मंथिकाManthika
Menu
IN REVIEW

Free while we finish the books. A book in review is free to read and use, and so is its PDF. Once the book is finished, its PDF is ₹299.

Undoing the derivative

A water tank on the roof, filled by a pipe at 20 litres a minute. How much water is inside it now?

Many homes have a water tank on the roof. A pipe fills it from below. Say you know how fast the water flows in, in litres a minute. But you cannot see inside the tank.

minutelitres a minute20 litres a minute
The flow, a steady 20 litres a minute for ten minutes. How much water came in? How much is in the tank now?

From the flow alone, can you tell how much water came in during ten minutes? Can you tell how much is in the tank now?

minutelitresW = 100 + 20t
minutelitres a minute20
Chapters 5 and 6 went down this page: from the water in the tank, 100 + 20t litres, to its rate, 20 litres a minute.

Chapters 5 and 6 went one way. You started from a quantity and found its rate of change. Here is the question to carry to the last page. Can you go back, from the rate to the quantity?

100 + 20t
→
d/dt
→
20
20
→
integrate
→
?
Differentiating takes the amount to its rate. INTEGRATION runs the other way, from the rate back to the amount.

Going back from a rate is called INTEGRATION. It is the opposite of a derivative: an ANTIDERIVATIVE. In this chapter you first go back from a rate to a quantity. Then you turn a rate into one number. That number is the total change between two moments. After that come the methods. Each one is a rule from Chapter 5, read backwards.

The tank comes back several times. Keep it in mind.

Kanada
A SCHOLAR INDIA REMEMBERS

Kanada lived in north India between the 6th and the 2nd century BCE. His school, the Vaisheshika, sorted all that exists into kinds. He taught that a large thing is built from very small parts, joined together. Ask the tank Kanada’s question: what is its water made of? Each minute of flow adds one small part. Going back from the rate to the amount means adding those parts up.

Check yourself
  1. What does integrating a function mean, most directly?
    1. finding the equation of the tangent line to the given function
    2. finding a function whose derivative equals the given function
    3. finding the value the given function approaches at a point
    4. finding the derivative of the given function
    Check your answer
    1. finding the equation of the tangent line to the given function — A tangent line is built FROM a derivative at one point; integration goes the other way, recovering the whole function $f$ from $f'$.
    2. ✓ finding a function whose derivative equals the given function — (B) Integration is the reverse of differentiation, start from $f'(x)$ and recover $f(x)$.
    3. finding the value the given function approaches at a point — A limit asks what value a function approaches; integration asks for an integral, two different questions.
    4. finding the derivative of the given function — Differentiating $f$ is the forward operation; integration is its reverse, recovering $f$ from $f'$.
  2. $x^2$ and $x^2+5$ share the same derivative, $2x$. What does this illustrate about integration?
    1. differentiation can never be undone, even up to a constant
    2. $x^2$ and $x^2+5$ must be equal for all $x$
    3. the derivative of a sum is always different from the derivative of each term
    4. a whole family of integrals shares one derivative
    Check your answer
    1. differentiation can never be undone, even up to a constant — Integration IS the reversal, it just returns the whole family $x^2+C$, not a single function; that is not irreversibility.
    2. $x^2$ and $x^2+5$ must be equal for all $x$ — $x^2$ and $x^2+5$ are different functions at every $x$, only their DERIVATIVES agree, not the functions themselves.
    3. the derivative of a sum is always different from the derivative of each term — The sum rule holds fine here, $d/d x (x^2+5) = 2x+0 = 2x$. The example does not challenge the sum rule at all.
    4. ✓ a whole family of integrals shares one derivative — (D) $x^2$ and $x^2+5$ differ only by a constant, and adding a constant never changes the derivative, so undoing $2x$ recovers a whole family, not one function.
  3. How does a definite integral differ from an indefinite integral, as this chapter frames it?
    1. a definite integral and an indefinite integral are two names for the same object
    2. a definite integral needs no integral to be found at all
    3. one number for definite; a family of integrals for indefinite
    4. a definite integral is always the derivative of the indefinite integral
    Check your answer
    1. a definite integral and an indefinite integral are two names for the same object — One is a family of functions; the other is a single number. They are two different things, not two names for one thing.
    2. a definite integral needs no integral to be found at all — The Fundamental Theorem is precisely what links a definite integral back to an indefinite integral, so one is still needed.
    3. ✓ one number for definite; a family of integrals for indefinite — (C) A definite integral turns an integral into one number, an area, a total , while the indefinite integral stays a whole family $F(x)+C$.
    4. a definite integral is always the derivative of the indefinite integral — The Fundamental Theorem connects the two through an integral relationship, not by differentiating one to get the other.

↑ Back to top

Before you start

You have already used a tool like each one below, long before now. Try each one below.

If any of these felt new, read the page named before going on.

↑ Back to top

Integrals come in families

▲CONCEPT
200 / 500
400 / 500
gain = 100
Two tanks on one pipe. In 5 minutes each gains 100 litres: one goes from 100 to 200, the other from 300 to 400.

Go back to the tank. Say the water flows in at a steady 20 litres a minute. After 5 minutes, 100 litres have come in. If the tank held 100 litres at the start, it now holds 200. If it held 300, it now holds 400. In both tanks the water rises at the same rate.

minutelitres100 + 20t300 + 20t
Both tanks rise 20 litres a minute: the same slope, two different lines. The flow alone cannot tell them apart.

So the flow tells you how the amount changes. It does not tell you where it started. Every starting amount fits the same flow.

100 + 20t
→
d/dt
→
20
300 + 20t
→
d/dt
→
20
Two inputs, one output: 100 + 20t and 300 + 20t both differentiate to 20. Each is an INTEGRAL of 20.

A function $F$ with $F'(x) = f(x)$ is an INTEGRAL of $f$. You write the whole family of integrals as

$\int f(x) \mathrm{d} x = F(x) + C$.

This is the INDEFINITE INTEGRAL of $f$. The function $f$ is the INTEGRAND. The number $C$ is the CONSTANT OF INTEGRATION. Each value of $C$ is one starting amount.

minutelitresC = 0C = 100C = 300
20t + C for three starting amounts, C = 0, 100 and 300 litres. Each value of C is one tank.

For the tank, the flow is $f(t) = 20$, so $\int 20 \mathrm{d} t = 20 t + C$. With $C = 100$ you get the first tank. With $C = 300$ you get the second.

y = x²y = x² + 5
difference 5
slope 0 all along
At x = 1, x² and x² + 5 stand 1 and 6 high, both with slope 2. Their difference is 5 everywhere: slope 0.

Why can two integrals of the same function differ only by a constant? Say $F$ and $G$ both have slope $f$ on an interval. Then $F - G$ has slope 0 at every point of it. A function with slope 0 all along an interval is constant there. Chapter 5’s mean value theorem is the reason.

Three curves, x squared plus 0, 2 and 4. At x equals 1.5 they sit 2 apart, and all three slopes are 3.

Let us check it on $x^2$ and $x^2 + 5$. Differentiate each one:

$d/d x (x^2) = 2 x$,

$d/d x (x^2 + 5) = 2 x$.

The two differ by 5 at every $x$, yet both have slope $2 x$. So each is an integral of $2 x$, and the $+ C$ covers the 5:

$\int 2 x \mathrm{d} x = x^2 + C$.

So never drop the $C$. Without it you have one tank, not all of them.

Your turn: differentiate $(x + 1)^2$ and $x^2 + 2 x$. Both are integrals of the same function. How far apart are they?

Four curves x² + C and their one slope graph, 2x. Going down to 2x loses C. Going back up gives the whole family.
Kanada
A SCHOLAR INDIA REMEMBERS

Two tanks on one flow. What makes this tank this tank, and not the other? Both take in $20$ litres a minute, and both rise along $20 t + C$. The one thing that tells them apart is where each one started, $C = 100$ or $C = 300$. It is the one fact the flow cannot give you.

Check yourself
  1. The indefinite integral $\int f(x) d x$ denotes \_\_\_\_\_\_.
    1. the single integral with $C=0$
    2. the derivative of $f$
    3. the area under the graph of $f$ from $0$ to $x$
    4. the family $F(x)+C$ of every integral
    Check your answer
    1. the single integral with $C=0$ — $\int f(x) d x$ names every integral at once, not just the one with $C=0$.
    2. the derivative of $f$ — $\int f(x) d x$ undoes differentiation; it does not compute $f$’s derivative.
    3. the area under the graph of $f$ from $0$ to $x$ — An indefinite integral is a family of functions, with no interval attached, area belongs to the definite integral.
    4. ✓ the family $F(x)+C$ of every integral — (D) $\int f(x) d x$ names the whole family $F(x)+C$, where $F'(x)=f(x)$, not one member of it.
  2. Why do two integrals of the same $f$ differ only by a constant?
    1. because every derivative rule happens to add a fixed number
    2. a zero derivative on an interval forces a constant
    3. because integration always rounds to the nearest whole number
    4. because $f$ itself must be a constant function
    Check your answer
    1. because every derivative rule happens to add a fixed number — No derivative rule adds a fixed number, the constant appears only because differentiating removes it, not because a rule inserts one.
    2. ✓ a zero derivative on an interval forces a constant — (B) If $F$ and $G$ both have derivative $f$, then $(F-G)'=0$ on the interval, so $F-G$ is constant there, this is a direct consequence of the Mean Value Theorem.
    3. because integration always rounds to the nearest whole number — Integration does not round anything; the constant-only difference comes from the Mean Value Theorem, not from rounding.
    4. because $f$ itself must be a constant function — The fact holds for ANY $f$ with an integral, not only for a constant $f$, the Mean Value Theorem argument never assumes that.
  3. $\int \cos x d x = \sin x + C$. Which of these IS a member of this integral family?
    1. $\cos x$
    2. $\sin x - 4$
    3. $2 \cdot \sin x$
    4. $-\sin x$
    Check your answer
    1. $\cos x$ — $\cos x$ is a different function with a different derivative ($-\sin x$), the family only allows adding a constant to $\sin x$, not swapping functions.
    2. ✓ $\sin x - 4$ — (B) $\sin x - 4$ differs from $\sin x$ only by the constant $-4$; differentiating it returns $\cos x$ exactly.
    3. $2 \cdot \sin x$ — $2 \sin x$ has derivative $2 \cos x$, only ADDING a constant preserves the derivative $\cos x$, not multiplying by one.
    4. $-\sin x$ — $-\sin x$ has derivative $-\cos x$, not $\cos x$, flipping the sign is not the same as adding a constant.
  4. Given $f'(x) = \cos x$ and $f(0) = 3$, what determines the value of $C$ in $f(x) = \sin x + C$?
    1. substitute $f(0)=3$ into the family, solve for $C$
    2. setting $C$ equal to the coefficient of $x$ in $f'(x)$
    3. setting $C$ equal to $f'(0)$
    4. differentiating the condition $f(0)=3$
    Check your answer
    1. ✓ substitute $f(0)=3$ into the family, solve for $C$ — (A) $f(0)=\sin(0)+C=C=3$, so the initial condition pins down $C=3$ directly.
    2. setting $C$ equal to the coefficient of $x$ in $f'(x)$ — $\cos x$ has no ’coefficient of $x$’ to extract, $C$ is found by solving the family equation at the given point, not by reading off a coefficient.
    3. setting $C$ equal to $f'(0)$ — $f'(0)=\cos(0)=1$ is a fact about the derivative, not the value of $C$, $C$ comes only from the initial condition on $f$ itself.
    4. differentiating the condition $f(0)=3$ — $f(0)=3$ is a single number, not a function, there is nothing to differentiate; it is substituted into the family instead.

↑ Back to top

The definite integral as a limit of totals

▲CONCEPT
Enrichment. A leaf traced on squared paper. Count the squares inside the outline to estimate its area. Smaller squares come closer.

Counting squares for area

In Class 6 or 7 you may have traced a leaf on squared paper. To find its area, you counted the squares inside the outline. A full square counted as one. A square more than half inside counted as one too, and the rest were left out. With smaller squares, the count came closer to the true area.

1 cm squares
cmcm
half-centimetre squares
cmcm
Enrichment. The same leaf: 1 cm squares count 8 cm², half-centimetre squares 11 cm². The true area is 12 cm².

So you can find an area by adding many thin pieces. The thinner the pieces, the closer the total.

y = x²/4 + 1
Enrichment. Four strips under x²/4 + 1, each 1 wide and as tall as the curve at its left end. Their total is 7.5.

The same idea gives the area under a graph. Let $f$ be continuous on $[a, b]$, with its graph above the axis. Split $[a, b]$ into $n$ equal strips, each of width $h = (b - a)/n$. On each strip, stand a rectangle as tall as $f$ at its left end. Add up the rectangles. As $n$ grows, the strips get thinner and the total settles on one number. That number is the DEFINITE INTEGRAL of $f$ from $a$ to $b$.

$\int_a^b f(x) \mathrm{d} x = \lim_{n \to \infty} h [f(a) + f(a + h) + \cdots + f(a + (n - 1) h)]$.

Under x squared over 4 plus 1: 4, 8 and 16 strips give 7.5, 8.375 and 8.84375. The true area is about 9.33.
nleft totalyyyyy
Enrichment. The totals for 4 to 64 strips: 7.5, 8.375, 8.844, 9.086, 9.209. They settle on 28/3, about 9.333.

The figure does this for $x^2/4 + 1$ on $[0, 4]$. With 4 strips the total is 7.5. With 8 strips it is 8.375, and with 16 it is 8.84375. The totals climb towards $28/3$, about 9.33.

The board exam does not set this section. It shows what a definite integral is. NCERT’s name for it is the limit of a sum. The fundamental theorem, in the next section, gives a much quicker way to find one.

Kanada
A SCHOLAR INDIA REMEMBERS

What are the smallest parts of an area? Ask it of the strips. Is there a thinnest strip, where you can stop? Each finer split gives a new total, and none of them is the last. So the area is not any one of these totals. It is the number they settle on, $28/3$.

Check yourself
  1. Before the Fundamental Theorem shortcut, $\int_a^b f(x) d x$ is defined as \_\_\_\_\_\_.
    1. a limit, $f$ times strip width, summed over the strips
    2. the slope of the line joining $(a,f(a))$ and $(b,f(b))$
    3. the largest value $f$ takes on $[a,b]$
    4. the difference $f(b)-f(a)$
    Check your answer
    1. ✓ a limit, $f$ times strip width, summed over the strips — (A) The definite integral’s actual definition is the Riemann sum’s limit, the integral shortcut (FTC) comes later.
    2. the slope of the line joining $(a,f(a))$ and $(b,f(b))$ — A slope compares two function values over an interval, the definite integral instead adds up infinitely many strip areas.
    3. the largest value $f$ takes on $[a,b]$ — The largest value of $f$ says nothing about the signed area under the whole curve, which is what the Riemann sum accumulates.
    4. the difference $f(b)-f(a)$ — $f(b)-f(a)$ is a fact about $f$’s values, not the definition of the integral, the definition is the Riemann-sum limit, computed by FTC’s shortcut afterward using an INTEGRAL, not $f$ itself.
  2. In the total $h[f(a)+f(a+h)+\cdots+f(a+(n-1)h)]$, what is $h$?
    1. the width of each strip, $(b-a)/n$
    2. the value of $f$ at the right end of the interval
    3. the total number of strips, $n$
    4. the height of the tallest strip
    Check your answer
    1. ✓ the width of each strip, $(b-a)/n$ — (A) $h$ is the strip width, $(b-a)/n$, obtained by splitting $[a,b]$ into $n$ equal pieces.
    2. the value of $f$ at the right end of the interval — $h$ is a length along the $x$-axis, not a value of $f$, $f(a+(n-1)h)$ IS a function value, but $h$ itself is not.
    3. the total number of strips, $n$ — $n$ counts the strips; $h=(b-a)/n$ is each strip’s width, the two shrink and grow in opposite directions as $n$ increases.
    4. the height of the tallest strip — The sum multiplies each $f$ value by the same fixed width $h$, no single strip’s height enters as $h$.
  3. As $n \to \infty$ in the Riemann sum, what happens, and why does it matter?
    1. the sum oscillates forever with no limiting value
    2. $h \to \infty$, so the strips grow wider and the sum becomes less accurate
    3. the number of strips stays fixed while their width shrinks to a set minimum
    4. $h \to 0$, and the strip-sum converges to the exact area
    Check your answer
    1. the sum oscillates forever with no limiting value — For any function this chapter deals with, the Riemann sum converges to a fixed value as $n \to \infty$ , that convergence is exactly what makes the definite integral well-defined.
    2. $h \to \infty$, so the strips grow wider and the sum becomes less accurate — As $n$ grows, $h=(b-a)/n$ SHRINKS toward zero, it does not grow, since more, narrower strips are being added, not fewer, wider ones.
    3. the number of strips stays fixed while their width shrinks to a set minimum — $n \to \infty$ means the strip COUNT itself is growing without bound, it cannot stay fixed while this happens.
    4. ✓ $h \to 0$, and the strip-sum converges to the exact area — (D) More strips means each is narrower, $h=(b-a)/n \to 0$, and in that limit the sum converges to the exact signed area, which is why a definite integral means area in the first place.

↑ Back to top

The fundamental theorem of calculus

▲CONCEPT
minutelitres a minute20 litres a minute
The flow graph, 20 litres a minute. Under it from 0 to 5 minutes, a 20 by 5 rectangle: 100 litres so far.

Go back to the tank, and draw its flow against time. Say the flow is a steady 20 litres a minute. The graph is a flat line at height 20. In 5 minutes, 100 litres come in. That is also the area under the line, a rectangle 20 tall and 5 wide. So the water so far is the area under the flow graph so far.

100 / 5000 minutes
120 / 5001 minute
140 / 5002 minutes
160 / 5003 minutes
The tank minute by minute: 100, 120, 140, 160 litres. Each minute adds one minute’s flow, 20 litres.
minutelitres20t
minutelitres a minute
Left, the water that came in, 20t litres; right, the flow. At 5 minutes the area on the right, 100, is the height on the left.

Now look at the next minute. The tank gains about one minute’s flow. So the water so far grows at the rate of the flow.

xAA(x)
xff = 1 + x²/8
Left, the area function A(x) from a = 1. Right, f, shaded up to x = 3. A(3) ≈ 3.083 is that area.

The same holds for any graph. Let $f$ be continuous on $[a, b]$. The AREA FUNCTION

$A(x) = \int_a^x f(t) \mathrm{d} t$

is the area under $f$ from $a$ up to $x$. The FIRST FUNDAMENTAL THEOREM says that $A'(x) = f(x)$ for every $x$ in $[a, b]$.

The area so far runs from 1 to 3. There the curve is 2.125 tall, so the next 0.5 strip adds about 1.0625.

Here is why. Move $x$ on by a small step $h$. The area gains a thin strip, about $f(x)$ tall and $h$ wide. So

$A(x + h) - A(x) \approx f(x) \cdot h$.

Divide both sides by $h$:

$(A(x + h) - A(x))/h \approx f(x)$.

*Now let $h$* shrink to 0. The left side becomes the derivative of $A$, as in Chapter 5, and the strip fits better and better:

$A'(x) = f(x)$.

In the figure, the curve is 2.125 tall at $x = 3$. A strip 0.5 wide adds about $2.125 \cdot 0.5 = 1.0625$.

x²
2t
The check. Under 2t from 0 to 3 lies a triangle of area half of 3 × 6, which is 9. So is 3².

Check it with $f(t) = 2 t$ from 0. The area up to $x$ is a triangle, $x$ wide and $2 x$ tall. Half of $x$ times $2 x$ gives

$A(x) = x^2$.

Differentiate it, as in Chapter 5:

$A'(x) = 2 x$, which is $f(x)$.

So the area function is an integral of $f$. Finding areas and finding integrals are one job.

Check yourself
  1. For $A(x) = \int_a^x f(t) d t$, what does the first fundamental theorem state?
    1. $A'(x) = f(x)$
    2. $A(x)$ is always equal to zero
    3. $A(x) = f'(x)$
    4. $A'(x) = \int f(x) d x$
    Check your answer
    1. ✓ $A'(x) = f(x)$ — (A) The derivative of the variable-upper-limit integral IS the integrand itself.
    2. $A(x)$ is always equal to zero — $A(x)$ is the running signed area from $a$ to $x$, there is no reason it should be zero except at $x=a$ itself.
    3. $A(x) = f'(x)$ — It is $A'(x)$, the DERIVATIVE of the area function, that equals $f(x)$, not $A(x)$ itself equaling $f'(x)$.
    4. $A'(x) = \int f(x) d x$ — $A'(x)$ is a single function, $f(x)$, re-integrating $f(x)$ on the right-hand side has no place in this statement.
  2. What is $A(x) = \int_a^x f(t) d t$, conceptually?
    1. the running signed area from $a$ up to $x$
    2. the derivative of $f$ at the point $x$
    3. a single fixed number, the total area under $f$ on $[a,b]$
    4. the average value of $f$ over $[a,x]$
    Check your answer
    1. ✓ the running signed area from $a$ up to $x$ — (A) $A(x)$ changes as $x$ moves, it is the accumulated signed area up to that moving point, not a fixed number.
    2. the derivative of $f$ at the point $x$ — $A(x)$ is built by INTEGRATING $f$ from $a$ to $x$, it is $A'(x)$, not $A(x)$ itself, that equals $f(x)$.
    3. a single fixed number, the total area under $f$ on $[a,b]$ — $A(x)$ changes as $x$ changes, it is a FUNCTION of the upper limit, not one fixed number like $\int_a^b f(t) d t$.
    4. the average value of $f$ over $[a,x]$ — $A(x)$ accumulates signed area directly, an average value would need dividing by the interval length, which $A(x)$ never does.
  3. Why is the first fundamental theorem the deep reason integration and differentiation are inverse operations?
    1. because it only works for polynomial functions
    2. because it proves every function is differentiable
    3. because it only applies when $f$ is a constant function
    4. because it holds for whole functions, not one point
    Check your answer
    1. because it only works for polynomial functions — The first fundamental theorem holds for any continuous function $f$, not only polynomials, the theorem places no such restriction.
    2. because it proves every function is differentiable — The first fundamental theorem shows $A(x)$ specifically is differentiable with derivative $f(x)$, it says nothing about whether every function whatsoever is differentiable.
    3. because it only applies when $f$ is a constant function — The first fundamental theorem holds for any continuous $f$, restricting it to constant functions strips away exactly the generality that makes it the deep reason for the inverse relationship.
    4. ✓ because it holds for whole functions, not one point — (D) $A'(x)=f(x)$ shows that differentiating the accumulated area exactly recovers the original function, the inverse relationship at the level of whole functions, not just one formula.
▲CONCEPT
The meter keeps a running total. At 6:00 it reads 1150 litres, at 6:10 it reads 1350. How much came in?

Many homes also have a water meter on the pipe. It keeps a running total of the litres that have passed. Say it reads 1150 litres at 6:00 and 1350 litres at 6:10. Then 200 litres came in between 6:00 and 6:10. You did not add up the flow minute by minute. Two readings were enough. And it does not matter how long ago the meter started counting.

200 litres
1150
1350
Two readings of the running total, 1150 and 1350 litres. The water between 6:00 and 6:10 is the difference, 200 litres.

So the total change over a stretch is the reading at the end minus the reading at the start.

minutes after 6:00litres
minutes after 6:00litres a minute
Left, the meter’s running total, 1150 to 1350 litres. Right, the flow. The area under the flow, 200, is the rise on the left.

The SECOND FUNDAMENTAL THEOREM says this in symbols. Let $f$ be continuous on $[a, b]$, and let $F$ be any integral of $f$. Then

$\int_a^b f(x) \mathrm{d} x = F(b) - F(a)$.

The short way to write $F(b) - F(a)$ is $[F(x)]_a^b$. Work out $F$ at the top limit, then take away its value at the bottom limit. So

$\int_a^b f(x) \mathrm{d} x = [F(x)]_a^b$.

y = x²y = x² + 5
From 0 to 3, x² climbs from 0 to 9 and x² + 5 from 5 to 14. Both rise 9: the constant cancels.

The constant cancels. With $F + C$ in place of $F$, you get

$(F(b) + C) - (F(a) + C) = F(b) - F(a)$.

So a definite integral is one number, with no $+ C$.

For the tank, the flow is 20 litres a minute, and $F(t) = 20 t + C$. Put in the end time 10 and the start time 0, and subtract:

$F(10) - F(0) = (200 + C) - (0 + C)$.

*The $C$ cancels* again:

$F(10) - F(0) = 200$.

That is the 200 litres on the meter.

Let us find $\int_0^3 2 x \mathrm{d} x$ with two different integrals.

Worked example

Find $\int_0^3 2 x \mathrm{d} x$ with two integrals

  1. $F(x) = x^2$
    An integral of $2 x$, since the derivative of $x^2$ is $2 x$.
  2. $\int_0^3 2 x \mathrm{d} x = F(3) - F(0)$
    The second fundamental theorem: the value at the top limit minus the value at the bottom.
  3. *$F(3) - F(0) = 9 - 0 = 9$*
    Work out each value, then subtract.
  4. $(9 + 5) - (0 + 5) = 14 - 5 = 9$
    Check with $F(x) = x^2 + 5$: the 5 cancels, and the answer is 9 again.
Left: the region under an arch. Right: 3x squared minus 2x from 1 to 2, heights 1 and 8. Its integral is 4.

So a definite integral needs one integral, $F$, and two values of it.

Your turn: find $\int_0^3 (2 x - 2) \mathrm{d} x$. Then sketch $y = 2 x - 2$. Where is the graph below the axis?

Check yourself
  1. The second fundamental theorem states that for any integral $F$ of $f$ on $[a,b]$, $\int_a^b f(x) d x$ equals \_\_\_\_\_\_.
    1. $F(a)-F(b)$
    2. $F(b)+F(a)$
    3. $F'(b)-F'(a)$
    4. $F(b)-F(a)$
    Check your answer
    1. $F(a)-F(b)$ — The formula is $F(b)-F(a)$, the UPPER limit’s value minus the lower, not the reverse.
    2. $F(b)+F(a)$ — The second fundamental theorem SUBTRACTS the lower-limit value from the upper-limit value, adding them is a different, wrong operation.
    3. $F'(b)-F'(a)$ — The second fundamental theorem evaluates the INTEGRAL $F$ at the endpoints, differentiating $F$ again would just return $f$, not the integral’s value.
    4. ✓ $F(b)-F(a)$ — (D) The second fundamental theorem gives the value as the integral evaluated at the upper endpoint minus the lower endpoint.
  2. What role does the second fundamental theorem play in this chapter?
    1. it proves that every function has an integral
    2. it is only used for definite integrals evaluated by substitution
    3. it bridges the limit-of-totals definition to a shortcut that skips adding the strips
    4. it defines the definite integral for the first time, replacing the limit of totals entirely
    Check your answer
    1. it proves that every function has an integral — The second fundamental theorem assumes an integral $F$ already exists and uses it, it says nothing about whether one exists for an arbitrary $f$.
    2. it is only used for definite integrals evaluated by substitution — The second fundamental theorem applies to any definite integral once an integral is found, by whatever method, it is not restricted to substitution-based ones.
    3. ✓ it bridges the limit-of-totals definition to a shortcut that skips adding the strips — (C) The second fundamental theorem is the bridge from the Riemann-sum definition to a practical shortcut , no limiting sum is needed once an integral is known.
    4. it defines the definite integral for the first time, replacing the limit of totals entirely — The Riemann sum remains the actual DEFINITION, the second fundamental theorem gives a faster way to COMPUTE the value it defines, not a new definition.
  3. Why does the constant of integration never appear in a definite-integral answer?
    1. because definite integrals never use an integral $F$ in the first place
    2. because $(F(b)+C)-(F(a)+C) = F(b)-F(a)$: the $C$ always cancels
    3. because $C$ must be set to zero before applying FTC
    4. because the Riemann sum removes $C$ automatically before FTC is used
    Check your answer
    1. because definite integrals never use an integral $F$ in the first place — The second fundamental theorem depends entirely on having an integral $F$, the claim that none is used contradicts the formula itself.
    2. ✓ because $(F(b)+C)-(F(a)+C) = F(b)-F(a)$: the $C$ always cancels — (B) Adding the same $C$ to both endpoint values and then subtracting cancels it out completely, regardless of what $C$ is.
    3. because $C$ must be set to zero before applying FTC — No such requirement exists, $(F(b)+C)-(F(a)+C)$ cancels $C$ for ANY value, so it need not be set to zero first.
    4. because the Riemann sum removes $C$ automatically before FTC is used — The Riemann sum is built from $f$ directly and has no constant of integration to remove, the cancellation happens inside FTC’s own subtraction, using an integral of $f$.
  4. To evaluate $\int_1^2 (3x^2-2x) d x$ using the second fundamental theorem, what is the first step?
    1. substitute $x=2$ and $x=1$ directly into $3x^2-2x$ without integrating
    2. find any one integral, such as $F(x)=x^3-x^2$
    3. differentiate $3x^2-2x$ to get a simpler integrand
    4. set up the Riemann sum and take the limit as $n \to \infty$
    Check your answer
    1. substitute $x=2$ and $x=1$ directly into $3x^2-2x$ without integrating — The second fundamental theorem evaluates an INTEGRAL $F$ at the endpoints, not the original integrand $f$ itself , this step skips finding $F$ altogether.
    2. ✓ find any one integral, such as $F(x)=x^3-x^2$ — (B) The second fundamental theorem’s first step is finding any single integral of the integrand.
    3. differentiate $3x^2-2x$ to get a simpler integrand — The second fundamental theorem needs an INTEGRAL of the integrand, differentiating it further moves away from, not toward, that goal.
    4. set up the Riemann sum and take the limit as $n \to \infty$ — The second fundamental theorem exists precisely to skip the limiting Riemann sum, reverting to it here abandons the shortcut entirely.
F = x³ − x²
3x² − 2x
Left, x³ − x² climbs from 0 to 4 between 1 and 2. Right, the area under 3x² − 2x is that rise, 4.
F
F + 7
The check with F + 7: the readings are 7 and 11. The difference is still 4.
Worked example

A definite integral via the Fundamental Theorem: $\int_1^2 (3x^2-2x) \mathrm{d} x$

  1. Find $\int_1^2 (3x^2 - 2x) \mathrm{d} x$
    The integrand is continuous on $[1, 2]$, so the second fundamental theorem applies. First find one integral.
  2. $F(x) = x^3 - x^2$
    Differentiate to test it. $F'(x) = 3 x^2 - 2 x$, the integrand.
  3. $\int_1^2 (3x^2-2x) \mathrm{d} x = F(2) - F(1)$
    Subtract the value at the lower limit from the value at the upper limit.
  4. $F(2)-F(1) = (8-4)-(1-1) = 4-0 = 4$
    Work out each value, then subtract.
  5. $\int_1^2 (3x^2-2x) \mathrm{d} x = 4$
    One number, with no $+ C$.
  6. $F(x) = x^3 - x^2 + 7$ gives $(8 - 4 + 7) - (1 - 1 + 7) = 11 - 7 = 4$
    Check with another integral. The 7 cancels, and the answer is 4 again.
Worked example

Find the reversible isothermal work done by an ideal gas, using the Fundamental Theorem

  1. $w_\text{rev} = -\int_{V_i}^{V_f} p \mathrm{d} V$
    Chemistry 11, Chapter 5, Thermodynamics, defines reversible work this way, the gas expanding or compressing against pressure $p$ from volume $V_i$ to volume $V_f$.
  2. for $n$ mol of an ideal gas at constant temperature $T$, $p = (n R T)/V$
    The ideal gas law, with temperature fixed throughout this isothermal process.
  3. $w_\text{rev} = -n R T \int_{V_i}^{V_f} 1/V \mathrm{d} V$
    Substitute $p$; $n$, $R$ and $T$ are all constants, so they come straight out of the integral.
  4. $F(V) = \ln V$
    The integral of $1/V$ is $\ln V$, as $1/x$ gives $\ln x$ in Chapter 5; the same pattern as the FTC example above; no substitution is needed here.
  5. $\int_{V_i}^{V_f} 1/V \mathrm{d} V = F(V_f)-F(V_i) = \ln V_f - \ln V_i$
    The second fundamental theorem.
  6. *$w_\text{rev} = -n R T (\ln V_f - \ln V_i) = -n R T \ln(V_f/V_i)$*
    Combine the two logs into one, matching the formula Chemistry 11 already states.
Check yourself
  1. For $\int_1^2 (3x^2-2x) d x$, an integral is $F(x)$ equal to \_\_\_\_\_\_.
    1. $x^2 - x$
    2. $6x - 2$
    3. $x^3 - x^2$
    4. $x^3 - x^2 + x$
    Check your answer
    1. $x^2 - x$ — Integrating $3x^2$ raises the exponent to $x^3$, not down to $x^2$, the power rule for integration always raises the exponent.
    2. $6x - 2$ — $6x-2$ is the DERIVATIVE of $3x^2-2x$, an integral needs to move the other way, raising powers, not lowering them.
    3. ✓ $x^3 - x^2$ — (C) $F(x)=x^3-x^2$ has $F'(x)=3x^2-2x$, matching the integrand exactly.
    4. $x^3 - x^2 + x$ — The derivative of $x^3 - x^2 + x$ is $3x^2 - 2x + 1$, one more than the integrand. Differentiate back to check.
  2. With $F(x)=x^3-x^2$, $F(2)-F(1)$ equals \_\_\_\_\_\_.
    1. $4$, since $F(2)=4$ and $F(1)=0$
    2. $-4$, subtracting in the reverse order
    3. $12$, since $F(2)=8$ and $F(1)=-4$
    4. $8$, using only $F(2)$
    Check your answer
    1. ✓ $4$, since $F(2)=4$ and $F(1)=0$ — (A) $F(2)=8-4=4$ and $F(1)=1-1=0$, so $F(2)-F(1)=4-0=4$.
    2. $-4$, subtracting in the reverse order — The second fundamental theorem subtracts the LOWER limit’s value from the upper, $F(2)-F(1)$, not the reverse order.
    3. $12$, since $F(2)=8$ and $F(1)=-4$ — $F(2)=2^3-2^2=8-4=4$, not $8$, and $F(1)=1^3-1^2=1-1=0$, not $-4$, both endpoint values here are wrong.
    4. $8$, using only $F(2)$ — The second fundamental theorem needs BOTH endpoint values subtracted, $F(2)$ alone is only half of the computation.
  3. Why does this worked example need no limiting Riemann sum?
    1. because definite integrals never use Riemann sums, only FTC
    2. because the interval $[1,2]$ is too short for a Riemann sum to apply
    3. The second fundamental theorem gives the value directly, from $F$
    4. because $3x^2-2x$ has no valid Riemann sum
    Check your answer
    1. because definite integrals never use Riemann sums, only FTC — The Riemann sum stays the DEFINITION of the definite integral, the second fundamental theorem gives a shortcut for computing it once an integral is known, not a replacement definition.
    2. because the interval $[1,2]$ is too short for a Riemann sum to apply — Riemann sums apply to intervals of any length, there is no minimum length requirement in the theory.
    3. ✓ The second fundamental theorem gives the value directly, from $F$ — (C) The second fundamental theorem’s whole purpose is to skip the limiting sum once an integral is known.
    4. because $3x^2-2x$ has no valid Riemann sum — $3x^2-2x$ is continuous on $[1,2]$, so its Riemann sum exists and converges perfectly well, the second fundamental theorem is simply a faster route to the same value.
✕MISCONCEPTION
Working
∫ from −1 to 2 of x dx
2 − 0.5 = 1.5
So
so the area is
1.5 ?
The working for the area under y = x from −1 to 2. Which step is wrong? Keep your answer, then look at the graph.

Here is a working for the area between $y = x$ and the $x$-axis, from $x = -1$ to $x = 2$. Which step is wrong?

y equals x from minus 1 to 2. The piece below the axis is 0.5, and the piece above is 2. Integral 1.5, area 2.5.

Keep your answer, then look at the graph. On $[-1, 0]$ the line is below the axis. That piece has area 0.5, but the integral counts it as $-0.5$. On $[0, 2]$ the line is above the axis, with area 2. The integral adds the two pieces with their signs:

$-0.5 + 2 = 1.5$.

The area adds their sizes:

$0.5 + 2 = 2.5$.

The second step is wrong. A definite integral counts a piece below the axis as negative. In the tank, water flowing out counts against the water flowing in.

y = x² − 1
Split at the crossing, x = 1. The pieces are −2/3 and 4/3: the integral is 2/3, the area 2/3 + 4/3 = 2.

To find an area, split the interval at every point where the graph crosses the axis. Then add the sizes of the pieces.

The same tank, tap open: 20 litres a minute in, 10 out, shown together; the graph below runs them one after the other.
minutelitres a minute
20 litres a minute in for 5 minutes, then 10 a minute out for 3. The tank gains 100 − 30 = 70 litres.
Kanada
A SCHOLAR INDIA REMEMBERS

You called the integral and the area, $1.5$ and $2.5$ from the last page, both “the area”. Are they the same kind of thing? The integral is a signed total. It counts the piece below the axis as $-0.5$, as the tank counts water going out. An area is a size, and a size is never below $0$. So on $[-1, 2]$ the integral is $1.5$, but the area is $2.5$. Before you answer, ask which kind of number the question wants.

Check yourself
  1. The claim ‘the definite integral $\int_a^b f(x) d x$ IS the area under the curve’ is \_\_\_\_\_\_.
    1. false: it can be negative or under-report area
    2. true for every function, with absolutely no exceptions
    3. false only when $a$ and $b$ are both negative numbers
    4. true whenever $f$ is a polynomial
    Check your answer
    1. ✓ false: it can be negative or under-report area — (A) The claim fails whenever $f$ is negative on part of $[a, b]$: where it crosses zero, or where it stays below the axis. There the integral is negative or under-reports the area.
    2. true for every function, with absolutely no exceptions — The claim fails whenever $f$ is negative on part of $[a, b]$: where it crosses zero, or where it stays below the axis. Saying there is no exception is the misconception itself.
    3. false only when $a$ and $b$ are both negative numbers — Whether $a$ and $b$ are negative has nothing to do with it. What matters is whether $f$ itself is negative somewhere in $[a, b]$.
    4. true whenever $f$ is a polynomial — Being a polynomial does not stop a function from crossing zero, $f(x)=x$ is a polynomial and is exactly the example where the claim fails.
  2. $\int_{-1}^2 x d x = 1.5$. What is the true AREA between $y=x$ and the $x$-axis on $[-1,2]$?
    1. $-1.5$, the negative of the raw integral
    2. $3.5$, found by adding the raw integral to the interval length
    3. $2.5$, splitting at $x=0$ and adding magnitudes
    4. $1.5$, the same as the raw integral
    Check your answer
    1. $-1.5$, the negative of the raw integral — Flipping the sign to $-1.5$ still treats the whole interval as one signed piece, the true area requires splitting at $x=0$ first.
    2. $3.5$, found by adding the raw integral to the interval length — The interval’s length, $3$, plays no role in finding the true area, the correct fix is splitting at the zero-crossing, not adding lengths.
    3. ✓ $2.5$, splitting at $x=0$ and adding magnitudes — (C) $f(x)=x$ is negative on $[-1,0]$ and positive on $[0,2]$; splitting there and adding magnitudes gives $0.5+2=2.5$.
    4. $1.5$, the same as the raw integral — $1.5$ is the raw SIGNED integral, where the negative part on $[-1,0]$ partly cancels the positive part on $[0,2]$, the true area needs those parts added as magnitudes instead.
  3. Why does $f$ crossing zero inside $[a,b]$ break the ’integral equals area’ claim?
    1. the negative and positive parts cancel each other by subtraction
    2. because $f$ must be continuous for the integral to exist, and crossing zero breaks continuity
    3. because the integral becomes undefined at the crossing point
    4. it does not break the claim: crossing zero has no effect
    Check your answer
    1. ✓ the negative and positive parts cancel each other by subtraction — (A) The signed integral lets a negative region partly cancel a positive one, while true area always adds every region as a positive contribution.
    2. because $f$ must be continuous for the integral to exist, and crossing zero breaks continuity — $f(x)=x$ is continuous everywhere, including at its zero crossing, crossing zero and losing continuity are unrelated ideas.
    3. because the integral becomes undefined at the crossing point — $\int_{-1}^2 x d x$ is perfectly well-defined through $x=0$, a zero crossing causes no breakdown in the integral’s existence.
    4. it does not break the claim: crossing zero has no effect — Crossing zero is exactly the condition under which the raw integral and the true area start to differ , denying any effect contradicts the worked example itself.
  4. To find the true area when $f$ crosses zero inside $[a,b]$, the fix is to \_\_\_\_\_\_.
    1. multiply the raw integral by $2$ to account for the crossing
    2. split the interval at each sign crossing, then add up the magnitudes
    3. evaluate $F(b)-F(a)$ as usual; the crossing does not change the procedure
    4. take the absolute value of the final answer $F(b) - F(a)$
    Check your answer
    1. multiply the raw integral by $2$ to account for the crossing — There is no fixed factor that repairs the cancellation in general, the fix depends on splitting at the actual crossing points, not on a universal multiplier.
    2. ✓ split the interval at each sign crossing, then add up the magnitudes — (B) Splitting at every zero-crossing and adding the magnitude of each piece turns the signed integral into the true, always-positive area.
    3. evaluate $F(b)-F(a)$ as usual; the crossing does not change the procedure — The plain $F(b)-F(a)$ evaluation is the raw signed integral that under-reports area when $f$ crosses zero, that is exactly the gap the fix must close.
    4. take the absolute value of the final answer $F(b) - F(a)$ — By the time you have $F(b) - F(a)$, the positive and negative pieces have already cancelled. Its absolute value is still too small: split first, then add the sizes.

↑ Back to top

The standard integrals table

▲CONCEPT

The Fundamental Theorem turned an area into one job: find an integral. This table does that job for the functions you know from Chapter 5. Three methods follow to do the rest. Substitution is the chain rule read backwards. By parts is the product rule read backwards. Partial fractions is adding fractions, undone. Between them sits a table of nine formulas found by those methods, kept for lookup.

sin x
→
d/dx
→
cos x
cos x
→
integrate
→
sin x + C
A Chapter 5 derivative read backwards. sin x differentiates to cos x, so the integral of cos x is sin x + C.

Each formula in this table is a Chapter 5 derivative, read from right to left. You know that

$d/d x (\sin x) = \cos x$.

*So $\sin x$* is an integral of $\cos x$:

$\int \cos x \mathrm{d} x = \sin x + C$.

y = x³/3
xy′y′ = x²
The power rule read backwards at n = 2. The slope of x³/3 is x², so the integral of x² is x³/3 + C.

The power rule needs one condition. For $n \neq -1$,

$\int x^n \mathrm{d} x = x^{n+1}/(n+1) + C$.

xy′y′ = 1/x
For x ≠ 0, ln|x| has slope 1/x on both sides of 0. So the integral of 1/x is ln|x| + C.

At $n = -1$ this would divide by 0. That case has its own formula. For $x \neq 0$, $\int 1/x \mathrm{d} x = \ln|x| + C$.

eˣ
→
d/dx
→
eˣ
−cos x
→
d/dx
→
sin x
tan x
→
d/dx
→
sec² x
Three more formulas, each a derivative read backwards: eˣ gives eˣ, −cos x gives sin x, and tan x gives sec² x.

The other formulas come the same way. You do not learn this table now. Each line is one Chapter 5 derivative read backwards, and you look it up when you need it.

You know, from Chapter 5So
$d/d x (e^x) = e^x$$\int e^x \mathrm{d} x = e^x + C$
$d/d x (a^x) = a^x \ln a$$\int a^x \mathrm{d} x = a^x/(\ln a) + C$, for $a > 0$, $a \neq 1$
$d/d x (-\cos x) = \sin x$$\int \sin x \mathrm{d} x = -\cos x + C$
$d/d x (\sin x) = \cos x$$\int \cos x \mathrm{d} x = \sin x + C$
$d/d x (\tan x) = \sec^2 x$$\int \sec^2 x \mathrm{d} x = \tan x + C$
$d/d x (-\cot x) = \text{cosec}^2 x$$\int \text{cosec}^2 x \mathrm{d} x = -\cot x + C$
$d/d x (\sec x) = \sec x \tan x$$\int \sec x \tan x \mathrm{d} x = \sec x + C$
$d/d x (-\text{cosec} x) = \text{cosec} x \cot x$$\int \text{cosec} x \cot x \mathrm{d} x = -\text{cosec} x + C$
$d/d x (\sec^{-1} x) = 1/(x \sqrt{x^2 - 1})$$\int 1/(x \sqrt{x^2 - 1}) \mathrm{d} x = \sec^{-1} x + C$, for $x > 1$
y = x² + sin x
xy′y′ = 2x + cos x
Two formulas used together. x² + sin x has slope 2x + cos x, so that integral is x² + sin x + C.

Two rules let you use the formulas together. The integral of a sum is the sum of the two integrals. A constant factor comes outside, so $\int k f(x) \mathrm{d} x = k \int f(x) \mathrm{d} x$.

−cot x
→
d/dx
→
cosec² x
The check. Differentiate −cot x and you get cosec² x, the integrand, so the formula is right.

Check one formula by differentiating. Take $\int \text{cosec}^2 x \mathrm{d} x = -\cot x + C$. The derivative of $\cot x$ is $-\text{cosec}^2 x$, so

$d/d x (-\cot x) = -(-\text{cosec}^2 x)$.

Two minus signs make a plus:

$d/d x (-\cot x) = \text{cosec}^2 x$.

That is the integrand, so the formula is right.

So you can check every answer the same way. Differentiate it, and you should get back the integrand.

Check yourself
  1. For $n \neq -1$, $\int x^n d x$ equals \_\_\_\_\_\_.
    1. $x^{n+1}/(n+1) + C$
    2. $x^n/n + C$
    3. $n x^{n-1} + C$
    4. $x^{n-1}/(n-1) + C$
    Check your answer
    1. ✓ $x^{n+1}/(n+1) + C$ — (A) The power rule for integration raises the exponent by one and divides by the new exponent.
    2. $x^n/n + C$ — The exponent must be raised to $n+1$ before dividing, dividing by the ORIGINAL exponent $n$ leaves the power unchanged, which is wrong.
    3. $n x^{n-1} + C$ — $n x^{n-1}$ is the DERIVATIVE of $x^n$; integrating goes the other way, raising the exponent, not lowering it.
    4. $x^{n-1}/(n-1) + C$ — Integration raises the exponent by one, to $n+1$; lowering it to $n-1$ moves in the wrong direction.
  2. Why does the power rule $\int x^n d x = x^{n+1}/(n+1)+C$ exclude $n=-1$?
    1. the power rule only ever applies to positive integers
    2. $n=-1$ makes the integrand equal to zero
    3. $x^{-1}$ has no integral at all
    4. at $n=-1$, the formula denominator $n+1$ becomes zero
    Check your answer
    1. the power rule only ever applies to positive integers — The power rule applies to any $n \neq -1$, including negative and fractional exponents, the only excluded value is $-1$.
    2. $n=-1$ makes the integrand equal to zero — $1/x$ is not zero for any nonzero $x$, the problem is a division by zero in the FORMULA, not the integrand vanishing.
    3. $x^{-1}$ has no integral at all — $1/x$ does have an integral, $\ln|x|+C$, it is just not reached by the power-rule formula itself.
    4. ✓ at $n=-1$, the formula denominator $n+1$ becomes zero — (D) $n=-1$ would make the denominator $n+1=0$; the case is handled separately by $\int 1/x d x = \ln|x|+C$.
  3. $\int e^x d x$ equals \_\_\_\_\_\_.
    1. $x e^{x-1} + C$
    2. $e^x + C$
    3. $e^x/x + C$
    4. $x e^x + C$
    Check your answer
    1. $x e^{x-1} + C$ — $e^x$ is an exponential, not a power of $x$, the power rule’s exponent-shifting has no application here.
    2. ✓ $e^x + C$ — (B) $e^x$ is its own integral: integrating it leaves it unchanged, plus $C$.
    3. $e^x/x + C$ — Nothing in $\int e^x d x$ introduces a division by $x$, the integral is simply $e^x$ itself.
    4. $x e^x + C$ — Nothing in $\int e^x d x$ introduces a factor of $x$, the integral is $e^x$ unchanged, not $x$ times it.
  4. $\int \sec^2 x d x$ equals \_\_\_\_\_\_.
    1. $\sec x + C$
    2. $-\cot x + C$
    3. $\tan x + C$
    4. $-\tan x + C$
    Check your answer
    1. $\sec x + C$ — $\sec x + C$ is the integral of $\sec x \tan x$, a different standard integral from $\sec^2 x$.
    2. $-\cot x + C$ — $-\cot x+C$ is the integral of $\csc^2 x$, a different table entry from $\sec^2 x$.
    3. ✓ $\tan x + C$ — (C) $\sec^2 x$ is exactly the derivative of $\tan x$, so $\tan x + C$ is the integral.
    4. $-\tan x + C$ — $\tan x$, not $-\tan x$, differentiates to $\sec^2 x$, the sign here is flipped from the correct entry.
  5. $\int (x^3+2x)/x d x$ needs which step before a table entry applies?
    1. split into $x^2+2$, then integrate term by term
    2. recognizing it as $\int e^x d x$ in disguise
    3. applying the power rule to the whole fraction at once, exponent and all
    4. recognizing it directly as $\int 1/x d x$
    Check your answer
    1. ✓ split into $x^2+2$, then integrate term by term — (A) $(x^3+2x)/x = x^2+2$ once the fraction is split; each piece is then a direct power-rule entry.
    2. recognizing it as $\int e^x d x$ in disguise — There is no exponential anywhere in $(x^3+2x)/x$, splitting it gives $x^2+2$, a polynomial, not $e^x$.
    3. applying the power rule to the whole fraction at once, exponent and all — The power rule applies term by term to a sum of powers, not to a fraction taken as one block, the fraction must be split first.
    4. recognizing it directly as $\int 1/x d x$ — The fraction is not $1/x$, it is $x^2+2$ once split, and neither piece matches the $1/x$ exception.

↑ Back to top

Integration by substitution

▲CONCEPT
x
→
x²
slope 2x
→
sin
slope cos t
→
sin(x²)
slope cos(x²) × 2x
A composite function: x gives t = x², then t gives sin t. Their slopes multiply, cos(x²) × 2x.

The chain rule from Chapter 5 is for a composite function, one function applied to the output of another. Their slopes multiply. So if $F' = f$, then

$d/d x (F(g(x))) = f(g(x)) \cdot g'(x)$.

Read it backwards: $F(g(x))$ is an integral of the right side. That is INTEGRATION BY SUBSTITUTION.

$\int f(g(x)) \cdot g'(x) \mathrm{d} x = F(g(x)) + C$.

To use it, put

$t = g(x)$.

Differentiate both sides with respect to $x$:

$d t/d x = g'(x)$.

Then write it with $\mathrm{d} t$ and $\mathrm{d} x$ apart. This is shorthand: wherever the integral has $g'(x) \mathrm{d} x$, write $\mathrm{d} t$. The chain rule above is what allows it.

$\mathrm{d} t = g'(x) \mathrm{d} x$.

So the integral becomes $\int f(t) \mathrm{d} t$. Integrate in $t$. Then put $g(x)$ back in place of $t$.

So look for a function whose derivative also appears in the integrand. It may appear there times a constant, and that is fine.

x
t
Put t = x², and each x has its t. Work the integral in t; at the end, put x² back.
y = sin(x²)
xy′y′ = 2x cos(x²)
sin(x²) and its slope 2x cos(x²). The right-hand curve is the integrand and the left-hand one the answer, so the check holds.

Let us try $\int 2 x \cos(x^2) \mathrm{d} x$, one line at a time.

Worked example

Find $\int 2 x \cos(x^2) \mathrm{d} x$ by substitution

  1. $t = x^2$
    The derivative of $x^2$ is $2 x$, and $2 x$ appears in the integrand.
  2. $\mathrm{d} t = 2 x \mathrm{d} x$
    Differentiate $t = x^2$, then write $\mathrm{d} t$ in place of $2 x \mathrm{d} x$.
  3. $\int 2 x \cos(x^2) \mathrm{d} x = \int \cos t \mathrm{d} t$
    Put in $t$ and $\mathrm{d} t$: now the integral is in $t$ alone.
  4. $\int \cos t \mathrm{d} t = \sin t + C$
    The table gives the integral of $\cos t$.
  5. *$\int 2 x \cos(x^2) \mathrm{d} x = \sin(x^2) + C$*
    Put $x^2$ back in place of $t$, so the answer is in $x$.
  6. $d/d x (\sin(x^2)) = \cos(x^2) \cdot 2 x$
    Check by differentiating with the chain rule: the integrand comes back.
Left in t
sin t + C
In x
sin(x²) + C
Left in t, the answer is not finished. Put x² back in place of t: sin(x²) + C.

The answer must be in $x$, the variable you started with. An answer left in $t$ is not finished.

Your turn: find $\int 2 x (x^2 + 3)^4 \mathrm{d} x$. Which $t$ turns it into a power of $t$?

Check yourself
  1. Substitution works by setting $t=g(x)$ to turn $\int f(g(x)) \cdot g'(x) d x$ into \_\_\_\_\_\_.
    1. $\int f(x) d t$
    2. $\int f(t) d x$
    3. $\int f(t) d t$
    4. $\int f(g(t)) d t$
    Check your answer
    1. $\int f(x) d t$ — Substitution replaces $x$ and $d x$ TOGETHER, leaving $f(x)$ while switching only the differential mixes the two variables.
    2. $\int f(t) d x$ — If $f$ is written in terms of $t$, the differential must be $d t$ too, mixing $f(t)$ with $d x$ is inconsistent.
    3. ✓ $\int f(t) d t$ — (C) Every $x$ and $d x$ in sight is replaced by $t$ and $d t$, the whole integral moves into the new variable.
    4. $\int f(g(t)) d t$ — Substitution should collapse $f(g(x))$ down to plain $f(t)$, leaving $g$ nested inside $f$ means the chain-rule pattern was never actually undone.
  2. What should you look for in the integrand to spot a good substitution $t=g(x)$?
    1. any function raised to a power
    2. the largest term in the integrand
    3. a function whose own derivative also appears
    4. a function that is already in the standard-integrals table
    Check your answer
    1. any function raised to a power — A power alone is not the signal, substitution needs the DERIVATIVE of that power’s inside function to also be present.
    2. the largest term in the integrand — Size has nothing to do with it, the signal is a function paired with its own derivative, regardless of which term looks largest.
    3. ✓ a function whose own derivative also appears — (C) The pattern that collapses under substitution is a function paired with its own derivative sitting in the same integrand.
    4. a function that is already in the standard-integrals table — Substitution is needed precisely when the integrand is NOT already a table entry , if it already matches, no substitution is required.
  3. After integrating $\int \sqrt{t} d t$ during a substitution, what must happen before the answer is complete?
    1. $t$ must be renamed $x$ without changing its value
    2. $t$ is replaced by $g(x)$, back in terms of $x$
    3. the constant of integration must be dropped
    4. nothing further: the answer in $t$ is already final
    Check your answer
    1. $t$ must be renamed $x$ without changing its value — $t$ is not simply renamed $x$, it must be replaced by the actual expression $g(x)$ that it stood for.
    2. ✓ $t$ is replaced by $g(x)$, back in terms of $x$ — (B) The original problem was posed in $x$, so the final integral must be substituted back into $x$, not left in the substitution variable.
    3. the constant of integration must be dropped — This is an indefinite integral, the $+C$ stays through the whole process, including after substituting back.
    4. nothing further: the answer in $t$ is already final — An answer left in $t$ answers a different, substituted problem, it must be converted back using $t=g(x)$ to answer the original one.
  4. $\int 2x \cos(x^2) d x$ needs substitution, but $\int \cos x d x$ does not. What is the difference?
    1. the first integral has a larger coefficient than the second
    2. the first integral uses degrees and the second uses radians
    3. the first has a composed function times its own derivative
    4. the two integrals are actually identical in difficulty
    Check your answer
    1. the first integral has a larger coefficient than the second — $2x$ is not just a coefficient, it is the derivative of $x^2$, which is the actual reason substitution helps here.
    2. the first integral uses degrees and the second uses radians — Neither integral depends on degrees versus radians, the real difference is a composed function paired with its inside’s derivative.
    3. ✓ the first has a composed function times its own derivative — (C) $\cos(x^2)$ is a composed function, and $2x$ is exactly the derivative of its inside, $x^2$, that pairing is the substitution signal; plain $\cos x$ has no such composition to undo.
    4. the two integrals are actually identical in difficulty — The two integrals are structurally different, one matches the table directly, the other needs the chain-rule pattern undone first.
x
t
Put t = x² + 1: x = 0, 1, 2 give t = 1, 2, 5. Then dt stands for 2x dx.
y = (2/3)(x² + 1)√(x² + 1)
xy′y′ = 2x√(x² + 1)
The worked answer, (2/3)(x² + 1)√(x² + 1), and its slope 2x√(x² + 1), the integrand. The check step, drawn.
Worked example

Substitution, start to finish: $\int 2x \sqrt{x^2+1} \mathrm{d} x$

  1. Find $\int 2x \sqrt{x^2+1} \mathrm{d} x$
    Look for a function whose derivative also appears in the integrand.
  2. $t = x^2+1$, so $\mathrm{d} t = 2x \mathrm{d} x$
    The derivative of $x^2 + 1$ is $2 x$, and $2 x$ is already there.
  3. $\int 2x \sqrt{x^2+1} \mathrm{d} x = \int \sqrt{t} \mathrm{d} t$
    Put in $t$ and $\mathrm{d} t$. Now the integral is in $t$ alone.
  4. $\int \sqrt{t} \mathrm{d} t = 2/3 \cdot t^{3/2} + C$
    This is the power rule, with $n = 1/2$.
  5. $\int 2x \sqrt{x^2+1} \mathrm{d} x = 2/3 \cdot (x^2+1)^{3/2} + C$
    Put $x^2 + 1$ back in place of $t$. The answer must be in $x$.
  6. Check by differentiating: $d/d x (2/3 \cdot (x^2+1)^{3/2}) = 2/3 \cdot 3/2 \cdot (x^2+1)^{1/2} \cdot 2x = 2x \sqrt{x^2+1}$
    Check by differentiating the answer with the chain rule. You get back the integrand.

Four more formulas come from substitution, for $\tan x$, $\cot x$, $\sec x$ and $\text{cosec} x$. Here is the one for $\tan x$, line by line.

Why it works

Find $\int \tan x \mathrm{d} x$

  1. $\int \tan x \mathrm{d} x = \int (\sin x)/(\cos x) \mathrm{d} x$
    Write $\tan x$ as $(\sin x)/(\cos x)$.
  2. $t = \cos x$, so $\mathrm{d} t = -\sin x \mathrm{d} x$
    The top is minus the derivative of the bottom, so put $t$ for the bottom.
  3. $\int (\sin x)/(\cos x) \mathrm{d} x = -\int (\mathrm{d} t)/t$
    Write $-\mathrm{d} t$ in place of $\sin x \mathrm{d} x$.
  4. $-\int (\mathrm{d} t)/t = -\ln \left|\cos x\right| + C$
    The integral of $1/t$ is $\ln \left|t\right|$, then put $\cos x$ back.
  5. *$\int \tan x \mathrm{d} x = \ln \left|\sec x\right| + C$*
    A minus sign before a log turns $\cos x$ into $1/(\cos x)$, which is $\sec x$.

The cot formula is the same, with $t = \sin x$: $\int \cot x \mathrm{d} x = \ln \left|\sin x\right| + C$.

The sec formula needs one more step. Multiply above and below by $\sec x + \tan x$.

Why it works

Find $\int \sec x \mathrm{d} x$

  1. $\int \sec x \mathrm{d} x = \int (\sec x (\sec x + \tan x))/(\sec x + \tan x) \mathrm{d} x$
    Multiply above and below by $\sec x + \tan x$, which changes nothing.
  2. $\int \sec x \mathrm{d} x = \int (\sec^2 x + \sec x \tan x)/(\sec x + \tan x) \mathrm{d} x$
    Open the bracket on top.
  3. $t = \sec x + \tan x$, so $\mathrm{d} t = (\sec x \tan x + \sec^2 x) \mathrm{d} x$
    The top is the derivative of the bottom, so put $t$ for the bottom.
  4. $\int (\mathrm{d} t)/t = \ln \left|t\right| + C$
    The integral of $1/t$ is $\ln \left|t\right|$.
  5. *$\int \sec x \mathrm{d} x = \ln \left|\sec x + \tan x\right| + C$*
    Put $\sec x + \tan x$ back in place of $t$.

The cosec formula works the same way, with $\text{cosec} x - \cot x$ above and below: its derivative is $\text{cosec} x (\text{cosec} x - \cot x)$, the new top. $\int \text{cosec} x \mathrm{d} x = \ln \left|\text{cosec} x - \cot x\right| + C$.

Check yourself
  1. For $\int 2x \sqrt{x^2+1} d x$, the natural substitution is \_\_\_\_\_\_.
    1. $t=\sqrt{x^2+1}$, since it is the more complicated-looking factor
    2. $t=2x$, since it is the coefficient in the integrand
    3. $t=x$, treating the whole integrand as a function of $x$ directly
    4. $t=x^2+1$, since its derivative matches the leftover factor $2x$
    Check your answer
    1. $t=\sqrt{x^2+1}$, since it is the more complicated-looking factor — $t=\sqrt{x^2+1}$ leaves a derivative that does not cleanly match the remaining $2x$ factor, the working substitution is the simpler $t=x^2+1$.
    2. $t=2x$, since it is the coefficient in the integrand — $t=2x$ has derivative $2$, which does not connect to $\sqrt{x^2+1}$ at all, this substitution collapses nothing.
    3. $t=x$, treating the whole integrand as a function of $x$ directly — Setting $t=x$ changes no variable at all, a useful substitution must simplify the integrand, which renaming $x$ as $t$ does not.
    4. ✓ $t=x^2+1$, since its derivative matches the leftover factor $2x$ — (D) $t=x^2+1$ gives $d t=2x d x$, which is exactly the factor already sitting in the integrand.
  2. With $t=x^2+1$, $\int 2x \sqrt{x^2+1} d x$ becomes \_\_\_\_\_\_.
    1. $\int \sqrt{t} d x$
    2. $\int 2x \sqrt{t} d t$
    3. $\int \sqrt{t} d t$
    4. $\int t \sqrt{t} d t$
    Check your answer
    1. $\int \sqrt{t} d x$ — Once $t=x^2+1$ is set, $d t=2x d x$ replaces the differential entirely, leaving $d x$ mixes the old and new variables.
    2. $\int 2x \sqrt{t} d t$ — $2x d x$ converts entirely to $d t$, leaving a bare $2x$ alongside $d t$ double-counts that factor.
    3. ✓ $\int \sqrt{t} d t$ — (C) Both the square-root factor and the differential convert to $t$ together, leaving a clean $\int \sqrt{t} d t$.
    4. $\int t \sqrt{t} d t$ — The integrand after substitution is exactly $\sqrt{t}$, with the $2x d x$ already absorbed into $d t$, no extra factor of $t$ belongs here.
  3. $\int \sqrt{t} d t = 2/3 \cdot t^{3/2} + C$. Substituting back, $\int 2x \sqrt{x^2+1} d x$ equals \_\_\_\_\_\_.
    1. $2/3 \cdot t^{3/2} + C$, left in terms of $t$
    2. $2/3 \cdot (x^2+1)^{3/2} + C$
    3. $2/3 \cdot x^{3/2} + C$
    4. $2/3 \cdot (2x+1)^{3/2} + C$
    Check your answer
    1. $2/3 \cdot t^{3/2} + C$, left in terms of $t$ — The original problem was posed in $x$, an answer left in $t$ has not yet been converted back to answer that original problem.
    2. ✓ $2/3 \cdot (x^2+1)^{3/2} + C$ — (B) Replacing $t$ with the full expression $x^2+1$ gives the final answer in terms of the original variable.
    3. $2/3 \cdot x^{3/2} + C$ — $t$ stood for $x^2+1$, not $x$ alone, replacing $t$ with bare $x$ drops the $+1$ and the squaring entirely.
    4. $2/3 \cdot (2x+1)^{3/2} + C$ — $t$ was defined as $x^2+1$, replacing it with $2x+1$ uses a different expression than the one actually substituted.

↑ Back to top

A toolkit of special-form integrals

▲CONCEPT
1/(x² + 4)
1/√(4 − x²)
√(4 − x²)
Three table shapes with a = 2: 1/(x² + 4), 1/√(4 − x²) and √(4 − x²). Match the shape, read the formula.

Some integrals come up again and again. For those you keep the formula and look it up. There are nine. You match the shape and read off the answer. In every formula $a > 0$. Each answer holds wherever its integrand is defined. Differentiate any answer and you get its integrand back. That is the check.

The nine are in one table, in three groups of three.

The integralThe answer
No square root
$\int 1/(x^2 + a^2) \mathrm{d} x$$(1/a) \tan^{-1}(x/a) + C$
$\int 1/(x^2 - a^2) \mathrm{d} x$$(1/(2a)) \ln|(x - a)/(x + a)| + C$
$\int 1/(a^2 - x^2) \mathrm{d} x$$(1/(2a)) \ln|(a + x)/(a - x)| + C$
A square root below the line
$\int 1/\sqrt{a^2 - x^2} \mathrm{d} x$$\sin^{-1}(x/a) + C$
$\int 1/\sqrt{x^2 + a^2} \mathrm{d} x$$\ln|x + \sqrt{x^2 + a^2}| + C$
$\int 1/\sqrt{x^2 - a^2} \mathrm{d} x$$\ln|x + \sqrt{x^2 - a^2}| + C$
The square root on top
$\int \sqrt{a^2 - x^2} \mathrm{d} x$$(x/2) \sqrt{a^2 - x^2} + (a^2/2) \sin^{-1}(x/a) + C$
$\int \sqrt{x^2 + a^2} \mathrm{d} x$$(x/2) \sqrt{x^2 + a^2} + (a^2/2) \ln|x + \sqrt{x^2 + a^2}| + C$
$\int \sqrt{x^2 - a^2} \mathrm{d} x$$(x/2) \sqrt{x^2 - a^2} - (a^2/2) \ln|x + \sqrt{x^2 - a^2}| + C$

To prove one formula, take $\int 1/\sqrt{a^2 - x^2} \mathrm{d} x$ and put $x = a \sin t$, with $t$ in $[-\pi/2, \pi/2]$. Here is why it gives the formula.

Why it works

Why $\int 1/\sqrt{a^2 - x^2} \mathrm{d} x = \sin^{-1}(x/a) + C$

  1. $x = a \sin t$, so $\mathrm{d} x = a \cos t \mathrm{d} t$
    Differentiate $x = a \sin t$, then write $\mathrm{d} x$ as $a \cos t \mathrm{d} t$.
  2. $\sqrt{a^2 - x^2} = \sqrt{a^2 - a^2 \sin^2 t}$
    Put $a \sin t$ in place of $x$.
  3. $\sqrt{a^2 - a^2 \sin^2 t} = \sqrt{a^2 \cos^2 t} = a \cos t$
    Use $1 - \sin^2 t = \cos^2 t$, and $\cos t \geq 0$ for $t$ in $[-\pi/2, \pi/2]$.
  4. $\int 1/\sqrt{a^2 - x^2} \mathrm{d} x = \int (a \cos t)/(a \cos t) \mathrm{d} t$
    Put both into the integral.
  5. $\int 1 \mathrm{d} t = t + C$
    The fraction is 1, and the integral of 1 is $t$.
  6. *$t + C = \sin^{-1}(x/a) + C$*
    $\sin t = x/a$ with $t$ on the principal branch, so $t = \sin^{-1}(x/a)$.
xy′y′ = 1/√(25 − x²)
a = 5. The slope of sin⁻¹(x/5) is 1/√(25 − x²), so the formula’s answer checks.

To use a formula, first find $a$. Take $\int 1/\sqrt{25 - x^2} \mathrm{d} x$.

Worked example

Find $\int 1/\sqrt{25 - x^2} \mathrm{d} x$ from the toolkit

  1. $a^2 = 25$, so $a = 5$
    Match the shape $1/\sqrt{a^2 - x^2}$: the number in place of $a^2$ is 25.
  2. *$\int 1/\sqrt{25 - x^2} \mathrm{d} x = \sin^{-1}(x/5) + C$*
    Put $a = 5$ into the formula $\sin^{-1}(x/a) + C$.
  3. $d/d x (\sin^{-1}(x/5)) = 1/\sqrt{1 - x^2/25} \cdot 1/5$
    Check by differentiating: the chain rule gives the extra factor $1/5$.
  4. $1/\sqrt{1 - x^2/25} \cdot 1/5 = 1/\sqrt{25 - x^2}$
    Take the 5 inside the root as 25: the integrand comes back.
y = x² + 4x + 13name
(x + 2)² + 9
Complete the square: x² + 4x + 13 is (x + 2)² + 9, lowest at x = −2. Then a = 3.

Sometimes the shape is hidden. Then complete the square first. Take $\int 1/(x^2 + 4 x + 13) \mathrm{d} x$. Half of 4 is 2, and $2^2$ is 4, so split 13 as $4 + 9$:

$x^2 + 4 x + 13 = (x^2 + 4 x + 4) + 9$.

The bracket is a perfect square:

$x^2 + 4 x + 13 = (x + 2)^2 + 9$.

Now put $t = x + 2$. Its derivative is 1, so $\mathrm{d} t = \mathrm{d} x$:

$\int 1/(x^2 + 4 x + 13) \mathrm{d} x = \int 1/(t^2 + 3^2) \mathrm{d} t$.

That is the first formula, with $a = 3$.

Your turn: find $\int 1/(x^2 - 36) \mathrm{d} x$. Which formula fits, and what is $a$?

Check yourself
  1. $\int 1/(x^2+a^2) d x$ equals \_\_\_\_\_\_.
    1. $1/a \cdot \tan^{-1}(x/a) + C$
    2. $\tan^{-1}(x/a) + C$
    3. $1/a \cdot \sin^{-1}(x/a) + C$
    4. $a \cdot \tan^{-1}(x/a) + C$
    Check your answer
    1. ✓ $1/a \cdot \tan^{-1}(x/a) + C$ — (A) The $x^2+a^2$ shape matches the $\tan^{-1}$ form, with a leading factor of $1/a$.
    2. $\tan^{-1}(x/a) + C$ — The formula needs the factor $1/a$ in front of $\tan^{-1}(x/a)$, dropping it changes the value whenever $a \neq 1$.
    3. $1/a \cdot \sin^{-1}(x/a) + C$ — $\sin^{-1}$ belongs to the $1/\sqrt{a^2-x^2}$ form, $x^2+a^2$ under no root matches $\tan^{-1}$, not $\sin^{-1}$.
    4. $a \cdot \tan^{-1}(x/a) + C$ — The leading factor is $1/a$, not $a$, multiplying by $a$ inverts the correct factor.
  2. $\int 1/(x^2-a^2) d x$ equals \_\_\_\_\_\_.
    1. $\ln|x^2-a^2| + C$
    2. $1/(2a) \cdot \ln|(x-a)/(x+a)| + C$
    3. $1/a \cdot \tan^{-1}(x/a) + C$
    4. $1/(2a) \cdot \ln|(x+a)/(x-a)| + C$
    Check your answer
    1. $\ln|x^2-a^2| + C$ — $\ln|x^2-a^2|$ is not the integral of $1/(x^2-a^2)$, the correct form comes from splitting into partial fractions first, giving the ratio form instead.
    2. ✓ $1/(2a) \cdot \ln|(x-a)/(x+a)| + C$ — (B) The $x^2-a^2$ shape (a difference, not a sum) matches its own log-based special form.
    3. $1/a \cdot \tan^{-1}(x/a) + C$ — $\tan^{-1}$ matches $x^2+a^2$, a sum under the fraction, $x^2-a^2$ is a difference, which matches the log form instead.
    4. $1/(2a) \cdot \ln|(x+a)/(x-a)| + C$ — The ratio inside the log is $(x-a)/(x+a)$, not $(x+a)/(x-a)$, flipping it changes the sign of the result.
  3. $\int 1/\sqrt{a^2-x^2} d x$ equals \_\_\_\_\_\_.
    1. $-\sin^{-1}(x/a) + C$
    2. $1/a \cdot \sin^{-1}(x/a) + C$
    3. $\sin^{-1}(x/a) + C$
    4. $\tan^{-1}(x/a) + C$
    Check your answer
    1. $-\sin^{-1}(x/a) + C$ — $\sin^{-1}(x/a)$, not its negative, differentiates to $1/\sqrt{a^2-x^2}$, the sign here is flipped.
    2. $1/a \cdot \sin^{-1}(x/a) + C$ — Unlike the $\tan^{-1}$ and log forms, $\sin^{-1}(x/a)$ needs no leading $1/a$ factor, adding one changes the value.
    3. ✓ $\sin^{-1}(x/a) + C$ — (C) The square root of $a^2-x^2$ in the denominator matches the $\sin^{-1}$ form, with no leading fraction needed.
    4. $\tan^{-1}(x/a) + C$ — $\tan^{-1}(x/a)$ matches $1/(x^2+a^2)$, a different shape with no square root, this integrand needs $\sin^{-1}$.
  4. $\int 1/\sqrt{9-x^2} d x$ matches which special form, and what does $a$ equal?
    1. the $1/(x^2+a^2)$ form, with $a=9$
    2. the $1/\sqrt{a^2-x^2}$ form, with $a=3$
    3. the $1/(x^2-a^2)$ form, with $a=3$
    4. the $1/\sqrt{a^2-x^2}$ form, with $a=9$
    Check your answer
    1. the $1/(x^2+a^2)$ form, with $a=9$ — The $\tan^{-1}$ form has no square root at all, $9-x^2$ under a root does not match it, whatever value of $a$ is used.
    2. ✓ the $1/\sqrt{a^2-x^2}$ form, with $a=3$ — (B) $9-x^2$ matches $a^2-x^2$ with $a^2=9$, so $a=3$, the $\sin^{-1}$ special form.
    3. the $1/(x^2-a^2)$ form, with $a=3$ — The log form is for $1/(x^2-a^2)$, with no square root, this integrand has a root, so it does not match that form either.
    4. the $1/\sqrt{a^2-x^2}$ form, with $a=9$ — $a^2=9$ means $a=3$, the square root of $9$, using $9$ itself as $a$ skips that last step.
xy′y′ = 1/(x² + 4)
The worked answer, (1/2) tan⁻¹(x/2), and its slope 1/(x² + 4). Differentiating gives back the integrand.
Worked example

A special form, matched by shape: $\int 1/(x^2+4) \mathrm{d} x$

  1. Find $\int 1/(x^2+4) \mathrm{d} x$
    Match the integrand to a formula in the toolkit.
  2. $\int 1/(x^2+a^2) \mathrm{d} x = 1/a \cdot \tan^{-1}(x/a) + C$, with $a^2=4$ so $a=2$
    The integrand has the shape $1/(x^2 + a^2)$, with $a^2 = 4$.
  3. $\int 1/(x^2+4) \mathrm{d} x = 1/2 \cdot \tan^{-1}(x/2) + C$
    Put $a = 2$ into the formula. There is nothing to derive.
  4. Check by differentiating: $d/d x (1/2 \cdot \tan^{-1}(x/2)) = 1/2 \cdot 1/(1+(x/2)^2) \cdot 1/2 = 1/(4+x^2)$
    Check by differentiating. The chain rule gives the extra factor $1/2$, and the result is the integrand.
Check yourself
  1. For $\int 1/(x^2+4) d x$, matching $1/(x^2+a^2)$ gives $a$ equal to \_\_\_\_\_\_.
    1. $4$
    2. $2$
    3. $\sqrt{2}$
    4. $16$
    Check your answer
    1. $4$ — $a^2=4$ names $a^2$, not $a$ itself, taking the square root of $4$ gives $a=2$.
    2. ✓ $2$ — (B) $a^2=4$, so $a=2$, the square root of $4$.
    3. $\sqrt{2}$ — $a^2=4$, not $2$, the square root should be taken of $4$, giving $a=2$, not $\sqrt{2}$.
    4. $16$ — $4$ needs a square ROOT to find $a$, not a further squaring, $\sqrt{4}=2$, not $16$.
  2. With $a=2$, $\int 1/(x^2+4) d x$ equals \_\_\_\_\_\_.
    1. $2 \cdot \tan^{-1}(x/2) + C$
    2. $(1/2) \cdot \tan^{-1}(2 x) + C$
    3. $1/2 \cdot \tan^{-1}(x/2) + C$
    4. $\tan^{-1}(x/2) + C$
    Check your answer
    1. $2 \cdot \tan^{-1}(x/2) + C$ — The leading factor is $1/a=1/2$, not $a=2$, multiplying by $a$ inverts the correct factor.
    2. $(1/2) \cdot \tan^{-1}(2 x) + C$ — The argument of $\tan^{-1}$ is $x/a=x/2$, not $a x=2x$, multiplying by $a$ inside $\tan^{-1}$ inverts the correct argument.
    3. ✓ $1/2 \cdot \tan^{-1}(x/2) + C$ — (C) Plugging $a=2$ into $1/a \cdot \tan^{-1}(x/a)+C$ gives $1/2 \cdot \tan^{-1}(x/2)+C$.
    4. $\tan^{-1}(x/2) + C$ — The formula needs the leading factor $1/a=1/2$, dropping it changes the value of the integral.
  3. Differentiating $1/2 \cdot \tan^{-1}(x/2)$ to check the answer gives \_\_\_\_\_\_.
    1. $1/2 \cdot 1/(1+(x/2)^2)$, which does not match the original integrand
    2. $\tan^{-1}(x/2)$, recovering the integral unchanged
    3. it simplifies back to $1/(4+x^2)$, confirming the original integrand
    4. $1/(x^2+4)^2$, a squared version of the integrand
    Check your answer
    1. $1/2 \cdot 1/(1+(x/2)^2)$, which does not match the original integrand — Differentiating $\tan^{-1}(x/2)$ needs the chain rule’s inner factor, $1/2$, from the derivative of $x/2$, stopping before that factor leaves the check incomplete.
    2. $\tan^{-1}(x/2)$, recovering the integral unchanged — Differentiating should return the ORIGINAL INTEGRAND, not the integral itself unchanged, that would mean no differentiation happened at all.
    3. ✓ it simplifies back to $1/(4+x^2)$, confirming the original integrand — (C) The chain rule contributes an extra factor of $1/2$ from differentiating the inner $x/2$, and simplifying recovers exactly $1/(4+x^2)$.
    4. $1/(x^2+4)^2$, a squared version of the integrand — Carrying out the chain rule correctly gives $1/(4+x^2)$, the plain integrand, nothing in the computation produces a squared denominator.

↑ Back to top

Integration by parts, and when to skip it

▲CONCEPT
x × (−cos x)
→
d/dx
product rule
→
−cos x + x sin x
The product rule on x and −cos x gives −cos x + x sin x. Read backwards, it finds the integral of x sin x.

The product rule from Chapter 5 says $d/d x (u w) = u' w + u w'$. Read it backwards and you get INTEGRATION BY PARTS.

A 3 by 3 rectangle, area 9, cut by a curve: below is 3, the integral of w du; above is 6, u dw.

Say you want $\int u \cdot v \mathrm{d} x$. The product rule gives a formula for it, one line at a time.

Why it works

From the product rule to by parts

  1. $w = \int v \mathrm{d} x$, so $w' = v$
    Call an integral of $v$ by the letter $w$: its derivative is $v$.
  2. $d/d x (u w) = u' w + u w'$
    The product rule from Chapter 5.
  3. $u w = \int u' w \mathrm{d} x + \int u w' \mathrm{d} x$
    Integrate both sides: the integral of a derivative is the function itself.
  4. $u w = \int u' w \mathrm{d} x + \int u v \mathrm{d} x$
    Put $v$ in place of $w'$.
  5. $\int u v \mathrm{d} x = u w - \int u' w \mathrm{d} x$
    Move $\int u' w \mathrm{d} x$ to the other side, changing its sign.
  6. *$\int u \cdot v \mathrm{d} x = u \int v \mathrm{d} x - \int (u' \int v \mathrm{d} x) \mathrm{d} x$*
    Put $\int v \mathrm{d} x$ back in place of $w$.

*The factor $u$ is the FIRST FUNCTION* and $v$ is the SECOND FUNCTION. You differentiate the first and integrate the second.

inverse trig
sin⁻¹ xtan⁻¹ x
log
ln x
algebraic
xx²
trig
sin xcos x
exponential
eˣ
ILATE, a rule of thumb: inverse trig, log, algebraic, trig, exponential. Take as u the factor from the leftmost box.

Choose them so that the new integral is easier. A factor that gets simpler when you differentiate it makes a good first function. So does a factor you cannot integrate easily, such as $\ln x$. Many teachers order the choice with the word ILATE. Its letters stand for inverse trig, log, algebraic, trig and exponential. Take as $u$ the factor that comes first. It is a rule of thumb, not part of the method.

y = −x cos x + sin x
xy′y′ = x sin x
−x cos x + sin x and its slope, x sin x. The answer from by parts, checked by its graph.

Let us try $\int x \sin x \mathrm{d} x$, with the formula.

Worked example

Find $\int x \sin x \mathrm{d} x$ by parts

  1. $u = x$, $v = \sin x$
    Differentiating $x$ gives 1, which is simpler, so $x$ is the first function.
  2. $\int v \mathrm{d} x = -\cos x$
    The formula needs an integral of $v$, and the table gives $-\cos x$.
  3. $\int x \sin x \mathrm{d} x = -x \cos x - \int 1 \cdot (-\cos x) \mathrm{d} x$
    Put $u = x$, $u' = 1$ and $\int v \mathrm{d} x = -\cos x$ into the formula.
  4. *$\int x \sin x \mathrm{d} x = -x \cos x + \sin x + C$*
    Two minus signs make a plus, and the integral of $\cos x$ is $\sin x$.
  5. $d/d x (-x \cos x + \sin x) = -\cos x + x \sin x + \cos x = x \sin x$
    Check by differentiating with the product rule: the integrand comes back.

So one hard integral becomes an easier one.

Your turn: find $\int x \ln x \mathrm{d} x$. Which factor has to be $u$?

Check yourself
  1. Integration by parts states $\int u \cdot v d x$ equals \_\_\_\_\_\_.
    1. $\int u d x \cdot \int v d x$
    2. $u \int v d x - \int (u' \int v d x) d x$
    3. $u \cdot v - \int u' \cdot v' d x$
    4. $u' \int v d x + \int (u \int v d x) d x$
    Check your answer
    1. $\int u d x \cdot \int v d x$ — $\int u \cdot v d x$ is not the product of two separate integrals, no rule lets a product’s integral split apart like that.
    2. ✓ $u \int v d x - \int (u' \int v d x) d x$ — (B) By-parts undoes the product rule: keep $u$ times the integral of $v$, then subtract the integral of $u'$ times that same running integral.
    3. $u \cdot v - \int u' \cdot v' d x$ — By-parts involves $\int v d x$, the integral of $v$, not $v'$, this invented formula does not match the product-rule reversal.
    4. $u' \int v d x + \int (u \int v d x) d x$ — The first term keeps $u$ itself (not $u'$), multiplied by $\int v d x$, and the second term is SUBTRACTED, not added.
  2. ILATE ranks factor types for choosing $u$. What is it actually for?
    1. a rule for choosing which factor to integrate, not which to differentiate
    2. a way to decide whether substitution or by-parts should be used
    3. a starting guess for $u$, to check against the new integral
    4. a fixed rule that must always be followed in order
    Check your answer
    1. a rule for choosing which factor to integrate, not which to differentiate — ILATE ranks candidates for $u$, the factor to differentiate, the factor to integrate, $v$, is whatever is left over.
    2. a way to decide whether substitution or by-parts should be used — ILATE operates entirely inside by-parts, ranking candidates for $u$, it says nothing about whether to use substitution instead.
    3. ✓ a starting guess for $u$, to check against the new integral — (C) ILATE is a heuristic ranking for $u$, a guess worth checking, since a factor earlier in the list usually simplifies when differentiated.
    4. a fixed rule that must always be followed in order — ILATE is a starting guess, not a rule to apply blind, treating it as fixed can mean unnecessary work a shortcut would skip.
  3. Choosing which factor is $u$ and which is $v$ in by-parts matters because \_\_\_\_\_\_.
    1. the answer’s numeric value changes depending on the choice
    2. only one choice gives a valid answer; the other choice is mathematically wrong
    3. one choice simplifies the new integral; the other can make it harder
    4. $v$ must always be the factor written second in the product
    Check your answer
    1. the answer’s numeric value changes depending on the choice — A correct application of either valid choice reaches the same true integral, just with more or less work along the way.
    2. only one choice gives a valid answer; the other choice is mathematically wrong — Either choice of $u$ and $v$ is mathematically valid, the difference is how much harder or easier the remaining integral becomes, not validity.
    3. ✓ one choice simplifies the new integral; the other can make it harder — (C) Both choices are valid in principle, but a poor choice can leave a harder integral than the one you started with, the good choice is the one that simplifies.
    4. $v$ must always be the factor written second in the product — Nothing about writing order fixes which factor is $u$, the choice is about which factor to differentiate, not where it sits in the product.
  4. For $\int \ln x \cdot x d x$, ILATE suggests \_\_\_\_\_\_.
    1. $u = x$, since algebraic ranks above logarithmic
    2. $u$ and $v$ can be chosen either way with no difference in difficulty
    3. neither factor should be $u$; by-parts does not apply here
    4. $u = \ln x$, since logarithmic ranks above algebraic
    Check your answer
    1. $u = x$, since algebraic ranks above logarithmic — ILATE ranks Logarithmic above Algebraic, $u=x$ reverses that order; ILATE suggests $u=\ln x$ instead.
    2. $u$ and $v$ can be chosen either way with no difference in difficulty — If either choice were equally easy, ILATE would serve no purpose, the ranking exists because $\ln x$ is far easier to differentiate than to integrate.
    3. neither factor should be $u$; by-parts does not apply here — A logarithm times an algebraic factor is a textbook by-parts case, ILATE exists to guide the choice of $u$ here, not to rule the method out.
    4. ✓ $u = \ln x$, since logarithmic ranks above algebraic — (D) ILATE ranks Logarithmic above Algebraic, so $\ln x$ should be differentiated as $u$ and $x$ integrated as $v$.
y = x eˣ − eˣ
xy′y′ = x eˣ
The worked answer, x eˣ − eˣ, and its slope x eˣ. Differentiating gives back the integrand.
Worked example

By parts, checked by differentiating back: $\int x \cdot e^x \mathrm{d} x$

  1. Find $\int x \cdot e^x \mathrm{d} x$
    A product of $x$ and $e^x$. Differentiating $x$ makes it simpler, so take $u = x$.
  2. $u = x$, $v = e^x$
    $x$ becomes 1 when you differentiate it. $e^x$ stays $e^x$ when you integrate it.
  3. $\int v \mathrm{d} x = e^x$
    The formula needs this integral of $v$.
  4. $\int x \cdot e^x \mathrm{d} x = x \cdot e^x - \int 1 \cdot e^x \mathrm{d} x$
    Put $u = x$, $u' = 1$ and $\int v \mathrm{d} x = e^x$ into the formula for by parts.
  5. $\int x \cdot e^x \mathrm{d} x = x \cdot e^x - e^x + C$
    The integral left over is in the standard table.
  6. Check by differentiating: $d/d x (x \cdot e^x - e^x) = e^x + x \cdot e^x - e^x = x \cdot e^x$
    Check by differentiating with the product rule. You get back $x \cdot e^x$.
Worked example

By parts twice: $\int x^2 e^x \mathrm{d} x$

  1. $u = x^2$, $v = e^x$
    Differentiating $x^2$ makes it simpler: take $x^2$ as the first function, same choice as $x \cdot e^x$.
  2. $\int v \mathrm{d} x = e^x$
    The formula needs this integral of $v$.
  3. $\int x^2 e^x \mathrm{d} x = x^2 e^x - \int 2 x \cdot e^x \mathrm{d} x$
    Put $u = x^2$, $u' = 2 x$ and $\int v \mathrm{d} x = e^x$ into the formula for by parts.
  4. $\int x \cdot e^x \mathrm{d} x = x e^x - e^x + C$
    A second round of by parts, on the leftover integral: the same working as $\int x \cdot e^x \mathrm{d} x$ above.
  5. *$\int x^2 e^x \mathrm{d} x = x^2 e^x - 2(x e^x - e^x) + C = e^x(x^2 - 2 x + 2) + C$*
    Put the second round’s answer back in and collect the $e^x$ terms.
  6. $d/d x (e^x(x^2 - 2 x + 2)) = e^x(x^2 - 2 x + 2) + e^x(2 x - 2) = x^2 e^x$
    Check by differentiating with the product rule: the integrand comes back.
Worked example

By parts twice, then solving for the integral: $\int e^x \cos x \mathrm{d} x$

  1. Let $I = \int e^x \cos x \mathrm{d} x$
    Name the integral $I$: it will reappear on the right side.
  2. $u = \cos x$, $v = e^x$
    Either factor can be the first function; this round takes $\cos x$.
  3. $I = e^x \cos x + \int e^x \sin x \mathrm{d} x$
    By parts: $\int v \mathrm{d} x = e^x$, and $u' = -\sin x$ turns the usual minus sign into a plus.
  4. $u = \sin x$, $v = e^x$, for $\int e^x \sin x \mathrm{d} x$
    A second round of by parts, on the new integral.
  5. $\int e^x \sin x \mathrm{d} x = e^x \sin x - I$
    The same formula; the leftover integral is $\int e^x \cos x \mathrm{d} x$ again, not a new one, so it is $I$.
  6. $I = e^x \cos x + e^x \sin x - I$
    Substitute the second round’s result back into the first line.
  7. *$2 I = e^x(\sin x + \cos x)$, so $I = e^x(\sin x + \cos x)/2 + C$*
    Move $I$ to one side and divide by 2. The $+C$ belongs to the final answer, not to $I$ on the way.
  8. $d/d x (e^x(\sin x + \cos x)/2) = e^x \cos x$
    Check by differentiating with the product rule: the integrand comes back.
Check yourself
  1. For $\int x \cdot e^x d x$, ILATE gives \_\_\_\_\_\_.
    1. $u=x$, $d v=e^x d x$
    2. $u=x$ and $u=e^x$ simultaneously
    3. neither factor should be $u$
    4. $u=e^x$, $d v = x d x$
    Check your answer
    1. ✓ $u=x$, $d v=e^x d x$ — (A) ILATE ranks Algebraic above Exponential, so $x$ is differentiated as $u$ and $e^x$ is integrated as $v$.
    2. $u=x$ and $u=e^x$ simultaneously — Only one factor can be $u$ at a time, assigning both $x$ and $e^x$ as $u$ simultaneously is not a valid by-parts setup.
    3. neither factor should be $u$ — $x \cdot e^x$ is a standard by-parts case, one algebraic factor, one exponential factor, exactly what ILATE is built to guide.
    4. $u=e^x$, $d v = x d x$ — ILATE ranks Algebraic above Exponential, $u$ should be $x$, not $e^x$.
  2. With $u=x$, $d v=e^x d x$, by-parts gives $\int x \cdot e^x d x$ equals \_\_\_\_\_\_.
    1. $x \cdot e^x - \int 1 \cdot e^x d x$
    2. $x \cdot e^x + \int 1 \cdot e^x d x$
    3. $e^x - \int x d x$
    4. $x \cdot e^x - \int x \cdot e^x d x$
    Check your answer
    1. ✓ $x \cdot e^x - \int 1 \cdot e^x d x$ — (A) The by-parts formula keeps $u \cdot v$ as the first term and subtracts the integral of $u' \cdot v$.
    2. $x \cdot e^x + \int 1 \cdot e^x d x$ — The by-parts formula SUBTRACTS the second integral, $x \cdot e^x - \int 1 \cdot e^x d x$, not a sum.
    3. $e^x - \int x d x$ — The first term is $u \cdot v = x \cdot e^x$, keeping the factor $x$, dropping it loses part of the by-parts formula’s first piece.
    4. $x \cdot e^x - \int x \cdot e^x d x$ — $u'=1$, so the remaining integral is $\int 1 \cdot e^x d x = \int e^x d x$, not the original $\int x \cdot e^x d x$ again.
  3. $\int x \cdot e^x d x$ equals \_\_\_\_\_\_.
    1. $e^x - x \cdot e^x + C$
    2. $x \cdot e^x + e^x + C$
    3. $x \cdot e^x + C$
    4. $x \cdot e^x - e^x + C$
    Check your answer
    1. $e^x - x \cdot e^x + C$ — Differentiating $e^x - x \cdot e^x$ does not return $x \cdot e^x$, the correct order and signs are $x \cdot e^x - e^x$.
    2. $x \cdot e^x + e^x + C$ — The second term is SUBTRACTED, $x \cdot e^x - e^x + C$, not $x \cdot e^x + e^x + C$.
    3. $x \cdot e^x + C$ — $\int 1 \cdot e^x d x = e^x$ must still be subtracted, leaving it out gives an incomplete answer.
    4. ✓ $x \cdot e^x - e^x + C$ — (D) $\int 1 \cdot e^x d x = e^x$, so $x \cdot e^x - \int 1 \cdot e^x d x = x \cdot e^x - e^x + C$.
▲CONCEPT
eˣ sin x
→
d/dx
product rule
→
eˣ (sin x + cos x)
The product rule on eˣ sin x gives eˣ(sin x + cos x). Read backwards, that integral needs no by parts.

One formula comes straight from the product rule. For a function $f$ with a derivative, differentiate $e^x f(x)$ by the product rule:

$d/d x (e^x f(x)) = e^x f(x) + e^x f'(x)$.

*Take out $e^x$* as a common factor:

$d/d x (e^x f(x)) = e^x (f(x) + f'(x))$.

Read it backwards.

$\int e^x (f(x) + f'(x)) \mathrm{d} x = e^x f(x) + C$.

sin x
→
d/dx
→
cos x
1/x
→
d/dx
→
−1/x²
x²
→
d/dx
→
2x
Three pairs f and f′: sin x and cos x, 1/x and −1/x², x² and 2x. eˣ times any such sum fits the formula.

Look for $e^x$ times a function plus its own derivative, $e^x (f(x) + f'(x))$. The answer is $e^x f(x) + C$, with no by parts.

Top: e to the x sin x is flat at x equals minus pi over four. Below: the integrand crosses zero there too.

Take $f(x) = \sin x$, so $f'(x) = \cos x$. The bracket is $f(x) + f'(x)$, so the shortcut gives

$\int e^x (\sin x + \cos x) \mathrm{d} x = e^x \sin x + C$.

Check by differentiating with the product rule:

$d/d x (e^x \sin x) = e^x \sin x + e^x \cos x$, the integrand.

The order of the two terms does not matter. What matters is which term is $f$. Its derivative must be the other term.

Your turn: find $\int e^x (\sec x + \sec x \tan x) \mathrm{d} x$. Which term is $f$?

Check yourself
  1. $\int e^x (f(x)+f'(x)) d x$ equals \_\_\_\_\_\_.
    1. $e^x (f(x)+f'(x)) + C$
    2. $e^x \cdot f(x) \cdot f'(x) + C$
    3. $e^x f'(x) + C$
    4. $e^x f(x) + C$
    Check your answer
    1. $e^x (f(x)+f'(x)) + C$ — Integrating collapses the sum $f(x)+f'(x)$ down to $f(x)$ alone, the unintegrated sum cannot be the final answer.
    2. $e^x \cdot f(x) \cdot f'(x) + C$ — The shortcut’s answer uses $f(x)$ alone, multiplied by $e^x$, no product of $f$ and $f'$ appears in it.
    3. $e^x f'(x) + C$ — The shortcut’s answer is $e^x f(x)+C$, it keeps $f$, not $f'$, in the final expression.
    4. ✓ $e^x f(x) + C$ — (D) The shortcut collapses the whole integral to $e^x$ times $f(x)$ alone, plus $C$.
  2. Why does $\int e^x (f(x)+f'(x)) d x = e^x f(x)+C$ hold?
    1. because $f(x)$ and $f'(x)$ always cancel each other out
    2. because every function multiplied by $e^x$ integrates to itself
    3. the product rule run backwards: $(e^x f(x))' = e^x (f(x)+f'(x))$
    4. because $e^x$ integrates to itself, so it can simply be factored out unchanged
    Check your answer
    1. because $f(x)$ and $f'(x)$ always cancel each other out — $f(x)$ and $f'(x)$ do not cancel, they are the two terms the product rule produces when differentiating $e^x f(x)$.
    2. because every function multiplied by $e^x$ integrates to itself — Not every function times $e^x$ integrates to itself, only the specific sum $f(x)+f'(x)$ triggers this shortcut.
    3. ✓ the product rule run backwards: $(e^x f(x))' = e^x (f(x)+f'(x))$ — (C) Differentiating $e^x f(x)$ by the product rule gives exactly $e^x (f+f')$, so integrating that sum must return $e^x f(x)$.
    4. because $e^x$ integrates to itself, so it can simply be factored out unchanged — $e^x$ integrating to itself is true but separate, the shortcut’s mechanism is the product rule run backwards on $e^x f(x)$, not that fact alone.
  3. Which integral fits the $e^x (f+f')$ shortcut?
    1. $\int e^x/x d x$
    2. $\int e^x \sin x d x$
    3. $\int (e^x) \cdot x^2 d x$
    4. $\int e^x (x^2+2x) d x$
    Check your answer
    1. $\int e^x/x d x$ — $1/x$ alone is not a sum of a function and its derivative, the shortcut’s pattern requires an addition, not a division.
    2. $\int e^x \sin x d x$ — $\sin x$ is not a sum of a function and its derivative multiplying $e^x$, this integral needs repeated by-parts instead.
    3. $\int (e^x) \cdot x^2 d x$ — The shortcut needs a SUM, $f(x)+f'(x)$, a lone term like $x^2$ has no such sum to match.
    4. ✓ $\int e^x (x^2+2x) d x$ — (D) $f(x)=x^2$ gives $f'(x)=2x$, exactly the second term, the shortcut applies directly, giving $x^2 e^x+C$.
  4. For $\int e^x (1/x - 1/x^2) d x$, how does spotting the shortcut compare to working through ILATE-by-rote by-parts?
    1. the shortcut cannot be used because the terms involve $1/x$, not a polynomial
    2. the shortcut is one line; by-parts needs separate work
    3. both take exactly the same number of steps
    4. ILATE-by-rote is actually faster here, since the shortcut only works for polynomials
    Check your answer
    1. the shortcut cannot be used because the terms involve $1/x$, not a polynomial — $1/x - 1/x^2$ is exactly $f(x)+f'(x)$ for $f(x)=1/x$, the shortcut applies here just as it would for a polynomial $f$.
    2. ✓ the shortcut is one line; by-parts needs separate work — (B) Spotting $f(x)=1/x$ with $f'(x)=-1/x^2$ collapses the whole problem instantly; by-parts would work through the same terms far more slowly.
    3. both take exactly the same number of steps — The shortcut resolves in one line once spotted; by-parts on the same integrand takes several separate steps, the effort is not equal.
    4. ILATE-by-rote is actually faster here, since the shortcut only works for polynomials — The shortcut needs only $f(x)+f'(x)$ in the integrand, $f(x)=1/x$ works exactly as well as any polynomial $f$.
y = eˣ/x
xy′y′ = eˣ(1/x − 1/x²)
The worked answer, eˣ/x, and its slope eˣ(1/x − 1/x²), for x > 0. On the right, the slope is 0 at x = 1.
Worked example

Spotting the shortcut inside a product: $\int e^x (1/x - 1/x^2) \mathrm{d} x$

  1. Find $\int e^x (1/x - 1/x^2) \mathrm{d} x$
    Before trying by parts, test whether the bracket is $f(x) + f'(x)$.
  2. $f(x) = 1/x$, so $f'(x) = -1/x^2$
    The second term is the derivative of the first.
  3. $\int e^x (f(x)+f'(x)) \mathrm{d} x = e^x f(x) + C$
    This formula comes from the product rule, read backwards.
  4. $\int e^x (1/x - 1/x^2) \mathrm{d} x = e^x/x + C$
    Put $f(x) = 1/x$ into the formula.
  5. Check by differentiating: $d/d x (e^x/x) = (e^x \cdot x - e^x)/x^2 = e^x (1/x - 1/x^2)$
    Check by differentiating with the quotient rule. You get back the integrand.
Check yourself
  1. For $\int e^x (1/x - 1/x^2) d x$, matching $e^x (f+f')$ requires $f(x)$ equal to \_\_\_\_\_\_.
    1. $1/x - 1/x^2$, the whole expression in parentheses
    2. $-1/x^2$, since it is the second term written directly
    3. $1/x$, since its derivative is the second term
    4. $x$, the variable itself
    Check your answer
    1. $1/x - 1/x^2$, the whole expression in parentheses — The whole expression $1/x-1/x^2$ is the SUM $f+f'$, $f$ alone is just the first piece, $1/x$.
    2. $-1/x^2$, since it is the second term written directly — $-1/x^2$ IS $f'(x)$, the shortcut needs $f(x)$ itself, which is $1/x$.
    3. ✓ $1/x$, since its derivative is the second term — (C) $f(x)=1/x$ has $f'(x)=-1/x^2$, exactly matching the second term of the sum.
    4. $x$, the variable itself — $x$ appears nowhere as $f$ in this integrand, the two terms present are $1/x$ and $-1/x^2$, not $x$.
  2. With $f(x)=1/x$, the shortcut $\int e^x (f+f') d x = e^x f(x)+C$ gives \_\_\_\_\_\_.
    1. $e^x \cdot (-1/x^2) + C$
    2. $e^x (1/x - 1/x^2) + C$
    3. $e^x/x + C$
    4. $e^x \cdot x + C$
    Check your answer
    1. $e^x \cdot (-1/x^2) + C$ — The shortcut’s answer uses $f(x)=1/x$, not $f'(x)=-1/x^2$, the final expression should be $e^x/x$, not $e^x \cdot (-1/x^2)$.
    2. $e^x (1/x - 1/x^2) + C$ — Integrating collapses $f+f'$ down to $f$ alone, leaving the full sum unintegrated is not the shortcut’s actual result.
    3. ✓ $e^x/x + C$ — (C) Plugging $f(x)=1/x$ into $e^x f(x)+C$ gives $e^x/x+C$.
    4. $e^x \cdot x + C$ — $f(x)=1/x$ stays as $1/x$ in the answer, inverting it to $x$ changes the value entirely.
  3. Differentiating $e^x/x$ to check gives $(e^x \cdot x - e^x)/x^2$, which simplifies to \_\_\_\_\_\_.
    1. $e^x (1/x - 1/x^2)$
    2. $e^x (x-1)/x$
    3. $e^x (x+1)/x^2$, sign flipped
    4. $e^x/x^2$
    Check your answer
    1. ✓ $e^x (1/x - 1/x^2)$ — (A) Factoring $e^x$ from the numerator and splitting the fraction recovers $e^x (1/x-1/x^2)$, exactly the original integrand.
    2. $e^x (x-1)/x$ — The denominator stays $x^2$ throughout, dropping to a single $x$ loses a factor introduced by the quotient rule.
    3. $e^x (x+1)/x^2$, sign flipped — The numerator factors as $e^x (x-1)$, with a MINUS sign, flipping it to $(x+1)$ does not match the actual quotient-rule result.
    4. $e^x/x^2$ — The numerator $e^x \cdot x - e^x$ factors as $e^x (x-1)$, dropping the $(x-1)$ factor loses part of the expression.
✕MISCONCEPTION
ILATE
u = x² + 2x,
v = eˣ
So
the only way is
by parts
By parts once
(x² + 2x)eˣ − ∫(2x + 2)eˣ dx
?
The working on eˣ(x² + 2x), step by step. Which step is wrong? Keep your answer: the bracket tests it.

Here is a working for $\int e^x (x^2 + 2 x) \mathrm{d} x$. Which step is wrong?

y = x² eˣ
xy′y′ = (x² + 2x)eˣ
x² eˣ and its slope (x² + 2x)eˣ. With f = x² the formula gives the answer in one line, no by parts.

Keep your answer, then look at the bracket. With $f(x) = x^2$ you get $f'(x) = 2 x$. So the bracket is $f(x) + f'(x)$, and the shortcut gives the answer in one line:

$\int e^x (x^2 + 2 x) \mathrm{d} x = x^2 e^x + C$.

Check by differentiating with the product rule:

$d/d x (x^2 e^x) = 2 x e^x + x^2 e^x$, the integrand.

eˣ(x² + 2x)
→
by parts
twice
→
x² eˣ + C
eˣ(x² + 2x)
→
the eˣ formula
f = x²
→
x² eˣ + C
Two ways, one answer: by parts twice, or the eˣ formula once. ILATE helps choose u; it does not choose the method.

The second step is wrong. ILATE only orders the choice of $u$. It does not say that by parts is the best way. By parts does reach $x^2 e^x + C$ too, but only after a second round. Each round is one more chance to drop a sign.

So look for a pattern before you reach for by parts.

First try

I saw $\int e^x (\sin x + \cos x) \mathrm{d} x$ and went straight to by parts, taking $u = \sin x$ and $v = e^x$ because ILATE puts the trigonometric factor ahead of the exponential.

Second look

Check the $e^x (f(x)+f'(x))$ shortcut FIRST, before reaching for by parts at all. Here $f(x) = \sin x$ gives $f'(x) = \cos x$, and the bracket $\sin x + \cos x$ is exactly $f(x)+f'(x)$.

$\int e^x (f(x)+f'(x)) \mathrm{d} x = e^x f(x) + C$ gives $e^x \sin x + C$ in one line. By parts reaches the same answer, but only after two full rounds, each one a fresh chance to drop a sign.

Check yourself
  1. A student sees $\int e^x (x^2+2x) d x$, checks ILATE, sets $u=x^2+2x$, and works through two rounds of by parts. What did the student miss?
    1. ILATE was applied in the wrong order
    2. by-parts cannot be used on this integral at all
    3. $e^x$ should have been chosen as $u$ instead
    4. the $e^x (f+f')$ shortcut applies here
    Check your answer
    1. ILATE was applied in the wrong order — No reordering of ILATE’s ranking would have been faster here, the real fix is checking for the $e^x (f+f')$ shortcut BEFORE reaching for by-parts at all.
    2. by-parts cannot be used on this integral at all — By-parts does reach the correct answer here, in two rounds, it is valid, just far slower than recognizing the shortcut first.
    3. $e^x$ should have been chosen as $u$ instead — Choosing $u=e^x$ instead still goes through by-parts, the shortcut avoids by-parts altogether, which no choice of $u$ can match for speed.
    4. ✓ the $e^x (f+f')$ shortcut applies here — (D) $x^2+2x$ splits as $f(x)+f'(x)$ with $f(x)=x^2$, so the shortcut gives $x^2 e^x+C$ instantly, skipping by-parts entirely.
  2. ILATE is best understood as \_\_\_\_\_\_.
    1. a rule that applies only to exponential integrands
    2. a starting guess for $u$, never a strict rule to follow
    3. a method for evaluating definite integrals specifically
    4. a fixed lookup that always names the correct $u$
    Check your answer
    1. a rule that applies only to exponential integrands — ILATE ranks any factor type, inverse trig, log, algebraic, trig, exponential, for choosing $u$ in ANY by-parts problem, not only exponential ones.
    2. ✓ a starting guess for $u$, never a strict rule to follow — (B) ILATE is a heuristic starting guess, not a rule to apply without first checking for a faster shortcut.
    3. a method for evaluating definite integrals specifically — ILATE has nothing to do with definite versus indefinite integrals, it only guides which factor to differentiate inside by-parts.
    4. a fixed lookup that always names the correct $u$ — Treating ILATE as an unbreakable lookup is exactly the trap, it is a starting guess, worth checking against faster options first.
  3. For $\int e^x (x^2+2x) d x$, compare the shortcut answer to the ILATE-by-rote path.
    1. the shortcut gives a wrong answer; only by-parts is reliable here
    2. both methods take the same number of steps and carry the same risk of error
    3. the shortcut is one line; by-parts needs two full rounds of integration
    4. ILATE-by-rote is actually shorter, since it avoids checking for a pattern first
    Check your answer
    1. the shortcut gives a wrong answer; only by-parts is reliable here — The shortcut gives the exact, verifiable answer $x^2 e^x+C$, it is by-parts that takes the longer route to the same correct value, not the shortcut that is unreliable.
    2. both methods take the same number of steps and carry the same risk of error — One line versus two rounds of by-parts is clearly not equal effort, each extra round of by-parts is also an extra chance to slip on a sign.
    3. ✓ the shortcut is one line; by-parts needs two full rounds of integration — (C) Spotting the pattern collapses the problem instantly; working through by parts twice multiplies both the effort and the chance of a sign error.
    4. ILATE-by-rote is actually shorter, since it avoids checking for a pattern first — Checking for the shortcut first is what SAVES the two rounds of by-parts, skipping the check is what costs the extra effort, not the other way round.

↑ Back to top

Partial fractions before integrating

▲CONCEPT
1/21/3
added
1/21/3
split back
One half plus one third on a strip of six sixths: 3 sixths and 2 sixths make 5 sixths. Run back, 5 sixths splits again.

In Class 7 you added fractions. One half plus one third is five sixths. You made one denominator, 6, out of the two denominators 2 and 3. Now run it the other way. Start from five sixths. Since 6 is 2 times 3, you can split five sixths back into a half and a third.

So a fraction whose denominator is a product can be split. You get simpler fractions, one for each factor.

yyy
At x = 0 the pieces 1/(x − 1) and 1/(x + 2) read −1 and 0.5. They add to −0.5, the whole fraction’s value.

The same works for functions. A RATIONAL FUNCTION is one polynomial divided by another. It is PROPER when the degree of the numerator is less than the degree of the denominator. Say the denominator has two different linear factors. Then a proper rational function splits into PARTIAL FRACTIONS.

$(p x + q)/((x - a)(x - b)) = A/(x - a) + B/(x - b)$, with $a \neq b$.

Each piece integrates to a log: $\int A/(x - a) \mathrm{d} x = A \ln|x - a| + C$.

x = 1
→
into 2x + 1 = A(x + 2) + B(x − 1)
→
÷ 3
→
A = 1
x = −2
→
the same
→
÷ (−3)
→
B = 1
Put in each root. At x = 1, 3 = 3A, so A = 1. At x = −2, B = 1 the same way.

Let us split $(2 x + 1)/((x - 1)(x + 2))$, then integrate it.

Worked example

Split $(2 x + 1)/((x - 1)(x + 2))$ and integrate

  1. $(2 x + 1)/((x - 1)(x + 2)) = A/(x - 1) + B/(x + 2)$
    Write one fraction for each factor of the denominator.
  2. $2 x + 1 = A(x + 2) + B(x - 1)$
    Multiply both sides by $(x - 1)(x + 2)$, which clears every denominator.
  3. At $x = 1$: $3 = 3 A$, so $A = 1$
    Put $x = 1$: it makes $x - 1$ zero, so the $B$ term drops out.
  4. At $x = -2$: $-3 = -3 B$, so $B = 1$
    Put $x = -2$: it makes $x + 2$ zero, so the $A$ term drops out.
  5. *$\int (2 x + 1)/((x - 1)(x + 2)) \mathrm{d} x = \ln|x - 1| + \ln|x + 2| + C$*
    Each piece integrates to a log.
  6. At $x = 0$: $-1 + 1/2 = -1/2$, and $1/(-2) = -1/2$
    Check the split at $x = 0$: the two pieces and the fraction agree.

Two more splits are needed for Exercise 7.5. A repeated linear factor needs a term for each power of $(x - a)$:

$(p x + q)/((x - a)^2(x - b)) = A/(x - a) + B/(x - a)^2 + D/(x - b)$.

An irreducible quadratic factor, one with no real linear factors, needs a linear numerator over it, not a bare constant:

$(p x + q)/((x - a)(x^2 + b x + c)) = A/(x - a) + (B x + D)/(x^2 + b x + c)$.

Worked example

A repeated factor: $\int (6 x + 2)/((x - 1)^2(x + 3)) \mathrm{d} x$

  1. $(6 x + 2)/((x - 1)^2(x + 3)) = A/(x - 1) + B/(x - 1)^2 + D/(x + 3)$
    One factor is repeated: it needs both a plain and a squared term.
  2. $6 x + 2 = A(x - 1)(x + 3) + B(x + 3) + D(x - 1)^2$
    Multiply both sides by $(x - 1)^2(x + 3)$, which clears every denominator.
  3. At $x = 1$: $8 = 4 B$, so $B = 2$
    Substituting the repeated root kills the $A$ and $D$ terms, leaving $B$ directly.
  4. At $x = -3$: $-16 = 16 D$, so $D = -1$
    Substituting the other root kills the $A$ and $B$ terms, leaving $D$ directly.
  5. Comparing $x^2$-coefficients: $A + D = 0$, so $A = 1$
    The left side has no $x^2$-term, so $A$ and $D$’s $x^2$-contributions must cancel.
  6. *$\int (6 x + 2)/((x - 1)^2(x + 3)) \mathrm{d} x = \ln|x - 1| - 2/(x - 1) - \ln|x + 3| + C$*
    The plain-factor terms use the log rule; the squared term uses the power rule with $n = -2$.
  7. At $x = 0$: $1/(0-1) + 2/(0-1)^2 - 1/(0+3) = -1 + 2 - 1/3 = 2/3$, and the fraction at $x = 0$ is $2/((-1)^2 \cdot 3) = 2/3$
    Check by putting $x = 0$ into the split and into the original fraction: both give the same value.
Worked example

A quadratic factor: $\int (3 x^2 - 5 x + 8)/((x - 3)(x^2 + 2 x + 5)) \mathrm{d} x$

  1. $(3 x^2 - 5 x + 8)/((x - 3)(x^2 + 2 x + 5)) = A/(x - 3) + (B x + D)/(x^2 + 2 x + 5)$
    An irreducible quadratic factor needs a linear numerator over it.
  2. $3 x^2 - 5 x + 8 = A(x^2 + 2 x + 5) + (B x + D)(x - 3)$
    Multiply both sides by $(x - 3)(x^2 + 2 x + 5)$.
  3. At $x = 3$: $20 = 20 A$, so $A = 1$
    Substituting the linear factor’s root kills the quadratic term, leaving $A$ directly.
  4. Comparing $x^2$-coefficients: $3 = A + B$, so $B = 2$
    The $x^2$-terms on both sides must match.
  5. Comparing constant terms: $8 = 5 A - 3 D$, so $D = -1$
    The constant terms on both sides must match, once $A$ is known.
  6. $\int [1/(x - 3) + (2 x - 1)/(x^2 + 2 x + 5)] \mathrm{d} x$
    Split the second term: $2 x - 1 = (2 x + 2) - 3$, the first part the derivative of $x^2 + 2 x + 5$.
  7. *$= \ln|x - 3| + \ln(x^2 + 2 x + 5) - 3/2 \cdot \tan^{-1}((x + 1)/2) + C$*
    The log term comes from $u = x^2 + 2 x + 5$; the last from $(x + 1)^2 + 4$, the $1/(x^2+a^2)$ special form with $a = 2$.
  8. At $x = 0$: $1/(0-3) + (2 \cdot 0 - 1)/(0 + 0 + 5) = -1/3 - 1/5 = -8/15$, and the fraction at $x = 0$ is $8/((-3) \cdot 5) = -8/15$
    Check by putting $x = 0$ into the split and into the original fraction: both give the same value.
proper: split
(2x + 1)/((x − 1)(x + 2))3/(x² − 4)
improper: divide first
(x² + 1)/(x² − 1)x³/(x + 2)
Proper: the top’s degree is less than the bottom’s, so split. Improper: it is not, so divide first, then split what is left.

If the degree of the numerator is not less, the function is IMPROPER. Divide first. What is left over is proper, and you split that.

Your turn: split $1/((x - 1)(x + 1))$ into partial fractions. Do not integrate yet.

Worked example

Dividing first: $\int x^3/(x + 2) \mathrm{d} x$

  1. $x^3/(x + 2)$
    Improper: the top’s degree, 3, is not less than the bottom’s, 1. Divide first.
  2. $x^3 = (x + 2)(x^2 - 2 x + 4) - 8$
    Long division of $x^3$ by $x + 2$: the quotient is $x^2 - 2 x + 4$, the remainder $-8$.
  3. $x^3/(x + 2) = x^2 - 2 x + 4 - 8/(x + 2)$
    Divide both sides by $x + 2$. The polynomial part splits off; what is left, $-8/(x+2)$, is proper.
  4. $\int x^3/(x + 2) \mathrm{d} x = \int (x^2 - 2 x + 4) \mathrm{d} x - 8 \int 1/(x + 2) \mathrm{d} x$
    Integrate the polynomial part and the proper part term by term.
  5. *$\int x^3/(x + 2) \mathrm{d} x = x^3/3 - x^2 + 4 x - 8 \ln|x + 2| + C$*
    Each piece is a standard integral; the log rule gives the last one.
  6. $(x + 2)(x^2 - 2 x + 4) - 8 = x^3$
    Check the division: multiply the quotient back by $x + 2$ and subtract $8$; the cubic returns.
At x equals 0 the dashed curves read minus 2 and 0.5. They add to minus 1.5, the value of the full fraction.
Check yourself
  1. For distinct linear factors, $P(x)/((x-a)(x-b))$ splits into \_\_\_\_\_\_.
    1. $A/(x-a) - B/(x-b)$, always with a minus
    2. $A/(x-a) + B/(x-b)$
    3. $A/(x-a) \cdot B/(x-b)$
    4. $(A+B)/((x-a)(x-b))$
    Check your answer
    1. $A/(x-a) - B/(x-b)$, always with a minus — The sign between the two fractions is a plus; any needed minus is absorbed into the VALUE of $B$ itself, not built into the form.
    2. ✓ $A/(x-a) + B/(x-b)$ — (B) Distinct linear factors split into a SUM of simpler fractions, one per factor.
    3. $A/(x-a) \cdot B/(x-b)$ — The decomposition is a SUM of two fractions, not a product, multiplying them does not reconstruct $P(x)/((x-a)(x-b))$.
    4. $(A+B)/((x-a)(x-b))$ — This just recombines the numerator over the same denominator, it is not a decomposition into simpler pieces at all.
  2. After clearing denominators to get $P(x) = A(x-b)+B(x-a)$, why does substituting $x=a$ quickly give $A$?
    1. because $x=a$ makes both terms on the right vanish
    2. because substituting $x=a$ changes the value of $A$ itself
    3. because $x=a$ is the only value where the equation is true
    4. it zeroes the $B$-term, so only $A$ is left to solve for
    Check your answer
    1. because $x=a$ makes both terms on the right vanish — Only $B(x-a)$ vanishes at $x=a$, the $A(x-b)$ term survives as $A(a-b)$, which is exactly how $A$ gets isolated.
    2. because substituting $x=a$ changes the value of $A$ itself — $A$ is a single fixed constant, determined once and for all, substituting different values of $x$ only reveals its value, never changes it.
    3. because $x=a$ is the only value where the equation is true — $P(x)=A(x-b)+B(x-a)$ must hold for all $x$, substituting $x=a$ is just a convenient CHOICE that isolates $A$, not the only valid point.
    4. ✓ it zeroes the $B$-term, so only $A$ is left to solve for — (D) At $x=a$, the factor $(x-a)$ in $B(x-a)$ becomes zero, leaving only $A(a-b)$ on the right, solvable for $A$ directly.
  3. Once $P(x)/((x-1)(x+2)) = 2/(x-1) + 1/(x+2)$ is found, integrating gives \_\_\_\_\_\_.
    1. $2 \ln|x-1| + \ln|x+2| + C$
    2. $2 \ln|x-1| \cdot \ln|x+2| + C$
    3. $(x-1)^2 + (x+2) + C$
    4. $\ln|2(x-1)| + \ln|x+2| + C$
    Check your answer
    1. ✓ $2 \ln|x-1| + \ln|x+2| + C$ — (A) Each piece $A/(x-a)$ integrates straight from the standard table to $A \cdot \ln|x-a|$.
    2. $2 \ln|x-1| \cdot \ln|x+2| + C$ — The two integrated pieces are ADDED, since they came from a sum, multiplying them does not match the term-by-term integration.
    3. $(x-1)^2 + (x+2) + C$ — $1/(x-a)$ is the $n=-1$ case of the power rule, excluded from it, it integrates to a logarithm, not a power.
    4. $\ln|2(x-1)| + \ln|x+2| + C$ — $\int 2/(x-1) d x = 2 \ln|x-1|$, with the $2$ OUTSIDE the log, folding it inside the absolute value changes the expression.
  4. A rational function has numerator degree 3 and denominator degree 2. What must happen before partial fractions apply?
    1. the numerator must be reduced to degree 2 by dropping its highest term
    2. the denominator must be factored into three linear factors
    3. nothing: partial fractions apply directly regardless of degree
    4. long division first, to leave a proper remainder
    Check your answer
    1. the numerator must be reduced to degree 2 by dropping its highest term — Dropping the numerator’s highest term changes the function into a different one, long division extracts a polynomial part correctly, without altering the value.
    2. the denominator must be factored into three linear factors — The denominator’s degree being 2 does not force three linear factors, the actual issue here is the numerator’s degree, not the factor count.
    3. nothing: partial fractions apply directly regardless of degree — Partial fractions in the standard form apply only to a PROPER function, numerator degree 3 against denominator degree 2 is improper and needs long division first.
    4. ✓ long division first, to leave a proper remainder — (D) Numerator degree at least denominator degree makes the function IMPROPER, which needs long division first, before the remainder is proper enough for partial fractions.
2/(x − 1) + 1/(x + 2)
→
add
over (x − 1)(x + 2)
→
(3x + 3)/((x − 1)(x + 2))
The check. Add the two pieces back over (x − 1)(x + 2) and you get the integrand, (3x + 3)/((x − 1)(x + 2)).
Worked example

Partial fractions, term by term: $\int (3x+3)/((x-1)(x+2)) \mathrm{d} x$

  1. Find $\int (3x+3)/((x-1)(x+2)) \mathrm{d} x$
    The numerator has degree 1 and the denominator degree 2, so the fraction is proper. The two factors are different.
  2. $(3x+3)/((x-1)(x+2)) = A/(x-1) + B/(x+2)$
    Write one fraction for each factor.
  3. $3x+3 = A(x+2) + B(x-1)$
    Multiply both sides by $(x - 1)(x + 2)$.
  4. At $x=1$: $6=3A$, so $A=2$. At $x=-2$: $-3=-3B$, so $B=1$
    Each root makes one bracket zero. So each leaves one unknown.
  5. $\int [2/(x-1) + 1/(x+2)] \mathrm{d} x = 2 \ln|x-1| + \ln|x+2| + C$
    Each piece is a log formula from the standard table.
  6. Check: $2/(x-1)+1/(x+2) = (2(x+2)+(x-1))/((x-1)(x+2)) = (3x+3)/((x-1)(x+2))$
    Check by adding the two fractions back. You get the integrand.
Check yourself
  1. Substituting $x=1$ into $3x+3=A(x+2)+B(x-1)$ gives $6=3A$, so $A$ equals \_\_\_\_\_\_.
    1. $1/2$
    2. $3$
    3. $2$
    4. $6$
    Check your answer
    1. $1/2$ — $A=6/3=2$, dividing $3$ by $6$ instead inverts the correct computation.
    2. $3$ — $6=3A$ still needs dividing by $3$ to isolate $A$, reporting $3$ skips that division.
    3. ✓ $2$ — (C) $6=3A$ divides through to give $A=2$.
    4. $6$ — $6$ is the value BEFORE dividing by $3$, $A$ itself is $6/3=2$, not $6$.
  2. Substituting $x=-2$ into $3x+3=A(x+2)+B(x-1)$ gives $-3=-3B$, so $B$ equals \_\_\_\_\_\_.
    1. $-1$
    2. $3$
    3. $-3$
    4. $1$
    Check your answer
    1. $-1$ — $-3/-3=1$, a positive value, the two negatives cancel, so the sign should not remain negative.
    2. $3$ — The equation is $-3=-3B$, dividing both sides by $-3$ (not $3$) is what correctly gives $B=1$.
    3. $-3$ — $-3$ is the value BEFORE dividing by $-3$, $B$ itself is $-3/-3=1$, not $-3$.
    4. ✓ $1$ — (D) $-3=-3B$ divides through to give $B=1$.
  3. With $A=2$, $B=1$, $\int (3x+3)/((x-1)(x+2)) d x$ equals \_\_\_\_\_\_.
    1. $\ln|x-1| + \ln|x+2| + C$
    2. $2 \ln|x-1| + \ln|x+2| + C$
    3. $\ln|2(x-1)(x+2)| + C$
    4. $2 \ln|x-1| \cdot \ln|x+2| + C$
    Check your answer
    1. $\ln|x-1| + \ln|x+2| + C$ — The first piece is $2/(x-1)$, which integrates to $2 \ln|x-1|$, dropping the $2$ loses that coefficient.
    2. ✓ $2 \ln|x-1| + \ln|x+2| + C$ — (B) Each piece integrates separately, keeping its own coefficient: $2 \ln|x-1| + \ln|x+2| + C$.
    3. $\ln|2(x-1)(x+2)| + C$ — Combining everything into a single $\ln$ of a product loses the coefficient $2$ that belongs only to the first term.
    4. $2 \ln|x-1| \cdot \ln|x+2| + C$ — The two pieces came from a SUM, $2/(x-1)+1/(x+2)$, so their integrals are added, not multiplied.

↑ Back to top

Definite integrals by substitution

▲CONCEPT
x
t
With t = x² the limits change too. x = 0 gives t = 0, and x = 2 gives t = 4.

So far a definite integral took two steps: find the integral, then put in the limits. When the integral needs a substitution, you have a choice. The long way: substitute, integrate in $t$, put $x$ back, and then put in the limits in $x$. That is the substitution section’s integral, finished with the Fundamental Theorem. The short way: change the limits to $t$ at the moment you substitute, and never come back to $x$. The board’s answers use the short way. Here it is.

Put $t = g(x)$. Then $\mathrm{d} t = g'(x) \mathrm{d} x$, as in the substitution section. Change each limit with $t$ too:

$x = a$ gives $t = g(a)$.

$x = b$ gives $t = g(b)$.

*Put in $t$*, $\mathrm{d} t$ and the new limits:

$\int_a^b f(g(x)) \cdot g'(x) \mathrm{d} x = \int_{g(a)}^{g(b)} f(t) \mathrm{d} t$.

After that you work in $t$ alone. There is no need to go back to $x$.

2x eˣ²
tyeᵗ
Under 2x times e to the x squared, from 0 to 2, and under eᵗ, from 0 to 4, each area is e⁴ − 1.

Let us find $\int_0^2 2 x e^{x^2} \mathrm{d} x$ the short way.

Worked example

Find $\int_0^2 2 x e^{x^2} \mathrm{d} x$ with new limits

  1. $t = x^2$, so $\mathrm{d} t = 2 x \mathrm{d} x$
    The derivative of $x^2$ is $2 x$, and $2 x \mathrm{d} x$ appears in the integral.
  2. $x = 0$ gives $t = 0$, and $x = 2$ gives $t = 4$
    Change each limit with $t = x^2$.
  3. $\int_0^2 2 x e^{x^2} \mathrm{d} x = \int_0^4 e^t \mathrm{d} t$
    Put in $t$, $\mathrm{d} t$ and the new limits: now everything is in $t$.
  4. $\int_0^4 e^t \mathrm{d} t = [e^t]_0^4 = e^4 - e^0$
    The integral of $e^t$ is $e^t$, at 4 minus at 0.
  5. *$\int_0^2 2 x e^{x^2} \mathrm{d} x = e^4 - 1$*
    Any number to the power 0 is 1.
tyeᵗ
tyeᵗ
Keep the old limits and you find the area under eᵗ from 0 to 2: e² − 1 ≈ 6.39, not e⁴ − 1.

If you keep the old limits 0 and 2 with $t$, you get $e^2 - 1$. That is wrong, because $t$ runs from 0 to 4. Check with $F(x) = e^{x^2}$ in $x$. It gives $e^4 - e^0$, the same $e^4 - 1$.

Your turn: find $\int_1^2 2 x (x^2 - 1)^3 \mathrm{d} x$ with $t = x^2 - 1$. What are the new limits?

Check yourself
  1. Substituting $t=g(x)$ in a definite integral $\int_a^b f(g(x)) \cdot g'(x) d x$ changes \_\_\_\_\_\_.
    1. only the variable, never the limits
    2. neither the variable nor the limits, since the numeric answer is unaffected either way
    3. only the limits, never the variable
    4. both the variable and the limits, together, at the same step
    Check your answer
    1. only the variable, never the limits — Changing only the variable while leaving the OLD $x$-limits in place is the standard slip, the limits must move to $g(a)$ and $g(b)$ at the same step.
    2. neither the variable nor the limits, since the numeric answer is unaffected either way — A valid substitution requires converting both the variable and the limits, denying any change is needed abandons the procedure altogether.
    3. only the limits, never the variable — Both the variable and the limits move together, changing only the limits while keeping $f(g(x)) \cdot g'(x)$ unconverted leaves the integral half-substituted.
    4. ✓ both the variable and the limits, together, at the same step — (D) The limits move to $t=g(a)$ and $t=g(b)$ at the same step the variable itself changes, there is no going back to $x$ afterward.
  2. For $\int_a^b f(g(x)) \cdot g'(x) d x$ with $t=g(x)$, the new limits are \_\_\_\_\_\_.
    1. $t=a$ and $t=b$, reusing the same numbers
    2. still $x=a$ and $x=b$, unchanged
    3. $t=g(b)$ for the lower limit and $t=g(a)$ for the upper limit
    4. $t=g(a)$ for the lower limit and $t=g(b)$ for the upper limit
    Check your answer
    1. $t=a$ and $t=b$, reusing the same numbers — The new limits are $g(a)$ and $g(b)$, not the bare numbers $a$ and $b$ themselves, $g$ must be applied to each old limit.
    2. still $x=a$ and $x=b$, unchanged — The limits must be converted to $t$-values through $g$, leaving them as $x=a$ and $x=b$ ignores the substitution entirely.
    3. $t=g(b)$ for the lower limit and $t=g(a)$ for the upper limit — The LOWER limit becomes $g(a)$ and the UPPER limit becomes $g(b)$, matching the original order, not swapped.
    4. ✓ $t=g(a)$ for the lower limit and $t=g(b)$ for the upper limit — (D) The lower limit becomes $g(a)$ and the upper limit becomes $g(b)$, each old limit passed through $g$.
  3. Once the limits are changed to $t=g(a)$ and $t=g(b)$, what happens at the end of a definite-integral substitution?
    1. the constant of integration must be evaluated using the new limits
    2. it finishes entirely in $t$; the old $x$-limits play no further role
    3. the integral in $t$ must be converted back to $x$ before evaluating
    4. the original $x$-limits must be plugged back in alongside the $t$-limits
    Check your answer
    1. the constant of integration must be evaluated using the new limits — A definite integral’s answer is a number with no constant of integration attached, there is no $C$ left to evaluate using the new limits.
    2. ✓ it finishes entirely in $t$; the old $x$-limits play no further role — (B) With the limits already converted to $t$, the whole evaluation finishes in $t$ , no return to $x$ is needed.
    3. the integral in $t$ must be converted back to $x$ before evaluating — An indefinite integral must return to $x$ since no numeric limits exist; a definite integral, with its limits already in $t$, needs no such return.
    4. the original $x$-limits must be plugged back in alongside the $t$-limits — Once the limits are converted to $t$, the old $x$-limits play no further role, plugging them back in mixes two different variables’ bounds.
x
t
With t = x² + 1, x = 0 and 1 give t = 1 and 2. The work stays in t.
2x√(x² + 1)
ty√t
Under 2x√(x² + 1) from 0 to 1, and under √t from 1 to 2: the same area, (2/3)(2√2 − 1) ≈ 1.219.
Worked example

A definite integral by substitution: $\int_0^1 2x \sqrt{x^2+1} \mathrm{d} x$

  1. Find $\int_0^1 2x \sqrt{x^2+1} \mathrm{d} x$
    Substitute, and change the limits along with the variable.
  2. $t = x^2+1$, $\mathrm{d} t = 2x \mathrm{d} x$; when $x=0$, $t=1$; when $x=1$, $t=2$
    Work out $t$ at each limit.
  3. $\int_0^1 2x \sqrt{x^2+1} \mathrm{d} x = \int_1^2 \sqrt{t} \mathrm{d} t$
    Now everything is in $t$, the limits too.
  4. $\int_1^2 \sqrt{t} \mathrm{d} t = [2/3 \cdot t^{3/2}]_1^2 = 2/3 \cdot (2 \sqrt{2} - 1)$
    Use the power rule in $t$. Then subtract its value at $t = 1$ from its value at $t = 2$.
  5. Check against the earlier integral $2/3 \cdot (x^2+1)^{3/2}$: at $x=1$ it gives $2/3 \cdot 2 \sqrt{2}$, at $x=0$ it gives $2/3$
    Check with the integral in $x$ from the first worked example. The difference is the same.
Check yourself
  1. With $t=x^2+1$, the limits for $\int_0^1 2x \sqrt{x^2+1} d x$ become \_\_\_\_\_\_.
    1. $t=1$ to $t=0$, reversed
    2. $t=0$ to $t=2$
    3. $t=0$ to $t=1$, unchanged
    4. $t=1$ to $t=2$
    Check your answer
    1. $t=1$ to $t=0$, reversed — The lower limit $x=0$ gives $t=1$, and the upper limit $x=1$ gives $t=2$, the order should match, not reverse.
    2. $t=0$ to $t=2$ — At $x=0$, $t=0^2+1=1$, not $0$, the lower limit in $t$ is $1$.
    3. $t=0$ to $t=1$, unchanged — The limits must be converted through $t=x^2+1$, at $x=0$ this gives $t=1$, not $t=0$.
    4. ✓ $t=1$ to $t=2$ — (D) At $x=0$, $t=0^2+1=1$; at $x=1$, $t=1^2+1=2$.
  2. $\int_1^2 \sqrt{t} d t = [2/3 \cdot t^{3/2}]_1^2$ equals \_\_\_\_\_\_.
    1. $2/3 \cdot 1^{3/2}$, using only the lower limit
    2. $2/3 \cdot 2^{3/2}$, using only the upper limit
    3. $2/3 \cdot (1^{3/2} - 2^{3/2})$, limits reversed
    4. $2/3 \cdot (2^{3/2} - 1)$
    Check your answer
    1. $2/3 \cdot 1^{3/2}$, using only the lower limit — The upper-limit value must be included and the lower-limit value subtracted from it, the lower term alone is not the answer.
    2. $2/3 \cdot 2^{3/2}$, using only the upper limit — FTC needs both endpoint values subtracted, the upper-limit term alone is only half the computation.
    3. $2/3 \cdot (1^{3/2} - 2^{3/2})$, limits reversed — The upper limit’s value is subtracted FROM, not subtracted BY, the lower limit’s value, $2^{3/2}-1^{3/2}$, not the reverse.
    4. ✓ $2/3 \cdot (2^{3/2} - 1)$ — (D) Evaluating at the upper limit $t=2$ and subtracting the value at the lower limit $t=1$ gives $2/3.(2^{3/2}-1)$.
  3. Why is $2/3 \cdot (2 \sqrt{2} - 1)$ the final answer, with no further step needed?
    1. it is already a number, as the limits were changed to $t$
    2. because the answer must still be rewritten as a function of $x$
    3. because the constant of integration still needs to be added
    4. because $t$ happens to equal $x$ at both limits
    Check your answer
    1. ✓ it is already a number, as the limits were changed to $t$ — (A) With the limits already converted to $t$, the evaluation is complete once a number comes out, no return to $x$ is needed for a definite integral.
    2. because the answer must still be rewritten as a function of $x$ — Returning to $x$ is the INDEFINITE-integral habit, once the limits are converted to $t$, a definite integral finishes entirely in $t$, with no return needed.
    3. because the constant of integration still needs to be added — A definite integral evaluates to a plain number with no constant of integration attached, there is no $C$ left to add.
    4. because $t$ happens to equal $x$ at both limits — $t=1,2$ at the limits while $x=0,1$, $t$ and $x$ are different numbers at both endpoints, so no such coincidence is at work.

↑ Back to top

Properties of definite integrals

▲CONCEPT
The same region under x squared over 3, from 0 to 3, area 3. Forwards the integral is 3. Backwards it is minus 3.

A few properties make some definite integrals easy, often without finding an integral at all. In each one, let $f$ be continuous on the interval.

y = x²
Split [0, 3] at 1. Under x² there is 1/3 from 0 to 1 and 26/3 from 1 to 3. Together they make 9.
y = x²
y = x³
x² is even: two halves of 8/3 make 16/3. x³ is odd: −4 and 4 make 0.
Left: the V of absolute x, two halves of 2, total 4. Right: y equals x, pieces of 2 and minus 2, total 0.

Check the even and odd rule with numbers. $x^2$ is even. From 0 to 2,

$\int_0^2 x^2 \mathrm{d} x = 8/3$.

The piece on $[-2, 0]$ is its mirror image, also $8/3$, so the whole is twice as much:

$\int_{-2}^2 x^2 \mathrm{d} x = 16/3$.

*Now $x^3$*, which is odd. Its piece on $[-2, 0]$ is $-4$ and its piece on $[0, 2]$ is 4, so they cancel:

$\int_{-2}^2 x^3 \mathrm{d} x = 0$.

y = sin x
y = cos x
On [0, π/2], putting π/2 − x for x turns sin x into cos x. The two areas are equal, 1 each.
Left: x squared over 4 on 0 to 3, area 2.25. Right: its mirror image, area 2.25. Both curves pass 0.5625 at 1.5.

The reflection rule helps when $f(a + b - x)$ is simpler, or pairs well with $f(x)$. On $[0, \pi/2]$, put $\pi/2 - x$ in place of $x$:

$\sin(\pi/2 - x) = \cos x$.

So the rule turns $\sin x$ into $\cos x$. The two have the same integral on $[0, \pi/2]$.

$\int_0^{\pi/2} \sin x \mathrm{d} x = \int_0^{\pi/2} \cos x \mathrm{d} x$.

Each one is 1. An integral of $\sin x$ is $-\cos x$, which is 0 at $\pi/2$ and $-1$ at 0, and $0 - (-1)$ is 1.

Worked example

The reflection move, added to itself: $\int_0^{\pi/2} \sin x/(\sin x + \cos x) \mathrm{d} x$

  1. Let $I = \int_0^{\pi/2} \sin x/(\sin x + \cos x) \mathrm{d} x$
    Name the integral $I$, so it can be combined with its own reflection.
  2. $I = \int_0^{\pi/2} \cos x/(\cos x + \sin x) \mathrm{d} x$
    Replace $x$ by $\pi/2 - x$: $\sin(\pi/2 - x) = \cos x$ and $\cos(\pi/2 - x) = \sin x$, and the limits stay $0$ to $\pi/2$.
  3. $I + I = \int_0^{\pi/2} (\sin x + \cos x)/(\sin x + \cos x) \mathrm{d} x$
    Add the two equal forms of $I$: the denominators match, so the fractions combine into one.
  4. *$2 I = \int_0^{\pi/2} 1 \mathrm{d} x = \pi/2$*
    The integrand is $1$ everywhere on $[0, \pi/2]$, so the integral is just the length of the interval.
  5. $I = \pi/4$
    Divide by $2$.
  6. $f(x) + f(\pi/2 - x) = 1$ for every $x$ in $[0, \pi/2]$
    Step 2 gives this for every $x$, not only at one point, so the average height over the interval is $1/2$: the area is $(1/2)(\pi/2) = \pi/4$, matching.
y = |sin x|
|sin x| repeats every π, and each arch has area 2. So from 0 to 3π the area is 3 × 2 = 6.

One more is true, though NCERT does not list it. If $f$ repeats every $p$, then $\int_0^{n p} f(x) \mathrm{d} x = n \int_0^p f(x) \mathrm{d} x$.

Your turn: find $\int_{-2}^2 x^2 \sin x \mathrm{d} x$ without integrating. Is $x^2 \sin x$ even or odd?

Check yourself
  1. The reversal property states $\int_a^b f(x) d x$ equals \_\_\_\_\_\_.
    1. $-\int_b^a f(x) d x$
    2. $\int_b^a f(x) d x$, with no sign change
    3. $2 \int_b^a f(x) d x$
    4. $\int_a^b f(x) d x$ unchanged, since swapping limits changes nothing
    Check your answer
    1. ✓ $-\int_b^a f(x) d x$ — (A) Swapping the limits flips the sign of the definite integral.
    2. $\int_b^a f(x) d x$, with no sign change — Swapping the limits changes the sign, reporting $\int_b^a f(x) d x$ with no minus sign misses exactly that flip.
    3. $2 \int_b^a f(x) d x$ — The reversal property only flips the sign, no factor of $2$ enters when the limits are simply swapped.
    4. $\int_a^b f(x) d x$ unchanged, since swapping limits changes nothing — Swapping the limits does change the value, it flips the sign, from $\int_a^b$ to $-\int_b^a$.
  2. The reflection property, $\int_a^b f(x) d x = \int_a^b f(a+b-x) d x$, is useful when \_\_\_\_\_\_.
    1. the interval $[a,b]$ is centered at zero
    2. $f(a+b-x)$ simplifies against $f(x)$
    3. the integrand is a constant function
    4. $f$ has no useful symmetry at all
    Check your answer
    1. the interval $[a,b]$ is centered at zero — The reflection property works on any interval $[a,b]$, not just one centered at zero, that centering condition belongs to a different property, the symmetric-interval rule.
    2. ✓ $f(a+b-x)$ simplifies against $f(x)$ — (B) The reflection is worth applying precisely when replacing $x$ by $a+b-x$ makes the integrand combine or simplify nicely against the original.
    3. the integrand is a constant function — A constant function integrates the same way with or without the reflection, the property earns its keep on integrands where the reflection actually simplifies something.
    4. $f$ has no useful symmetry at all — The reflection property is a tool for exploiting a symmetry in $f$, claiming it helps when there is NO such symmetry gets the property backwards.
  3. For $f$ ODD on $[-a,a]$, $\int_{-a}^a f(x) d x$ equals \_\_\_\_\_\_.
    1. $0$
    2. $\int_0^a f(x) d x$
    3. $2 \int_0^a f(x) d x$
    4. $a$
    Check your answer
    1. ✓ $0$ — (A) An odd function’s symmetric-interval integral vanishes completely, the negative and positive halves cancel exactly.
    2. $\int_0^a f(x) d x$ — Halving the even-function formula still uses the wrong rule for an odd function, the odd case is not a fraction of the even one, it is $0$.
    3. $2 \int_0^a f(x) d x$ — $2 \int_0^a f(x) d x$ is the rule for an EVEN $f$, an ODD $f$ instead integrates to exactly $0$ on a symmetric interval.
    4. $a$ — Neither the even nor the odd symmetric-interval rule produces a bare value of $a$, the odd case gives exactly $0$.
  4. If $f$ has period $p$, the periodicity property gives $\int_0^{n p} f(x) d x$ as \_\_\_\_\_\_.
    1. $\int_0^p f(x) d x$, regardless of $n$
    2. $p \int_0^n f(x) d x$
    3. $n \cdot p \cdot \int_0^1 f(x) d x$
    4. $n \int_0^p f(x) d x$
    Check your answer
    1. $\int_0^p f(x) d x$, regardless of $n$ — Covering $n$ periods means $n$ copies of the same one-period integral are added, dropping $n$ ignores how many periods are actually spanned.
    2. $p \int_0^n f(x) d x$ — The one-period integral runs from $0$ to $p$, scaled by the period COUNT $n$, swapping these roles integrates over the wrong range.
    3. $n \cdot p \cdot \int_0^1 f(x) d x$ — The property is built entirely from the one-period integral over $[0,p]$, an integral over $[0,1]$ has no role in it unless $p=1$.
    4. ✓ $n \int_0^p f(x) d x$ — (D) Each full period contributes the same integral, so $n$ periods give $n$ times the one-period integral.

↑ Back to top

Every method is the same question, asked differently

■RECAP
Integrals + C Integrals come in families The definite integral as a limit of totals A(x) + f(x)h The fundamental theorem of calculus d dx ∫ The standard integrals table g f t t = g(x) dt = g′(x) dx Integration by substitution √(a² − x²) A toolkit of special-form integrals u dw w du Integration by parts, and when to skip it + Partial fractions before integrating 0 1 x 1 2 t Definite integrals by substitution Properties of definite integrals
Ten idea cards: finding a family on top, one number below, four methods between. An arrow means the later idea uses the earlier.

The question at the start asked if you can go back, from the rate to the quantity.

2x cos(x²)
→
t = x²
→
sin(x²) + C
x sin x
→
by parts
→
−x cos x + sin x + C
(2x + 1)/((x − 1)(x + 2))
→
split
→
ln|x − 1| + ln|x + 2| + C
Three integrals from the chapter, each a Chapter 5 rule run backwards: substitution, by parts, and adding fractions undone.
a family
tWC = 100C = 300
16 strips
F(2) − F(1)
Answered: going back gives a family, 20t + C. A definite integral is a limit of totals of strips, and it equals F(b) − F(a), here 4.

Here is the answer. Yes, you can go back, but you get a family of functions. There is one for every starting amount, and that is the $+ C$. Each method is a rule of differentiation read backwards. Substitution is the chain rule, by parts is the product rule, and partial fractions is adding fractions undone. The total change between two moments is the area under the graph of the rate. It is the limit of totals of thin strips. The fundamental theorem says it equals $F(b) - F(a)$, so the $C$ cancels.

Check yourself
  1. Every technique in this chapter answers one central question. What is it?
    1. which derivative rule, run backwards, fits it
    2. whether the function is continuous on the interval
    3. how many terms the integrand contains
    4. which numeric value the integral evaluates to
    Check your answer
    1. ✓ which derivative rule, run backwards, fits it — (A) Every method, table, substitution, by parts, partial fractions, is really asking which derivative rule, reversed, matches the integrand.
    2. whether the function is continuous on the interval — Continuity is not the question this chapter’s techniques are built around, matching the integrand to a reversed derivative rule is.
    3. how many terms the integrand contains — The number of terms in an integrand says nothing about which technique applies, recognizing the SHAPE against a derivative rule is what matters.
    4. which numeric value the integral evaluates to — A numeric value is what a DEFINITE integral eventually produces, the recognition question that decides HOW to get there is about matching a derivative rule, not the resulting number.
  2. When no technique matches an integrand directly, the next step is to \_\_\_\_\_\_.
    1. give up and treat the integral as impossible to evaluate
    2. guess an answer and check it by differentiating, with no other work
    3. always default to integration by parts, regardless of shape
    4. transform it: try substitution, by parts, or partial fractions
    Check your answer
    1. give up and treat the integral as impossible to evaluate — The method is the opposite: keep transforming the integral until a technique matches, rather than declaring it unsolvable.
    2. guess an answer and check it by differentiating, with no other work — Transform the integral first, by substitution, by parts or partial fractions, rather than guessing an answer with no other work.
    3. always default to integration by parts, regardless of shape — Different integrand shapes call for different transforms, always defaulting to by-parts ignores cases better suited to substitution or partial fractions.
    4. ✓ transform it: try substitution, by parts, or partial fractions — (D) The method is to keep transforming the integral, by substitution, by parts, or by partial fractions, until a match appears.
  3. A definite integral adds one more question on top of finding an integral. What is it?
    1. whether the interval $[a,b]$ is long enough
    2. whether it’s signed value or true area
    3. whether substitution or by-parts was used
    4. whether the integral found is unique
    Check your answer
    1. whether the interval $[a,b]$ is long enough — Interval length has no bearing on the extra question. It is about sign versus area, not length.
    2. ✓ whether it’s signed value or true area — (B) Sign versus area is the extra layer a definite integral adds on top of finding the integral.
    3. whether substitution or by-parts was used — Which technique found the integral is settled before the definite integral is even evaluated. The extra question comes after, about sign versus area.
    4. whether the integral found is unique — Integral non-uniqueness is a fact about INDEFINITE integrals, already handled by the constant of integration. The extra question for DEFINITE integrals is about sign versus area.
Check yourself: the whole chapter
  1. A student proposes $\int 3x^2 d x = x^3 + C$. What is the correct way to check whether this is right?
    1. Integrate $x^3 + C$ again and see if the result matches $3x^2$.
    2. Substitute one value of $x$ into both sides and check if the numbers agree.
    3. Compare $x^3 + C$ to a memorised list of similar integrals and see if it looks right.
    4. Differentiate $x^3 + C$ and see if the result matches $3x^2$.
    Check your answer
    1. Integrate $x^3 + C$ again and see if the result matches $3x^2$. — Integrating $x^3+C$ again moves further from the original problem, not back to it. The check must run the other way: differentiate.
    2. Substitute one value of $x$ into both sides and check if the numbers agree. — One matching value does not prove two functions agree everywhere. Differentiating checks the rate at every $x$, all at once.
    3. Compare $x^3 + C$ to a memorised list of similar integrals and see if it looks right. — Resemblance to memory is not proof. Differentiating $x^3+C$ gives $3x^2$ exactly, which settles the question without guessing.
    4. ✓ Differentiate $x^3 + C$ and see if the result matches $3x^2$. — (D) Differentiate the proposed answer. $d/d x (x^3+C) = 3x^2$, exactly the original integrand, so the integral is confirmed correct.
  2. Which of these correctly completes the standard integrals table for $\int \sin x d x$?
    1. $-\cos x + C$, since differentiating $-\cos x$ gives $\sin x$.
    2. $\cos x + C$, since integrating undoes the derivative-of-sin rule directly.
    3. $-\sin x + C$, since a sign flip belongs somewhere in the answer.
    4. $\sin x + C$, since integrating a trig function leaves it unchanged.
    Check your answer
    1. ✓ $-\cos x + C$, since differentiating $-\cos x$ gives $\sin x$. — (A) Differentiating $-\cos x$ gives $\sin x$, so $\int \sin x d x = -\cos x + C$. Check any table entry the same way, by differentiating the right side.
    2. $\cos x + C$, since integrating undoes the derivative-of-sin rule directly. — $\cos x$ is the derivative of $\sin x$, not its integral. Differentiating $-\cos x$ recovers $\sin x$, so $-\cos x + C$ is the integral instead.
    3. $-\sin x + C$, since a sign flip belongs somewhere in the answer. — The sign is right, but the function is not. Differentiating $-\sin x$ gives $-\cos x$, not $\sin x$, so $-\sin x$ cannot be the integral.
    4. $\sin x + C$, since integrating a trig function leaves it unchanged. — Integrating does change the function here. Differentiating $\sin x$ gives $\cos x$, not $\sin x$, so $\sin x$ cannot be its own integral.
  3. For $\int x \cos(x^2) d x$, take $t = x^2$, so $d t = 2x d x$. But the integrand only has $x d x$, not $2x d x$. What should be done?
    1. Substitution cannot be used here, since the exact derivative $2x d x$ is missing.
    2. Replace $x d x$ with $d t$ directly, treating the missing factor of $2$ as unimportant.
    3. Write $x d x = (1/2) d t$, and carry the constant $1/2$ through the integral.
    4. Insert an extra factor of $2$ into the original integrand to make the substitution exact.
    Check your answer
    1. Substitution cannot be used here, since the exact derivative $2x d x$ is missing. — A missing constant multiple does not block substitution. Solve $d t = 2x d x$ for $x d x$ instead, giving $x d x = (1/2) d t$.
    2. Replace $x d x$ with $d t$ directly, treating the missing factor of $2$ as unimportant. — Dropping the $1/2$ changes the value of the integral. A constant factor must be carried through exactly, not ignored.
    3. ✓ Write $x d x = (1/2) d t$, and carry the constant $1/2$ through the integral. — (C) From $d t = 2x d x$, solve $x d x = (1/2) d t$. The integral becomes $(1/2) \int \cos t d t$, carrying the constant through cleanly.
    4. Insert an extra factor of $2$ into the original integrand to make the substitution exact. — The original integrand is not the problem, it is already correct. The adjustment belongs in how $x d x$ is rewritten in terms of $d t$.
  4. Integration by parts subtracts a second integral: $\int u \cdot v d x = u \int v d x - \int (d/(d x) (u) \cdot \int v d x) d x$. For $\int x \cdot e^x d x$ with $u = x$ and $v = e^x$, a student ends with $x \cdot e^x + e^x + C$ instead of $x \cdot e^x - e^x + C$. What went wrong?
    1. The choice of $u$ was wrong, since $e^x$ should have been differentiated instead of $x$.
    2. The minus sign in the formula became a plus sign somewhere in the working.
    3. $e^x$ was never actually integrated before the terms were combined.
    4. Nothing is wrong. $+e^x$ and $-e^x$ differ only by a constant that $+C$ absorbs.
    Check your answer
    1. The choice of $u$ was wrong, since $e^x$ should have been differentiated instead of $x$. — $u=x$ is the correct choice. Differentiating it gives $1$, exactly what ILATE expects. The error is the sign in front of the second integral, not the choice of $u$.
    2. ✓ The minus sign in the formula became a plus sign somewhere in the working. — (B) By parts subtracts the second integral. Writing $+e^x$ instead of $-e^x$ flips that sign, so the correct answer is $x \cdot e^x - e^x + C$.
    3. $e^x$ was never actually integrated before the terms were combined. — $e^x$ is integrated correctly in both versions, it appears in each. Only the sign attached to it is different.
    4. Nothing is wrong. $+e^x$ and $-e^x$ differ only by a constant that $+C$ absorbs. — $2e^x$ is a function of $x$, not a fixed number. Two answers differing by a non-constant function are genuinely different, and $+C$ cannot repair that.
  5. Using the second fundamental theorem on $\int_0^3 2x d x$ with integral $F(x) = x^2$, a student writes the answer as $F(0) - F(3) = 0 - 9 = -9$. What went wrong?
    1. Nothing. The second fundamental theorem allows either subtraction order, since the constant of integration cancels regardless.
    2. The integral itself is wrong, since it should be $F(x) = 2x^2$, not $x^2$.
    3. The limits $0$ and $3$ should have been swapped in the original integral, not in the subtraction.
    4. The order was reversed. The second fundamental theorem requires $F(3) - F(0) = 9$, the upper limit first.
    Check your answer
    1. Nothing. The second fundamental theorem allows either subtraction order, since the constant of integration cancels regardless. — The constant of integration cancels regardless of order, but the endpoint order does not. Reversing $F(b)-F(a)$ to $F(a)-F(b)$ flips the sign of the whole answer.
    2. The integral itself is wrong, since it should be $F(x) = 2x^2$, not $x^2$. — Differentiating $x^2$ gives $2x$, exactly the integrand. The integral is correct. Only the subtraction order was reversed.
    3. The limits $0$ and $3$ should have been swapped in the original integral, not in the subtraction. — The integral $\int_0^3 2x d x$ is stated correctly, with $0$ as the lower limit. The error sits only in which value is subtracted from which.
    4. ✓ The order was reversed. The second fundamental theorem requires $F(3) - F(0) = 9$, the upper limit first. — (D) The second fundamental theorem always takes upper limit minus lower limit: $F(3)-F(0) = 9-0=9$. Reversing this flips the sign to $-9$, which is wrong.

↑ Back to top

Where you will meet this

Each one below is the same idea you already use, just named differently.

Your turn: A café’s cash box starts the day with ₹⁠$500$ and takes in ₹⁠$200$ every hour it is open. How much is in the cash box after $5$ hours?

LATER

Where integrals go next

First-year engineering mathematics extends this chapter’s integration to functions of two or more variables, as a multiple integral taken over a region instead of an interval. It also meets the Laplace transform, which integrates a function against $e^{-s t}$ to turn a differential equation into an algebraic one. Both methods build on the single-variable integral this chapter sets up.

↑ Back to top

Practice set: Exercise 7.1

Exercise 7.1 practises the formulas of the standard table. Some questions need a little algebra first, such as opening a bracket or splitting a fraction. Check each answer by differentiating it.

Exercise 7.1: Standard integrals
  1. practice Find $\int \sin 5x \mathrm{d} x$.
  2. practice Find $\int \cos 7x \mathrm{d} x$.
  3. practice Find $\int e^{4x} \mathrm{d} x$.
  4. practice Find $\int x^6 \mathrm{d} x$.
  5. practice Which of these equals $\int 3/x \mathrm{d} x$?
    1. $3/x^2 + C$
    2. $-3/x^2 + C$
    3. $3 \ln x + C$
    4. $3 \ln|x| + C$
  6. practice Find $\int \sec^2 3x \mathrm{d} x$.
  7. practice Find $\int \text{cosec}^2 x \mathrm{d} x$.
  8. practice Find $\int \sec x \tan x \mathrm{d} x$.
  9. practice Find $\int \text{cosec} x \cot x \mathrm{d} x$.
  10. practice Find $\int 4 e^x \mathrm{d} x$.
  11. practice Find $\int (3 \sin x - 2 \cos x) \mathrm{d} x$.
  12. practice Find $\int (2x - 5)^2 \mathrm{d} x$.
  13. practice Find $\int x^2 (x^3 + 1) \mathrm{d} x$.
  14. practice Find $\int (x^2 + 3x + 2)/x \mathrm{d} x$.
  15. practice Find $\int (x^3 - 2x + 1)/x^2 \mathrm{d} x$.
  16. practice Find $\int \text{cosec} x (\text{cosec} x - \cot x) \mathrm{d} x$.
  17. practice Find $\int (x+1)(x+2) \mathrm{d} x$.
  18. practice Find $\int (\sqrt{x} + 1/\sqrt{x})^2 \mathrm{d} x$.
  19. practice Find $\int (x^3 - 3x^2 + 5)/\sqrt{x} \mathrm{d} x$.
  20. practice Find $\int (2x+3)(x-1) \mathrm{d} x$.
  21. practice Which of these equals $\int (x^4 + 1/x^4) \mathrm{d} x$?
    1. $x^5/5 + 1/(3x^3) + C$
    2. $x^5/5 - 1/(3x^3) + C$
    3. $x^5/5 - 3/x^3 + C$
    4. $4x^3 - 4/x^5 + C$
  22. practice If $d y/d x = 4x^3 - 3/x^4$ and $y = 2$ when $x = 1$, find $y$ as a function of $x$.
  23. practice Find $\int 3/(x \sqrt{x^2 - 1}) \mathrm{d} x$, for $x > 1$.
Answers
  1. $-(1/5) \cos 5x + C$
  2. $(1/7) \sin 7x + C$
  3. $(1/4) e^{4x} + C$
  4. $x^7/7 + C$
  5. (D) $3 \ln|x| + C$.
  6. $(1/3) \tan 3x + C$
  7. $-\cot x + C$
  8. $\sec x + C$
  9. $-\text{cosec} x + C$
  10. $4 e^x + C$
  11. $-3 \cos x - 2 \sin x + C$
  12. $(4/3) x^3 - 10 x^2 + 25 x + C$
  13. $x^6/6 + x^3/3 + C$
  14. $x^2/2 + 3x + 2 \ln|x| + C$
  15. $x^2/2 - 2 \ln|x| - 1/x + C$
  16. $-\cot x + \text{cosec} x + C$
  17. $x^3/3 + (3/2) x^2 + 2x + C$
  18. $x^2/2 + 2x + \ln|x| + C$
  19. $(2/7) x^{7/2} - (6/5) x^{5/2} + 10 x^{1/2} + C$
  20. $(2/3) x^3 + (1/2) x^2 - 3x + C$
  21. (B) $x^5/5 - 1/(3x^3) + C$.
  22. $y = x^4 + 1/x^3$
  23. $3 \sec^{-1} x + C$

↑ Back to top

Practice set: Exercise 7.2

Exercise 7.2 practises substitution. In each question, look for a function whose derivative also appears in the integrand.

Exercise 7.2: Substitution
  1. practice Find $\int 2x (x^2+5)^6 \mathrm{d} x$.
  2. practice Find $\int 4x^3 (x^4+1)^4 \mathrm{d} x$.
  3. practice Find $\int x/(x^2+9) \mathrm{d} x$.
  4. practice Find $\int x^2/(x^3+4) \mathrm{d} x$.
  5. practice Find $\int (\ln x)^3/x \mathrm{d} x$.
  6. practice Which of these equals $\int 1/(x \ln x) \mathrm{d} x$?
    1. $\ln|\ln x| + C$
    2. $1/(\ln x) + C$
    3. $(\ln x)^2/2 + C$
    4. $\ln x + C$
  7. practice Find $\int e^x/(1+e^x) \mathrm{d} x$.
  8. practice Find $\int e^x \sin(e^x) \mathrm{d} x$.
  9. practice Find $\int e^x \cos(e^x) \mathrm{d} x$.
  10. practice Find $\int \sin x \cos^5 x \mathrm{d} x$.
  11. practice Find $\int \cos x \sin^3 x \mathrm{d} x$.
  12. practice Find $\int \sec^2 x \tan^4 x \mathrm{d} x$.
  13. practice Find $\int \text{cosec}^2 x \cot^3 x \mathrm{d} x$.
  14. practice Find $\int \sec x \tan x \sqrt{\sec x} \mathrm{d} x$.
  15. practice Find $\int \text{cosec} x \cot x \sqrt{\text{cosec} x} \mathrm{d} x$.
  16. practice Find $\int x e^{x^2} \mathrm{d} x$.
  17. practice Find $\int x^3 e^{x^4} \mathrm{d} x$.
  18. practice Find $\int x \sin(x^2) \mathrm{d} x$.
  19. practice Find $\int x \cos(x^2) \mathrm{d} x$.
  20. practice Find $\int (2x+1)/(x^2+x+3) \mathrm{d} x$.
  21. practice Find $\int (3x^2+2x)/(x^3+x^2+1) \mathrm{d} x$.
  22. practice Find $\int x \sqrt{x^2+4} \mathrm{d} x$.
  23. practice Find $\int x \sqrt{1-x^2} \mathrm{d} x$.
  24. practice Find $\int x^3 \sqrt{x^4+1} \mathrm{d} x$.
  25. practice Find $\int (x+1)/\sqrt{x^2+2x+5} \mathrm{d} x$.
  26. practice Find $\int (2x-1)/\sqrt{x^2-x+1} \mathrm{d} x$.
  27. practice Find $\int x/(1+x^4) \mathrm{d} x$.
  28. practice Find $\int e^{2x}/(1+e^{4x}) \mathrm{d} x$.
  29. practice Find $\int 1/(x (1+(\ln x)^2)) \mathrm{d} x$.
  30. practice Find $\int \sin(\sqrt{x})/\sqrt{x} \mathrm{d} x$.
  31. practice Find $\int \cos(\sqrt{x})/\sqrt{x} \mathrm{d} x$.
  32. practice Find $\int x^2/(x^3+1)^2 \mathrm{d} x$.
  33. practice Find $\int (x-1)/(x^2-2x+5) \mathrm{d} x$.
  34. practice Find $\int x^3/(x^2+1) \mathrm{d} x$.
  35. practice Find $\int x^3 \sqrt{x^2+1} \mathrm{d} x$.
  36. practice Find $\int \sec^2 x/(\tan^2 x+4 \tan x+5) \mathrm{d} x$.
  37. practice Find $\int 1/(\sqrt{x} (1+x)) \mathrm{d} x$.
  38. practice Find $\int 1/(\sqrt{x} (1+\sqrt{x})^2) \mathrm{d} x$.
  39. practice Which of these equals $\int 2x/(1+x^2)^2 \mathrm{d} x$?
    1. $-1/(1+x^2) + C$
    2. $1/(1+x^2) + C$
    3. $\ln(1+x^2) + C$
    4. $-2/(1+x^2)^2 + C$
  40. practice Find $\int x^3/\sqrt{1-x^2} \mathrm{d} x$.
  41. practice Find $\int \tan x \mathrm{d} x$.
  42. practice Find $\int \cot x \mathrm{d} x$.
  43. practice Find $\int \sec x \mathrm{d} x$.
  44. practice Find $\int \text{cosec} x \mathrm{d} x$.
  45. practice Find $\int 2^x \mathrm{d} x$.
Answers
  1. $(x^2+5)^7/7 + C$
  2. $(x^4+1)^5/5 + C$
  3. $(1/2) \ln|x^2+9| + C$
  4. $(1/3) \ln|x^3+4| + C$
  5. $(\ln x)^4/4 + C$
  6. (A) $\ln|\ln x| + C$.
  7. $\ln(1+e^x) + C$
  8. $-\cos(e^x) + C$
  9. $\sin(e^x) + C$
  10. $-\cos^6 x/6 + C$
  11. $\sin^4 x/4 + C$
  12. $\tan^5 x/5 + C$
  13. $-\cot^4 x/4 + C$
  14. $(2/3) (\sec x)^{3/2} + C$
  15. $-(2/3) (\text{cosec} x)^{3/2} + C$
  16. $(1/2) e^{x^2} + C$
  17. $(1/4) e^{x^4} + C$
  18. $-(1/2) \cos(x^2) + C$
  19. $(1/2) \sin(x^2) + C$
  20. $\ln|x^2+x+3| + C$
  21. $\ln|x^3+x^2+1| + C$
  22. $(1/3) (x^2+4)^{3/2} + C$
  23. $-(1/3) (1-x^2)^{3/2} + C$
  24. $(1/6) (x^4+1)^{3/2} + C$
  25. $\sqrt{x^2+2x+5} + C$
  26. $2 \sqrt{x^2-x+1} + C$
  27. $(1/2) \tan^{-1}(x^2) + C$
  28. $(1/2) \tan^{-1}(e^{2x}) + C$
  29. $\tan^{-1}(\ln x) + C$
  30. $-2 \cos(\sqrt{x}) + C$
  31. $2 \sin(\sqrt{x}) + C$
  32. $-1/(3(x^3+1)) + C$
  33. $(1/2) \ln|x^2-2x+5| + C$
  34. $x^2/2 - (1/2) \ln|x^2+1| + C$
  35. $(1/5) (x^2+1)^{5/2} - (1/3) (x^2+1)^{3/2} + C$
  36. $\tan^{-1}(\tan x + 2) + C$
  37. $2 \tan^{-1}(\sqrt{x}) + C$
  38. $-2/(1+\sqrt{x}) + C$
  39. (A) $-1/(1+x^2) + C$.
  40. $-\sqrt{1-x^2} + (1/3) (1-x^2)^{3/2} + C$
  41. $\ln|\sec x| + C$
  42. $\ln|\sin x| + C$
  43. $\ln|\sec x + \tan x| + C$
  44. $\ln|\text{cosec} x - \cot x| + C$
  45. $2^x/(\ln 2) + C$

↑ Back to top

Practice set: Exercise 7.3

Exercise 7.3 practises integrals of trigonometric functions. First rewrite the integrand with an identity, such as $\sin^2 x = (1 - \cos 2 x)/2$. Then a formula from the table applies.

Exercise 7.3: Trig identities before integrating
  1. practice Find $\int \sin^2 x \mathrm{d} x$.
  2. practice Find $\int \cos^2 x \mathrm{d} x$.
  3. practice Find $\int \sin^2 (3x) \mathrm{d} x$.
  4. practice Find $\int \tan^2 x \mathrm{d} x$.
  5. practice Find $\int \cot^2 x \mathrm{d} x$.
  6. practice Find $\int \sec^2 x \tan x \mathrm{d} x$.
  7. practice Find $\int \text{cosec}^2 x \cot x \mathrm{d} x$.
  8. practice Find $\int \sin 4x \cos 2x \mathrm{d} x$.
  9. practice Find $\int \cos 3x \cos x \mathrm{d} x$.
  10. practice Find $\int \sin 5x \sin x \mathrm{d} x$.
  11. practice Find $\int \sin^3 x \mathrm{d} x$.
  12. practice Find $\int \cos^3 x \mathrm{d} x$.
  13. practice Find $\int \sin^4 x \mathrm{d} x$.
  14. practice Find $\int \cos^4 x \mathrm{d} x$.
  15. practice Find $\int \sin^2 x \cos^2 x \mathrm{d} x$.
  16. practice Find $\int (\sin x + \cos x)^2 \mathrm{d} x$.
  17. practice Find $\int \sec^4 x \mathrm{d} x$.
  18. practice Find $\int \text{cosec}^4 x \mathrm{d} x$.
  19. practice Find $\int (1-\cos x)/(1+\cos x) \mathrm{d} x$.
  20. practice Find $\int \sin^3 x \cos^2 x \mathrm{d} x$.
  21. practice Find $\int \cos^3 x \sin^2 x \mathrm{d} x$.
  22. practice Find $\int \sin x \sin 2x \sin 3x \mathrm{d} x$.
  23. practice To evaluate $\int \sin x \cos 5x \mathrm{d} x$, which identity is applied first?
    1. $\sin A \cos B = (1/2) [\sin(A+B)+\sin(A-B)]$
    2. $\cos A \cos B = (1/2) [\cos(A-B)+\cos(A+B)]$
    3. $\sin A \sin B = (1/2) [\cos(A-B)-\cos(A+B)]$
    4. $\sin^2 x = (1-\cos 2x)/2$
  24. practice $\int 2 \sin^2 x \mathrm{d} x$ equals
    1. $x - \sin(2x)/2 + C$
    2. $2x - \sin(2x) + C$
    3. $x - \sin(2x) + C$
    4. $2x - 2 \sin(2x) + C$
Answers
  1. $x/2 - \sin(2x)/4 + C$
  2. $x/2 + \sin(2x)/4 + C$
  3. $x/2 - \sin(6x)/12 + C$
  4. $\tan x - x + C$
  5. $-\cot x - x + C$
  6. $\tan^2 x/2 + C$
  7. $-\cot^2 x/2 + C$
  8. $-\cos(6x)/12 - \cos(2x)/4 + C$
  9. $\sin(2x)/4 + \sin(4x)/8 + C$
  10. $\sin(4x)/8 - \sin(6x)/12 + C$
  11. $-\cos x + \cos^3 x/3 + C$
  12. $\sin x - \sin^3 x/3 + C$
  13. $3x/8 - \sin(2x)/4 + \sin(4x)/32 + C$
  14. $3x/8 + \sin(2x)/4 + \sin(4x)/32 + C$
  15. $x/8 - \sin(4x)/32 + C$
  16. $x - \cos(2x)/2 + C$
  17. $\tan x + \tan^3 x/3 + C$
  18. $-\cot x - \cot^3 x/3 + C$
  19. $2 \tan(x/2) - x + C$
  20. $-\cos^3 x/3 + \cos^5 x/5 + C$
  21. $\sin^3 x/3 - \sin^5 x/5 + C$
  22. $-\cos(4x)/16 + \cos(6x)/24 - \cos(2x)/8 + C$
  23. (A) $\sin A \cos B = (1/2) [\sin(A+B)+\sin(A-B)]$.
  24. (A) $x - \sin(2x)/2 + C$.

↑ Back to top

Practice set: Exercise 7.4, part 1

These questions are the first part of Exercise 7.4. They use the formulas for $1/(x^2 + a^2)$ and $1/(x^2 - a^2)$. Some need you to complete the square first.

Exercise 7.4, part 1: The x² + a² and x² − a² special forms
  1. practice Find $\int 1/(x^2+9) \mathrm{d} x$.
  2. practice Find $\int 1/(x^2+16) \mathrm{d} x$.
  3. practice Find $\int 1/(x^2+1) \mathrm{d} x$.
  4. practice Find $\int 1/(x^2-9) \mathrm{d} x$.
  5. practice Find $\int 1/(x^2-25) \mathrm{d} x$.
  6. practice Find $\int 1/(x^2-1) \mathrm{d} x$.
  7. practice Find $\int 1/(x^2+36) \mathrm{d} x$.
  8. practice Find $\int 1/(x^2-16) \mathrm{d} x$.
  9. practice Find $\int 1/(x^2+49) \mathrm{d} x$.
  10. practice Find $\int 1/(4x^2+9) \mathrm{d} x$.
  11. practice Find $\int 1/(9x^2-4) \mathrm{d} x$.
  12. practice Find $\int 1/(25x^2+4) \mathrm{d} x$.
  13. practice Find $\int 1/(16x^2-9) \mathrm{d} x$.
  14. practice Find $\int 1/(x^2+2x+5) \mathrm{d} x$.
  15. practice Find $\int 1/(x^2-4x-5) \mathrm{d} x$.
  16. practice Find $\int 1/(x^2+6x+13) \mathrm{d} x$.
  17. practice Find $\int 1/(x^2-8x+7) \mathrm{d} x$.
  18. practice Find $\int 1/(x^2+4x+13) \mathrm{d} x$.
  19. practice Find $\int 1/(x^2-6x+5) \mathrm{d} x$.
  20. practice Find $\int 1/(x^2+10x+34) \mathrm{d} x$.
  21. practice Find $\int 1/(x^2-2x-8) \mathrm{d} x$.
  22. practice Find $\int 1/(x^2+8x+20) \mathrm{d} x$.
  23. practice Find $\int 1/(4x^2+4x+5) \mathrm{d} x$.
  24. practice Find $\int 1/(9x^2-12x-5) \mathrm{d} x$.
  25. practice Find $\int 1/(2x^2+4x+10) \mathrm{d} x$.
  26. practice Find $\int 1/(16x^2-8x-3) \mathrm{d} x$.
  27. practice $\int 1/(x^2+64) \mathrm{d} x$ equals
    1. $1/8 \cdot \tan^{-1}(x/8) + C$
    2. $\tan^{-1}(x/8) + C$
    3. $1/64 \cdot \tan^{-1}(x/8) + C$
    4. $8 \cdot \tan^{-1}(x/8) + C$
  28. practice $\int 1/(x^2-64) \mathrm{d} x$ equals
    1. $1/16 \cdot \ln|(x-8)/(x+8)| + C$
    2. $1/8 \cdot \ln|(x-8)/(x+8)| + C$
    3. $1/16 \cdot \ln|(x+8)/(x-8)| + C$
    4. $\ln|(x-8)/(x+8)| + C$
  29. practice Find $\int (x+2)/(2 x^2+6 x+5) \mathrm{d} x$.
  30. practice Find $\int (x+3)/\sqrt{x^2+4 x+5} \mathrm{d} x$.
  31. practice Find $\int 1/(8 - 2 x - x^2) \mathrm{d} x$.
  32. practice Find $\int (x + 1)/(x^2 + 4 x + 5) \mathrm{d} x$.
  33. practice Find $\int (x + 2)/\sqrt{x^2 + 2 x + 3} \mathrm{d} x$.
Answers
  1. $1/3 \cdot \tan^{-1}(x/3) + C$
  2. $1/4 \cdot \tan^{-1}(x/4) + C$
  3. $\tan^{-1}(x) + C$
  4. $1/6 \cdot \ln|(x-3)/(x+3)| + C$
  5. $1/10 \cdot \ln|(x-5)/(x+5)| + C$
  6. $1/2 \cdot \ln|(x-1)/(x+1)| + C$
  7. $1/6 \cdot \tan^{-1}(x/6) + C$
  8. $1/8 \cdot \ln|(x-4)/(x+4)| + C$
  9. $1/7 \cdot \tan^{-1}(x/7) + C$
  10. $1/6 \cdot \tan^{-1}(2x/3) + C$
  11. $1/12 \cdot \ln|(3x-2)/(3x+2)| + C$
  12. $1/10 \cdot \tan^{-1}(5x/2) + C$
  13. $1/24 \cdot \ln|(4x-3)/(4x+3)| + C$
  14. $1/2 \cdot \tan^{-1}((x+1)/2) + C$
  15. $1/6 \cdot \ln|(x-5)/(x+1)| + C$
  16. $1/2 \cdot \tan^{-1}((x+3)/2) + C$
  17. $1/6 \cdot \ln|(x-7)/(x-1)| + C$
  18. $1/3 \cdot \tan^{-1}((x+2)/3) + C$
  19. $1/4 \cdot \ln|(x-5)/(x-1)| + C$
  20. $1/3 \cdot \tan^{-1}((x+5)/3) + C$
  21. $1/6 \cdot \ln|(x-4)/(x+2)| + C$
  22. $1/2 \cdot \tan^{-1}((x+4)/2) + C$
  23. $1/4 \cdot \tan^{-1}(x+1/2) + C$
  24. $1/18 \cdot \ln|(3x-5)/(3x+1)| + C$
  25. $1/4 \cdot \tan^{-1}((x+1)/2) + C$
  26. $1/16 \cdot \ln|(4x-3)/(4x+1)| + C$
  27. (A) $1/8 \cdot \tan^{-1}(x/8) + C$.
  28. (A) $1/16 \cdot \ln|(x-8)/(x+8)| + C$.
  29. $(1/4) \ln|2 x^2+6 x+5| + (1/2) \tan^{-1}(2 x+3) + C$
  30. $\sqrt{x^2+4 x+5} + \ln|x+2+\sqrt{x^2+4 x+5}| + C$
  31. $(1/6) \ln|(x + 4)/(2 - x)| + C$
  32. $(1/2) \ln(x^2 + 4 x + 5) - \tan^{-1}(x + 2) + C$
  33. $\sqrt{x^2 + 2 x + 3} + \ln|x + 1 + \sqrt{x^2 + 2 x + 3}| + C$

↑ Back to top

Practice set: Exercise 7.4, part 2, and Exercise 7.7

These questions finish Exercise 7.4 and cover Exercise 7.7. They use the formulas with a square root, such as $1/\sqrt{a^2 - x^2}$ and $\sqrt{x^2 + a^2}$. Some need a substitution or a completed square first.

Exercise 7.4, part 2, and Exercise 7.7: Special forms, part 2
  1. practice Find $\int 1/\sqrt{9-x^2} \mathrm{d} x$.
  2. practice Find $\int 1/\sqrt{16-x^2} \mathrm{d} x$.
  3. practice Find $\int 1/\sqrt{25-x^2} \mathrm{d} x$.
  4. practice Find $\int 5/\sqrt{4-x^2} \mathrm{d} x$.
  5. practice Find $\int \sqrt{4-x^2} \mathrm{d} x$.
  6. practice Find $\int \sqrt{9-x^2} \mathrm{d} x$.
  7. practice Find $\int \sqrt{1-x^2} \mathrm{d} x$.
  8. practice Find $\int 1/\sqrt{x^2+9} \mathrm{d} x$.
  9. practice Find $\int 1/\sqrt{x^2+16} \mathrm{d} x$.
  10. practice Find $\int 3/\sqrt{x^2+4} \mathrm{d} x$.
  11. practice Find $\int 1/\sqrt{x^2+1} \mathrm{d} x$.
  12. practice Find $\int 4/\sqrt{1-x^2} \mathrm{d} x$.
  13. practice Find $\int 1/\sqrt{4-9x^2} \mathrm{d} x$.
  14. practice Find $\int 1/\sqrt{9-4x^2} \mathrm{d} x$.
  15. practice Find $\int \sqrt{4-9x^2} \mathrm{d} x$.
  16. practice Find $\int 1/\sqrt{x^2+2x+10} \mathrm{d} x$.
  17. practice Find $\int 1/\sqrt{4x-x^2} \mathrm{d} x$.
  18. practice Find $\int 1/\sqrt{2x-x^2} \mathrm{d} x$.
  19. practice Find $\int 1/\sqrt{x^2-6x+18} \mathrm{d} x$.
  20. practice Find $\int 1/\sqrt{5-4x-x^2} \mathrm{d} x$.
  21. practice Find $\int \sqrt{8-2x-x^2} \mathrm{d} x$.
  22. practice Find $\int 1/\sqrt{9-8x-x^2} \mathrm{d} x$.
  23. practice Which of these is $\int 1/\sqrt{4-x^2} \mathrm{d} x$?
    1. $\sin^{-1}(x/2) + C$
    2. $\sin^{-1}(2x) + C$
    3. $1/2 \cdot \sin^{-1}(x/2) + C$
    4. $2 \cdot \sin^{-1}(x/2) + C$
  24. practice Which of these is $\int 1/\sqrt{x^2+4} \mathrm{d} x$?
    1. $\ln|x+\sqrt{x^2+4}| + C$
    2. $1/2 \cdot \ln|x+\sqrt{x^2+4}| + C$
    3. $\sin^{-1}(x/2) + C$
    4. $2 \cdot \ln|x+\sqrt{x^2+4}| + C$
  25. practice Find $\int \sqrt{x^2 + 4 x + 6} \mathrm{d} x$.
  26. practice Find $\int 1/\sqrt{x^2 - 9} \mathrm{d} x$.
  27. practice Find $\int \sqrt{x^2 - 4} \mathrm{d} x$.
Answers
  1. $\sin^{-1}(x/3) + C$
  2. $\sin^{-1}(x/4) + C$
  3. $\sin^{-1}(x/5) + C$
  4. $5 \cdot \sin^{-1}(x/2) + C$
  5. $x/2 \cdot \sqrt{4-x^2} + 2 \cdot \sin^{-1}(x/2) + C$
  6. $x/2 \cdot \sqrt{9-x^2} + 9/2 \cdot \sin^{-1}(x/3) + C$
  7. $x/2 \cdot \sqrt{1-x^2} + 1/2 \cdot \sin^{-1}(x) + C$
  8. $\ln|x+\sqrt{x^2+9}| + C$
  9. $\ln|x+\sqrt{x^2+16}| + C$
  10. $3 \cdot \ln|x+\sqrt{x^2+4}| + C$
  11. $\ln|x+\sqrt{x^2+1}| + C$
  12. $4 \cdot \sin^{-1}(x) + C$
  13. $1/3 \cdot \sin^{-1}(3x/2) + C$
  14. $1/2 \cdot \sin^{-1}(2x/3) + C$
  15. $x/2 \cdot \sqrt{4-9x^2} + 2/3 \cdot \sin^{-1}(3x/2) + C$
  16. $\ln|x+1+\sqrt{x^2+2x+10}| + C$
  17. $\sin^{-1}((x-2)/2) + C$
  18. $\sin^{-1}(x-1) + C$
  19. $\ln|x-3+\sqrt{x^2-6x+18}| + C$
  20. $\sin^{-1}((x+2)/3) + C$
  21. $(x+1)/2 \cdot \sqrt{8-2x-x^2} + 9/2 \cdot \sin^{-1}((x+1)/3) + C$
  22. $\sin^{-1}((x+4)/5) + C$
  23. (A) $\sin^{-1}(x/2) + C$.
  24. (A) $\ln|x+\sqrt{x^2+4}| + C$.
  25. $((x + 2)/2) \sqrt{x^2 + 4 x + 6} + \ln|x + 2 + \sqrt{x^2 + 4 x + 6}| + C$
  26. $\ln|x + \sqrt{x^2 - 9}| + C$
  27. $(x/2) \sqrt{x^2 - 4} - 2 \ln|x + \sqrt{x^2 - 4}| + C$

↑ Back to top

Practice set: Exercise 7.5

Exercise 7.5 practises partial fractions. Split the integrand, find the constants, then integrate each piece. A repeated factor or a quadratic factor needs a different split.

Exercise 7.5: Partial fractions
  1. practice Find $\int (5x+4)/((x-1)(x+2)) \mathrm{d} x$.
  2. practice Find $\int (5x-13)/((x-2)(x-5)) \mathrm{d} x$.
  3. practice Find $\int (x-11)/((x+3)(x-4)) \mathrm{d} x$.
  4. practice Find $\int 1/(x^2-9) \mathrm{d} x$.
  5. practice Find $\int (x+7)/((x+1)(x+4)) \mathrm{d} x$.
  6. practice Find $\int (4x-8)/((x-3)(x+1)) \mathrm{d} x$.
  7. practice Find $\int (x-1)/((x+1)(x+4)) \mathrm{d} x$.
  8. practice Find $\int (2x+3)/(x-1)^2 \mathrm{d} x$.
  9. practice Find $\int (5x-4)/(x+2)^2 \mathrm{d} x$.
  10. practice Find $\int (x-5)/(x+3)^2 \mathrm{d} x$.
  11. practice Find $\int 1/(x(x^2+9)) \mathrm{d} x$.
  12. practice Find $\int (x+8)/(x^2+x-6) \mathrm{d} x$.
  13. practice Find $\int 2/((x-1)(x-2)(x-3)) \mathrm{d} x$.
  14. practice Find $\int (x+1)/((x-1)(x-2)(x-3)) \mathrm{d} x$.
  15. practice Find $\int (5x+1)/((x-1)^2(x+2)) \mathrm{d} x$.
  16. practice Find $\int (10x-8)/((x-2)^2(x+1)) \mathrm{d} x$.
  17. practice Find $\int x/((x+3)(x^2+9)) \mathrm{d} x$.
  18. practice Find $\int 8/((x-2)(x^2+4)) \mathrm{d} x$.
  19. practice Which of these is $\int x/((x-1)(x-2)) \mathrm{d} x$?
    1. $\ln|x-1| - 2 \cdot \ln|x-2| + C$
    2. $-\ln|x-1| + 2 \cdot \ln|x-2| + C$
    3. $2 \cdot \ln|x-1| - \ln|x-2| + C$
    4. $-2 \cdot \ln|x-1| + \ln|x-2| + C$
  20. practice Which of these is $\int (3x-1)/(x+1)^2 \mathrm{d} x$?
    1. $3 \cdot \ln|x+1| + 4/(x+1) + C$
    2. $3 \cdot \ln|x+1| - 4/(x+1) + C$
    3. $-3 \cdot \ln|x+1| + 4/(x+1) + C$
    4. $3 \cdot \ln|x+1| + 4(x+1) + C$
  21. practice Find $\int (x^2+7x+4)/((x-1)(x+2)) \mathrm{d} x$.
  22. practice Find $\int (10-3x)/((x+2)(x^2+4)) \mathrm{d} x$.
  23. practice Find $\int (x^2+x+1)/((x-1)(x^2+1)) \mathrm{d} x$.
  24. practice Which of these is $\int 1/(x(x^2+4)) \mathrm{d} x$?
    1. $1/4 \cdot \ln|x| - 1/8 \cdot \ln(x^2+4) + C$
    2. $1/4 \cdot \ln|x| + 1/8 \cdot \ln(x^2+4) + C$
    3. $\ln|x| - 1/2 \cdot \ln(x^2+4) + C$
    4. $1/8 \cdot \ln|x| - 1/4 \cdot \ln(x^2+4) + C$
Answers
  1. $3 \cdot \ln|x-1| + 2 \cdot \ln|x+2| + C$
  2. $\ln|x-2| + 4 \cdot \ln|x-5| + C$
  3. $2 \cdot \ln|x+3| - \ln|x-4| + C$
  4. $1/6 \cdot \ln|x-3| - 1/6 \cdot \ln|x+3| + C$
  5. $2 \cdot \ln|x+1| - \ln|x+4| + C$
  6. $\ln|x-3| + 3 \cdot \ln|x+1| + C$
  7. $-2/3 \cdot \ln|x+1| + 5/3 \cdot \ln|x+4| + C$
  8. $2 \cdot \ln|x-1| - 5/(x-1) + C$
  9. $5 \cdot \ln|x+2| + 14/(x+2) + C$
  10. $\ln|x+3| + 8/(x+3) + C$
  11. $1/9 \cdot \ln|x| - 1/18 \cdot \ln(x^2+9) + C$
  12. $-\ln|x+3| + 2 \cdot \ln|x-2| + C$
  13. $\ln|x-1| - 2 \cdot \ln|x-2| + \ln|x-3| + C$
  14. $\ln|x-1| - 3 \cdot \ln|x-2| + 2 \cdot \ln|x-3| + C$
  15. $\ln|x-1| - 2/(x-1) - \ln|x+2| + C$
  16. $2 \cdot \ln|x-2| - 4/(x-2) - 2 \cdot \ln|x+1| + C$
  17. $-1/6 \cdot \ln|x+3| + 1/12 \cdot \ln(x^2+9) + 1/6 \cdot \tan^{-1}(x/3) + C$
  18. $\ln|x-2| - 1/2 \cdot \ln(x^2+4) - \tan^{-1}(x/2) + C$
  19. (B) $-\ln|x-1| + 2 \cdot \ln|x-2| + C$.
  20. (A) $3 \cdot \ln|x+1| + 4/(x+1) + C$.
  21. $x + 4 \cdot \ln|x-1| + 2 \cdot \ln|x+2| + C$
  22. $2 \cdot \ln|x+2| - \ln(x^2+4) + 1/2 \cdot \tan^{-1}(x/2) + C$
  23. $3/2 \cdot \ln|x-1| - 1/4 \cdot \ln(x^2+1) + 1/2 \cdot \tan^{-1}(x) + C$
  24. (A) $1/4 \cdot \ln|x| - 1/8 \cdot \ln(x^2+4) + C$.

↑ Back to top

Practice set: Exercise 7.6

Exercise 7.6 practises integration by parts. Some questions need it twice. Others fit the formula $\int e^x (f(x) + f'(x)) \mathrm{d} x = e^x f(x) + C$, so look for that first.

Exercise 7.6: Integration by parts
  1. practice Find $\int x \cdot \sin 2 x \mathrm{d} x$.
  2. practice Find $\int x \cdot e^{3 x} \mathrm{d} x$.
  3. practice Find $\int x \cdot \cos x \mathrm{d} x$.
  4. practice Find $\int \ln x \mathrm{d} x$.
  5. practice Find $\int (x+1) \cdot e^x \mathrm{d} x$.
  6. practice Find $\int \tan^{-1}(2 x) \mathrm{d} x$.
  7. practice Find $\int x \cdot \ln(3 x) \mathrm{d} x$.
  8. practice Find $\int (2 x+1) \cdot \ln x \mathrm{d} x$.
  9. practice Find $\int x \cdot e^{-2 x} \mathrm{d} x$.
  10. practice Which of these equals $\int x \cdot e^{-x} \mathrm{d} x$?
    1. $-x e^{-x} - e^{-x} + C$
    2. $-x e^{-x} + e^{-x} + C$
    3. $x e^{-x} - e^{-x} + C$
    4. $x e^{-x} + e^{-x} + C$
  11. practice Find $\int x^2 \cdot e^{2 x} \mathrm{d} x$.
  12. practice Find $\int x^2 \cdot \sin x \mathrm{d} x$.
  13. practice Find $\int x^3 \cdot \ln x \mathrm{d} x$.
  14. practice Find $\int (\ln x)^2 \mathrm{d} x$.
  15. practice Find $\int x \cdot \cos^{-1}(x) \mathrm{d} x$.
  16. practice Find $\int e^x (\sqrt{x} + 1/(2 \sqrt{x})) \mathrm{d} x$.
  17. practice Find $\int e^x (\cos x - \sin x) \mathrm{d} x$.
  18. practice Find $\int e^x (x^3+3 x^2) \mathrm{d} x$.
  19. practice Find $\int e^{3 x} \cdot \cos x \mathrm{d} x$.
  20. practice Which of these equals $\int e^x (\ln x + 1/x) \mathrm{d} x$?
    1. $e^x \ln x + C$
    2. $e^x/x + C$
    3. $(e^x \ln x)/x + C$
    4. $e^x + \ln x + C$
  21. practice Find $\int x^2 e^{-x} \mathrm{d} x$.
  22. practice Find $\int x \sec^2 x \mathrm{d} x$.
  23. practice Find $\int x \cos(3 x) \mathrm{d} x$.
  24. practice Find $\int \sin^{-1}(x) \mathrm{d} x$.
Answers
  1. $-(x \cos 2 x)/2 + (\sin 2 x)/4 + C$
  2. $(x e^{3 x})/3 - e^{3 x}/9 + C$
  3. $x \sin x + \cos x + C$
  4. $x \ln x - x + C$
  5. $x e^x + C$
  6. $x \tan^{-1}(2 x) - (1/4) \ln(1+4 x^2) + C$
  7. $(x^2/2) \ln(3 x) - x^2/4 + C$
  8. $(x^2+x) \ln x - x^2/2 - x + C$
  9. $-(x e^{-2 x})/2 - e^{-2 x}/4 + C$
  10. (A) $-x e^{-x} - e^{-x} + C$.
  11. $(x^2 e^{2 x})/2 - (x e^{2 x})/2 + e^{2 x}/4 + C$
  12. $-x^2 \cos x + 2 x \sin x + 2 \cos x + C$
  13. $(x^4/4) \ln x - x^4/16 + C$
  14. $x (\ln x)^2 - 2 x \ln x + 2 x + C$
  15. $(x^2 \cos^{-1}(x))/2 + (\sin^{-1}(x))/4 - (x \sqrt{1-x^2})/4 + C$
  16. $e^x \sqrt{x} + C$
  17. $e^x \cos x + C$
  18. $e^x x^3 + C$
  19. $(e^{3 x}(\sin x+3 \cos x))/10 + C$
  20. (A) $e^x \ln x + C$.
  21. $-e^{-x} (x^2+2 x+2) + C$
  22. $x \tan x + \ln|\cos x| + C$
  23. $(x \sin(3 x))/3 + (\cos(3 x))/9 + C$
  24. $x \sin^{-1}(x) + \sqrt{1-x^2} + C$

↑ Back to top

Practice set: the definite integral from strips (beyond the board syllabus)

These questions find a definite integral straight from the strips, as a limit of totals. The board exam does not set these questions, and NCERT no longer has this exercise. Try a few to see the work the fundamental theorem saves.

Beyond the board syllabus: the definite integral from strips, as a limit of totals
  1. practice Evaluate $\int_1^4 5 \mathrm{d} x$ directly from the definition, as a limit of totals.
  2. practice Evaluate $\int_0^2 x \mathrm{d} x$ directly from the definition, as a limit of totals.
  3. practice Evaluate $\int_1^3 x \mathrm{d} x$ directly from the definition, as a limit of totals.
  4. practice Evaluate $\int_0^1 2 x \mathrm{d} x$ directly from the definition, as a limit of totals.
  5. practice Evaluate $\int_0^2 (3 x+1) \mathrm{d} x$ directly from the definition, as a limit of totals.
  6. practice Evaluate $\int_1^3 (x+2) \mathrm{d} x$ directly from the definition, as a limit of totals.
  7. practice Evaluate $\int_{-1}^1 x \mathrm{d} x$ directly from the definition, as a limit of totals.
  8. practice Evaluate $\int_0^4 (5-x) \mathrm{d} x$ directly from the definition, as a limit of totals.
  9. practice Evaluate $\int_0^1 x^2 \mathrm{d} x$ directly from the definition, as a limit of totals.
  10. practice Evaluate $\int_0^3 x^2 \mathrm{d} x$ directly from the definition, as a limit of totals.
  11. practice Evaluate $\int_1^2 x^2 \mathrm{d} x$ directly from the definition, as a limit of totals.
  12. practice Which value does $\int_{-1}^2 x^2 \mathrm{d} x$ tend to, once its limit is taken?
    1. $3$
    2. $9$
    3. $7/3$
    4. $-3$
  13. practice Evaluate $\int_0^2 (x^2+1) \mathrm{d} x$ directly from the definition, as a limit of totals.
  14. practice Evaluate $\int_0^2 (2 x^2-x) \mathrm{d} x$ directly from the definition, as a limit of totals.
  15. practice Evaluate $\int_0^3 (x^2+2 x) \mathrm{d} x$ directly from the definition, as a limit of totals.
  16. practice Evaluate $\int_0^1 x^3 \mathrm{d} x$ directly from the definition, as a limit of totals.
Answers
  1. $15$
  2. $2$
  3. $4$
  4. $1$
  5. $8$
  6. $8$
  7. $0$
  8. $12$
  9. $1/3$
  10. $9$
  11. $7/3$
  12. (A) $3$.
  13. $14/3$
  14. $10/3$
  15. $18$
  16. $1/4$

↑ Back to top

Practice set: Exercise 7.8

Exercise 7.8 practises the second fundamental theorem. Find one integral, then subtract its value at the lower limit from its value at the upper limit. Some need a method from earlier, such as substitution or by parts.

Exercise 7.8: Definite integrals via the Fundamental Theorem
  1. practice Evaluate $\int_0^2 (x+3) \mathrm{d} x$.
  2. practice Evaluate $\int_1^3 1/x \mathrm{d} x$.
  3. practice Evaluate $\int_1^2 (4 x^3-3 x^2+2 x+1) \mathrm{d} x$.
  4. practice Evaluate $\int_0^{\pi/2} \sin x \mathrm{d} x$.
  5. practice Evaluate $\int_0^{\pi/4} \cos 2 x \mathrm{d} x$.
  6. practice Evaluate $\int_2^3 e^x \mathrm{d} x$.
  7. practice Evaluate $\int_0^1 x/(1+x^2) \mathrm{d} x$.
  8. practice Evaluate $\int_3^4 1/((x-1)(x-2)) \mathrm{d} x$.
  9. practice Evaluate $\int_0^1 x \cdot e^x \mathrm{d} x$.
  10. practice Which value does $\int_0^1 1/(1+x^2) \mathrm{d} x$ equal?
    1. $\pi/4$
    2. $\pi/2$
    3. $\pi/3$
    4. $\pi/6$
  11. practice Evaluate $\int_0^{1/2} 1/\sqrt{1-x^2} \mathrm{d} x$.
  12. practice Evaluate $\int_0^{\pi/4} (\sec^2 x+x) \mathrm{d} x$.
  13. practice Evaluate $\int_0^1 (3 x^2+2 x) \mathrm{d} x$.
  14. practice Evaluate $\int_2^4 1/x^2 \mathrm{d} x$.
  15. practice Evaluate $\int_0^{\pi/3} \sec^2 x \mathrm{d} x$.
  16. practice Evaluate $\int_0^1 x^2 e^x \mathrm{d} x$.
  17. practice Evaluate $\int_0^2 1/(x^2+4) \mathrm{d} x$.
  18. practice Evaluate $\int_0^1 1/(x^2-4) \mathrm{d} x$.
  19. practice Evaluate $\int_0^{\sqrt{3}} 1/\sqrt{4-x^2} \mathrm{d} x$.
  20. practice Evaluate $\int_0^{\pi/2} x \cos x \mathrm{d} x$.
  21. practice Evaluate $\int_1^9 1/\sqrt{x} \mathrm{d} x$.
  22. practice Evaluate $\int_2^3 1/(x^2-1) \mathrm{d} x$.
Answers
  1. $8$
  2. $\ln 3$
  3. $12$
  4. $1$
  5. $1/2$
  6. $e^3-e^2$
  7. $(\ln 2)/2$
  8. $\ln(4/3)$
  9. $1$
  10. (A) $\pi/4$.
  11. $\pi/6$
  12. $1+\pi^2/32$
  13. $2$
  14. $1/4$
  15. $\sqrt{3}$
  16. $e-2$
  17. $\pi/8$
  18. $-(1/4) \ln 3$
  19. $\pi/3$
  20. $\pi/2-1$
  21. $4$
  22. $(1/2) \ln(3/2)$

↑ Back to top

Practice set: Exercise 7.9

Exercise 7.9 practises substitution in definite integrals. Change the limits in the same step as the variable. Then stay in the new variable to the end.

Exercise 7.9: Definite integrals by substitution
  1. practice Evaluate $\int_0^1 (2 x)/(x^2+1) \mathrm{d} x$.
  2. practice Evaluate $\int_0^2 (3 x^2)/(x^3+1) \mathrm{d} x$.
  3. practice Evaluate $\int_0^{\pi/2} \sin x \cos^2 x \mathrm{d} x$.
  4. practice Evaluate $\int_0^{\pi/4} \sec^2 x \tan x \mathrm{d} x$.
  5. practice Evaluate $\int_0^1 x e^{x^2} \mathrm{d} x$.
  6. practice Evaluate $\int_1^2 (2 x+1)/(x^2+x+1) \mathrm{d} x$.
  7. practice Evaluate $\int_0^1 x/(x^2+1)^2 \mathrm{d} x$.
  8. practice Evaluate $\int_1^e (\ln x)/x \mathrm{d} x$.
  9. practice Evaluate $\int_0^1 x^2/(x^3+8) \mathrm{d} x$.
  10. practice Evaluate $\int_0^{\pi/2} (\cos x)/(1+\sin^2 x) \mathrm{d} x$.
  11. practice Evaluate $\int_0^1 x \sqrt{1-x^2} \mathrm{d} x$.
  12. practice Evaluate $\int_0^1 x^3/(x^4+1) \mathrm{d} x$.
  13. practice Evaluate $\int_0^1 x(1-x)^4 \mathrm{d} x$.
  14. practice Evaluate $\int_0^{\pi/4} \tan^3 x \sec^2 x \mathrm{d} x$.
  15. practice Evaluate $\int_0^1 (x+1)/\sqrt{x^2+2 x+3} \mathrm{d} x$.
  16. practice Evaluate $\int_e^{e^2} 1/(x \ln x) \mathrm{d} x$.
  17. practice The value of $\int_0^{\pi/2} (\sin x)/(1+\cos^2 x) \mathrm{d} x$ is
    1. $\pi/4$
    2. $\pi/2$
    3. $0$
    4. $1$
  18. practice The value of $\int_0^1 (2 x+3)/(x^2+3 x+1) \mathrm{d} x$ is
    1. $\ln 5$
    2. $\ln 4$
    3. $5$
    4. $0$
  19. practice Evaluate $\int_0^1 (3 x^2-2 x)/(x^3-x^2+1) \mathrm{d} x$.
  20. practice Evaluate $\int_0^1 x^3/\sqrt{1+x^4} \mathrm{d} x$.
Answers
  1. $\ln 2$
  2. $\ln 9$
  3. $1/3$
  4. $1/2$
  5. $(1/2) (e-1)$
  6. $\ln(7/3)$
  7. $1/4$
  8. $1/2$
  9. $(1/3) \ln(9/8)$
  10. $\pi/4$
  11. $1/3$
  12. $(1/4) \ln 2$
  13. $1/30$
  14. $1/4$
  15. $\sqrt{6} - \sqrt{3}$
  16. $\ln 2$
  17. (A) $\pi/4$.
  18. (A) $\ln 5$.
  19. $0$
  20. $(1/2) (\sqrt{2}-1)$

↑ Back to top

Practice set: Exercise 7.10 and Miscellaneous

Exercise 7.10 practises the properties of definite integrals. The Miscellaneous Exercise then mixes every method you have met. For an area, split the interval where the graph crosses the axis.

Exercise 7.10 and Miscellaneous: Properties of definite integrals, and a chapter-wide review
  1. practice If $\int_2^5 f(x) \mathrm{d} x = 7$, find $\int_5^2 f(x) \mathrm{d} x$.
  2. practice If $\int_1^4 g(x) \mathrm{d} x = 10$, evaluate $\int_4^1 3 g(x) \mathrm{d} x$.
  3. practice Evaluate $\int_{-\pi/2}^{\pi/2} \sin^3 x \mathrm{d} x$.
  4. practice Evaluate $\int_{-1}^1 x^3 \cos x \mathrm{d} x$.
  5. practice Evaluate $\int_{-3}^3 (x^5+2 x^3-x) \mathrm{d} x$.
  6. practice Evaluate $\int_{-2}^2 x^4 \mathrm{d} x$.
  7. practice Given that $\sin^2 x$ has period $\pi$, evaluate $\int_0^{6 \pi} \sin^2 x \mathrm{d} x$.
  8. practice Evaluate $\int_0^{\pi/2} \sin^2 x \mathrm{d} x$ using the reflection $x \to \pi/2 - x$.
  9. practice Evaluate $\int_0^1 \ln((1-x)/x) \mathrm{d} x$.
  10. practice Evaluate $\int_0^{\pi/2} \sqrt{\sin x}/(\sqrt{\sin x}+\sqrt{\cos x}) \mathrm{d} x$.
  11. practice Evaluate $\int_0^\pi x (\sin x)/(1+\cos^2 x) \mathrm{d} x$.
  12. practice Evaluate $\int_0^\pi x \sin^3 x \mathrm{d} x$.
  13. practice Show that $\int_0^1 x(1-x)^n \mathrm{d} x = \int_0^1 x^n \cdot (1-x) \mathrm{d} x$ for any positive integer $n$.
  14. practice Show that $\int_{-a}^a f(x) g(x) \mathrm{d} x = 0$ if $f$ is even and $g$ is odd.
  15. practice The value of $\int_{-2}^2 (x^5+x \cos 3 x+\sin^3 x) \mathrm{d} x$ is
    1. $0$
    2. $4$
    3. $16$
    4. cannot be determined
  16. practice If $\int_0^a 1/(4+x^2) \mathrm{d} x = \pi/8$, then $a$ equals
    1. $1$
    2. $2$
    3. $4$
    4. $\pi$
  17. practice Evaluate the definite integral $\int_0^3 (x-1) \mathrm{d} x$.
  18. practice The line $y=x-1$ crosses the $x$-axis at $x=1$. Find the true area enclosed between the line and the $x$-axis on $[0,3]$.
  19. practice True or false, with a reason: a definite integral that evaluates to $0$ means the function is zero everywhere on the interval.
  20. practice Evaluate $\int_0^1 x e^x \mathrm{d} x$.
  21. practice Evaluate $\int_2^3 (3 x+3)/((x-1)(x+2)) \mathrm{d} x$.
  22. practice Evaluate $\int_0^{\pi/2} \sin^2 x \mathrm{d} x$.
  23. practice Evaluate $\int_0^2 1/(x^2+4) \mathrm{d} x$.
  24. practice Evaluate $\int_1^2 (x^2+1)/x \mathrm{d} x$.
  25. practice The value of $\int_0^1 (2 x+1) e^{x^2+x} \mathrm{d} x$ is
    1. $e^2-1$
    2. $e^2$
    3. $e-1$
    4. $1$
  26. practice The value of $\int_{-1}^1 \left|x\right| \mathrm{d} x$ is
    1. $0$
    2. $1$
    3. $2$
    4. $-1$
  27. practice Prove that $\int_{-\pi}^\pi x^3 \sin^2 x \mathrm{d} x = 0$.
  28. practice Prove that $\int_0^{\pi/2} \sin^3 x \mathrm{d} x = 2/3$.
  29. practice Prove that $\int_0^1 1/((x+1)(x+2)) \mathrm{d} x = \ln(4/3)$.
  30. practice Evaluate $\int_0^{\pi/4} \tan x \mathrm{d} x$.
  31. practice Evaluate $\int_0^{\pi/2} 1/(1+\tan x) \mathrm{d} x$.
  32. practice Evaluate $\int_0^1 e^x (1+x) \mathrm{d} x$.
  33. practice Evaluate $\int_0^{\pi/4} \sin^2 x \cos^2 x \mathrm{d} x$.
  34. practice Evaluate $\int_0^1 \sqrt{1-x^2} \mathrm{d} x$.
  35. practice Evaluate $\int_0^1 1/(x^2-4) \mathrm{d} x$.
  36. practice Evaluate $\int_0^1 (2 x)/((x^2+1)(x^2+2)) \mathrm{d} x$.
  37. practice Evaluate $\int_0^{\pi/2} \sin 2 x/(1+\sin^2 x) \mathrm{d} x$.
  38. practice Evaluate $\int_0^1 x^2 e^x \mathrm{d} x$.
  39. practice Evaluate $\int_0^1 1/((x+1)(x^2+1)) \mathrm{d} x$.
  40. practice Evaluate $\int_0^{\pi/2} x \sin x \cos x \mathrm{d} x$.
  41. practice Evaluate $\int_0^{2 \pi} \cos^5 x \mathrm{d} x$.
  42. practice Evaluate $\int_0^4 \left|x - 1\right| \mathrm{d} x$.
Answers
  1. $-7$
  2. $-30$
  3. $0$
  4. $0$
  5. $0$
  6. $64/5$
  7. $3 \pi$
  8. $\pi/4$
  9. $0$
  10. $\pi/4$
  11. $\pi^2/4$
  12. $2 \pi/3$
  13. Shown: by the reflection property $\int_a^b f(x) \mathrm{d} x = \int_a^b f(a+b-x) \mathrm{d} x$ with $a=0$, $b=1$; $a+b-x=1-x$ turns $x(1-x)^n$ into $(1-x) x^n$.
  14. Shown: an even function times an odd function is odd, and the integral of an odd function over an interval symmetric about $0$ is $0$.
  15. (A) $0$.
  16. (B) $2$.
  17. $3/2$
  18. $5/2$
  19. False: positive and negative contributions can cancel to give a net integral of $0$ even though $f$ is not the zero function on the interval.
  20. $1$
  21. $\ln 5$
  22. $\pi/4$
  23. $\pi/8$
  24. $3/2 + \ln 2$
  25. (A) $e^2-1$.
  26. (B) $1$.
  27. True: $x^3$ is odd and $\sin^2 x$ is even, so the product is odd, and an odd function integrates to $0$ over an interval symmetric about $0$.
  28. True: using $\sin^3 x = \sin x(1-\cos^2 x)$ and the substitution $t=\cos x$.
  29. True: the partial-fraction split $1/((x+1)(x+2)) = 1/(x+1) - 1/(x+2)$, integrated and evaluated at the limits, gives $\ln(4/3)$.
  30. $(1/2) \ln 2$
  31. $\pi/4$
  32. $e$
  33. $\pi/32$
  34. $\pi/4$
  35. $-(1/4) \ln 3$
  36. $\ln(4/3)$
  37. $\ln 2$
  38. $e-2$
  39. $(1/4) \ln 2 + \pi/8$
  40. $\pi/8$
  41. $0$
  42. $5$
Board paper practice
  1. practice $\int e^x (\tan x + \ln(\sec x)) \mathrm{d} x$ equals ______. (1 mark)
    1. $e^x \tan x + C$
    2. $e^x \ln(\sec x) + C$
    3. $e^x \sec x + C$
    4. $e^x (\tan x - \ln(\sec x)) + C$
  2. practice Assertion (A). $\int_{-\pi/2}^{\pi/2} \sin x \mathrm{d} x = 0$. Reason (R). $\sin x$ is an odd function, and $\int_{-a}^a f(x) \mathrm{d} x = 0$ whenever $f$ is odd. (1 mark)
    1. Both A and R are true and R is the correct explanation of A.
    2. Both A and R are true but R is not the correct explanation of A.
    3. A is true but R is false.
    4. A is false but R is true.
  3. practice Find $\int \sec^2 x/(1+\tan x) \mathrm{d} x$. (2 marks)
  4. practice Find $a > 0$ such that $\int_0^a (2 x+1) \mathrm{d} x = 12$. (3 marks)
  5. practice Evaluate $\int 1/((x-1)(x^2+1)) \mathrm{d} x$. (5 marks)
  6. practice A firm’s marginal cost of producing $x$ units is $C'(x) = 3 x^2-24 x+50$ rupees per unit, and the fixed cost is $C(0) = 500$ rupees. (i) Find $C(x)$. (1 mark) (ii) Find the cost of producing $10$ units. (1 mark) (iii) Find the extra cost of producing units $11$ to $20$. (2 marks)
  7. practice Evaluate $\int_0^{\pi/2} \sin x/(\sin x+\cos x) \mathrm{d} x$. (3 marks)
Answers
  1. (B) $e^x \ln(\sec x) + C$.
  2. (A) Both A and R are true and R is the correct explanation of A.
  3. $\ln|1+\tan x| + C$.
  4. $a = 3$.
  5. $(1/2) \ln|x-1| - (1/4) \ln(x^2+1) - (1/2) \tan^{-1}(x) + C$.
  6. (i) $C(x) = x^3-12 x^2+50 x+500$. (ii) ₹⁠$800$. (iii) ₹⁠$3900$.
  7. $\pi/4$.

↑ Back to top