My IB Annoted IA S 3 PDF

Example 1: Student work
Breaking the Code
“To keep your secret is wisdom; but to expect others to keep it is folly.”
Samuel Johnson
“Secrets are made to be found out with time”

Charles Sanford
Codes have been used by the military to keep secrets from the enemy for thousands
of years. In the information age, when many people use the internet for banking and
shopping, they do not want the information they enter to be available to unauthorised C Student could
people so such information is encrypted on secure websites.
have related this
to personal interest.
This report looks at some of the codes which have been used over time and how
mathematics is used in making and breaking them.
A No aim or
The Caesar Shift Cipher rationale in the
introduction
The first documented use of codes for military purposes was by Julius Caesar (100-
44BC). The type of code he used is commonly known as a Caesar shift. In the
following table the middle row gives plain text and the bottom row gives the
corresponding cipher text for a Caesar shift of 2 places. It is conventional to write
uncoded plain text using lower case letters and coded text using upper case.
Position 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Plain a b c d e f g h i j k l m n o p q r s t u v w x y z
Code C D E F G H I J K L M N O P Q R S T U V W X Y Z A B
This code can be represented using the mapping x → x + 2 where x stands for the
position of the letter in the alphabet. So k, with postion number 10, becomes M with
position number 12. Towards the end of the alphabet, y would go to the letter with
position number 24 + 2 = 26 but 25 is the maximum position number as there are
only 26 letters in the alphabet and we used 0 for a. 26 is subtracted to give 0. This is
called addition modulo 26.
Positions 0 to 25 have been used (rather than 1 to 26) because another way to think
of modular arithmetic is as a remainder from division. 26 is equivalent to 0 modulo 26
because 26 ÷ 26 = 1 rem 0 .
Using this code, the second quote at the start of this report would be encoded as
follows:
C Student's
secretsaremadetobefoundoutwithtime own example
UGETGVUCTGOCFGVQDGHQWPFQWVYKVJVKOG with connection
to the topic
Spaces between words have been removed to make it more difficult to break the
code but if someone found this message and knew that it was a Caesar shift code it
would not take long to break. There are only 25 possible Caesar shifts; 26 if we
include the one that does not move the letters at all but that one is not very secret!
Mathematics SL and HL teacher support material 1

It is possible to write out all 25 possibilities using nothing more complicated than pen
and paper and once you have a message that makes sense you know you have
decoded it. Using a computer to do all 25 possible decodings would make the
process of decoding very quick.
Substitution ciphers
In a substitution cipher each letter is replaced by another letter. A Caesar shift is one
type of substitution cipher but the pattern of substitutions in a Caesar shift makes it
easy to break. A general subsitution cipher can be made by rearranging the 26 letters
of the alphabet and writing this random permutation below the ordinary alphabet. B Good use of
There would be 26! ≈ 4 ×1026 possible codes. It would take a long time to go through approximation notation
all the possible decodings to find the right one! However, some of the possible
substitution codes are too easy to break; the one where each letter is encoded by the
same letter is no use but nor are those where a large number of letters do not
change.
Permutations where nothing stays in its original place are called derangements. The
number of derangements of 26 letters is the subfactorial of 26, denoted !26. To see
how to work this out, consider a much shorter alphabet of 4 letters. The total number
of permutations is 4! = 24. These are shown below.
abcd abdc acbd acdb adbc adcb
bacd badc bcad bcda bdac bdca
cabd cadb cbad cbda cdab cdba
dabc dacb dbac dbca dcab dcba
The 9 derangements are shown in bold, with a box round them. To calculate the
number of derangements, start with the total number of permutations and take away
the ones which have at least one letter in its original position.
There are 6 arrangements with a in its original position, these are on a grey
background. This is because, with a in its original position, there are 3 other letters
which can be arranged in 3! = 6 ways. Likewise, there are 3! permutations with b in its
original place and the same for c and d. However, subtracting 4 × 3! will be too much
as arrangements with both a and b in their original places have been counted twice.
B Mathematical
So 4!− 4 × 3! (1.1) is too small. explanation is
well presented.
If a and b are both in their places, there are 2 letters left which can be arranged in 2!
ways. The same is true for other pairs of letters. There are 4C2 pairs of letters and so
4
we need to add on C2 × 2! to (1.1) to give
4!− 4 × 3!+ 4C2 × 2! (1.2)
However, arrangements which have 3 letters in the same places have been counted
more than once. In fact, it is not possible for 3 letters to be in the same place and the
4th letter to be in a different place so it is only the one arrangement with all four letters
in the same place that has been counted too many times. abcd has been counted 4
times but it should only be counted once.

The number of derangements is
4!− 4×3!+4C2 × 2!− 4 +1 (1.3)
This can be written as
C Good demonstration
4! 4! 4! 4!  1 1 1 1
!4 =4!− + − + =4!1 − + − +  (1.4) of learning and describing
1! 2! 3! 4!  1! 2! 3! 4!  unfamiliar mathematics
In general
B Good mathematical
 1 1 1 1 definition of terms
n n !1 − + − + ....... + (−1) n  (1.5)
!=
 1! 2! 3! n ! E Work here is beyond
syllabus requirements.
The Maclaurin series for ex is
x 2 x3 xr
e x =1 + x + + + ......... + + .........
2! 3! r!
1 1 1 n 1 E Good mathematical
so 1 − + − + ...... + (−1) ≈ e −1
1! 2! 3! n! understanding
demonstrated here
n!
!n ≈
e
n!
In fact, rounding gives !n so the number of derangements of the alphabet is
e
26!
!26 ≈ ≈ 1.5 ×1026 . This is still far more possibilities than could be run through in a
e
reasonable time but substitution codes can be broken quickly and easily.
Breaking a substitution cipher
In English, some letters are used more often than others. The following graph shows
how often the different letters are used.
% frequency in English
14
12
10
8
6
4
2
0
a b c d e f g h i j k l m n o p q r s t u v w x y z
If a substitution code is used then the letter that has been used to encode e will be
the most frequent letter. Using computers, counting frequencies of letters is quick and
accurate so substitution codes are not secure. The earliest known description of
frequency analysis for breaking codes dates from the 9th century and is by the Arabic

scholar Al-Kindi. In addition to looking at frequencies of single letters, the frequencies C Opportunity for
of pairs of letters can be compared with commonly occuring pairs such as th, er, on
and so on.
student to demonstrate
personal engagement
with an example
The Vigenère Cipher
In the 16th century Blaise de Vigenère built on the work of others to develop a cipher
which was not vulnerable to frequency analysis but which was simple to use. It
makes use of Caesar shifts but does not use the same shift all the time. To use the
cipher, a keyword is needed. The person doing the coding and the person receiving
the message need to know the keyword but that is all they need to remember. By
contrast, for a substitiution cipher which is not a Caesar shift, the whole table for
coding needs to be either remembered or written down.
The method is illustrated using the keyword STELLA. Caesar shift alphabets which
start with S, T, E, L and A will be used.
Plain a b c d e f g h i j k l m n o p q r s t u v w x y z
Code S T U V W X Y Z A B C D E F G H I J K L M N O P Q R
Code T U V W X Y Z A B C D E F G H I J K L M N O P Q R S
Code E F G H I J K L M N O P Q R S T U V W X Y Z A B C D
Code L M N O P Q R S T U V W X Y Z A B C D E F G H I J K
Code A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
To code the initial quote from this report, the keyword is written repeatedly above the
text to be coded:
stellastellastellastellastellastellastellastellastellast C Use of own

tokeepyoursecretiswisdombuttoexpectotherstokeepitisfolly
example by student
The first letter of the message is encoded using the Caesar alphabet which starts
with S so it becomes L, the second letter of the message is encoded using the
alphabet starting with T so it becomes H. The encoded message is as follows:
LHOPPPQHYCDEUKIETSOBWOZMTNXEZEPIINEOLAICDTGDIPAILBWQZLDR
The Vigenère cipher is not susceptible to breaking by frequency analysis as each

letter is coded in different ways. However, it was not widely used because to code
each message manually is time consuming as you have to choose the alphabet to
use and then find the appropriate encoded letter. Mechanising the process would
make it much easier but the use of technology was also what led to the code being
broken around 400 years after it was first developed.
Breaking the Vigenère cipher
Charles Babbage, who developed an early computer, realised that to break the
Vigenère code the first thing to do was to decide the length of the keyword.
Consider the following encoded text:
JERFIWFLXFIWFLXVHXEAMCNWVHXHIWFLX C Further example
Repeated sets of letters in the encoded message are shown in red. They are likely to
come from the same letters in the original message being encoded in the same way

and, in this case, it is likely that the keyword is 3 letters long because the distance
between repeats is a multiple of 3.
123123123123123123123123123123123
JERFIWFLXFIWFLXVHXEAMCNWVHXHIWFLX
If the code word has 3 letters then all the letters numbered 1 above have been coded
with the same Caesar shift. The same applies for letters numbered 2 and letters
numbered 3. Frequency analysis can now be used separately for each set of letters
with the same number to break the code. There is a useful computerised tool for
breaking Vigenère codes at
http://www.simonsingh.net/The_Black_Chamber/cracking_tool.html
The Enigma Machine
To make adaptations of the Vigenère cipher more secure, a longer keyword needs to
be used and it should not be a real word so that guesswork cannot be used in working
it out. This makes it much harder to use the Vigenère cipher unless the process can
be automated. In 1918, a German inventor, Arthur Scherbius, invented the Enigma
machine. It was further developed and used by the German army during the second
world war.
The Enigma machine had a keyboard. Operators typed in a message and the
machine encoded it. The basic idea is that a rotor is used to move on and encode the
next letter using a different substitution alphabet. With one rotor and 26 positions, this
would give 26 possible initial starting postions and so 26 possible codes. However 3
rotors were used and each was set to a different starting position every day giving
26 × 26 × 26 = 17 576 possible codes.
Operators also chose 3 rotors from 5 possible rotors, ie 5P3 = 60 ways of positioning
the rotors so this now gives 60×17 576 = 1 054 560 possible codes.
In addition to the 3 rotors, pairs of letters could be connected on the keyboard. If a

was connected to b, this would swap the encryption of a with the encryption of b. 10
connections were made, choosing 10 pairs of letters from 26.
The number of ways of choosing the first pair then the second pair and so on is
26 26! 24! 22! 8! E Good demonstration of

C2 × 24C2 × 22C2 × .... × 8C=
2 × × × ..... ×
24!× 2! 22!× 2! 20!× 2! 6!× 2! mathematical understanding
(1.6)
26!
=
6!× 210
However, the order of selecting the pairs does not matter so each set of 10 pairs has
been counted too many times; the answer in (1.6) needs to be divided by 10!
26!
≈ 1.5 ×1014
6!× 210 ×10!
Taken together with the options for the rotors, this gives a total of
1.5 ×1014 ×1 054 560 ≈ 1.59 ×1020 possible codes. With no possibility of using
frequency analysis, the Enigma code was considered to be unbreakable.

Breaking the Enigma Code
Operators of Enigma machines were issued with books which gave them the settings
for the day. They were told which letters to connect on the keyboard, which rotors to
use and the initial setting for the rotors. If the code breakers had an Enigma machine
and could find out the settings for the day, they would be able to decipher all the
messages for that day.
Messages would start with the Enigma machine in this position and then the setting of
the rotors for the message would be transmitted; the rotors were then adjusted before
transmitting the rest of the message. This meant that even if the code breakers had
all the messages for the day, each one would use a different code (apart from the first
3 letters of each message). This did not give the code breakers very much to go on.
To break the Enigma code, the code breakers used a copy of an Enigma machine and
constructed machines, called bombes, which could work through a large number of
possible codes. The machines could not work through anything like all possible codes
in one day so further information and deduction was necessary.
The Enigma machine never encoded a letter as itself. This helped with the decoding
of some messages. One day a message came through which had no letter L in it.
The operator had been sending a test message and had just pressed the letter L
every time so it was the only letter that did not feature in the encoded message.
Knowing that messages which were sent at particular times of day were about the
weather enabled the code breakers to guess part of the message. Knowing that a
letter could not be encoded as itself, helped them narrow down the possibilities for
which of the possible codes was being used that day and the bombes could then
search for that day’s settings.
Suppose the word “weather” is known to be part of the plain text for the following
coded message. The word can be tried in different positions, as follows
Coded message: QZWPM ZHVVG YGQOZ UQZX

Positioning plain text: weath er
weat her C Another personal example
wea ther
The third of these possible positions cannot be the place where the word “weather”
was in the text because w cannot code to W and h cannot code to H.
D Student could have

Public key cryptography
reflected on the implications
of breaking this code.
All the above systems of encoding rely on the key being kept secret from anyone
unauthorised who might try to decode the message. The key might be the shift used
for a Caesar shift, the keyword for a Vigenère cipher or the initial settings for Enigma
machine but once someone knows it they can decode a message.
In the 1970s, Rivest, Shamir and Adleman came up with a way to encode messages
which did not rely on the key being kept secret. The system they invented is
sometimes known as RSA. The system encodes numbers rather than letters but it is
possible to have a system for making letters or words into numbers. The table below
shows the steps in the process by using a simple example.

Step Simple example

Choose 2 prime numbers, p and q. For p = 2 , q = 11
security, these need to be large.
m = pq m = 22
Work out A = ( p −1)(q −1) and choose a A = (p −1)(q −1) = 1×10 = 10
number E which is smaller than this and E=3
has no factors in common with it.
Find a number D such that DE − 1 is a 3 D −1 must be a muliple of 10
multiple of A D=7
To encipher a number M, work out
C = M E (modulo m)
Examples are given below C Own example
To decipher a number C, work out
M = C D (modulo m)
E and m need not be kept secret as long as D is kept secret.
Enciphering
M C
1 13 (mod 22) = 1
2 23 (mod 22) = 8 (mod 22) = 8
3 33 (mod 22) = 27 (mod 22) = 5
4 43 (mod 22) = 64 (mod 22) = 20
Remember from the section on Caesar ciphers that to find 64 (mod 22) you can find
the remainder when 64 is divided by 22. However, working with powers of numbers
can lead to calculations with very large numbers. Fortunately mathematics can help.
Suppose you want to work out ab (modulo 15). Suppose a = r (modulo 15) and
b = R (modulo 15) .
So a = 15x + r and b = 15 y + R where x and y are integers.
ab = (15x + r )(15 y + R ) = 152 xy + 15xR + 15r + rR

The remainder when dividing ab by 15 is the same as the remainder when dividing
rR by 15. So to find what a large number is modulo 15, it is easier to write the large
number as a product of numbers and find each of those modulo 15 and then
multiply the answers. This can be summed up as follows:
ab (modulo m) = ((a modulo m )(b modulo m )) modulo m .

E Good mathematics
3 2
15 (mod 22) = (15 ×15)(mod 22) =
(225 mod 22) ×15(mod 22) demonstrated here and
So
= (5 ×15) mod 22 =75 mod 22 =9 throughout the next page

Deciphering
M C
1 17 (mod 22) = 1
8 87 (mod 22) = (82 mod22) × (82 mod22) × (82 mod22) × (8 mod 22) (mod 22)
= ((64 mod 22)3 × 8) mod 22
= (203 × 8) mod 22
= 64 000 mod 22
=2
5 57 (mod 22) = (52 mod22) × (52 mod22) × (52 mod22) × (5 mod 22) (mod 22)
= ((33 mod 22) × 5) mod 22
= ((27 mod 22) × 5) mod 22
= (5 × 5) mod 22
=3
20 207 ≡ 202 × 202 × 202 × 20

≡ 4 × 4 × 4 × 20
≡ 64 × 20
≡ 20 × 20
≡ 400 ≡ 4
To keep writing “mod 22” is making this look more complicated than necessary so, for
deciphering 20, a different notation has been used. ≡ means “has the same
remainder when divided by 22”.
The tables above demonstrate that deciphering takes you back to the original number
for the particular examples tested. This will always work. The proof depends on
Euler’s theorem, which states that
M φ(m) ≡ 1mod m
φ(m) is the number of integers smaller than m and having no common factor with with
m. If m = pq where p and q are prime, ϕ(m) = ( p −1)(q −1) and this is what we called
A in the description of the RSA algorithm.
So Euler’s theorem implies that M A ≡ 1mod m
Using ≡ to mean “has the same remainder when divided by m” and remembering the
formula for deciphering is M = C D (modulo m)
C≡ME
C D ≡ (M E )D
≡ M ED = M ED−1 × M
ED −1 is a multiple of A so ED −1 = kA where k is an integer.

So
C D ≡ M kA × M = (M A )k × M
≡ 1k × M = M

Breaking the RSA code
m = pq is known; to break the code needs the individual primes, p and q to be known.
m = 22 does not give a secure code as it is obvious that the primes were 2 and 11.
The primes need to be much larger for the code to be secure.
655 427 is the product of two primes. To find the primes a spreadsheet could be
used. This is the start of a spreadsheet showing one way to find the primes:
n 655427/n Integer part of 655427/n Factor?

1 655427 655427 yes
2 327713.5 327713 no
3 218475.6667 218475 no
4 163856.75 163856 no
5 131085.4 131085 no
6 109237.8333 109237 no
7 93632.42857 93632 no
Rather than just trying to divide by primes, all integers are tested so 1 does go into
655 427 but this does not help. The prime factors are found further down:
438 1496.408676 1496 no

439 1493 1493 yes
440 1489.606818 1489 no
Constructing the spreadsheet and finding the factors took about 1 minute so 655 427
would not give a secure code.
C Timed self
doing spreadsheet
The factorising can be speeded up by using Fermat’s factorising method. This
depends on the following result. D Reflection on
2 2 significance of
 p+q  p−q
m pq
= =   −  personal timing
 2   2 
Multiplying out the brackets on the right hand side gives
p 2 + q 2 + 2 pq p 2 + q 2 − 2 pq
− pq
=
4 4
p+q p−q
If p and q are large prime numbers, they will both be odd so and are
2 2
whole numbers.
Suppose m = x2 − y2 then x2 − m = y 2. This means x2 ≥ m as square numbers cannot

be negative so the smallest m could be is m.
We are trying to factorise m = 655 427 . 655 427 = 809.58... so the smallest possible
value of x is 810.
For each possible value of x find the value of x2 − m . When this is a square number,
y2 , the factors are ( x − y )( x + y ) . Doing the calculations on a spreadsheet gives the
following:

n 810+n (810+n)^2-655427=Y SQRT(Y) Factor Factor

1 811 2294 47.8957 no no
2 812 3917 62.5859 no no
3 813 5542 74.4446 no no
4 814 7169 84.6699 no no
5 815 8798 93.7977 no no
6 816 10429 102.122 no no
155 965 275798 525.165 no no

156 966 277729 527 1493 439
157 967 279662 528.831 no no
The factors are found after testing 156 numbers whereas the previous method took
439 numbers. It is so quick to do for a 6 digit value of m that it makes little difference
but it shows that computing speed and efficient methods of factorising can enable the
RSA code to be broken unless the prime numbers used are very large. The current
record for successfully factoring a number is 200 digits.
“In the eighties it was generally held that prime numbers of fifty odd digits would
suffice. However, developments went much faster than foreseen, and it is a
precarious matter to venture upon quantitative forecasts in this field. When Rivest
challenged the world in 1977 to factor RSA-129, a 129 digit number (from a special
list), he estimated that on the basis of contemporary computational methods and A No conclusion
computer systems this would take about 1016 years of computing time. Seventeen
years later it took only eight months in a world-wide cooperative effort to do the job.”
D Many opportunities for
http://www.cwi.nl/en/RSA reflection: What about
the future? Will faster
Bibliography
computers eventually
Cryptology timeline www.math.cornell.edu/~morris/135/timeline.html
eliminate the ability to
encode messages? What
An introduction to cryptography other implications could
http://www.math.sunysb.edu/~scott/papers/MSTP/crypto/crypto.html there be?
The Code Book, Simon Singh, Fourth Estate (2000)
Derangements http://www.mathlab.mtu.edu/~eewestlu/ma3210_lecture19.pdf
Codes and ciphers http://www.bletchleypark.org.uk/edu/teachers/ccresources.rhtm
http://en.wikipedia.org/wiki/Cryptanalysis_of_the_Enigma
Euler function and theorem http://www.cut-the-knot.org/blue/Euler.shtml
Mathematical formulae http://www.po28.dial.pipex.com/maths/formulae.htm
http://www.cwi.nl/en/RSA

M AT H S C O U R S E WO R K
Proving Euler’s Totient Theorem

TA B L E O F C O N T E N T S
Maths Coursework 3
Introduction 3
The Theorem 3
Euler’s Totient Function 3
Fermat’s Little Theorem 4
Proving Fermat’s Little Theorem 5
Proving Euler’s Totient Theorem 6
Applications 7
Conclusion 7

M AT H S C O U R S E WO R K
Proving Euler’s Totient Theorem A Aim
Introduction
My interest in Euler as a mathematician was first sparked when, on completing a listener
crossword, the hidden message “Read Euler, he is the master of us all” was revealed, so
when I saw the inclusion of his name on the list of prompt words there was really no
option but to go for him. Euler was a mathematician in the 18th century and is responsible
for the first proofs of many great many number of conjectures and problems. In number
theory alone his accomplishments include proving the two square theorem and Fermat’s
little theorem as well as doing a great deal of work that later led to the first proof of the
four square theorem. His achievement that I am going to focus on though is less well
known, it is a generalisation of Fermat’s little theorem that has come to be known as
A Introduction
Euler’s totient theorem.
and rationale
The Theorem
Euler’s totient theorem1 states that for relatively prime a and n:
B Class knew
about modular
aΦn ≡ 1 (mod n) arithmetic,
so this didn't
need definition.
Where Φn is Euler’s totient function
Euler’s Totient Function

Euler’s totient function2, or Φn, is a count of the numbers that are less than n and
relatively prime to n. For example Φ10 is 4 as there are four number less than ten that are
relatively prime to 10 { 1, 3, 7, 9 }, Φ11 is 10 as 11 is prime all numbers less than it are
relatively prime to it and Φ6 is 2 as 1 and 5 are relatively prime to 6 but 2,3 and 4 are not.
1
http://en.wikipedia.org/wiki/Euler's_theorem
2
http://mathworld.wolfram.com/TotientFunction.html

B Small n in
Below is a table of the totients of the numbers up to 20. text. Condoned
N Φ N
2 1
3 2
4 2
5 4
6 2
7 6
8 4
9 6
10 4
11 10
12 4
13 12
14 6
15 8
16 8
17 16
18 6
19 18
20 8
Some examples will serve to demonstrate Euler’s totient theorem.

Let n = 10 and a = 3. Note that 10 and 3 are relatively prime. From the table Φ10 = 4.
Then, 34 = 81 ≡1(mod 10).
Also, if n = 15 and a = 2 we see that 28 = 256 ≡ 1 (mod 15).
Fermat’s Little Theorem

Euler’s totient theorem is a generalisation of Fermat’s little theorem3 and works for all n
relatively prime to a. Fermat’s little theorem only works for a and p relatively prime
3 http://mathworld.wolfram.com/FermatsLittleTheorem.html

where p is itself prime and states:
ap ≡ a (mod p)
or
ap-1 ≡ 1 (mod p)
It is immediately apparent that this fits in with Euler’s totient theorem for primes p, as we
have seen Φp, where p is a prime, is always p-1.
As an introduction to Euler’s totient theorem I shall prove Fermat’s little theorem.
Proving Fermat’s Little Theorem

RTP: ap ≡ a (mod p)
Take two numbers a and p which are relatively prime, and where p itself is prime.
Consider the set of the multiples of a { a, 2a, 3a, 4a, 5a ..... (p-1)a }
Consider the set of numbers { 1, 2, 3, 4, 5 ..... (p-1) }

If taken to the modulus p each element of the first set will be congruent to an element in
the second, there will be one to one correspondence between the two sets and this is
proven as lemma 1.
If we take the product of the first set { a x 2a x 3a x 4a x 5a ...... (p-1)a } and the product
of the second { 1 x 2 x 3 x 4 x 5 ..... (p-1) } we can see that they are congruent to one
another (as each element in the first is congruent to an element in the second)
Therefore { a x 2a x 3a x 4a x 5a ...... (p-1)a } ≡ { 1 x 2 x 3 x 4 x 5 ..... (p-1) } (mod p)
We can take out a factor of ap-1 from the left hand side
Giving ap-1 { 1 x 2 x 3 x 4 x 5 ..... (p-1) } ≡ { 1 x 2 x 3 x 4 x 5 ..... (p-1) } (mod p)

By dividing each side by { 1 x 2 x 3 x 4 x 5 ..... (p-1) } which is valid as p is prime we get
ap-1 ≡ 1 (mod p)
or
ap ≡ a (mod p)
QED.

Lemma 1: Each number in the first set must be congruent to one and only one number in
the second and each number in the second set must be congruent to one and only one
number in the first. This may not be obvious at first but can be proved through three
logical steps.
(1) Each number in the first set must be congruent to one of the elements in the second as
all possible congruences save 0 are present, none will be congruent to 0 as a and p are
relatively prime.
(2) A number cannot be congruent to two numbers in the second set as a number can only
be congruent to numbers which differ by a multiple of p, as all elements of the second
set are smaller than p a number can only be congruent to one of them.
(3) No two numbers in the first set, call them ba and ca, can be congruent to the same
number in the second. This would indicate that the two numbers were congruent to
each other ba ≡ ca (mod p) which would indicate that b ≡ c (mod p) which is not true
as they are both different and less than p itself.
Therefore, through these three steps Lemma 1 is proven.

As Fermat’s little theorem is a special case of Euler’s totient theorem (where n is prime)
the two proofs are quite similar and in fact only slight adjustments need to be made to the
proof of Fermat’s little theorem to give you Euler’s totient theorem4.
RTP: aΦn ≡ 1 (mod n)

Take two numbers, a and n which are relatively prime
Consider the set N of numbers that are relatively prime to n { 1, n1, n2...nΦn }
This set will have Φn elements (Φn is defined as the number of numbers relatively prime to n)
Consider the set aN, where each element is the product of a and an element of N { a, an1,
an2... anΦn }
Each element in set aN will be congruent to an element in set N (mod n), this is follows by
the same argument as in lemma 1 and so the two sets will be congruent to each other
Therefore { a x an1 x an2 x ... x anΦn } ≡ { 1 x n1 x n2 x ... x nΦn } (mod n)
4
http://planetmath.org/?op=getobj&from=objects&id=335

By taking out a factor of aΦn from the left hand side we get
aΦn { 1 x n1 x n2 x ... x nΦn } ≡ { 1 x n1 x n2 x ... x nΦn } (mod n)
If we then divide through by { 1 x n1 x n2 x ... x nΦn } which is valid as all elements are
relatively prime to n we get
A Complete.
aΦn ≡ 1 (mod n) Achieves aim.
Concise!
QED.
Applications
Unlike some of Euler’s other work in number theory such as his proof of the two square
theorem Euler’s totient theorem has very real uses and applications in the world and like
much of number theory those uses are almost exclusively in the world of cryptography
and cryptanalysis. Both Fermat’s little theorem and Euler’s totient theorem are used in the
encryption and decryption of data, specifically in the RSA encryption system5, whose
protection revolves around large prime numbers raised to large powers being difficult to
factorise.
D Reflection. Linking
mathematical ideas
Conclusion
This theorem may not be Euler’s most elegant piece of mathematics (my personal
favourite is his proof of the two square theorem by infinite descent) or at the time seemed
like his most important piece of work at the time but this, in number theory at least, is
probably his most useful piece of mathematics to the world today.
This proof has given me a chance to link up some of the work I have done in the largely
separate discrete mathematics and sets relations and groups options. These two options
appear to me to be the purest sections of mathematics that I have studied but are for
whatever reason seldom linked in class, this project has allowed me to explore the links
between them and use knowledge from one in relation to the other, broadening my view
of maths.
A Conclusion
C Personally
involved
5
http://www.muppetlabs.com/~breadbox/txt/rsa.html#7

Mathematical Exploration: Minesweeper
Used with permission from Microsoft
‘’Minesweeper’’, the common household computer game installed in many

Microsoft PC’s today. The game itself is on-par with Sudoku for being one of the worlds
most time-consuming mathematical puzzle game. As simple as the game may be,
mathematicians al of the world toil over this game, hopes of looking strategies,
theorems and perhaps solutions that lead to solving problems of mathematical
applications. The ones who call themselves ‘’experts’’ of this game truly possess great
deduction, analytical and logical skills that can be applied to today’s mathematics.
A Introduction but
no aim or rationale
Overview
The original minesweeper computer game whose objective is to ‘’clear a minefield
without detonating a mine’’. There are many strategies involved with this common
household game.
When the game is started, the player is presented a grid of squares. The size of the grid
varies from the difficulty level chosen by the player. The larger the grid the higher the
difficulty level and skills required. The number of mines also varies from difficulty level;
usually in the beginner level there are 10 mines in a 9x9 grid. The game progresses by
the player’s clicks on a square. The square clicked will either a mine or reveal a square
with a digit. If the square revealed turns out to be a mine, the game is over. The game
progresses as the player clicks on squares without mines and reveals digits with the
highest digit being 8. The digit indicates the number of adjacent squares that contain
mines. Using this information, players can deduce the certainties of other squares on
whether they are mine-filled or free and continue on with the game by either flagging a
square or proceed to clearing other squares. For example, if you left-clicked a square,

and an 8 digit appeared, then you would know that the square is surrounded by 8
mines, all 8 of its adjacent squares. You would then have to go elsewhere to test for
non-mine squares. The player is declared the winner when either all mine’s are flagged
or you cleared all the squares that does not contain mines.
The game can be played for two basic purposes; play for speed or play to just win. To
play for speed simply try to clear the minefield as fast as you can without detonating a
mine and to play to win; simply clear the minefield. Each contradicts each other as
focusing for speed decreases your chances of winning and focusing on just winning
decreases your speed. D Some reflection
but very simplistic
Controls
For a player to clear a mine simply left click a square.
Players can then ‘’flag’’ a square by right clicking on a square they believe to inhabit a
mine thus eliminating it.
Right-clicking again on the flag symbol of a square will change the flag into a question-
mark graphic to indicate the possibility of the square may or may not be a mine.
Right clicking once more on a question mark graphic transforms the square back to its
original form.
The player can press down both left and right clicks at the same time on a square to A Lack of
show the square’s adjacent squares. coherency
through limited
explanation
Basic Strategy
The key of minesweeper is logical deduction from the digits of the squares. Simple
logical strategies cases such as if there is a 3 in a corner, all the squares adjacent to is
100% mines. Another example would be if an uncovered square is with a digit 1, and
there is only one covered square touching it, that square must be a mine. In general,
clearing a board with speed requires a form of pattern recognition and the brain to B, E No mathematical
quickly identify as well as make connections in identifying the clues which state that strategy outlined.
there’s a 100% probability in a square being a mine. Missed opportunity
for presentation using
mathematical language
Element of Luck
The first few moves of minesweeper without a doubt require luck. The player can only
give wild guesses because of the fact the player have either no digits or anything to refer

to. The chance of a player clicking on a mine on the first move is 0% as the mines
generate randomly after your first click. On your second click however, on a beginner’s
A No explanation
difficulty would be around 12% and 38% on expert difficulty. Because the placement of
the mines in every game is random, the choice of your starting move can’t be based on of why these
anything but luck. Even if you start in the corner, you may either be lucky enough to get values were used
a 3 or be unlucky and get a 1 which really doesn’t help a whole lot for your first move.
Experts do however recommend beginning the middle area of the grid opposed to the
corners.
It is only after the first few moves of minesweeper is complete when the player can truly
deduct and rely on logic of where the mines are for the rest of the game.
Minesweeper also have the factor of sheer luck in cases where the player has no choice
but to take a guess in clearing a square as the digits of the squares show all the
surrounding squares each having the same probability of possessing a mine. C Opportunity for
an example here
There are times when a grid generated can only be solved with the element of guessing.
Refer to Figure 3.
Figure 3
The player must guess whether or is the mine. The probability of A and B being
B Incorrect
mines are equal, 50/50.
terminology used
Another scenario that requires guessing is when an un-clicked square is completely or
partially surrounded by mines. The player has no numbers touching the un-clicked
square thus the player has again no information about the probability of whether or not
the square will be a mine.
C, D An
Strategies when facing this scenario that will allow the player to avoid simple guesses appropriate
are to play the rest of the game and ignore the square. In the game when the number of logical
un-clicked and un-flagged squares equals the number of mines left unmarked then all implication
the remaining spaces are mines. Some versions of the game automatically flag these
spaces once all the other squares in the game have been clicked as well as reveal them
once all the flags have been placed.

Single Square Strategies
If the number of un-clicked squares adjacent to the number of square equal to the
number on that square, all the un-clicked squares are definitely all mines.
For any square with a digit, if the digit of mines you've found adjacent to that square is
equal to the number of the square, all other squares adjacent to that numbered square
must be safe. If you know the square to the right of a 1 is a mine, then you can deduce
that all the other squares next to that 1 do not contain mines and you can clear them to
obtain more digits to progress in the game. B, E Student
work is descriptive
rather than
mathematical.
Example case Example case
1 2
C Use of
a and b are personal examples
safe to open,
as the 3 is
already a and b must
satisfied with be mines; the
3 adjacent only squares
mines. that can
provide those
demanded by
the 3 are a
and b.
Multiple Squares Strategies A Lack of coherency
Multiple square strategies are a great for players who want to win in beginner
difficulties quickly, and a must in the higher difficulties. For these complex puzzles, the
player would need to consider more than one square a time
If you have two adjacent numbers, the difference between those numbers is equal to
the difference in the number of mines for the 3 squares adjacent to each that are not
adjacent to the other number. For example: if these numbers differ by 3, all of the

adjacent squares to the higher number not shared by the other are mines, and all the
opposite ones are safe. C Opportunity for
an example
Sometimes it can be known that there are a certain number of mines in a certain
number of squares (without necessarily knowing which the mines are and which are
safe) and you can often utilize this information to find out information about other
squares.
A method used by Minesweeper Computers of artificial intelligence is to label the game

as Constraint Satisfaction Problem; ‘’mathematical problems defined as a set of objects
whose state must satisfy a number of constraints or limitations’’
http://en.wikipedia.org/wiki/Constraint_satisfaction_problem
One would recognize the variables in the game are the unopened squares, and the
constraints are the adjacent squares that are opened.
The algorithm consists of trying every combination of mines that satisfies all the
numbers in the adjacent squares, and making a conclusion from there.
In expert difficulty, this would be a time-consuming process for even the most powerful
computer, but minesweeper players who are experienced at the expert level would be
able to quickly see which squares need this procedure, and where one might expect it to
succeed.
Analysis
According to Robert Kaye, Minesweeper is a NP-Complete Problem.
A problem is said to be NP-Complete if all other problems in NP can be solved easily

using a solution to this problem. The main method for proving that a problem is NP-
Complete is reduction. By converting one problem into another one like finding a
solution to the latter problem will let you solve the former, in other words ‘’breaking it
down’’ or using ‘’logical sense.’’. If one problem we call A is reducible to problem B, then
solving B is just as difficult as solving for A as the solution for B gives the solution to B. If
we reduce an NP-Complete problem to a new problem then we know that the new
problem is NP-Complete as well.
C Unfamiliar mathematics, but not
Possible Applications explained well enough to justify
understanding and not used
A computer powerful enough to win a minesweeper game would earn the title of one of
the most powerful computers and best achievements of man-kind.

C General and dramatic

statement without any
support or evidence
The result of the powerful artificial intelligence is a solution that could enhance security
measures on the web and data files in general.
E Lack of
It's surprising that such a simple game would put us at such a frontier of mathematics. mathematics
But the big questions in math are not very far below the surface of everyday life. actually presented
One of the outstanding problems in logic and computer science is determining whether A Lack of
questions exist whose answer can be quickly checked by an artificial intelligence, coherency
however requires a much longer time to solve without knowing the answer. So far no
one has proved that any of them really does require a long time to solve.
Again, referring to Robert Kaye, Minesweeper is a NP-complete problem. The definition

of a NP-complete problem is that all other NP problems can be reduced to using
polynomial-time many one deduction. The problem attempts to determine whether
questions that seem to be unsolvable within a reasonable period of time might actually
have a relatively simple way of being solved, possibly by computer. Thusly algorithm
that can solve minesweeper is speculated to be able to solve any and all NP problems.
D Limited
‘’If there was a way of playing Minesweeper efficiently, then there would also be a way reflection
of cracking codes efficiently.’’ Kaye said.
If someone were to come up with an algorithm for successfully beating Minesweeper,

that algorithm would also be instrumental in vastly improving encryption and code-
cracking.
Conclusion
Through development in ‘’Minesweeping strategies’’, as represented by Mathematicians

Patti Frazer Lock of St. Lawrence University in Canton, N.Y., and Allan A. Struthers of
Michigan Technological University in Houghton, the game can be used to introduce
students to formal mathematical proofs.
In the school’s courses, students play a few games and attempts to deduct and evaluate
the various positions the squares that definitely possess mines and the squares
definitely are safe. The exercise is proven to give students a sense of given the rules, the
A Implicit
evidence and conjecture, then prove or test. Students who become ‘’masters’’ of this
game learn truly useful reasoning, logic and deductive techniques such as proof by aim here but
contradiction or the role of counter, all which can be applied to high levels of incorrectly placed
mathematics such as proving theorems.
"Putting a flag on a square is a theorem--you know there's a bomb there," Struthers

explains.

It is indeed interesting how gaming can contribute “practical” things to people's lives
and how this simple game such as minesweeper might just be the thing that can a solve D Superficial
today’s mathematical mystery. It truly is no surprise, with the amount of logical reflection
deductions, time and brainpower required to play this game.
Bibliography
Loy, Jim. "Minesweeper Strategy." Minesweeper Strategy. 1996. Web. 6 Jan. 2010.
<http://www.jimloy.com/games/mine.htm>.
Peterson, Ivars. "Minesweeper Logic."

Http://www.maa.org/mathland/mathtrek_5_3_99.html. Ivars Pertereson's Mathtrek, 3
May 1999. Web. 6 Jan. 2010. <http://www.maa.org/mathland/mathtrek_5_3_99.html>.
Geek Squad Team. "Minesweeper + Math = 1 million." Geek.com. GeekSquad, 3 Nov.

2000. Web. 6 Jan. 2010. <http://www.geek.com/articles/games/minesweeper-math-1-
million-2000113/>.
Lipman, Lisa. "Games may help solve math mysteries." MSNBC. Associated Press, 2000.
Web. 6 Jan. 2010.
<http://www.math.hawaii.edu/lab/Game%20may%20help%20solve%20math%20myste
ries.htm>.
Wikipedia. "Minesweeper (Computer Game)." Wikipedia. The Free Encyclopedia. Web. 6

Jan. 2010. <http://en.wikipedia.org/wiki/Minesweeper_%28computer_game%29>.

Modelling musical chords using sine waves
Introduction A Rationale
included in
From the stimulus word “Harmony”, I chose to look at the transmission of sound introduction
waves in music. As a keen musician myself, I was curious to understand more about how
electronic instruments emit sound. Sound travels as a transverse wave, so takes the same
basic shape as a mathematical sine wave. Electronic instruments use these waves to
produce sounds of different frequencies, volume and timbre.
Oscillators 1 are used in most electronic instruments to produce sound waves. These
sound waves are sine waves, with different frequencies, changing the pitch of the note,
and different amplitudes, varying the volume of the note. The timbre of the sound is
changed once the wave has left the oscillator. It travels through various filters which
change the type of sound. This allows for a great range of sounds, that can imitate C Personal
different instruments. It is this technology that is used in electronic keyboards. interest
in subject
I decided to focus on the changing of frequencies to produce different notes. By area, but it
modelling chords, and series of notes using sine waves, I can demonstrate how the sound is limited
waves travel in the air, and how the different frequency waves relate to each other in
different chords. I will look specifically at the relationships within each type of chord,
and whether there is any difference between chords that are perceived to be in harmony A Aim not really
and those that are perceived as discordant. clear compared
with what is done
Investigation A Undefined
term and not clear
I started by working out how to model a single note. I decided to use “Middle” A, as for non-musicians
it has an exact frequency of 440 Hz 2, and is the note that Western orchestras tune to. I
wanted to make the sound wave of this note equal to the sine wave, which meant making
220Hz equivalent toππ. This produces the following graph:
Typing error
4 y
−2π −π π 2π
−2
1
http://en.wikipedia.org/wiki/Synthesizer
2
http://www.phy.mtu.edu/~suits/notefreqs.html

A Lack of clarity
1. The Major Chord
for non-musicians
Using an A major chord, I can look at the relationship between the notes in a major
chord. The following graph demonstrates how this chord travels through air:
y
−π −π/2 π/2 π
−1
−2
A No mathematical explanation
of how these are determined
Equation 1: y=sinx
Equation 2: y=sin(554.37/440x)
B Erroneous notation. This means
Equation 4: y=sin2x 554.37 but should be 554.37x .
440x 440
I altered the wavelengths, or frequencies, of the graphs by changing the coefficient of
x. In order to get the most accurate graph, I used the original frequencies of the other
notes 3. The table below shows more information about the notes in this chord.
Note Semitones away Frequency (Hz) Approximate B No explanation

from A frequency ratio of how these values
(Hz/440) are determined
A 0 440.00 1
C# 4 554.37 1.25 B Graphs
E 7 659.26 1.50 not related
to tables
A 12 880.00 2
The frequency ratios in this chord appear to be neatly spaced, with the third gap in
the chord equalling the sum of the first 2 gaps in the chord. The ratio of the gaps is 1:1:2. C Student
This may indicate the start of a pattern. Similarly, it could be that the third gap is double created own
the first or second gap. I will now look at different chords to see if this conjecture holds way of looking
for patterns in
3
the chord.
http://www.phy.mtu.edu/~suits/notefreqs.html

true. The gaps in the ratios do not directly relate to the difference in semitones, as the
semitones do not have equal frequencies-the frequencies of semitones diverge as the
frequencies increase, and the notes get higher in pitch.
2. The Minor Chord
Minor chords are also considered to be in harmony. Again using an A chord, I have
modelled a minor chord and looked at its features. These waves look very similar to the
major chord waves, as only one note has changed; the C# has gone down by a semitone.
Therefore the graphs are very similar. The graph for the minor chord looks like this:
y
−π/2 π/2 π 3π/2
−1
−2
Equation 1: y=sinx E Using trigonometric graphs,

Equation 2: y=sin(523.25/440x) but nothing is done with
Equation 3: y=sin(659.26/440x) them. Limited understanding
Equation 4: y=sin2x of notes played together
This graph looks very similar to the major graph, but the actual values, shown in the
following table, are slightly different.
Note Semitones away Frequency (Hz) Approximate

from A frequency ratio
(Hz/440)
A 0 440.00 1
C 3 523.25 1.20
E 7 659.26 1.50
A 12 880.00 2
Again, because the E in the middle of the chord has not moved, the sum of the first 2
gaps in the chord is equal to the third gap. The ratio of the gaps is now 2:3:5. However,
to test this theory properly, we need to look at chords that differ from the original major
chord more dramatically.

3. The First inversion major chord
An inversion chord is the same as a major chord, but the notes are in a slightly
different order. A first inversion chord starting on A will actually be in the key of F
major. The graph of such a chord looks like this:
2 y
−π/2 π/2 π 3π/2 2π
−1
−2
Equation 1: y=sinx
Equation 2: y=sin(523.25/440x) Incorrect value
Equation 3: y=sin(698.46/440x) (698.46) compared
Equation 4: y=sin2x with table

(Hz/440)
A 0 440.00 1
C 3 523.25 1.20
F 8 659.26 1.60
A 12 880.00 2
The table of values for this chord shows that my original conjecture does not hold
true for all chords. However, the gaps are still in a neat ratio of 1:2:2.
4. The Second inversion major chord
This chord, like the first inversion, is another different rearrangement of the notes in
a major chord.

−π/2 π/2 π 3π/2
−1
−2
Equation 1: y=sinx A Repetition of

Equation 2: y=sin(587.33/440x) the same work
Equation 4: y=sin2x

(Hz/440)
A 0 440.00 1
D 5 587.33 1.33
F# 9 739.99 1.67
A 12 880.00 2
Interestingly, in this chord, the frequencies of the notes are evenly spaced, as the ratio
of the gaps is 1:1:1.
I will now look at some discordant notes, and see how they compare with the chords
I have already looked at.
5. The Augmented 4th
This chord is also known as “The Devil’s Chord”, as it is considered to be the most
unpleasant sounding chord in music. As it is only 2 notes, it cannot be modelled in
exactly the same way as the other chords, but can still be graphed:

2 y
−π/2 π/2 π 3π/2 2π
−1
−2
Equation 1: y=sinx

(Hz/440)
A 0 440.00 1
D# 6 622.25 1.41
B Poor use
of terminology
The approximate ratio for this chord shows that the ratio is not as precise as with the
other chords. However, there is not necessarily a difference in chords that are in
harmony and discordant chords. This chord is not directly comparable to the other
chords, as it does not have 3 notes like the others.
6. The Augmented 6th D Some reflection
I will now look at the augmented 6th, another discordant chord. This chord however,
has 3 notes, so can be more easily compared to the other chords I have looked at. An
Augmented 6th is also known as a dominant 7th, which was one of the first discordant
chords to be used in music in the West.

2 y
−π/2 π/2 π 3π/2 2π
−1
−2
Equation 1: y=sinx
Equation 4: y=sin2x

(Hz/440)
A 0 440.00 1
C 3 523.25 1.2
F# 9 739.99 1.67
A 12 880.00 2
In this chord, the gaps in the ratios are not easily relatable to each other, suggesting
that discordant sets of notes are not as natural as those that sound harmonic.
“Perfect Intervals” D Reflection on results
The intervals known as a “perfect fourth” and “perfect fifth” are, as the name
suggests, considered to be the most harmonic intervals. Interestingly, the graphs of these
intervals show that the notes in the perfect fourth cross at exactly 3π, and the notes in the
perfect fifth cross at 2π.

The graph of a perfect fourth:
y
1
0.5
x
0
3π/2 2π 5π/2
−0.5
Equation 1: y=sinx
The graph of a perfect fifth:
0.5
x
0
9π/4 11π/4 13π/4
−0.5
−1
A, D Lack of explanation
Equation 1: y=sinx
and a lost opportunity for
reflection on how results
fit with conjecture
Conclusion
This evidence would suggest that notes which are traditionally considered to be “in
harmony” are mathematically “neater” than those that are not in harmony. The
mathematical nature of the notes in traditional harmony suggests that the sound waves are
more pleasing to our brain because of the way the notes fit together. The ratios of
frequencies of chords in harmony are more concise than the discordant notes. The sound
B More poor use
of terminology

waves produced by oscillators are in the same form as notes produced by mechanical
instruments, so the same note ultimately sounds the same, or very similar, to our ear. For
this reason, this theory does not only apply to sound coming to electronic instruments, but
every kind of instrument.
Evaluation
It is clear then, that there is a difference in the mathematical relationship of notes

within chords in harmony and discordant chords. This means, that perhaps in the future
music can be created entirely mathematically. It may be that maths can help determine
which chords, and sequences of chords, will work together and be most pleasing to the
ear.
This investigation was very limited, as I was not able to research properly how
electronic instruments work, and how they differ from mechanical instruments. I also
would have liked to tried to find a mathematical formula connecting the number of
semitones to the frequency ratios. The evidence here suggests that notes in harmony fit
into a mathematical context, whereas discordant notes do not. However, this is just a
theory, and would have to be tested more thoroughly. To extend this investigation, I
would like to look at the area between the waves in relation to the semitone difference
and frequency ratios.
D Superficial reflection

Bibliography
Suits, B.H., 2010. Frequencies of musical notes [Website] Available at:

http://www.phy.mtu.edu/~suits/notefreqs.html [Accessed: 23rd June 2010]
Wikipedia Synthesiser. Anon [Website] Available at:

http://en.wikipedia.org/wiki/Synthesizer [Accessed: 23rd June 2010]

Maths Exploration
Newton-Raphson method
A Introduction
includes aim
and rationale
Rationale- For this project I chose to research and analyse the Newton-Raphson method,
where calculus is used to approximate roots. I chose this topic because it looked extremely
interesting and the idea of using calculus to approximate roots, seemed intriguing.
The aim of this exploration is to find out how to use the Newton-Raphson method, and in
what situations this method is used
Explanation of the Newton-Raphson method
The Newton-Raphson or Newton’s method is an iterative process to approximate roots.

We know simple roots for rational numbers such as 4 or 9 , but what about irrational
numbers such as 3 or 5 . This method was discovered in 1736 by Isaac Newton after
being published in the ‘Method of Fluxions’, this method was also described by Joseph
Raphson in 1690 in ‘Analysis Aequationum’.
The Newton-Raphson Process:
In the Newton-Raphson process the following formula is used:

A, B Lack of
f ( xn )
xn +=
1 xn − , f '( xn ) ≠ 0 mathematical explanation
f '( xn )
of this terminology
This is called the Newton-Raphson formula
How this formula was derived: B Student needs to be careful

how the term "root" is used. Here
The square root of a number (n) can be found by using the function:
it means the square root of "n", but
the solution of f(x)=0 is also a root.
) x2 − n
f ( x=
B No explanation of how an
The root of n is the value of x when f(x) = 0 initial estimate is made
The first thing you do when approximating a root is to make an initial estimate in terms of
a positive integer and find the tangent of the function at that point.
B Poor use of subscripts
Let the initial estimate = x₁
throughout exploration
The intercept of the tangent at x₁ with the x-axis will be closer to the root than the initial
estimate. This intercept can be called x₂. E No
explanation of
Calculating x₂ is done by studying the triangle bounded by the x-axis, the line x=x₁, and how student
the tangent to the line at x₁. This only works when the gradient of the function at x₁ is not knows it will
be closer
equal to zero.
An example of a triangle is shown in the following graph (where the square root of 7 is
being estimated):
B Graph could
be larger and
clearer, zooming in
on the pertinent
domain and range
In this case x₁=3

This results in the following formula:

A Lack of
y1 explanation of
x2= x1 −
gradient at x1 where this
comes from
Therefore
f ( x1 ) E Method is applied after this,

x2= x1 −
f '( x1 ) but no evidence of understanding
why the method works
Now that we have a better estimate, the same method can be used again to get closer to
the answer. This time x₃ will be found from x₂ represented by the formula below:
f ( x2 )
x=
3 x2 −
f '( x2 )
x₃ will be even closer to the root than x₂ was, and this procedure can be repeated an
infinite number of times.
This is an explanation to the Newton-Raphson formula:
f ( xn )
xn +=
1 xn − A Not an
f '( xn )
explanation
Where n= the number of iterations
The more iterations that are carried out the more accurate the answer is. In practice the
iteration is usually only repeated to x₁₀ or less.
For example, if we use this method to approximate the value of root 7 , if the equation
y x 2 − 7 is used.
=
C Good demonstration
The formula for this function is: of applying unfamiliar
mathematics
f ( xn )
xn +=
1 xn −
f '( xn )
xn 2 − 7 E Differentiation
xn +=
1 xn −
2 xn within syllabus

Using the Newton-Raphson formula for this function the following table showing the
value of x for different numbers of iterations:
x₁ 3.0000000000
x₂ 2.6666666667
x₃ 2.6458333333
x₄ 2.6457513123 A Lack of detailed
x₅ 2.6457513111 explanation
x₆ 2.6457513111
x₇ 2.6457513111
x₈ 2.6457513111
x₉ 2.6457513111
x₁₀ 2.6457513111
The table shows that this method will only need a few iterations and you already have the
value of x to 3 decimal places and after the fifth iteration you have the values of x up to 10
decimal places.
Graph showing Newton-Raphson method for =

y x 2 − 7 , when approximating 7
A Graph
does not aid
coherence and
adds nothing to
previous graph.

I will now do the same for y x2 − 3

3 , and I’ll use the equation =
The formula for this function is:
f ( xn )
xn +=
1 xn −
f '( xn )
xn 2 − 3
xn +=
1 xn −
2 xn
Using the Newton-Raphson formula for this function the following table showing the
value of x for different numbers of iterations:
x₁ 2.0000000000
x₂ 1.7500000000
x₃ 1.7321428571
x₄ 1.7320508100
x₅ 1.7320508076
x₆ 1.7320508076
x₇ 1.7320508076
x₈ 1.7320508076
x₉ 1.7320508076
x₁₀ 1.7320508076
The table again shows that this method will only need a few iterations and you already
have the value of x to 3 decimal places and after the fifth iteration you have the values of x
up to 10 decimal places.
.Graph showing Newton-Raphson method for , when approximating

Using a spreadsheet, I repeated the same method for all roots x where x ∈ Z + in the
domain 1 ≤ x ≤ 7 , and a table of iterations was produced: B Good use
of notation
1 2 3 4 5 6 7
x₁ 1.0000000000 1.5000000000 2.0000000000 2.0000000000 2.5000000000 2.5000000000 3.0000000000
x₂ 1.0000000000 1.4166666667 1.7500000000 2.0000000000 2.2500000000 2.4500000000 2.6666666667
x₃ 1.0000000000 1.4142156863 1.7321428571 2.0000000000 2.2361111111 2.4494897959 2.6458333333
x₄ 1.0000000000 1.4142135624 1.7320508100 2.0000000000 2.2360679779 2.4494897428 2.6457513123
x₅ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
x₆ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
x₇ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
x₈ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
x₉ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
x₁₀ 1.0000000000 1.4142135624 1.7320508076 2.0000000000 2.2360679775 2.4494897428 2.6457513111
The table shows that after the first few iterations you get a very precise value of x
Although the main function of this method is to approximate roots it can also be used
for non-linear equations. C, E Student could have mentioned
other roots, such as cube roots.
For example the Newton-Raphson method can be used to find when tan(x) = x or cos(x)= x
Where f(x) = x – tan x or f(x) = x – cos x
The Newton Raphson formula can easily be adjusted for these equations:
f ( xn )
xn +=
1 xn − C Student created
f '( xn ) own example, but did
not follow through to
When f ( x)= x − tan x complete application
with graphical analysis
to see how the process
worked for this case.

Becomes:
xn − tan( xn )
xn +=
1 xn −
1 − sec 2 ( xn )
D Opportunity for
And when f ( x)= x − cos x reflection on any
special implications to
non-linear functions
Becomes:
xn − cos ( xn )
xn +=
1 xn −
1 + sin( xn )
Reflection:
Evaluation and conclusion
This was a very interesting exploration, as at the start I had no idea what the formula meant
and the logic behind it. The idea of using the x-intercept of the gradient to get closer and
closer to the correct value of x was amazing. Up to now, I always had a calculator to find
B Confused
the root or zero of a number. If I didn’t I would just leave it as the 3 or 5 . Although,use of "root"
this method may not be practical to use on a maths test, it has quite a few advantages. This and "zero"
method gives us the chance to approximate roots up to thousands of decimal places. Even
after the first five iterations the value of x is given to 10 decimal places. When finding the
root of x, it is quite easy to work out x₂ in your head, which is very close to the precise
value. Furthermore, this formula is implemented in technology such as autograph, where
Newton-Raphson Iteration is very easy to use and find, as well as by using spreadsheets
where it is simple to implement the formula for any number of iterations. Nevertheless,
it’s easy to find the zero or root by finding the x-intercept, which is much easier than using
the Newton-Raphson method on the calculator or any graphing program. This method
would be useful though in any field involving zeros or roots where a high-level of
precision is needed.
D Superficial
Bibliography: reflection with
no depth
Internet Sources:
"Newton-Raphson method." knowledgerush, n.d. Web. 20 Jun 2010.

<http://www.knowledgerush.com/kr/encyclopedia/Newton-Raphson_method/>.

"Newton-Raphson Method." N.p., n.d. Web. 20 Jun 2010.

<http://www.shodor.org/unchem/math/newton/>.
"The Newton-Raphson Method ." S.O.S. Math. N.p., n.d. Web. 20 Jun 2010.
<http://www.sosmath.com/calculus/diff/der07/der07.html>.
Image:
Web. 20 Jun 2010. <http://gilkalai.files.wordpress.com/2010/01/fa666.png>.

The Polar Area Diagrams of Florence Nightingale A Rationale

If you read the article on Florence Nightingale in “The Children’s Book of Famous and aim
Lives”1 you will not learn that she had to battle with her parents to be allowed to study included in
Mathematics. If you read the Ladybird book “Florence Nightingale”2 you will not introduction
discover that she was the first woman to be elected as a Fellow of the Royal Statistical
Society. In looking around for an area of research I was intrigued to discover that
Florence Nightingale, who I always thought of as the “lady with the lamp”, was a
competent Mathematician who created her own type of statistical diagram which she used
to save thousands of soldiers from needless death.
Florence Nightingale headed a group of 38 nurses who went to clean up the hospitals for
the British soldiers in the Crimea in 1854. She found that most of the deaths were due to
diseases which could be prevented by basic hygiene, such as typhus and cholera. Her
improvements were simple but they had an enormous effect:
“She and her nurses washed and bathed the soldiers, laundered their linens, gave
them clean beds to lie in, and fed them”3.
When she returned to Britain she made a detailed report to the Government setting out
what conditions were like and what needed to be done to reduce deaths in the hospitals.
Nothing was done, so she tried again, making another statistical report and included in it
three new statistical diagrams to make data collated by William Farr more accessible to
people who could not get their minds around tables of figures. These were her polar area
diagrams or rose diagrams, sometimes also known as ‘coxcombs’. The first showed how
many men had died over the two years 1854-5, the second showed what proportions of
men had died from wounds in battle, from disease and from other causes, the third
showed how the number of deaths had decreased once “sanitary improvements”4 had
been introduced.
I decided I would try to recreate the second of these diagrams which is the most
complicated and the most shocking. It is called “Diagram of the causes of mortality in the
army in the east”. A copy of it is below:

Figure 1
The basic ideas are very simple. The blue area represents deaths due to disease, the red
area represents death due to wounds in battle and the black area represents death due to
other causes.
I tried to find a copy of the data which this diagram represented, but I had no luck, so I
C Student
decided to make sure I understood exactly how the diagram was made and to make my
own version of some data which I did have to hand. Once I tried to understand the
demonstrates
diagram in detail I found there were some problems. personal interest
in topic.
The First Problem
I wasn’t sure whether the black area in a shape such as this:

was supposed to be
D Good critical
reflection on
problems related
to objective
this area or this area

In other words, were the colours separate, or overlapping? The articles I read didn’t make
it clear. O’Connor says that “The area of each coloured wedge, measured from the centre
A Good research
as a common point, is in proportion to the statistic it represents”5, which makes it seem
demonstrated
that all colours are wedged shaped, or sectors, so the colours overlap. However, Lienhard
commented that in the November 1854 section “battle deaths take up a very small portion
of each slice”6, which makes it sound as though the slice has three separate portions, and
Brasseur says that “she also divided the areas within each of the wedges to show which
portion of the mortality data for that month could be allotted to each cause of death”4. I
decided to construct polar area diagrams for a set of data with the colours separate and
with the colours overlapping to see if putting theory into practice would make it clearer to
C Created
me. The data I used was taken from the IB grade distribution statistics for the past 15
personal example
years at my own institution. I used the numbers taking Higher Mathematics, Standard
Mathematics and Mathematical Studies to be represented by my three colours. I took the
old Mathematical Methods course to be the same as Standard Mathematics.
2π
To fit 15 sectors into the circle I needed each arc to subtend an angle of radians at
15 E Area of sector
1 2π 2 π 2 using radians is
the centre. The area of each sector would then be A = × r =r where r is the
2 15 15 in syllabus and
radius of the sector. Since the area needs to be proportional to the statistic, I needed to is relevant to
15A exploration.
find the radius, so I used r = and just used a scale which would allow me to draw
π
a reasonable sized diagram. To create a polar area diagram with overlapping sectors I just
used this formula on each of the numbers of students taking the various options.
Numbers taking Mathematics year on year

Numbers (A) Radius ( r )
Higher Studies Standard Higher Studies Standard
1995 1 24 0 2.2 10.7 0.0
1996 4 15 0 4.4 8.5 0.0
1997 8 10 0 6.2 6.9 0.0
1998 6 31 0 5.4 12.2 0.0
1999 9 17 0 6.6 9.0 0.0
2000 10 20 0 6.9 9.8 0.0
2001 4 31 1 4.4 12.2 2.2
2002 5 21 2 4.9 10.0 3.1
2003 4 15 4 4.4 8.5 4.4
2004 5 29 5 4.9 11.8 4.9
2005 1 28 0 2.2 11.6 0.0
2006 3 16 2 3.8 8.7 3.1
2007 8 13 0 6.2 7.9 0.0
2008 11 29 14 7.2 11.8 8.2
2009 10 23 15 6.9 10.5 8.5

I then used a geometric program (GeoGebra) to draw the sectors all with a common
2π
centre, each with an angle of radians, and with the radii as given in the table. I drew
15
the Higher sectors first with the Studies on top of these, and the Standard on top of these.
This was the result:
Figure 2
Polar area diagram to show students
taking Mathematics at one school
(colours overlapping)
Blue represents the number of students taking
Higher Maths.
Brown represents the number of students
taking Mathematical Studies.
Green represents the number of students
taking Standard Maths.
The colours are not solid, so where colours overlap there is a different colour. The blue
overlapping the brown makes a pink here, and the green overlapping the blue makes a
darker green. In 2003 and in 2004 there were an equal number of students taking Higher
and Standard so three separate colours cannot be seen on the diagram.
Next I worked out the radii needed if the colours were not to overlap. For this I used
15 A1 15 ( A1 + A2 )
cumulative areas to work out the radii. R1 = , R2 = and
π π B Good use
15 ( A1 + A2 + A3) of notation
R3 = .
π
Numbers taking Mathematics year on year
Numbers (A) Radii
Higher (A1) Studies (A2) Standard (A3) R1 R2 R3
1995 1 24 0 2.2 10.9 10.9
1996 4 15 0 4.4 9.5 9.5
1997 8 10 0 6.2 9.3 9.3
1998 6 31 0 5.4 13.3 13.3
1999 9 17 0 6.6 11.1 11.1
2000 10 20 0 6.9 12.0 12.0
2001 4 31 1 4.4 12.9 13.1
2002 5 21 2 4.9 11.1 11.6
2003 4 15 4 4.4 9.5 10.5
2004 5 29 5 4.9 12.7 13.6
2005 1 28 0 2.2 11.8 11.8
2006 3 16 2 3.8 9.5 10.0
2007 8 13 0 6.2 10.0 10.0
2008 11 29 14 7.2 13.8 16.1
2009 10 23 15 6.9 12.6 15.1

This gave a diagram with Higher numbers at the centre and Standard numbers at the
edge, like this:
Figure 3
(colours separate)
Higher Maths.
This diagram is incomplete in that it has not got the dates on it, but I was interested in the
basic shape it would make rather than seeing it as a finished article to represent the data.
I decided to do the same thing but with Studies in the middle and Higher at the edge to
see how different it would look.
Figure 4
(colours separate)
Higher Maths.
C Good
demonstration of
personal interest
This feels very different. The blue section actually looks less significant, to my eye, being
put at the edges. This made me think of something else I had read in Brasseur’s article, D Critical
“Nightingale arranged these colored areas so that the main cause of death (and the largest reflection on
sections)—deaths by disease—would be at the end of the wedges and would be more personal work
easily noticed.”4 I am sure that Brasseur thought that the colours were separate, and not
overlapped. However, comparing my diagrams to Nightingale’s original in Figure 1, I

became sure that she did mean them to be overlapped. I noticed that in the lefthand rose
in figure 1 (representing the second year) there is a wedge with blue at the edge followed
by a wedge with blue at the edge: "Blue" is a
typing error.
Figure 5
A zoom in of part of figure 1
This can happen in a diagram like my figure 2 of overlapping colours, but would be
impossible if the colours are separate as in figures 3 and 4. From this I deduced that the
colours on the diagram must be overlapping.
D Very good
The Second Problem example of
critical reflection
My diagrams were unlike Nightingale’s ones in that the total area of the sectors in figure
2 represented the total number of students taking the IB at this school over the 15 years.
Nightingale’s statistics were rates of mortality. Basically they can be thought of as
percentages of soldiers who died, but, as before, when I read through the articles again, I
was unsure what they were percentages of. Gill and Gill have table (Table 2) in their
article with headings “No. of soldiers admitted to the hospital” and “No. (%) of soldiers
who died”3. This might suggest that Nightingale was working with percentages of
soldiers who were admitted into hospital. Lewi is more definite and refers to the actual
statistic of one wedge of the third of Nightingale’s polar area diagrams as follows: “The
mortality during the first period was 192 per 1,000 hospitalized soldiers (on a yearly
basis)”9. However, Brasseur refers to the statistic in a wedge of Nightingale’s first
diagram as being “the ratio of mortality for every 1,000 soldiers per annum in the field”4,
in other words a percentage of the army actually on duty.
I decided to create a polar area diagram to act as an analogy to the possible situations as
follows:
Nightingale’s data My data
Number of soldiers in the army in a month Number of students taking the IB in a year
Number of soldiers taken to hospital Number of students taking Maths Studies
Number of soldiers dying of wounds Number of students gaining grade 7
Number of soldiers dying of disease Number of students gaining grade 6
C Another good
Number of soldiers dying for other reasons Number of students gaining grade 5
personal example
My analogy of drawing a diagram showing the numbers of soldiers dying as a percentage
of those admitted to hospital would then be the number of students gaining a grade above
4 as a percentage of those taking Mathematical Studies. I decided to do this one by hand,
partly to prove I could, and partly to see if it would throw any extra light on the
construction of the diagrams.

I gathered the data, found the percentages and used the percentages as A in the usual
15A
formula r = to find the radii needed to construct the diagram. The data is here:
π
Numbers gaining top three grades in Mathematical Studies
As percentage of those taking
Studies Radius required for each
Grade 7 Grade 6 Grade 5 Taking Total
Studies in year % grade 7 % grade 6 % grade 5 R7 R6 R5
1995 7 10 4 24 25 29.16667 41.66667 16.66667 11.80 14.10 8.92
1996 2 9 3 15 19 13.33333 60.00000 20.00000 7.98 16.93 9.77
1997 1 4 2 10 18 10.00000 40.00000 20.00000 6.91 13.82 9.77
1998 5 12 11 31 37 16.12903 38.70968 35.48387 8.78 13.60 13.02
1999 2 6 7 17 26 11.76471 35.29412 41.17647 7.49 12.98 14.02
2000 3 4 7 20 30 15.00000 20.00000 35.00000 8.46 9.77 12.93
2001 3 8 8 31 36 9.67742 25.80645 25.80645 6.80 11.10 11.10
2002 1 8 4 21 28 4.76190 38.09524 19.04762 4.77 13.49 9.54
2003 0 1 8 15 23 0.00000 6.66667 53.33333 0.00 5.64 15.96
2004 3 9 7 29 34 10.34483 31.03448 24.13793 7.03 12.17 10.74
2005 1 11 9 28 29 3.57143 39.28571 32.14286 4.13 13.70 12.39
2006 2 4 5 16 21 12.50000 25.00000 31.25000 7.73 10.93 12.22
2007 1 8 3 13 22 7.69231 61.53846 23.07692 6.06 17.14 10.50
2008 0 3 17 29 54 0.00000 10.34483 58.62069 0.00 7.03 16.73
2009 0 5 5 23 48 0.00000 21.73913 21.73913 0.00 10.19 10.19
And the diagram came out like this:
Figure 6
Polar area diagram to show
percentages of students taking
Mathematical Studies who gained
grades above 4
Red represents the number of students
gaining grade 7.
Blue represents the number of students
gaining grade 6.
gaining grade 5.
The purple areas represent coinciding B Student
numbers of students gaining grade 5 and 6. needs to watch
consistency of
location of colours
used here.

One thing which I learnt from this exercise is that you have to be very careful about your
scale and think through every move before you start if you don’t want to fall off the edge
of the paper! It is a far more tense experience drawing a diagram by hand because you
know that one slip will make the whole diagram flawed. A computer slip can be corrected
before you print out the result. My admiration for Florence Nightingale’s draftsmanship C This section
was heightened by doing this. shows a lot of
The other thing which drawing by hand brought out was that, if you draw the arcs in in personal interest.
the appropriate colours, the colouring of the sectors sorts itself out. You colour from the
arc inwards until you come to another arc or the centre. The only problem came when
two arcs of different colours came in exactly the same place. I got around this problem by
colouring these areas in a totally different colour and saying so at the side.
At this point in my research someone suggested some more possible websites to me, and
following these up I found a copy of Nightingale’s second diagram which was clear
enough for me to read her notes, and a copy of the original data she used. The first of
these was in a letter by Henry Woodbury suggesting that Nightingale got her calculations
wrong and the radii represented the statistics rather than the area.7 The letter had a
comment posted by Ian Short which led me to an article by him8 giving the data for the
second diagram and explaining how it was created.
The very clear reproduction of Nightingale’s second diagram in Woodbury’s letter7

shows that Miss Nightingale wrote beside it: “The areas of the blue, red and black
wedges are each measured from the centre as the common vertex”. This makes it quite
clear that the colours are overlapped and so solves my first problem. She also wrote “In
October 1854 & April 1855 the black area coincides with the red”. She coloured the first
of these in red and the second in black, but just commented on it beside the diagram to
make it clear.
The article by Short8 was a joy to read, although I could only work out the mathematical
equations, which were written out in a way which is strange to me ( for example
“$$ \text{Area of sector B} = \frac{\pi r_B^2}{3}=3$$”8 ) because I already knew what
they were (The example had a sector B in a diagram which I could see had
1 2π 2 π 2
=areaB = rB rB ). The two things I found exciting from this article were the
2 3 3
table of data which Nightingale used to create the second diagram, and an explanation of
what rates of mortality she used. She described these as follows; “The ratios of deaths
and admissions to Force per 1000 per annum are calculated from the monthly ratios
given in Dr. Smith’s Table B”4 and I had not been able to understand the meaning of this
from the other articles. (Brasseur adds that “Dr. Smith was the late director-general of the
army.”4). Using Short’s article I was able to work out what it meant.
I will use an example of data taken from the table in Short’s article, which is in turn taken
from “A contribution to the sanitary history of the British army during the late war with
Russia” by Florence Nightingale of 18598. In February 1855 the average size of the army
was 30919. Of these 2120 died of ‘zymotic diseases’, 42 died of ‘wounds & injuries’ and
361 died of ‘all other causes’. This gives a total of 2120 + 42 + 361 = 2523 deaths. 2523

2523
out of 30919 means that ×1000 = 81.6003 men died per 1000 men in the army in
30919
that month. If the size of the army had stayed at 30919, with no more men being shipped
in or out, and the death rate had continued at 81.6 deaths per 1000 men per month over 12
months, the number of deaths per annum would have been 81.6003 ×12 = 979.2 per 1000
men in the army. In other words 979.2 deaths per 1000 per annum.
This understanding of the units used allowed me to finally understand why O’Connor
says of the death rate in January 1855, “if this rate had continued, and troops had not been
replaced frequently, then disease alone would have killed the entire British army in the C, D Interesting
Crimea.”5 The number of deaths due to disease in January 1855 was 2761 and the
background
2761
average size of the army was 32393. This gives a rate of ×1000 ×12 = 1022.8 information and
32393 reflection
deaths from disease per 1000 per annum. Another way of looking at it is that if 2761 had
dies each month from disease, 2761×12 = 33132 would have died in 12 months, but there
were only 32393 in the army!
As an aside, I noticed that O’Connor quoted the mortality rate for January 1855 as “1,023
per 10,000 being from zymotic diseases”5. Another example that we should not trust
everything we see in print.
Having sorted this out I was ready to attempt my recreation of figure 1. I decided to do
the right hand rose only, covering April 1854 to March 1855. The following table shows
the data taken from Short’s article in blue and my calculations in black:
Average Wounds
size of Zymotic & Z/S*1000*12 Radius W/S*1000*12 Radius O/S*1000*12 Radius
army diseases injuries Other (Az) for (Aw) for (Ao) for
Month (S) (Z) (W) (O) (1 d.p.) Zymtotic (1 d.p.) Wounds (1 d.p.) Other
Apr-54 8571 1 0 5 1.4 2.3 0.0 0.0 7.0 5.2
May-54 23333 12 0 9 6.2 4.9 0.0 0.0 4.6 4.2
Jun-54 28333 11 0 6 4.7 4.2 0.0 0.0 2.5 3.1
Jul-54 28722 359 0 23 150.0 23.9 0.0 0.0 9.6 6.1
Aug-54 30246 828 1 30 328.5 35.4 0.4 1.2 11.9 6.7
Sep-54 30290 788 81 70 312.2 34.5 32.1 11.1 27.7 10.3
Oct-54 30643 503 132 128 197.0 27.4 51.7 14.1 50.1 13.8
Nov-54 29736 844 287 106 340.6 36.1 115.8 21.0 42.8 12.8
Dec-54 32779 1725 114 131 631.5 49.1 41.7 12.6 48.0 13.5
Jan-55 32393 2761 83 324 1022.8 62.5 30.7 10.8 120.0 21.4
Feb-55 30919 2120 42 361 822.8 56.1 16.3 7.9 140.1 23.1
Mar-55 30107 1205 32 172 480.3 42.8 12.8 7.0 68.6 16.2
Az is the death rate per 1000 per annum from disease, Aw is the death rate per 1000 per
annum from wounds and Ao is the death rate per 1000 per annum from other causes. For
2π π
this diagram there are 12 divisions so each sector has an angle of = and an area of
12 6
1π 2 π 2 12A
r = r . So for each radius r = .
26 12 π

I will show my final polar area diagram side by side with Nightingale’s original version:
A Demonstration
Figure 7. of the fulfillment
Nightingale’s original of the aim
“Diagram of the causes of
mortality in the army in the
east” and my recreation.
I have to admit that I felt rather proud once I had done this! However, looking at the
C Personal
September 1854 wedge I realised that the two diagrams didn’t correspond. In
Nightingale’s original diagram I can see that there are more deaths from other causes than ownership
from wounds. In my version there are fewer deaths from other causes than from wounds.
All other versions of the original in other articles I looked at ( Gill and Gill3, Brasseur4,
O’Connor5, Woodbury7, Riddle10, Small11, Lienhard6) are as the original, but the table in
Short definitely shows fewer deaths from other causes than from wounds8.
Conclusion
I started out to try to lean how to recreate the polar area diagram which Florence
Nightingale made to communicate to other people just how bad the situation was in army
C Good link to
hospitals. This diagram shouts a need for reform. Look at it. The blue represents deaths
which could be avoided with a bit of organisation and care. The red represents deaths due human factor
to the actual battles. Florence Nightingale had copies of her report containing her
diagrams published at her own expense and sent them to doctors, army officers, members
of parliament and the Queen. Following her persistent lobbying a commission was set up
to improve military barracks and hospitals, sanitary codes were established and
procedures were put in place for more organised collection of medical statistics4. It is a
very shocking picture with a huge snowball of social change behind it. It has been an
A Very good
exciting adventure to drill down to a real understanding of its construction.
conclusion

However, the biggest lesson I have learnt from this research is that you can’t trust what
you read. As I have argued in the main text, I am moderately sure that Brasseur thought
the colours of the second diagram did not overlap4, I think O’Connor got his death rates
wrong for January 18555, and I think Short may have transcribed the data incorrectly for
September 18548. According to Brasseur, Florence Nightingale cross checked her data D Reflection
and was systematic about addressing objections to her analysis4. Everyone can make on what has
mistakes, and errors can propagate if we just quote what someone else says without been learned
looking for corroboration.
I have been left with a desire to find out more about this tenacious woman who wouldn’t
let society mould her into a genteel wife. Also, if I ever get the chance, I would like to get
a look at one of the 2000 copies of “Notes on Matters Affecting the Health, Effiency and
Hospital Administration of the British Army. Founded Chiefly on the Experience of the
Late War” which Florence Nightingale had published in 1858, to see the actual table of
data and check the numbers for September 1854.

References/Bibliography
1.Duthie, Eric ed. The Children’s Book of Famous Lives.Odhams Press Ltd, London 1957
2. Du Garde Peach, L. Florence Nightingale. Wills & Hepworth Ltd, Loughborough,

1959
3. Gill, Christopher J. and Gill, Gillian C. Nightingale in Scutari: Her Legacy

Reexamined Center for Internatinal Health, Boston University School of Public Health,
Boston, Massachusetts, <http://www.countryjoe.com/nightingale/FN%20in%20CID%20final.pdf >
viewed 26th July 2009
4. Brasseur, Lee, Florence Nightingale’s Visual Rhetoric in the Rose Diagrams.

Technical Communication Quarterly, 14(2), 161-182, Lawrence Erlbaum Associates, Inc,
2005, <http://benninghoff.emich.edu/424W06/handouts/Brasseur_Florence_Nightengale.pdf> viewed
26th July 2009
5. O’Connor, J.J. and Robertson, E.F., Florence Nightingale. <http://www-groups.dcs.st-

th
and.ac.uk/~history/Biographies/Nightingale.html> viewed 26 July 2009
6. Lienhard, John H., Nightingale’s Graph, The Engines of Our Ingenuity. 2002
<http://www.uh.edu/engines/epi1712.htm> viewed 26th July 2009
7. Woodbury, Henry, Nightingale’s Rose. American Physical SocietyLaunches Dynamic

Diagrams Redesign of Physical Review Letters, January 9, 2008, 4:05 pm, filed under
Information Design, Visual Explanation <http://dd.dynamicdiagrams.com/2008/01/nightingales-
th
rose/> Viewed 30 July 2009
8. Short, Ian, Mathematics of the Coxcombs. November 5th, 2008

<http://understandinguncertainty.org/node/214> viewed 30th July 2009
9. Lewi, Paul J. Florence Nightingale and Polar Area Diagrams, Speaking of Graphics.
2006 < www.datascope.be/sog/SOG-Chapter5.pdf> viewed 26th July 2009
10. Riddle, Larry, Polar-Area Diagram. 2006 ,

<http://www.scottlan.edu/lriddle/women/nightpiechart.htm> viewed 26th July 2009
11. Small, Hugh, Florence Nightingale’s statistical diagrams. Presentation to Research

Conference organized by the Florence Nightingale Museum St. Thomas’s Hospital, 18th
March 1998 <http://www.florence-nightingale-avenging-angel.co.uk/GraphicsPaper/Graphics.htm>
viewed 26th July 2009

Example 7: Annotated student work
Mathematical Exploration
Modelling Rainfall A Introduction

includes rationale
Rationale: and aim.
I have decided to centre this mathematical exploration around the topic of weather. C A good
To begin, I looked at the stimuli given to me by my teacher. While looking at different ideas, personal
weather was a topic that stood out to me, because never before, had I thought about the perspective
mathematical significance in weather. Would it be possible to predict tomorrow’s weather using
Mathematics? I did some research about the significance in meteorology, but did not find anything
very interesting in particular. Nevertheless, I wanted to stick to the topic of weather and decided to
relate it to a more personal matter. Having lived in a country like Germany for nearly all my life,
one is face to face with rain almost every day of the year. It comes and goes, no matter what. I thus
decided to write this exploration about rain, that is falling objects. Is it possible to model rainfall?
What would I need to consider? I first decided on the factors, which I wanted to account for in my
model and then looked a little closer at rain itself: There are different kinds of rain, and one would
have quite different experiences walking through a light drizzle than being outside on the street A Clarity
during a thunderstorm. needed on
I thus decided to explore different sizes of raindrops, first very small drops with diameter how these
d ≤ 0.008cm and then larger drops with diameter d ≥ 0.125cm . In addition, falling objects are not values were
only dependent upon gravity and mass itself, but on air resistance. Raindrops thus experience an chosen
upward drag, which must be accounted for when modelling rainfall. The velocity upon impact, thus
terminal velocity of the raindrop, and the time it takes the raindrop to reach the ground were
determined.
Finally, the model was extended to a parachute jump, which in a way is a very, very large raindrop
falling down from the sky. However, there is one essential difference: the parachute itself, as a
skydiver will always experience two periods of falling: the free-fall and the slow descent, after the
parachute is activated. This must be accounted for when developing a model.
Introduction:
Rain is a phenomenon we come across on a regularly basis in our day-to-day life. But what do we
actually know about it? We know it is liquid precipitation and that rain can be of different intensity,
such as light rain (a drizzle) or heavy rain (a storm). This exploration aims to develop a model for
rainfall.
In order to correctly model rainfall, I decided to first determine the factors related to rainfall, which
are thus of importance to the model: As mentioned above, the intensity of rain is never the same. It
is a measure of the amount of rain that falls over time, thus measured in the height of the water
layer covering the ground in a period of time, millimetres per hour. But not only the intensity of rain
matters. In order to determine the other factors, I found it easiest to just think of a falling object,
such as a pen, which is dropped at a certain height. What factors will have an impact upon the pen?
Gravitational force would be the first thing that springs to mind, but also acceleration and velocity
are important factors to consider, as is the distance of the pen from the ground and the time it takes
the pen to reach the ground.

First Model (not accounting for air resistance):
To develop the first model taking into account the factors mentioned above, I did some research
about the forces upon falling objects and found Newton’s Second law of motion, which states that
force equals mass times acceleration. 1 This can be written as:
F = ma , where F is force, m the mass of the object and a its acceleration.

Newton’s Law of Motion can be used to model a raindrop with constant acceleration, that is,
ignoring the air resistance. It is known that acceleration, a, is the first derivative of velocity, v, and
the second derivative of position, s. This can be written as:
dv d 2 s
a= =
dt dt 2
For a falling object, we can also state that a = g . We then get:

dv
=a=g
dt
We thus obtain a differential equation for a raindrop with constant acceleration, for which, in order
A Further
to solve it, we first need to set some initial conditions: The raindrop is starting at rest from 1000
metres in the air, thus the origin. By doing some research, it was obtained that the constant clarity needed
acceleration of falling objects, on which the only significant force in gravitational force, is
g = 9.81ms −2 . 2 It can also be stated that, at origin, v(0) = 0, s(0) = 0,t = 0 .
The differential equation can now be solved using basic integration:
dv
=g
dt
dv = gdt
∫ dv = ∫ g dt
v = gt + k
To find the constant of integration, the initial conditions explained above must be used, thus
v(0) = 0,t = 0
0 = g(0) + k
k=0
Thus:
v = gt
ds
=v
We can now use another position equation, dt , to continue. Again, basic integration is used to
solve the differential equation.
1
“Newton’s Three Laws of Motion“, Astronomy 162 Stars, Galaxies, and Cosmology,
<http://csep10.phys.utk.edu/astr161/lect/history/newton3laws.html>, 1 May 2010
2
“Acceleration of Gravity and Newton’s Second Law of Motion“, The Engineering ToolBox,
<http://www.engineeringtoolbox.com/accelaration-gravity-d_340.html>, 3 May 2010

v = gt
ds
v=
dt
ds
= gt
dt
ds = gtdt
∫ ds = ∫ gt dt
g 2
s= t +k
2
To determine the constant of integration, the initial conditions have to be used again, thus
s(0) = 0,t(0) = 0 . Thus,
B Explanation
needed
g 2 why t(0) is
0= (0) + k
2 included here
k=0
We now get:
g
s = t2
2
−2
As stated above, s = 1000m and g = 9.81ms . These values can be substituted in the above
equation to solve for time, t:
9.81 2
1000 = t
2
2000 = 9.81t 2
2000
t2 =
9.81
2000 E Accuracy
t=
9.81 is too great
t = 14.278 for three
significant
To determine the velocity of the raindrop upon arriving on the ground we can use the first equation, figures.
v = gt .
v = 9.81 × 14.278 E Accuracy

again—this will
v = 140.0714
limit criterion
We have now modelled a raindrop falling from a given distance, 1000 metres, which experience E grade.
−2
constant acceleration, g = 9.81ms . As aimed, the results obtained show the time, how long it takes
the raindrop to reach the ground, and its velocity upon reaching the ground. As shown above it
D Good
would take the raindrop approximately 14.278 seconds and it would hit the ground at around
140.0714 metres per second. reflection
on results
The value of the velocity seems rather big. In fact, this model is an unrealistic one, as the raindrop
was said to fall form a distance of one kilometre. In addition, only gravitational force was taken into
account. In reality, however, the raindrop is subject to air resistance and thus dragged upwards.

However, when observing the constant acceleration model, it can be said that it will work for any
size of raindrop, since neither acceleration, nor velocity is dependent on mass.
dv
When taking a closer look at the equation for acceleration, = g , we find that it is not possible to
dt
determine the terminal velocity of the raindrop, as g = 9.81ms −2 and acceleration can thus never be
zero. This means, that the raindrop will not stop accelerating, which in reality, is not the case. 3
Second Model (small raindrops, accounting for air resistance): D Good

comments
Having found this first model for rainfall, we now want to develop a more realistic one, thus taking
into account the air resistance, which is the force that acts against anything that moves through the
atmosphere or air. The amount of air resistance is dependent upon a variety of factors, most
important are the speed of the object and its cross sectional area. 4 Thus both increased speed and
cross sectional area of a moving object leads to an increase in the amount of air resistance.
To account for the air resistance encountered by air drops one must first find some further
information about the upward drag. Research shows that for very small drop with diameter
d ≤ 0.008cm experience a drag force which is proportional to the velocity. Given that force is
proportional to acceleration we get:
Fdrag ∝ adrag ∝ v
Thus,
adrag = kv
−1
Experiment shows that k = 12.2s .
−2
As stated above acceleration, when not accounting for air resistance, is constant, g = 9.81ms .
In order to find a new differential equation for this model, we now need to account for the constant
acceleration and the upward drag. Acceleration is thus given by:
dv
a=
dt
dv
= g − kv
dt
Using this differential equation we can easily find the terminal velocity of the raindrop. Terminal
velocity of a falling object is a constant velocity reached when there is no resultant acceleration,
acceleration is thus zero. We can thus obtain the terminal velocity from the model:
0 = g − kv
g = kv
g
v=
k
3
Emerson, Brandon, “Raindrop Modeling“, <bpemer.people.wm.edu/Raindrops.doc>, 28 April
2010
4
“The Forces at Work – Gravity“, < http://ffden-
2.phys.uaf.edu/211.fall2000.web.projects/vlad%20paverman/forces.htm>, 22 May 2010

Substituting in the values for g and k we get:

B Poor
9.81
v= use of
12.2 terminology—
v = 0.804 "verify" would
be better.
As we continue working on the model, thus integrating it, we can prove the above result by
graphing the equation found for velocity. Also, as we find the time it takes the raindrop to reach the
floor, the velocity equation, which takes time into account, should give us the same result as above.
To continue, we now integrate the differential equation:
dv
∫ g − kv = ∫ dt
 1 E Good
 −  ln (g − kv ) = t + a mathematics
k
demonstrated
ln(g − kv) = −kt + a
g − kv = e− kt + a B Incorrect
g − kv = e− kt × ea notation for
ea is a constant and can thus be written as b to simplify the equation. constant
We now get:
g − kv = b × e− kt
To obtain the constant of integration, b, we consider the initial conditions, which are the same as in
the first model, v(0) = 0,t = 0 .
Thus,
B Poor use of
g − k(0) = b × e− k (0) mathematical language
g=b
Plugging in the constant of integration and solving for the velocity, v we get:
B Poor
g − kv = ge − kt use of
terminology
g − ge− kt = kv ("prove")
g − ge− kt
v= E Good
k
g example of
v = (1 − e− kt ) connecting
k
different areas
We can now graph the equation above to prove that the terminal velocity, which was determined of mathematics
earlier, is correct. The graph should, at one point, have a gradient of zero. Thus, once the graph
reaches a horizontal plateau, we can determine the terminal velocity. E Incorrect
analysis
Point at which the gradient of
the curve is firstly equal to zero,
x=1, y=0.804
v 9.81
y= (1 − e−12.2t )
12.2

B Poor
display of
graph and
associated
data
As seen above, the graph is horizontal, and the point at which the gradient is equal to zero is
E
reached very fast, within a minute. It can thus be assumed that the droplets travel with terminal
velocity for almost the entire fall. This seems rather odd at first, why would a falling body Incorrect
experience zero acceleration for such a long time? This is due to the retarding force of air analysis
resistance, which exists because air molecules collide into a falling body and create an upward force
opposite gravity. Eventually, this upward drag balances the weight of the falling object, which will
thus continue to fall at constant, terminal velocity. 5
ds
=v
As before, we now use a second position equation, dt . Thus:
g
v= (1 − e− kt )
k
ds g
= (1 − e− kt )
dt k
g
ds = (1 − e− kt )dt
k
We can now integrate this problem:
5
Elert, Glenn, “Speed of a Skydiver (Terminal Velocity)“, The Physics
Factbookhttp://hypertextbook.com/facts/JianHuang.shtml, 10 May 2010

g
∫ ds = ∫ k (1 − e
− kt
)dt
g 1 
⇒s=  t + e− kt  + c
k k 
Solving for the constant of integration, c, we use the initial conditions, s(0) = 0,t(0) = 0 . We then
get:
g 1
0= (0 + ) + c
k k B Poor
g use of
⇒c=− 2
k mathematical
language
Plugging in the constant of integration:
g 1 g
s= (t + kt
)− 2
k ke k
g 1 1
s = (t + − )
k kekt k
We can now substitute in the known values, s = 1000, g = 9.81, k = 12.2 , to obtain:
9.81 1 1
1000 = (t + − )
12.2 12.2e12.2t 12.2
Instead of solving the equation algebraically, it is better to just graph the equation
9.81 1 1
y= (t + − )
12.2 12.2e 12.2t
12.2 and find the intersection with the equation y = 1000 . This can be
done on the graphing display calculator. However, to avoid complicated fractions it is best to first
simplify:
9.81 9.81
12200 = 9.81t + −
12.2e12.2t
12.2
We thus graph the two equations and find the point of intersection of the two graphs.
y Point of
intersection of
the two
graphs:
(1243.7109,
y = 12000 Typing mistake
12200)
condoned
9.81 9.81
y = 9.81t + 12.2t
−
12.2e 12.2
x

We thus get obtain that, at y=12200, x=1243.711. This number now needs to be divided by sixty, to
determine the time in minutes it will take the raindrop to reach the ground.
1243.711
= 20.729
60
It thus takes the raindrop approximately 21 minutes to reach the ground. Having found the time, we
can now determine the raindrop’s velocity upon reaching the ground. We found that
g
v= (1 − e− kt )
k
Thus:
9.81
v= (1 − e(−12.2 × 20.729) )
12.2
v = 0.804m / s
As mentioned above, the result is the same as the terminal velocity.
Third model (larger drops, accounting for air resistance):
We will now again extend the model of rainfall. It is unreasonable to assume that all droplets are
smaller than 0.008cm. We will thus consider larger raindrops with a diameter d ≥ 0.125cm . To
correctly model the rainfall of the larger drops, we will immediately account for the upward drag, B Error:
which in this case, experiment shows, is proportional to velocity squared and the constant of mentioned as
proportionality is c = 0.0097 . "c" here but
We can thus alter the model for smaller raindrops according to the new conditions. Thus: referred to as
"k" from here on
dv
= g − kv 2
dt
Again, we will integrate to continue:
dv
∫ g − kv 2
= ∫ dt

The integral on the left seems rather difficult, as we need to use partial fractions to integrate. Using
partial fractions, we know that:
1 1 1 C New mathematics
2
= − to this student and
g − kv 2 g × (v k + g ) 2 g × (v k − g )
not covered in class.
We can now use this to integrate, and, as g is a constant, we can rewrite it as: Good demonstration of
personal involvement
dv dv
∫2 g × (v k + g )
−∫
2 g × (v k − g ) ∫
= dt
1 dv 1 dv
×∫ ×∫
v k− g ∫
− = dt
2 g k v+ g 2 g
Integrating: Typing error

condoned
1 ln v k + g ln v k − g 
× −  =t
2 g  k k  B Lost variable
t on left-hand
1 ln v k + g − ln v k − g  side. Typing
×  =t
2 g  k  error condoned
as it reappears
correctly later on
Using the laws of logs, we get:
v k+ g B Notation
ln
1 v k− g error—
× =t +C constant
2 g k
appears late
v k+ g
ln = (t + C) × 2 g k
v k− g
The initial conditions are v = 0,t = 0 , thus C = 0 and:
ln
kv + g
kv − g
( )
= 2 kg t
Simplifying:
e
(2 kg )t = kv + g
E Good
kv − g
mathematical
e
(2 kg )t
( kv − g = kv + g ) manipulation of
the equations and
 (2 ) − 1 =
kg t  (2 ) + 1
kg t
kv  e  ge  expressions here
 
 (2 ) + 1
kg t
g e
v=  
k  e(2 ) − 1
kg t


We have now found the formula for the velocity. This is a differential equation for the distance s.
Simplifying further we get:
g  e(2 kg )t + 1 
v=
k  e(2 kg )t − 1 
ds g  e(2 kg )t − 1 + 2 
=
dt k  e(2 kg )t − 1 
ds g 2 
=  1 + (2 kg )t 
dt k e − 1
g  2 
∫ ds = k ∫  1 + e(2 kg )t − 1 dt
In order to integrate this fraction we use the substitution:
u=e
(2 kg )t
du = 2 kge
(2 kg )t dt B Variable t
is here, but
du obscured by the
⇒ dt =
2 kgu brackets—possible
 g  g problem with the
du
s=  t + 2 ∫ equation editor
 k  k  2 kgu(u − 1)
 g g du
2 ∫
s= t+
 k k g u(u − 1)
1 1 1
We can now use partial fractions to solve this last integral. We know that = − .
u(u − 1) (u − 1) u
Therefore:
du  1 1 E Sophisticated
∫ u(u − 1) = ∫  (u − 1) − u  dt level of mathematics.
, which can be written as: Integration involves
 1 1 dt dt partial fractions,
∫  (u − 1) − u  dt = ∫ (u − 1) − ∫ u substitution, and multiple
This can now be integrated. stages. Could be tidied
dt dt further, so not elegant,
∫ (u − 1) − ∫ u = ln u − 1 − ln u but impressive
As we used a substitution for u above, we now need to substitute u back in.
u = e( )
2 kg t
We thus obtain:
ln e( ) − 1 − ln e( )
2 kg t 2 kg t
Substituting the part back into the equation above we get a solution for s:
s=
 g
 k
g
 t + k 2 g ln e

( )
(2 kg )t − 1 − ln e(2 kg )t
We can now substitute in the known values, s = 1000, g = 9.81, k = 0.0097 , to obtain:

 9.81 
1000 =  
 0.0097 
t+
9.81
(0.0097)2 × 9.81
ln e(
2
( 0.0097 × 9.81 t ) − 1 − ln e(2 )
0.0097 × 9.81 t
)
In order to solve for t, we can graph the two equations y = 1000 and
y=
 9.81 
 0.0097 
 t+
9.81
(0.0097)2 × 9.81
ln e( (
2 0.0097 × 9.81 )t
− 1 − ln e(
2 0.0097 × 9.81 )t
)
and find the point of
intersection of the two graphs.
y
Point of
y = 1000 B x and y in
intersection:
x=1000, wrong order.
y=31.445 Correctly given
below, so condoned
 9.81 
y=
9.81
(
 t + (0.0097)2 × 9.81 ln e
 0.0097 
(2 ) − 1 − ln e(2
0.0097 × 9.81 t )
0.0097 × 9.81 t
)
x
We thus obtain that at y=1000, x=31.445. The value we found for x is the time it takes the raindrop
to reach the ground. This value can now be substituted into the equation for velocity:
g  e(2 kg )t + 1 
v=
k  e(2 kg )t − 1 
We thus obtain:
9.81  e(2 9.81× 0.0097 )× 31.445 + 1 
v=
0.0097  e(2 9.81× 0.0097 )× 31.445 − 1 
v = 31.802m / s
Again, we find that this is the terminal velocity as
dv
= g − kv 2
dt , as explained above. To find the terminal velocity we calculate v when acceleration is
equal to zero. Thus:
g = kv 2
g
v2 =
k
g
v=
k

9.81
v=
0.0097
v = 31.802
It is thus reasonable to assume that the large raindrops also travel at terminal velocity for most of
the time.
The Parachute Problem:
To apply the concept of raindrop modelling to human beings, one could now consider a parachute
jump. As I do not know much about skydiving, I first did some research around the basics of it.
A typical parachute jump consists of the individuals jumping out of an aircraft at approximately
4,400 metres altitude. The period of free-fall takes about a minute, then the parachute is opened to
slow the landing down, which takes between five and seven minutes.
To model a parachute jump, we can say that the parachutist, just like any other falling object,
−2
experiences constant acceleration g = 9.81ms , wince all objects fall with the same rate of
Typing error
acceleration regardless of their mass.
We also need to account for air resistance and can thus use the same model as before, when
modelling the fall of small raindrops.
dv
a=
dt
dv
= g − kv
dt
However, as the size of humans is not comparable to the size of small raindrops, we need to find a
different value for k, as the air resistance is not the same in this case. Also, the significance of air
resistance will ultimately depend on the position of the skydiver in the air. For example, a skydiver
in the spread eagle position will encounter more air resistance than a parachutist, who falls feet or
D Meaningful
head first, which related back to the cross sectional area of the falling object as explained earlier.
As explained earlier, the amount of air resistance is subject to change throughout the free fall, as it reflection
is also dependent upon speed. The air resistance is small at the start of the jump and the parachutist
accelerates downward. As the speed of the free faller increases, the amount of air resistance goes up
as well, thus slowing down acceleration. The air resistance force will eventually be equal and
opposite to the downward force due to gravity and at this point the skydiver’s speed remains
constant. Terminal velocity has thus been reached. 6
In order to develop a model for the parachute problem, one would thus have to divide the free fall
into three parts, each with a different amount of air resistance and then also develop a model for the
slow descent after the parachute is deployed. The mathematical process however is very simple and
the same as in the second model for small raindrops, k is just exchanged for the respective values of
the air resistance, that is the constant of proportionality in reference to velocity.
Having extended the mathematical exploration to the problem of a parachute jump, one can now
critically reflect upon the work done. We can start by looking at the first model, which merely looks
at raindrops as falling objects. It is an unrealistic model as such, as it does not account for air A Good conclusion
6
„KS§ 9K Air resistance and friction“, slideshare,
<http://www.slideshare.net/scienceinteractive/ks3-9k-air-resistance-and-friction>, 23 May 2010

resistance. The results could thus be used as a rough idea on how to develop the other models, but
other than that, were hypothetical.
The second model, which looks at smaller raindrops, takes air resistance into account and thus gets
a lot closer to reality. However, there are certain uncertainties to this model. To begin, we put all
raindrops with a diameter smaller that 0.008cm into one group, the results are thus not specific to
the respective size of the raindrop. In addition, we made the assumption that the amount of air
resistance is always that same. In reality, the amount of air resistance is subject to a number of
factors. As stated earlier, it depends upon the cross sectional area and speed of the object and thus
changes throughout the fall. However, this is only a minor inaccuracy, as we found out that
raindrops reach terminal velocity after a very short period of time and thus travel at terminal
velocity, and thus the same speed, for almost the entire fall. The air resistance, however, can also be
influenced by environmental factors, such as a strong wind.
These uncertainties are the same for the third model, looking at larger raindrops.
Bibliography:
Elert, Glenn. "Speed of a Skydiver (Terminal Velocity)." The Physics Factbook. Web. 10 May
2010. <http://hypertextbook.com/facts/JianHuang.shtml>.
Emerson, Brandon. "Raindrop Modeling." 2005. Web. 28 Apr. 2010.

<bpemer.people.wm.edu/Raindrops.doc>.
"The Forces at Work - Gravity." Web. 22 May 2010. <http://ffden-

2.phys.uaf.edu/211.fall2000.web.projects/vlad%20paverman/forces.htm>.
"Intensity of Rainfall." FLOODsite. Web. 07 May 2010.

<http://www.floodsite.net/juniorfloodsite/html/en/student/thingstoknow/hydrology/rainfallin
tensity.html>.
"KS3 9K Air Resistance and Friction." Slideshare. Web. 23 May 2010.

<http://www.slideshare.net/scienceinteractive/ks3-9k-air-resistance-and-friction>.
Meade, Douglas B. "Ode Models for the Parachute Problem." Web. 16 May 2010.
<http://www.math.sc.edu/~meade/papers/sr-parachute.pdf>.
Phoebus, Ronald, and Cole Reilly. "Differential Equations and the Parachute Problem." 10 May
2004. Web. 15 May 2010.
<http://online.redwoods.cc.ca.us/instruct/darnold/DEProj/sp04/coleron/paper1.pdf>.
Smith, David A., and Lawrence C. Moore. "Raindrops." Math DL The MAA Mathematical Science
Digital Library. The Mathematical Association of America. Web. 05 May 2010.
<http://mathdl.maa.org/mathDL/?pa=content&sa=viewDocument&nodeId=315&pf=1>.

"Student Project: Skydiving." Pure Mathematics 30. Alberta Education, Feb. 2009. Web. 18 May
2010. <http://education.alberta.ca/media/930712/03%20pure30%20feb-
09%20skydiving%20student%20project_pdf%20final.pdf>.

Spirals in Nature
A Aim and
rationale
When researching about mathematics in nature I found that certain spirals are found in shell
shapes. The Nautilus is a marine mollusk with a spiral shell with partitions to create buoyancy. included in
The adult mollusk can grow up to 25-30 centimetres across and the shell can withstand depths of introduction
up to 650 metres underwater. The chambers of the shell are separated but are interconnected via
a tube running through them. The tube pours gas or liquid through the tube to move the creature
around to sink or float respectively). The Nautilus shell’s curves are logarithmic and equiangular
with slightly different proportions to other spirals such as the Golden ratio.
I have a fossil of an ammonite which has a spiral in it. I want to find out if this has an
Archimedian spiral (a spiral described in polar coordinates by r = aθ , where a is a constant), or a
logarithmic spiral (a spiral described in polar coordinates by r = kecθ where c = cot φ ), or
something else altogether.
Here I have scanned the fossil and will attempt to model the spiral shown. In order to model the
spiral, I have marked several points along it, according to the picture below and have computed
C Developed
the x- and y-coordinates.
own example

There were 31 points. These are the coordinates and a plot of the coordinates to make sure I have
the right shape.
x y
648 -639
434 -741
221 -786
7 -773 C Good
-212 -742 example of
-421 -586 personal
-639 -216 engagement
-635 105
-537 323
-421 461
-207 581
2 599
221 537
434 336 C Applying unfamiliar
506 114
mathematics of polar
429 -158
225 -332 curves and polar points
2 -349
-212 -212
-283 7
-198 203
7 305
221 198
221 -46.8 B Very nice
7 -126 linking between
-105 11
the scan, the data
11 167
73 96 and the graph
96 33
7 -47
-25 74

Now I want to see if this spiral can be described as r = aθ , so I need to find the radius and angle
for each point. I will find the angle in radians. The radius is easy:=r x 2 + y 2 . The angle is a
bit more complicated because I want the point ( −25, 74 ) to have an angle of less than π, but I
want the point ( −105,11) to have an angle of nearly 3π because you can see from the plotting of
 y
the points that you have to go round more than a whole revolution to get there. arctan   gives
x E Good
an angle in the 1st or 4th quadrant so I worked out that if I know which quadrant a point is in and I understanding
keep the points all in order I can work out the angle by thinking about how many times I need to of the use
 y of the
add
= π, so θ arctan   + nπ where I define n in th following spreadsheet:
x periodic
x y quadrant n θ r logθ logr nature of
648 -639 4 8 24.35433596 910.0687 1.386576 2.959074 tangents
434 -741 4 8 24.09177962 858.7415 1.381869 2.933862 above SL
221 -786 4 8 23.83603866 816.4784 1.377234 2.911945 student
7 -773 4 8 23.57100028 773.0317 1.372378 2.888197 expectations
-212 -742 3 7 23.28364524 771.6916 1.367051 2.887444
-421 -586 3 7 22.93895658 721.5518 1.360574 2.858268
-639 -216 3 7 22.31711851 674.5198 1.348638 2.828995
-635 105 2 7 21.82727704 643.6226 1.339 2.808631
-537 323 2 7 21.44963438 626.6562 1.33142 2.797029
-421 461 2 7 21.16042999 624.3092 1.325524 2.7954
-207 581 2 7 20.76261275 616.7739 1.317282 2.790126
2 599 1 6 20.41701336 599.0033 1.309992 2.777429
221 537 1 6 20.02993254 580.6979 1.301679 2.76395
434 336 1 6 19.50836196 548.8643 1.290221 2.739465
506 114 1 6 19.0711525 518.6829 1.280377 2.714902
429 -158 4 6 18.49667356 457.1706 1.267094 2.660078
225 -332 4 6 17.87436924 401.0598 1.252231 2.603209
2 -349 4 6 17.28449019 349.0057 1.237657 2.542833
-212 -212 3 5 16.49336143 299.8133 1.217309 2.476851
-283 7 2 5 15.68323333 283.0866 1.195436 2.451919
-198 203 2 5 14.91009692 283.5719 1.17348 2.452663
7 305 1 4 14.11422015 305.0803 1.149657 2.484414
221 198 1 4 13.29693121 296.7238 1.123751 2.472352
221 -46.8 4 4 12.35768885 225.901 1.091937 2.353918
7 -126 4 4 11.05107279 126.1943 1.043404 2.10104
-105 11 2 3 9.320396808 105.5746 0.969434 2.02356
11 167 1 2 7.788208383 167.3619 0.891438 2.223657
73 96 1 2 7.203847123 120.6027 0.857564 2.081357
96 33 1 2 6.614281384 101.5135 0.820483 2.006524
7 -47 4 2 4.860238346 47.51842 0.686658 1.676862
-25 74 2 1 1.896595442 78.1089 0.277975 1.892701
If the spiral follows an Archemedian spiral, r = aθ , so plotting r against θ should give a straight
line of gradient a intersecting the vertical axis at the origin. I have plotted rand θ and fitted a line
of best fit by computer.

This doesn’t look at all promising!

Anyway I’ll try using a computer to draw a polar curve r = 38.9346 θ on top of the data.
Well, it’s a spiral, but it doesn’t fit the points well.

I have looked at the graph of r against θ and used the computer to fit a quadratic curve to the
points and a cubic curve to the points. The cubic seems to fit quite well. It looks like this:

I will try using the computer to draw r= 0.041132 θ 3 + 0.287964 θ 2 + 1.73643θ + 64.6669 on top
of the data.
E Good
understanding
of relationship
between the
graphs and
polar graphs
This does seem to follow the spiral more closely.
Now I will try fitting a logarithmic spiral. In the Encyclopaedia Britannica it said that the curve
was of the form r = kecθ where c = cot φ . I think c is a constant, so
ln r = ln kecθ
= ln k + ln ecθ
= ln k + cθ ln e
= ln k + cθ
So if I plot log r against θ I should get a straight line with gradient c and y-intercept ln k .

θ r lnr
24.35434 910.0687 6.81352
24.09178 858.7415 6.755468
23.83604 816.4784 6.705 lnr
23.571 773.0317 6.65032
23.28365 771.6916 6.648585
22.93896 721.5518 6.581404
22.31712 674.5198 6.514001
21.82728 643.6226 6.467112
21.44963 626.6562 6.440398
21.16043 624.3092 6.436646
20.76261 616.7739 6.424502
20.41701 599.0033 6.395267
20.02993 580.6979 6.364231
19.50836 548.8643 6.307851
19.07115 518.6829 6.251293
18.49667 457.1706 6.125057
17.87437 401.0598 5.994111
θ
17.28449 349.0057 5.855088
16.49336 299.8133 5.70316
15.68323 283.0866 5.645753
14.9101 283.5719 5.647466
14.11422 305.0803 5.720575
13.29693 296.7238 5.692802
12.35769 225.901 5.420097
11.05107 126.1943 4.837823
9.320397 105.5746 4.659418
7.788208 167.3619 5.120158
7.203847 120.6027 4.792501
6.614281 101.5135 4.620192
4.860238 47.51842 3.861117
1.896595 78.1089 4.358104
That looks a bit more likely! I will try c = 0.123083 , ln k = 3.80866 so
So
= r 45.09 × e0.123083θ
D Opportunities for
considering accuracy
and whether
using less accurate
parameters would
impact reasonableness

D Meaningful
reflection on
which is the
best approach
I think this looks very good. The cubic example looked very good too, however, so now I am
going to try to compare them. Every point on the spiral has a value of theta and a value of r and
each model has an approximate value of r for that theta. I have decided to find the absolute error
in each approximation of r and I will sum these errors for each of my models to see which has
the least sum and hence is the closest to the real points on the spiral.
θ r 38.9346θ abs(r-approx) cubic abs(r-approx) logarithmic abs(r-approx)

24.35434 910.0687 948.22633 38.15765008 871.9249 38.14378905 903.4901656 6.578513127
24.09178 858.7415 938.0038 79.2622816 848.796 9.945556819 874.7595101 16.01798906
23.83604 816.4784 928.04663 111.5682173 826.6997 10.22125627 847.6532419 31.1748282
23.571 773.0317 917.72747 144.6957735 804.2457 31.21404107 820.4474975 47.4158035
23.28365 771.6916 906.53941 134.8477658 780.4084 8.716744071 791.9366489 20.24500065
22.93896 721.5518 893.1191 171.5672989 752.5031 30.95132975 759.0412207 37.48942073
22.31712 674.5198 868.90808 194.3882533 704.0279 29.50807324 703.1135427 28.59371351
21.82728 643.6226 849.8363 206.2137403 667.5018 23.87923796 661.97463 18.35206976
21.44963 626.6562 835.13293 208.4767293 640.3202 13.66404316 631.9092812 5.25307558
21.16043 624.3092 823.87288 199.5636593 620.0709 4.238321177 609.8113805 14.49783772
20.76261 616.7739 808.38402 191.6101577 593.0079 23.76599034 580.6714795 36.10238507
20.41701 599.0033 794.92825 195.9249096 570.23 28.77338712 556.4891914 42.51414745
20.02993 580.6979 779.85741 199.1595555 545.5137 35.18410793 530.5980037 50.09985234
19.50836 548.8643 759.55027 210.6859875 513.5152 35.34907515 497.605805 51.25847699
19.07115 518.6829 742.52769 223.8447468 487.8235 30.85943987 471.5358609 47.14708662
18.49667 457.1706 720.16059 262.9899399 455.5973 1.573385921 439.3456954 17.82495108
17.87437 401.0598 695.93142 294.8715709 422.601 21.54119205 406.9504697 5.890623833
17.28449 349.0057 672.96471 323.9589812 393.1081 44.10233203 378.4512999 29.44556931
16.49336 299.8133 642.16243 342.3491548 356.1887 56.37547441 343.337146 43.52387082
15.68323 283.0866 610.62042 327.5338571 321.3955 38.30890977 310.753471 27.66691183
14.9101 283.5719 580.51866 296.9467993 290.9142 7.342293638 282.5455996 1.026260809
14.11422 305.0803 549.53152 244.4511986 262.1922 42.88811306 256.1801867 48.9001306
13.29693 296.7238 517.7107 220.9869207 235.3722 61.35154229 231.6636828 65.06009451
12.35769 225.901 481.14167 255.2407206 207.7241 18.17688581 206.3723186 19.52863313
11.05107 126.1943 430.2691 304.0748041 174.5371 48.34277322 175.7143535 49.52005884
9.320397 105.5746 362.88592 257.3113034 139.1695 33.59484783 142.0022486 36.42763044
A Huge,
detailed
tables stop the
exploration from
being concise.

7.788208 167.3619 303.23078 135.8688947 115.0883 52.27362354 117.5961792 49.7657042

7.203847 120.6027 280.47891 159.8762528 107.4969 13.10570694 109.4350922 11.1675612
6.614281 101.5135 257.5244 156.0108541 100.6524 0.861137386 101.7751622 0.261616296
4.860238 47.51842 189.23144 141.7130184 84.63091 37.11249648 82.01262581 34.49420833
1.896595 78.1089 73.843185 4.265713442 69.27664 8.832254711 56.94575507 21.16314327
sum 6238.41671 sum 840.1973621 sum 914.4071688
The Archemedian spiral had an absolute error sum of over 6000. The logarithmic spiral had an
absolute error sum of just over 914 and the cubic spiral which I just made up because the points
when I plotted r against θ looked as if they could follow a polynomial curve, has the smallest
error sum of about 840. I expected the logarithmic spiral to fit best because I had read that
A Good
Nautilus shells follow these curves, but that doesn’t seem to be the case for my fossil.
conclusion
It would be interesting to find other examples of ammonites from the many photographs
available on the web, and to model their curves by cubic spirals and logarithmic spirals and to D Limited
see which seem to fit more closely to find out whether the cubic shape is a good model for these reflection
spirals in general or whether it was a coincidence in this one case. on how the
exploration could
be extended
Bibliography
http://www.bsu.edu/web/math/exchange/01-01/allen.pdf
Britannica 2002 Deluxe Edition, Britannica.com Inc., 6 Jan. 2010
Murphy, Amanarth. "Maths of Nature and Nature of Maths Chapter 1." Scribd, 11
Jan. 2009. Web. 6 Jan. 2010. <http://www.scribd.com/doc/21990600/Maths-of-
Nature-and-Nature-of-Maths-Chapter-1>.
Images
Chambered Nautilus. Digital image. Inquiry By Design, Proportion. 7 Nov. 2006.
Web. 6 Jan. 2010.
<http://blog.lib.umn.edu/myee/architecture/Nautilus%20Shell%202.gif>.
Shell. Digital image. FH Perry Builder. Web. 6 Jan. 2010.

<http://www.fhperry.com/pages/relationships.html>.

The Legend (Introduction)

The Tower of Hanoi (also known as The Tower of Brahma and The A Good
introduction
End of The World Puzzle) was invented in 1883 by Edouard Lucas, a
French mathematician. It is said that he designed the puzzle based on a
legend of a Hindu Temple. In the beginning of time the priests in the
temple were given a tower of 64 gold disks, each smaller in size then the
disk beneath. They were to transfer the disks from one of three poles to
another without allowing any disk to be placed on top of a smaller one
(as the weight from the disk will crush the one beneath it). It is said that
when the day the priests successfully transfer the 64 disks from one pole
to another, the world will crumble and vanish. If this legend was true,
could there be a way of predicting the end of the world?
The Tower of Hanoi is a classic puzzle for all ages as the number of
disks creates endless levels of difficulty and fun. Though the aim of this
game is simple, it reveals many mathematical concepts and patterns
through the process of playing the puzzle. These patterns will be explored
A Aim
and analyzed and the legend of The Tower of Hanoi will be put to the test.
identified
Finding a Pattern
The task is to find out how many moves it takes 64 disks to
transfer from one pole to another pole of the Tower of Hanoi. Solving the
puzzle using a smaller number of disks will be easier to analyze and
understand. Let’s look at how the Tower of Hanoi is solved using 1, 2,
and 3 discs.
Figure A. One Move

Moves 1: move disk 1 to post c

B These diagrams
Figure B. Three Moves
are helpful, but
better clarity could
Move 1: move disk 2 to post B be achieved by
Move 2: move disk 1 to post C identifying which
Move 3: move disk 2 to post C is disk 1 and 2.
Figure C. Seven Moves

Move 1: move disk 3 to post C
Move 2: move disk 2 to post B
Move 3: move disk 3 to post B
Move 5: move disk 3 to post A
The number of moves it takes to complete the puzzle for an

amount of discs is a recursive pattern. A recursive pattern requires
information from the previous term of the pattern to determine the
following term. In this situation, it is the number of moves required for n
discs to transfer from A to C. A Unsure
what this
Figure 1-3 shows that one must first transfer n-1 discs (n being
refers to
the number of discs in the puzzle) to pole B. In Figure C it takes three

moves to transfer two discs to pole B (step 3 of Figure C). The number of
moves required shall be the variable M.
The next step is to transfer the remaining disc from A to C (step 4
of Figure C).
The last step is to transfer the discs from pole B to C. It will require
the same number of moves required in the first step also known as the
variable M. In Figure C it takes three moves to transfer n-1 discs from
pole B to pole C (step 7 of Figure C).
Reflecting on the steps you can see that to solve the puzzle you
move M amount of moves twice (from A to B, and from B to C). You also
make a single move from A to C. Mathematically the recursive pattern
should look like this:
B Not an
# of discs Total Moves Equation 2M + 1
equation
1 1 2(0) + 1 = 1
2 3 2(1) + 1 = 3
3 7 2(3) + 1 = 7
Table A.
In Table A under the Equation column, you are able to clearly see
why there is a recursive pattern present. The total moves from the 1 disc
puzzle become the M of the second puzzle as shown in the bold and
black numbers. The total moves for the 2 disc puzzle become the M of
the third puzzle as shown in the bold and grey numbers. Using the
recursive pattern, you are able to find the number of moves it takes for a
4, 5 disc puzzle and for an n disc puzzle as shown in Table B on page 4.
C More diagrams
would help here.
# of discs Total Moves Equation 2M + 1

1 1 2(0) + 1 = 1
2 3 2(1) + 1 = 3
3 7 2(3) + 1 = 7
4 15 2(7) + 1 = 15
5 31 2(15) + 1 = 31
n T 2(M) + 1 = T
Table B.
E Poor display
of knowledge—not
Referring back to the original task, the number of moves it takes to a recognized
solve a 64 disc puzzle is still unknown. It is possible to use the recursive geometric sequence
in differences
pattern and equation 2M + 1 to find the answer but in order to do this
one needs to know what M is. This reveals a weakness in recursive
patterns because in order to know what M is for 64 discs, one must
know M for 63 discs, M for 62 discs, and so on. This process would be
very time consuming and therefore not an efficient way of solving the
D Some
problem. reflection
here
Another problem with using the recursive pattern is if you wanted
to graph the data to compare the # of discs with the Total Moves. If one
wanted to use the recursive equation 2M + 1 one would discover that it is
not possible. This is because of the nature of a recursive equation. The M
variable is unknown unless the previous terms are known. Therefore it is
not possible to graph data using a recursive equation.
Perhaps a non-recursive way of looking at the puzzle would be
more efficient and useful. Using the data obtained from the recursive
pattern we can find an explicit pattern in the puzzle. An explicit pattern is
E Explicit versus
basically a non-recursive pattern meaning it doesn’t need information
recursive recognizes
from previous terms to obtain an answer. two important
mathematical ideas
related to patterns.

Creating a Formula
First, take the values from the first two columns of Table B. These values
are displayed below.
# of discs Total Moves
1 1
2 3
C Trying to find
3 7
a formula by using
4 15
graphical means
5 31
indicates independent
thinking.
n T
Table C
At first glance, there doesn’t seem to be an explicit relationship

between # of discs and total moves. One way to find a relationship
between the values is by guess-and-test until a working function is made.
But this method can be very time consuming and inefficient like the
previous one. A more efficient way to finding a function for two sets of
data is by graphing them. Below is a graph of # of discs versus total
moves.
B Well-labelled
graph; however,
student should try
to avoid wasted
open space by
adjusting the domain.
Graph A.

Graph A appears to have a trend which suggests there is a

relationship between the data graphed. From past knowledge of common
functions, we can try different functions and compare them to determine
which is best suited for this graph. Graph A has a growth trend because
the points move towards the right and upwards. There are two common
functions that show a similar shape to Graph A. The functions are
graphed and compared to the points in Graph A below.
Graph B compares the data points with the quadratic function y=x2.
B Student doesn't
acknowledge that
discrete data are
being modelled
with a continuous
function.
Graph B.
In graph B, the quadratic function y=x2 has a similar trend to the
data points except the quadratic function forms a parabola. This is
unsuitable to the data points because they do not form a trend in the E Confusing why
negatives are
shape of a parabola. Even when negative numbers are used (though this considered here
is unrealistic as it’s impossible to have negative # of moves), the trend is
E Not good knowledge

demonstrated—seems as though y=x2
is only possible parabola considered

still not that of a parabola. Therefore the quadratic function is not a

function of the data points.
Graph C compares the data points with the exponential function y=2x.
Graph C.
In Graph C, the exponential function y=2x has a very similar trend

to the trend of the data points. It is important to note that the
exponential function is actually y=ax where a > 0. Using y=1x would only
produce a horizontal line. Therefore the next value used was y=2x.
In order to make the function fit better with the points we make
transformations that slightly change the function. The first step is to
lower the line so that the line can become closer to the data points. This
is done by a transformation of 1 down.
D Good ideas
linking recursive
formula and
graph through
transformations

Graph D shows the function y=2x - 1 compared with the data points.
Graph D.
The function in graph D appears to completely fit with the data

points after the transformation. This means that there is a possibility E Good link
between the
that the relation between # of moves and Total moves is 2x - 1. We can pattern and
confirm this hypothesis by calculating the # of Discs with Total Moves the function
using the formula 2x – 1.
# of discs (x) Total Moves (y) Equation 2x – 1= y
1 1 21 -1 = 1
2 3 22 -1 = 3
3 7 23 -1 = 7
4 15 24 -1 = 15
5 31 25 -1 = 31
Table D.

Applying the Formula

The formula for finding the number of moves it takes an amount of
discs to move from pole A to C of the Tower of Hanoi is y = 2x – 1 where x
is the # of discs and y is the total amount of moves.
Now the formula has been found, it can be applied to a task. In
this case, the task is to find the number of moves it takes to transfer 64
disks from pole A to C of the Tower of Hanoi.
Total moves = 264 -1

B Inconsistent
=1.844674407 x 1019 total moves to transfer 64 discs from A to C
terminology—
approximate here
(without saying so)
To summarize, the legend of the Tower of Hanoi was that when all
but uses something
the discs have been transferred from A to C, the world will come to an for "approximately
equal to" later
end. Suppose one disc was transferred every second, how many years
will it take to complete the puzzle?
To solve this question, we take the total moves calculated and
divide them by units of time until we obtain the number of years it takes
to complete the puzzle.
Total moves=264 -1 = 1.844674407 x 1019

1 move = 1 second ∴ 60 moves = 1 minute of time passed
Converting seconds to years:

Total moves/60 seconds/60 minutes/24 hours/365 days= years passed
1.844674407 x 1019 /60/60/24/365 = 5.849424174 x 1011 years passed

=584,942,417,400 years passed ≅ 580 billion years
B Not
correct use of
If one disc was transferred every second, it would take approximately 580 "approximately
billion years to transfer 64 discs from pole A to C of The Tower of Hanoi. equal to" sign

Reflection A Rationale
identified at end
I chose this topic because I remember playing this simple but
challenging game at Science World. While playing the game I did not
naturally think of it as a mathematical concept but naturally I would
identify patterns in the game as I solve the puzzle and the difficulty
increases. I chose this topic because the history of it fascinates me. Also I
knew that the game had several mathematical concepts that were I
recently learned. I wanted to see if my mathematical knowledge of
transformations, functions, and patterns could be put to use and solve
the legend behind The Tower of Hanoi.
When exploring the mathematics of this game, I learned to see
things in life in a mathematical point-of-view as well as seeing
mathematics in a realistic point-of-view. The process of solving the
legend showed the relationships that mathematical concepts have with
each other as well as with life. I learned that everything is connected
mathematically and even the simplest things can be dissected and
analyzed in a mathematical way.
Conclusion D Rather
simple reflections
According to the legend of the Tower of Hanoi, if one disc was

transferred every second since the beginning of time, it would take about
580 billion years until the puzzle is solved and the world comes to an end.
If this were true, then the world still has many more years to live.
This solution was discovered through the recognition of a recursive
pattern in the puzzle. Through this recursive pattern the function
y=2x-1was created. Using this formula, the number of moves it takes to
solve a 64 disc Tower of Hanoi puzzle was obtained. This function is
useful for obtaining the number of moves for any amount of discs in The
Tower of Hanoi. This is useful when trying to obtain the minimal number
of moves to complete the puzzle as a way to challenge one’s intellectual
A Good
strength. conclusion—we
can see the aim
is fulfilled, so
work is complete.

Words Cited
Bogomolny, A. "Tower of Hanoi from Interactive Mathematics Miscellany and
Puzzles." Interactive Mathematics Miscellany and Puzzles, Index.
Alexander Bogomolny. Web. 04 Jan. 2010. <http://www.cut-the-
knot.org/recurrence/hanoi.shtml>.
Dr. Math. "Math Forum: Ask Dr. Math FAQ: Tower of Hanoi." The Math Forum @
Drexel University. Drexel University. Web. 04 Jan. 2010.
<http://mathforum.org/dr.math/faq/faq.tower.hanoi.html>.
"LHS: Tower of Hanoi Facts." Lawrence Hall of Science. Lawrence Hall of
Science. Web. 08 Jan. 2010.
<http://lawrencehallofscience.org/java/tower/towerhistory.html>.
Weisstein, Eric W. "Tower of Hanoi -- from Wolfram MathWorld." Wolfram
MathWorld: The Web's Most Extensive Mathematics Resource. Wolfram
Research, Inc., 07 Jan. 2010. Web. 08 Jan. 2010.
<http://mathworld.wolfram.com/TowerofHanoi.html>.

A The introduction
includes a clear and
explicit description of
the aim, as well as a
rationale. The plan of
action is not explicitly
explained.
B Constraints relative to
this specific study are
appropriately defined.
C The Mathematical
exploration relates to
personal interest.
C A specific model is
chosen.
A/B Useful explanation,

appropriately supported
by the diagram.
B The variable P
used in the formula is
explained at this point.
The definition (fluid
density) is made at the
next paragraph.

E Research has been

conducted on the
literature of the domain
of the study, in order
to provide the equations
relevant to the problem.
B The (differential) ∆
operator is not defined
in the equations provided.
It is implicitly explained
at the next line.
A Any particular reason

for choosing these
values?
A, B Interesting
explanations on the
method and process
that will be used.
C Research has been

conducted on different
models and a choice is
made.
E Very good
understanding of the
complexity of the
modelling.

B How was this

equation determined any
particular reason?
E Appropriate use of
mathematics to calculate
the area under the
curve.
A The task would
benefit from relevant
explanations, in order
to explain the process
to readers who are
not familiar with the
problem or mathematics
involved.
B The results provided

are correct to three
significant figures.
E The vertex lines on

the axis of symmetry, a
point cannot be an axis
of symmetry (improper
terminology).
A, B, C Although the
explanations are not
correctly expressed,
the symmetry of the
quadratic function is
used in order to avoid
repetition of calculations.
This helps making the
work concise.

C Personal example.
A Appropriate
explanations.
B, E Technology is used
in order to determine
this function.
B, E The diagram displays

the function modelling the
aerofoil, It is correctly
presented with title,
labels, units. The use of
the space is not optimal.

E Interesting idea.
E Appropriate use of
mathematics, beyond the
scope of the Math HL
syllabus.
E Correct use of
calculus.
B The data presented

in table are expressed
in decimetre. This value
(5cm) corresponds to the
first row.
A Extended research
has been conducted.
D Critical reflection on
the research made and
its applications to the
current problem.
E Good understanding
of the subject is
demonstrated.

B Good use of
technology, in order to
generate (correct) data.

E Good reasoning,
supported by appropriate
explanations
B, E Correct and
appropriate data are
generated.

A, B, E Final results
for the drag force are
produced. The lines of
reasoning are consistent
and easy to follow
through.
B Appropriate display of
the results.

E The method used

to find the model is
similar to that used in
Physics.
B, E The data
manipulated are
displayed and the linear
relationship is visible.
The unit of the y-axis
is not properly expressed
(confusing notation). The
line of best fit is also
displayed.

B ...function...
D Critical analysis of the

function obtained.
D Critical analysis of the

results obtained for the
lift force.
D Very good and

interesting analysis of the
results in the light of
further research.

B Interesting
display of the data
which offers a visual
representation of the
data presented in the
table.
B The use of arrows

needs to be explained.
B The values used

below correspond to
those for a thickness
factor h equal to 1.32
(see table p.8).
B It is important
to mention that this
function is an estimation
as more calculations are
necessary to prove it.

D Critical analysis of
the initial assumptions/
simplifications made for
the investigation.
B, D Interesting
suggestion for possible and
realistic improvements.


No explanation of
notation.

The mathematics
requires more clarity and
explanation.
How do you know?

Exactly?
Source?

Probability of what?


Some attempt at reflection.

Aim and
rationale
Citation?




Evidence of reflection
Labels?

Evidence of understanding








A1

See bibliography 2
wikipedia

Good use
of geogebra



B2
B2
A2

Some reflection but

not very meaningful

E1
B3
B4


B4
E2

B7 OK - student
tabulates results


Aim and rationale not clearly stated.


Points A,B,C need to be

indicated on the diagram.
Also X,Y,Z.
A Confusion between
explanationanddiagram.
This is evidence of personal engagement.

Drawing these correctly is challenging.



Correctly noted but

perhaps this is obvious.

Clear well
presented
photo.
Consideration of error is a
good example of meaningful
reflection.


Should be 0.361 not 0.344 but

does not detract from flow.
Error calculation evidence of

meaningful reflection.


This is reflection but not

critical reflection.




Personal
engagement.
Aim.

Defines variables.
variables
Key terms are explained.

Not well defined.


Personal
engagement.
What is N?

More explanation for the

derivation of ∆ and у would
help here.
The graph is defined clearly by the data in

the appendix. The long table would detract
from communication.

Good observation



A little
too wordy.
Good
observation.



Reflection.
Personal
engagement


Personal
example.




Personal
example.



Limited data
Rationale for the

investigation
A+ Introduction and
background information
In bibliography

Personal engagement
Good use of table to

organize data.

Discussed limitations of data.
Good use of tables to

show data.

D+
B - Representation of fractions.
D + Good analysis.

More analyses

Student does a good job in comparing

different parts of the body

C + Good - more data.

Student analyses the findings.

B - Why n-1?

B - Difficult to follow reasoning -

how does this help to reach the
conclusion?

D - The student attempts to use

another mathematical process of
comparing the data but the results
are not clear.
Why?


Good point.
Student reflects
on exploration.


A Rationale.
D+
Images referenced in sources.
A Defines types
of domes.

D+
B - ∏ symbol. KM2 in formula.
B - Used rounded value for r.

C+
B+
3 sig fig (rounding error).

E +, B - Correct but repeated rounding in work. D + Superficial
D + Meaningful



Does it?

B - Difficult to read. E Why these functions?

Could be clearer.
D+

D -, B - Why aren’t the dome

types integers?

B - Define x
2.44?
D+

D+
D+

A+
C+

D+
B+

B+


C + Devises own method.

B+

C + New to student.

E+
D+

D+

C+
E+
D+
D+

Period, not multiply - ok. 1/0.5

D+
— 35.479...

D+
B - Missing dx
D+
D+

Good conclusion



B - Why less than or equal to?

B + x- and y- defined.





What is B?
D -, E - Why do you
think a linear regression
makes sense?
B Units explained earlier
A - This is very difficult to read.


It seems like a linear function is not

the best way to model this data.


D-



C+
D+
A + Aim
C+

B+
Not a scatter graph
Units in table.
E+

Parameters
E+
B - Two “a’s”


D+
Typing mistake? (b)


D+
C+


B+

D+

B+
D+

D+
A + Returns to aim.
D+




A + Good introduction.
C + Topic is personal.

A + Clear aim.

C+
D+
D+

B + Clear graph.
y-axis explained.




B + Variables defined.
Condone * for
multiplication
C + Deriving own formula.

x condoned here

mathematics correct




My IB Annoted IA S 3 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

My IB Annoted IA S 3 PDF

Uploaded by

Copyright:

Available Formats

Example 1: Student work

Breaking the Code

“Secrets are made to be found out with time”

Mathematics SL and HL teacher support material 1

abcd abdc acbd acdb adbc adcb

bacd badc bcad bcda bdac bdca

cabd cadb cbad cbda cdab cdba

dabc dacb dbac dbca dcab dcba

4!− 4 × 3!+ 4C2 × 2! (1.2)

Mathematics SL and HL teacher support material 2

The number of derangements is

4!− 4×3!+4C2 × 2!− 4 +1 (1.3)

This can be written as

Breaking a substitution cipher

Mathematics SL and HL teacher support material 3

stellastellastellastellastellastellastellastellastellast C Use of own

The Vigenère cipher is not susceptible to breaking by frequency analysis as each

Breaking the Vigenère cipher

Consider the following encoded text:

JERFIWFLXFIWFLXVHXEAMCNWVHXHIWFLX C Further example

Mathematics SL and HL teacher support material 4

The Enigma Machine

In addition to the 3 rotors, pairs of letters could be connected on the keyboard. If a

26 26! 24! 22! 8! E Good demonstration of

Mathematics SL and HL teacher support material 5

Breaking the Enigma Code

Coded message: QZWPM ZHVVG YGQOZ UQZX

D Student could have

Mathematics SL and HL teacher support material 6

Step Simple example

So a = 15x + r and b = 15 y + R where x and y are integers.

ab = (15x + r )(15 y + R ) = 152 xy + 15xR + 15r + rR

ab (modulo m) = ((a modulo m )(b modulo m )) modulo m .

Mathematics SL and HL teacher support material 7

20 207 ≡ 202 × 202 × 202 × 20

So Euler’s theorem implies that M A ≡ 1mod m

ED −1 is a multiple of A so ED −1 = kA where k is an integer.

Mathematics SL and HL teacher support material 8

Breaking the RSA code

n 655427/n Integer part of 655427/n Factor?

438 1496.408676 1496 no

Suppose m = x2 − y2 then x2 − m = y 2. This means x2 ≥ m as square numbers cannot

Mathematics SL and HL teacher support material 9

n 810+n (810+n)^2-655427=Y SQRT(Y) Factor Factor

155 965 275798 525.165 no no

The Code Book, Simon Singh, Fourth Estate (2000)

Codes and ciphers http://www.bletchleypark.org.uk/edu/teachers/ccresources.rhtm

Euler function and theorem http://www.cut-the-knot.org/blue/Euler.shtml

Mathematical formulae http://www.po28.dial.pipex.com/maths/formulae.htm

Mathematics SL and HL teacher support material 10

Mathematics SL and HL teacher support material 1

Proving Euler’s Totient Theorem

Euler’s Totient Function 3

Fermat’s Little Theorem 4

Proving Fermat’s Little Theorem 5

Proving Euler’s Totient Theorem 6

Proving Euler’s Totient Theorem

Mathematics SL and HL teacher support material 2

Euler’s Totient Function

Mathematics SL and HL teacher support material 3

Some examples will serve to demonstrate Euler’s totient theorem.

Also, if n = 15 and a = 2 we see that 28 = 256 ≡ 1 (mod 15).

Fermat’s Little Theorem