Mathematics Learning Centre

Mathematical Induction

Mary Barnes
Sue Gordon

1987
c University of Sydney
Contents
1 Mathematical Induction 1
1.1 Why do we need proof by induction? . . . . . . . . . . . . . . . . . . . . . 1
1.2 What is proof by induction? . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 The Binomial Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Finding the formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6 Further exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2 Solutions to Exercises 11
Mathematics Learning Centre, University of Sydney 1

1 Mathematical Induction
Mathematical Induction is a powerful and elegant technique for proving certain types of
mathematical statements: general propositions which assert that something is true for all
positive integers or for all positive integers from some point on.

Let us look at some examples of the type of result that can be proved by induction.

Proposition 1. The sum of the first n positive integers (1, 2, 3, . . .) is 12 n(n + 1).

Proposition 2. In a convex polygon with n vertices, the greatest number of diagonal

that can be drawn is 12 n(n − 3).

Note, we give an example of a convex polygon together with one that is not convex in
Figure 1.

This is a convex polygon This polygon is not convex

Figure 1: Examples of polygons

A polygon is said to be convex if any line joining two vertices lies within the polygon or
on its boundary.
By a diagonal, we mean a line joining any two non-adjacent vertices.

As you see, the subject matter of the statements can vary widely. It can include algebra,
geometry and many other topics. What is common to all the examples is the number n
that appears in the statement. In all cases it is either stated, or implicitly assumed, that
n can be any positive integer.

1.1 Why do we need proof by induction?

A natural starting point for proving many mathematical results is to look at a few simple
cases. This helps us understand what is being claimed and may even give us some pointers
for finding a proof.

Let’s do this with Proposition 1. The results are recorded in the following table.
Mathematics Learning Centre, University of Sydney 2

n 1 2 3 4
sum of first 1 1+2=3 1 + 2 + 3 = 6 1 + 2 + 3 + 4 = 10
n numbers
1
2
n(n + 1) 1
2
×1×2=1 1
2
×2×3=3 1
2
×3×4=6 1
2
× 4 × 5 = 10

You may find this quite convincing and wonder whether any further proof is needed. If
a statement is true for all numbers we have tested, can we conclude that it is true for all
values of n?

To answer this, let’s look at another example.

Proposition 3. If p is any prime number, 2p − 1 is also a prime. Let us try some special
cases here too.

p 2 3 5 7
2p − 1 3 7 31 127

Since 3, 7, 31, 127 are all primes, we may be satisfied the result is always true. But if we
try the next prime, 11, we find that

211 − 1 = 2047 = 23 × 89.

So it is not a prime, and our general assertion is therefore FALSE.

Let us turn again to Proposition 1, and ask how many cases we would need to check,
before we could say for certain that it is true. Imagine getting a computer on the job and
setting it the task. No matter how many values of n we found the proposition to be true
for, we could never be sure that there was not an ever bigger value for which it was false.
This explains the need for a general proof which covers all values of n. Mathematical
induction is one way of doing this.

1.2 What is proof by induction?

One way of thinking about mathematical induction is to regard the statement we are
trying to prove as not one proposition, but a whole sequence of propositions, one for
each n. The trick used in mathematical induction is to prove the first statement in the
sequence, and then prove that if any particular statement is true, then the one after it is
also true. This enables us to conclude that all the statements are true.

Let’s state these two steps in more formal language.

The initial step

Prove the proposition is true for n = 1.
(Or, if the assertion is that the proposition is true for n ≥ a, prove it for n = a.)
Mathematics Learning Centre, University of Sydney 3

Inductive step
Prove that if the proposition is true for n = k, then it must also be true for n = k + 1.
This step is the difficult part, and it may help you if we break it up into several stages.

Stage 1 Write down what the proposition asserts for the case n = k. This is what you
are going to assume. It is often called the inductive hypothesis.

Stage 2 Write down what the proposition asserts for the case n = k + 1. This is what
you have to prove. Keep this clearly in mind as you go.

Stage 3 Prove the statement in Stage 2, using the assumption in Stage 1. We can’t give
you any recipe for how to do this. It varies from problem to problem, depending on the
mathematical content. You have to use your ingenuity, common sense and knowledge of
mathematics here. The question to ask is “how can I get from Stage 1 to Stage 2?”

Once the initial and the inductive step have been carried out, we can conclude immediately
that the proposition is true for all n ≥ 1 (or for all n ≥ a if we started at n = a.)

To explain this, it may help to think of mathematical induction as an authomatic “state-

ment proving” machine.
We have proved the proposition for n = 1.
By the inductive step, since it is true for n = 1, it is also true for n = 2. Again, by the
inductive step, since it is true for n = 2, it is also true for n = 3. And since it is true for
n = 3, it is also true for n = 4, and so on.

Because we have proved the inductive step, this process will never come to an end.
We could set the “machine” running, and it would keep going forever, eventually reaching
any n, no matter how big it may be.

Suppose there was a number N for which the statement was false. Then when we get to
the number N − 1, we would have the following situation:

The statement is true for n = N − 1, but false for n = N .

This contradicts the inductive step, so it cannot possible happen. Hence the statement
must be true for all positive integers n.

If you are familiar with computer programming, it may be helpful for you to compare
this argument with a looping process, in which a computation is carried out, an indexing
variable is advanced by one, and the computation is repeated.

The two processes have much in common. In a computer program you must begin by
setting the initial value of your variables (this is analogous to our initial step). Then you
must set up the loop, calling on the previous values of your variables to calculate new
values (this is analogous to our inductive step).
Mathematics Learning Centre, University of Sydney 4

There is one other thing necessary in a computer program: you must set up a “stop”
condition otherwise your program will run forever. That has no analogy in our process –
our theoretical machine will run forever! That is why we can be certain our result is true
for all positive integers.

Now let’s see how this works in practice, by proving Proposition 1.

Proposition 1. The sum of the first n positive integers is 12 n(n + 1).

Initial step: If n = 1, the sum is simply 1.

Now, for n = 1, 12 n(n + 1) = 1
2
× 1 × 2 = 1. So the result is true for n = 1.

Inductive step:
Stage 1: Our assumption (the inductive hypothesis) asserts that
1
1 + 2 + 3 + · · · + k = k(k + 1).
2

Stage 2: We want to prove that

1 1
1 + 2 + 3 + · · · + (k + 1) = (k + 1)[(k + 1) + 1] = (k + 1)(k + 2).
2 2

Stage 3: How can we get to stage 2 from stage 1?

The answer here is that we get the left hand side of stage 2 from the left hand side of
stage 1 by adding (k + 1).
So,
1 + 2 + 3 + · · · + (k + 1) = 1 + 2 + 3 + · · · + k + (k + 1)
1
= k(k + 1) + (k + 1) using the inductive hypothesis
2
1
= (k + 1)( k + 1) factorising
2
1
= (k + 1)(k + 2) which is what we wanted.
2

This completes the inductive step.

Hence, the result is true for all n ≥ 1.

As a further example, let’s try proving Proposition 2.

Proposition 2. The number of diagonals of a convex polygon with n vertices is 12 n(n−3),
for n ≥ 4.

Initial step: If n = 4, the polygon is a quadrilateral, which has two diagonals as shown
in Figure 2.

Also, for n = 4, 12 n(n − 3) = 12 (4)(1) = 2.

Mathematics Learning Centre, University of Sydney 5

Figure 2: A quadrilateral showing 2 diagonals.

Inductive step:
Stage 1: The inductive hypothesis asserts that the number of diagonals of a polygon
with k vertices is 12 k(k − 3).

Stage 2: We want to prove that the number of diagonals of a polygon with (k + 1)

vertices is 12 (k + 1)[(k + 1) − 3] = 12 (k + 1))(k + 2).

Stage 3: How can we get to stage 2 from stage 1?

The answer here is to “add another vertex”. Let’s do this and see if we can count how
many additional diagonals can be drawn as a result. Figure 3 will help here.

Pk+1
Pk Pk

P1 P1

Pk-1 P
k-1
P2 P2

P3 P3
P4 P4

Polygon with k vertices Polygon with (k +1) vertices

Figure 3: Adding another vertex to a polygon with k vertices.

When we add an extra vertex to a polygon with n vertices, all the lines which were
diagonals of the original polygon will still be diagonals of the new one. The inductive
Mathematics Learning Centre, University of Sydney 6

hypothesis says that there are 12 k(k − 3) existing diagonals. In addition, new diagonals
can be drawn from the extra vertex Pk+1 to all other vertices except the two adjoining it
(P1 and Pk ) giving us k − 2 extra diagonals. Finally, the line joining P1 and Pk , which
used to be a side of the polygon, is now a diagonal.

This gives us a total of

1
k(k − 3) + (k − 2) + 1 diagonals.
2
But,
1 1
k(k − 3) + (k − 2) + 1 = [k(k − 3) + 2k − 2]
2 2
1 2
= (k − k − 2)
2
1
= (k + 1)(k + 2), as required.
2

This completes the inductive step. Hence the result is true for all n ≥ 4.

At this point, it would be a good idea to go back and read over the explanation of the
process of mathematical induction thinking about the general explanation in the light of
the two examples we have just completed.

Next, we illustrate this process again, by using mathematical induction to give a proof of
an important result, which is frequently used in algebra, calculus, probability and other
topics.

1.3 The Binomial Theorem

The Binomial Theorem states that if n is an integer greater than 0,
n(n − 1) n−2 2 n(n − 1)(n − 2) n−3 3
(x+a)n = xn +nxn−1 a+ x a+ x a +· · ·+nxan−1 +an . (1)
2! 3!

For example,

(x + a)3 = x3 + 3x2 a + 3xa2 + a3 .

This useful theorem can be more compactly expressed as

n
(x + a)n = nCrxr an−r
r=0

but if you are not familiar with this notation, use equation (1).

Let’s prove this theorem by mathematical induction.

Initial step: Let n = 1. Then the left hand side of (1) is (x + a)1 and the right hand
side of (1) is x1 + a1 which both equal x + a.
Mathematics Learning Centre, University of Sydney 7

Inductive step:
Step 1: Assume the theorem is true for n = k, ie that

k(k − 1) k−2 2
(x + a)k = xk + kxk−1 a + x a + · · · + kxak−1 + ak .
2!

Stage 2: We want to prove that the theorem is true for n = k + 1, ie that

(k + 1)(k) k−1 2
(x + a)k+1 = xk+1 + (k + 1)xk a + x a + · · · + (k + 1)xak + ak+1 . (2)
2!

Stage 3: How do we get to stage 2 from stage 1?

Look at the left hand side of (2),

(x + a)k+1 = (x + a)(x + a)k .

but by the inductive hypothesis,

k(k − 1) k−2 2
(x + a)(x + a)k = (x + a)[xk + kxk−1 a + x a + · · · + kxak−1 + ak ]
2!
k(k − 1) k−1 2
= xk+1 + kxk a + x a + · · · + kx2 ak−1 + xak
2!

+xk a + kxk−1 a2 + ··· + kxak + ak+1

(k + 1)(k) k−1 2
(adding) = xk+1 + (k + 1)xk a + x a + · · · + (k + 1)xak + ak+1 .
2!

This is the right hand side as required. Hence the result is true for all n ≥ 1.
Now try some of the exercises for yourself.

1.4 Exercises

Prove the following statements by mathematical induction.

Note that many of these statements can also be proved by other methods, and sometimes
these other proofs will be neater and simpler. However, the purpose of these exercises is
to practice proof by induction, so please try to use this method, even if if you can see an
easier way.
1. The nth odd number is 2n − 1.

2. The sum of the first n odd numbers is n2 .

3. a. For every positive integer n, n(n + 1) is even.

b. For every positve integer n ≥ 2, n3 − n is a multiple of 6.
Mathematics Learning Centre, University of Sydney 8

4. For all n ≥ 1,

xn+1 − 1
= 1 + x + x2 + · · · + x n where x = 1.
x−1

5. For all n ≥ 1, (1 + x)n ≥ 1 + nx where x is any real number greater than −1.

6. For all ≥ 1,


n
1
r(r + 1) = n(n + 1)(n + 2).
r=1 3
ie,
1
1(2) + 2(3) + 3(4) + · · · + n(n + 1) = n(n + 1)(n + 2).
3

7. For all n ≥ 3, n2 ≥ 2n + 1.

8. For all n ≥ 4, 2n ≥ n2 .
(Hint: Use the previous exercise to help you in the inductive step.)

9. For all n ≥ 3, the sum of the interior angles of a polygon with n vertices is 180(n−2)◦ .

10. The number of non-empty subsets of a set of n elements is 2n − 1.

1.5 Finding the formula

The problems in the exercises were stated in a form which made it clear that mathematical
induction could be used to prove the results: they nearly all began with the words “for
all n ≥ . . .”. Sometimes you will have a more open-ended question to answer, such as the
following.

Exercise

What is the maximum numbeer of regions into which a plane can be divided by n straight
lines?

In this case, we have to analyse the problem carefully, step by step, looking for a pattern,
and then use that pattern as the basis for a proof by induction. Try this yourself, before
reading the solution.

Solution

First of all, if two of the lines were parallel, by moving one so that they were no longer
parallel, we could increase the number of regions. This is illustrated in Figure 4.
Mathematics Learning Centre, University of Sydney 9

3 regions 4 regions

Figure 4: Dividing the plane with two lines.

Secondly, if 3 of the lines were concurrent, we could increase the number of regions by
moving one so that they were no longer concurrent. This is illustrated in Figure 5.

6 regions 7 regions

Figure 5: Dividing the plane with three lines.

Hence, we conclude that, to get the maximum number of regions, none of the lines can
be parallel, and no 3 can be concurrent. We can now work out the first few cases which
are illustrated in Figure 6.

1 line, 2 regions 2 lines, 4 regions 3 lines, 7 regions 4 lines, 11 regions

Figure 6: Dividing the plane with one to four lines.

There is no obvious pattern in this, so we need to analyse more carefully what is going
on. What happens when we add an extra line?
When we added the third line, it cut each of the other lines at a distinct point. The new
line thus passed through three of the existing regions, dividing each in two and creating
3 new regions.

When we added the fourth line, it cut each of the others at a distinct point, and passed
through 4 of the existing regions, creating 4 new regions.
Mathematics Learning Centre, University of Sydney 10

And so on . . ..

When we add an nth line, it will cut through each of the existing (n − 1) lines at a distinct
point, and will pass through n of the existing regions creating n additional regions.

We can put this information in a table.

Number of lines 1 2 3 4 ··· n

Number of regions 2 2+2 2+2+3 2+2+3+4 2 + 2 + 3 + 4 + ··· + n

From this, we can write down an answer.

Maximum no. of regions formed by n lines = 2 + 2 + 3 + 4 + · · · + n

= 1 + (1 + 2 + 3 + 4 + · · · + n)
1
= 1 + n(n + 1) by Proposition 1
2
1
= [2 + n(n + 1)]
2
1 2
= (n + n + 2).
2

Not only have we obtained an answer to the question, we also have worked out how
to progress from one step to the next, and so we have the basis for a formal proof by
induction.

1.6 Further exercises

1. Write out a formal proof by mathematical induction that the maximum number of
regions the plane is divided into by n lines is 12 (n2 + n + 2).

2. Each of n famous scientists who meet at a conference (where n ≥ 2) wants to shake

hands with all the others. Work out how many handshakes there will be and prove
your result by mathematical induction.
Mathematics Learning Centre, University of Sydney 11

2 Solutions to Exercises

Solutions to exercises

Remember there may be may different ways of presenting a solution. The working given
is just one way.

1. The nth odd number is 2n − 1.

Initial step: The first odd number is 1 which is 2(1) − 1 so the assertion is true for
n = 1.

Inductive step:
Stage 1: The inductive hypothesis asserts that the nth odd number is 2n − 1.

Stage 2: We want to prove that the (n + 1)th odd number is 2(n + 1) − 1, ie 2n + 1.

Stage 3: How can we get to Stage 2 from Stage 1?

By the inductive hypothesis, the nth odd number is 2n − 1. The next odd number is
2 more than this, ie 2n − 1 + 2 = 2n + 1 as required.
Thus the result is true for all n ≥ 1.

2. The sum of the first n odd numbers is n2 .

Initial step: For n = 1, the sum is 1 which is 12 .

Inductive step:
Stage 1: The inductive hypothesis asserts that

1 + 3 + 5 + · · · + (2n − 1) = n2 .

(Note, we are using the result of question 1 here.)

Stage 2: We want to prove that the sum of the first (n + 1) odd numbers is (n + 1)2 .

Stage 3: Getting to stage 2 from stage 1.

The sum of the first (n + 1) odd numbers is equal to the sum of the first n odd
numbers plus the next odd number, namely, 1 + 3 + 5 + · · · + (2n − 1) + (2n + 1), as
each odd number is two more than the previous one. So,
1 + 3 + 5 + · · · + (2n − 1) + (2n + 1) = n2 + (2n + 1) the inductive hypothesis
2
= n + 2n + 1
= (n + 1)2 as required.

The result is therefore true for all n ≥ 1.

Mathematics Learning Centre, University of Sydney 12

3. a. For every positive integer n, n(n + 1) is even (ie is divisible by 2).

Initial step: Let n = 1. Then n(n + 1) = 2, which is even, so the proposition is

true for n = 1.

Inductive Step:
Assume that for n = k, k(k + 1) is even. If a number is even, it is a multiple of 2.
So, we can assume that k(k + 1) = k 2 + k = 2m for some integer m.

We want to prove that the theorem is true for n = k + 1, ie that (k + 1)[(k + 1) + 1]

is even.

Now
(k + 1)[(k + 1) + 1] = (k + 1)(k + 2)
= k 2 + k + 2k + 2
= 2m + 2k + 2 using the inductive hypothesis
= 2(m + k + 1) which is even as required.
Hence the result is true for all n ≥ 1.

b. For every positive integer n ≥ 2, n3 − n is a multiple of 6.

Initial step: For n = 2, n3 − n = 23 − 2 = 6 which is a multiple of 6 (6 × 1).

Inductive step:
The inductive hypothesis asserts that n3 − n is a multiple of 6, so let n3 − n = 6m
for some integer m. We want to prove that (n + 1)3 − (n + 1) is a multiple of 6.

(n + 1)3 − (n + 1) = n3 + 3n2 + 3n + 1 − n − 1
= (n3 − n) + (3n2 + 3n) rearranging the previous line
= 6m + 3n(n + 1) using the inductive hypothesis
= 6m + 3 × 2l we proved n(n + 1) even in 3(a)
= 6(m + l) which is divisible by 6.
The result is therefore true for all n ≥ 2.

4. For all n ≥ 1,

xn+1 − 1
= 1 + x + x2 + · · · + x n where x = 1. (1)
x−1

(The reason why x must not equal 1 is to avoid dividing by zero and has nothing to
do with the inductive process.)

Initial step: For n = 1, the left hand side of equation (1) is

x2 − 1 (x + 1)(x − 1)
= = x + 1, x = 1.
x−1 x−1
Mathematics Learning Centre, University of Sydney 13

The right hand side of (1) is 1 + x1 , so the proposition is true for n = 1.

Inductive step:
We assume the proposition is true for n = k, ie
xk+1 − 1
= 1 + x + x2 + · · · + x k x = 1.
x−1

We want to show that the proposition is true for n = k + 1. That is,

xk+2 − 1
= 1 + x + x2 + · · · + xk + xk+1 x = 1.
x−1

Let us look at the right hand side of the equation first.

xk+1 − 1
1 + x + x2 + · · · + xk + xk+1 = + xk+1 using inductive hypothesis
x−1
xk+1 − 1 + xk+1 (x − 1)
=
x−1
xk+1 − 1 + xk+2 − xk+1
=
x−1
xk+2 − 1
= the left hand side as required.
x−1
The proposition is thus true for all n ≥ 1.

5. For all n ≥ 1, (1 + x)n ≥ 1 + nx, where x is any real number greater than −1. Note
this means that (1 + x) > 0.

Initial step: For n = 1, (1+x)1 = 1+x and 1+nx = 1+x, so (1+x)n = 1+nx ≥ 1+nx
is true for n = 1.

Inductive step:
We assume that the proposition is true for n = k. That is, (1 + x)k ≥ 1 + kx, when
x ≥ −1.

We want to prove that (1 + x)k+1 ≥ 1 + (k + 1)x. (1)

Let us look at the inequality and try to relate it to the inductive hypothesis.

Now the left hand side of (1) is

(1 + x)k+1 = (1 + x)(1 + x)k
≥ (1 + x)(1 + kx) using the inductive hypothesis and (1 + x) > 0
= 1 + kx + x + kx2
≥ 1 + kx + x since kx2 ≥ 0
= 1 + (k + 1)x.
Hence (1 + x)k+1 ≥ 1 + (k + 1)x, as required.

So the result is true for all n ≥ 1.

Mathematics Learning Centre, University of Sydney 14

6. For all n ≥ 1,


n
1
r(r + 1) = n(n + 1)(n + 2).
r=1 3

That is,
1
1(2) + 2(3) + · · · + n(n + 1) = n(n + 1)(n + 2).
3

Initial step: For n = 1, the left hand side is 1(2) = 2 and the right hand side is
1
3
1(2)(3) = 2, so the assertion is true for n = 1.

Inductive step:
Assume that
1
1(2) + 2(3) + · · · + k(k + 1) = k(k + 1)(k + 2).
3

We want to prove that

1
1(2) + 2(3) + · · · + k(k + 1) + (k + 1)(k + 2) = (k + 1)(k + 2)(k + 3).
3

Starting with the left hand side,

1
1(2) + 2(3) + · · · + k(k + 1) + (k + 1)(k + 2) = k(k + 1)(k + 2) + (k + 1)(k + 2)
3
1
= (k + 1)(k + 2)[ k + 1]
3
1
= (k + 1)(k + 2)(k + 3) as required.
3
Thus the proposition is true for all n ≥ 1.

7. For all n ≥ 3, n2 ≥ 2n + 1.

Initial step: For n = 3, n2 = 9, 2n + 1 = 7, so n2 ≥ 2n + 1 for n = 3.

Inductive step:
Assume k 2 ≥ 2k + 1.
We want to prove that (k + 1)2 ≥ 2(k + 1) + 1.
Now, starting with the left hand side
(k + 1)2 = k 2 + 2k + 1
≥ (2k + 1) + 2k + 1 using inductive hypothesis
= 2(k + 1) + 2k
≥ 2(k + 1) + 1 since 2k ≥ 1 for k ≥ 1.
This is the right hand side.
Thus the result is true for all n ≥ 3.
Mathematics Learning Centre, University of Sydney 15

8. For all n ≥ 4, 2n ≥ n2 .

Initial step: Let n = 4. Then 2n = 16 and n2 = 16 so the proposition is true for

n = 4.

Inductive step:
Assume that 2k ≥ k 2 .

We want to prove that 2k+1 ≥ (k + 1)2 .

Now,
2k+1 = 2 × 2k
≥ 2 × k2 using inductive hypothesis
= k2 + k2
≥ k 2 + 2k + 1 k 2 ≥ 2k + 1 for all k ≥ 3 from Question 7
= (k + 1)2
Thus the proposition is true for all n ≥ 4.

9. For n ≥ 3, the sum of the interior angles of a polygon with n vertices is 180(n − 2)◦ .

Initial step: For n = 3, the polygon is a triangle and the sum of the angles of a
triangle is 180◦ which is 180(3 − 2)◦ , as required.

Inductive step:
Assume that the sum of the interior angles of a polygon with k vertices is 180(k − 2)◦ .

We want to prove that if a polygon has k + 1 vertices, the sum of its interior angles
is 180(k − 1)◦ .

Pk+1

Pk

P1

Pk-1

P2

P3
P4

Figure 7: A polygon with k + 1 vertices.

Mathematics Learning Centre, University of Sydney 16

Let us join the first and k th vertex in the polygon with k+1 vertices shown in Figure 7.
This divides the polygon into a triangle (P1 Pk Pk+1 ) and a polygon with k vertices.

The sum of the interior angles of the polygon so formed is 180(k−2)◦ , by our inductive
hypothesis, and the sum of the angles of the triangle is 180◦ .

Thus the sum of the interior angles of a polygon with k + 1 vertices is

180(k − 2)◦ + 180◦ = 180(k − 1)◦ as required.

Thus the proposition is true for all n ≥ 3.

10. The number of non-empty subsets of a set of n elements is 2n − 1, for any positive
integer n.

Initial step: If a set has one element in it then the only non-empty subset is the set
itself. Thus there is one non-empty subset and 21 − 1 = 1, so the assertion is true for
n = 1.

Inductive step:
Assume that if a set has k elements, then it has 2k − 1 non-empty subsets.

We want to prove that if a set has k + 1 elements, it has 2k+1 − 1 non-empty subsets.

Let us suppose the set has k elements in it plus a new element we will denote by p as
shown in Figure 8.

c
a

d
b

Figure 8: A set with k elements and a new element p.

Now the number of non-empty subsets that can be formed that do not contain p is
2k − 1 by the inductive hypothesis.
We can adjoin the element p to each of these subsets to form 2k − 1 new subsets.
Finally there is one subset whose only element is p.
Hence the total number of subsets is
(2k − 1) + (2k − 1) + 1 = 2(2k ) − 1
= 2k+1 − 1 as required.

Thus the proposition is true for all n ≥ 1.

Mathematics Learning Centre, University of Sydney 17

Short solutions or hints to further exercises

1. We want to prove that the maximum number of regions a plane can be subdivided
into by n lines is 12 (n2 + n + 2).

Initial step: If n = 1, the plane is divided into 2 regions and 12 (12 + 1 + 2) = 2, so the
assertion is true for n = 1.

Inductive step:
1
Now suppose we have n lines and 2
(n2 + n + 2) regions. (This is the inductive
hypothesis.)
Add one more line, which meets the existing lines at n distinct points. These n points
subdivide the new line into n + 1 segments. Each segment divides one of the existing
regions into two, so the number of regions is increased by n + 1.

Now check that

1 2 1
(n + n + 2) + (n + 1) = [(n + 1)2 + (n + 1) + 2].
2 2
This completes the proof.

2. (Hint) If there are only 2 scientists, there will be only one handshake. If one more
arrives, he or she will shake hands with both of the others, making 3 handshakes in
all.

Now imagine there are n people in the room, and one more arrives and shakes hands
with all the others. There will be n additional handshakes.