Calculus Chapter 2 13

Calculus Chapter 2
Differentiation

Slopes of curves

Figure 2.1. Tangent to a curve

We have observed that if you calculate the slope between any two points on a line,
then you get the same value. In other words, a (straight) line has constant slope. This is
obviously not true for a general curve. Here the best we can do is to talk about the slope
at any particular point on that curve, which we shall define to be the slope of the tangent
line to the curve at that point. Thus

slope of curve at a point = slope of tangent at that point.

(See Figure 2.1.)

HaL HbL

Figure 2.2. (a) Not continuous (b) continuous but not smooth
It follows that for the slope to be defined, the tangent must exist and must not be
vertical. For the tangent to exist, the curve must be continuous, i.e. not have a jump,
and smooth, i.e. not have a sharp corner or kink. (See Figure 2.2.) From now on we shall
assume that all curves are continuous and smooth, whenever we need them to be. Just as
14 MATH1041

with lines, we shall say that a curve is increasing at a point if its slope is positive there,
and decreasing if its slope is negative.
How can we evaluate the slope of a curve at a point? Clearly we cannot do so by
considering that point alone; we must look at all nearby points as well. We are aiming
at finding the slope of the tangent (or touching line), but we need at least two points to
calculate a slope, and the tangent has (in this neighbourhood, at least) only one point in
common with the curve. Thus we first consider the secants (or cutting lines) that cut the
curve in two distinct points, as illustrated in Figure 2.3.

y=fHxL
Hx2 , y2 L

Hx1 , y1 L

HaL HbL

Figure 2.3. (a) Secant (b) tangent

Suppose (x1 , y1 ) and (x2 , y2 ) are two points on the curve y = f (x). The slope of the secant
y2 − y1
is .
x2 − x1
If x2 approaches x1 , then the secant approaches the tangent at (x1 , y1 ).
Therefore as x2 approaches x1 , so the slope of the secant approaches the slope of the
tangent.
y2 − y1
Therefore as x2 approaches x1 , so approaches the slope of the curve at (x1 , y1 ).
x2 − x1
y2 − y1
Therefore as x2 → x1 , so → the slope of the curve at (x1 , y1 ).
x2 − x1
y2 − y 1
Therefore, at (x1 , y1 ) slope of curve = lim .
x2 →x1 x2 − x1

This process of finding slopes of curves is called differentiation, and there are several
notations for it. If the curve has equation y = f (x), then the slope of the curve at the
point (x, y) is called its derivative with respect to x, and it is written as
dy df (x)
y ′ or or f ′ (x) or .
dx dx

Hx+Dx, y+DyL Hx+Dx, fHx+DxL

Hx, yL Hx, fHxLL

Figure 2.4. Different notations for the derivative

 Calculus Chapter 2 15

The definition of derivative, as given above, can be rewritten using different notation,
as shown in Figure 2.4.
• Put (x1 , y1 ) = (x, y) and (x2 , y2 ) = (x + ∆x, y + ∆y). Then x2 − x1 = ∆x, also
y2 − y1 = ∆y, and x2 → x1 is the same as ∆x → 0, so the definition becomes
∆y dy
lim = = y′ .
∆x→0 ∆x dx
• Since (x, y) and (x + ∆x, y + ∆y) are both points on the curve, it follows that y = f (x)
and y + ∆y = f (x + ∆x). Thus ∆y = f (x + ∆x) − f (x) and the above definition becomes
f (x + ∆x) − f (x) df (x)
lim = = f ′ (x).
∆x→0 ∆x dx
Tutorial questions — Slopes of curves
D 1. From first principles (i.e. from the definition of the derivative) find:
dy 1
(a) if y = x2 + x + 1 (b) y ′ if y =
dx x
√ df (t) 1
(c) f (x) if f (x) = x2 + 1
′
(d) if f (t) = √ .
dt 2+ t
Simple rules for differentiation
We do not want to go back to first principles, i.e. the definition, every time we differentiate
something, so we establish some useful general results.
Theorem. The derivative of a constant is zero.
Proof. Suppose y = f (x) = c (constant). Then y + ∆y = f (x + ∆x) = c also, so ∆y = 0.
∆y
Thus ∆x = 0 (for every non-zero value of ∆x), and by taking limits as ∆x → 0 we obtain
dy
dx = 0.
This theorem simply expresses the well known result that a horizontal line has zero
slope.
Theorem. The derivative of a sum is the sum of the derivatives.
Proof. Suppose y = f (x) + g(x). Then y + ∆y = f (x + ∆x) + g(x + ∆x), so
∆y = f (x + ∆x) + g(x + ∆x) − {f (x) + g(x)}
= {f (x + ∆x) − f (x)} + {g(x + ∆x) − g(x)}.
∆y f (x + ∆x) − f (x) g(x + ∆x) − g(x)
Therefore = + .
∆x ∆x ∆x
If we take limits as ∆x → 0, then we obtain
dy df (x) dg(x)
= + ,
dx dx dx
as required.
 16 MATH1041

Theorem. The derivative of a constant times a function is that constant times the deriv-
ative.

Proof. Let y = cf (x). Then y + ∆y = cf (x + ∆x). Therefore

∆y cf (x + ∆x) − cf (x) f (x + ∆x) − f (x)

= =c .
∆x ∆x ∆x

If we take limits as ∆x → 0, then we obtain

dy
= cf ′ (x),
dx

as required.

Theorem. For any constant n, the derivative of xn is nxn−1 .

Proof. (This assumes n is a positive integer — more general proofs come later). First note
that if a 6= b =
6 0, then

a n − bn n a a a o
= bn−1 1 + ( ) + ( )2 + · · · + ( )n−1 ,
a−b b b b
a
since the expression in brackets is a geometric series with n terms and common ratio b.
Now suppose y = xn and put a = x + ∆x and b = x. If x 6= 0 and ∆x 6= 0, then

∆y (x + ∆x)n − xn
=
∆x ∆x
 
n−1 x + ∆x x + ∆x 2 x + ∆x n−1
=x 1+( )+( ) + ··· + ( ) .
x x x

If ∆x → 0, then x+∆xx → 1 (since x 6= 0), so the geometric series in brackets tends to n

since it has n terms. Therefore, taking limits as ∆x → 0, we have

dxn
= nxn−1 .
dx

This can easily shown to be true when x = 0 as well.

Note that with the above four results we can differentiate any polynomial.

Tutorial questions — Simple rules for differentiation

D 2. Use the rules for differentiation to write down the derivatives of:
(a) x4 + 2x3 − 5x + 2 with respect to x (b) y log10 2 + log10 3 with respect to y
(c) 22 + 33 − 44 with respect to x (d) t3 − t2 + 5 with respect to t
2 1 3 √ 1
(e) − 2 + + 3 with respect to x (f) x x − √ with respect to x.
x x x x
Calculus Chapter 2 17

An important limit
RH1,tanΘL
PHcosΘ,sinΘL

Θ
O QH1,0L

sin θ
Figure 2.5. Limit of θ as θ → 0
In Figure 2.5, where P is the point (cos θ, sin θ) in the first quadrant, it is clear that

Area of △OP Q < area of sector OP Q < area of △OQR

1 1 2 1
.˙. 2 × 1 × sin θ < 21 θ < 2 × 1 × tan θ
.˙. sin θ < θ < tan θ
θ 1
.˙. 1 < sin θ < cos θ
sin θ
.˙. cos θ < θ < 1.

sin θ
As θ approaches 0 or tends to 0, so cos θ tends to 1. Therefore θ must also tend to 1,
because it is squeezed between cos θ and 1. We write an arrow for “tends to”, so we have

sin θ
→ 1 as θ → 0.
θ

sin θ
Another way is to say that the limit as θ → 0 of θ is 1, which is written

sin θ
lim = 1.
θ→0 θ

(Do not confuse the equals sign in the “lim” notation with the arrow in the “tends to”
notation.) In the above we considered θ > 0 only. But if θ < 0, say θ = −φ, then

sin θ sin(−φ) − sin φ sin φ

= = = ,
θ −φ −φ φ

so the result is the same as before.

 18 MATH1041

Differentiation of sine and cosine

Theorem. If x is given in radians, then

d sin x d cos x
= cos x and = − sin x.
dx dx

Proof. If y = sin x, then

∆y sin(x + ∆x) − sin x

=
∆x ∆x
2 cos(x + 21 ∆x) sin( 21 ∆x)
= (using the formula for sin C − sin D)
∆x
sin( 1 ∆x)
= cos(x + 21 ∆x) 1 2 .
2 ∆x

sin θ
We have just proved that θ → 1 as θ → 0, so if we take limits on both sides as ∆x → 0,
dy
then we obtain dx = cos(x + 0) 1 = cos x, as required.
The result for the derivative of cos x is left as an exercise. Note the minus sign: all co-ratios
have minus signs in their derivatives.

Tutorial questions — Differentiation of sine and cosine

D 3. Find the limits as θ → 0 of

θ sec θ − 1 sin 2θ sin kθ

, θ cot θ, , , .
sin θ θ2 θ θ

(Hint: multiply top and bottom of the last one by k.)

D 4. From first principles find the derivatives with respect to θ of:
(a) cos θ (b) tan θ (c) sin2 θ E (d) sin(θ2 ).
D 5. Write down the derivatives of:
√
(a) − sin x + 3 cos x with respect to x (b) θ − sin θ + π with respect to θ.
d
T 6. Show that R cos(x − α) = −R sin(x − α), where R and α are constant, by first
dx
expanding both sides, using the addition formulae for sin and cos.

Figure 2.6. Tangent and normal

 Calculus Chapter 2 19

Tangents and normals

The tangent to the curve y = f (x) at the point where x = a is (by definition) the line with
slope f ′ (a) that passes through the point (a, f (a)), as illustrated in Figure 2.6. Thus we
can find its equation by the usual method for lines with a given slope passing through a
given point (see Algebra Chapter 1).
The normal to a curve at any point is the line through that point perpendicular to the
tangent there. Since the product of the slopes of perpendicular lines is −1, it follows that
−1
the normal to y = f (x) at the point where x = a has slope ′ and passes through the
f (a)
point (a, f (a)). Thus we can find its equation as before.

Tutorial questions — Tangents and normals

D 7. Find the equations of the tangents and normals to the curve y = 3x − x2 at the points
where the curve cuts the x axis (i.e. x = 0 and x = 3).
D 8. Find the equations of the tangents and normals to the curve y = x3 − 4x at the
points where the curve cuts the x axis. Do these tangents and normals cut the curve
again?
−1
D 9. Show that the slope of the normal to the curve y = x2 at the point where x = t is ,
2t
and find the equation of this normal. If the normal cuts the y axis at the point P (0, c),
show that t2 − c + 12 = 0. Deduce that if c > 21 then there are two solutions for t, and
hence two (non-vertical) normals from P to the curve, but that if c ≤ 21 then there are
no non-vertical normals. (The y axis is also a normal, for which t = 0 but the slope is
undefined.)

Maxima and minima

A (local) maximum on a curve is a point where the y value is greater than or equal to
the y values of all nearby points on the curve. Similarly for (local) minimum, replacing
“greater” by “less.” Maxima or minima are also called turning points or critical points.
At a turning point on a smooth curve the tangent must be horizontal, so the derivative
equals zero. Thus in order to find the turning points on the curve y = f (x), first solve the
equation f ′ (x) = 0.

local maximum

positive derivative negative derivative positive derivative

local minimum

Figure 2.7. Turning points and the sign of the derivative

 20 MATH1041

Finally, to determine which turning points are maxima and which are minima, note that
in passing from left to right through a maximum the derivative changes sign from positive
to negative. Through a minimum the derivative changes from negative to positive. These
are illustrated in Figure 2.7. If the derivative does not change sign in passing through a
critical point, then it is a point of inflexion, which will be discussed in Calculus Chapter 5.
Tutorial questions — Maxima and minima
D 10. A manufacturer wishes to produce cylindrical metal tanks (with lid) having a volume of
6 cubic metres. If the radius of the tank is r metres and the height is h metres, express
h in terms of r. The metal for the curved portion of the tank costs R20 per square
metre, while the metal for the base and lid costs R40 per square metre. Show that the
total cost in rands is 80(πr2 + 3r ), and find the dimensions and cost of the cheapest
tank.
D 11. Find the points on the graph xy 2 = 16 which are nearest the origin. (Hint: take a
general point (x, y) on the graph, and express its distance from the origin in terms of
x alone. Then find the value of x for which the square of the distance has a minimum
value.)

Figure 2.8. Box made from a sheet of tin

D 12. A box without a lid is made by cutting corners from a sheet of tin 160mm × 100mm
and folding up the edges as shown in Figure 2.8. Express the volume in terms of h only
and find the dimensions of the box of maximum volume.
Higher derivatives
We have shown that if y = f (x) then we can differentiate to find the derivative, which is
denoted
dy df (x)
y ′ or f ′ (x) oror .
dx dx
This process of differentiation can in principle be repeated; that is, we can differentiate
the derivative and obtain what is called the second derivative and denoted
d2 y d2 f (x)
y ′′ or f ′′ (x) or or .
dx2 dx2
 Calculus Chapter 2 21

In fact we can differentiate n times and obtain the n-th derivative, denoted

dn y dn f (x)
y (n) or f (n) (x) or or .
dxn dxn
The second derivative is often useful (for example, we shall show later that if f ′′ (x) > 0
at a critical point, then the critical point is a minimum, while if f ′′ (x) < 0, then it is a maxi-
d2 Displacement
mum), and the second derivative has physical significance (for example, =
dTime2
Acceleration). Third and higher derivatives do not often appear in physical formulae, but
are essential in finding accurate approximations to functions.

Tutorial questions — Higher derivatives

D 13. The displacement s metres of a particle at time t seconds is given by the equation
s = 30t − 5t2 . Find the velocity and acceleration at time t. What is the maximum
displacement? What is the initial velocity?
d2 sin x d4 sin x
D 14. Show that = − sin x and deduce that = sin x. Find (quickly) sin(227) x.
dx2 dx4

T 15. Prove by induction that the n-th derivative of xn is n!. (Hint: the (k + 1)-th derivative
is the k-th derivative of the first derivative.)

Answers
1. (a) 2x + 1 (b) − x12 (c) x
(x2 +1)1/2
(d) √ −1√
2 t(2+ t)2
.
√
4 1 9 3 x 1√
2. (a) 4x3 + 6x2 − 5 (b) log10 2 (c) 0 (d) 3t2 − 2t (e) x3 − x2 − x4 (f) 2 + 2x x
.
3. 1, 1, 12 , 2, k.
4. (a) − sin θ (b) sec2 θ (c) 2 sin θ cos θ (d) 2θ cos(θ2 ).
√
5. (a) − cos x − 3 sin x (b) 1 − cos θ.
7. At (0, 0) tangent is y = 3x and normal is y = − 13 x.
At (3, 0) tangent is y = −3x + 9 and normal is y = 31 x − 1.
8. Tangents: y = −4x, No;y = 8x − 16, Yes, at x = −4; y = 8x + 16, Yes, at x = 4.
√ √
Normals: y = 14 x, Yes, at x = ± 12 17; 8y + x − 2 = 0, Yes, at x = −1 ± 41 14;
√
8y + x + 2 = 0, Yes, at x = 1 ± 41 14.
6 3 1/3 6 2π 2/3
10. h = πr√2, r = ( 2π ) , h= π( 3 ) , C = 360( 2π
3 )
1/3
.
11. (2, ±2 2).
12. h = 20, w = 60, l = 120.
ds dv
13. v = dt= 30 − 10t, acceleration = dt = −10, maximum displacement = 45 m, initial
velocity = 30 ms−1 .
d3
14. − cos x = dx3 sin x, since 227 = 56 × 4 + 3.