Speed Arithmetic

- (Based on Vedic Word-sede as 2

‘Second Edition |
 }
:i.)al ae
a
;vo 4
Mey |a ct
pi
Speed Arithmetic
(BASED ON VEDIC WORD-FORMULAS)
Second Edition
 Digitized by the Internet Archive
in 2022 with funding from
Kahle/Austin Foundation

https ://archive.org/details/spoeedarithmetico0O000rama
 Speed Arithmetic
(BASED ON VEDIC WORD-FORMULAS)
Second Edition

Prof. Dr. T.A. RAMASUBBAN, msc. (Bom); Ph.D. (Lond)

Foreword by

Dr. K. Venkatasubramanian
Former Vice-Chancellor
Central University
Pondicherry

Published by

The Author
No.1, Mooparappan Street, T.Nagar, Chennai — 600017.
Phone : 4345027
E-mail : [email protected]
First Published : August,1999
Reprinted : June, 2000
Second Reprint : December, 2000
Second Edition : September, 2001
Reprinted : June 2002, Nov 2002, May 2003

ISBN : 81-7525-344-4

© Copyright with the author

This book, or any part thereof, must not be reproduced or stored in a retrieval system or
transmitted in any form or by any means electronic, mechanical, photocopy, recording or
otherwise, without written permission from the author.

Available at :

(1) No.1, Mooparappan Street, T.Nagar, Chennai — 600017.

Phone : 4345027
E-mail : [email protected]

(2) All leading book stores.

Price : Rs.120/-

Printed at: — Udayam Offsets, Chennai - 2.

® : 8534808; 8520906
 FOREWORD

In the past. education was for the elite and then knowledge was less important than
values. But today we have to recognise rightly the fact that education is for the masses and
this has resulted in a situation where knowledge is more important than values. The Gurukula
system emphasised values as per the great Indian cultural traditions. Modern pragmatic
systems feel that results are more sacred than values. Education commission of communist
China has emphasised that universities are the purveyors of knowledge and not values. Perhaps
this is the naked truth.

But in India we have to find a mix on Gandhian lines that vaiue based education is
most important and to us the path is more relevant than the end result. “Do your duty and
result is not thy concern” - so said Lord Krishna in the famous Bhagavad Gita.

Writing in those days in ‘Harijan’, the Father of the Nation Mahatma Gandhi said
that value oriented education alone mattered and illiteracy could be prefered to education
without a soul.

We find today that the universities are managed with yesterday’s tools for tomorrow’s
life. It is our duty today to give a new orientation to education to face the new challenge.

The future of India is very clear. India is the Guru of the wor'd. The future
structure of the world depends on India. India is the living soul of the globe, so said the
Great Master Sri Aurobindo. However what is the present position?

Inspite of the material affluency, the western civilization is unable to solve some of
the basic problems of existence. Crime is on the increase, police bandobast is proportionately
increasing. Still the problem of peace is eluding. The developing countries are blindly imitating
the west. Consumerism has come to stay. There is no end to human desires. Humanity
has entered into a vicious circle. There is no way out. Complete break-through is needed.
Once upon a time, the religions were the gifts of the Divine to the humanity. In the beginning,
religions gave solace and peace of mind. They were the instruments of progress. Now the
same religions have become instruments of retardation. Individual freedom is discounted.
Freedom of expression is stifled. Medievalism is raising its head. The echoes of inquisition
are re-echoing. Inspite of science and technology, the followers of religions are not able to
throw away the worn-out and meaningless customs.
 (iv)

What is the way out? We have to go back to the original fountains of knowledge.

Our Vedas are perennial sources of inspiration. The concept of Vedic Mathematics
is an astounding phenomenon. We are stunned at the modernity of the ancients.

Prof. Dr. T.A. Ramasubban has done an yeoman service to the cause of education
in general and to mathematics in particular by bringing out a wonderful publication entitled,
“SPEED ARITHMETIC” based on Vedic word formulae.

The nine chapters so profoundly written by the author gives the reader an accurate
picture of the marvels and mysteries of numbers. Where possible, the author tries to give
simpler methods of derivations and has added exercises at the end of many chapters. This
is a valuable addition as it makes the book complete for the sincere student.

Every para in this learned book emphasises the seminal idea of Speed Mathematics as
borne by the Vedic traditions.

It was Max Mueller who said “the Hindu Vedas are the great repositories of rare
wisdom’. You find everything in the Vedas and mathematics is no exception.

Mathematicians all over the world owe a deep debt of gratitutde to Prof. Dr.
T.A.Ramasubban for this very valuable ‘Vademecum’.

Chennai, India Dr. K. VENKATASUBRAMANIAN

August, 1999. Former Vice-Chancellor,
Central University,
Pondicherry.
 PREFACE TO THE SECOND EDITION

The overwhelming and enthusiastic response from the public towards the first
edition of his book “SPEED ARITHMETIC (Based on Vedic Word-Formulas)” has

prompted the author to bring out the present second edition.

In this edition, two new chapters namely “Addition and Subtraction” (Chapter 3)
and “Auxiliary Fractions” (Chapter 9) have been added. A few additional results, not too
widely known, have been included in section 1.2 of Chapter | (Marvels and Mysteries of
Numbers) and Chapter 4 (Multiplication) has been extended to include a new section.

Chapters 10 and 11 (Extraction of Square Roots and Cube Roots) have been revised to
enable the users to employ the methods discussed even more simply and quickly.

The author hopes that with these newly added and revised features the readers, at
all levels, would find working with numbers an enjoyable experience.

Chennai, India. TAR

September, 2001.
 < « caeeaasi co es
ow aba ail velt,marlNode, perey

ane (pincneet nica caine poet . 2g atk

aH me A Rs pea ihebe oneal a Wed

Need 9A Rj
ogeware ane} "2 aeeeatirs 17 particular Dy cringare mire oocsiuheprtt

« waagE SY”abide Bis HHMI tial ASA Won vst a

oasosethutegn Jomgaiditibag. eH ds. does svrgchyp ah ( voiged Dh apenas
At ae aie ty OD neiio we Zant
onbuberpinebdhe orig a
4 re
,Aak sc e a :a e
igi oi abnioiae>

ara Se at a
ee
1o¥s RGes Saoilb t r
. "aa worn

sp wilont wee fortis bobbie Neary “oeatywiesey eS

. " 09hT94/43 ain iohe evademusn iia sachow bl
7 ae 4 A _ » =
. ne = ad % 5 i
- * ~
oat }
ars 7 oe ®..

AT

yj

te
 roi

PREFACE TO THE FIRST EDITION

About Vedic Mathematics.

"Why were we not taught this before"
"It is all fun, it is so easy; you have several ways to do the problems
and you can choose the easiest way"
"It is our most favourite subject"
"We call out the answer even before the teacher can write the sum
on the board"
"How many ways can you come up with to solve the problem; we
enjoy this challenge”.
These are some of the spontaneous reactions of students in schools and colleges in some
western countries where Vedic Mathematics has been initiated.
Well, what is Vedic Mathematics that it should receive such an acclamation abroad.

Its place of origin is India (the land that gave birth arguably to the present system of
numerals, decimals and zero) and dates back to prehistoric times. In essence, it is based on
sixteen formulas (sutras) made up of barely 120 words in all. It was discovered by Jagadguru
Swami Sri Bharathi Krsna Tirtha Maharaja, Sankaracharya of Govardhana Matha , Puri and
propagated in his pioneering work in the year 1965.
It must be emphasized that although because of its origin, it is rightly called "Vedic"
Mathematics, the mathematics it deals with is universal in that it does not require the knowledge
of Sanskrit nor any acquaintance with the vedas, nor even the literal meanings of the word-
formulas on which the various methods rest. Any lay person who is conversant with the basics
of mathematics can understand and follow these methods.
This being so, it is sad that Vedic Mathematics has been neglected in its own country
of birth and that it is the other countries which have appreciated its potential and applicability.
In the last few years, students, teachers and mathematicians in countries as diverse as UK
(where some of the schools teach it even exclusively), USA, Holland, Canada, Spain, Hongkong,
Taiwan, Amsterdam and Mexico continue to accept it increasingly and have started teaching
Vedic Mathematics in their schools. Recently, the scientists in NASA have come to recognise
its scope and importance in the field of artificial intelligence. Yet, we in this country, fight shy
to teach, learn and use these methods and continue to be wedded to the existing conventional
methods. We still have to attempt a concerted beginning to make these methods widely known
and practiced.
 (vill)

Let us now see why Vedic Mathematics has become so popular and is rated so high by
the users abroad.

e It is very original, totally unconventional and provides a new thinking and approach to
mathematics that is so powerful, yet simple and fun to work with.
Its word-formulas lead to efficient and superfast calculations.
It encourages mental calculations since it handles the numbers involved digit by digit
of small magnitudes.
It is easy, simple,direct and straight forward. The students are freed from committing
to memory the multiplication tables and the like. They are enthusiastic in learning these
methods and eagerly look forward to the mathematics classes. They also enjoy sharing
their new-found knowledge with others.
Every one loves variety and vedic mathematics offers it in ample measure.
It is very comprehensive and is based on pattern recognition.
It is exceedingly flexible and offers a large number of alternatives for almost every kind
of problem. It is this aspect that captivates the students and others the most. The
weaker students gain mastery of the subject slowly and steadily by taking advantage of
the flexibility and availability of alternatives for a given problem; in contrast, presently
available blanket methods provide only two options: "to learn them or leave them". On
the other hand, the bright students take full advantage of the flexibility by learning to be
creative and challengeful in choosing the most appropriate and time-saving method that
is consistent with the underlying pattern of a given problem. +
Finally, all areas of modern mathematics, be it pure or applied, come under the scope
of vedic mathematics. This is because it provides a unified approach to different areas
of mathematics by the application of the same word-formulas to diverse fields like
trigonometry, analytical geometry, calculus, differential equations and so on. The present
day researchers in vedic mathematics have confirmed. the enormous versatility that it has
in this respect. The spin-off of this extendability is that students who have been trained
in the application of vedic mathematics to arithmetic can easily move over to college and
higher mathematics deriving the same benefits as they did in the case of arithmetic.

About the Book.

The book is based on the word-formulas of vedic mathematics discovered by Jagatguru

Swami Sri Bharathi Krsna Thirtha Maharaja, Sankaracharya of Govardhana Matha, Puri, and
draws inspriation for its techniques from the pioneering work "Vedic Mathematics" written by
 (ix)

him and which has come to be the fountain head for authors of this subject. I have provided
a step-by-step procedure for every illustrative example to facilitate easy reading and understanding
by the readers in general and the new comers in particular. I have also taken the opportunity
to develop some new alternative methods and present the same. These are identified by
double asterisks(**) at the beginning and end of such methods.
The main objective of the present book is, to make the method of this "new mathematics"
widely known and popular amongst two groups of readers.
1. The school going children who have to study mathematics compulsorily as
well as amongst college students whose studies include mathematics directly or remotely.
With about fifty years of experience of teaching mathematics and allied subjects both in
India and abroad and sharing the genuine fear and handicaps of these students in facing
mathematics, Iam _ of the firm view that the present treatise would go a long way to
remove from their mind, all anxieties ( and even hatred in some extreme cases) of
mathematics and instead help to implant in them a joyful feeling of acceptance of and
admiration for the subject.
2. Another group to whom the book is targetted is those who have completed
their formal education and wish to enter professional courses and services like MBA etc.
It has now-a-days become unavoidable to do so unless they succeed in their respective
entrance examinations. These examinations have become extremely competitive mainly
because of the necessity to answer a large number of questions in arithmetic and
mathematics in the least permissible time and this too without any calculating aids like
logarithmic tables, slide rules or pocket calculators. The examinees exposed to the
conventional methods of calculations soon realise that while these methods can assure
accuracy of answers, it is often at the cost of speed. On the other hand, those exposed
to vedic mathematics would consider it a unique and distinct advantage as it has both
speed and accuracy built into it. For this reason, the latter students would have a
winning edge over the former in these competitive examinations.
Naturally, I have taken extreme care to see that the book retains fully the original flavour
and the various attributes of vedic mathematics that are described in the previous section. It
has nine chapters in all and includes all the basic topics in arithmetic such as multiplication,
division, fractions, percentages and decimals, squares, cubes, square roots and cube roots of
numbers. A chapter has been devoted to complements and vinculums, at the beginning of the
book to enable the readers to be familiar with them and use the same in the later chapters with
ease and confidence. More than one method is discussed for many of the topics as the table
of contents would testify. Full descriptions explanations of each method are presented using
a variety of illustrative examples. Ample exercises are provided at the end of the chapters for
readers to practice on their own and develop their skills.
 (x)

The book, as stated in its title, is based on vedic word-formulas also refered to as
aphorisms or sutras. I have grouped these formulas in Appendix 2 (at the end of the book)
which lists them in English along with their translations and the arithmetic operations to which
they relate. Readers interested in them may refer to it. On the other hand, those concerned more
with methods and procedures than with word-formulas persemay'continue with the text without
the digression into the details of the word-formulas on which they are based. In any case, I
have identified the different methods by names which truly reflect the respective word-formulas.
The text assumes that the readers have the ability to perform the four basic operations
of arithmetic. Students from about the 8th standard may form the appropriate audience since,
by then, they would have gone through the traditional methods of arithmetic sufficient enough
to enable them to compare and appreciate the merits of the non-traditional methods offered
in this book.
As regards the time duration required by the students to become well conversant with
the material presented here, they may need 8-12 months at an average of two hours per day.
In view of their simplicity and variety, the methods may be learnt in a spirit of fun and challenge
outside the class hours perhaps in the company of two or more persons. It would be even
more desirable if schools can allot an hour or so for their students to learn this mathematics
as an extra-curricular activity preferably under the guidance of able and inspired teachers who
are themselves familiar with it.
My aim in writing this book is to draw attention of students, teachers and educationists
to an alternative system of arithmetic which is more variant, vibrant and a lot more speedier
in comparison to the current one. It is earnestly hoped that it will find its due place in
mathematics-curricula of various schools and institutions even if not exclusively as it has in
other countries but at least as a companion to the existing one.

Acknowledgements
I am greateful to Dr. K. Venkatasubramanian, former Vice-Chancellor of Pondicherry
Central University for having written the foreword to the book. I wish to acknowledge my
sincere thanks to Ms. Gayatri Anand, my daughter-in-law, who has assisted me in reviewing
the manuscript, suggesting revisions, wherever necessary, for easy understanding and compiling
exercises. Her assistance has been invaluable. I also owe my indebtedness to my wife Ms.
Vijayakumari who has so generously offered to decipher my illegible manuscript to a readable
computerscript.

Chennai, India. TAR

August, 1999.
 CONTENTS
Page No.
Foreword (111) — (iv)
Preface to the Second Edition (v)
Preface to the First Edition (vil) — (x)

CHAPTER 1 MARVELS AND MYSTERIES OF NUMBERS 1-10

1.1. Digital Root of a Number l
1.2. Marvels of Some Numbers 3
1.3. Binary Method of Multiplying Two or More Numbers +
1.4. Prime Numbers 5
1.5. Integral Powers of Numbers d.
Ome tre briples 3.6.9. 8
1.7. Mysteries of Number 9 9
1.8. Divisibility Tests of Numbers )
1.9. Reciprocals of Some Prime Numbers 10

CHAPTER 2 COMPLEMENT AND VINCULUM 11-18

2.1 Complement
2.2 Vinculum 12
(a) Conversion of digits of a number to their vinculums
(b) Conversion of vinculums digits of a number to
non-vinculum digits. 14
(c) Removing a vinculum or shifting of avinculum. 14
Exercises 2 (a) - 2(b) 14
2.3 Addition, Subtraction, Multiplication and Division of
Vinculum Numbers. iS
Exercises 2(c) - 2(g). 18

CHPTER 3 ADDITION AND SUBTRACTION 19-24

3.1 Addition 19
3.2 Subtraction ZA
3.3. Mixture of Addition and Subtraction Za

Exercises 3(a) — 3(c) 23

CHAPTER 4 MULTIPLICATION 25-48

4.1 Multiplication Using Base 10 and it Powers Ds
(a) When the multiplicands are both below a base 2D
 Page No.
(b) When the multiplicands are both above a base Zt
(c) When one of the numbers is below and the other
above base 10 and its powers. oe

4.2 Miultiplication of more than Two Numbers j 28

4.3 Multiplication when Numbers are not near the ‘Base 10, 100 etc. 30
4.4 Multiplication when Numbers have Different Bases 33
4.5 Method using Vertical and Cross-Wise Multiplication (VCP Method) 34
4.6 Binary Method 38
4.7 Some Special Cases of Multiplication 38
4.8 Multiplication when One of the Numbers is a Series of the Same Digit 41
4.9 Summary of Multiplication Methods 47
Exercises 4(a) - 4(i) 48

CHAPTER 5. SQUARES, CUBES AND HIGHER POWERS 49-60

5.1 Squaring of Numbers ending in 5 49
5.2. Squaring of Numbers ending in a Digit other than 5 49
5.3. A Combination of Methods 52
5.4 Another Method of Squaring ays
Exercises 5(a) - 5(c) 54
5.5 Sums and Differences of Squares 54
(a) To express the square of a number as the sum of
two squared numbers 54
(b) To express a number as the difference of two :
squared numbers 5D
5.6 Cubing of Numbers 56
(a) Method 1 56
(b) Method 2 7
5.7 Raising to Fourth and Higher Powers 59
Exercises 5(d) - 5(e) 60
CHAPTER 6. DIVISION 61-86
6.1 A Simple Method 61
6.2 Method of Complements ~ 62
6.3. Method of Cumulation 63
6.4 Division when Divisors are Different from 9’s
but are Close to the Powers of 10 65
6.5 The Multiples Method 67
Exercises 6(a) - 6(b) 68
6.6 “Change of Sign” Method 68
 Page No.
6.7 Combination of Methods 10:
Exercise 6(c) 74
6.8 Vertical and Cross-Product Method ft)
A variation using vinculum Vat
Exercises 6(d) - 6(2) 86
CHAPTER 7. DIVISIBILITY 87-98
7.1. Method of Determining Positive Osculators 87
7.2 Method of Osculation Using Positive Osculators 88
(a) Chain method 88
(b) Another method 89
7.3 Method of Finding Negative Osculators 91
7.4 Method of Negative Osculation D2
7.5 Group Osculation 94
7.6 Types of Divisors in Group Osculation and
their Corresponding Osculators US
(a) Osculation of the first type aS
(b) Osculation of the second type 96
(c) Osculation of the third type 97
Exercise 7 98
CHAPTER 8. FRACTIONS, PERCENTAGES AND RECURRING
DECIMALS 99-128
8.1 Three Types of Decimals 29
8.2 An Easy Method of Converting Fractions to Decimals 100
8.3. An Easy Method of Converting Fractions to Percentages 101
8.4 Multiplication Method for Converting Fractions to
Recurring Decimals 104
(a) Using positive multipliers 104
(b) Using negative multipliers 110
8.5 Division Method for Converting Fractions to
Recurring Decimals EZ
(a) Using positive divisors ig
(b) Using negative divisors 114
8.6 Remainder Digit Method 115
8.7 Geometric Progression Method 118
(a) Alternative G.P. Method 120
(b) Using vinculums |
8.8 Arithmetic Progression Method 122
 Page No.
8.9 Determining Multiples of Fractions 124
(a) Method 1 124
(b) Method 2 124
(c) Method 3 |
aR)
(d) Method 4 ‘ 25
8.10 Conversion of Recurring Decimals to Fractions a

CHAPTER 9. AUXILARY FRACTIONS 129-137

9.1 Type I: Denominator of the Fraction Ending in 9’s ee!
9.2 Type H: Denominators of Fractions Ending in 1’s 130
9.3 Type lil: General 134
Exercise 9 137

CHAPTER 10. EXTRACTION OF SQUARE ROOTS 138-153

10.1 Some Properties of Square Roots 138
10.2 Duplex and a Method of Finding It 139
10.3 Method of Extracting Square Roots 141
10.4 A Simpler Method Using Multipliers 148
10.5 Method Based on Regrouping 150
10.6 A Thumb-Rule for Finding if a Number is a Perfect Square tS2
Exercise 10 £53

CHAPTER 11. EXTRACTION OF CUBE ROOTS 154-167

eat Some Properties of Cube Roots 154
1h Method of Finding Cube Roots of Exact Cubes 195
(a) An alternative method 158
(b) Another use of the alternative method [59
Es Method of Finding the Cube Root of Any Number 164
Exercises: 11(a) - 11(b) 167
ANSWERS 169-172
APPENDIX 1 173
APPENDIX 2 174-175
 CHAPTER 1

MARVELS AND MYSTERIES OF NUMBERS

The ordinary numbers in arithmetic with which we are all familiar from the school
days are expressed in terms of powers of 10.
Examples:
S29 10") 42 10") = 6 x 10) FO « 1) =30 42:
3417, = (3.x 10°) + (4x 109.4 ix 10') +7 « 10°
= (3 x 1000)+(4 x 100)+(1x 10) +(7 « 1)
3000 + 400 +1 +7.
This method of representing numbers is called the decimal system and is based on
the deanery scale. The number 10 is called the radix. The ten numbers belonging to the
system are the digits 0 to9.
(The other common scale, which has become vogue after the advent of computers is
the binary scale whose radix is 2 and has 2 digits: 0 and 1. In fact, numbers can be
constructed to any scale with different radixes. Thus it is possible to deal with numbers
having radix 3,4,5,.........., 12, 16, and so on. The ordinary operations of arithmetic can be
performed in any scale taking into account the radixes involved). However in the present
treatise, we shall be dealing with the decimal system of numbers only.
It is important to recognize that every digit of a number has two values viz. a digit
value and a place value; for example in 32, the number 2 has its digit value 2 and the
place value, unity and number 3 has its digit value 3 and place value 10 so that (3 x 10) +
(2 x 1) =30+2=32. For this reason 32 is not the same as 23 although the same numbers
2 and 3 are involved. Their place values are now changed: 3 has unit’s place and 2 has
Mesmiace and gives (2.x 10) (3.x 1)=20 + 3.= 23.
Similarly in the number 3417, the digit 7 has the unit’s place, 1 the 10th, 4, the
100th and 3 the 1000th making it different from, say 7143.
With this background, let us consider some of the amazing features of numbers and
operations with the same.
1.1. Digital Root of a Number
We start with the digital root of a number. If we continue to add the digits of a
number until only a single digit is left, then that digit is called the digita! root (d.r.) of the
number.
2 SPEED ARITHMETIC

Eg 2351068~ adding 2+ 3-1 5S 1+ 0 6578 255 2

therefore 7 is the digital root. ;
A quicker way of finding a digital root is by the method of “casting out the nines”.

Eg. 7328614. Here we find 7+2=9, 6+3=9and 8+ 1 =9; therefore

casting these digits viz. 7,2,6,3,8 and 1 , we are left with only 4
which is the digital root of the given number.
Note:
(1) If, on adding the digits of a number, we end up in 0 , this may be taken as equivalent
to the digit 9 written as 0 ~ 9.
(2) Every number has a unique digital root whereas a given digital root may be
associated with more than one number. For example, 3254 has the digital root 5
whereas the digital root 5 may have the numbers 32, 41,536,2327 and so on.
Use of digital roots
When numbers are added, subtracted, multiplied or divided, the correctness of the
answers obtained from such operations may be checked with the help of the digital roots.
) Addition
Number Digital root
1] ) The digital root of the sum 4444 of the
246 3 numbers is 7 which is also the digital
3078 5 root of the sum of the drgital roots of
1 SOUS 6 the summands.
4444 16

(i) Subtraction
Number Digital root
6392 2 The subtracted result 3989 has the dr = 2.
-2403 -9 (x 0) Noting that the equivalent of 9 is zero,
3989 2 the subtraction of the digital roots of the
two numbers is also 2.
(iii) Multiplication
Number Digital root
6347 2 The d.r. of the product of the numbers
x 206 % Ses is 7, same as the d.r. of the product of
1307482 16 the respective d.r.s.
 MARVELS AND MYSTERIESOF NUMBERS >)

(iv) Division
In this case, the remainder of a number divided by 9 is the same as the digital root of
the dividend.
Eg. (i) 371427 divided by 9 gives the remainder 6, the same as the digital root of the
dividend 371427.
(ii) 4356 divided by 9 leaves the remainder 0 ( ~ 9) and the digital root of 4356
is also 9.
Thus the correctness of the division of a number by 9 too can be established with
the help of the digital roots.

In the case of division, in general, the digital root of the divisor multiplied by that
of the quotient added to the digital root of the remainder is equal to the digital root of the
dividend.
Fg. 6432 + 23 gives 279 as the quotient and 15 as the remainder. The product (d.r. of
the divisor multiplied by d.r. of the quotient) + the d.r. of the remainder equals (5 x 9) +
6 = 51 whose d.r. is 6, the same as the d.r. of the dividend 6432.

1.2. Marvels of Some Numbers

(i) Take a number M. Form another number N by altering the digits of M in any
manner. Then the difference (M ~ N ) is divisible by 9. (The result, in fact, holds for
numbers of any system like binary etc.)
bes (1) “M = 203
N = 032

Difference M ~N = 171 which is divisible by 9 (the digital root of 171)

(2). Meee 2576

Nee 7256
Difference M ~N = 4680 is divisible by 9, the digital root of 4680
(il) A number formed by 3 digits and repeated in the same order once more is divisible
by 7, 1 f and 13.
Hes 150150, 167167
(iii) Let an odd number of digits, n, of a number be repeated once more. Then the
number so formed is divisible by 9090 ........... 9091 containing (n-1) digits.
Eg. (i) Take n= 3. Consider the number 254254. This is divisible
by 91 (containing n - | = 2 digits).
 SPEED ARITHMETIC

(ii) Let n= 5 and the number 35021 35021. This is divisible by

9091 (containing n- | = 4 digits).
(iv) The sum of all n digit numbers, (n>2), equals 494 99---9 55 00----0.
Sas Lpeeuaes
‘(n-3) (n-2)
For example, the sum of all 3-digit numbers is 494550 and the sum of all 5-digit
numbers is 4949955000.

(v) If one of the two numbers 2"-1 and 2°+1, (n>2), is prime, then the other number is
composite. For n=2, however, 2" -1 = 3 and 2" + 1 = 5, both being primes.

(vi) The sum of the cubes of consecutive integers is divisible by the sum of the
integers; for example, (4° + 5° + 6°) is divisible by (4+5+6).
(vil) If k is odd, then (1* + 2" + ...... + n‘) is divisible by (1 + 2 +...... +n), where

(vill) Any number in the sequence 49, 4489, 444889, 44448889......... is got by
inserting 48 in the middle of the just preceding number. For example, 4489 is
obtained by placing 48 between 4 and 9 of the preceding number 49. Then, each
number in the sequence is the square of an integer, namely 7,67,667, 6667.

1.3. Binary Method of Multiplying Two or More Numbers

Eg. 41 x 57
Step (1) ALES 7:
Divide the Left Hand Side (LHS) number (41) 20 x 114
by 2, omitting fractions and multiply the Right 10 x 228
Hand Side (RHS) number (57) by 2. Continue 5 x 456
till the LHS equals 1. ek OT?
1 x 1824
Step (11)
Delete all even numbers on the LHS and the 4] x 57
corresponding numbers on the RHS and retain the 5 x 456
remaining ones. : 1 x 824
Step (111)
Add the RHS of step (ii) viz. 57 + 456 + 1824 This equals 2337 which
is the required product.
 MARVELS AND MYSTERIES OF NUMBERS 5

Eg. 74 x 68
Step (1) 74 x 68
37136
1Six 272
9 x 544
4 x 1088
22 TO
1 x 4352
Step (ii) 37 x 136
9 x 544
1 x 4352
Step (iii) 136 + 544 + 4352 = 5032
Logic of the method
Consider the RHS numbers in step (ii) of eg. (1) above
Si =e? 4 of isi.
456 =2? x 57
[8040-2 92 pd 57
Sum: 2337 = (2° 42°42?
) x57-=¢:1--84-32)>57-=41-« 57
Next consider the RHS numbers in step (ii) of eg. (2) above
136 = 2' x 68
544 = 2? x 68
4352 = 2° x 68
Sum 5032 =(2+2?+2°)
x 68 = (2+8+64)
x68 = 74x68
1.4. Prime Numbers :
A number which is not exactly divisible by any number other than unity and itself is
called a prime number, eg. 37. The number of primes is infinity.
A number which is divisible by number (s) besides unity and itself is called a
composite number, eg. 54 since it is divisible by 2,3,6,9,18 and 27 (apart from | and 54)
These are called the factors of the number 54. A number having factors (other than unity
and itself) is called a composite number. Numbers which may be primes or composites
themselves and having no common factor (except unity) are said to be prime to one
another: egs. 22 and 35 or 23 and 64.
If we call a group of ten successive numbers, a decade, the first decade being 1-10,
the second 11-20 and so on, we find that the number of primes in each decade is
different, the maximum number in any decade being not more than 4. For example, in the
6 SPEED ARITHMETIC

decade 11-20, there are 4 prime numbers viz. 11,13 17 and 19. In the first decade 1-10
however, there are 5 primes 1,2,3,5,7 but this is because 2 (an even number) is considered
a prime. In fact, 2 is the only even number amongst primes. The maximum number of
primes in a decade (except the first) cannot exceed 4, since out of the 10 numbers in a
decade, there are 4 even numbers, one number ending in 5 and one in 0.
So long as a number is small, one can easily check, by means of factorisation,
whether it is prime or not but for a larger number, the task of finding factors may become
difficult.
While there are possibly no formal methods to identify if a given number
(particularly a large one) is prime or not, we describe two working rules below:-
(i) One method is the grid - method which has come to be known as “the sieve of
Eratosthanes, a Greek mathematician who lived in 276 - 194 B.C. This consists of a
10x10 grid and writing in the numbers | -100 in the 100 cells thus formed. We then start
from 2, cross out every second number excepting 2, then proceed to 3 and cross out every
third number excepting 3 and so on till we reach the end of the grid. The numbers that
remain uncrossed are the primes. We illustrate this with the help of a 8 x 8 grid.

This method may be extended to any grid - size. A similar method was in practice
by Indian mathematicians who used the aphorism “elimination and retention” However,
this method is of limited practical use.
(2) Another method to check if a given large number is prime or not is to take its
square root and see if any of the primes less than or equal to the integral part of the square
root divides the given number. If so, the number is not prime and if not, it is prime. E.g.
let us take the number 237 whose integral part of the square root is 15. The prime
 MARVELS AND MYSTERIES OF NUMBERS

numbers less than 15 are 2,3 5, 7, 11 and 13, of which 3 divides the number and therefore
237 is not a prime number. Let us now take 577. The integral part of its square root is 24
and therefore we consider the primes 2,3, 5,7, 11, 13, 17,19 and 23 none of which divides
577 exactly. Hence 577 is a prime number.
1.5. Integral Powers of Numbers
Let us construct the following table, the body of which gives the ending-digit when
a number ( shown on the extreme left) is raised to the power given at the top row.
Table 1

pared

( The ending digits for other decades 11 to 20, 21 to 30....... are the same as for the
above decade. )
The first thing to note is that the powers of odd numbers are all odd and of even
numbers, all even.

We summarise other main features of the table as follows:

Column-wise
(i) In the square column, the see digits (for the numbers | - 9 ) bear a symmetry
around five viz., 1,4,9,6, 5 ,6,9,4,1.
(11) In the cube column, ilsi digits 1- 9 appear in the order 1,8,7,4,5,6,3,2,9
corresponding to the numbers | to 9. The extreme pairs (1,9) , (8,2), (7,3), and (4.6)
are such that one digit in each pair is the complement of 10 of the other digit in the
same pair; Eg., 9 is the complement of 1 in (1,9) and so on.....
(iii) For the 4th power, the endings are 1,6,1,6, 5 , 6,1,6,1 so that there are only two
digits involved viz., 1 and 6 which bear symmetry on either side of five.
8 SPEED ARITHMETIC

(iv) The endings in respect of Sth and 9th powers follow the sequence 1,2,3,4,5,6,7,8,9
the same as the ending digits of the respective numbers.
(v) The columns under 6th, 7th, 8th and 9th powers aree exactly the same as those under
2nd, 3rd, 4th and Sth respectively.
Row-wise
(i) The ending digits of numbers 1, 5, 6 and 10 for all powers are 1, 5, 6 and 0.

(ii) Ending digits of number 4 for all powers repeat in pairs (6,4) and of number 9 in
pairs (1,9) and again , one digit in the pair being the complement of the other.
1.6. The Triples : 3, 6, 9.
(a) Addition
(i) The addition of the 3 numbers 3 + 6 + 9 = 18 whose digital root 9.
(ii) The additions in pairs give rise to digital roots 3 or 6 or 9 as below.
ce i ee coc
34+6=9 6+6=3 9+6=6
3+9=3 64+9=6 94+9=9
(b) Subtraction:
The subtracted value of any unequal pair out of the triples 3,6,9 has a digital root
which is either 3 or 6.
6-3-3
9-3" ="6
P26 t= 93 :
(c) Multiplication
The digital roots of multiplication of any two numbers amongst the triples are
shown below.
3x1 =3 6x1=6 9x1=9
3x2 =6 6x2=3 9x2=9
3x3 =9 6x3=9 9x3=9
3x4 =3 6x4=6 9x4=9
3x5 =6 6x5=3 9x5=9
3x 6=9 6x 6=9 9x6=9

(3,6,9) repeat (6,3,9) repeat = ( 9 throughout)

Also 3 x 6 x 9 = 162 whose digital root is 9
 MARVELS AND MYSTERIES OF NUMBERS 9

Mysteries of Number 9
Sums of the 9 consecutive numbers | +2 +.....+.9, 11 +12 +....4+19,
21 +22 +... +29, etc. have the same digital root 9.
The product of any number multiplied by 9 has the digital root 9.
The digital root of the addition of any number with 9 is the same as the number
itself (if itis a single number ) or the digital root of the number if it has 2 or more
digits.
4+9 =13 (d.sr.4, same as the number 4), 16+9 = 25 (d.r.=7, same as the dr.
OT. bO.):
230+9 = 239 (dr. = 5, same as that of 230) and so on.
Similarly if any number is added to 99 or 999 etc. the digital root of the sum will be
the digital root of the number.

(4) Consider the vulgar fraction 1/7 whose decimal is recurring viz.
0.142857 142857......; we find that the digital root of the cycle 142857 is 9.
Similarly, for the fraction 1/17 which equals the recurring decimal
0.05882352117647, the digital root is again 9.
The recurring decimals of other fractions like 1/13, 1/17, 1/19, 1/23, 1/29, 1/47,
1/49 also have the same feature. Other fascinating mysteries of 9 are brought out in
later chapters.
1.8. Divisibility Tests of Numbers:
Divisible by When
Z a number is even or has 0 at its end
3 the sum of the digits of anumber is divisible by 3
4 the last two digits are divisible by 4
5 a number ends in 5 or 0
6 a number is even and divisible by 3
8 the last 3 digits of a number are divisible by 8
) the sum of the digits of a number is divisible by 9
10 a number ends in 0
1] the difference between
(1) the sum of the Ist , 3rd ,5th digit .....and
(11) the sum of the 2nd, 4th ,6th digits.....
isa multiple of 11 or 0.
We examine the question of divisibility very extensively in chapter 6.
10 SPEED ARITHMETIC

1.9 Reciprocals of Some Prime Numbers

While we shall discuss these numbers in considerable depth in a later chapter, we
shall highlight here some of the simple and interesting PEOpoTues with the help of the
fraction 1/7.
(i) Its decimal equivalent is 0.142857 142857, ... , and the group of 6 digits 142857,
repeats itself in that order any number of times.
(ii) The cycle 142857 has 6 digits, the same as the difference between the denominator
and the numerator of the fraction.

(iii) If one multiplies the group of digits 142857 by any number from | to 6, the product
has the same set of digits which follow the same order except that the starting digit
is different. Thus

1 x 142857 = 142857
2 x 142857 = 285714
3 x 142857 = 428571
4 x 142857 = 571428
5 x 142857 = 714285
6 x 142857 = 857142
and finally 7 x 142857 II= 999999

Thus, if one remembers the sequence of digits 142857 in that order, one can then
shoot off the answer to a question like: what is the value of 3/7? Since the first two digits
of the sequence 142857 are 14, multiplying the same by 3 gives 42, so that we know that
the starting digits of the answer are 42 and we reel off the answer as 428571 (maintaining
the order of the digits in the sequence 142857). Similarly, when asked the value of 5/7, we
multiply 14 by 5 to get 70 and the nearest double digit in the sequence is 71 which
therefore would be the starting digits for 5/7. Thus the answer would be 714285 (again
retaining the same cyclical order of digits in the group).

(iv) It is also of interest and help (as we shall see later) to know that the first 3 digits
viz., 142 and the second 3 EUS 857 Re add up to 999.
 11

CHAPTER 2

COMPLEMENT AND VINCULUM

In the chapters that follow, we discuss and illustrate various techniques that speed
up the operations such as multiplication, division, conversions of fractions to decimals
etc. on numbers. We shall however, begin with two important aids viz. complement
and vinculum which play a major role in later chapters in quickening the arithmetic
operations even further. It is therefore in the interest of the readers to be well conversant
with the handling of these aids.

2.1. Complement
We discuss how a complement of a number is found under two situations:
(1) When the last digit (.e. the unit’s digit ) is non zero
In this case, we subtract all the digits, excepting the last one, from 9 and the last
digit from 10.
Fig, “7022 927 a2
9 a0iges9
10-2= 8
The complement of 702 = 298

(2) When the last digits end in zeros

Here, a slight variation is made namely, the last non - zero digit is taken as the last
digit from which 10 is taken away and as many zeros are added to the right of it as there
are zeros in the given number.

Eg. 81900; 9-8 = 1

9-1 =8
10-9.=1
Add two zeros ( as in the number) at the end.
.. complement = 18100

Eg. 9050; 9-9 =0

9-0 =9
10-5 =5
Add one zero at the end (as in the given number )
*. complement = 950
 SPEED ARITHMETIC

It will be seen from the above that the sum of a number and its complement has
unity at the extreme left followed by as many zeros as the number of digits in the original
number.

2.2. Vinculum
A vinculum is a very handy tool as the user of the book will soon realise as he
advances to later chapters. He should however be cautioned, that extreme care and a large
amount of practice is needed, especially by a beginner, since it can easily lead to wrong
answers if not handled carefully.

In algebra, a vinculum is a line or a bar drawn over several terms of an expression to

indicate that they have a certain common relation to what follows or precedes. (in fact, a
vinculum means a chain or a bond).

Eg. a+bxc = ac+be or a-b+c = a-b-c.

In arithmetic, it represents a minus sign or a deficiency.

Eg. -342 = - 3(100) - 4(10) - 2(1) and in vinculum notation this is written as Br 4 3)
i.e. all the digits carry a negative (-ve) sign before it. But suppose we wish to attacha .
minus connotation to one or two digits only. This also can be easily handled by using the
vinculums ; in fact, the advantage of vinculums lies here.

fg. (1) 3.42: This isa short-hand for

-3(100) + 4(10) + 2(1) = -300 + 40 + 2 = -258
(Pe pee) 42: This is equivalent to
3(100) - 410) +2 (1) = 300-40 +2 = 262
(3) 34 2 = 3(100) + 410) -2(1) = 300+ 40-2 = 338
(4) 3 42 = -3(100)-4(10)+2 = -300 -40 +2 = -338
(5) 34 2 = -3(100)+4(10) -2(1) = -300+40-2 = -262
(6) 342 = 3(100) -4(10)-2(1) = 300-40-2 = 258
It follows that the negative 4(10) of a number consisting of any number of digits is
the same as placing vinculum over every digit of that number.

Eg -6732= 6 73 2,
Note. A vinculum over the digit 0 is inconsequential. Thus
4 0 56 = 4(1000) - 0(106) - 5(10) + 6 = 4000 - 50 + 6 = 3956 and
40 56 = 4(1000) + 0(100) -5(10) + 6 = 4000 - 50 + 6 = 3956
 COMPLEMENT AND VINCULUM 13

The above examples suggest the rules for connecting the digit(s) of a number to
their corresponding vinculum(s) and vice versa. We now consider them in detail.

(a) Conversion of digits of a number to their vinculums.

Take eg. (2) above : 262 =3 42
We see that if we wish to change the middle digit 6 to its vinculum, we write the
last digit as it is (since no conversion is involved), take the complement of the middle
digit 6 from 10 viz. 4 and place a bar over it and finally increase the first digit 2 to 3. This
gives the result.

The rule is : Take the complement of a digit for which vinculum is desired and increase
the digit immediately to its left by unity .

Let us have a few more examples.

Egs. (1) ’Convert 7 into a vinculum digit <7 = 07 ="1 3 GCheek: 7 = 10.73)
(The complement of 7 is 3 and we increase the preceding digit (0 in this case)
to 1).
(2) Convert the middle two digits of 3778 to their vinculums :
31 poe se 44 SS:
(3) Convert 9 and eS) of 49251 into their vinculums:
AIT ale 13 1S Sl
(4) Convert 0 and6 of 2063 to vinculums
2063 = 3 9 43
(5) Convert 0 and 3
of 2063 to vinculums
2063 = 207 7 (since ‘0’ is neutral to vinculum)
(6) Convert 10 of 5410 into their vinculums
5410 =55 90 (since ‘0’ is neutral to vinculum)
(7) Convert 7 of 7321 to vinculum
7321 = 13321 (since we can write 73 2 1 as 07 3 2 1 so that the
vinculum on 7 is the complement of 7 from 10 and we increase its preceding
digit 0 to L)

Finally, a given number can have many alternative vinculum representations. For
example,
1407 can be writen as 1413, 159 3, 2607, 259 3.
14 SPEED ARITHMETIC

(b) Conversion of vinculum digits of anumber_to non-vinculum digits.

In this case, the rule is to take the complement of the vinculum digit and decrease the
digit just preceding the vinculum digit by unity. ;
Consider one of the earlier examples, say 3 ey = 258.
Here, the complements of the vinculum digits 4 2 are 5 8 and we decrease the
preceding (non-vinculum) digit 3 to2 to get3 4 2=258.

Egs. Convert the vinculum digits into non-vinculum ones in the following numbers.
(teas. 68
a a6
GB)\it-3.5 Se GSB 015 3
‘c) Removing a vinculum or shiftingof a vinculum.
We shall illustrate the procedure by means of a variety of examples.
Egs. (1) 3 = 17;since 3 = 03 and the complement of 3 is 7 and
decreasing 0 by | gives - 1 1.e., a
(2) 753 =e43162
(3) FS Das 8 ame 3810882
(4) 9804= 11816
6) B202004%=" iy t 016
(6) 3 5 6 2= 1643

Exercises
2(a). Convert into vinculums the digits indicated in the brackets against each number
CL) 71598: 7,9)
(2) 931028): ©:3,0;8)
(3) 259101: (9, both 1’s)
(4) 8670: (All the digits)
(3): -91865': (9.7,8;))

2(b). Convert vinculums into ordinary numbers.

(1) 10708
(Qeoele9.03
3) 56001
(4) 1953602
(5) 106646024
 COMPLEMENT AND VINCULUM 15

2.3. Addition, Subtraction, Multiplication and Division of Vinculum Numbers

The operations of addition, subtraction, multiplication and division can be carried
out on vinculum numbers just as on the ordinary numbers (except that extreme care must
be taken to the minus character of the vinculum digits). We illustrate these operations
with the help of a few examples below. (Remember that vinculum on zero has no effect).

(a) Addition
Egs. (1) 6+4 =2; 6+ 4= 1 0= 10
(2) 016 424 6 {= 22
By 1547 =547=1
Qo 54 7 =54 7=
Stet 7 =o. 8
(6) 549467 =5494+53=602
(7) 5494 67 =549+4+(60)+7=496
(8) 549467 =549 +(-67)=482
(9) 5 494+67=469
+67 = 536
(M0 549 467 =-500b49 +67="41 6
(11) 54 94+67=4514+67=518
(12) 549+67 = 536
(13) 549467 =-500+49 +53 = -500+102= -398= 402
(14) 54 9+67= 40 2= 598
(15) 549+67= 53 6= 664
16 1*4 94+ 6 7 = 6162 «784
G7). 70 92142 903-=69121+110 3 370204

(b) Subtraction
Bes. 0) 26-2. =.8 (6-2=-8.= 8)
s4iGaGet2e-6+2=-4=-A4)
= 8 (2-(6)=2+6=8)
(4) 17-6 =23 (17-(-6)=17+6=23)
6 9 ((10-7) -(-6) =3+6=9)
pat 29(-17-(-6)=-17+6=-ll=1 1= 29)
9 -4 0 81 =3921 - 3921 =0
16 SPEED ARITHMETIC

8 = 122)
9- 14 207= 2941- 13807 = (02941- 13807) = 9134)

(c) Combination of addition and subtraction.

Egs. (1) 42-63+ 160 = 42+ 63+ 160= 141
(2) 7+ 1007 -2654+1 63 = 7+ 1007+ 2654+ 1 63 = 1 603
(3) "324:567 101 2 45213 = 324-0 5-6 7410 A2Ge21S de OD
+

(d) Multiplication
Egs. (1) 7x6= 42= 58 (-7x6 =-42 = 4 2)
OV I es eu
8x l= 8
2A
Gy 32 Deer ae
oes
5.6
64
896 = 904

(4) 122x59=122 ox 32 =18 1+ B3E2 Bl) £0

x 59 9x2= 18 5x2 = 10
TOs 9xl= 9 5x1 =5
Gd fF A144 lt Os-—=-es-2
C4180 08 ee
Big 2b: TOTS 6 10
5178 = 4838

(5) “022%. 0:= 18s 7

78 9x8= 72 £5x8= 40
~ 59 9x7= 63 5x7 = 35
a eG aces cerrone ©gens ee
3.90 7 ay 390
 COMPLEMENT AND VINCULUM 17

Whenever we have digits greater than 5, we may employ vinculums to render them
small, perform the operation required and then reconvert the result into the original form.
Let us illustrate this procedure for the multiplication.

(6) 94x87=114x93=114x113

(e) Division :
Fos. (1) 1243 =-12+3=-4= 4
Q) 26+7-— = 20-6=14+7=2
(3) 80 3+13 =-803 +13 >-61 is the quotient (Q) and -10 is the remainder (R)
i.e. 79 is the quotient and 10 is the remainder. The same division can be carried out
without converting the vinculum numbers into corresponding negatives as follows.

13 9 18kO)FCF 79 BOL onh = 6049111" =

of fr Serre 11 34.1 17e* To
(i3:
= ares
10
18 SPEED ARITHMETIC

As acheck, we have: dividend = (divisor x Q)+R 79

= (13x 79)+10 13
= Sorpayio Poery
= 817 . 19
= 803 | $07
(4). 9621 4=5 (904 6) (10-4) = - 84+ 6= -14= 14 = 26
(5) 32922 7 23 = 0024290) +(- 720 +3)
= (2712 +-717) > Q=-3 = 3, R=56l
(6) 2000 6+ 300 2= -20006 + -3002 > Q=6,R=-19944=1994
()oPI-89°70 2+4 1 4 2=28beRh 3958 > O a7, R= 1292.

To summarise, the main advantage in using vinculums lies in the fact that digits
greater than five can be converted to smaller ones by taking the vinculums of the former.
The operations on these vinculum digits are far more simpler and quicker since they
would involve digits that are smaller than five.

Exercises
2(c). Addition
(1), 8-6 44 327 CSS: 2 eo eae
(3) 760 8+100 9 2 (4) 100 6 2+ 9 938
(5), 8355 0.40: 7 50 (6)45-38 645226
2(d). Subtraction .
(W603 - D486 (2) 226-39 8 ele ee ae ee
(AyiGc854= 5040 - w(5) dr 44 ed (6)38 6-12 5
2(e) Addition and Subtraction
(1) 6 3 = 2 LS 1 ae (2) 1004 + 1022 Lat
(3) 9694+7- 3.4301 =30 (4) 8 1 70-3 20+ 100-99+ 101 35
(5) 2345 - 1131-1 3 1+99 - 206 7
2(f). Multiplication
(1) 86x 24 (2) 385x206 3)836x164
(4) 10 8x90 5 (5)~2009 6x5 3 2
2(g). Division (Strokes stand for division +)
(1) 6 26/1 8 (2) 32 2/ 3 (3) 10 4 8/23
(4) 10 5 5/0 5 (5) 3 43 4/2 7
 CHAPTER 3

ADDITION AND SUBTRACTION

Having studied the complements and vinculums in the previous chapter which
are essential to the understanding of this and later chapters ,we now start with the two
most basic operations of arithmetic viz . addition and subtraction of numbers.

3.1. Addition

In the conventional addition taught presently in schools and practiced in

colleges, we start the addition from the extreme right of the numbers to be added i.e. with
the unit digits of the numbers and then keep moving to the left taking the “carry-overs”
obtained in the unit’s digits to the ten’s digits and then add the “carry-over” along with
the numbers in the tens-digits-column, then taking the “carry-over” obtained here to the
hundredth columns and add the carry-over and all the numbers in this column and so on
until we have exhausted all the columns of the addition.
In the above method, the carry-overs can be very large particularly if the addition
involves too many rows with large numbers and therefore can lead to error in the final
answer.
We now present a method in which’ “carry-overs” as practiced in the
conventional method can be avoided. The method proposed here also offers other
advantages: (1) each column can be handled individually either from right to left or vice
versa and (ii) we need not add numbers which are more than 10 in any column, thus
making the addition easy, simple, quick and also much more error free.
Let us now present some examples in which addition is made using the proposed
method.

Eg. (1) Adding the units digits, we have 2+9=11:

nes
(11) Since it exceeds 10, we continue the addition by taking rsa
only the | in the unit’s digit. To remember that we 346 2
have omitted the other one in the ten’s digitof 11, we Total 3 5 2 4
place a_ star (*) alongside 9 as shown.

(iii) | Adding the one in the unit’s digit of 11 to 3 we have 1 + 3 = 4 and write it
in the last row namely the total row.
20 SPEED ARITHMETIC

(iv) We now proceed to the ten’s column: 6 + 4 = 10 and take only O for
adding the 2 in that column which gives 0.+ 2 = 2. We write the 2 in the
total and place a star along side 4.

(Vv) Adding the numbers in the hundred’s‘column, we have 4 + | = 5

which has no carry-over. Therefore we place the 5 in the hundred’s
column of the total with no stars on any number of this column since there
are no Catry-overs.

(v1) Finally, we have 3 in the last column and place it in the total.

(vil) We next, count the number of stars in each column. We have one star each
from the unit’s and the ten’s columns.

(Vill) We place these counts in either of the two ways as below:

(a)? “Potal'’? °3° Qe (b) Total : 3 yi2 Wied
No. of stars : fo or No. of stars : ay

and add to get 3634 as the final total. Since the carry over |’s obtained in
the addition of the columns really belong to the ten’s and hundred’s digit
column, we place them as above before the final addition.

It can be easily verified that we get the same results as above even if we start from left
and move towards the right t.e. from the thousand’s column to the unit’s column.

Eg. i
Ae)3*
Gy lt 2
Pero hae a @)
Oe Oe
Siac le
ft TRG AS
Total: 1o0a.-9 9 or 9/ 9
No ofstars: leeSel ee

3) 2 a2 Liles? ini vedehign

 ADDITION AND SUBTRACTION OM

Eg. SAG ee Ee.

2 OLS Rs. Es:
6 al: tiowat Pee ea a, | ake
Sor | ETE Gis
7*0 “6* 9 9*6. 470
13. 2* +2212 2.7 36.9
pre eee eS Total: EO a) ees
Total: Se O—3. 3. | No. of stars: ‘bay et eee
No. of stars: fas eae: Sa Cis a eae
= ee) ay eee: |

Eg. Meters Millimeters Note: The row “total” has a zero in the 3rd
Sa | column from right. Since 1 meter equals
ffs 52 1000 millimeters, the entries under
A¥O 1 Ae 3* millimeters have three columns like 007.
40 OF 052, 043, 079 and 008.Once we keep this
+1704 8 in mind in this and other similar
Total: ©14-79 069 examples, we can delete the 0’s in the left
No. of stars> 1 1 le of the significant digits under the
2189 18 9 millimeter and proceed as shown in the
example so as to save considerable time.
3.2. Subtraction
We now present a method of subtraction which again dispenses with the
“carry-overs”’ as learnt in the traditional method.
In doing so, we employ vinculums which we elaborated in the previous chapter.
The procedure involved is very simple and quick without having to deal with numbers
larger than 9. Let us explain by means of a few examples.
Lg,
(i) Whenever a digit in the first row is larger than the
corresponding digit in the second row, subtract the be: 5 bh
two digits as usual. Thus we have 7-6=1. - 986
(11) Whenever a digit in the top row is lesser than the 1631 =371
corresponding digit in the second row, subtract the
smaller digit from the larger one and place a vinculum on the subtracted digit.
Thus, we have for the ten’s digit, 5 in the top row and 8 in the second row,
so that we have5-—8= 3.
2 SPEED ARITHMETIC

(iii) | Follow the same method for the other columns.

(iv) | Now convert all the vinculums in the final row into regular numbers using the
rule of conversion stated in chapter 2 to get the final answer.

Eg. AO pelt oe 0 Eg OD B*7 OLD

racine. 6 9 re ss p 9 3.4
ha Peet. 88 | 1-9-3
T= 77

Eg. Kilograms Grams

370 ay)
eet. 270
Since | kilogram = 1000 grams, this is equivalent to writing
Kgms Grams
3 S10 OM 9
ee ag aa)
3 0.0. SiaMRS)
RRO 1449
It will be seen that, as in the case of addition earlier, subtraction under the present
method can too be effected either from left to right or from right to left.

3.3. Mixture of Addition and Subtraction

Very often, we frequent the situation of a series of entries some of which are to be
added and others to be subtracted to arrive at a final answer. In such a case, we separate
all the entries to be added to form one group from those to be subtracted which forms
another group; and we subtract the sum of the second group from that of the first group.
In carrying out these additions and subtractions, we follow the same methods as in
sections | and 2 above.

Eg. + (-73
07
- LS
+ 1392
 ADDITION AND SUBTRACTION iB

(1) Separating the entries with + signs and — signs, we have

1s) = 207
Pek Oo 2. <2 gol5
+e fli) = 24.99

(ii) | Adding the positive entries in the first group gives

aes
pegs Oe
| Aes |

Total: tats 3G
No. of stars: l

loaves 6

(ii) Adding the negative entries, the second group gives

PT
i 5%
ATo
Total: 6 0]
No. of stars: baz
y #2

(iv) We now subtract 721 from 1586

1 8876
2 l
l 6 5 = 865

(v) The final answer is thus 865.

Exercises:

3(a). Add the following entries in each sum using the method of this chapter.
(1) 61+ 2+ 349 +76 +3453+ 56+3 (2)32714+271+71+4 109
(3) 2.39 + 171.671 + 81 .8823 + 222.02 + 99.9
24 SPEED ARITHMETIC

(4) Rs Ps. (5) Meters Cms

Wl 4] ld! 4°}
18 3 233 a
229 91 7 fe 99
Si 8 8 . 1433 10
27 80 17147 ]
345 1 a oat
92001 76 S113 7

3(b). Subtract using the vinculum method of this chapter.

(1) 1701 — 989 (2) 362.47 - 74.589 (3) 7651.92 - 398.08
(4) 18000 — 97.6254 .
(5) A train runs from station A to station B, a distance of 365 kms. and 9 meters.
If starting from A it has yet to cover a distance of 97 kms. and 375 meters to
reach B, how far has it run from A. (1 kilometer = 1000 meters).

3(C): Find the final sum of the following entries using the method of this chapter:
(1) 37- 200-429 + 1000 + 769 (2) - 1203 + 252 + 1139 — 527 —212
(3) 3 + 62 — 562 — 2200 + 3607 (4) 78 + 91.9+177.1 - 200.9 — 75.9
(5) — 212.61 — 1.117 + 4.0009 — 2171.37 + 2567.99
 20

CHAPTER 4

MULTIPLICATION

We now take up the operation of multiplication and discuss with illustrations

various non conventional methods for the same.

4.1. Multiplication Using Base 10 and its Powers

(a) When the multiplicands are both below a base
(i) Base 10
Eg. Multiply 9x9 (Of course, the students give the answer after having
committed to memory the multiplication table of 9. Weshall do the multiplication
without resorting to this) .
Being a single digit number we use the base 10.
Number 9 is one less than the base. 9-1]
So we write as x 9-1
2

(i) Cross subtract: 9-1 =8 (either of the two cross-

subtractions must give the same answer) and write
it on the extreme left and place a stroke after it. it
(11) Multiply vertically the two digits
(-1) x (-1) = 1 and place it after the stroke.
Now remove the stroke to get the answer 81.
All these steps can be performed mentally.
Eg. 6 = 6x6 (base 10)
6-4=2 6-4
(-4) (-4) = 16 6-44.
HO
Be is 6 =(2+1)/6 = 36

Since we are working to base 10 containing a single zero, we retain one digit of the RHS
of the stroke and carry over the other digit to be added to the digit on the LHS of the
stroke. .. we write 16 with 6 on the top line and | on the second line to be carried over
to the other side of the stroke. This gives (2+1) | 6 > 36 as the product.
Eg. 6 x 4 (base 10) 6-4
4-6
0/14 =(0+2)|4=24
26 SPEED ARITHMETIC

(ii) Base 100

Let us now consider the multiplication of double digits close to and below 100.

Eg. 91x95 (base 100):

(i) Since, 91 and 95 are 9 and 5 short of 91-9
100 respectively, we write them as SO ee
(ity 91-35 = 95-9 =86
Giii) 9X5 = 45 86/45 = 8645

In this case, we do not write 45 as 45. (as we did previously) Since we are working
to base 100 (which contains 2 zeros), we retain the two digits 45 as such on the RHS of
the stroke ( this is an easy way to remember as to how many digits should be retained on
the right and how many to carry over to the left).

Eg. 93° = 93 x93 (base 100): 2 ee

Eye es
86/49 = 8649

Eg. 88 x 89 (base 100) 88 - 12

89 - 11
771432 = 78/32 = 71832

(Here since 132 has 3 digits and only 2 digits are allowed on the RHS by virtue of the
base 100, the two digits 3 and 2 are retained on the RHS and the digit 1 is carried to the
left)

(iii) Base 1000

Let us now consider the product of numbers close to and below 1000.
Eg. 988x991: We write 988 - 12
991- 9
. 979/108 = 979108

Eg. 898 x 985: 898 - 102

985 - 15
883 / ;530 > 884 / 530 = 884530
 MULTIPLICATION a7

A comparison of this multiplication method with the conventional method will show how
easy and quick the above is without taxing the memory on multiplication tables.
Technically, this method can be extended to bases 10° F 10° etc.

(b) When the multiplicands are both above a base.

In the above examples, we have considered numbers less than a base. We can apply
the same procedure for numbers above the bases 10, 100, etc. also, but with a difference.
Let us illustrate:

Ee. 16X18: (Base 103 Be 105: x1 bly. (Basesl00)

16+6 105+ 5
18+8 111 +11
24/48 = 288 LiG,A oJ Ray aa O))

Eg. 118 x 109: (Base 100) Eg. 1012 x 1008: (Base 1000)
118+18: 1012+12
109+ 9 1008 +_ 8
SiO) = 12802 1020/096 = 1020096

On the RHS , when we get 96 (only 2 digits) we

adda zero to its left to make it have the three
digits 096 since we have a base of 1000 (of 3
Zeros).

Eg. 1068 x 1012: (Base 1000) Eg. 100097 (Base 10000)

1068 + 68 : 10009 + 9
1012+ 12 10009 +9
1080/816 = 1080816 10018/0081 = 100180081

(c) When one of the numbers is below and the other above base 10 and its powers.

Eg. 7 x 12 (base 10):

(i) Write the signs as shown to correspond
with the deficit and excess of the given
numbers from the base. ees
Gy 74+42=9 S71249 12+2
(iii) (-3) x (2) = (6)
9/-6 = (9x 10)-6 = 84
28 SPEED ARITHMETIC

Eg. 111 x 84 (base 100) : bikewolt

84 - 16
95/-176 = 95x 100- 176
roe = 9500-1=76 9324
We may adopt a vinculum for the -ve part on the RHS. This gives
95 Ta hute 94c7 209309324}
1
Eg. 988 x 1023 ( base 1000) : 988 - 12
1023" + 28
1011 /-276 = 1011 x 1000
- 276

= 1010724
or using vinculums as in the above example, we have
[Ot 26 = 1011 2-7 6.=101074.
4.2. Multiplication of More than Two numbers.
We have so far considered the products of two numbers, using various methods .
Let us now extend them to obtain the products of more than two numbers.
Bg, 9X [x5
(1) Write the numbers as shown with their 9-1]
deficits from 10 b= 3
Gi), orm (1) (3) 2) -=.-6 = 6 8-2
(111) Form (-1) (-3) + (-1) (-2) + (-3) (-2)
= 34246 = 14 4/il/ 6
QWs 8 Aa (ol aes 3) = as 6
(since the base is 10 for all the 1]
three numbers, we retain one = 516
digit in each stage) = 504
Eg. 103x97x92
(1) Write the numbers as shown with 103 +3
their deficits or excesses from 100 97 -3
Gp BiGs)es es 2 92-8
(iii) (3) (3) #(-3)-8)+3)(8) a
9S 24 dr Oe Ono 92/0 9/72
(iv) 103-3- 8=92 = 920 972
(since we have the base 100 , 919172
we retain two digits between each stroke)
 MULTIPLICATION

Eg. 8x 91 x 105
(1) Here , since 8 is closer to base 10 96-4
and the other 2 numbers are nearer 91-9
to 100, we multiply the former by 105+ 5
12 to get 96 which is now closer
to 100. We now proceed as before. 92/ 2 9/180
(ii) (-4) (-9) (5) = 180 = °2/D of80
(iii) AVA) ea) >) 9) (5)
=S0' 2045S “279 = 92 2 880
(iv) 96-945. = 92 = 917280
(v) Since we have used the multiplier D17280 12
12, we divide 917280 by 12. = 76440
|
Rep 996 x 1007 x 992
(i) We now work with base 1000. 996 - 4
(ii) (-4) (7) (-8) = 224 1007 + 7
(iii) Sy Et GS) + (7)4-8) 992 -8
== 8 $F32en90.=: OZ
(iv) 996+7-8 =995 995/ 5 2/224
= 995/0 5 2/224
= 9950 5 2224
994948224
Eg. 332 x 1007 x 62
Here, 332 and 62 are far from any of the base to power of 10, whereas 1007 is
close to base 1000. We therefore multiply 332 by 3 and 62 by 16 which results in 996
and 992 so that we now have the three numbers 996, 1007 and 992. These are the same
as in the previous example and we have seen that its product is 994948224. To account
for the multiplication of 3 and 16 above, we divide 994948224 by (3 x 16) = 48 which
gives the required answer 20728088.
Eg. 992 x 998 x 995 x 1002
(i) Since the common base can 992 - 8
be taken as 1000, we have P= 2
(ii) (-8) (-2)(-5) (2) = -160 = 160 G95-= 5
(iii) (-8)(-2)(-5) + (-8) (-2) (2) 1002+ 2
+ (-2)(-5)(2) + C5) (2) C8) 987/036/052/ 1 60
= -80 + 32+ 20+ 80 =52 4052 987036051840
30 SPEED ARITHMETIC

(iv) (-8)(-2) + (-8)(-5) + C8) (2)

+ (-2)(-5) + (-2) 2) +(3)@)
= 16+40-16+10-4-10= 36 > 036
(v) 992-2-54+2 = 987
4.3. Multiplication when Numbers are not Near the Bases 10, 100, etc.
Up to now, we have considered numbers which are not very far off a given base so
as to get the vertical multiplication of the RHS reasonably manageable.
We now remove this constraint and multiply numbers which are not near any base.
Eg. 48 x 39:
If now we take the base as 100 we will have 48- 2
48-52 and 39-61 which involves the vertical 39 - 11
multiplication of numbers viz. 52 and 61 ———_——
which are very large. To avoid this, we take OM ie OE 92
50 as the working base (w.b.) so that we 18 422
write 18 | ('/ x 100) + 22
1ST 72
= 1872
Since we have taken 50 = (100/2) as the w. b. we divide the LHS (only) by 2
which gives 18 MY, We retain 18 onthe LHS and take the fraction JF to the RHS
i.e. we carry over '/, x 100 (since 100 is the theoretical base) =50 to the RHS and
add 22 (already there) giving (50 + 22 = 72) onthe RHS. Thus, the product is 1872.
Eg nnd? X43. (t= 1007 wb = 50 = at00/2)
39-11
Ae
DO ISE EIT wma, alitt lk eae BO

Peo 10 (th: b:=100, wibs=

20 =.100/5)
29+ 9
19-1
 MULTIPLICATION 3]

Let us take again the earlier example 48 x 39. Let the th.b. now be 10 ( and not 100
as before) and the w. b. 50.
48 x 39 may then be written as 48- 2
39 - 11
ge ety ie Bu 2°= 1872
Z Z 2

Since 50=5 x 10, we multiply 37 by 5 to get 185 so that we have,

seh 2s 672:
2
ew Ze. Let the (th.b.= 10 and the w.b. = 60)
So we write 72+12
Joe 2

YP Ml ae ay 4
2 2 2
Stat 34 o=s4 176
Since 60 = 10 x 6 wehave onthe LHS of the stroke 70 x 6 = 420
giving 420 / 4 = 4184 = 4176
Z
Eg: Let us now work out the above eg. with the theoretical base as 100 and working base
as 60.
*. we now have 72+ 12
58- 2

W/o TOK. | 2 4a A904 = 4 2 A

Now 60 + 100 =3/5. -.70x3/5 =

*. The answer is42/ 24 =42 24 = 4176.
Thus, any convenient theoretical and working bases can be adopted to effect the
multiplication.
We shall now apply any of the above methods to get the product of 3 digit numbers.
It may be mentioned however, that for numbers with 3 or more digits, the method
developed later in section 3.5 is more suitable and easy.
Goin) SPEED ARITHMETIC

Eg. 478x512 (th.b; 1000'and

w.b. 500; w.b 500 = 1000 x 1/2)
AVES S22

-. We have 245| 2 64 = 244736

Eg. 478x512 (th. b. 100 and w.b. 500 ; 500 = 5 x 100)

478 - 22
312422

430/64
2
490x 5 = 2450 / 64 = 244736
2

Eg. 721 x 687 (th. b. 100, w.b. 700; 700 = 100 x7)
721 +21
687- 13

i) _73 >708x7/ 73 = 4956

2 2 2
4 73 = 49547 3
= 4954/

Eg. 404x516 (th. b. 1000, w.b. 500 ; 500 = 1/2 x 1000)

404 - 96
516
+ 16

ag /_536
]

—> 420~x 1/2 y

1
= 209 53 6 = 208464
 MULTIPLICATION 33

Eg. 279x241 (th. b. 100, w.b. 250 ; 250 = 5/2 x 100)

279 +29
241 - 09

A |ee eet 6 ao lI on= D| al

y yy a
“S25

= 6/239

Eg. 586x641 (th. b. 1000, w.b. 600 ; 600 = 3/5 x 1000)

586 -14
641 +41

6270S 7A Oreo
= 376 1/5/ 574
= 376/(1/5x 1000) +574
376/ 374 = 375626

4.4, Multiplication when Numbers have Different Bases

Let us now do some multiplication of two numbers whose bases are different.
Ee 29387 x 102
Here 102 isclose to 100; so we 987 - 13
multiply it by 9 to get 102 x9=918 918 - 82
and consider 987 x 918 both of which eee
are close to the base 1000. 905/ ,066 = 906/066 = 906066

We now adjust this product to account for the multiplication of 102 by 9, by dividing it
by 9, so that
906066+9 = 100674

pe, 1016x 93
Multiply 93 by 10 to have 930 which 1016 + 16
is closer to the base 1000, same as that 930 - 70
of the other number 1016. Then we SSeS geo .
proceed as usual viz. 946 / 20 = 945/ 120
l
34 SPEED ARITHMETIC

As in the previous example and for the same reason, we divide 944880 by 10 to arrive at
the final answer 94488.
Eg. 244 x 503
Neither of the 2 numbers is in the proximity 976 - 24
of aconvenient base of power of 10. 1006 + 6
However, we multiply 244 by 4 to give 976 ee AAs
and 503 by 2 to get 1006, both of which are 982/ 144 = 982 144
now near to the base 1000. a
We now proceed as usual. . = 951896

Now, since we have used the multiplies 4 and 2 to bring the 2 given numbers nearer
to 1000, we divide the product 981856 by (2 x 4) to obtain the final result as 122732.

4.5. Method Using Vertical and Cross - Wise Multiplication (VCP Method).
Till now we have considered multiplication of numbers, one of which at least was
close to achosen base like 10, 100 , 50, 250 and so on. This method is only a special
case of a more general one which is based on “vertical and cross-wise” multiplication. Let
us consider a few examples to explain the procedure.
(a) Two - digit numbers
The procedure to follow may be shown diagrammatically by the scheme:
a b a a D b
> | »< |
c d Cc Cc d d
ac ad + be bd
Eg. A2x% 54
(i) “Take a=4ob=2,c=5 andd=4 42
Multiply 2 by 4 to give 8 x 54
and place a stroke to its left
Ss

(ii) Multiply cross - wise (2 x 5) and (4 x 4) 42

and add 1.e. (2x5) + (4x4) = 10 +16 = 26. x 54
Place 6 in line with 8 on thé RHS of the Vis gt
stroke and carry over 2 on the bottom 68
line as shown; Place a stroke again. fhe /
 MULTIPLICATION

(iii) Multiply 5 x 4 = 20 which is to be added 42

to the carry over 2 Se ey!

ay aye
2
83 x6 = 83 x 06
3x6=18; soretain 8 and carry over | 83

ue
(6 x 8 ) + (0 +3) = 48; retain 8 and carry over 4 x 06
0x8=0

4/1 =4 oo

Three - digit numbers.

The corresponding scheme is:
a b C
=>
d e f
a b a b Cc Aa C

d
< e d
Se
e £ e ii
|
f
ad ae+bd af+cd+be bf+ce Ch

Eg. 793 x 166

(1) 3x6=18 — 8 with carry over | ios
(ii) (6x9)+(6x3)=54+18=72 “166
2 and carry over 7
(iii) (7x6) + (3 x 1) + (9x6) = 42 +3 +54 = 99 TNS OLS
— 9 and carry over 9. iy fs dayfocie hl
(iv) (9 x 1) + (7 x 6) =51 — 1 and carry over 5.
(v) TEL 131638
Eg. 692 x 87 = 692 x 087
(i) 27a14
(ii) (9 x 7) + (8 x 2) =79
(i11) (6x7)+(2x0)+(9x 8)=114
(iv) (6 x 8)+(9 x 0) = 48
(v) 6x0 =0
36 SPEED ARITHMETIC

(c) Four - digit numbers

The corresponding scheme is as follows.
a b Cc d
=> \

e f g h
a SC oot a b “e d

© e f e f g ie OK h
ae af + be ag + ce + bf ah + de + bg + cf
b Cc d >< d

ig
mere
g h g h
|
h
bh + df + cg ch + dg

Eg. 7981 x 3652 ay a 3 o/yh Vie

(Beer Sy
(ii) (8x2)+(5xb)=21
(iii) (9x 2)+(6x 1)+(8
x 5) = 64 = 29146612
(iv) (7x2)+(3x1)+(9x5)+(8 x 6)
=144+3+45+48 =110
(v) (7x5)+(3x8)+(9
x 6) =35 4+ 244+ 54= 113
(vi) (7X6)+(3
x9) =424+27=69
(vii) 7 xX 321
When we have numbers whose most of the digits are greater than 5, we can adopt the
same method of multiplication using vinculums for such digits and finally convert them
into non-vinculum digits.
Eg. 819% 780= elale? Lx 1214
(i) Ix 4=4 i
(i) a620644) cht CIx 1) =82AL eo algae ERs Deane
(iii) (1x 4)+(1x 2)+( 2x 1)=44+24+2=8 iA
Sneee
(iv) (x 4)4+( Exl+( 1x D+( 2x 2) hi. Ave be Odea oy 4
= 4+ 14+1+4=0
(vy) TOXIC 2X DCL x eeet Dae: 1 13 10894
(vi) (1x 2)+( 10a) 24 1= 3
(vii) 1x1=1 = 690894
 MULTIPLICATION ay
oa

As may be seen, the use of vinculum helps to avoid digits of larger magnitudes
and thereby simplify the task of multiplication and addition; also most of the latter can
be carried out mentally. Thus, not only is the tediousness of dealing with large
digits removed, it also helps to save considerable time.

Eg. 8738 x 908

Since many of the digits involved are > 5, we resort to the vinculum transformation.
The corresponding vinculum multiplication is | Usa a4 OED
However, since 5 digits are involved now in one of the multiplicands, we first
develop a scheme of cross-multiplication involving 5 digits.
a b c d e
=>
f g h i j

a arsed oe. a Cy aid eer

| woe g h f HY 295 ae <p Macatee heal j

af ag + bf ah + cf + bg ai + df + bh + cg aj + ef + bi + dg+ch

ee
Bg Dennel ol)
ee
ES ar A
ol
es j
bj +eg+ci+dh cj + eh + di dj + e1 ej

Now, consider the product. 1 1 342 x 1 11 2. Prefix a zero to bring the second
number to 5 digits. Perform the above scheme.
Oeil 24 me 1 dl, 4
Gi) (4x 2)+(1x 2) 1 1
0 rp|
WI]
eet2= 10 nico
Git} (3x 2)+¢€1x. 2) Oiihw
2y Bhdesifafi2if 20/4
= 4x 1)=6+2+4=12 ne
ivy VES 24a 7 Tht TOTS Sara,
(v) 2+0+ 14+44+3=4 12146104
(vi) 1+0+1+ 3= 1 7934104
ies Oe fe. 2
38 SPEED ARITHMETIC

(viii) 1+0= 1
(ix) =O

4.6. Binary Method

Although this method is not based on vedic want formula, we have included it all
the same, as an unconventional method of multiplication. It may be recalled that we have
demonstrated it earlier in chapter 1. For the same example 879 x 786, which we have
considered just a while ago, this method runs as follows:
(i) Take the smaller of the two numbers whose product 786 x 879
is required (so as to minimize the time of calculation), 393, SA SSE
(ii) Divide the numbers on the LHS by 2 (omitting 196 x 3516
fractions) and multiply the numbers on the 98 x 7032
corresponding RHS by 2 (until we reach | on the LHS) 49 x 14064*
(i111) Delete the RHS values which correspond to the even 24 x 28128
numbers on the LHS 12 ‘< 56256
(iv) Sum the numbers on the right, which 67x Figs
have not been deleted the starred ones. BF KEI Z 0A
We have the required product viz. 690894. 1 x 450048*
= 690894
Thus we see that apart from the usual method of multiplication, we have four
methods, the choice of which may be made on the magnitudes of digits involved, the
time and the easiness of calculations (preferably mentally).

4.7. Some Special Cases of Multiplication

Case (i): Suppose, we wish to multiply two numbers whose last digits add up to base
10 (or its powers viz 100, 1000,............... ) and the remaining digits in both the numbers
are the same.

Eg. 69 x 61
(1) Multiply the 2 digits 9 and 1 (which add to 10)
1.e.9 x 1 =9 and since we have 2 digits 9 and |
as multiplicands, we write this as 09 so as to
have 2 digits /09
(i) Multiply the common number 6 by its next
number i.€. oC. 42/09
| 4209
 MULTIPLICATION 39

Eg. 78x72
(i) 8x2 (2 digits) / 16
(ii) 7x8 56/ 16

5616
Eg. 294 x 206
(i) 94x06 (4 digits) / 0564
(ii) 2x3 6/0564

60564
Eg. 686 x 614
(i) 86x 14 (4 digits) /1204
(ii) 6x7 42/1204
421204
Eg. 2984 x 2016
(i) 984 x 016 (6 digits) /015744
(ii) 2x3 6/015744
6015744

Eg. 139995 x 130005

(i) 9995 x 0005 (8 digits) /00049975
(ii) 13x14 182/00049975
18200049975

Case (ii): We have seen already some of the mysteries of the number 9 in chapter 1.
Here we have a few more of the same in the multiplication of a number by another whose
digits are 9.

|
Se,
(i) Take one less of 5i.e.5- 1 4/
(ii) Subtract the number in (1) from 9 1.e. 9-4 4/5

45
40 SPEED ARITHMETIC

Eg. 547 x999

@). 547-1 : | 546/
(ii) 999 -546 : 546 /453

546453.

Eg. 730x 9999

(iy 730" 4 729/
(ii) 9999 -729 729/ 9270

7299270

**Case_ (iii): In the above examples, we have considered multiplications whose number
of digits equals or is less than the number of 9’s in the multiplier. Ler us now consider
the situation where this is not so 1.e. where the number of digits in the multiplicand
exceeds the number of 9’s in the multiplier.

Eg. “T2969
(1) As before, take one less than the
multiplicand i.e. 72 -1 viy
(ii) Subtract 71 from 9 to get-62 = 62. Th sb, 2

(iii) Now since the multiplier has only one 9, ‘

retain 2 inthe right andtake 6 as carry rY 2
over to the left. 6

+> 652 = 648

Eg. 737x99
(Gy 73h 736 /
(ii) 99-736 3G fosGe ak

736, ete:
. 6

Eg. 84034 x 999

(i) 84034- 1 84033/
 MULTIPLICATION AI

(11) 999 - 84033 84033/ 8303 4

= 83949966 ss
4.8. Multiplication when One of the Numbers is a Series of the Same Digit.

Case (it): Multiplication of a number by a multiplier consisting of a series of 1’s

We now consider a simple method (which can be carried out, in most cases,
mentally) of multiplying two numbers one of which contains 1’s only.

He a3 eit
Let us do this sum by the current method which will help us to understand better
the proposed short-cut method.
Doe ey
x tl
(1) 5 resis ZL
(2) See ey
(3) ye O77
| = 5907

The first and second rows identified as (1) and (2) are simply the multiplicand 537
itself, row (2) being shifted to one place to the left to correspond to the ten’s place of 7.

Now let us look at row (3), which is the addition of the two rows (1) and (2).

The unit’s digit 7 in row (3) 1s the same as 1n row (1).

The ten’s digit of the sum.is 3 + 7 = 10 1.e. the ten’s digit of 537 added to 7, the
immediate right neighbor of 3 in the multiplicand 537.

The hundred’s digit of the sum is 5 + 3 = 8 1.e. the hundred’s digit of 537 added
to its immediate right neighbor 3.

Finally, the thousand’s digit of the sum is 5 which can be obtained as the sum of
0+5, 0 being added as the thousand’s digit of 537.
 42 SPEED ARITHMETIC

Thus the product (row 3) is a result of a series of group-wise additions of digits

starting from the unit’s digit of the multiplication, the size of each group being
determined by the number of ones in the multiplier. Since-11 has two ones, we have
groups of two adjacent digits to be taken each time. The gore can be summarized now as
follows to provide the proposed method.

Few 7 530 as
(1) Add one zero (always one less than the number of 1’s in the multiplier)
at the extreme left of the multiplicand) to form 0537
(ii) Take the unit’s digit 7 first /7
(ii1) For the ten’s digit of the product, add the two digits 3 and 0
7 (since there are two 1|’s in the multiplier) and write the |
sum 10 as shown.

(iv) For the hundred’s digit of the product, add the two digits 5 + 3. /8
(v) Finally add 0 and 5. 5

.. Therequired product is 53/8/ 0/7

|y = 5907.
Eg 17269 x ‘11
(i) 017269 :

(11) Unit’s digit: /y

Gil) Tens dite 64 9: /:

(iv) Hundred’s digit: 2 +6 . js

(v) Next digit: 7 +2 /s

(yi), - Next digit: 1+ 7 e | hk

(vii) Next digit: 0+ 1 1

“. The product = ty Df
| De = 189959.
 MULTIPLICATION 43

1 oo20)er 111

(1) 0035201 (Two zeros are added at the extreme left since we have three 1’s in the
multiplier).

(ii) Unit’s digit: j

aie Ten’s digit: 0+ 1 A

(iv) _ We now add 3 adjacent digits every time (since the multiplier has three 1’s so that
the subsequent digits are

ce), 34542 /5+24+0/2+041 = yy

‘. The final product = LTT | = 39073110

Eg. 29129" x F111

(1) 00027129

(11) Unit’s digit: | hy

|
(iii) Ten s‘digit: ~2-+ 9

(iv) Hundred’s digit: 1+2+9 h

1

(v) Subsequent digits: 04+0+0+2 /0404247 Uh0424741 i2474142 y,7414249

. Assembling all these, we get the product = Lhd) /9 = 30140319.
1A Woy cia) Aa

Eg. 39036, x M1111

(1) 000055036

(11) Unit’s digit: |s

(111) Ten’s digit: .3 +6 hs

(iv) Hundred’s digit: 0+3+6 /9

44 SPEED ARITHMETIC

(v) Thousand’s digit! 5+0+3+ 6 of

.

(vi) Other digits:

0+0+0+0+45 /0+0+0+5+45 /040454540/ 0+545+043/ 54+5+04+346

RS TL AAA. = 611504996.
VA tA

Although we have written the various steps in detail for easy understanding of the
procedure for a beginner, the additions involved in these steps can all be done mentally
and the respective sums written down in one line as we have done in the last step using
strokes and the final answer therefrom.
In this method, only two points are to be remembered: (1) to add at the extreme
left, number of zeros which is one less than the number of |’s in the multiplier, and (11) to
form successive groups, each group containing as many 1’s as there are in the multiplier
(except for the units, tens etc. digits as the case may be).
Case (ii) Multiplication of a number by a multiplier consisting of a series of the same
digits (other than | and 9)
It might be noted that the above procedure will apply for multiplication of two
numbers, one of which consists of a series of digits not necessarily unity like eg. 22,
pape 2 0°SCS ete etc. The only difference is that we have first to multiply the
number (not containing the series of the same digits) by the single digit like 2 or 7 etc. as
the case may be and then proceed exactly as in the above examples.

Be \ 18312550333
(1) TOIL K\39S 8312 <3) XVII) = 234930. x Fl
Now to this last product we apply the above procedure of case (i). Thus, we have

(ii) 00234936
(iii) Unit’s digit: /s
(iv) Ten’s digit: 3+ 6 jh
(v) Subsequent digits: 0+0+2 jo
04+2+3 /24+34+4 /3+44+9 / a 9+3+6

. The product yy
VA,9,‘4;,
ie = 26077896.
 MULTIPLICATION 45

In similar examples, multiplication of a given number by a single digit alone is

involved and this is not difficult. One only needs to memorize multiplication tables upto 8
(even this may not be necessary, if one applies the corresponding vinculums, in which
case, one needs multiplication only upto 5).

The single digit multiplier 9 is however special, its complement from 10 being
unity. For this multiplier we adopt the procedure as explained in the next case.

Case (iti) Multiplication of a number by a multiplier consisting of a series of 9’s.

Earlier in this chapter, we have considered a few methods when the multiplier is
of the form 9,99,999.....Let us provide yet another method for the same. Before we
proceed with this method, let us take the example 3121 x 9 and carry out the
multiplication by the method of section 4.7, case (111).

Ee 3121 «9
(i) 21d A

(ii) 9-3120=- (3120-9): oe asd

Gin. “2: The product 331207Vo hase 21 1 P= 28089.

cae ei

Or
es eal 3120/
Giy 9= 3120 = 0009 -3120= 3 1 29
(following the vinculum method of subtraction described in chapter 3)

(ii)... Theproduct=3120/- 9 =3 21 29 = 28089.

‘se alae

Eg. ra WA ae opeJe)
(1) 3121 * 99 =G121 x9). If =280892x\ 11 (from the above example).
This multiplication can be carried out now using the procedure that we have
described in detail under case (i) above. The various steps are:
46 SPEED ARITHMETIC

(11) 028089

(111) Unit’s digit: “ /s

(iv) Subsequent digits: 0+2 (EO a) ony 8+9

. The product vy Va vehi = 308979.

] ]
Ee: B12 99

(i) . 3121 x 999 =(3121x 9) x 111

(11) The product 3121 x9 = 28089 as obtained above.

We therefore consider 28089 x 111

(iii) 0028089

(iv) Unit’s digit:

(v) Ten’s.digit: 8 +9

(v1) Subsequent digits: 0+0+2 if0+2+8 /'2+8+0 if8+0+8 rh0+8+9

. The required product= 2shYh,

vyVy)I y 9 = 3117879.
way ae se.

30718. x 9999
= (56718. «.9)>< 1111
First, consider the product: (56718 x 9)

56718 — 1:

9 ~ 56717:

.. The first product ll onon—] oe [oe)

On| O|
ae
I Nn a OO~ Vu
co, ~I|
nA
 MULTIPLICATION 47

Or
(i) 56718 —1= 56717

(11) 9 — 56717 = 00009 - 56717 = Si6uM12

(following the vinculum method of subtraction in chapter 3)

(iii) .. The first product = At mask.” =: 511: 662 = 51046)

> Ot
(b) The required product = 510462 x 1111
(iv) 000510462

(Vv) Unit’s digit: . hb

(vi) Ten’s digit: 6+ 2 /s

(vii) Hundred’s digit: 4+6+2 /[?

I
(viii) Subsequent digits:
0+0+045 0+04+5+ V/04+5+1+0 /5414044 ye1404446 we0444642
“. The required product = Vaasa By)to 2
Lost 7eh/ "1
= S67123252.

4.9. Summary of Multiplication Methods.

(1) When both the numbers are near a common theoretical base 10, 100, 1000 etc.,
we may use vertical multiplication and cross-addition or cross-subtraction.
(2) When both the numbers are not near the common theoretical base 10, 100, 1000
etc., we may find out acommon working base which when multiplied or divided by
a suitable number will give a common theoretical base. The result obtained with
this working base will have to be adjusted to conform to the theoretical base.
(3) We can use vertical and cross-wise multiplication method for any type of numbers.

(4) We may use binary method but it is often lengthy for large multiplication.

(5) When all or some of the digits are more than5, convert such digits to their
vinculums, apply the vinculum multiplication and then revert the vinculums to
their original numbers.
48 SPEED ARITHMETIC

(6) Various special cases of multiplication are also available as explained in sections
4.7 and 4.8.
Exercises:
4(a). Multiplication using base 10 , 100, 1000 etc. ix
(1)9x8 (2) 12x 18 (3) 13 x9 (4) 96 x 89
(5) 93 x 105 (6) 114 x 107 (7) 856 x 968 (8) 1103 x 974
(9) 1075 x 1052 (10) 986 x 1099 (11) 9065 x 1049 (12) 84x 113
(13) 7699 (14) 123 x 108 (5) 2052x923
4(b). Multiplication of more than two numbers.
(1) 92 x 89 x 90 (2) 23 17 xS0'« 16T (3) 992 x 99 x 99]
(4) 996 x 1002 x 1010. () 990 x 992 x 1001 x 1005
4(c). Multiplication of numbers not near the base 10 , 100 etc.
(1) 563 x 495 (2) 303 x 288 (Q)a15 x 755 (4) 214x 189
(S23 eZ (6) 633 x.612 (7) 8074 x 7963 (8) 5006 x 504
(9) 64 x 663 CFO) 388 x 4972 = (11):895:x TOFS (D2 302 VET?
(13) 512 x 501 (14) 621x583 (15) 88 x 485
A(d). Multiplication using vertical and cross-wise method.
(1): AS x93 (2) 98 x 76 (3) 153 x 965. (4) 893 x 26 (5) 301 x 1064
(6) 1875.x 72. (774. 1253 \(8) 564 «49 =9(9) 255\« 1279 (10) 1305. x 127
4(e). Multiplication using binary method.
(1) 76x 23 (2) 52.%95 (3) 116x218 (4)92x214 (5) 389 x 403
(6) 508 x 151 © (7) 118 x.983° (8) 107.x.586.. (9) 382 x87. ~(1Gy96 x48
4(f). Multiplication using vinculums.
(1) 989 x 1011 (2) 786x958 (3) 87%) 4135 »(4) 2312.x 93) / \(S) 692«:589
4(g). Multiplication when the last digits sum to base 10, 100 etc., and the remaining
digits are the same.
@h) 738)x 74 (2) 253 x 647 (3) 801 x 899 (4) 395 x 305
(5) 503 x 507 (6) 1076 x 1024 (7) 1653 x 1657 (8) 1235 x 1265
(9) 73823 x 73177 (10) 50026 x 50974
4(h). Multiplication when one of the multiplicands consists of only 9’s
(1).25,x99 (2) 19.x,99.9'~... (GB) 83,.x9., (4), 75.x:99 (5) 1016 x 99
(6) 2568 x 999 (7) 80056 x 999 (8)63x99 (9)483x99 (10) 483 x 999.
4(i). Multiplication when one of the multiplicands is a series of the same digits (using the
method of section 4.8).
(TP SIRT” =a) OI xa eo (4) 4.05 x 88
(5) 9.9969 x 99.99
 49

CHAPTER 5

SQUARES, CUBES AND HIGHER POWERS

Squaring, cubing and raising to higher integral powers of a number are only special
cases Of multiplication which we have discussed at length in chapter 4. However, we
devote a separate chapter for this topic since it has certain novel features of its own which
help to describe special techniques and in turn to increase the speed of calculations even
more.

5.1. Squaring of Numbers Ending in 5

The squaring of numbers ending in 5 (with 2 or more digits) is the simplest of all,
that can be carried out most often mentally.
Eg. 35°: (i) Square 5, to get DD/;
(i) Multiply the penultimate
number (3) by a number
which is one more than it Is)
= 33+1)=3 x4 1225

Bios 1G). /25

(ii) 16,16 4N)= 16x17 272/25
27225

we 305 (1) ak /25

(ii) 30(30% 1) =30 x 31 930/25
93025

9G. 5. D5 /25
(ii) 399 x 400 159600/25
15960025

5.2. Squaring of Numbers Ending in a Digit Other Than 5.

(a) We first consider a method which is closer to base 10 and its higher powers. The
logic of this method lies in the fact that
a -b’ =(at+b)(a-b)
= a =(atb)(a-b)+b°
where we choose a and b in such a way that (a+b) or (a-b) equals 10 or 100 or 1000.
50 SPEED ARITHMETIC

Egs. (4) 16° : we can write this as (16-6) (16+6) + 6°, where a= 16 and b=6
=a’ = 10 (22) + 36 = 220 + 36 = 256
(2) 1097 = (109- 9) (109 + 9) + 9? = 100 (118) + 81 = 11800 + 81=11881
(3) 9927 = (992 + 8) (992 - 8) + 8°= 1000 ( 984 ) + 64 = 984064

(b) We may also apply the two earlier methods that we discussed in chapter 4 in the
context of multiplying two numbers. In the first method, we considered
multiplication of numbers which are nearer to bases 10, 100,1000, 500, 250 and so
on. In the second method, we considered multiplication based on “vertical and
cross-wise” principle. We have also seen that this method is independent of any
base and is, for that reason, even more general than the forrner.

Let us now illustrate the vertical and cross-wise method in the context of squaring.

(i) Two-digit numbers: In the case of two equal multiplicands, the diagrammatic
scheme becomes:
a b
=>. a /2ab Jb
a b

Eo, 5A OE 25/2 620/16 = 25/40) 165 25 yh 6

4/7 |
— (25 + 4)/(0+1)/6
> 29/1/65 2916

Eg TA SS 2 I 4) 4? > 49156716 > Aik 6

th a
—»54/7/6 = 5476

(ii) Three digit numbers :The corresponding scheme is

a b Cc
Siva J 2ab/ 2act b’sf2besfe
a b 3

Ee. 3177 4 37]

& ae UCD Xa Tita) Dadhoe iad

> 9/6/43/14/49 re/SNS 100489

4/1 /4
 SQUARES, CUBES AND HIGHER POWERS J]

Eg. 5097 feach ots W/o cpa 259081

9 8

(111) Four digitnumbers: The scheme is

a b C d
—
a b c d

a’ / 2ab / 2ac+b’ / 2ad+2be / 2bd +c? / 2cd / a

Eg. 49367 2/2x4x9/

(2x4x3) + 9? /2(4x6+9x3) /2(9x6)+37
/2x3x6 / 67
— 16/72/105/102/117/ 36/36
-> St ea ee
date 107 1A Ase 3

Eg. 2497? : In this case, the squaring can be done conveniently with the help of
vinculums by writing 24977 = 25 0 3?

250 3° > 4/20/25/1 2/ 3 0/0/9

is, y)vpvere? ig — 624 5 009 = 6235009
IV ASL ee 3

(iv) Five digit numbers:

a? / 2ab / 2ac+b* / 2ad+2be / 2ae+2bd+c’ / 2be+2cd / 2ce+d? / 2de / e”

Be 0798S. 1032

10933 2 —- 1/0) 6G/6/5/ APB QV yenlso4

1/0/ 6/ 6/ 5 ak 1 eae) 4
info27 1

+» 10 664 60 24 = 94633984
a2 SPEED ARITHMETIC

Eg. 12.342°: In decimal multiplication, proceed as if there is no decimal and in

the final answer, place the decimal in the usual way.
123427 01 £4. 410/2 297 0;32/28 L1Gh4
> 4f00/f OF fe defo8 16 / 4h 211324064
i, Me fee IW he OPOSG 88

5.3. A Combination of Methods

For numbers ending in 5, we may partially apply the earlier method of squaring the
last digit and suffixing the product:
[ remaining digits x (remaining digits + 1)] and partially the method of “vertical and
cross-wise”. Some examples illustrating this approach are given below.
Eg. 315° Let us treat 3 and 15 as 2 groups.
— (3/15° ) = 3° / 2x3x15 /157 > >|ay: 25 99225
2 oN;
Alternatively, we treat 31 and 5 as 2 groups so that
315" (Bid
ale edd?
x315 15° 3 9617 310725
=> Ro ahi Dy voees
31 A"
The above two examples, show how the carry-over is done depending upon the
partitioning of the given number viz. that the number of digits in the top line within each
stroke on the right is the same as the number of digits on the extreme left after
partitioning.
Eg. 455° -—> (4/55)
>firstlye<55° = 5641) 157 = 30/25.
next, we have 4° / 2x4x55/ 55° and using 30/25 for55°,
=> fwe wget 4/55° 95> eat 25, 4555207025
4 730

Bo t625>3 (18) 95)" ee 25 =. 3330625

9 46

5.4. Another Method of Squaring

(a) Using base 10 and its higher powers
Eg. 93°: (i) Base 100; deficit = 93-100 = =
(ii) No. + (deficit) = 93 +(-7) = 86
(iii) Sq. of the deficit = (-7)° = 49
-. 93° = 8649 (using the results of steps (ii) and (iii))
 SQUARES, CUBES AND HIGHER POWERS 53

Eg. 108°: (i) Base 100; excess =

(1) 1084+8 =
(iii) 8° =
». 108° = 11664

Eg. 984° : (ty pasey = 1000, elicit: =

(i) 984 -16 = 968
(iii) 167 =
Sq. = 968256
Eg.1021°: (i) eBase= 10003 excess =
aiyY “O22 1042
(iii) 21° = 44|
“. Sq. = 1042441

(b) Using bases other than 10 and its higher powers.

Eg. 257°: (i) theoretical base (th.b.) = 1000;
working base (w.b.) = 250
“. w. b./th.b .= 250/1000 = '/y.
Excess = 25 = 250 =

(Gi) 257+7 = 264 ; 264 x (w.b./th.b)

S164. ia
Git)! FE 490e 049
(since three zeros are present in the th.b.).
Ans. 66049 (from steps (11) and (111))

Eg. 264°: (i) th. base 1000, working base = 250 ;

“. 250/1000 = '/4
Excess =.264-=250 = 14

(ii) 264+ 14=278;278x'/4 = 69 '/,

Now '/, of 1000 = 500; this is to be
added to the result of (i11) below.
(iii) 14° = 196
. Ans. 69/ (500+ 196) = 69696
 54 SPEED ARITHMETIC

Eg. 516°: (i) th. base = 1000, working base = 500;

| 500/1000 = '/>
Excess = 516 - 500 = 16
Gi) 546416253256 5 =, * 266
(iii) 16° = 256
*. Ans. 266256

Exercises
Find the square of the numbers given below:
5(a). Numbers ending in 5
(10252, (2), 85° (3). 1057= - 164) x275900 (5) ALS
5(b). Numbers near to base 10, 100, 1000, etc.
Gupce (Q):03*— "Gy 987 (Ay 1014) 127
5(c). Any other numbers :
(Ijse63" Gy sie (3) 1062 (4): 983? (5) 10167 = (6) 19877
(Ty 20047 -(8) 992" (Oy S1d* 0). 250.2 4(11), Olek ec Gla) 699°

5.5. Sums and Differences of Squares

As an extension of squares, we now study some simpler techniques for finding the
sums and differences of squares.
The occurrence of sums or differences of squared numbers is very common in
mathematics; for example, the Pythogoras or the Apollonius’ theorems, the squared
distance of two points whose coordinates are given, or the trigonometric identities like
sin’ 8 + cos’ @ = 1 etc. We shall therefore see if some simple working rules are available
to express a given number as the sum/difference of two other numbers.

_ (a) To express the square of a number as the sum of two squared numbers.
We may divide this into 2 cases:
(i) Case of odd numbers:
The square of an odd number is also odd and will have 2 consecutive middle
numbers whose sum equals the square of the odd number. For example, 3° = 9= 445,
5? = 25 = 12+13, 7° = 49 = 24 + 25 and so on. The first of the two middle numbers of a
given number (particularly if it is large) may be got easily by halving the given number
and taking only its integral part. The other middle number is obtained by adding unity
to the latter. We then express the square of the larger odd number as the sum of the other
two squared numbers. Thus, taking the above examples 5° =3°+4? or 137=5?+4 12? or
25S TOA
 SQUARES, CUBES AND HIGHER POWERS 55

Egs. (1) 33°? = 1089( 1089/2 = 544'/,)

. 1089 = 544 +545 leading to
337 +544? = 545°
(2) 147° = 21609;
(21609/2 = 10804 '/,)
”. 21609= 10804 + 10805
so that 1477+ 108047 = 108057.
From these, we can easily form the “Pythogorean triples” viz (3,4,5), (5,12,13), (7,24,25),
(33,544,545), (147, 10804,10805) and so on.

(ii) Case of even numbers:

An even number when squared gives an even number and therefore it is not
possible to have two consecutive middle numbers. However, we can divide the even
number (to be squared) by 2 or 2’ or 2° and so on until we get an odd number. We then
apply the method of case (i) to it and having got the two required squared numbers,
multiply by the same multiple of 2 used for dividing the even number initially.
os. ol: 6° ; 6/2 =3 and we have seen above that 3°44’ = ite now
multiply both sides by 2 to get 6° +8°=10
2. 12° ; 12/4 =3 and again 37447 =5’; multiply by 4 so that
12? 4 16r 200
3. 112?;112/16=7 and 77 = 49 = 24 + 25 giving 7° + 24° = 25”.
This on multiplication by 16 gives 112 + 384° = 4007 and so on.
(b) To express a given number as the difference of two squared numbers.
~ Given anumbern we write it as a product of two numbers, say 6 and g so that
n = pq and then express nas the difference : ( (p+q)/2)” - ((p-q)/2)’.
The underlying logic is that (( p+q)/2) - ((p-q/2)" =pg =n

Eg. 3 =3 x1 so that p=3, q=1

*, 3 = (GHY2) -(G-AV2y =(2? = 17
We can write a given number as the difference of two squared numbers in a number
of ways.

SIO9wH o~—yS, 6x2 = ((6+2)/2)° - ((6-2)/2)* = 4°-2° orwecan have

— = 4x3 = ((443)/2)? - ((4-3)/2)? = (3.5)
oa)
rte - (0.5)°
(2). Similarly, 108 can be written as various squared differences given below:
108 = 36x3 = (19.5) - (16.5)
= 18x6 = 12°-6
12x9 = (10.5)? - (1.5)
56 SPEED ARITHMETIC

5.6. Cubing of Numbers

(a) Method 1: This is based on the result:
(a+b) = qg + 3a°b +3ab’ +b° which can be written in 2 rows as:

Qa pe oe a aa
+ Dab 4 gab

Sumacsre a Ga bate Sab abs

As can be seen, the first row consists of a’, a’b ; ab’, b> where the succeesive numbers
are in the ratio of b/a if one moves from right to left and in the ratio of a/b if we move
from left to right. (That direction is chosen which gives a simple value for the ratio). We
next write two times the middle terms of the first row below the middle terms as shown.
We finally add these two rows to get the cube.

Eg. 83°: Here a=8, b=3. We move from left to right so that the ratio b/a = 3/8.
This gives
8° =512, 3/8 x 512 = 192, 3/8 x 192 = 72, 3/8 x 72 = 27 (check: 27 = 3°)
The two middle digits are 192 & 72 so that 2 x 192 = 384, 2x72 = 144
We write these as oa ee)

by SD 6 6 fea ;
Me fod: B/Q = 571787

Eg. 143°:Here a= 14 and b=3. In tliis case we move from right to

left with ratio a/b= 14/3

14/3 x 588 = 2744 14/3 x 126= 588 b4fo Reet = T6 Aj

LROOS = LITO 2X $20 =a2o2

— 2744 4 8 7
Aes Wh = _ 2924207
The method also allows the use of vinculum in case the digits of a given number whose
cube is required are greater than 5.
 SQUARES, CUBES AND HIGHER POWERS 57

SS sy 707 415077
04 T= 93039

(b) Method 2 : This is similar to excess/deficit method used for squaring but with a
slight difference. Let us, as usual, illustrate with examples.

Bo 106 : (i) Base 100, Ist excess = 106-100. =, 6

(1) No. + 2 (1st excess) a MOGaE bo a 18
(i111) 2nd excess = (11) - Base =i behS re l OO repehs
(iv) 2nd excessx Ist excess SLL OL Me SONOS
(v) Cube of Ist excess = 6 =216

“. We have 106° > 118/ 108 /216 from steps (11),(iv) and (v).

+> 118 / 08 16
| 2 — 1191016

Eg 88°: (i) Base0051Istdeficit-88-100. = - 12

(iy. 88+ 2(- 12) 88 424 = 64
(iu1) 2nd deficit
= (11) - Base =—64- 100 = =36
Gy (36a) = 432 na sa
Gane iy 1728 = 1728
giving 88°> 64/432 /onie2-8!
CUR Oiles Th Zn8) 2a
4 lay, > 6815 2 8 = 681472
58. SPEED ARITHMETIC

In the above examples we have considered numbers which are close to 10 and its
powers. Let us now take up some examples in which this is not the case.
Eg. 192°: (i) th. base = 100, working base (w.b.) = 200
Deficit (from w.b. ) = . -8
(ii) 192+2(-8) = 192-16 = 176
Now since (w.b. /th.b.) = 2, we multiply 176 by 2” and
get 176x4 = 704.
(iii) -2nd-defieit-=-(176~200)-= - 24
(iv) Ist deficit x 2nd deficit = (-8) (-24) = 192
We multiply this by 2(= w.b./th.b.) so that 192 x2 = 384
(v) Cube of the Ist deficit = (-8)° =-512= - shied
Putting the values of steps (11), (iv) and (v)
together, we have pw 2h
1927-5 gf Se 2—> 70779 1 2 > 7077888
3 5
Eg. 2483 (i) th. base 100; working base 250; deficit = 248 - 250 = -2
(ii) 2484+2(-2) = 244
244 (w.b./th.b.)? = 244 (2-5) = 1525
(iii) 2nd deficit (244 - 250) = -6
(iv) (-2)(-6) = 12 and(12/2,5) |= 30
(v) (-2)° = 8
248° 1525/30/ 0 8 = 1525300 8 = 15252992
Finally, we work out some examples involving vinculums to see how simple and
quick the calculations can be.
Eg. (96) :96= 104; Base = 100
(ij) ishexcesse=
(ii) 10 44+2( 4) =104+4 8oo I —

(iii) 2ndexcess =
(iv) 4x 12 = 4
(vy) (4)
Thus (96)° > 1
> |

Eg. (9898): 9898= 10 1 0 2; Base = 10000

(i) Ist excess = a
(Gi) 10° 102 4+2(10 2) 2 10102 eee 103
 SQUARES, CUBES AND HIGHER POWERS ah)

(il) “2ndexcess =. 306

(iv) (10 2)( 30 6)a= 2 24 2:
(vy) (10 2) = UD) Guth Mls:
so that (9898) 310 30 6/31212/ 1061208
eS aia bal Tre.
3 10 6
Bs AG5003 iid ail 0 85.10: 305 3.141 + 1 2028
= 969711058792

5.7. Raising to Fourth and Higher Powers.

Basing the logic on the binomial theorem as we did for squaring and cubing, we
have, for the 4th power.
(a+b)* = a’ + 4a°b + 6a°b?+4ab°+ b*, so that we have the following two rows:

abe Fabs tab = “b:

3a°b Sab? 3ab°
which on addition gives (a+b)*
The terms in the first row are in the ratio of b/a if we move from left to right and
a/b if from right to left (same as in the case of cubing). The second row consists of
excess terms.
Eg. 12*: here a=1,b=2, direction: left to right; ratio is b/a = 2/1=2
Thus 1* =1, 2x1=2, 2x2=4,.2x4=8, 2x8=16
and 3x2=6, 3x4=20,.3x8=24
Putting together and adding, we get

i =

— 20736
Libs)
We 8/4 ee

Eg. 122*: a= 12;b=2, direction : right to left .. ratio= a/b = 6. Thus,

6x3456 = 20736 |6x576 = 3456 | 6x96 = 576 | 6xl16= 96 |2*=16

SX3450 = 10368. 15x5/6-= 2880 | 3x96 = 288
60 SPEED ARITHMETIC

Adding these two rows gives

(422° —+ 20736 / 13824 / 3456 / 384 / 16

7 ON on 5 [a]
For avoiding errors, this may be written as:
Av
3456
13824
+ 20736

221533456
sothat.(4 220 =) 221530456, ood) lainttor
Powers higher than four of a number can be obtained similarly by making use of the
binomial expansion of the required power and proceeding as above.

Exercises
5(d) Find the cubes of the following numbers:
(1) 96 ye lO2seg3)a153- (4) 981 OWe25

5(e) Extract the fourth powers of :

(1). 26 (2) LOSiS(3)32 (4) 312 (5) 99

 61

CHAPTER 6

DIVISION

We now consider the reverse process of multiplication viz. division. In this context,
it may be remembered that the number unity is neutral to both multiplication and division
in that when a number; is multiplied or divided by unity, the same number is returned.

As in the case of multiplication in the earlier chapter, we present more than one
unconventional method of division all of which are shorter and quicker in comparison to
the conventional method.

We start with the easy case of division by 9, whose base is 10, using three
alternative methods.

6.1. A Simple Method.

Fg. 231+ 9
(1) Split the dividend by a stroke such that, the number 9 | PAS | 1
of digits on the RHS of the stroke equals the number
of 9’s inthe divisor. The stroke also identifies the
quotient digits (Q) appearing on the LHS of the
stroke and the remainder digits R on its RHS in the
final answer.

ai) Bring down 2 of the dividend below, as shown SU CES a ER

2

(ii1) Take the Q- digit 2 below 3 and take their addition Se aoe
1.e. (2+3) = 5 which becomes the next Q-digit. 2

(iv) Write the 5 below | on the RHS of the stroke and Oe ok |

add to give (1+5) = 6 which is the remainder a Ne
digit. 6
(v) Thus the quotient is Q=25 and the remainder
R= 6.
62 SPEED ARITHMETIC

6.2. Method of Complements.

Eg. 231 + 9 (same as above)
(i) Write, below the divisor 9, its complement viz. 1. 2 9;2 3}1
1

(ii) Bring down the first digit2 of the dividend to ey eee

form the first Q- digit; multiply it by 1 (the 1 Z
complement of 9) and write the product 2 x 1 =2
below 3 of the dividend; Add 3 and 2 to give 5
and write it as the next Q - digit.

(iii) Multiply the new Q- digit 5 by | and write it below | On?

on the RHS of the stroke; add | and 5 to give the i
remainder 6
Thus"O'=25"-and#k'=6

Eg. 3227+ 9
“. (@t=s3 2 and Riss

Eg.316 + 9
*Heteak =s1 0 >9)
.. Add 1| to the Q and take away (1 x 9) =9
from the remainder to give
O=35.and R=1

Fe 299 9
Proceeding as before, we have [1 on adding 9+2
we retain the right digit | as the second Q-digit
and carry over the left digit 1 (as shown) to be
added to 2. Accordingly, we have (2+1=3) as the
first Q-digit and | as the second Q-digit.
Next we take (11 x 1) = 11 to the RHS ofthe
stroke as usual and add to give 20 as the
remainder. Since 20 > 9, we add 2 to the Q- digit
and subtract (2 x 9) = 18 from 20 to obtain finally ©
“ QO=33 andR=2.
 DIVISION: 63

** 6.3. Method of Cumulation.

(a) Dividing by 9.
Of all these methods, this is the most simplest, quickest and on most occasions, can
be done mentally.

ee 250 => 2) |5 tO = 2, RES.

Write the dividend as 2 | 5 as shown placing one digit on the RHS of the stroke to
correspond to only one 9 of the divisor. The Q - digit is 2 and (2 + 5) =7 is the
remainder giving Q=2, R=5.

eae) > 23) 13> 21 223=5 |54+1=6 SOUR <6

Take 2 as the first Q - digit, (2+3) = 5 as the second
Q - digit and (5+1) = 6 as the
remainder so that Q = 25 and R=6.

Peto 52 |? -3.31349=5.|542=7 3s. 0=55 Bed

Ist O digit = 3, 2nd QO digit =3 +2 =5, and the remainder =3 +2 =7
Q=35 R =7
Fae
es a
. Ist Qdigit=3, 2nd Qdigit 3+1=4,andR=4+6=10 .. Q=34
But R=10is >9. .. Add 1 to the Q and take away 9 from R to give
Q = 344+1 =35 andR=1..

Eg. 59839 + 9 > 5983 |9

Adding xiaa cumulatively, we have
5,5+9= 1448=22, 224+3=25 | 25+9=34.
— S42) 5/4
ROO

On adding the carry-over digits with the corresponding digits, we have

5+1, 442, 2+2, 54+314+3 — 664817 so that Q= 6648 and R=7.

Note: The last carry-over digit 3 is added to the Q-digit 5 which is before the stroke
as well as to the remainder -digit 4 which 1s after the stroke.
64 SPEED ARITHMETIC

Eg. 613428 + 9
61342|8 — 6, 6+1 =7, 743= 10, 1044= 14, 1442= 16 | 16+8=24
6m 6 amiigyad
opishy WPlgivhesor othai,
= 6 Fel =8) 041 <1. Aeliage 642 =3|
42 —0
> 68158|6
giving QO= 68158, R=6.

ig. 910019 + 9

Pee
hconyen
> 91001|9 9, 941=10, 10+0=10, 10+0=10, 10+1=11 | 11+9 =20

12AxeteitiZita-A02
58931510; 7041=1, 0+1=1, 0+1=1, 142=3 |.6m
25 oniig4| 2
giving Q =101113, R=2

(b) Dividing by 99
Although we may carry out the division of a number by 99, 999 etc. by any one of
the first three methods as for 9, division by 99, 999 etc.may be accomplished more easily
and quickly by adopting the method that we described just above.
Eg. 252 + 99 — 2|52 (Note: the stroke is now placed so that the last 2 digits of
the dividend are on its RHS since the divisor 99 has 2 digits.)+
> 2|2+52> 2/54 +. Q=2,R=54.
Eg. 3067 + 99 — 30|67 — 30|30+67 > 30|97
-. Q=30, R=97.
Eg. 4059 +99 — 40|59-— 40 |40+59-40 | 99
> 40+1 | 0(since99 =(1 x 99) +0)
ay | | 0
Q =41,R =0.
Eg 37912
+99 — 379 |12 :
Since the divisor 99 has 2 digits, we would need even number of digits on the LHS
of the stroke, so that we can form the Q-digits in pairs. To enable this, we add a 0 to
the left of 379 and write itas 0379. Then, we have 03, 03+79 | 03+79+12 giving
03 82|94 sothat Q =382, R=94.
 DIVISION 65

Eg. 15704 + 99 — 0157104501, 01457 1014+57+04 > 01 58

| 62
soshat OH 158) R= 62. °

Eg. 368391 + 99 —+ 3683191 > 36, 36+83 | 36+83+91 —36,1191210

= of a

—
37 a
eiviag © = 3721, KR =

(c) Dividing by 999, 9999 and so on

We may extend the just described method to divide by 999 or 9999 and so on.
Fg. 312341 + 999 -— 3121341 (Note: 3 digits after the stroke to account for the
three 9’s in the divisor)
@-=312, and R= 312 + 341-='655

Ee.. Billo -9999 >» A1A721

O=4,-and R= 4+ 1721= 1725

Eg. 149999 + 9999 — 1419999

Q = 14, and aa 14 +9999 = 10013 =( 1 x 9999)
+ 14
Q=14+4+1 5 and R = 14
Thus one can seeare simple and easy it is to divide any number by divisors
consisting of 9’s without resorting to laborious and time consuming long division. KK

6.4. Division when Divisors are Different from 9’s but are Close to the
Powers of 10.
Eg, 108°+.7
(i) As before, we leave one digit on the right of POM | ATA RE:
the stroke since the divisor has a single digit.

(ii) As we explained for the divisor 9, we J/gled 0 8

now take the complement of 7 viz. 3 from 3
10, as the working divisor.

(iii) We bring | of the dividend below, multiply dale Ue ae

it by 3 and place the product below 0 (the next 3 3 9
digit) and add to get the next Q-digit at 3. Tele by.
66 SPEED ARITHMETIC

(iv) Again, we multiply this by the divisor

to get 9 and place it below 8 on the
RHS of the stroke and add to obtain 17.

This gives Q= 13° andR =I7> ° (>°7).

Now 17 = (2x7) +3 and so, as in the earlier cases,
we add 2 to the Q and subtract 2 x7 = 14 from 17
so:that O.=13 + 2= 15 and R= 17-14 = 3.

Ee. OOo se=

(i) Take the complement of 7 viz. 3 and write
it below 7.
(ii) Bring down 8 of the dividend and multiply
it by 3 to get 24; write 24 as shown and
add 6 to 24 to get 30: Take 0 on the Q-line
and 3 on the carry over line.
(111) Multiply 30 by 3 and write the product 90
on the RHS of the stroke and add it to 3
to'get 93: (>'7);AS 93.= (Gx 134-25 we cet
Q = 110413 = 123, andsR = 93-91 = 2.
Fe5* 029 226 Eg, =259 2087
fe D9
YOryS 26
Ms 85

141 (6x23 + 3) Eh Et ay(ole ot

-138

O = 116 andskR:=3.

Eg. 40854 + 878

222*
244. > (2
 DIVISION 67

*Here, we get 2222 as the first remainder which we can deal with as if it is the new
dividend and proceed to divide as shown to get Q= 2 and R = 466 first. However, this Q
has to be added to the Ist Q viz. 44 to get the final answer as
Q =44+4+2=46
andR. = 466.

OR the first R = 2222 = (2 x 878) + 466 so that we add 2 to the old 44 to get
Q=46 and R= 466.

The above method is recommended only when divisors are large as in the above
example. If it is pursued for small divisors, the division process becomes extremely
lengthy and cumbersome and is no better than the conventional long division. The
question naturally arises: what, if the divisors are small. In such a case, we follow the
method described below. .

6.5. The Multiples Method.

| This method overcomes the lengthiness and makes the calculations short and quick.
The method brings the divisor closer to a power of 10 and is somewhat similar to the one
that we adopted in the previous chapter on multiplication under the theoretical base and
the working base. Let us explain by means of the following illustrations.

Eg. 1216327
(1) Multiply the divisor 27 by a suitable number to
bring it as close as possible to the appropriate
power of 10. Here we can take 27x 3 =81
and its complement from 100 to give 19.
(ji) We proceed exactly as before with the new
dividend 81 until we reach the step marked(*).
Now the remainder 163 has 3 digits (one more
than the divisor 81). So we deal with 163 as if
itis the dividend to be divided by 81. Thus
we bring down | of 163, multiply it by 19, take
the product to the right of the stroke and add
to 63 to give | | 82. The new remainder
82 =(1 x 81) + 1, and so we add | to the Q-side
and retain | on the remainder side. Adding all
the numbers on the Q - side, we have
O= 13:4+2=15 and-R=1.
68 SPEED ARITHMETIC

(iii) We finally multiply only the quotient 15

(and not the remainder) by 3 (the multiplier
with which we multiplied the divisor 27 in
step (i)) to get Q=15x3=45 andR=1. . <

Eg. 3467 + 24 Eg. 4128 +62

62x16
992:
008

w O=(G4+2) x4=144, R=I11 . O=.66, R= 36

Exercises. (/ indicates division) (to 3 decimals)

6(a).. Division by#9499"099:
(1) 100/9 2) 509/9 (3) 309 (4) 9100/9
(9) 1995/9 (6) 805/9 (7) 8105/9 (8) 298/9
(9) 720/9 (10) 8975/9 CPE) U8999/9 (12) 8001/9
(13) 19080/9 (14) 12605/99 1S} 98 T89280g, (16) 99999/999
(17 99989/99 (18) 9016832/99 (19) 1254011/999 = (20) 93507/89
(21) 3683491/99 (22) 3461021/999 (23) 454/99 (24) 71395/99
6(b). Division by other numbers.
(1) 7836/6 (2) 83259/8 (3) 80532/6 (4) 43829/86
(5) 9613528/96 (6) 832515/85 (/)., 095201792 (8) 32865/78
(9) 9084655/985 (10) 832169/90

6.6. “Change of Sign’? Method .

We have seen that the method of complements discussed at the beginning is
preferable whenever the division has large divisors. In the event that this is not so, the
“multiples” method is available. However, the latter method is convenient if (i) a multiple
on multiplication with the original small divisor leads to a number that is close to a power
of 10 and (ii) such a multiple is an integer.
 DIVISION 69

We now consider a method - Change of Sign (C-S) method- which is particularly

suitable for divisions with small divisors when the multiples method is not feasible
eo 113.4 12
(1) From the divisor 12, drop the extreme left
digit viz. 1 and change the sign of the other
digit (s) so that 2 of the divisor becomes - 2
(11) Place the stroke on the dividend so that its
RHS contains the same number of digits
as the divisor but modified as in (i) above.

(i111) Proceed as before, with the modified divisor -2

(iv) The quotient Qis 1 1 which

equals 2.4... O.59:R = 5,

2... dO 71> 122

We proceed as in the example above
until we have reached the step marked(*) ORESE, BEG Sad
Having got 14 on the Q- side and
-3 -7 on the Rem. side, convert
the latter to -37; in order to render it
+ ve, take away one time the original 14 -57
divisor viz. 1 x 122 = 122 and add Les err ee
it to -37 to get R = 85. Thus, the final ie: O39
answer iS Sd HRS 85:

Bene 514) = 150

ty a0 S145
70 SPEED ARITHMETIC

Eg. 13579 + 1141.

1141
et

= 11, R=1028.

As might have been noted from the above examples, this method requires that the
first digit of the divisor be unity. If this is not so in the original divisor, then it has to be
brought to that form either by converting the divisor into its corresponding vinculum or by
mutiplying the original divisor by a suitable multiple so as to have digit one at the start of
the product or by means of a combination which would bring unity as the opening digit.
Eg. (i) _ if the-original divisor is 823, then its equivalent in vinculum is 1 223 which
has 1 as the starting digit.
(ii) if the original divisor is 326, we multiply it by 4 to get 1304 whose first
digit is 1.
Let us take an example in which the divisor is not one and ignoring it , we use the
change of sign method.
Eg. 3257 + 829

+ Qe 25, R= 7/32
The answer is obviously wrong
Let us now take the same example and use the vinculum cum C-S methods and
multiples cum C-S methods
 DIVISION ° 71

Eg. 3257 + 829 (using vinculum and C-S methods)

Here we first obtain the vinculum of the
2 large digits 8 and 9 to get 1 2251,
This has | as the starting digit in the
modified divisor and so we can apply
the C-S method.

Eg. 3257 + 829 (using multiples and C-S methods) 829 x 2 vm 8, 7

Here we multiply the divisor by 2 1658 -18 -15 -24
to have | as the first digit in the -6-5 -8 -16_ -10 -17
modified divisor.

Of=ae) Ra=2 770".

Since in the above example, the divisor is large we need not use C-S method but apply
either the (i) complements method, or (ii) a combination of complements and vinculum.

* O=324R=77/0 Q=3, R=770

heel 21Oe2] 2x 2 LEG
Here the starting digit of the divisor 108
is not1 and therefore we choose the -0 -8
multiple 4 to make the first digit of the
modified divisor to be 1 and then adopt
the C-S method.
a2 SPEED ARITHMETIC

Eg. 70169 + 8989 |

As seen earlier. direct C-S method would fail since the divisor does not start with |
and we therefore adopt either the complement method or a combination of vinculum
and C - S method, both of which lead to the same modified divisor as can be seen below.

Complement method Vinculum and C-S method

8989|7 | 0169 Bs ly ha 70169
1011 TOET ja ae8 Se
7246 POQALS (modified divisor, the same as
that of the complement method)
re ae 2G
6.7. Combination of Methods.
In the above examples, we have considered a combination of methods in order to
restore unity as a starting digit of a divisor. We now consider some examples where we
combine “multiples” and C-S methods to our advantage in making the division simpler.
Eeod371y s2 2584

Eg. 4561 + 484

 DIVISION

Eg. 33167 + 2787

Eg. 631056+ 389

x2
t 6270
+ 2

“. Q= 1622, R= 98
74 SPEED ARITHMETIC

Eg. 1736
+ 242
Dic 04 wala
or
-2: =1

Eg. 1732+ 42

Exercise.
6(c) Perform divisions (to 3 decimals) using any of the above methods.
Cys OS ara? (2) 3 318e2d 1.3 (3) 70086/84
(4)-9983 3/109 (5) 83469/612 (6) F362 s.19
(7) 89506/1016 (8)103856/1154 (9): 3 15:6 2/9:85
(10) 3°6 5 23/256 (1b) <8 81905 A/3,3 (12) 92365-1/687
(13) 3493:3°7/2003" KIS) 8:5 24esas 7 (E35) 1-6 S02
POS 7
 DIVISION. 75

6.8. Vertical and Cross-Product Method.

We shall now describe a method of division which compared to the earlier methods
is far more simpler and less time consuming and more often than not involves lesser
number of digits to add, subtract or multiply. For purposes of comparison of methods let
us consider the division of 31407 by 81.
(1) Complement Method (ii) C-S Method
81
19
BL
las
2-1

Q = 387, R=60

= 375 +104
2= 3875: R= 60
Now, let us take the same example and perform the division by the Vertical and
Cross Product (VCP, for short) method.

Eg. 31407 + 81
(1) Write 8 and | as shown where the number at
the top i.e. | is the index or the flag digit and
the number below it viz.8, the working divisor.

(11) Place a vertical line on the dividend so that

as many digits are present at its right as there
are digits in the index.

(111) Since 3 cannot be divided by 8, we take 31

of the dividend and divide it by 8 to get
Q =3 and R=7. Place the Q on the 3rd row
(below | of 31) and the R, between | and 4
of the dividend.
76 SPEED ARITHMETIC

(iv) Take 74 as the new dividend, subtract from

it, the product of the index and the previous
Onerla- (exo Te
Divide 71 by the working divisor 1.e.8 to get
Q = 8; R = 7. Place 8..below. 4 and_7
between 4 and 0 as shown.
(v) The new dividend is 70; subtract from it the
product of | and 8 i.e. 1 x 8 = 8 so that 70-8=
62. Divide it by 8 to give Q=7 and R=6
and place them as shown.
(vi) Take 67 as the new dividend and subtract
from it, (index x previous Q)= 1x 7= 7. This
gives 67 - 7 = 60. But now, we have already
covered all the dividend digits to the left of
the vertical line so that the above 60 goes to
the RHS of the vertical line as the final
remainder. Thus Q = 387 and R= 60.
(Of course , because of the detailed explanations, the method appears longer than
what it really is).
Comparing this method with the regular division in which we would have the
double digit number 81 as the divisor with the attendant multiplication and subtraction
with numbers having large number of digits, we have only the single digit divisor viz. 8
to handle, thus reducing very considerably the laboriousness of the operation as well as
the time.
Also, the present method scores over the earlier two methods viz.the complement
and the C-S methods, in terms of its simplicity.

Let us take another example.

Eg. 246098 + 72
(i) 24+ 75 Q=3,R=3
(ii) 36-(2x3)=30
and 30 + 7:5 Q=4,R=2
(For short, we shall write these two
steps as 36-(2 x 3) = 30+7 = 4/2. = 3418, R=2
(iii): 20. -(2 x 4) 2 + 7 = 1/5
(ivy) D932 x lai = sit
(vy) 18-2 x8) =" 2 eR
 DIVISION vi

Eg. 1480 + 32
G) 14+3 > 4/2
(ti) 128 =(2 <4) = 20.+ 3:=6/2
Gil) 420-2 <6) =S8=R.

Eg. 67+ 71 (to 4 decimal places).

GO Tie9/4
(ii) °40-9=31-+7 =" 4/3
(it) 30 -4=26.2 7 = 3/5
iv) -0=35=47+ 7 =-6/5
(v) 50-6=44+ 7 = 6/2

Eg. 4267 + 63
(i) 42 + 6 =7/0
so that the new dividend is 06. Now
06 - (3 x 7) = 6-21 which is negative
“. We take.42 + 6 = 6/6 tee OR Wind ew (os
thus reducing the Q from 7 to 6 and
increasing the remainder from 0 to 6.
(ii)- 66 - @x'6) = 48. +6 =8/ 0.
(iii) This gives us the new dividend as 07.
Again 07 - (8 x 3) is -ve.
.. Wetake 48+ 6=7/6 i.e. decrease
the Q to 7 and increase the R to 6.
(iv) 67-(3x7)=46= R since we have
exhausted all the digits to the left of the
vertical line.

** A Variation using Vinculum

The above method incurs a few wasteful calculations involving -ve values. To
avoid this, we adopt the vinculum method as illustrated below which avoids this waste
and is therefore more economical and faster.
Let us take the previous example and adopt this new method.
78 SPEED ARITHMETIC

(i) 42+6=7/0
(ji) 06-21= -15+6= 2/3
(iii) 37-3( 2)= 37+6 = 23
Thus,we have Q=7 2, R= 23 ie. Q=68,R=-17
To make R positive, we subtract | from Q and add 63 (=1 x 63 i.e. 1 x the divisor)
toR
to give Q = 68-1=67 andR=- 17+
63 =46
Thus, Q = 67, R = 46
One can see. how simple and quick, this method is.

Eg. 610 + 62 (to 3 decimal places).

20; ae Oe Oe

(i) 6 +6=1/0
Gi) Ole 2.1) fo = le 6 ==)1,
(iii) 10-(2x0) = 10+6= 4
(iV S02 er ce re Ol,O
(v) 20-(2x 6) = 8+6= I2
(vi) -h20.-"(2 xe SS 6S 53/0
.. Answer = 10.1613
= 9. 8387
9 . 839 (to 3 decimal places)

Eg. 493.76 + 74 (to four decimals)

(i) 49+ 7/0

(ii) 03-4x7 = -25+7 = -3/-4
 DIVISION 79

(iii) 47-4x 3 = AT ENDL 3927 So 3po

(iv) 06-4x 3 = 18+7 = 2/4
(v) 40-8 = 32+7= 4/4
(vi) 40-16 = 24+7 = 3/1 Se Atiss => 7 ee 024" nGro4
Let us now consider a few examples having 3-digit divisors.
Split the divisor so that 1 and 3 are the 2 flag or the index digits and the remaining
one viz., 8, the working divisor. As before, draw a vertical line in the dividend so that
its RHS has the same number of digits as in the index.
Eg. 7039+ 813

(i) 70 + 8 =8/6
(ii) Now take 63 and subtract from it the product
(first digit of the index x first Q - digit) i.e. 63 - (1 x 8) =55

(ii1) Divide this result by 8 to give 55 + 8 = 6/7

(iv) From 79, subtract the sum of the cross-product digits formed by the index and the
two recent Q-digits so that

(v) SS
|baleen = -14=
ee
TO A113 = 7T9%- (6% 1+8 x3) = 79-30 = 49-8 = GM

14+8= 1/6
66
(since 66 are the two most recent-digits)

(vi) eae |” 6 0-( 1+18) = 60-17= 7728 = 95

6
(vii) 50) ee) I WI o| ' a \o|+ YI II n| o| i} at N| II 00 | co HT}
| | ~~ a
19
Answer = 8.66194 = 8.65806
In place of decimals in the answer, we may write the remainder as :
639 - A x 10 - ffie (the remainder digits after the vertical line ) - ten times (the
0 8
 80 SPEED ARITHMETIC

cross product of the flag-digits and the quotient digits before the decimal) - (the last flag
digit x the last obtained Q-digit). This in our example equals 639 - 80 - 24 = 535 so that
the answer is Q = 8, R= 535.

Eg. 1872 +144

(i) 1+1=1/0
(ii) 08 -(4x 1) =4+1=4/0
(iii) pele =7-20 = -13
14
13s
1 3/0 (a2-digit quotient)
.. we take 4+ = 3/1

(vy iT
al = 17 =16 =1+1=1/0
13 = ——
(v) 02 ibe 16= 14+1= 1.4/0 (a2 digit quotient)
13
*. we take" L-=-1-= 0/1:
(vi) 12- ea = 0
230)
ee) hose0):
Eg. Let us take the previous example with the working divisor having two digits viz. 14.
(i) 18+14 = 1/4
Gi) 47-4x1= 43.+14=3/1
(ii) abe 4x Sue)
: O= 13, R=0
Evidently, the larger the number of digits in the working divisor, the quicker are
the calculations and vice versa.
Eg. 672 +412 (to 2 decimal places)
(yt 67 © 4 1S 21/26
(ii) 262-2 = 260+41 = 6/14
(iii) 140-12 = 128+41 = 3/5
(iv) 50-6 = 44+41 = 1/3
.. Ans. = 4263
or Q =I and
R = 262 - i 260
1
 DIVISION ~ 81

We now consider a few examples in which the original divisor contains digits which are
greater than 5. In these cases we convert the relevant digits to the corresponding
vinculums and then perform the division as above.

Eg. 583. + 38 (to 2 decimals)

The divisor 38 is written as 4 2. We can now carry out the division with the
modified divisor as we have done above.
GQ) Pras S11
(ii) 18- 2x1 =18+2=20+4=5/0
(iii) 03- 2x5=34+10=13+ 4=3/1
(iv) 10- 2x3=10+6=16+4 = 4/0
(v) 00O- 2x4 =8+4=2/l
- Ans. 15.34 or Q=15 and R= al 2 = 03+10 = 13
5
Eg. 31254 + 478
478 =522

(i) 31+ 5=6/1

(i) 12- 2x6=12+12=24+5=4/4
(iii) 45- be? |= 65+ 5 = 13/0
64
(iv) 04- tap SOF bes
4 13
(v) 30- ee = 70+5= 14/0
oa
(vi) 00- le J 42 +5=8/2
714
(vii) 20- Fawt GAb5e
Af914
4 8
(viii) 40 -& Zi 80 +5 16/0
8 12

7
82 SPEED ARITHMETIC

(ix) 0 - L5 El tguiod
{2 16
Ans. = 6 ge ee eae
I I iy ales
= 65,3849371
bt Oe Gand R= 454 i)5 2 |xi0- B 2454 ete o = borane
Ay 4
662 > 478 (the divisor) we have.
Q= 644+1=65, R= 662 -(1 x 478) =184

Let us now take up divisions with 4 digit divisors.

Eg, -66550.+'3025
There are again two ways of handling this division: (i) take 30 as the divisor (since
it is not difficult to deal with 30) and 25 as the index or (ii) take 3 as the divisor and 025
as the index. Let us illustrate both of them.
The procedure here is exactly as before.
(i) 66+ 30° =2/6
i) 63 x2) =6l- 30 =
(iii) 15 - P| S130 = 0/1
22
(iv) 10- [25]=0+30=0/0
20
». O.=22 and R=150- Bi x ele 150- 140-10 = 0
22 2
Alternatively, we may work out the same example as follows.

Ee. 66550 + 3025

G) 6+3=2/0
(ii)06-(0x2)=6+3=20 ~
(iii)05-| 02 |= 1+3=0/1
22
(iv) 15- gal = 15-(04+ 1044) = 143 =0/1
20
 DIVISION ~ 83

fy), 0 = 11025 = 310 = 10.=0

200
-. Q=22 and R= 0550- 025) x 100- 3 |x 10- |
022 22 2
= 0550-(400 +140 +10) = 0
Note: In step (ii), we subtract from 06, the product (1 st digit of the index and the Ist Q-
digit )so that 06 - (0 x 2) =6
In step (iii), we subtract from 05, the cross product formed by the first 2 digits of
the index and the first two Q digit so that we have 05- |02 |= 05-(0x2+2x2) =1
2D
In step (iv), we subrtract from 15, the cross product formed by the 3 digits of the
index and the first three Q-digits i.e. 15 - 22 i]= 15- (0x0 + 5x2 + 2x2) =15-14=1
220

In step (v), we subtract 10 from the cross product of the 3 digits of the index and
the last obtained three Q digits so that we have 10 - I2 ‘|= 10 -(04+5X2+0) <0.
200

In problems where we may have to continue the division, we proceed as in the last
step (v) ; viz. from the last obtained remainder subtract the cross-product found by the
three digits of the index and the three last obtained Q - digits.
We can similarly have divisors with indexes having 4 or 5 digits; in these cases, the
cross-products are the same as the schemes we have given in chapter 3 on
“multiplication”.
Eg. 2314 + 34022
In this example, we take the number of digits in the index as four and therefore
allowing for four digits in the dividend, we see that the vertical stroke seperating the Q
and R digits coincides with the line at the right of the divisor.

()-23 + 3 = 7/2
Gi) 21-40 |= 743= @/ 1
07
84 SPEED \RITHMETIC

(iii) Pea 44+8=2+3=0/2

072
(iv) 2 2 ea L
O72 3
(v) 00- ee cme ae 1 8+3= 6/0
7 202
(vi) 00- naa os rte 28 +3 =9/1
2026
(vii) 10- [40 2 2]=10- (6 +4) =-30+3= 1 0/0
0269
(viii) 00- [4 02 2 7]=-(-40+4-12)=48+
3= 16/0
ete 10
SUATEC SeOROT I vbr Dune VE:
tf
= 0.068014816
Eg. 145 + 8989; 8989 = 1 10 1 1 (The number of digits in the index is 4 and
accordingly we take 4 digits 0145 as the dividend and draw the vertical line as shown.
Qe IlGal Ob 2 de Ae Oe) Dice Se eee

(i) 1+1=1/0
(ji) 04- 10 =5+1=5/0
01
(iii) O5-[ 10 1 ]=10+1=10/0
015
(iv) 00- Es il i =-(-10-1)=11+1=11/0
015 10 ~

2
1
 DIVISION

(vii) 00- ho O01 Ere ae ee Bi

ih eke
(viii) OO-| 1 0 1 1] =-(-53-11-17)=81+1=81/0
fete 32°53
—uAns. SU ahd oF oe Sohal
Pyliey ilwy ke ars
= (016130811
or Sa ee es aia rece balone
0000 006 00 0
R= 145.
Finally, let us take an example with a divisor having 6-digits.

Eg. 68997 + 414321

P30 Oy aSa Uaioe Sf a0) 0. 4)
4 I Galion 0 ty 0: 2

(i) 6+ 4=1/2
(ii) 28-1x1= 27+4 = 6/3
(iii) 39-| 14 |= 29+4 = 7/1
16

(iv) 19- Eclone ae een teal yee

167
(v) 37 berate Oe E18) 31-45 37 4+.4,5:
1 = 72+4= 1 7/0
(vi) ieee 372 73 Omi
lei) +10 421) = 5 408. t/a
6 7 cs et
10), 10 Adit adel -10-(-1+6-68+
14-9) = 48+4=12/0
op criesDeel a7
86 SPEED ARITHMETIC

sede iE tgs = 42+4 = 10/2

oa: | = 0-(124+7-4-6-51)
ae ee
gaMfiseg=e Us IP Eom ed eer 2fe
ara
II 0.1697 4.7030 eK

= 0,.16653030

Exercises
Perform the following divisions using the VCP method
6(d) 2 and 3- digit divisors
(1) 93752/689 (2) 987654/877 (3) 86352/27
(4) 395627/84 (5) 3925/16 (6) 41903/96
(7) 10836/98 - (8) 10836/113 (9) -938256/876
(10) 938256/965 (11) 57321/437 (12) 12345/609
(13) 93752/1657 (14) 267354/758 (15) 101526/873
(16) 728316/318

6(e) 4- digit divisors.

(1) 103649/4874 (2) 763591/1234 (3) 920165/5252
(4) 3658329/8765 (5). 294965/3927 (6) 646942/3491
 87

CHAPTER 7

DIVISIBILITY

We have already considered in Chapter 1, some of the tests of divisibility of

numbers by small divisors like 3, 9, 11, etc. We have also noted that no such tests are
provided for numbers like 7, 13, 19 and so on.

We now present a comprehensive and a general procedure which can examine the
question of any large or small number being divisible by any other number.
The procedure rests on what is called osculation and employs the tool of
osculators.

7.1. Method of Determining Positive Osculators.

There are two types of osculators, one a positive osculator and the other a negative
osculator for a given divisor. Here we will describe the method for finding the positive
osculators and take up the case of negative osculators later.
(a) For divisors ending in 9, the +ve osculator is one more than the remaining digits of
the divisors.
Thus 9, 19, 29, 59, 119 have their osculators (0+1)=1, (1+1)=2, (2+1)=3, (5+1)=6,
(114+1)=12.
(b) For divisors ending in 3, we multiply them by 3 so that their ending digits are 9; and
then apply the above rule for finding the osculators.
Ee. 13,23, 33, 63, .193°on multiplication by 3 gives 39, 69; 99, IX ard 579 so
that the respective osculators are 3+1=4, 6+1=7, 94+1=10, 18+!=19 and
57+1=58.
(c) For divisors ending in 7, we multiply them by 7 so that the resulting ending digits
are 9 and then apply rule (a).
Eg. 17, 27,97 on multiplying by 7 gives 119, 189, 679 and the respective positive
osculators are (11+1) = 12, (184+1)=19, (67+1)=68.
(d) For divisors with ending digits 1, we multiply by 9 and then apply rule (a).
Beso Ady 3t61, 121 on multiplying by 9 give 99, 279, 549.,.1089.so:that the
corresponding osculators are (9+1) = 10, (27+1) = 28, (54+1) = 55 and
(1084+1) = 109.
(e) For divisors ending in 0, 5 and even numbers, we divide them by suitable powers
of 2 or 5 until the final digits end in 1,3, 7 or 9 and then apply one of the above
appropriate rules.
88 SPEED ARITHMETIC

Eg. Divisor 35; when divided by 5 gives 7 so that its osculator by rule (c) is 5.
Eg. Divisor 44 on division by 4 gives 11 so that its osculator by rule (d) is 10.
Eg. Divisor 310, when divided by 10 gives 31 and its osculator by rule (d) is 28.

7.2. Method of Osculation Using Positive Osculators. —

(a) Chain method
Let us start with some simple examples to demonstrate how osculation is carried
out.
Eg. Is 434 divisible by 7?
The osculator of 7 is 5. Using it, the process of osculation for the dividend 434
runs as follows:-
(1) Multiply the last digit 4 by the osculator 5
and add the remaining digits 43 to the product. 43+(4x5) 63
(ii) At this stage we know that 63 is
divisible by 7 and therefore we can conclude
that 434 is divisible by 7; or we may proceed
further as in step (111) below.
(1) Take the last digit (1.e. 3) of 63, multiply by the OH Ge SPST
osculator 5 and add the remaining digit 6 to the
product.
(iv) Multiply the first digit | of 21 by the osculator DE IG RESy = 7
5 and add to the last digit.
The resulting number is the same as the osculator itself viz. 7. We therefore
conclude that the number 434 1s divisible by 7. .

Eg. Is 266 divisible by 19?

(The osculator of 19 is 2)
(i) 26+(6x2)=38
Gi) 3+(8x2)=19 (Same as the divisor)
. Divisible.

Eg. 431 by 23
(The osculator of 23 is 7)
i> ‘ASC1L xR) =L5O
“

Gi)-2 SF (OX 7) 25
In this example, the last number is 5 and it being less than the divisor (23) and not
being a sub multiple of the same, we say that 431 is not divisible by 23.
 DIVISIBILITY 89

Let us take another example with the same divisor 23 whose osculator is 7.

Eg. 1748 by 23
(i) 174+ (8 x 7) = 230. Since 230 is divisible by 23, we stop the osculation and
conclude divisibility.
Let us, however, proceed.
(ii) 23+(0 x7) = 23, same as the divisor; therefore we stop the osculation and
conclude divisibility.

Let us now summarise the circumstances under which an osculation may be terminated.

Rules for Stopping the Osculation

These rules relate to the divisors which are odd numbers.
(i) The osculation may be continued as long as the osculator is larger than the
divisor.
(11) If at any stage of continuing with the osculation, we have the divisor itself
or zero or its multiple or its submultiple, the osculation may be stopped
concluding that the dividend is exactly divisible by the divisor.
(111) The osculation may be stopped if it results in repeated values; in such a case,
we conclude exact divisibility.
(iv) We can stop the osculation whenever we reach a small number which by
mental division of the dividend or otherwise can let us know if divisibility is
possible or not.

It is of interest to find that if we osculate a number by its own osculator, we will

end up with the same number or its multiple or sub-multiple.
Egs. (1) 71 osculated by its own osculator 64 gives 7 + (1 x 64) = 71 (the number
itself).
(2) 46 osculated by its own osculator 7 gives 4 + (6 x 7) = 46
(3) 245 by its own osculator 5 gives 24 + (5 x 5) = 49 (sub-multiple of 245)

(b) Another method .

Here the osculator is carried out using single.digits for forming products.
Let us illustrate by means of a few examples.

Eg, Is 3795 divisible by 23°?

(i) Determine the osculator of the divisor by any of
the above rules. Here the osculator of 23 is 7.
90 SPEED ARITHMETIC

(ii) Osculate the last digit 5 by 7 and add to the next Ste BAGDAD
previous digit 9 to get 44. Ad

(iii) Osculate 44 by 7 to get (4x 7) +4 = 32. Addsthe BO) F PYO: 5

next previous digit 7 to it to give 32+7 = 39. 39 44

(iv) Osculate 39 by 7 to get 3 +(9 x 7) = 66. Add the

next previous digit 3 (i.e. the first digit in our 3 VPOROHDS
case) to have 66+3 = 69. 69 39. 44 5

Now 69 is divisible by 23 so that the given

number 3795 is divisible by 23.

(v) On the other hand, if we cannot mentally verify the last step, we can osculate
the last obtained number 69 by the osculator (7) to get 6 + (9 x 7) = 69, thus
repeating itself and confirming divisibility. If we apply the chain method to
this example, we would have
3795> 379+ (5x7) =414- 41+ (4x7) =69 564+(9x7) = 69.

Eg. 32451 + 49
(i) Osculator is 5. Osculate 1 (the last digit) a ges Seay tee
as usual so that (1 x 5) =5 and add to 5 A027) Sele
to get 10 .
Gi) Osculate 1047 1+(0x5)=17341+4=5
Giii) Osculate 5=05 50+ (5 x5) =25 3 25+2=27
(iv) Osculate 27 — 2+ (7x 5)=37 7 37+3=40
Since 40 < the divisor 49 and is not divisible
by 49, we conclude that 32451 is not divisible
by 49.

Eg. 6407 + 43
(i) Osculator is 13. Osculate 7 by 13 and 6 4 0 aS
add zero to it .{(7x13)} +0=91 86 §626 91
(ii) Osculate 91 and add 4 — 26
(111) Osculate 26 and add 6 —> 86
As 86 is divisible by 43
Number divisible by 43.
 DIVISIBILITY 9]

Eg. ‘T488 32
(1) The osculator of 32 (an even number) is |
Gi? Aad SAK O= 9
and the osculator is 1)
Gi) (8x1)+8 =16
Gu) 1+(6x1)75 74+4=11 7 7 aa Sa
(iv) (lx1l)4+41=23524+7=9 oO M6
(v) The last number 9 is the same as the osculator.
*. Divisible.

7.3. Method of Finding Negative Osculators.

It might have been noted that the method of determining +ve osculators may end
up with large osculators . For example , the osculator of 27 is 19, of 31 it 1s 28 and so on.
This would render the process of osculation tedious and make mental calculations
difficult.
The question therefore arises if it is possible for such divisors to have osculators
which are smaller. Fortunately, it is possible to find such osculators and these are called
negative osculators. They are so called because the osculation using these osculators
involves subtraction instead of addition. For convenience, we symbolize the +ve
osculators as P, (for plus) and the -ve ones as M (for minus). The method of finding a -
ve osculator is :
(1) in respect of divisors ending in 1, we simply drop the last digit, namely the unity
and take the remaining digits as the -ve osculator; for example, for the divisors 21,
zs te glpideWA Come Ao , the -ve osculators are respectively 2, 4,5, 12, .......

In respect of other divisors, multiply the divisor by a suitable number so that the
product ends in digit | and then apply rule (1) above. Thus,
(2) for divisors ending in 3, multiply by 7.
Examples: 13) 25,49) sac.-.53 we have (on multiplication by 7) 91, 161, 301, .......
and the corresponding -ve osculators are 9, 16, 30, ......

(3) for divisors ending in 7, multiply by 3. Examples: 7, 17, 37, 57,......We have, on
multiplication by 3, the products 21, 51, 111, 171 and the corresponding -ve
osculators are 2,5, 11 and 17.

(4) for divisors whose ending digit is 9, we multiply by 9. Examples: 9, 29, 39, 109, the
products by 9 are 81, 261, 351, 981 and thus the -ve osculators are 8, 26, 35, 98.
92 SPEED ARITHMETIC

Two things might have been noted from the above examples.
(a) In all the cases, the sum of the +ve and -ve osculators equals the divisor.
(b) For divisors ending in | and 7, the -ve osculators are smaller than the +ve ones,
whereas, for those ending in 3 and 9, +ve osculators-are smaller than the respective
-ve osculators. Thus, for osculation purposes, we may adopt the -ve osculators for
divisors ending in | and 7 and +ve osculators for divisors ending in 3 and 9.

Finally, for even divisors and those ending in 5, we may, as before, divide them by
appropriate powers of 2 and/or 5 so as to result in 1,3,7 or 9 as the ending digits and
then apply the above rules for finding the -ve osculators.

** 7.4, Method of Negative Osculation.

Multiply the last digit of the number by the osculator and subtract the product from
the remaining digits. (In +ve. osculation, we add the product to the remaining digits).

Bas, (lye “7s osculated. Dy: 7 3= 07 o(30/) 5, aed

(2) 450 oscualted by5 = 45-(0x5) = 45
(3) 3743 osculated by 9 = 374-(3 x9) = 347
(4) -642 osculated by3 = -64-(-2 x3) = -58

Let us now take up some examples to illustrate the use of -ve osculation to test
divisibility of a number.
Eg. Is 4346 divisible by 41 ?
Since the divisor ends 1n 1, we choose .
the -ve osculator viz. 4.
(i) Osculate 6 by 4 i.e. 6 x (-4) to get - 24 Al BS. 6s
and add to 4; write the sum-below 4 -20

(11) Osculate -20 by 4 to get -2-(-0 x 4) 4 3 4 6

= -2 and add it to 3 to get 1 ies

(111) Osculate 1 by 4 to get -4 and add to ie aa Sis

4 which gives 0 0 1. -20

Since the last number is 0, wé conclude that 4346 is divisible by 41.

Alternatively, the method of chain osculation gives
4346 > 434-(6x4) = 410 41-(0x4) = 41 54-(1x4) = 0,
 DIVISIBILITY

Eg. 345681 by 27?

(Since the divisor ends in 7, we choose
the -ve osculator which is 8).
(i) Osculate 1 = 01 by 8 i.e. 0- 1 x 8 and add dake vile 69878
to the previous digit 8 and write the 0
addition below 8.
(11) Osculate 0 by 8 to give 0 - (0 x 8) = 0 Le is)
and add to 6 to give 6. 6 0
(iii) Osculate 6 by 8 (= -48) and add to 5. RE RO ERT:
to obtain - 43 43-56, 0
(iv) Osculate - 43 by 8 so that Ad Saoz6ib 8
-4-(-3 x 8) = 20 and addto 4 AmeAse 6 0
to result in 24
(v) Osculate 24 by 8 to give 2- (4 x 8) ii Maal oy:
= - 30 and add to 3 to give -27 243 6)
The last number (-27) is divisible by the divisor 27 and therefore the given number
is divisible by 27.

Eg. Divisibility of 3239998 by 311.

Again we choose the -ve osculator 31 for
the reason that the divisor ends in 1.
(i) O-(8 x31) = -248+9 = -239 Oe ae Sn0iGuseO quid
-239
(ii) -23-(-9x31) = 256+9 = 265 Saag 4 3 pr eS Rael A’
265 -239
(iii) Bftai = -129+9 = -120 CMe ey, 3 RE ee)
-120 265 -239
(iv) -12-(-0x31) = -12+3 = -9 ee 2 EL a Me)
-9 -120 265 -239
(v) 0-(-9x31) = 279+2 = 281 edi tations faeIPs!
281 -9 -120 265 -239
(vi) 28-(1 x31) I= -3+3 =0 Sih Ae i eines jai” ae
0 281 -9 -120 265 -239
Since 0 is the last number, the given number is divisible by 311.
94 SPEED ARITHMETIC

The chain osculation gives 3239998 —>.323999 - (8 x 31) = 323751 — 32375 - (1 x 31)
= 32344> 3234-(4x 31) = 3110 = 311 CO XBDY = 311-931 (Px =0

Eg. Is 57532 divisible by 68 ? ( 68 is an even number.and so we divide by 4 and get

17; and since it ends in 7,we take the -ve osculator as 5).
5 7 5 3 ps
-200 Il 40 -7
Since the last number is not zero or the divisor or a multiple of the divisor, we
conclude that the number is not divisible by 68.
| The chain osculation gives
57532 >' 5753 -(2 x 5) = 5743 — 574-( x5) = 5595.55 -(@ x5) = od
(< the divisor). Therefore we stop the process here with the same conclusion.

Fe Js 410230 divisible by 710?

(710 is an even number, so we divide by 10 to get the odd number 71;
this ends in | and so we take the -ve osculator as 7). The result of osculation is
st I 0 y 3 0
53 -7 C218 sD 3
For the same reasons as in the above example, the number 1s not divisible.
Chain method: 410230 > 41023 - (0x7) = 41023 — 4102 - (3 x7)
= 4081 — 408 - (1 x 7) = 401 — 40- (1 x 7) = 33 (< the divisor) . Therefore the chain
stops here with the same conclusion as above. =F

7.5. Group Osculation.

Until now we have considered the test of divisibility of numbers where the divisors
and consequently the osculators are not very large. We now extend the method so as to
include large divisors and hence large osculators. This is done by lumping the digits into
appropriate groups and considering them as osculators unlike in the earlier method
where we osculated using only the single ending-digit of the divisor. The process of
osculation using group osculators may be called the group osculation.

Eg. (1) Letus osculate 765 by 9 for 2 (right-most) digits.

The +ve osculation gives 7+ (65 x 9) = 592
and the -ve osculation give’ 7 - (65 x9) = -578.
(2) 9428 osculated by 11 for 3 (last) digits
+ve osculation: 9 + (428 x11) = 4717
-ve osculation: 9 - (428 x 11) = -4699
 DIVISIBILITY 95

(3) 625 osculated by 7 for 4 last digits

Since 625 has 3 digits, we take it as 0625 for osculation.
+ve osculation : 0 + (0625 x 7) = 4375
-ve osculation : 0 - (0625 x 7) = -4375
(4) 539 osculated by 11 for 5 last digits
As above, we consider 00539 for osculation.
+ve osculation : 0 + (00539 x11) = 5929
-ve osculation : 0 - (00539 x11) =- 5929 taking 539 = 000539

7.6. Types of Divisors in Group Osculation and Their Corresponding Osculators.

We may encounter the following 3 types of divisors in group osculation:-

(1) Dinisors: ending in-9 or a-series of 9 ‘s like 99,999 cfc. Here, thespositive
osculation must be employed.
(2) Divisors ending in | or a series ending in I| like 01, 001, and so on. In this case, -ve
osculation is the one to be used.

(3) Divisors not belonging to (1) or (2) above but may be multiplied by suitable
numbers so as to bring them to type (1) or (2) above.

(a) Osculation of the first type.

Eg: Let the divisor be 6999. Then by the above rule (1), we do the osculation by a +ve
osculator i.e. by P. But since the divisor has three 9’s as the ending digits, we carry out
the +ve osculation in groups of 3. For short, we shall indicate the corresponding osculator
by the symbol P; which equals (6 +1) =7.
Let the number to be tested for divisibility by 6999 be 4934295.
(1) We group the given number into 3’s starting + D3 diniin295
from right
(11) Osculate 295 by 7 (the last groups of 3 digits) 4 934 295
to get 295 x 7 = 2065; add 934 to give 2999 2999
(iii) Osculate 2999 by 7 (in group of 3 digits so 4 934270295
that we have 2 + (999 x 7) = 6995 and adding 4 6999 2999
to it gives 6999 which 1s the same as the divisor
“. Divisible.

The chain method of osculation gives 4934 + (295 x 7) = 6999.

96 SPEED ARITHMETIC

Eg. Is 74645093455 divisible by 49999?

Since the divisor has 4 9’s at the end, we have,
in our usual notation, Py =5.
(i) We group the divisors in 4 digits “S746 4509 3455
starting from extreme right
(ii) Osculate 3455 in group of 4 digits 746 4509 3455
by 5 and add 4509 to it to give 21784
{(3455 x 5) + 4509} = 21784
(iii) Osculate 21784 in group of 4 digits 746 4509 3455
by 5 and add 746 to it 9668 21784
{2 + (1784 x 5)} +746 = 9668
Since the last number is 9668 (< the divisor), and is not a submultiple of the
divisor, we conclude non-divisibility.
The chain method of osculation gives
7464509 + (3455 x 5) = 7481784 — 748 + (1784 x 5) = 9668.
(b) Osculation of the second type .
Eg. Suppose the divisor is 16001. We have as ending digits two 0’s followed by 1. We
therefore group the given number in 3’s. Also, since the ending digit is 1, we employ the
-ve osculator. Put together, we have, in our symbols M3 = 16 as the osculator.

Let us now apply this to test the divisibility of 69124320 by the divisor 16001.

(i) We group the dividend in groups of 3 digits 69 124 320

(11) Osculate 320 in 3-digits group by the -ve

osculator M3 = 16 to give - (16x320) = -5120 69 124 320
and add 124 to it which gives -4996 -4996

(iii) Again osculate -4996 in 3-digits, group by

M; = 16 to get - 4 - (-996 x 16) = 15932; 69 124 320
adding it to 69 gives 16001 which is the 16001 -4996
same as the divisor.
“. The number is divisible by 160Q1.
The chain Osculation gives : 69124 - (320 x16) = 64004 — 64 - (004 x 16) = 0.

Eg. Is 7005945219 divisible by 90001?

We have the osculator My, = 9
(i) Group the digits in 4’s 70 0594 5219
 DIVISIBILITY eal

(ii) -(5219 x 9) + 0594 = - 46377 57459 -46377

(iii) {- 4 - (- 6377 x 9)} + 70 = 57459
.. Not divisible

c) Osculation of the third type.

Eg. Consider the divisor 211; This does not conform to either of the two previous types;
however, on multiplying it by 9, we get 211 x 9 = 1899 which comes under type 1. Thus,
we have P> = 19.
To test the divisibility of any number, say 91322699
by 211, we proceed as before.
(1) Group the dividend into digits of 2. 91 Ae 2 Orin DD
(ii) Osculate 99 by P2= 19 and add 26: 91 32 20 $i 199
(99 x 19) + 26 = 1907 1907
(iii) Osculate 1907 by 19 in 2-digit
group and add 32 of oe 2 YO
{ 19 + (07 x 19)} +32 = 184 184 1907
(iv) Osculate 184 by 19 in 2 digit group oh De. 20, 5.99
and add 91: {1 +(84x19)}+91=1688 1688 184 1907
It is easily seen that the last number 1688 is a multiple of 211 (1688 = 211 x 8).
.. Thus the given number is divisible by 211.
The chain osculation gives
913226 + (99 x 19) = 915107 > 9151 + (07 x 19) = 9284 — 92 + (84 x 19) = 1688.

Eg. Is 315699437 divisible by 359?

Now 359 x 39 = 14001 so that the corresponding -ve osculator is M3 = 14.

(1) Form groups of 3 digits 315 699 437

(ii) -ve osculate 437 by 14 and add 699:

- (437x14)+699 = - 5419 eH: 699 437
- 5419
(iii) -5-(-419 x 14)} +315 = 6176 oybe 699 437
. 6176 - 5419
Not divisible.
The chain osculation :
315699 - (437 x 14) = 309581 — 309 - (581 x 14) = -7825 — -7-( -825 x 14)
= 11543 > 11- (543 x 14) = -7591 > -7-(-591 x 14)

28
98 SPEED ARITHMETIC

= 10217> 10- (217 x 14). = - 3028 > (- 028 x 14) = 389.

-3-

Since the osculation continues indefinitely without arriving at the same value as the
divisor or its multiple or zero , we conclude “ Non divisible”.

Eg. Is 95151873 divisible by 359?

(osculator is M3 = 14 as above).

(i) {- (873 x 14) + 151) =- 12071 0) 15] 873

1077. -12071
(ii) {- 12 -(-O71 x 14)} +95 = 1077
and: 1077 is a multiple of 359 (1077 = 359 x 3)
The given number is divisible by 359.
The chain osculation gives: 95151 - (873 x 14) = 82929 —
82 - (929 x 14) =- 12924 — -12 - (-924 x 14) = 12924
. the value repeats; hence divisible.

A short - coming: — It might have been noted from the above examples under type (c)
that it is not always a simple task to find a suitable multiplier of the divisor which will
render the latter to have a string of 9’s or a string of 0’s followed by 1; and if we do not
have a sufficiently long string, the osculator value can be very large. For example, if we
have 3589 as the divisor and multiply it by 9 to get the end-digit 1, we would have the
product 3589 x 9 = 32301 so that the osculator would be a large value viz. Mz = 323.
Also, there are no integer multiples of 3589 which can give 001 or 0001 etc. The
multiples 109, 209, 309, etc. give products that end in 01, but make the osculators even
larger than 323 obtained above.

Exercise 7
Test if the following divisions are exact: (Stroke indicates division)
(1) 4125/55 (2) 4165/55 (3) 29859/81
(4) 10878/11 (5) 29859/111 (6) 17784/24
(7) 86862/42 (8) 36156/92 (9) 17846/48
(10) 7362109/3163 (11) 8652090/360 (12) 732468/973
*
 CHAPTER 8

FRACTIONS, PERCENTAGES AND RECURRING DECIMALS

8.1. Three Types of Decimals

We generally come across one of the following three kinds of decimals.
(a) Arecurring decimal i.e. a decimal consisting of un-ending digits which repeat
or recur themselves asin 1/3 = 0.3333..... or 1/9 =0.1111... or repeat in groups as in
1/99= 0.01010101.....or 1/333= 0.003003..... .Recurrence is often denoted by a bar over
the digits that recur or by means of dots over the beginning and the ending ene that
recur. Thus 1/3 =0. 3 or 0.3, 1/99 =0.01 or 0.01.
These decimals occur whenever the denominator of the proper fraction (or vulgar
fraction as it is sometimes called) has prime numbers like 3,7,11,13.... as factors.

gs. 1/3 =0.341/99= 1/3x3x11)= 0.01 or 1/7 = 0.142857.

In the case of a recurring decimal whose proper fraction has the numerator unity,
the product of the last digit of the denominator of the fraction and the recurring digits of
the decimal gives a series of 9’s; and in particular, the last digit of the product is 9. (This
property is important and will be used later).
Hes. Gy) W99-=-0l0lisaeithathoox 01 = 99
(ii) 1/7 = 0.142857 and 7 x 142857 = 999999

(b) A non-recurring decimal. This occurs whenever the denominator of a given

rational fraction, has 2 and/or 5 as factors. These decimals, unlike the above, terminate
after a certain number of digits. Also, every 2,5 or 10 in the denominator gives rise to one
significant digit in the decimal.

Bess Ah/d = O25 1/125 = 1/5° = 0.008; ee, 10.3: W/4s= 0.0625,
(1 digit) (3 digits) (1 digit) (4 digits)

1/2x5=1/10 = 0.1; 1/100=1/10° = 0.01; 1/800 = 1/(10?

x 2°) = 0.00125.
(1 digit) (2 digts) (5 digits)

*Note: Take for example 1/800, the denominator of which factors into 2x2x2x5x2x5x2
= 2°x5*; the way that these 2’s, 5’s, and 10’s can be combined so that the sum of their
indexes is minimum is 10°x2° and this gives rise to 5 significant digits in the decimal
form of 1/800.
100 SPEED ARITHMETIC

In the case of a (finitely ending) non-recurring decimal whose corresponding proper

fraction has one as the numerator, the product of the last oe of the denominator_and the
last digit of the decimal ends in zero.

Eg. 1/125 =0.008 and 125 x 008 ends in zero. 1/16 = 0.0625 and
16 x 0625 ends in zero and so on.
(c) Anon recurring and unending decimal.
This occurs when the numbers are irrational as in :
= LAdAQtO Ste... or’ in .constants likerqwj=: 331415920335 Sie4e3:::
In this chapter, we will be concerned, almost always, with the recurring aecimals.

8.2. An Easy Method of Converting Fractions to Decimals

(a) Fractions whose denominators are 9’s.
Lev usttaremtne cet tor tractions’ 1/97 17/99, L999 ea. which on multiplying by the
corresponding bases 10, 100, 1000,... gives the quotient 1 and remainder also 1.
Accordingly, the decimals will haye 1 as their non-zero digit as given below :-
1/9 = 0.1, 1/99= 0.01, 1/999= 0.001 ......
This gives a simple way of writing a fraction in its corresponding recurring decimal as
follows. j
1D eee OF
8/99 > 8x01 0.08; 61/999 61x01 0.61;
24/999 + 24x 001= 0.024 ; 919/999 > 919 x 001 > 0.919
and so on. (Note that, the number of significant digits in the decimals is the same as the
number of 9’s in the denominators of the fractions). )

(b) Fractions whose denominators are 3’s

Let: us consider thes fractions f/3,-1/33> 17333 424 The common feature of these
fractions is that multiplying each of them respectively by 10, 100, 1000, .... gives
quotient = 3 and remainder = | again. For this reason, the decimals will have 3 as their
digits. This helps us to find an easy method of conversion to decimals as follows:
We have 1/3 = 0.3 , 1/33 = 0.03 , 1/333 = 0.003 the number of significant digits in each
case being the same as the number of digits in the denominator of the corresponding
fractions. Now if we want to convert 2/3 to its decimal, what we do is to simply
multiply 3 by 2 and place a dot overthe product.
Thng,-23)-%2:x3'50:64¢p
2/33 > 2x03 = 0.06; 19/33 = 19 x 0.3=0.57;
17/333 — 17 x 003 = 0.657 and so on.
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 101

(c) Fractions whose denominators are 1’s

Considerthesrdcset: 1/11y W141) 4b ass: Multiplying them respectively by the
bases 100, 1000, 10000....... we have the quotient 9 and remainder | again as in the above.
So the decimals would carry the digit 9 giving 1/11=0. 09, 1/111=0. 609, 1/1111=0.0008.
(Again the recurring decimals will have the same number of digits as the number of 1’s in
the denominators).
Thus, 9/11 9 09 — 0.81, 9/TII > 9x 009 — 0.081,
107/111 — 107 x 009 > 0.963, 87/1111 > 87 x 0009 0. 0783
1040/1111 > 1040 x 0009 — 0.9360 and so on..

**8.3. An Easy Method of Converting Fractions to Percentages

Most often, students find the conversion of fractions into their respective
percentages difficult. We now present some simple methods of doing so, when the
fractions have 9’s., 3’s or |’s in their denominators.

(a) Fractions with denominators 9,99,999, etc.

oe. IIo.
(1) Multiply the numerator by 11
el be all
(ii) Annex the product to the given fraction 11 '/%
which is the answer required.
Eg. 7/9
(i) 7x11=77
(ii) 77 "bo%
Eg. 8/99
Simply annex the numerator to the given fraction. Thus, 8/99 = 8 §/99%

Eg. 43/99
As in the above, annex 43 to 43/99 to give 43 3/y9%.

Eg. 34/999
(i) Write the numerator 34 as 034; take the first two digits 03
(ii) Bring the first two digits after the last digit 403
(iii) Annex 03 to 403/999 to give 03 4“/o99%
= 343/55 %
Eg. 140/999
(i) 14
(ii) O14
(iit) 14 f/o99 = 14 fo99%
102 SPEED ARITHMETIC

Eg. 9517/9999
(i) Write 931 as 0931 and take the first two digits: 09
(ii) Bring 09 to the right of the remaining ES in the
numerator: al 3109
(iii) Annex 09 to 3109/9999 —> 09 31 fog = 9 71" /o999%

Eg. 7307/9999
(i) Take the first two digits. 1s
(ii) Write the first two digits after the last two digits: 0773
(iii) Annex 73 to 0773/9999 to give 73 7? lo999%
Eg. 10112/99999
(i) Take the first two digits : 10
(ii) Write the first two digits to follow the last 3 digits 11210
(iii) Annex 10 to 11210/99999 to give 10 1 /o9999%
We can repeat the above procedure for any fraction with a string of 9’s in the
denominator.

(b) Fractions with denominators 3,33,333 etc.

Eg. 2/3
(i) Multiply the fraction 2/3 by 1/3 213) X13 = 219
to give 9 in the denominator :
(11) From the previous section (on fractions
with 9 as the denominator)its percentage
is simply (2x11) + 2/9 = 22%
(iii) Multiply the results of step (ii) by 3 22 Io x3
(reciprocal of the multiplier 1/3 of step (i)) = (22x3) + (2/9x3)
which gives the required percentage = 66 7/3%
(or more simply, multiply the numerator 2 by 33 (since 33 x the denominator = 99) and
annex the product to the fraction 2/3 to give 66 */3%).

Eg. 17/33
(ie S17/33"RD/S S100
(ii) percentage = 17 '’/99 (based on the method of previous section).
(iii) (17 x 3)+ (17/99 x 3) =51 743%
(or, multiply the numerator by .(since 3 x denominator= 99) and annex the product 51
to the fraction 17/33 to get 51 '7/33%).
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 103

Eg. 14/333
(i) Write the numerator as 014 and multiply the
same by 1/3 (as in the above) 333.x%13 -=,014/999
(ii) Ol See ll Ae iss (based on the previous section)
Gu) (1 x 3)+( 401/999 x 3) = 3+ 401/333 = 3 +(14+ 68/333) = 4 68/33%

Eg. 140/333
(i) 140/333 x 1/3 = 140/999
(ii) 14 wean 5 = (4 as fxs (based on the previous section).
(iii) (14 x 3) + 14/999 x3 = 42 '4/433%
and so on for denominators 3333, ...........

(c) Fractions with denominators 1,11,111 etc.

Fg. Ti)
(1) Multiply the fraction by 1/9: TVA Oe 9
(ii) Percentage of 7/99 is-7 ’/o9
(from the section on fractions
with 9 as denominator).
(iii) Multiply 7 "/o9 by 11: (7 x 11)+(7/99 x 11)
= II
OR more simply
(1) Multiply the numerator by 9 iba as |
(ii) Annex the result of (1) to the
fraction to get the percentage 77 11;\%

he fst |
(i) 73/999 — 073/999
fii 207 oO ets (following the method under 9’s)
(iii) (07 x 9) + (307/999 x 9)
= 634+ 307/111 = 63+(2+ 85/111) = 65 ST

Eg. 107/111
(i) 107/999
(ii) 10 "°/o9o
(iii) (10 x 9) + (710/999 x9) = 90+710/111 = 90+6“%/,,,)
= 96 4/14;% seo
104 SPEED ARITHMETIC

8.4. Multiplication Method for Converting Fractions to Recurring Decimals.

(a) Using positive multipliers. Let us first consider fractions with denominators ending
in 9. We have already seen 1/9. We shall therefore take, up the conversion of fractions
like 1/19, 1/29, whose denominators contain only primes and also fractions 1/39 and 1/49
whose denominators have primes and non-primes as factors. It may be recalled that the
two special features of all these fractions are that one, their decimal counterparts are
recurring and two, the last digit of the recurrence cycle is 1 (since as we have noted, in
the beginning of this chapter, the product of the last digit of the denominator namely 9 of
the fraction and the last digit of the recurring decimal must equal 9).
Before we describe various short-cut methods, let us observe that the long method
in current use of dividing directly the numerator viz.1 by the denominator involves, most
often, as many steps as the number in the denominator less one. Thus the current method
of finding the decimal of 1/29 would involve (29-1) = 28 steps. Also, it involves,
invariably, the multiplication and subtraction of large digits. All these render the method
in vogue time-consuming, tedious and risky of committing errors.

In comparison, the methods that follow involve (i) only small digits in the
arithmetic operations, most of which can be carried out mentally (ii) one to three lines
and consequently are easy to carry out and most of all least time consuming.

Let us now describe the various short-cut methods by way of illustrations.

Eg. 1/19. The last digit of the recurring decimal is 1.

(1) Write 1 at the extreme right and place a bar ,
or dot above it to show that it is the last digit
of the recurring cycle.
19 = l
(11) The left-most digit of the denominator 19 is 1.
So the multiplier is one more than this digit viz.2.
(This corresponds to the +ve osculator of the last
chapter)
(iil) We multiply | of step (i) by this multiplier and
write 2, moving from right to left. Thus
19 = i 21
(iv) We now continue doing the multiplication until
we come across a 2-digit product. :
1/199 8421
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 105

(v) Now we have on multiplication of 8 by 2, the

2-digit number 16. So we write 6 on the top line
and write 1 on the line below the first line, to
indicate that it is to be carried over. Thus 3
ho s= 68421
l
(vi) We next multiply 6 (in the top line) by 2 to get
12, and add the carry over | (on the bottom line)
which gives 13, again a double digit number; }
so as before, we retain 3 on the top line and 3 68421
write 1 on the bottom line to be carried over. hed

We continue with this process, until we have reached 0 on the top line and 1
on the bottom line as below. :
1/19 = gen Diresheets
er Ree
1 1 LL phd) Ob 1 leet

(vi) If the next multiplication is made, we will have (2 x 0) + 1= 1, the same

number as at the beginning and any further continuation would throw up the
same digits. This suggests, that we have reached the recurring cycle with 0
at the extreme left and we therefore place a dot over it. Thus the fraction
1/19 has 18 (=19-1) digits in the recurring cycle starting with 0 and
ending in | as above.

It can thus be seen how elegant and quick this method is. It calls for no more than
multiplication of single digits (all of which can be done mentally) and entails just one -
line containing the answer.

As if this simplicity is not enough, there is a further short-cut to the above. It would
be seen that as we continue with the multiplication, we would come across 18 (the
denominator 19 minus the numerator 1) as the product (shown within the strokes //). This
indicates that we have reached the half way mark; and more importantly if we take the 9
digits that we have obtained thus far, viz. 947368421 and subtract it from 999999999 we
get the first 9 digits viz. 052631578 of the decimal. This would mean, that we need not
even continue with multiplication till we get all the digits of the decimal, but carry on
with multiplication only upto the half way mark and subtract the 9 digits got thus far from
a string of 9’s to get the first half .
106 SPEED ARITHMETIC

Eg. 1/29. (The normal method would involve 29-1= 28 steps of operations). As before,
we start with the right-most digit 1 and place a dot over it; continue to multiply left-
wards using the multiplier 3 (which is one more than the number to the left of 9).
Carrying out the multiplication on the same lines as before, we get,
L29. deed 4,£80) 9 Gri S Samed yuhetvA-gd faced. Panay i
/ /1 2 1 bei, 2
At this stage, we have reached the number 28 (29-1) which means that we have reached
one-half of the total multiplication required. Also we have 1/2 (29-1) = 14 digits viz.
9:6) 505 thel a Ack 37 Teeaek
This, on subtraction from a string of 14, 9’s gives the remaining 14 beginning digits viz.
03 44 St247 3 8.6 2076'S
Thus, the recurring decimal of 1/29 is:
129 = 0344827586206896551724137931

Eg. 1/39
(ends in 9 and therefore the last digit of the recurring decimal is 1: also, the digit
preceding 9 is 3 and therefore the multiplier is 3 +1 = 4). Proceding as above, we have :
30% yee’p) 6 Gur dginy
1 1
After 0 at the left, we can 1 which is the same as the’ one with which we started.
Therefore there is a repetition of digits after 0 and so the recurring decimal is 0.025641.
This, unlike the above ones, does not have 39-1 = 38 digits and does not exhibit 38 as
the half -way number. One of the reasons is that although the denominator ends in 9, it
is not a prime like 19 or 29. At any rate, we shall come back to this fraction again later.

Eg. 1/49
(the ending digit of the recurring decimal is 1 and the multiplier is 4+1 = 5).
By the above method of multiplication by 5, we can show that
1/49= 020408163265306122448979591836734693877551
(Left as an excercise)
[It may be verified that unlike in 1/39, we see here that 49-1=48 occurs as the half-
way mark (1.e.8 in the top line and 4 in the bottom line.)] However, the number of digits
in the recurring decimal is not 48 but only 42. Such a peculiarity occurs since once again
as in 1/39, the denominator 49 of thé fraction is not a prime but has 7 and 7 as prime
factors. We shall therefore return to this fraction after we have considered the recurring
decimal 1/7.
A similar procedure of multiplication can be adopted for converting, into recurring
decimals, fractions whose denominators end in 9.
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 107

Let us now consider fractions with denominators ending in 3. As a general rule,

the ending digit of the recurring decimal in such fractions is 3. Also, the multiplier can
be found in exactly the same way as we determined the +ve osculator in the previous
chapter, viz. to multiply the denominator by a factor such that the product ends in 9 and
take as the multiplier the penultimate digit increased by 1. Thus for 3, we have 3x3 = 9
and the multiplier is 0+1 = 1; for 13 we have 13x3 = 49 and so the multiplier is 4+1 = 5
and so on.

Ege t/3
(the last recurring digit is 3 and the multiplier is 1)
1/3 = 3/3 so that 3 repeats at the second multiplication itself.
1/3 =0.3.

Fg. WAS
(the last digit is 3 and the multiplier is 4)
LS = aD ee 223
Shier ates: 1 "
so that the recurring decimal is 0.076923
Again, it may be seen that since the denominator 13 of the fraction is prime, we find that
(i) the half-way mark is 36 which corresponds to the (denominator - numerator) of
the fraction 3/39 (the denominator of which we used for finding the multiplier);
and ;
(ii) the sum of the digits of the first half viz. 076 and of the second half, viz.923 is 999.

Eg. 1/23 .
(The ending digit of the recurring decimal is 3 and the multiplier is 7 (23x3=69; 6+1=7))
29" 23 LUee Oey oO syed me S*OND 261) SEOUPES
Sazuor os tee ee 2
so that 1/23= 0434782608695652173913
Once again, tha half way mark is 69 - 3 (of 3/69 = 1/23) = 66; the sum of the first-
half digit and the second-half digits is a string of 9’s.

Eg. 1/33
(the multiplieris 10: (33x3= 99; 9+1= 10))
1/33 = 3/0 3 “. Recurring decimal= 0.03
3
108 SPEED ARITHMETIC

Since 33 of 1/33 is not a prime, the features of half-way mark and the sum to 9’s of the
first-half and second-half digtts do not hold. In fact 1/33'= (1/3) x Lh =ah/39 CLA) =
0.3 x 0.01 = 0.03 so that the recurring decimal of 1/33 is determind by the recurring
decimals of its composite fractions.

Let us now take up fractions with denominators ending in 7. Here, the ending digit
of the recurring decimal is 7 and the multiplier is one more than the penultimate digits of
9 after the denominator is multiplied by 7 so that the product ends in 9.

Eee.
(the multiplier is 5, since 7X7 = 49 and 4+1=5)
ive Jes vi ve St bee 17
ZAM BAS 227 53
so that L/7- 2-0. 142857
Since the denominator 7 of the fraction is prime we have the half way mark 42( 1/7 =
7/49 — 49 - 7) and the first-half 142 added to the second-half 857 gives 999.
We have earlier seen that the recurring decimal 1/49 = (1/7)° has 42 digits and not
48 = (49-1) digits as we have for 1/19 and 1/29. The reason for this is two - fold.

(1) 49 has a factor 7 and 1/7 has 6 recurring digits. It may be seen that 1/7 = 0.142857
when divided by 7 (equivalent to 1/49) is not exactly divisible until we have taken
7 cycles each of the 6 digits 0.142857. This accounts for 6x7 = 42 digits in 1/49.
In the case of 1/39 = 0.025641, where also we do not have 39-1 = 38 digits but only
6 digits, we see that 1/13 = 076923, but is exactly divisible by 3 to give (1/13) +3
= 1/39. This makes 1/39 to have the same number of recurring digits as 1/13.
(ii) The six fractions 7/49, 14/49, 21/49, 28/49, 35/49 and 42/49, all of which
contains the denominator 49 are multiplies of 1/7 and do not belong to the group of
fractions like 2/49, 9/49, 37/49 etc. For this reason too, the number of recurring
digits in 1/49 is (49 -1)-6=42.
Thus, the number of digits in a fraction with a denominator which is composite may
be determined by one or both of the above peculiarities.

Ee ANT
(multiplier 12)
/l7 =» 7/0 5 8108) 2 lbgest DyAGwe 411 Jk CO
F-10009 Due dian CL te oe On een
so that 1/17 = 0.0588235294117647
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 109

Since the denominator 17 is a prime, both the features (i) the mid-value being 112
(=119-7 of 7/119 = 1/17) and (ii) the first-half digits and the second-half digits summing
to 9’s are reflected.

Eg, 1/27
eaeipiaer 19)
12d sa fllind
13
Hevesi 27 = 1/(3x9). so.that. 1/9. =0. | when divided by 3 takes 3 recurring digits 0.111
(of 1/9) to be exactly divisible; similarly 1/3= 0. 3 takes 3 digits.0.333 to be exactly
divisible by 9. Thus 1/27 has only 3 recurring digits.

Eegud {37
(multiplier 26).
KS Mt Yn Ore ee|
Wi 18
That 27 and 37 are complementary with reference to 9’s can be seen from the fact that
27 x 37 = 999 and correspondingly 27 x 0. 037= 0.999 and 37 x 0.027= 0.999.

Finally, let us turn to conversion to recurring decimals of fractions with

denominators ending in J. Here the ending-digit of the recurring decimal is 9 and the
multiplier is obtained by adding unity to the penultimate digit of | the product of the
denominator and 9. Thus eg. the multiplier for 1/31 is 28 (31x9 = 279 and 27 + 1 =28).

Eg. 1/11
(multiplier 10)
a oe 2
2
Wil =. 0.09

eA)
(multiplier 19)
Cie Oe OG ue arm Ben yttlig
he ways 47
-. 121 = 647619
1/21 has 6 digits, (the same as that for,1/7) since 1/21 = (1/3). ( 1/7)
and the recurring decimal of 1/7 =0. 142857 is exactly divisible by 3.
 110 SPEED ARITHMETIC

Eg. 1/31
(multiplier 28) :
pepemies eanaltwie serMreal.oc bis Gculimi2ghe
OGmaile 1st? el 1812. 14 aol ea 8-905
As might have been noted, in quite a few examples (like the last one), the multiplier
gets larger in line with the denominator and consequently the operations of multiplication
and addition get more difficult and tedious. To overcome this situation, we now present
an alternative method.

**(b) Using negative multipliers.

We _ have just seen that for fractions whose denominators end in 1,3,7 or 9, the
multipliers are respectively 9,3,7 or 1. It may be remembered from the previous chapter
that these are also the +ve osculators for the numbers 1,3,7 or 9. Also, we have
remarked in that chapter that for the numbers 3 and 9, _ the +ve osculators are smaller
than their -ve counterparts whereas for numbers | and 7, the -ve osculators are smaller
than the +ve ones. Since in our present case, the osculators play the role of multipliers, it
is preferable to use the +ve osculators as +ve multipliers for fractions with
denominators ending in 3 and 9 (as we have done above) and the -ve osculators as -ve
multipliers for fractions having 1 and 7 as ending digits in the denominators. This will
very much lighten the computational work.
Let us now work through 1/7 with the -ve multiplier. 1/7 = 7/49 so that the +ve
multiplier is 5 which we had employed earlier. The -ve multiplier = (the denominator -
the positive multiplier) = 7 -5 = 2 (the same as the -ve osculators for number 7 as we had
in the previous chapter). Since 2 < 5, it is preferable to employ 2 for our continued
multiplication method. However, 2 being a -ve multiplier, we denote it by 2 and
proceed as follows :

fia a
Be fie waibu (ere) aoe

(1) The ending digit of the recurring decimal is 7 for 1/7. So we take its vinculum ‘viz.
3 as the starting point for multiplication.
(ii) Multiply 3 by 2, we get6:
(1i1) Multiply 6 by » gives 1 2 and as usual, we write it as with 2 on the Q-line
(i.e. the top line) and 1 on the carry-over (c.0.) line (i.e. the bottom).
(iv) We now multiply 2 by 2 to give 4 and add the c.o. 1 to it. This gives 3.
(v) Multiply 3by 2to get 6
(vi) Multiply 6 by 2 to get 12.
(vii) Multiply 2 by 2 and add | to it to give 3.
 FRACTIONS, PERCENTAGES AND-RECURRING DECIMALS 111

At this stage we find that 3 (the same as the last digit with which we started)
repeats. We therefore conclude that the recurring cycle is complete and the cycle consists
of 2 63 26 3 which corresponds to 142857. Thus 1/7= 0. 142857, the same as the
one we obtained earlier with the +ve multiplier 5.
Also, two features may be noted:
(i) The sum of the first half of digits and the second half viz.2 63+ 26 3 give 000
(like the string of 9’s that we had with +ve multiplier).
(ii) When we have reached the half-way mark (i..e. after getting the 3rd digit), the next
digit is 3 which is the same as the starting digit but without a vinculum; in other
words, the starting digit and the mid-digit have the same value but with their signs
reversed.

Let us consider two more examples.

Eg. UT (= 7/119).
The +ve multiplier is 12 (which we have used before) and the -ve multiplier is 5.
The last digit in the recurring decimal for 1/17 is 7 whose vinculum is 3. Sowe take 3
as the starting digit for multiplication and 5 as the multiplier. Proceeding as before,
we get: - oe > © a)
Sy a gk Fuca’ gual fae» dees re ad
3 ig Rae arte es
Since we have arrived at 3 (the -ve of the starting digit 3, we know that we have reached
the mid-mark constituting the digits 0 62 98 45 3. The first half is easily obtained by
removing the vinculums from the digits which have them and placing vinculums on those
which do not have them. Thus the first half consists of 06 29 84 53 (so that the sum
of the two halves consists of zeros.

che LIT 006 29 34.530 6298 45.4

= 0.0588235294117647
(Same as before, as it should be).

Let us take up one last example: a fraction with its denominator ending in 1.

Be 1/21 (= 9189)
The ending digit of the recurring cycle should be 9 and therefore the ending digit of
the corresponding vinculum should be 1. The +ve multiplier is 19 ( a very large number)
and the corresponding -ve multiplier is 21-19 = 2; so we take 2 as the required
multiplier. Proceeding as before, we have:
12 SPEED ARITHMETIC

0 ile Oa
[re
Since, 1 has repeated (exactly as the starting digit) we have reached the recurring cycle
so that ; . i ye
21 benOmsle6.84 Qik ehOw 0 dicen e
(The above calculations may be compared with the heaviness of computations involved
by taking the +ve multiplier 19).
The advantage of working with a -ve multiplier becomes particularly transparent
as the fractions (with ending-digit denominators lor7) become larger and larger. a

8.5. Division Method for Converting Fractions to Recurring Decimals.

(a) Using positive divisors.
We can accomplish the same result as above, if we perform division instead of
multiplication but starting from left and moving towards right and using the same
multiplier (now calling it the divisor) as before. However, the first (starting) digit has to
be ascertained by a quick division of the numerator by the divisor. Let us now consider
some examples to illustrate this method.
Eg. Let us, as before, start with 1/19
(1) The multiplier (in fact, itis the divisor) is 2. Dividing | (the numerator) by 2, we
have Q as the quotient which we write on the top line (1.e. the quotient line) and |
in the bottom line (i.e. the remainder line) as shown below:
(11) Divide 10 by 2 to give 5 as the quotient which we write to the right of 0.
(iui) Divide 5 by 2 to get 2 as QandlasR
(iv) Diyide 12 by2 > Q=6andR=1 .
Continuing this way, we have the following -
VA OO. D2 O63) Slee Dy SS Ode oe ee heey ee
l I ‘ar! E / | | en jj
We note that apart from the fact that we have the same answer as that obtained by
the multiplication method , we also find the same dual features viz. (i) the recurring
decimal has a cycle of 19 -1 = 18 digits and (ii) the mid-way number is 18 as in the case
of multiplication.

Eg» 1/39
(divisor 4 ; starting quotient 0and rem. 1).
SEER A? SNE PASS Baty APRS. [yo
LDS 2ST Rl gugg yn
The same features are present as noted under the multiplication method.
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 113

Ee A/1B
Here, for determining the divisor, the starting and the remainder digits, we write
1/13 as its equivalent 3/39 which has the denominator ending in 9 so that the divisor is 4;
Next, we divide the numerator 3 of 3/39 by this divisor and get the starting quotient digit
as 0 and the remainder digit as 3.
Be AA? Gh Pan 35/0
Pabvilp ud sly 048
so that 1/13 = 0.076923
Eg. 1/23
(1/23 = 3/69 so that the divisor is 7, the startingQ - digit
= 0 and the remainder
digit = 3).
foes ot 4.5 4 7-5 2 Oo 0 y Grigg 5 Cyn jt pIgGgd 1993 BIG
Soe eg yk a). 4 yi Sale Nall Sale Each Meyaie tkWE aS cual aah
(The mid-digits are 66 (= 69-3))

Eg. A/T
(1/7 = 7/49; therefore the divisor is 5, the starting Q digit is 1 and the remainder
ise), ;
DAT oe botol2 wi 3 ifi
2 tt de Gay 2
giving 1/7 = 0.142857
Eg. 1/37
aia t= W257; therefore divisor = 26, starting Q-digit = 0 and remainder = 7)
P37 = > ery 3
NGdd 10 Loh
4 ASF = 0027

Eg. 1/21
(1/21 = 9/189; divisor = 19, starting Q =0 and remainder
= 9)
Vee Ded DCO, fed?
DR ot 72
so that 1/21 = 047619

Thus, we have demonstrated that for any fraction whose denominator ends in 1,3,7
or 9, the above methods of multiplication or division can be applied.
In the case of fractions ending in even numbers, five or zero, we adopt the usual
technique of multiplying such a fraction by powers of 2 and/or 5, so as to reduce it to a
fraction with its denominator ending in one of the numbers 1, 3, 7 or 9, carry out the

29
114 SPEED ARITHMETIC

conversion into recurring decimal by either the multiplication or division methods as

above; and finally divide the recurring decimal so obtained by the same powers of
2 and/ or 5 (as orginally used) to get the required decimal equivalent of the fraction.

Eg. 1/56 = 1/(8x7) .

Multiplying it by 8, we have 1/7. We therefore consider 1/7 for conversion into
recurring decimal. This, as we have seen, is 0. 142857. We now divide the Pa by 8 to
obtain 0. 017857142857.

**(b) Using negative divisors.

We shall now demonstrate that the method of division can as well be employed
using -ve divisors. It may be remembered that we used the same multiplier for both the
multiplication and the division methods. The same applies for - multipliers too; i.e. the -
ve multipliers used earlier under multiplication method will be the ones that we will
employ for the division process also; only, we shall call them -ve divisors. Secondly, the
starting digit of the process is also the same as the last digit of the multiplication process;
however in this case, we shall omit this starting digit and take the remaining digits in the
quotient line to constitute the required answer.
We now consider the same examples as before for illustrating the division
procedure with -ve divisors.

Be. AGT
The -ve divisor is 2 i.e. 2 ; the starting digit is 3 (the same as the ending digit
under multiplication); of course, we now proceed from right to left.
(i) Dividing 3 by bsgives quotient 2 and rem.= 1. 3 oe
So, as before, we write 20n the Q line and 1
on the carry over (i.e. the bottom line). We also
separate 3 and | by a stroke to remember that
3 is not to be included in the required answer.
(ii) Dividing 12 by > gives 6 Ny wa ie
(111) Dividing 6 by 2 gives 3 |
(iv) Now that we have reached 3 (i.e. the -ve of the starting digit 3) and have one-
half of the 7-1 = 6 digits, we conclude that we have reached the mid-mark i.e. the
first -half of the digits required namely 2 63. The second-half is obtained by
subtracting these digits from 000 which is the same thing as placing vinculums on
the digits that do not have them and removing the vinculums from those that have
them, thus giving 26 3. Sothat 1/7 = 0.2 63 26 35 0. 142857
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 115

Eg. 1/17
(The -ve divisor is 5 and the starting digit is 3. Continuing the division as before
we arrive at
Sie P2006! te? 099 68 104. 42513
3 APeerie oe |
Again we have 8 (= (17-1)/2) digits and the digit 3 (-ve of 3) to conclude that the above
8 quotient digits are the first-half digits of the answer. The other-half is obtained, as
usual, by subtraction of these digits from the zeros.

Thus 1/17 0.(ORGU 29. Sek a5, BehkO Con ek Oy Sa Acie

c
OED Ousted 22sec (944 1 lal
.6) AA)

Eg. 1/21 (Divisor is 2 and the starting digit is 1 and the rem. is 1)
So, we have 1/21 = 1 ] we Beads Oe oi
| l
Since Tis the S ame asthe starting digit, we conclude the division and write
1/21 = 0. 1 6 8 4 1=0.047619 asthe recurring decimals. Ges

8.6. Remainder-Digit Method.

The two earlier methods just considered provides the Q-digits directly. The present
method, however helps to find the remainder-digits (rem-digits) initially from which the
required Q-digits can be found . Yet, this two-stage method is sometimes more quicker
and computationally simpler.
Let us illustrate the method by means of examples.

Fg. \/7
(1) The first remainder (which can be mentally obtained) is 3.
(ii) Multiply 3 by itself and since the product exceeds 7 (the denominator of the
fraction), subtract
a suitable multiple of 7. Thus (3x3)-7 = 2.
(iii) Multiply the latest remainder 2 by 3 to give 6 which is less than 7 (and therefore
no subtraction of any multiple of 7 is required.) Thus we continue with the process
as shown below.
(iv) 6x3-2x7 = 4
(v) 4x3-7 =5
116 SPEED ARITHMETIC

(V1) 23
=2K) 4
(vii) 1x3 = 3 (same as the first remainder)
and therefore we need to consider only the first 6 remainders : 3, 2, 6, 4, 5, 1.

Here we may note that the first- half remainders viz. 3,2,6 added to the second-half
remainders viz. 4,5,1 gives the digits which are the same as that of the denominator of the
fraction i.e. 326+451=777 (just as we had the sum 99....9 for the Q - digits earlier).
This means that we need not go through all the above 6 steps to find the 6
remainders but only carry out one-half the number of steps.
It may also be noted, that 6 is the largest remainder (denominator of the fraction
being 7) and so we need to continue with the process of finding the remainders only till
we get a remainder which is one less than the denominator.
(viii) We now multiply these 6 remainder digits by a digit (denoted by c) which when
multiplied by the last digit of the denominator would give 9 as the last digit of the
product.
(Thus for denominator 7 , c = 7, for denominator 1, c = 9, for denominator = 3, c=3
and for denominator = 9 ,c = 1)

After multiplying the rem. digits by c, we retain only the last digit of the
multiplications. These last digits are the required Q - digits.
In the above we have the rem. digits 3,2,6,4,5,2. in step (iii) Multiplying them by
c= 7 and retaining only the last digits of the products we gt | 4, 2, 8,5, 7 so that
V7 9.4:2)8: 57 ®

Fe, 1/13
(1) First remainder = 10.
(ii) LOxdOess
x7 95. 9X 10213
f= 1D
Now we have reached the remainder 12 which is one U9 12
less than the denominator 13 of the fraction. So we a gear |
obtain the other remaining digits as complements of
the above remainders from 13. Sid 13

Therefore the remainder digits are 10, 9, 12, 3, 4, 1 (in that order)

(iii) We now multiply each of these digits by c =3 and retain only the
last digits. This gives the Q-digits as 0,7,6,9,2,3 so that
1/13 = 0.0769 23
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS Pr?

Let us now illustrate another variation of the above which is even more simple,
' using the fraction 1/19.
(i) The first remainder digit is 10.
diy .ine-second one is 1) x 10-5 * 19 =5:
Now, since 5 is less than 10 (the first remainder), we may henceforth use 5 for deriving
the rest of the remainders, starting with the first remainder = 10. This variation, unlike the
above, avoids large numbers in the subsequent computations.
Thus SX?X19 = Fi
» Xo-- 19 = 6
al 5 KD) =
Now 3 is even lesser than 5 and we make use of it to continue the process (Again as
above, we start with the first remainder 10).
3x 10-19 =i
SxS + “$5
3 x12 - 19 = 17.
7x6 =" 13
Since we have reached 18 (one less than 19, the denominator), we obtain the
remaining recurring digits through the subtraction of these digits from 19 to give
1 lO quid tOnel SD: glOrshO SF 1 Sind 9
LO Goon TZ, wi Op 30° bleed Skok ES

ee ly tS5 tO o, a

.. The rem. digits are

1015 12°6 Sela io ei) eta
9: 53. 168142

(iii) Multiplying these by c = 1, and taking only the last digits, gives
0, 522, O.9, lesen tao, ase 240, x4 2,11, so'that
1/19. = 0.052631578/947368421
EO 1/31 I! ,
(We have considered this example earlier using the multiplication method with 28
as the multiplier. This is nearly as large as the denominator of the fraction, thus requiring
more than mental process to carry out the required multiplications).
The first two remaining digits of 1/31 are 10 and 7 (= 10x 10-3 x 31)
We now take 7 as the multiplier (since it is lesser than 10) and start the process of
obtaining the other rem. digits with 10 as the first rem. digit.
 11g SPEED ARITHMETIC

Thus,
: Jd SAOi =e Ba> FERS
1M De BA =<18
Pek SoReal ais
TeXalL8 Aad KSidhiad= 2
Now 2 being the smallest remainder, thus far, we switch over to 2 (but starting with the
first remainder 10).
2D = 20; 207 > ci4
2 8 enkO. ¢: 2618-3 =e
2x 25 Si =O: - 22 = 4
Dx a Agee Oh 2x14 = Lo
OB ghey Pee eae peas = 10 (same as the first remainder)
We stop with the preceding step and take the rem. di a as
LOS eS ehOne2),
2-20. 914 OLD eameee
Multiplying the above pyec = ‘9 and taking oe the last digits of the products, we get
We AU) 0322580645 16129
Let us take one last example under this method.

Ege l/h].
The first rem. digit is 10 We switch over to 4
ROR OO PPS 4x10-2x17= 6
PO Se TS’ = 85017 S14 4.1523 xXTESS
10x 14-8x17= 4 4x14-3x17= 5 ,
4x4 = 16
Now that we have reached 16 (one less than the denominator 17), we conclude that the
half-way mark is here and that we can obtain the remaining digits by subtracting from 17
(the denominator of the fraction) to get
Wd, Sock Sel 1 Re ae
_ On multiplying these 16 rem. digits by c = 7 and_ keeping only the last digits of the
products, we have
1/17 = 005882352 /941 17647

8.7. Geometric Progression Method.

Another way of extracting the Sequence of remainder digits of a fraction is to use
the geometric progression (G.P., for short) method. Knowing the first two remaining
digits of the fraction, we obtain the ratio (called the common ratio, r ) of the second
remaining digit to the first and employ it ta derive the required sequence of the rem.
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 119

digits. Without going into the theory of the method, we illustrate the various steps of the
procedure with the help of, as usual, a few examples.

Eg. Take the case of the fraction 1/7.

(i) The first two rem.digits (which can be obtained mentally) are 3,2.
(ii) The common ratio r of the second to the first digit is 2/3;
(iii) To get the next rem. digit, multiply the second rem. digit 2 by 2/3 to get 4/3.
Add a suitable multiple of the denominator 7 of the given fraction to the numerator
4 of 4/3. so that the same is exactly divisible by the denominator 3. Thus
{(4+(2x7)}/3 = 18/3 = 6 which is the next rem. digit in the sequence.
(iv) Multiply 6 by r (=2/3) to get 4. This is a whole number by itself and is therefore the
next rem. digit.
(v) Multiply 4 by r which gives 8/3. So we add (1x7) =7 to 8 and divide the sum by 3.
This gives {8+(1x7)}/3 =5.
(VTEINEXtS Dim = 652x95 2/3 n=110/3cand» 110 -i(2x7)Won =rS8eand Since; 8H (the
denominator),we take away 7 from it and have 8 - 7 = 1. We can now constder that
we have reached the end of the sequence of the rem. digits. For, if we proceed
further, we would have 1 x 2/3 = 2/3 and {2 + (1x7)}/3 =3, the same as the
beginning rem. digit i.e. the sequence begins to repeat.

Thus, we have the cycle of rem. digits as 3,2,6,4,5,1.

Again as pointed out before, we have the rem. digit 6 at step (111) which is one less
than the denominator of the fraction; therefore we can as well stop the above werk with
the rem. digits as 3,2, and 6 and obtain the other rem. digits by subtraction from 7 (the
denominator of the fraction) to give 4,5 and 1. This would save one -half of the work as
well as time.
As before we multiply each of the above by c=7 and drop all other digits of
multiplication except the last one. This gives the quotient digits of the fraction as 1,4,
2,8,5,7 so that 1/7 = 0.142857.

een vii h3:

(i) The first two rem. digits are 10 and 9 ; r= 9/10
(ii) 9x 9/10 = 81/10 > (81+ (3 x 13)}/10 = 12
As in the above case we have now reached 12 which is one less than the
denominator. Therefore we subtract 10, 9 and 12 from 13 to get 3,4 and 1, so that the
cycle of rem. digits is 10, 9, 12, 3,4,1.
120 SPEED ARITHMETIC

Finally the quotient digits are obtained, as usual , by multiplying the above digits by
c = 3 and retaining the lost digits only which gives the required Q-digits 0! NG, SUR se
that 1/13 = O:0 76973;

(a) Alternative GP method ave

In the above, we see that when we take the first two rem. digits, the common ratio
is a fraction which can make the calculations tedious.
This can be minimised if we start with the last two rem. digits, as the following
examples will demonstrate.
Eg. 1/19
The last digit, as we know, is 1 and the penultimate digit is one more than the
digit preceding 9 in the fraction (this is true for all fractions with the denominators ending
in 9) Thus, the last two rem. digits are 2,1 in that order with the common ratio r = 2/1 = 2
(which is now a whole number making the computations far easier). For getting the rem.
digits left of 2, we multiply the rem. digit 2 by r giving 4 . We now have the last three
rem. digits as 4,2,1. So multiplying the last obtained rem. digits by r and subtracting, if
necessary, a suitable multiple of the denominator, we get the required rem. digits starting
from extreme right and moving leftwards.
1, 2x2 = 4, 4x2=8; 8X2=16, 16x2'= 32°and 32°- 19'="13), 13x2'= 26 and’ 26-19 =7,
7x2 = 14,14 x 2 = 28 and 28-19 = 9. 9x2 = 18 (number which ‘is one less than the
denominator) showing that we have reached the half-way mark moving leftwards.
Writing these rem. digits in the proper sequence, we have
/9, 14, 7, 13, 16, 8, 4, 2.1 .
The beginning rem. digits can now be obtained by subtracting the above digits from the
denominator 19 to give 10, 5, 12, 6, 3, 11, 15, 17, 18.

Let us take another example whose denominator does not end in 9, say 1/7. In such
a case we multiply the denominator by a suitable number whose product ends in 9. Thus,
in our case, we multiply the denominator 7 by 7 to get 49. This gives the last two digits as
5 and | (where 5 is one more than the digit preceding 49); Also, the common ratio r = 5.
So we proceed as in the above example to get the leftward rem. digits as
L, ExS*=" 5, S83 S925 > 25 “6x77 = 4,
4x5 = 20— 20 - (2x7) =6 (digit one less than the denominator 7).
Therefore stopping with 4, we have the set of last rem. digits as : 4,5,1. The first half is
obtained by subtraction of the above from7 to give the entire set of rem. digits as
3,2,0.1- 45951.
 FRACTIONS, PERCENTAGES AND RECURRING DECIMALS 121

Eg. 1/13 : A suitable multiplier of the denominator to give a number ending in 9 is 3 as

13 x 3 = 39. This gives the last two rem. digits as 4 and | in that order; also r = 4. We
thus have the sequence :
1, 1x4 =4, 4x4=16 — 16-13 = 3, 3x4 = 12 (a number that is one less than 13).
Thus, the second half of the rem. digits are ...,-,/ 3,4,1. The first half of the rem. digits are
obtained by subtraction from 13 to give the complete sequence as 10. 9, 12/3,4,1.

2 th)2bes
Here the multiplier is 9, the last two digits are 19 and | ; alsor = 19. So,
1x19=19, 19x19 = 361 — 361 -(21x17) =4, 4x19 = 76 >
76 -(21x3) = 13, 13x19 = 247 — 247 - (21x11) = 16, 16x19 = 304 >
304 - (21x14) = 10, 10x19=190 — 190 - (219) = 1. Therefore the sequence
of rem. digits is 10, 16, 13, 4, 19,1.

**(b) Using vinculums

We see in the above example, that since r= 19, the calculations have become
heavy. In this connection, we may recollect from the earlier chapter that number 19
corresponds to the positive osculator of 21. It may also be remembered that for numbers
ending in | and 7, it is more economical to operate with the negative osculators which are
nothing more than the given number minus the positive osculator. Thus, in our present
case, the negative ratio is 21 - 19 = 2; and to convey that it is negative, we write it with a
vinculum as 2 and carry out the above operations with r= 2 instead of 19.

This gives: 1x 2= 2-21 -2=19;

2x 2=4; 4x 2= 8521-8=13;
8 x 2=16; 16x 2= 3 25(21x2)-32 =10
so that we have the sequence 10, 16, 13, 4, 19, 1 as before but obtained much more easily.

Fe. Uh]:
: The multiplier (@.e. the positive osculator) is 12: therefore the negative
osculator is 5 ( much less than 12 ) .The last two rem. digits are 12 and 1. We prefer to
work with r= 5 to obtain the following sequence of rem. digits.
ee ae oe le x 5S 2S20-1 = 8,
8x '5= 40 — (3x17) - 40 =11, llx 5 = 5 5 3 (4x17)-55 = 13.
13x 5= 6 5 (4x17)-65 =3.3x 5 = 1 5-(1x17)- 15 =2,
2x 5= 10-(1x17)-10 =7,7x 5 = 3 5 (3x17)-35
= 16
(a number which is one less than the denominator).
122 SPEED ARITHMETIC

Therefore the second half ‘of tem. digits 18 TLS EQS NBM saa

As usual, the first set is got after subtracting the above from the denominator 17 so that
the whole sequence of the rem. digitsis :
10,15,14,4,6,9,5,16/7,2,3,13,11,8,12,1 p25 a

8.8. Arithmetic Progression Method

We will finally discuss one more method called the arithmetic progression (AP, for
short) method for finding the rem. digits. This is primarily applicable for fractions whose
denominators end in 9; in the case of other denominators, they may be converted by
suitable multipliers to have their ending digit as 9. Having got such a denominator, we
determine the common difference d by subtracting the ending digit 9, from the
penultimate digit of the denominator. Thus, if 1/79 is the given fraction, then d= 9-7 = 2;
if 1/109, then d = 9-10 = -1, if 1/7 then since 1/7 = 7/49, d= 9-4 = 5 and so on. Using the
first two rem. digits of the fraction as before and the value of d, we develop the sequence
of the rem. digits through an iterative process explained in words below
3rd rem. digit = (2nd rem digit x d) + first rem. digit
4th rem. digit = (3rd rem. digit x d ) + 2nd rem. digit
5th rem. digit = (4th rem. digit x d) + 3rd rem. digit
and so on.
In case, the resulting number.on the RHS of any of the above values exceeds the
denominator of the fraction, we subtract suitable smallest multiples of the denominator so
as to obtain a number which is less than the denominator.
®

Hg. Nii = 7/49;

The first rem. digit = 3, the second rem. digit is 2 andd =9 -4=5.
Now, using the above relations, we have

3rd rem. digit = (2x5) + 3 = 13 > 13 - (1x7) =6

4th rem. digit = (6x5) +2 = 32 — 32 - (4x7) =4
5th rem. digit = (4x5) +6 = 26 — 26 - (3x7) =
6th rem. digit = (5x5) +4 =29 — 29 - (4x7) = 1

Thus, as earlier, the rem. digits are 3,2,6,4,5,1. Of course, as we did previously, we can
stop the iteration half-way when we>get the rem. digit = 6 (which is one less than the
denominator 7) and obtain the other rem. digits by a process of subtraction from the
denominator 7.
 FRACTIONS, PERCENTAGES AND-RECURRING DECIMALS 123

E¢. 39%:
The first two rem. digits are 10 and 22, d =9-3 =6. Therefore,
3rd rem. digit = (22x6) +10 = 142 — 142 - (3x39) = 25
4th rem. digit = (25x6) + 22 = 172— 172 - (4x39) = 16
5th rem. digit = (16x6) +25 = 121 — 121 - (3x39) =4
6th rem. digit = (4x6) +16 = 40 > 40-39 =1.
Since we have reached 1, we know that we have come to the end of the sequence. (As a
check, if we continue with the 7th rem. digit, we would have it as(1x6) +4 = 10 ie. the
same as the starting rem. digit ).
Thus, the rem. digits of 1/39 are 10,22,25,16,4,1.

Eg. 1/109:
Here d = 9-10 = -1 and the first two rem. digits are 10 and 100. A few of the rem. digits
are obtained below as illustration.
3rd rem. digit = 100 (-1) + 10= -90 — (1x109) -90 = 19 (we add 109 here
to convert the -ve value (-90) into a +ve one.)
4th rem. digit = 19 (-1)+100 = 81
5th rem. digit = 81 (-1)+19 = -62— 109-62 = 47
6th rem. digit = 47(-1)+81 = 34
7th rem. digit’ = 34 (-1) +47 = 13
8th rem: digit = 13 (-1)+34 = 21
9th rem. digit = 21 (-1) +13 -8 > -8+109= 101
10th rem. digit = 101 (-1)+21 = -80+109 = 29
llth rem. digit = 29 (-1) +101 =72 and so on.

Eg. \/13.;
1/13 = 3/39 therefore d = 9-3 = 6; also the first two rem. digits are 10 and 9.
3rd rem. digit = 9x6 +10 =64 > 64 - (4x13) = 12
4th rem. digit = 12x6+9= 81--— 81- (6x13) =3
5th rem. digit = 3x6+12 = 30 — 30- (2x13) =4
6th rem. digit = 4x6 +3 =27 > 27 - (2x13) =1
Since we have reached 1, the cycle is complete.

Be 1/89:
This has a convenient feature. We have d = 9-8 = 1 and the opening two rem.
digits are 10 and 11. Since d = 1, we get the succeeding rem. digits without much of
calcilavons.as 10; bly CLOTLL) sli a (2 + 11) 32, 32 4:2))=53,- (53 +32) = 85,
(85 + 53) = 138 > 138-89 = 49
124 SPEED ARITHMETIC

and so on i.e. by successively adding the rem. digits (of course, taking care to subtract by
a suitable multiple of 89 whenever a rem. digit exceeds 89).

8.9. Determining Multiples of Fractions.

We have so far discussed methods of decimalising. a basic fraction i.e. a fraction
whose numerator is unity. We shall now consider methods of converting into decimals,
fractions having numerators other than unity (and, of course, less than the denominator).

(a) Method 1: The method that immediately comes to mind is to apply any of the
previous procedures such as the multiplication/division or rem. digits methods to such a
fraction. Thus, for example, we may apply the division method to the fraction 8/13.
Since 8/13 =24/39, the divisor is 4 and we also know from 8/13 that the starting digit is 6.
So we apply the division method taking 4 as the divisor to give
SAB = AOU Gob SHBTSYAi 1G
a lsBs ad ¥ }
Since 6 ( the starting digit ) has appeared, we stop here so that 8/13 = 0.615384.

(b) > Method 2: Consider the fraction 1/7= 0. 142857, ,it is easy to get its multiples as
2/7 =0.285714, 3/7= 0.428571, 4/7 =0.571428, 5/7 =0.714285 and 6/7 =0.857142
A close observation of the above would reveal the following:-
(i) All the 6 decimal - digits are different; also the same 6 digits appear in all the
multiples but with different starting digits.
(11) The digits follow the same sequence as in the basic fraction 1/7 ; and
(iii) Given a starting digit, the other digits follow the same clockwise order.

These features help to devise the following method.

(Gi) Writes: lean 2°85 3. 7 (the-sanie-sequerncé asin: W7)
(1). Rankethesepdisitsase(1)s(2)oe..... ,(6) in increasing order, ranking the smallest digit
namely | as (1) and the largest digit 8 as (6) as below:
| 4 pe 8 3 %
CURE poi eile Ret lego!
(iil) If now, we require, say 4/7, we start with the digit ranked (4) i.e. 5Sand take the
. digits following it cyclically so that 4/7= 0. 571428: or, if we want 6/7, then we start
with the digit ranked (6) and move cyclically to get 6/7= 0. 857142.

A variation of the above : Since 1/7 = 0.142857 ,we just remember its first two digits
namely 14 so that the starting digits of say 2/7 should be 28 and_ noting the cyclicality of
the sequence 142857, we can immediately say that 2/7= 0. 285714. It may be remembered
that we adopted this method in chapter | (under the section “Reciprocal of Prime
 FRACTIONS, PERCENTAGES AND-RECURRING DECIMALS 125

Numbers”). It avoids the task of ranking the digits as we did above so that the value of the
fraction can be given instantly so long as we remember the sequence of digits in 1/7.

(c) Method 3:
In the above example, we have, as just mentioned, the three peculiarities viz. (i) all
the 6 digits are different (ii) the multiples of the fraction are also the same six digits and
(111) these digits move cyclically.

But, what about a fraction in which any one of these special features breakdown?
Can we still write the decimal equivalence of a multiple of such a fraction, knowing only
the decimal digits of the basic fraction ?

Let us take the fraction 1/19 which has 18 multiples: 1/19, 2/19......,18/19.
Now 1/19 = 0.0 2) 2 6 3 1 5 7. 8
Gh) nnllO}: elds Kh2).4 (Ow (S)la ids OSGS90a) 4LD

: 4 7 & 6 8e~4). 2 i
| Cs) {1 )osis) (7) (13) (16)) () (4) @)
Here, we have 18 digits some of which repeat (unlike in the previous case). In such a
situation, we can rank them by comparing a given digit with its immediate nght hand
neighbour. For example, if we take 6, it has two right hand neighbours 3 and 8 and since
63 < 68, we rank the 6 of 63 lower than the 6 of 68. Following this procedure. the
rankings of the various digits are given in brackets below the respective digits. It may be
noted that due to clockwise cyclicality of the recurring decimals, the digit 1 at the end of
the cycle will have 0 (the starting digit ) as its right hand neighbour and the other digit 1
will have 5 as its neighbour so that 1 of zero will have a smaller rank than the 1 of 5.
Using these ranks we can now write down the multiples of 1/19 as follows:
1/19 =0. 052631578947368421
2/19=0.105263157894736842

4/19 = 0.2 10526315789473684

18/19 = 0.94736842 105263 1578

(d) Method 4: Finally, we consider a case where unlike the above. the number of
decimal digits in a fraction is far less than the denominator and has no chance to account
for all the multiples of the fraction. Take for instance, 1/13 = 0.076923. While we can
have 12 multiples, we have a cycle of only 6 digits. In order to account for the 12
126 SPEED ARITHMETIC

multiples, we consider 1/13 and 2/13 together, each providing for 6 multiples . Thus, we
proceed to rank these 12 digits together as shown below:-
Lire AO sae aes een Zio ly gees Se 4 4a 96
COME Ye) AG ae Rees, 2) (7) ©) A) ©) &)
This gives: ; i
1/13, = 0.076923 NOU: 153846
3/13. = 0.230769 2)files = 0.384615
4/13 = 0.307692 6/13 = 0.461538
O13 = 0.692307 7/13 = 0.538461
10/13. = 0.769230 8/13 = 0.615384
12/13 = 0.923076 11/13 = 0.846153
**Now, let us consider an example similar to the one above but with a difference. Take
the fraction 1/21= 0.047619 which again has a cycle of 6 decimals but has 20 multiples
in order to accommodate these 20 multiples with a basic fraction having only 6 decimals,
we write (as we do above), the decimal equivalent for the first 4 multiples
ite A Oe Tee poTsS oO Ow sve O82
(1) (10) (16) (13) (4) (19) 2) 20) 1)'eG)- (8) AY)

BOR aoldentlr 2 AAG? HOI We Wakes

0% Broan wee
(3) (9) (6) (18) (12) C15)

We now find that the 6 digits of 4/21 are the same as the 6 digits of 1/21 (even if they are
in different order) and we therefore do not consider them for ranking, thus allowing for
only 18 digits. This would mean that the ranking of these 18 digits-would account for 18
multiples out of 20. The two multiples that the rankings would miss correspond to 7/21 =
0.3 and 14/21 = 0.6, each of which is different from the rest in that each has a cycle of
only one digit unlike 6 decimals for others and hence do not belong to the rest of the
family of the fraction 1/21. This feature has to be kept in mind and these two multiple
fractions have to be suitably accommodated while ranking. Doing so, we have the
rankings as shown above. As can be seen, the fractions 7/21 and 14/21 hold the ranks 7
and 14 respectively. Using these ranks, we can now write, the decimal equivalents of the
20 multiple fractions as follows:-
1/21 = 0.047619 13/21 =0.619047
2/21 = 0.095238 1421 0.623)
3/21 = 0.142857 15/21=0.714285
6/21=0.285714
H21 =0.3 (= 1/3) : :
8/21 =0.380952 20/21= 0.952380 we
 FRACTIONS, PERCENTAGES AND ‘RECURRING DECIMALS 127

8.10. Conversion of Recurring Decimals to Fractions

Until now, weshave seen methods to change a given fraction to its corresponding
recurring decimals. Let us now consider the reverse process of obtaining the fraction,
given a recurring cycle of decimals.
In this task, we note the following interesting features which play an important role
in the conversions.

(i) Let N=0.9

» 1ON=9.9 ;
Subtracting we have 9 N =9 giving N = J so that 0.9 = |
Take another example: Let N = 0.9999 c
Then LOOOON = 9999°9999 ee
Subtraction gives 9999N = 9999 giving N = | and 9999 = |

In other words, a string of recurring 9’s equals unity. The implication of this important
aspect is that if a given set of recurring decimals is multiplied by a suitable number to
give a string of 9’s, then the conversion of the set of recurring decimals to its
corresponding fraction is immediate.

Egs. , i
(i) 0.09 : Multiplying by 11 we have 0.99 which by the above, equals 1.
i.e. 0.09 x 11 = 0.99=1 ». 0.09=1/11
(11) 0.142857 which on multiplying by 7 gives 0.999999 so that
OL42857 = A/T.
(i11) 0.047619 : Since we want the last digit to be 9 we multiply by 1.
Thus 0.047619 x 1 = 0.047619.
To get the second digit as 9 we multiply this by 2 so that

0.047619 x 21
047619
95238
. D99999

giving 0.047619 x 21 = 0.599999 = 1. Thus 0.047619 = 1/21.

Let us now adopt another method. This is applicable whenever the first half of the
decimal digits added to the second half gives a string of 9’s.
128 SPEED ARITHMETIC

(i) Take for example 0.142857. We have 142 +857 = 999.

Let N= 142857
- N+143 = 143000 > N= 143 (1000-1) = 143 x999
SN

Thus 142857 = 143 x 999

Now 0.142857 = 142857 (since 0. 999999 = | as we have seen already)

999999
9
= i sc (since sum of the halves = 999)
999999

143 Pie 3 1
Pe 1O0leeht (5 oe

(11) Consider 0.076923 (in which the two halves add upto 9’s)

77 x 999 2) TOC EL 1
0.076923 = 999999 ~~ —-:1001.—oa7x11xK13..— 13

We have so far considered cases where the ending digits of a recurring cycle is 1, 3,
7 or 9. However, in cases where the ending digit is 2 or 5 or 0, we divide the given decimal
by appropriate powers of 2 and or 5 so that the recurring decimal has a digit ending in 1, 3,
7 or9. After having applied any one of the above procedures, we get a corresponding fraction
which on multiplication by the multiple (which we have used above), gives the required
fraction.

Eg. 0.857142 ; dividing it by 2, we have 0.428571

‘ © SOK 999
0.428571 = —.——.— (since sum of the halves equal 999)
999999

Multiplying 3/7 by 2 (a factor with which we divided initially) gives 6/7 as the equivalent -
fraction.

Eg. 0. 380952: Dividing by 2 gives 0. 190476; again dividing by 2 gives 0. 095238: one more
division by 2 gives 0. 047619 which we have seen earlier corresponds to the fraction 1/21;
therefore on multiplying it by 2? = 8, we have the required fraction as 8/21.
 129

CHAPTER 9

AUXILIARY FRACTIONS

In the chapter on Division (chapter 6) and Conversion of Fractions to Percentages

(chapter 8), we have shown how certain proper fractions can be converted to their
equivalent decimals using the vedic method and thus avoiding the long and tedious
method practiced by current methods.

The above vedic method of division and conversion can be shortened even more
and the computational burden reduced by the use of auxiliary fractions (AF, for short) .

We now describe various methods available under the AF when the denominators
‘of fractions end in 9’s and in any other digits, with the help of a few typical examples.

9.1. Type I. Denominator of the Fraction Ending in 9’s.

Eg. 6/49

(1) To start with, we carry out two preliminary operations:-

(a) Since the denominator has a single 9, we shift the decimal in the numerator by
one place to the left and (b) drop 9 and increase the 4 of the denominator by one
so that the AF of 6/49 is 0.6/5 .

(11) We now divide 0.6 by 5 which is the working divisor giving the quotient Q = 0.1
and the remainder R = |. It is important to note that we have performed the
division partially to only one place, the same as that of the dividend.

(111) We next write the above Q and R yeye take the dividend as 1.1 and divide by
|
the working divisor 5 to get Q = 2 and R = 1, dropping out the decimal of the
resulting division from now on and write the same ey vi
/
l

(iv) As in (iii), we now take the dividend as 12 and divide by 5 to get Q=2, R=2
and write it as wa;
2.

30
130 SPEED ARITHMETIC

(v) We carry on this operation until we get the required number of decimals. Thus
from the above we have, on successive divisions,

Bi ae aN EW RV ROS ME ABI LySatay Se

iA. / Dae 4< [AF oe 7 2 fe] O :

(vi) | We take the final Q’s as the first line in the above, ignoring the second line of
remainders so that the final answer is 6/49 = 0.12244897959 to 11 decimal
places (dp’s)

Eg. 11/149
(i) AF = 11/15
(ii) O48 Wis 3S ASA 24S diSs/D0f Ss 7 BftS
1 de 51257 Bie /A8if- et Oni Sib Bh Sb Aa

197149. =-0,07382550335 to 11-dp.’s.

Eg. 21365/17999
(i) This is a mixed fraction and so we write it as an integer plus a proper fraction as
1 + 3366/17999 and apply the AF to the proper fraction 3366/17999
®

(11) AF. = 3.366/18 = .187 /010/389 /465

0 7 8 4

Note that we take the Q’s to 3 digits in every division since the number 3.366 of the
auxiliary fraction has 3 decimal places. This is to be kept in mind in dealing with the
AF’s in general.

Thus the final answer is 21365/17999 = 1.187010389465 to 12 dp’s.

9.2. Type II. Denominators of Fractions Ending in 1’s

We next consider proper fractions whose last digits of the denominators are
ones.
 AUXILIARY FRACTIONS 131

Eg. 8/31.
(i) In this case we (a) subtract one from the denominator and (b) reduce the
numerator by 1. Thus, 8/31 changes to 7/30. We then adopt the following
steps.

(11) Shift the decimal in the numerator to as many places to the left as there are
zeros in the denominator and drop the zeros in the denominator. Thus, write
7/30 as 0.7/3.

(11) Proceed with the division as before so that we now have Q=0.2 and R=1.
(Remember that the division is performed such that the number of digits in the Q
is as many as those in the dividend so that we have op for this example.
,

(iv) Now, unlike in the previous type, take the complement of Q from 9. We thus
now have JiThis gives 17 as the new dividend to be divided by 3 resulting in
l
Q=5 andR=2. The above steps are carried out in sequence until we have the
required number of decimals for the fraction.

As a matter of routinely carrying out this procedure, we have written below the
steps of the sequence using the format described below.
At any stage, the top and bottom numbers following the equality sign are
respectively the Q and R obtained by dividing the dividend in that step by the
denominator of the AF. The top number following an arrow is the complement (from 9,
99, 999 etc. as the case may be) of the Q whereas the bottom number is the same as the
R obtained ~ prior to the arrow.
Thus, ay, ew have

ma aL AS
2/ denotes that Q =5 and R = 2 on dividing 17 by the divisor 3 of the AF and/ 2 /after
the arrow indicates that 4 is the complement (from 9) of Q=5 and 2 is the R obtained
8 4
earlier, / 0 that follows the equality sign just te 2 shows that Q = 8 and R=0
on dividing 24 by 3 and so on. We thus have the following sequence (to the cgi
number of decimal places):
132 SPEED ARITHMETIC

ha ho aa
WAAL AY:
MAP AD AL Wo
We have enclosed the Q-R s appearing after the equality signs within strokes for
easy identification and clarity and also for the reason that these are the ones which we
will consider for the final answer.
Collecting all the numbers after the equality signs in the above, we have

SPAKE
KRAA /o/ofap/
which, after deleting the bottom R’s as before, gives the required value as
8/31 = 0.258064516129

Eg. 4325/8001

(1) AF. = 4324/8000= 4.324/8

“ynsee fess 442 “ 569”_ [2
(ii) 4.324/8= /, wh
Note that the Q s obtained immediately after the arrows are the complements of the
immediately preceding Q s from 999 in this example, since the Q s are in blocks of
threes. Thus, 459 = 999 — 540, 442 = 999 — 557, 569 = 999 — 430 and so on.
Collecting the Q—R’s after the equality signs in the above, we have

3407 © SST /? 4380/7321

4 3 2 l

so that the required fraction is 0.540557430321.

 AUXILIARY FRACTIONS 133

Eg. 21/73
(1) Multiplying both the numerator and the denominator by 7 to yield | as the ending
digit in the denominator gives 189/511 — 188/510 = 18.8/51 = AF.

(11) 1Ss/o1). =

5 6 6 3 9 0) 8 1 6
—S = > = => = => =
35 39 50 50 44 44 52 a2 15

6 5 3 6 0) 9 l 8 3
= = -5 = = = =, Fi
{5 15 0) 0 6 6 18 18 Be)

3 6 6 3 9 0) 8
> = > = > =
CRY Ammo ny 218) 50 44 44 32 / and so on.

As before, collecting all the Q-R’s after the equality signs from the above gives

pad 8, 2)memes©)ll fey ctl affateh byall gy eli sage)

397) DO)44" SIP AS 07 OP 18/337 50/44 7327 ‘atid’
so on

so that the required fraction (from the top line) is 0.369863013698.

ke. 2322/667-=3 + 321/667
On taking the proper fraction 321/667, multiplying its numerator and
denominator by 3 to get | as the ending digit in the denominator, we have 963/2001
(1) BES 9627/2000 sO 62/2
(11) = a ae A Vig ve 629 so

685 42 5a

a as uewe so on, so that

2322/667= 3.4812593703 14842578. (on collecting the numbers in the top line
appearing after the equality signs)
134 SPEED ARITHMETIC

WL eis
Eg. (i) 4/37 = 4x3/37x3 = 12/111 73 11/110=1.1/11
= AF

AL AL VAL mtn
, 4/37= 0..
108-----
108 (a recurring decimal).

9.3, Type III.. General

In the above two types, the digits consisted of 9’s and 1’s as the ae ale in
the denominators of the fractions. These are one below or one above a
base. We now consider fractions whose digits are slightly far off such a base eg. 2, 3, 7 or:
8 as the ending digits of the denominators.

Eg.. 17/58.
Here, the denominator 58 is one less than 59 whose ending digit is 9 which comes under
type I. To account for this deficit of 1, we multiply the Q obtained at each step by | and
add the same to the dividend at that step to form the new dividend . Let us work out some
examples to illustrate the procedure. Take the above example 17/58
17/58 > 17/59 — 1.7/6 = AF. i
(1) As in Type I, we divide 1.7 by 6 and write the Q-R as 2/
5
(ii) Take Q=2 and add the product | x 2 = 2 to the dividend 52 so that we have
(52+2)/6 = 54/6 whose Q-Ris:/ 9
0
(111) | Proceed as in (11) taking the new dividend 09;

we thus have (09+ 1x9)) /6= 18/3¢ hs/

(iv) (03 +(1 x3) /6 : ry /

(vy) (O1+(1x1)/6 = 2/6 a:

| 2
 AUXILIARY FRACTIONS 135

(vi) (20+ (1 x 0))/6 20/6 suse?

2
(vil) (23+(1 x3))/6 I DMGlooised?! 4ii
2 /

Proceeding in this way, we have

[sh fo/ofoh/sfofa]sp
Ph]
We see that the last two Q’s in the above have 2 digits instead of a single digit as
they should have and we therefore add | of 12 to the preceding Q=7 and 3 of 37 to2
of the preceding Q = 12. so that we have

shi)olo/2/2hih offeror)
which gives the required decimal form of the given fraction as 0.293103448285.....
(taking only the top Q’s line and ignoring the R’s as in types I and II earlier.)

Eg. 117/101;= 1+ 16/101

16/101 — 16/99 -— 16/100 =0.16/1 = AF
Excess of the denominator 101 from 99 is 2.
Therefore, we subtract 2 times the quotient at every step as follows.

(1) 0.16/1 : [0.16

0

(ii) (16-(2x16)/1= - 16/1 we yy

0
Civphyelehelr ea me i /,
0
(iv) (16-(2x 16))/1 = - 16/1 J1 67 and so on (recurring)
0
~16/101= 0.16 1 616 1 6.... =0.15841584...
*, 117/101 = 1.15841584 .... (recurring)
136 SPEED ARITHMETIC

Eg. 213/703
213/703 — 213/699 — 213/700 =2.13/7 = AF
The denominator 703 is in excess of 699 by 4 Therefore, we subtract 4 times the Q at
every stage as follows. .

(1) Jal oit 30

3

(11) (330- ( 4 x 30))/7 = 210/7 ikpy

0

|2> =| |
a
Gui) = (30—(4 x 30))/7 II a~

Guwe eG NPR eset Oph Shed 4/0 vNS Oe

4

(v= C4-8-0=
"(4X B0)'7 =F 6 0/7 [22
6

(vi) (6 N| | | iN x
Ss
iw) N]SS —~ I
a Wn | = bet ay and so on
2

i993/703.=1,2,0 43030.1 9. 8029 7 6 = 0 3000R

IGT iee

Eg. 223/593
223/593 — 223/599 — 223/600 = 2.23/6 = AF
Deficit of the denominator 593 from 599 is 6. ... we add 6 times the Q at every stage .

(i) 2.2316 ve Day

(ii) -(1374(6 B76 =359/6 - ih”/

j/
v/s
(iit) (596 X59))/6=913S/6" / nes
|
 AUXILIARY FRACTIONS 137

Here Q = 152 has 3 digits instead of 2 digits as it should have and since the steps
to follow would consequently entail larger numbers of digits for the Q (as additions
obtained from 6 times the Q’s are involved at every step), we go back to step (ii) and
write

(ii) 359/6 fsYh

(iii) ( 160 +(6 x60))/6 320/6 -7

(iv) (253+ (6 x53))/6 =571/6

(v) (195.6 <95))/6 =765/6
This again contains 3 digits in the Q.
“. we go to Step (iv) to rewrite 571/6 as

(v) (596 +(6x96))/6 = 172/6 fs

/

(vi) (428 + (6 x 28))/6= 596/6

(vii) (29946 x99))/6= 893/6 ==f148 sail <8 on
“123/503. =11037 2160 53096928109 ves
|
= 0.37605396290048
The method of AF as the above illustrations show can thus be one of the
alternative tools for converting a fraction into decimals besides the methods discussed in
chapters 6 and 8. These methods are unquestionably far shorter and quicker involving
only mental calculations compared to the long, tedious and cumbersome method of long
division currently in practice.

Exercise 9

Express the following fractions in decimals (to 9 decimal places ) using the methods of
this chapter.
(1) 7/99 yo 999 (3) 146522/89799 (4) 9/91 (5) 1425/601
(6) 951/1007 (7) 3500/3993 = (8)_ 123/68 (9) 1259/694 = (10) 309/502
 138

CHAPTER 10

EXTRACTION OF SQUARE ROOTS

An area of arithmetic operation that is frequently encountered is finding power

roots of numbers and particularly square roots and cube roots. We shall therefore discuss
methods for extracting them by methods which are by far simpler and time saving
compared to the prevailing methods. The methods explained here would be found to be
handy especially when very large numbers are involved for which even the aids like four-
or seven figures logarithmic tables, or the commonly available 10-digit calculators and
such others may not be sufficient.
In the present chapter, we consider the square roots and this is followed by cube
roots in the next chapter.
Before describing the method, let us list down the various properties associated
with square roots.

10.1. Some Properties of Square Roots.

(1) For taking a square-root, the given number is split into 2- digit groups from right
to left; if in this pairing, a single digit is left at the extreme left, then it, by itself is
considered to form a group. In the case of a pure decimal number, the grouping in
pairs is done from left of the decimal to right; and in the case of a mixed number,
consisting of both integers and decimals, pairing is done from right to left (starting
from the decimal ) for the integral part and from left to right ( again starting from
the decimal) for the decimal part.
(2) The number of digits in the square roots will be the same as the number of groups
; made up of 2- digits including the single digit if any; for example the square root
of 3946 will have 2 digits and 17143 will have 3 digits.
In general, if a given number contains N digits, its square root will have N/2 or
(N +1) /2 digits.
(3) The number of digits in the square of a number of pure decimals is always double
the number in the square root. Thus
(.8)° = 0.64, (.0339)* = 0.00114921
Also, it is worthwhile to remember some of the features that we have enumerated
in table 2 of the first chapter when we discussed the squares, cubes etc. of numbers. In
particular,
 EXTRACTION OF SQUARE ROOTS 139

(a) If the ending — digit of a number is 0, 1, 5 or 6, the square of that number too ends
in 0,1,5 or 6 respectively.

(b) The square of a number and its complement from 10 has the same ending digits,
example 3° and 7° or 4° and 6’ etc.
(c) A perfect square cannot end
(1) in’ 2;3,7 of:8
(41) with odd number of zeros
(111) in 6 with its previous digit being even.

10.2. Duplex and a Method of Finding It.

In the procedure of extracting square roots that follow, we will use what are called
the Duplexes (D’s) of numbers. We shall therefore explain the method of finding them
before taking up the square roots.
The table below gives the duplexes for extracting five digits of the square root of a
given number. The general method of finding them is discussed immediately after the
table.
Table 1: Duplexes for finding square roots

No. of digitsin Symbol for Duplex (D) Example

the square root the digits
of a number

One a a” a=6,D=6°
= 36
Two ab 2ab a=4,b = 2
Di Kaka eo)

Three abc 2ac +b? a=1,b=3,c=2

D=(2x 1x2)+3°=44+9=13
Four abcd 2(ad+bc) A=Ssebe= Ore e4dad=d
DE23 214-10 x 4)]=6

Five abcde 2(aet+bd)+c* ee Wipe area? 2d Sifie.=3

D =2 {{2x 3)+(4x 1)} +2°=24
140 SPEED ARITHMETIC

A simple method of finding the duplexes listed in column 3 of the above table for one,
(WOR ts. digits 1s as follows:

One digit a ed
(a)
a

Two digits a b 2ab

(a,b)
a mae b

Three digits geet ote 2ac +b’

(a,b,c) Tetatedt
a bere

Four digits a b Cc d 2ad + 2bc

a b Cc d

Five digits a b c 9d =e 2(ae + bd) + c?

(a,b,c,d,e) es <et

and so on.

As may be seen, the method of finding a duplex is similar to the one using the
vertical and cross products (VCP) of digits and employ the same diagrams as explained
in section 4.5 under the chapter “Multiplication”. The only difference now is that the
digits above and below the lines are the same.

Using the fact that the elements in the top and bottom lines are the same we can
obtain the duplexes in another form: (14) when the number of elements a,b,c,d ....is even,
write them along a row as before. Then the duplex is got as twice the sum of products
formed by multiplying the corresponding pairs of elements placed symmetrically on the
left and right about the middle digits starting with the two outer- most elements and
moving inwards as shown below.
 EXTRACTION OF SQUARE ROOTS 14]

Eg. a ; Duplex = 2ab

Vitek boil
a P¢ d Duplex = 2 (ad + bc)

Se.
a si ral f Duplex
=2 (af +be+cd) andsoon.

(11) when the number of elements is odd, we first form twice the sum of the products
of elements exactly as above leaving out the middle element . To this, we add the
square of the middle element. This final sum is the required duplex.

Egs. a Duplex =a’

eg ae Duplex =2ac + b°
| ;
per
) eid
aes Duplex
P = (2(ae
+ bd) ) +c”

a HCAs te ap Tes Duplex = 2 (ag + bf +ce) + ee and so on.

10.3. Method of Extracting Square Roots.

The method starts with grouping the digits of a whole number whose square root
is required, in pairs, commencing from the extreme right of the number. It then uses the
fact that the duplexes at each step for one, two, ..... digits of the square root equals the
gross dividend (g.d.) at that step. The method of finding the g.d. is better illustrated with
the help of examples that follow than through verbal description.

To begin with, the exact or approximate square root of the first pair of digits at the
extreme left of the given number is found which corresponds to the first digit ‘a’ of the
square root. This is then employed for determining the first g.d. The next duplex 2ab
equals this g.d. On dividing it by 2a, we get a quotient (Q) which corresponds to ‘b’
(the second digit of the square root) and a remainder (R) which is used for finding the
second g.d.
142 SPEED ARITHMETIC

This second g.d. is then equated to the next duplex 2ac + b* from which we get the
third digit ‘c’ of the square root and a new g.d. and so on.
We repeat this process as often as necessary until we have exhausted all the pairs
of digits of the given number if it is a perfect square .or until we have obtained the
required number of decimal places of the square root if the given number is not an exact
square.
It may be noted that the above procedure has 2a as the divisor at each stage of
finding a new digit for the square root.
We now illustrate the method with the help of a number of examples. Wherever
possible, we also provide shorter methods as alternatives.

Eg. V6241
(i) Grouping the number into pairs of digits,
we have the first pair 62 (on the left) / 9 °
Thus, a= V62 =7 approximately so that 2a= 14 anda’ =49
(11) 62 =a" = 13.
To find the first g.d., we annex the first digit 4 of the second pair of digits
41 so that the required g.d. = 134.
(iii) 2ab = 134 > b=9/8 (meaning Q=9 and R= 8)
The next g.d. is got by annexing the second digit 1 of the second pair 41
to R=8 so that we have 81 as the g.d. at this step.
Thethird duplex for finding c is 2ac +b’. This equated to 81, (the g.d. of step
(11)), gives 2ac = 81 — b? (called the net dividend, n.d. for short) = 0 >c = 0/0.
. The square root = 79,

a= 7 (as above)

(il) The second pair 41 of the given number ends in | so that the square root may end
in 1 or its complement 9. To decide which one, we-take the help of the d.rs. (the
digital roots), whenever the given number is a perfect square. The d.r of 71 (if we
choose |) = 8 and the d.r. of its square is 1. The d.r. of 79 (if we choose 9) = 7
and the d.r. of its square is 4, the same as that of the given number 6241. .. We
select 9 so that the required square root = 79.
 EXTRACTION OF SQUARE ROOTS 143
a.

Check: For numbers which are perfect squares (and this may be ascertained by
employing the thumb-rule described in section 10.6 at the end of the chapter), we may
check the correctness of their square roots as follows: We apply the VCP to the square
root 79. Thus,

gives Ay Gy/ +1
7 9 12 /8 =6241 (the given number)

Eg. 45369 (For this and other examples to follow, we proceed to find the duplexes
for determining a,b,c,d,e at each stage on the same lines as we have demonstrated in the
above example.)
(1) a=V4 =2:; 2a=4and a’ =4

(ji) 4-a°=0; «gd. =05

(iii) 2ab=05=> b=I1/1 «. gd.=13
(iv) 2ac=gd.—b’=13-1=12 >c= 3/0
= bne square root — Z21'3

(1) a= 2 andb =! (as above)

(11) the ending digit of the pair 69 of the given number is 9 so that the root ends in 3
or 7. As before, using the d.rs. we choose c = 3 so that the square root 1s 213.

By VCP, we have (213) =

2: 53 4 /4/ 3/6/ 9 = 45369 (the given number)

a
ome
oe:
Eg. V1521 ani :
@) a=V15-=3; «. 2a=6anda’ =9
(ii) I5-a°=6 -. gd. =62
fie. 2ab = 62,5 )= 1072" gd= 2)
Since b= 10 is a double digit number, we alter it to the single digit 9 so that
Zab = 62 => b= 9/8 and the g.d.= 81
(iv) 2ac =g.d.—b* >c=0/0
”. The square root = 39.
 144 SPEED ARITHMETIC

The earlier VCP check can be made again to verify the correctness of the answer.
Or
(1) a = 3 (as above)
(11) The ending digit 21 of the number 1521 pee} | which leads to b=
1 or 9. Again, using the d.r. method, we select 9.

Eg. 289444
(ii ni tatsN280m 55 22al=10-andai=25
(ii 28 ag => 3 cord 30
(ili)o 2abt= 39 2>ib=3/9 -. g.d.=94
(iv)e 2act=g.de—b? =94-9=85 >c=8/5 «. gd. =54
(v) oad be) = 54 => d= O/6. .-. ¢.d = 64
(vi) 2(ae+bd)+c° = 64> e=0/0.
.. The square root = 538
Or

The last pair of digits of the given number is 44 whose ending digit is 4 giving rise to c =
2 or 8. The application of the d.r. method leads to the choice 8 so that the square root is
538.

og. 410225
(i) a=v21 =4; 2a = 8and a =16
Gip. Qe =5° siedi= 56
(iii) 2ab=56=> b=7/Oandg.d.=02
(iv) Zac=02-b?=2-49 =-47(nd)= 47 =
(v) 2(ad+be)= 72 => 2ad = 72-2be= 72- ii
yy oR bE ae)
(vi) 2(ae+bd)
+c? =25 => 2 (ae + bd) = 25-25=0>e€=0/0
Thus, the square root is 47 5 = 465
 EXTRACTION OF SQUARE ROOTS 145
4

[Note : In step (iv), we have 2ac = 2 — 49 = - 47 (a negative n.d.) which means that
we
have to go back to step (iii) to write b = 6/8 so that we maintain 2ab = 56. To avoid
moving backwards, and thus to cut short the number of steps, we can employ the
vinculums as we have done by writing -47 as_ 4 7 so that we keep moving forwards. To
understand steps (iv) and (v) which involve vinculums, we have written them down in
full details.|
Or
The last pair of digits is 25 whose square root is 5. This gives the square root as 47 5 =
465 straightaway as above but without the need for finding d and e.
Eg. V7592
We know from section 10.1 that a number ending in 2 is not a perfect square. Let us
therefore find the square root of 7592 to (say) two decimal places.

(i) a=V75 = 8, 2a=l6anda’ = 64

Gi) -a=11
« gd =119
(iii) ~~2ab=119
> b=7/7 «. gd. =72
(iv) 2ac+b*=
72 >c=1/7 ..g.d.=70
Now we may add as many zeros to the right of the last digit 2 as are required to
get the square root to specified number of decimal places.
(v) Z(ad + be) = 10 => -d =3/8 ~. g.d. = 80

(vi) 2(ae+bd)+c°
=80 Se = 2/5 ». g.d. =50.
We could carry on with the g.d. = SO to find digits beyond e. However since the
Square root is to be found only to two decimal places we stop going further and write the
square root as 87.132 which to the desired decimal places is 87.13
[Note: The rule for the placement of decimals with square roots of non-square numbers is
exactly the same as 1s well-known].
Eg. V23409

(ij) az=v2 =1; 2a=2,a =1

i). 2—-a =1 od =13
(ii) «6Zab= [3- =b=0/1 “i g.d.=14
146 SPEED ARITHMETIC

(iv) Jac+b?=14 = 2ac= 14-36=-22(nd.)= 22 >c= 1 1/0.

However since the Q is a double digit number, which is not permissible, we go to
step (iii) and write b = 5/3 .. g.d. = 34.
We now return to step (iv).

(iv) 2ac +b? =34>c=4/1 «. g.d.=10

(v) 2(ad+bc)=10 > 2ad= 30d= 1 5/0.
Once again Q = 1 5 which is a double digit number . Therefore we rewrite the
value of c is step (iv) as 3/3 with g.d. = 30.
Returning to step (v) we have

(v) ZA Oe) = Ut CO = 0 ang. 0.Ga— 9

(vi) 2(ae + bd) +c? = 9 Se =0/0 so that the square root is 153.
Or
Having got the value of the digit b = 5 in the revised step (111), the last digit c may be
found by noting that the ending digit of the last pair 09 of the given number is 9 leading
to 3 or 7, and we decide on 3 using the d.r. approach.
The shunting back and forth in the later steps in this problem is a result of the fact
that the first pair of digits is 2 which is very small. To circumvent this difficulty, we
resort either (1) to the method of multipliers (in the next section 10.4) or (ii) to the
method of regrouping (section 10.5), wherein we have reworked the same example to
demonstrate how either of these two methods cuts short the above lengthy and tedious
procedure.

Eg. \0.09004513 (to 6 digits)

(i) a = VO9 =3:2a =6and a =9 a ibe dene

(ii) 09-a°=09-9=0 «gd. =0 21.0ow

0) 0 scfiea Sst) 2
(iii) 2ab=0 >b=0/0 ». gd. =0
(1) 2ac=0-b'=0>c=0/0 «. g.d.=04 =4
(11) Qad=4-2be=4 Sd= 0/4+. gd. =45
(iii) 2ae = 45 — (2bd +c*) =45 Se =7/3 «. gd.
=3)
(iv) 2af
=31 - (2be + 2cd) = 31-0=31 > f=5/1 «. g.d.= 13
 EXTRACTION OF SQUARE ROOTS 147
4

(v) 2ag = 13 - [2(bf +ce) +d? ] = 13> g = 2/1 and so on.

.. Square root = 0.300075 (to 6 digits)

Eg. V0.007811 = V.7811/100 = (1/10) V.7811
.. We find the square root of 0.7811 and divide the result by 10 to get the square root
of the given number.
V0.7811
(1) a= V78 = 8;2a=16, a =64
(ii) 78 =a = 1400. gd 141
(iii) 2ab = 141 > b=8/13 ». gd.= 131
(iv) 2ac + b° = 131 > c =4/3 «. g.d.
=30
(v) 2(ad + bc) = 30 = 2ad = 30-64=-34= 34
Se od 320
(vi) 2(ae
+bd) +c? = 20>¢€ =0/ 4 «. gd.= 40
(Vil) Qaf+bet+cd)= 40>f= 1/8 «gd.= 80
(Vili) Qagt+tbf+ce)t+d’= 80>¢g= 4/4 »«gd.= 40
(ix) 2(ah + bg +cf +de)= 40 >h =2/0.
-. Square root of 0.7811 = 0.884 20 1 42 =0.88379862.
.. Square root of the given number = 0.088379862. (to 9 decimal places)

V92.3521
a = V92 =9; 2a = 18, c =81
92-97 =11 «. gd.=113

2ab= 113 > b=6/5 ». g.d. =55

Qact+b?=55=> 1/1 «. g.d.=12
2(ad +bc) = 12 > d=0/0 «. g.d.=01
Q(ae
+bd) +c’ =1 >e=0/0
.. The Square root = 9.61
 148 SPEED ARITHMETIC

10.4. A Simpler Method Using Multipliers.

The method considered above can be sometimes tedious (as we have noted
above) in view of the fact that large instances of negative n.d.’s may be encountered
entailing fresh calculations and more time. It will be particularly so, if the starting pair
has a single digit or two small digits. This can be avoided to a sizeable extent either by
resorting to vinculums as we have done in that section or by the method to be described
now.
Given the number whose square root is to be found, first multiply the same by
integer powers of 4,9,16 etc....... so that, a reasonably large number (whose square root
may be worked out mentally) is available on the extreme left. The square root of such a
modified number can then be found on the same lines as described in the last section;
and the square root so obtained may be divided by the square root of the multiplier used
at the start.
Let us now work out some examples which we have considered in the last section
by the present method, to enable comparison in terms of speed and the number of
negatives.
Eg. 23209
As we have noted in the previous section, the first pair of digits is 2 which is so
small, that it makes the regular method very lengthy, as we have seen, as a result of
negative n.d.’s occurring too often. Even the adoption of vinculums does not reduce the
burden of work. :
We therefore resort to multiplying the given number by say 16 so that we have
23409 x 16 = 374544 so that its first pair of digits is 37, relatively a large number.
Having found its square root, we divide it by the square root of the multiplier 16 namely
by 4 to get the square root of the original number.

374544
(i) a = V37=6 ;.. 2a= 12, a? = 36
ii) “aia mile Se dela
(iii) %2ab=a14 Sb=1/2- 4 2.de=25
(iv) 2ac+b’=25
>c=2/0 «gd. =04
 EXTRACTION OF SQUARE ROOTS 149 .

(v) 2(ad+be)=4 >d=0/0..d=04

(vi) 2(ae+bd)+c?=4 >e=0/0
so that the square root of 374544 is 612. On dividing it by 4, we get 153 which
is the square root of the original number .23409.
The above shows how simple and quick this alternative method is, in
comparison to the one adopted in the previous section.

Eg. 216225
Let us now take this example which we have considered earlier and which also
involved negative n.d’s and consequently a longer time.

Multiplying 216225 by 4 we get 864900. We find its square root in the usual
manner as follows:

(i) a= V86 ~9;2a=18 andc’=81

(ii) . 86-81=5 «gd. =54
(iii) 2ab=54 >b=3/0 «. g.d.=09
(iv) 2ac+b?=09>c=0/0
“. The square root = 930.

Or
864900 = 8649 x 100. .. The square root of 864900 = square root of 8649 x 10
so that we need to find the square root of 8649 only. As before

(i) V86 =9
(il) The ending digit of 49 is 9 so that the square root may end in 3 or 7. The
application of the d.r. method selects 3 so that the square root of 8649 is 93. This,
on multiplying by 10, gives 930 as before. On dividing this by the square root of
the multiplier 4 viz by 2, we get the square root of the original number as 465.

This method once again shows the elegance and speed with which the extraction
can be effected.
150 SPEED ARITHMETIC

10.5. Method Based on Regrouping.

Another method that will simplify the work considerably is to consider not just
the beginning group of digits but to adjoin with it, the next group as well and to take the
square root for the number so regrouped. Let us take one of the previous examples:

Eg. 23409

(i) a= V234 =15 .. 2a=30, a*=225

(ii) 234-a7=9 ». gd. =90
(iii) 2ab=90 >b =3/0 «. gid. =09
(iii) 2ac+b?=09>c=0/0
The required square root = 153. Thus this method too, demonstrates how
speedily the calculations can be effected as compared to the ones made in section 10.3.

Eg. v45369
We now redo this example (considered in section 10.4) by the method of
regrouping of the digits to illustrate how speedily we can obtain its square root compared
to the straight forward procedure.
Take the first 3 digits together of the number, so that

(i) V453~ 21; ..2a=42; a=441

(ii) 453-217 =12 -. gd.= 126
Gian -2abis3126 = TE Oresetater 9

(iv) 2ac+b°=9>c=0/0
. The square root = 213.

Or

The last pair of digit is 69 whose ending digit is 9 giving the choice c as 3 or 7.
Again applying the d.r. tool, we find that c = 3 is appropriate giving the square root as
pates
 EXTRACTION OF SQUARE ROOTS 1S]

Eg. V6017209
(vi) a= V601 ~ 24; 2a= 48 anda? =576
(vii) 601 —a’ = 601-576=25 «. g.d.=257
(viii) ga= 257 b= 517 © od. = 172
(iv) 2ac=g.d.- b’=172-25= 147 >c =3/3 «. g.d. = 30, and so on.
Or

The square root of the last pair of digits is 09 is 3 .. c=3

.. The square root = 2453.
Eg. \V79816356
As above, we take the first 2 pairs of digits 7981 to start with.
However, since the square root of 7981 cannot be guessed easily as in the earlier
examples, we do a little preliminary work as follows:
Taking the pair 79 of 7981 we have

(a) a = V79 ~8 so that 2a= 16 anda’ = 64

(b) T4-a = 15; also. the.¢.d. = 158

(c) 2ab = 158 > b=9/14 «. g.d.= 141

(d) Pie = ol),

With these, one has the nearest square root of 7981 as 89 and the remzinder as
7981 — (89) = 60. Using these results, we now proceed as usual to find the square root of
the given number 7981 63 56.

(1) a= V7981 = 89; 2a= 178

(11) 7981 —a’ = 60 .. g.d. = 606
(These two results are from the above preliminary work).

(iii) Zap O06 =) Di= 3/72... g.d, = 723

(iv) Qac+ be? = 723 >c=4/2 «. g.d.=25

(v) 2d + bey= 29 > d= 0/1 v. g:d-=16
 152 SPEED ARITHMETIC

(vi) 2(ae+bd)c* = 16 >e=0/0.

Thus the required square root = 8934

Or

The ending digit of the last pair of digits 56 is 6 giving d= 4 or 6. The d.r. method settles
for d = 4 giving the square root as 8934.

**10.6. A Thumb-Rule for Finding if a Number is a Perfect Square.

- To know if a number is a square or not is of some importance in determining the
stage at which the extraction of its square root may be terminated. For this reason, we
outline below some useful and simple guide-lines.
We have noted already in section 10.1 that a number ending:
(i) in 2,3,7,8 or (ii) with odd number of zeros or (ili) in 6 with its previous digit as even
cannot be a perfect square; and also that if the ending digit of a number is 0,1,5,6, its
square also ends in 0,1,5,6 respectively.
In addition, we now present below, a thumb-rule in the form of a sequence of
procedure:-

(1) Find, by the usual method of pairing, the number of integral digits (i.e. the non-
decimal digits) that the square root of a given number would have. As we have
said in section 10.1, the number of such digits would be, in general, N/2 or
(N+1)/2 according as the number of digits N in the given number is even or odd.
(i1) Find’ the digital root of the given number.
(iii) | When all the digits of the given number have been exhausted in extracting the
square root, find, the d.r. af the square root so obtained . Square the same and find
the d.r. of the squared quantity. If this latter d.r. is the same as that of the given
number one may, in general, conclude that the given number is a perfect square.
Let us now apply the above to a couple of examples.

Eg. 25210441. Its dr. is 1. Its square root would contain 4 integral digits which by
any of the methods discussed earlier can be found to be 5021. The d.r. of 5021 is 8 and
the d.r. of its square viz. 64 is 1. This is the same as that of the given number and which
therefore is a square.
 EXTRACTION OF SQUARE ROOTS 153

Eg. 702396 whose d.r. is 9. Its square root has three integral digits which are obtained
as 838 at the end of exhausting the last digit 6 of the given number. The d.r. of 838 is |
whose square is also 1, and this is different from the d.r. of the given number thus leading
to the conclusion that the given number is not an exact square. We may therefore
proceed to find the decimal digits of the square root on the same lines as illustrated in
section 10.3.

KK

Exercise 10
Find the square root of the following:
(1) 4096 (2) 6879 (3) er23795 (4) 295342
(5) 83267 (6) 998001 (7) 11881 (8) 16777216
(9) 0.032549 (10) 0.00173889 (11) 0.066954 (12) 7272.5078
(13) 196.175561 (14) 819.9062 (15) 782.8804 (16) 365.8279]
154

CHAPTER 11

EXTRACTION OF CUBE ROOTS

If finding of square roots of numbers by the conventional method is difficult and

tedious, finding of cube roots is even. more so. In this chapter, we present a few
comparatively simple methods (as usual, with plenty of illustrations) which the readers
will find interesting to follow and useful to practice.

11.1. Some Properties of Cube Roots.

Let us, as in the case of square roots, begin by listing out some of the features of
cubes of numbers, some of which we have noted in table 1 of chapter 1! under the
column of cubes.
(1) The cubes of the digits 0 to 9 have unique endings (unlike the squares) in that no
two digits have the same endings.
(2) The cubes of numbers ending in 0, 1, 4, 5, 6 and 9 have their cube roots also
ending in the same digits respectively.
(3) The cubes of numbers ending in 2, 3, 7 and 8 have their respective cube roots
ending in the digits which are complements from 10.

Thus, cubes ending in 2, 3, 7 and 8 have their cube roots ending in 8, 7, 3 and 2.

(4) The number of digits in a cube root of a number is the saime as the number of 3-
digit groups in the given number including a single or a two-digit group left out in
the 3- digit grouping. Thus,

the number 437 240 will have 2 digits in its cube root
25 072 306 will have 3 digits in its cube root
45 will have | digit in its cube root.

(5) As in the case of square roots, the grouping of digits is done from right to left for
non- decimals and from left of the decimal towards right in the case of decimals.
(6) Finally, the opening digit of a cube root of a number can be found from the first
3-digit group of the number.
Let us now show how the above features help us to determine the first and the last
digits of anumber whose cube root is exact.
 EXTRACTION OF CUBE ROOTS 155

Number Number of First digit in the Last digit in the

(exact digits in the cube root (obtained cube root
cube) cube root by guessing) (L)
(n) (F)
493 039 Z 7 9 (same as the ending digit)
216 l 6 6 (same as the ending digit)
2 744 2 1 4 (same as the ending digit)
[27 1G? 2 2 3 (complement of the ending digit 7)
5 000 211 3 1 1 (same as the ending digit)
2 875 403 448 4 1 2 (complement of the ending digits)
and so on
The above table shows that for numbers which are exact cubes, we can identify the
first and the last digits of its cube root as well as the number of digits that it will have.
even before extracting the cube root.
11.2. Method of Finding Cube Roots of Exact Cubes.
This may be carried out either by proceeding from right to left or from left to right
as we have done in the case of square roots in the previous chapter or in the case of
extracting the cube roots of any numbers as we have done in the next section of the
present chapter.
(1) To start with, we discuss the procedure for finding the cube root of a given
number by moving from right to left. To do so, we need certain algebraic results
(which we call excesses),similar to the duplexes (D’s) that we employed in the previous
chapter for square roots. The algebraic basis on which these are determined is
provided in Appendix | at the end of the book for readers interested in the same. These
excesses are listed in table 1 below for ready reference.
Table 1: Excesses for cube roots when proceeding from right to left of a given number.
Last Penultimate Third Second First digit
digit digit digit digit
GM) dh); .@) (T) (S) (F)
hee) Cs = : 3L°F
Ore Hn 8 -3LP : : 3LP743L7F
Avellile 3. 3EP : 3LP7+3L7S 3L7F+P°+6LPS
See SEP 3L°T+3LP* 3L7S+6LPT+P? 3L°F+3TP7+ 3LT°+6SPL
156 SPEED ARITHMETIC

As may be noted, these are the top three expressions (for n = 3) and the top four
expressions (for n= 4) given in Appendix 1
With this table, let us illustrate the procedure of finding a cube root of an exact cube. To
check if a given number is an exact cube, we may apply the thumb-rule given in section
10.6 of the previous chapter.

Eg. Cube root of 2744; heren=2 andL=4.

(i), samo 2 744
Subtracting 64, we have ( omitting the last 64
zero after subtraction) 2 168

(ii) Form 3L’F. This equals 48F (since L=4) and

should end in 8 (the last digit of 268)
.. F = 1 or 6. However F = 6 is obviously ruled out since the first group
consists of the digit 2.
*. The cube root = 14

Eg. -Cube root of 314 432);n=2,L=8.

Ges Te oe Sade subtracting ( and as usual 314 432
omitting 0 at the end) S12
(ii) 3L°F = 192F and this should end in 2. ais ace 2
IF = Lor 6.

Now F = | gives F’ = | and this is far removed from 314 (the first 3 digit group of
the given number) whereas with F = 6, we have F’=216 whichis close to 314. ». We
take F = 6 (and since L = 8),
The cube root = 68.

Eg. 5 000 211

Here n= 3, F =1 and L = 1. Once we have found the penultimate (i.e. the middle)
digit of the cube root, the work is complete.
Cia ee Pye Se Siaraere 5 000 241
(ii) Next form 3L’? P. This equals 3P ' l
(since L=1) which should end in | 5 000 21
- P=7. Subtract
3P = 21, 21
(iii) We already know that the last digit L = 1. 5 000
.. The cube root = 171
 EXTRACTION OF CUBE ROOTS 157

As we have said earlier, it may happen in some cases that one may not be able to
guess the first digit correctly. In such a case, we may proceed further as follows.
(iv) Form 3L*F + 3LP’. This equals 3 x 1°>x F+3x1x7?=3F + 147.
Now this should end in 0 (since we have 5 000 above).
*<, 3F ends in 3 and'so F = 1.

Eg. Cube root of 1367631000

Here, since we have the last3 digits as 0’s, we divide the given number by
(10)? = 1000 to get the modified number 1367631 (If this is not done, we will take L=0
and we cannot proceed further). Let us therefore find the cube root of the modified
number 1367631 and multiply the resulting cube root by 10 (the cube root of 1000 with
which we divided the original number to start with).
Forsl56/63', we have n= 3, L=1 and F=1
@) [2 =1: Subjract-J ha367 263!
(ii) 3L°P = 3P which should end in3 J
giving P= | li 367: 63
a Subtrace SP’ =3 3
(ii1) We have already noted that F = | or it 1 367 6
may be found from 3L’F+3LP” = 3F + 3;
this should end in 6;
1.e. F should end in 1.
.. The cube root of the modified number 1s 111 and the cube root of the required
number 1s 19h KA OHA TIO:
Let us now take up an example of exact cubes having n=4(1.e.4 digits in their
cube roots).
Here corresponding to n = 4, we shall have 4 letters, F (first digit), P (penultimate
digit), L (last digit), all as before and S (to stand for second digit).

Eg. Cube root 1902014919

Here 7 = 41 = 9 =)
(i) L’>=729; Subtract 729 . 1 902 014 919
729
(ii) 3L’?P = 243P ending in 9 1 902 014 19
-. P = 3. Subtract 243P = 729 pins 29
1 902 006 9
(iii) 3L°S +3LP* = 2438S + 243 72 9
and should end in 9 ie 90 934
”, S = 2. Subtract 24384243 = 729
158 SPEED ARITHMETIC

(iv) We have already taken F = 1; but this can be found, for n = 4, from 3L7F + 6
LPS + P® =2 = 243F + 324 + 27 = 243F+351 and begets end in' 4. tex PF =1
The cube root_= 1239
Finally, we shall describe the procedure for finding the cube root with n= 5 digits,
of an exact cube. We shall now use the symbol T to stand for the third digit, the other 4
digits F, S , P and L having the old meanings.
Eg. Cube root of 1003604321728 ;
here tee. Fo= 1 Le 2.
1, Q03,.,,004....32 le aeetee
(i) L? =8; subtract ; 8
(ii) Form 3L’P. This equals 12P A542 003) 9 604, oAB2dbiv F2
and its ending digit is 2 so that 12
P= PCr 6; [25'' 003 > "S604" Ss Z Te 'G
(i)+-Bbetusstake- P= I-so that 12P=12 6
which we subtract. 1 003 604 321
(iv) Form 3L7T + 3LP? which, with I

Pe tequals i216. Lins 1 003 604 a4

should end in 61.e. 12T should
end ins@ tso.tnate| =<OmnmSaicleet
us take T = 0 giving 12T +6=6
which we subtract.
(v) Form 3L’S + 6LPT + P*. With L = 2, P=1.T=0, this ates
128 +0+1= 12S +1 and this should end in 1 so that 12S shouldend in 0O
i.e., S=Oor5. Subtract 128 +1 =1,
(vi) Form 3FL* + 3TP’ + 3LT? + 6SPL. With L=2, P=1, S=0=T, this gives
12F + 0+0+0 = 12F which should end in 2. i.e. F should be 1| or 6.
Obviously, F should be equal to 1.
Thus, we are certain of the first and the last digits only ; whereas P , T and S can
take any One orthe 2values cach (viza) F = 1.or 6, 1 = O'or Stand os = 0 or 5) giving
rise to 2° = 8 possible combinations. Evidently this is not practical. We therefore seek
an alternative method, which we now describe.
(a) An alternative method: 5
This method is somewhat similar to the one that we have considered in the earlier
chapter on square roots. It may be remembered that we multiply the given number by a
suitable multiplier so as to get large digits in the extreme left; and having got the
 EXTRACTION OF CUBE ROOTS 159

required root of the modified number, we divide it by the multiplier initially chosen so as
get the final answer.
We adopt the same procedure here too, except for the difference that instead of
multiplying the given number to begin with, we divide the same and extract the cube root
for the number so divided. Let us illustrate with the same previous example where we
have not been able to pinpoint the values of S,T, and P (unless we are ready to work out
all the possible 8 combinations).
We take the given number 1003604321728. Using the test of divisibility, we find
that the number is divisible by 8 successively. The first division by 8 gives
125450540216 and a further division by 8 gives 15681317527 which is an odd number.
We would now have n=4, F=2, L=3 for its cube root. We proceed, as before, with
the latter number as follows.

Gi) L=3;L* =27 PSs ius ep ~527

Subtract 27 2]

(ii) 3L°P = 27P which should end in 0. 15 681 317 50

SP =Ossuotracto 7 =O: 0

(iii) 3L°S + 3LP? = 27S which should 15° 681 90399755

end in 5, i.e. 27S should end in 5 1Ba*5

sothat S=5 .. 27S

= 135. 15 681 304 O
which we subtract.
(iv) And, finally F = 2 (by observation)
The cube root = 2503

Since we require the cube root of the original number whereas the above is the
cube root of the number modified by dividing it by 64, we therefore multiply the above
cube root 2503 by 4= (8x8)' to get 10012 as the cube root of the given number.
We also note that of the ambiguous values: P= Lior. 6; S$ =.0°0r5;/T. =, O°or’5, the
correct cheicé Should be P= 1,.S=0, T=0.

(b) Another use of the alternative method

The method just described can also be used whenever a given number has large
number of digits and consequently a large value of n. Let us explain by means of an
example.
160 SPEED ARITHMETIC

Eg. Cube root of 2299968 ; here n= 3.

Dividing the number by 8, we have 287496 and one more division by 8 gives
35937 which has n = 2. We can now easily (in fact, mentaHy) find its cube root since n
= 2. We see that F = 3 and L = 3 by observation so that the cube root of 35937 = 33;
and on multiplying this by 4, we have the cube root of the original number as 33 x 4 =
Use
The above examples may suggest that reductions such as the above can be done
only when a given number is exactly divisible by 8 and / or its integral powers. But this is
not so. The requirement is that so long as a given number is divisible by any other
number which itself is a cube (like 8 being a cube of 2), the method succeeds.
Eg. Cube root of 66923416
Dividing this by 7° (i.e. successively by 7, three times) we get 195112 where n = 2;
also L=8 and F = 5 (which can be mentally worked out) so that the cube root of
195112 is 58, on multiplication by 7 this gives the cube root of the original number viz.
58 x7 = 406.
Before ending this section, let us consider an example which is of a type that 1s
somewhat different from the above.

Eg. An exact cube has 9 digits and its last 4 digits are 1344. Find the number and its
cube root.
We have n= 3 (grouping 9 digits of the number into 3’s ). Also, L = 4 (since the
last digit of 1344).
(i) L* = 64; subtract 64 ; 1344
Penmaes
(ii) 3L°P = 48 P which should end in 8. 128
be P= OF. 48
(a) Let us take P = | ; then 08

(iii) 3L°F + 3LP? = 48F + 12 and this should

end in 8 .. 48 F should end in 6 i.e. F=2.
If F=2, the cube root is 214 and the given number is (214)° = 9800344, the last
4 digits of which is not 1344 as given: on the other hand, if F = 7, the cube root is 714
whose cube ends in 4344 as required.
(b) Letus take P= 6; then 48P = 288 128
which we subtract ; 288
(iii) next, 3L7F + 3LP* = 48F +432 andthis 84
should end in 4 1.e. 48F should end
 EXTRACTION OF CUBE ROOTS 161

in 2 1.e. F = 4 or 9 If F = 4, the cube root is 464 and its cube has 7344 as its last 4 digits
and not 1344 as required.
. If we take F = 9, we have the cube root 964 whose cube is 895841344 and its
last 4 digits are 1344 as given.
Thus, the number is 895841344 and its cube root is 964.

(ii) | Let us now consider the method of finding the cube root when we move from
left to right. The corresponding table of excesses for different digits (n) of the cube root
is given below.
Table 2: Excesses for extracting cube roots from left to right of a given number

First Second Third Penultimate Last

digit digit digit digit digit
(n) (F) (S) (T) (P) (L)
Q) oF = a ce 3F°L
G3) ak = =e 3F°P 3F°L + 3FP?
(4). ee 3F°S = 3F°P+3FS* 3F°L+S? +6FSP
fo tapas ve 3F°S)3F°T+3FS* -3F°P+6FST+S°> 9 3F°L+3S°T+3FT?+6FSP

The above entries correspond to the bottom three expressions (for n = 3) and the
bottom four expressions (for n= 4) given in Appendix 1 at the end of the book.
The procedure of finding the cube root in the present case ts similar to the one
that we followed for the square root. Broadly speaking, it is as follows:
(1) Having grouped the numbers into three digits as usual, we determine the cube
root (exact or approximate) of the first group in the left and designate it as F.
(11) We subtract the first group of numbers from F° and append the first digit of the
second group to the above subtracted value. This gives the first g.d.
(iii) The next digit S of the cube root (say for n= 4) can be obtained by equating the
excess 3F’S to the above g.d. where F is known in step (i). Division of the g.d. by
3F° gives Q = S and a remainder R.

(iv) Appending the R to the next digit of the second group of the three digits in the
given number, we have the second g.d.

(v) Equating this g.d. to the next excess 3F°P + 3FS’, subtracting 3FS* from the g.d.
and dividing again by 3F*, we have the next digit of the cube root viz. P and so on
till we finally get the last digit L of the cube root.
162 SPEED ARITHMETIC

It may be noted that in every division, the divisor is 3F° (similar to the
corresponding divisor 2a in the case of the square root).

While the above description is a general one, and most applicable to numbers
which are not perfect cubes, some modifications of the same (similar to the ones that we
employed in the case of square roots) can be adopted for numbers which are perfect
cubes.

Let us now demonstrate this method of moving from left to right by taking the
same examples (on perfect cubes) that we considered earlier when extracting from right
to left,

Eg. cube root (c.r.) of 2744

Bere i ="2, eee Ome approximately
Next, 744 ends in 4 . Therefore as mentioned under (2) in section 11.1, L=4
.. Cube root =14.

Check: d.r. of 2744 = 8; dr. of the cube root 14 =5; 53 = 125 and its d.r. = 8 = dr. of
the
given number.
Eg. cr. of 314 432
Here n = 2, F = 6 (since F? = 216 which is the nearest integer < 314). Also, the ©
next group 432 ends in 2 and therefore L = the complement of 2 = 8 (as noted |
under (3) in section 11.1).
8

., Cube root = 68.

S
Check: d.r. of the given number 314 432 =8; dr. ofits cube root = 68 is 5 and5 = 125
whose d.r. = 8 (same as that of the number).
Eg. cea oof 1 367 P6351 2000 1 367 alo de
Phe required'e. Percoll! 367 \634jS10
Consider 1 367 631; n=3
(je. 3
RCS Sa i ieee hee
ii oe
(iv) Ending digit of the last group 63lis1 .. L=1
. cr. of 1 367 631=111
“. Cube root of the given number is 1110.
 EXTRACTION OF CUBE ROOTS 163

Eg. cr. of 1 902 014 919

We may take the first two groups | 902 whose F = 12 (approximately); “.n=3
Pay A359
(i) 2b 1902-1798 =174, - gd.=1740
(11) 3F P=1740 => P= 4/12... od. =121
(iii) 3FP? +3F°L=121 = 3F-L=121-576=- 455= 455 >L= 1/23

Since we know that the given number is an exact cube and we have exhausted
all the digits of the c.r., we write c.r. = 124 1= 1239. (Also since the ending digit of the
last group 919 is 9, we take L = 9 as obtained above.
Eg. c.r. of 1 003 604 321 728; n=4 (since we take the two groups 1 003)

(1) =10, F’=1000, 3F’ = 300

(ii) (00G 2 = 2, 22d = 36
(111) BER =36°2> S =006 .. gd: = 360
(iv) 3F°P + 3FS* = 360 = 3F’P = 360 — 3FS’ = 360 > P=1/60
The ending digit of the last group 728is8 .. L= S2
‘. Cube root = 10012

eriofils 68h 317°527 -n=4:

Heh UMS HES cro? oe 1) =

Bot) eb
3F'S=76=> S=6/4 ». g.d.=48 ‘
3F’P= g.d.—3FS’= 48-216= 16 8 >P= 9 60
The last digit of 527 is 7 .. L=compliment of 3= 7
-. Cube root =26 9 7 = 2503

Since 3F°P <0, we take S (of step (111)) = 5/16 .. g.d. = 168
. 3F°P = g.d.—3FS*= 168 - 150=18 => P= 1/6
“. Cube root = 251 7 = 2503

cr. of 66 923 416; n=3

F=4: 3F* =48; F = 64

(i1) S662 Fo - se 29
(iii) 3F°P = 29 => P=0/29
The ending digit of 416is6 .. L=6, .. Cube root = 406.
164 SPEED ARITHMETIC

CT.O1s5 -000R2 Tika

Bal Eel ein
5 Ap eae ee = 40
3F°P = 40 = P= 13/1 and since P= 13 isa doubleMicit we take P=9/13
y ed pe t30
3F’L = 130 —3FP2 <0: we go back to step (iii) and write
P=8/f6~-and ¢.d-— 160
3F'L= 160-192 <0 .. we go back to step (iv) and write
PS Pros ance di 1 90
SF BOATS Oe Pa
Lastidiotor2ilis. | ooL=T
“. Cube root
= 171.

dit08 3.000 21 ivis 9

d.r. of c.r. has to be 9. We know (by inspection) that F= 1 and L=1 .. P has to
be 7 fords. =9 giving c.r.= 17).

er of 362, 46/¢ 097° a3 3

F= °V362 =7 (approximate); F° = 343; 3F° = 147
(11) 362-F=19 «. gd. =194
(iii) 3F’P=194 => P=1/47... gd. =476
(iv) 3F’L = 476 — 3FP’ which is > 0 so that there is no need to change the value of this
P. Having ensured this, we may proceed to find L from the equation in step
(iv).Or, since the ending digit of the last group 097 is 7%, we have L equal to the
complement of 7=3.
Cube roor= /13
11.3. Method of finding the Cube Root of Any Number.
We now take up examples of cube roots of any number. While in the case of exact
cubes, cube roots may be found either from right to left or vice versa as we have done
above, in the case of non-cubes, the last digit cannot be known since its end would extend
to decimals as well. Therefore we are forced to start with the first digit F and move from
left to right until we get the cube root to the required number of decimal places. Thus the
procedure that we just employed above applies.

Eg. 646 (to 2 places of decimals)’ snail, R=8.

 EXTRACTION OF CUBE ROOTS 165

(i)
The divisor d=3xF*=3x8?= 192
(ii)
646-F = 646-512 = 134
(iii)
g.d. = 1340 + 192 = 6/188
(iv)g.d. = 1880 ; n.d. = 1880 - 3FP” = 1880 - 3(8 x6?) = 1016:
1016 + 192 = 5/56
(v) g.d.=560; n.d. =560-(6FPL+P°) where F=8, P=6,L=5
giving 560 - (1440+216 ) which is -ve
*. We take 1016 + 192 = 4/248.
Now g.d. = 2480 and n.d. = 2480 - (6FPL + P*) where F=8, P=6
and L=4. .. n.d. = 2480 - 1368 = 1112; 1112 + 192 = 5/152 and so on.
As in the case of square roots, it is quite often that we come across -ve net
dividends at any given step in which case we have to go back to the previous step as often
as necessary until a +ve net dividend is ensured in the current step. To avoid this, two
approaches are possible.
Approach |: Multiply the given number by 2° , 3° or 4° etc. as many times as
necessary to find the first group of digits manageably large. (It may be remembered that
this is what we do for finding the square root under a similar situation) after finding the
cube root of such a modified number, divide it by 2, 3 or 4 respectively to get back the
cube root of the given number.
Eg. (2410)'”
Since the extreme-left group consists of only one digit viz. 1, we may multiply the
given number by 6° = 216 to get 520560

(i) divisord= 3x 8° =192

(ii) gd. =85+ 192 =0/85
(iii) g.d. = 856; n.d. = 856 - 3FP” (where F = 8, P = 0) = 856;
856 + 192 = 4/88
(iv) g.d. = 880 ; n.d. = 880 - ( 6FPL +P”) where L=4
-- n.d. = 880; 880+ 192 = 4/112.
 166 SPEED ARITHMETIC

(v)°. g.d.= 1120; n.d-= 1120 - (3FL? + 3P°L) = 1120 - 384 = 736;
736 + 192 = 3/160.
. cube root of the modified number 520560 = 80 . 443 ( to 3 decimals)
. cube root of the given number is 80 . 443+ 6 = 13.407.

Approach 2. This again is similar to what we do in the case of a square root. We take
more than one digit group to the extreme group if the former is made up of one or two
digits.

Eg. Cube root of 1124912

The extreme- left group has only one digit viz. 1. We therefore adjoin the next
group 124 to form 1124 as the starting digits to give 1124 912.

Since it may be difficult to guess the approximate cube root of the extreme left
group thus formed, we may do some preliminary work as follows.(We did similar work
in the case of square root too). Take 1124 and split it into 3-digit groups to form 1 | 124.
Proceeding as usual with this, we have 01+3 — 0/1

So, we take the approximate cube root of 1124 as 10 with remainder 124 (of course,
in the given example it is easy to get these results but the method ds general and applies in
difficult cases also).
So we proceed with 1124 | 912 using the cube root of 1124 as 10 and
the rem. = 124:

(i) divisor d= 3 x 10° = 300

(ii) 1249+ 300 = 4/49
(iii) 491 - 3FP* (where F = 10, P= 4) = 491 - 480 = 11; 11 + 300=0/11
(iv) 112 -(6FPL+P° with L=0)=112-64 =48; 48 + 300 =0/48.
Thus to 2 decimal places, the cube root of the required number
1124912 is 104.00.
 EXTRACTION OF CUBE ROOTS 167

Exercises.

Find the cube root of the following:

11(a).Exact cubes:
(1) 491169069 (2) 731432701 (3)e7 19323136 (4) 7301384
(5) 545338513 (6) 274.625 (7) 611960.049 (8) 804.357
(9) 0.830584 (10) 206425.071
11(b).Non-exact cubes (to 3 decimals)
(1) 36501298 (2) 63891352 (3) 19283746 (4) 44234387
(ay 0632 751 (6) G7S3 1217 (7) 3064327.8 (8) 183444.09
(9), 21801815 (10) 53879.564
 $34 ~ _ BPOGg

: i? ad + 0 @ttee votes <8

7 if. 899 .% 1a), .

: ile poore? tq sonddified eal suse 5

BBELOET (BL BEI sreit fe)

» Cameos nig: Eee"8
wae shies one Gigit Grouys tr Dre. ChILCHE oHE
a bei sale a *
7
et os cs

woke ert GPU saree é)

e aS, -_PROMENERI (8 <p soanPhp sper
Lt ‘ H a tyFi hey a 434-95 0 eS utprt TigMOS
Gigits
Ta be£0 POET

ee * Seite ot camghe dhe airvol t guess the approx) mate cube Foot af."he:“outs
-_ puy itep farmed: tee Aen) Go ea nem elis‘ ninary Werk as fotlsroi We did: situilas wor
ae i ae caadeae!¢ Chest
28 40) ") abet a wel splitters |rien"Bene tefei:4 1 24

enePres oting rm3 kena

—
- : =
as
i
F
ib.

5
sactigtbe
we ows

;
a. Bx
&
4
pit.
'
ay ine E -
%
be
~
‘Be 7me 'Li t: :
eas oor y| “et meee
i i i -
x Hi ~ 7 . F ‘6 ra Ts == ane
co ein ) ae a
7 : - ae «

oe” 2 ete Ge calles ee appre ites calle redst rate ~ ie 48 wilh remamndey take
In =} Cer < - ; a | be asy iO wet aes r erlhs, tess the risen jgreeay me

a Lae pe eity 118) 912 uring dhecsLoube Port oT lindas nad!

: : “they ki @ ar rs oe ae

% 7 ne. 7 ye a. 2 - é <

Se
*. 1. i9y ;
ee Ps Sa
;
ws :
: 4 ie P : es
- a a ee :
. <~
ae
5
7

« 1G HK _
~ <: Vau
ms ay ie .
oe » ae

Sen: SB sac ok; L-- 390=

me pe oA:ath Lee lis -¢ i ils 38; €20th = O/aR
u > detent. places, ite (3b beacheae
sr aa ©M. -

~ ie : ‘ i

. ; wal : \
~ :
ANSWERS
170 SPEED ARITHMETIC

CHAPTER 2
Exercise 2(a):
(1) 131618 (2) 1071032 (3) 260919 (4) 11330 (5) 102275
Exercise 2(b):
(1) 190708 (2) 20803 (3) 144001 (4) 19947678 (5) 114634184
Exercise 2(c):
(1) 543 (2) 11928 (3) 17500 (4) 19936 (5) 17500 (6) 10552
Exercise 2(d):
(1) 763 (2) 12 (3) 471 (4) 12496 (5) 4790 (6) 261
Exercise 2(e):
(1) 1252 (2) 1114 (3) 850 (4) 18574 (5) 1191
Exercise 2(f):
(1) 1824 (2) 73250 (3)32504 (4) 97740 (5) 9400712
Exercise 2(g):
(Ps 303 (2) 106 (3)64 (4) 211 (5) 202
CHAPTER 3
Exercise 3(a):
(1) 4000 (2) 3722 (3) 577.8633 (4) Rs.94651 Ps.80 (5) 27146.86
Exercise 3(b):
(1) 712 (2) 287.881 (3) 7253.84 (4) 17902.3746 (5) Kms.267 Mtrs. 634
Exercise3(c):
Cy tig 7 (2) 551 (3) 910 (4) 70.2 (5) 186.8939
CHAPTER 4
Exercise 4(a):
Cy 12 (2) 216 (3) 117 (4) 8544 (5) 9765 (6) 12198
(7) 828608 (8) 1074322 (9) 1130900 (10) 1083614 (11) 9509185 (12) 9492
(13) 7524 (14) 13284 = (15) 96915
Exercise 4(b):
(1) 736920 (2) 1184730 (3) 97324128 (4) 1007971920 (5) 987977390400
Exercise 4(c):
(1) 278685 (2) 87264 (3) 538395 (4) 40446 (5) 61335 (6) 387396
(7) 64293262 (8) 2523024 (9) 42432 (10) 765136 (11) 962125 (12) 402544
(13) 2563512 (14) 362043. ~—-(15) 42680
 ANSWERS 171

Exercise 4(d):
(1) 3999 (2) 7448 (3) 147645 (4) 23218 (5) 320264
(6) 135000 (7) 92722 (8) 27636 (9) 326145 (10) 165735
Exercise 4(e):
(1) 1748 (2) 4940 (3) 25288 (4) 19688 (5)56 767
(6) 76708 (7) 115994 =(8) 62702 (9) 33234 (10) 4608
Exercise 4(f):
(1) 999879 (2) 752988 (3) 98745 (4) 215016 (5) 40641
Exercise 4(g):
(1) > 5621 (2) 163691 (3) 720099 (4) 120475 (6) 2550271
(6) 1101824 (7) 2739021 (8) 156227 (9) 5402145671 (10) 2550025324
Exercise 4(h):
(1) 2475 (2) 18981 (3) 747 (4) 7425 (5) 100584
(6) 2565432 (7) 79975944 (8) 6237 (9) 47817 (10) 482517
Exercise 4(i):
(1) 35409 (2) 33732996 (3) 104362.2 (4) 356.4 (5) 999.590031
CHAPTER 5
Exercise 5(a):
CL) -623 (2) 7225 (Sy T1025 (4) 75625 (5S) 17598025

Exercise 5 (b):
(1) 64 (2) 529 (3) 974169 (4) 1028196 (5) 14884

Exercise 5(c):
(1) 3969 (2) 6561 (3),11236 (4) 966289 (5) 1032256 (6) 3948169
(7) 4016016 (8) 627264 (9) 261121 (10) 63504 (11) 381924 (12) 159201
Exercise 5(d):
(1) 884736 (2) 1061208 (3) 3581577 (4) 944076141 (5) 12167
Exercise 5(e):
(1) 456976 (2) 121550625 (3) 1048576 (4) 9475854336 (5)96059601

CHAPTER 6
Exercise 6(a):
G1) LA (2) 56.556 (3) 33.444 4yT019111 (5) '21:667 (6) 89.444
(7) 900.556 (8) 33.111 (9) 80.000 (10) 997.222 (11) 999.889 (12) 889.000
(13) 2120 Cray ties 23 (1a) 1097 222 (16) 100.099. (17) 1009-990 (18) 91079.111
(19) 1255.266 (20) 1050.640 (21) 37206.980 (22) 3464.485 (23) 4.586 (24) 721.162
2 SPEED ARITHMETIC

Exercise 6 (b):
(1) 1306 (2) 10407.375 (3) 13422 (4) 509.640 (5) 100140.917
(6) 9794.294 = (7) 7556.533 (8) 421.346 (9) 9223 (10) 9246.322

Exercise 6(c):
(1)1504.571 (2) 275.338 (3) 834.357 ~ (4) 915.917 (5) 1363.882 (6)83.755
(7) 88.096 (8)89.997 (9)32.043 (10)1404.923 (11) 1678.321 (12) 1344.470
(13)274.257 (14) 2249 (15) 83.056
Exercise 6(d):
(1)136.070 (2)1126.173 (3)3 198.222 (4) 4709.845 (5) 245.312 (6) 436.490
(7)110.571 (8) 95.894 (9) 1071.068 (10) 972.286 (1jak3t.169-. (12)20:271
(13) 56.579 =(14)352.710 (15) 116.296 (16) 2290.302
Exercise 6(e):
(1) 21.266 (2) 618.793 (3) 175.203 (4) 417.379 (5) 75.112 (6) 185.317

CHAPTER 7
Exercise 7:
(1) YES (2) NO (3) NO (4) NO (5) YES (6) YES
(7) NO (8) YES (9) NO (10) NO (11) NO (12) NO

CHAPTER 9
Exercise 9:
(1) 0.07 (recurring) (2) 0.053017672 (3) 1.631666277 (4) 0.098901099 (5) 2.371048253
(6) 0.944389275 = (7) 0.876533934 (8) 1.808823529 (9) 1.814121037 (10) 0.615537849
CHAPTER 10
Exercise 10: .
(1) 64 (2) 82.940 (3) 154.256 (4) 453.454 (5) 288.560 (6) 999
(7) 109 (8) 4096 (9) 0.180 (10) 0.042 (11) 0.259 (12) 85.279
(13) 14.006 (14) 28.634 (15) 27.98 Clo) 19 127
CHAPTER 11
Exercise 11(a):
(1) 789 (2) 901 (3) 896 (4) 194 (5) 817
(6) 6.5 (7) 84.9 (8) 9.3 (9) 0.94 (10) 59.1
Exercise 11 (b):
(1) 33 17133 (2) 399.7735 (3) 268.162 (4) 353.661 (5) 177.9263 «
(6) 40.783 (7) 145.248 (8) 56.82002 (9) 1.2967 (10) 37.7695
 o is
APPENDIX - 1

Here, we provide the algebraic basis on which tables land 2 on excesses in

Chapter 11 have been prepared.

Let us first consider the case where the cube root has 3-digits i.e. n= 3. Let, as we
have stated in the text, F denote the first of these 3 digits, P the penultimate (or the
middle digit in this case) and L the last digit. The number that forms these digits can then
be written as 100 F + 10P + L and its cube is (100F + 10P + LY’ = (10°F + 10F + L)’.
Expanding it by the binomial theorem, we have
(10°F+10P+L) = 10°F? + 10°P? + L?+ 3x10°F°P + 3x10°FP? + 3x10°F7L +
3x10° L°F + 3x10° LP* + 3x10L7P + 6x10°LPF
From this we see that the terms occupying different places in the expansion are
(omitting the powers of 10) as follows:-
The units place i3
The tens place 3L’P
The 10° ‘s place 3LP* + 3L°F
The 10°’s place P? + 6LPF
The 10*’s place 3FP* + 3F°L
The 10°’s place 3F°P
The 10°’s place F

Next, let us consider the case when the number of digits in the cube root is 4 1.e.
when n= 4. Let us designate the first digit by F, the second by S, the third or the
penultimate by P and the last by L (same as in the text). We then have to expand ( 10°F
+ 10°S + 10P +L)’. Expanding as before and collecting the terms in various places, we
get
The units place be
The tens place 3L°P
The 10° ‘s place 3LP* + 3L’S
The 10°’s place 3L°F + P? + 6LPS
The 10*’s place 3LS? + 3P°S + 6LPF
The 10°’s place 3FP + 3S°P + 6FSL
The 10°’s place 3F°L + S° + 6FSP
The 10’’s place 3F°P + 3FS?
The 10°’s place 3F°S
The 10”’s Place F

A similar method may be followed for n=5 with F representing the first digit,
S the second, T the third , P the penultimate and L the last.
174

APPENDIX 2

In his pioneering work “Vedic Mathematics”, (originally published in

1965), the former Shankaracharya Jagadguru Swami Sri. Bharati Krsna
Thirthaji Maharaja (1884-1960) of Govardan Mutt, Puri (Orissa) had used 16
Sutras and Sub Sutras (Corollaries) which are listed in the book. These Sutras are
in Sanskrit, for some of which English translations are not given but have to be
inferred from the context in which they have been used.
Sutra literally means “thread” but Thirthaji Maharaja employs the word
“aphorism” for the same since it is indeed, short, terse and condensed and yet it is
a forcible maxim. It can be likened to a “word-formula” (unlike a mathematical
formula with which we are accustomed). It does not show how a calculation is to
be made but only throws up a pointer or direction in which the calculation can
proceed. For this reason, the same Sutra can be employed for a variety of
applications.
In this Appendix, we present, out of 16 Sutras, those which we have
employed in our book. We first give the English version of the Sutra, followed by
the translation to the extent available. The area of its application is given by the
side of the Sutra or its Corollary.
Sutra(Word-formula)/Sub-Sutra(Corollary) Used in
1. Ekadhikena Pirvena | Multiplication
“By one more than the previous one” Divisibility
‘Recurring decimals
Auxiliary Fractions
Anuripyena (Corollary) Multiplication
“Proportionately” Division
Cubing
Auxiliary Fractions
2. Nikhilam Navatascaranam Dasatah Multiplication
“All from nine and the last from ten” Division
’ ,

Sisyate Segasamjnah (corollary) Multiplication

“Reminder remains constant”

3. Urdhava-tiryagbhyam Multiplication
“Vertically and Cross-Wise”’ Division
4. Paravartya Yojayet Division
“Transpose and divide”
 4
175

5. Vestanam(Corollary) Divisibility
"Osculation"

6. |Yavadunam Tavadinam(Corollary) Multiplication

"Whatever the extent of its deficiency, lessen it
still further to that very extent"

7. Antyayordasake api(Corollary) Multiplication

"(i) Squaring of anumber ending in 5;
(ii) Multiplication of two numbers whose last
digits add to 10 and whose previous part is
exactly the same"

8. Yavadunam Squaring
"Deficiency" Cubing

9. Sesanyankena Caramena Recurring decimals

"The remainder by the last digit"

Vilokanam(Corollary) Division
"By inspection or observation.",

10. Ekanydnena pirvena Multiplication

"Multiplication wherever the multiplied - digit
consist entirely of 9's"

11. Dhvajanka Division

"On top of the flag”

Meanings of Symbols for Pronounciation

A >] a asin far

U 9 u asinrule
S, s asin shirt
c as inchart
nah asin aha
 we —S ae ce

oa be ; es a APYEDOOLA 2

; ieee ay Wee Veilic Mi ih

ihe iat. Fee. cttw ya Bon
x = ai: - Si gcmmage | LR Pa) )* : . a ou
<u capastabs rae pom tarwest7 Whitteare fistsheers ah
— fe Saar’: ae eerie of watch’ Bo ich eranvles
ie pot 1 ~ feeanscaicy ste t it: eh Shey ao aa
. wa FugeLiterally maths “ihresd” Gat ae
2 | i Sagi bae aeeae tne ela SIP oats ‘i
(Paes S =uxinte HON My. 28 ck“a be TeReaiteMz RO!oe
Z a= <irp a tt StireeS GIS BCCHS ood Hire ‘
ei deta > Guess we aerdantcr Gr tied iat ihe.see
c
- -) repo - Pac geminata soneSl sahe “ano
gaviexticon. pik’)
7 =< —p

weseri OU; a $6 - Sulras, °thas

the Engliske yersinrees
4)

rst give
- The spes = 8

—7 irs ~ 7 : ‘ :

- » ai - “4

ire Taf Pinesi 3 es

ea an a : ae j 5 ¥ TS?2

a ie hl =i — ’ ir fh ri * : ae : "

=z si : oo Ms
Fn o
te ee ngibig - bedgitlunt
: aft} ts TV
srt
ae ot
ae a

mt: Peday

be 2 oT..veu Ao sGeranhes (igen

nt 37 ibe bast from:

iat) {Cory iaky

oi ; : y aS OU rst

hae
 The author has studied Mathematics and Statistics in Bombay and nad taken
his M.Sc. degree from Bombay University. Later, he obtained his Ph.D. degree
from London University. He had held senior teaching, research and administrative
- positions in various academic institutions of repute both in India as well as in
_ U.S. A., Australia and the Republic of the Phillippines. He has to his credit three
books and about 65 research publications in Indian and international journals.
The author has been teaching vedic mathematics for the past few years and his
interest in the present publication arose out of the students’ response and
enthusiasm,

BOOK REVIEWS:
Times of India (Education Supplement), dated 15.5: 2000.
onn= The book covers a wide variety of common as well as specialised problems
pertaining to multiplication, division, percentages, ratio--proportion, square roots,
and cube roots among others. The manner in which these new methods are
taught is what makes the book most striking. Solved examples as well as step-
wise explanations as to how to mptement the new methods is given in simple
and easy-to-understand language. oe
“.--- This book due to its vast coverage of topics and simplicity in approach
should be of great use to students in schools and colleges alike. ----”

Science Express (Supplement of The New Indian Express), dated 18.1.2000

“An exciting book for mathematics students and those interested in numbers. “
“The author demonstrates all these skills through ample examples in each
__ chapter. Starting with marvels and mysteries with numbers, the pages are full
of numbers EVare| calculations that are interesting, eS Oe and astoundingly
simple. cece
“THe book is well structured and easy to ioflee: A must for those who love
numbers. It is an oe opener of our ancient-wisdom.”:

Felix Bileski, Canada (User of the Book), dated 15. 2.2000.

“| received your new vedic arithmetic book oer and ayneery, pleased that it
» is a lot (o) value. for the price. ----”
2 “Itis valuable knowledge that should be Propagated as Meas as ponentsa