 2 3
= 2 2 5/2
= 2
(x + y ) (x + y ) (x + y 2 )5/2
2y 2 − x2
fyy =
(x2 + y 2 )5/2
and:
3xy
fxy = .
(x2 + y 2 )5/2
Hence:
2x2 − y 2 + 2y 2 − x2 x2 + y 2 1
fxx + fyy = 2 2 5/2
= 2 = 2 = (f (x, y))3 .
(x + y ) (x + y 2 )5/2 (x + y 2 )3/2
23

05a Mathematics 1
Important note
This commentary reflects the examination and assessment arrangements for this unit in the
Specific comments on questions – Zone B
Section A
Question 1
The functions f and g are defined by

f (x) = x2 + 5x + 6, g(x) = 2 − x2 + x.
Show that the curves y = f (x) and y = g(x) do not intersect and sketch the curves.
For which values of c will the graph of f intersect the graph of the function
h(x) = g(x) + c? For which particular value of c will there be precisely one point at
which they intersect?
This question concerns curve sketching, discussed in Chapter 2 of the subject guide.
The equation x2 + 5x + 6 = −x2 + x + 2 is equivalent to 2x2 + 4x + 4 = 0. This has no solution.

We can see this either by observing that its discriminant is negative, or by noting that it’s
equivalent to (x + 1)2 = −1. So the two curves do not intersect, because an intersection of
y = f (x) and y = g(x) corresponds to a solution of f (x) = g(x).
To find where the curves cross the x-axis, note that f (x) = (x + 2)(x + 3) and
g(x) = (−x + 2)(x + 1), so f crosses the x-axis at −2, −3 and g crosses at 2, −1. Clearly, f
crosses the y-axis at 6 and g at 2.
The minimum value of f is at x = −5/2 and the maximum value of g is at x = 1/2. The curves
are illustrated in Figure 2
The graphs of f and h intersect if and only if there’s a solution to

x2 + 5x + 6 = −x2 + x + 2 + c.
This equation is equivalent to 2x2 + 4x + (4 − c) = 0.
This has discriminant

∆ = 16 − 8(4 − c) = 8(c − 2).
For solutions (and hence intersection) we require ∆ ≥ 0, meaning c ≥ 2.
There will be exactly one solution when ∆ = 0; that is, when c = 2.
1
05a Mathematics 1
25
20
15
10
–4 –2 2 4
x
–5
Figure 2: Sketch of the curves f (x) = x2 + 5x + 6 and g(x) = 2 − x2 + x.
Question 2
matrix method.
4x + y − 2z = 4,
2x + 3y − 2z = 4,
2x + 5y + 2z = 8.
suffice.
The system expressed in matrix form is

    
4 1 −2 x 4
2 3 −2 y  = 4 .
2 5 2 z 8
4 1 −2 4
 2 3 −2 4  .
2 5 2 8

     
4 1 −2 4 4 1 −2 4 4 1 −2 4
 2 3 −2 4  →  4 6 −4 8 → 0 5 −2 4 
2 5 2 8 4 10 4 16 0 9 6 12
     
4 1 −2 4 4 1 −2 4 4 1 −2 4
→  0 45 −18 36  →  0 45 −18 36  →  0 5 −2 4  .
0 45 30 60 0 0 48 24 0 0 2 1
2

4x + y − 2z = 4
5y − 2z = 4
2z = 1.
So, z = 1/2, y = (1/5)(4 + 2z) = 1 and x = (1/4)(4 + −y + 2z) = 1.
Question 3

x ln x
Z Z
dx, √ dx.
x2 + 4x + 3 x
partial fractions.
x
Z
For 2
dx, we use partial fractions.
x + 4x + 3
We have
x x 3/2 1/2
= = − .
x2 + 4x + 3 (x + 3)(x + 1) x+3 x+1
So,
x 3 1
Z
dx = ln |x + 3| − ln |x + 1| + c.
x2 + 4x + 3 2 2
ln x
Z
For √ dx, we use integration by parts. We have
x
ln x
Z Z
√ dx = x−1/2 ln x dx
x
1
Z
= 2x1/2 ln x − 2x1/2 dx
x
√ √
= 2 x ln x − 4 x + c.
3
05a Mathematics 1
Question 4
The function f is defined for positive x and all y by
f (x, y) = y 2 ln x − x ln x.
Find the critical (or stationary) points of f and determine, for each, whether it is a
local maximum, a local minimum, or a saddle point.
This is a fairly standard type of question, using the material in Chapter 5 of the subject guide.
With f (x, y) = y 2 ln x − x ln x, we have
y2
fx = − ln x − 1, fy = 2y ln x.
x
So we solve fx = fy = 0.
From 2y ln x = 0 we have y = 0 or x = 1.
If y = 0, then fx = 0 becomes ln x = −1, so x = e−1 = 1/e.
If x = 1 then fx = 0 becomes y 2 − 1 = 0, so y = ±1.
So the critical points are

(1/e, 0), (1, 1), (1, −1).
To determine the nature of the critical points, we use the second-derivative test. We have
y2 1 2y
fxx = − 2
− , fyy = 2 ln x, fxy = .
x x x
2
Consider first (1/e, 0). Here, fxx = −e, fyy = −2, fxy = 0. So, fxx fyy − fxy = 2e > 0 and
fxx < 0, so the point is a local maximum.
2
Consider (1, 1). Here, fxx = −2, fyy = 0, fxy = 2. So, fxx fyy − fxy = −4 < 0, so the point is a
saddle point.
2
Consider (1, −1). Here, fxx = −2, fyy = 0, fxy = −2. So, fxx fyy − fxy = −4 < 0, so the point is
a saddle point.
Question 5
The production of an amount q of a high technology good requires the input of

capital, k, and labour, l, where
q = 10k1/4 l1/4 .
If the price of capital is $400 per unit and the cost of labour is $100 per unit, find
the minimum cost, C(q), of producing q.
In determining its production schedule, the manufacturer produces an amount q

that maximises 2000q − C(q). What is this optimal value of q?
Lagrange multiplier method, discussed in chapter 5 of the subject guide.
For this, we need to minimise 400k + 100l subject to 10k 1/4 l1/4 = q.
4
The Lagrangian is
L = 400k + 100l − λ(10k 1/4 l1/4 − q).
Here, one can also use +λ in place of −λ, or M − 100k − 400l in place of 100k + 400l − M .
We have
10 −3/4 1/4
Lk = 400 − λk l =0
4
10 1/4 −3/4
Ll = 100 − λk l =0
4
10k 1/4 l1/4 = q.

(Here, Lk denotes ∂L/∂k and Ll denotes ∂L/∂l.)
Eliminating λ,
100k −1/4 l3/4 = 400k 3/4 l−1/4 ,
l = 4k.
Then, 10k 1/4 l1/4 = q implies

1/4
k 1/4 (4k) = q/10.
So √ 1/2
2k = q/10.
Then,
k = (k 1/2 )2 = q 2 /200
and
l = 4k = 2q 2 /100.
The minimum cost is the cost with these values of k and l, which is
C(q) = 400(q 2 /200) + 100(2q 2 /100) = 4q 2 .
For the next part, we need to maximise
f (q) = 2000q − 4q 2 .
We have f ′ (q) = 2000 − 8q, and f ′ (q) = 0 ⇐⇒ q = 250.
We also note f ′′ (q) < 0, so q = 250 maximises the profit.
5
05a Mathematics 1
Question 6
An investor saves money in a bank account paying interest at a fixed rate of 5%,
where the interest is paid once per year, at the end of the year. She makes an initial
deposit of 20000 dollars and, then, at the end of each of the next N years, just after
the interest has been paid, she withdraws an amount of 500 dollars. Find an
expression, in terms of N , for the amount of money in the account at the end of N
years, just after the N th withdrawal has been made.
chapter 7 of the subject guide gives the required background material.
Let yn be the amount of money after the nth withdrawal. Then:
y1 = 20000(1.05) − 500,
y2 = (1.05)y1 − 500 = 20000(1.05)2 − 500(1.05) − 500,

y3 = y2 (1.05) − 500 = 20000(1.05)3 − 500(1.05)2 − 500(1.05) − 500.
Spotting the pattern,
yN = 20000(1.05)N − 500(1.05)N −1 − 500(1.05)N −2 − · · · − 500(1.05) − 500 (∗)

(1.05)N − 1
= 20000(1.05)N − 500
(1.05) − 1
= 20000(1.05)N − 10000((1.05)N − 1)
= 10000(1.05)N + 10000.
Section B
Question 7
(a) A firm is a monopoly and it has fixed costs of 5 and average variable cost function
q 4 − q − 5q 2 . The demand equation for the good it produces is p + q = 20, where p
is the selling price. Find an expression for the firm’s profit, as a function of q, and
determine the production level that maximises the firm’s profit.
This is essentially a one-variable optimisation problem, and the underlying material can be found
in Chapter 3 of the subject guide.
Because the average cost is AV C = (T C − F C)/q, where F C is the fixed cost and T C is the
total cost, it follows that the total cost is
T C = q(AC) + F C = q 5 − q 2 − 5q 3 + 5.
The total revenue is T R = pq = (20 − q)q. Here, we’ve used the fact that the firm is a monopoly,
so that we can use the demand equation to obtain an expression for the price in terms of the
firm’s production level, q.
So Π = T R − T C = 20q − q 5 + 5q 3 − 5.
We solve Π′ = 0. This is 20 − 5q 4 + 15q 2 = 0, which we note is a quadratic in q 2 and simplifies as
−5(q 2 − 4)(q 2 + 1) = 0, to which the only economically meaningful solution is q = 2.
Further, since Π′′ = −20q 3 + 30q, we have Π′′ (2) < 0 and therefore we maximise profit.
6
(b) The function f is given, for x, y > 0, by

y
f (x, y) = (x + y) ln .
x
Determine the first and second-order partial derivatives of f and show that
∂ 2f ∂ 2f ∂ 2f
x2 + 2xy + y2 = 0.
∂x2 ∂x∂y ∂y 2

It would be useful to note that
y
f (x, y) = (x + y) ln = (x + y)(ln y − ln x).
x
But suppose we just proceed directly. Then:
y y x y x + y y y
fx = ln + (x + y) − 2 = ln − = ln −1− ,
x x y x x x x
y 1x y x
fy = ln + (x + y) = ln +1+ ,
x xy x y
x y y 1 y
fxx = − 2 + 2 = − + 2,
y x x x x

x 1 1 1 1
fxy = fyx = − = − ,
y x x y x

x 1 x 1 x
fyy = − 2 = − 2.
y x y y y
So,
2
∂2f 2

2∂
f 2∂ f 2 1 y 1 1 2 1 x
x + 2xy + y = x − + 2 + 2xy − +y −
∂x2 ∂x∂y ∂y 2 x x y x y y2
= −x + y + 2x − 2y + y − x
= 0.
Question 8
A consumer has utility function

xy
u(x, y) = ,
x+y
for two goods, X and Y . Here, x > 0 denotes the amount of X consumed and y > 0
the amount of Y consumed. Each unit of x costs p dollars, each unit of Y costs q
dollars, and the consumer has a budget for X and Y of M dollars.
Use the Lagrange multiplier method to find the quantities x∗ of X and y ∗ of Y the
consumer will consume in order to maximise his utility subject to the budget
constraint. (Your answers will depend on p, q and M .)
Show that
M
x∗ + y ∗ = √ √ .
p q
Hence find the optimal value, λ∗ , of the Lagrange multiplier, simplifying your
expression as much as possible.
7
05a Mathematics 1
Suppose that V = u(x∗ , y ∗ ) is the maximum achievable utility. Find an explicit

expression for V in terms of p, q and M , simplifying your answer as much as
possible.
∂V
Find and verify that it is equal to λ∗ .
∂M
∂V
Find . Use this to find an approximate expression for the change in V if the
∂p
price of X increases by one dollar.
This question primarily concerns the Lagrange multiplier method discussed in Chapter 5 of the
subject guide.
To determine optimal affordable consumption, the consumer maximises u(x, y) subject to the
budget constraint px + qy = M .
The Lagrangian is
xy
L= − λ(px + qy − M ).
x+y
We have
y(x + y) − xy y2
Lx = − pλ = − pλ = 0
(x + y)2 (x + y)2
x(x + y) − xy x2
Ly = − qλ = − qλ = 0
(x + y)2 (x + y)2
px + qy = M.
Eliminating λ from the first two of these equations, we see that
y2 x2
= ,
p q
so that
p
r
y= x.
q
Then, px + qy = M implies
√
px + pqx = M,
∗
so x , the solution to this, is given by
M
x∗ = √ .
p + pq
Then,
p ∗ p M M
r r
y∗ = x = √ = √ .
q q p + pq q + pq
8
Now,

1 1
x∗ + y ∗ = M √ + √
p + pq q + pq
√
M (p + q + 2 pq)
= √ √
(p + pq)(q + pq)
√ √
M ( p + q)2
= √ √ √ √
p q( p + q)2
M
= √ √ .
p q
The corresponding value of λ∗ is

(y ∗ )2
λ∗ = .
p(x∗ + y ∗ )2
So,
(y ∗ )2 pq
λ∗ =
pM 2
M 2q
= √
M 2 (q + pq)2
1
= √ √ .
( q + p)2
We also have
x∗ y ∗
V =
x∗ + y∗
√ √
M M p q
= √ √ √ √ √ √
p( p + q) q( q + p) M
M
= √ √ .
( p + q)2
So
∂V 1
= √ √ = λ∗ .
∂M ( p + q)2
We have
∂V M
= −√ √ √ .
∂p p( q + p)3
√ √ √
So if p increases by 1 dollar, V decreases by approximately M/( p( q + p)3 ). This last part
relies on us understanding that the derivative of V with respect to p means the rate of change of
V when p is changed.

Math 1 All Answers

Uploaded by

Copyright:

Available Formats

Math 1 All Answers

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Math 1 All Answers

Uploaded by

Copyright:

Available Formats

Examination papers and

Examiner’s report 2004

Exam technique: general remarks

Speciﬁc comments on questions

2. We ﬁnd f (x) = e−x/2 − (1/2)(1 + x)e−x/2 = (1/2)e−x/2 (1 − x). So f (x) = 0 means

4. We have total revenue given, as a function of x, y, by

T R = xpX + ypY = x(100 − x/2) + y(100 − y),

and the proﬁt is

From the ﬁrst two equations,

6. The system expressed in matrix form (a step some forgot) is

So, z = 1/2, y = 2 − z = 1 and x = 4 + z − 3y = 2. Simply manipulating equations

The solutions are d = 1/2 and a = 2.

8. (a) We have T C = V C + F C = q(AV C) + F C = q 3 + q 2 + eq + 10 and

fxx (1, 1) = 2, fxy (1, 1) = −2, fyy (1, 1) = 4,

so, since fxx (1, 1) > 0 and

fxx (1, 1)fyy (1, 1) − (fxy (1, 1))2 > 0,

we see that (1, 1) is a local minimum.

4p1 − 2p2 + 2p3 = 50

The equations to be solved are

From the ﬁrst two equations,

(b) Let yN be the required number. Then we have

y3 = (0.95) (0.95)2 100 + (0.95)3 + 3 + 3 = (0.95)3 100 + (0.95)2 3 + (0.95)3 + 3,

yN = (0.95)N 100 + (0.95)N −1 3 + (0.95)N −2 3 + · · · + (0.95)3 + 3

= (0.95)N 100 + 3 + 3(0.95) + 3(0.95)2 + · · · + 3(0.95)N −1

This decreases with N , converging to 60.

Examination paper for 2005

Examiner’s report 2006

Examiners’ commentaries 2010

Specific comments on questions – Zone A

The equation x2 − x − 2 = x − 4 is equivalent to x2 − 2x + 2 = 0. This has no solution. We can

This equation is equivalent to x2 − 2x + (a − 2) = 0.

This has discriminant

There will be exactly one solution when ∆ = 0; that is, when a = 3.

Figure 1: Sketch of the curves y = x − 4 and y = x2 − x − 2.

The system expressed in matrix form is

Here is one way. (There are others, equally valid.)

It follows, from this last matrix, that

Determine the integrals

The function f is defined for positive y and all x by

With f (x, y) = x2 ln y − y ln y, we have

If x = 0, then fy = 0 becomes ln y = −1, so y = e−1 = 1/e.

If y = 1 then fx = 0 becomes x2 − 1 = 0, so x = ±1.

So the critical points are

The production of an amount q of a very specialized high technology good requires

In determining its production schedule, the manufacturer spends an amount M that

For this, we need to maximise k 1/4 l1/4 subject to 100k + 400l = M .

Then, 100k + 400l = M implies 800l = M.

So l = M/800 and k = 4l = 4M/800.

Then, the maximum output is

For the next part, we need to maximise

We also note f ′′ (M ) < 0, so M = 16 is a maximum.

Chapter 7 of the subject guide gives the required background material.

Let yn be the amount of money after the nth deposit. Then:

Spotting the pattern,

yN = D(1 + r)N −1 + D(1 + r)N −2 + · · · + D(1 + r) + D (∗)

Answer both questions from this section (20 marks each).

(b) The function f is given, for x, y > 0, by

Partial differentiation is discussed in Chapter 5 of the subject guide.