Japanese Abacus

The japanese abacus, its use and theory, by Takashi Kojima
The
JAPANESE ABACUS
Its Use and Theory
BY
TAKASHI KOJIMA
CHARLES E. TUTTLE COMPANY

TOKYO - JAPAN
First edition, April 1954
1
FOREWORD
It gives me great pleasure that this book in English on the abacus is ready for publication.
Japanese abacus operators have long cherished the desire, here finally realised, of
introducing the Japanese abacus to other countries in view of the remarkable advances
and developments which have been made in the instrument and its use during the past
quarter of a century.
The Japanese abacus, simple and primitive looking though it be, can be operated with
greater speed and efficiency than even the electric calculating machine -a fact proven in
numerous tests and well documented by Mr. Kojima in his first chapter. This is particulary
so in addition and subtraction, where the abacus can handle figures of any number of digits
twice as fast as the electric machine. To explain the instrument’s incredible speed and
mystifying efficiency it is essential not only to introduce the newest improved methods of
operation but also to elucidate the most advanced theories of rational bases and of bead
manipulation.
In writing this practical book Mr. Kojima has kept these two requirements well in mind.
The Abacus Research Institute of the Japan Chamber of Commerce and Industry has been
most pleased to cooperate with him by making available its research data and correcting
his manuscript in the light of all the latest information.
YOEMON YAMAZAKI
Vice-Chairman, Abacus Research Institute
Professor of Economics, Nippon University
Vice-President, All-Japan Federation of Abacus Operators
AUTHOR’S FOREWORD
This book has been written as a guide for those who, though interested in knowing more
about the use and theory of the Japanese abacus, have until now been unable to find any
full explanation in the English language. Chapter I presents the most important facts about
the speed and efficiency of abacus calculation, with special reference to a comparison of
the abacus and the electric calculating machine. Chapter II gives a brief survey of the
history and development of the abacus, and Chapter III introduces the basic principles of
abacus calculation.
The next three chapters explain in detail, with numerous examples, how the four
processes of arithmetic are worked out on the abacus. Particular attention should be given
to Chapter IV, on addition and subtraction, as it embodies the essential rules of bead
manipulation. Many notes have been included to give a theoretical and scientific
explanation of the rules and fundamental principles as such knowledge is not only of
interest but will prove of great aid in the actual operation of the instrument. The book
concludes with short chapters on decimals and mental calculation, and a selection of
exercises.
Among many who kindly gave me information and suggestions, I am particularly grateful
to Mr. Yoemon Yamazaki, who kindly wrote the foreword and provided me with many
2
valuable suggestions and a large part of the information in Chapter I. He is the

Vice-Chairman of the Abacus Research Institute and Advisor to the Central Committee of
the Federation of Abacus Workers (hereafter referred to simply as the Abacus Committee),
both organizations being under the sponsorship of the Japan Chamber of Commerce and
Industry.
I also wish to express my special gratitude to Mr. Shinji Ishikawa, President of the Japan
Association of Abacus Calculation, who spared himself no trouble in reading the whole of
the manuscript and furnishing much important up-to-date information.
I also extend my grateful acknowledgements to Mr. Zenji Arai, Chairman of the Abacus
Research Committee of the Japan Federation of Abacus Education, and Mr. Miyokichi Ban,
of the above-mentioned Abacus Committee. They kindly read the whole of the manuscript
and provided me with many necessary and valuable suggestions.
My grateful acknowledgements are also due to Mr. Takeo Uno on the Abacus Committee
and Mr. Tadao Yamamoto, who conducts his own abacus school. They kindly read the
manuscript in parts and gave me valuable suggestions.
I also wish to thank Mr. Kiyoshi Matsuzaki, of the Savings Bureau of the Ministry of Postal
Administration, who kindly furnished the table on page 5.
Finally I must express my sincere thanks for many invaluable suggestions on English style
from Mr. C. G. Wells, Chief Writer for the Far East Network; Mr. Harold Gosling, of the
British Commonwealth Public Relations; Mr. Richard D. Lane, formerly of the Far East
Network; and above all from Mr. Meredith Weatherby, of the Charles E. Tuttle Company,
without whose painstaking efforts this book could not have become what it is now.
T. K.
CONTENTS
Foreword by Yoemon Yamazaki…………………..2

Author’s Foreword………………………………………..2
I. Abacus versus Electric Calculator…………….… 4
II. Brief History of the Abacus…………………………11
III. Basic Principles of Calculation…………………… 13
IV. Addition and Subtraction…………………………… 16
V. Multiplication………………………………………….…. 30
VI. Division……………………………………………………….. 36
VII. Decimals……………………………………………………… 44
VIII. Mental Calculation…………………………………….. 48
Exercises…………………………………………………….. 49
3
I - ABACUS VERSUS ELECTRIC CALCULATOR

The abacus, or soroban as it is called in Japan, is one of the first objects that strongly
attract the attention of the foreigner in Japan. When he buys a few trifling articles at some
store, he soon notices that the tradesman does not perplex himself with mental arithmetic,
but instead seizes his soroban, prepares it by a tilt and a rattling sweep of his hand, and
after a deft manipulation of rapid clicks, reads off the price. It is true that the Japanese
tradesman often uses his board and beads even when the problem is simple enough to be
done in one’s head, but this is only because the use of the abacus has become a habit with
him. If he tried, he could no doubt easily add 37 and 48 in his head. But such is the force of
habit that he does not try to recognize the simplicity of any problem; instead, following
the line of least resistance, he adjusts his soroban for manipulation, and begins clicking
the beads, thus escaping any need of mental effort.
Doubtlessly the Westerner, with his belief in the powers of mental arithmetic and the
modern calculating machine, often mistrusts the efficiency of such a primitive looking
instrument. However, his mistrust of the soroban is likely to be transformed into
admiration when he gains some knowledge concerning it. For the soroban, which can
perform in a fraction of time a difficult arithmetic calculation that the Westerner could do
laboriously only by means of pencil and paper, possesses distinct advantages over mental
and written arithmetic. in a competition in arithmetic problems, an ordinary Japanese
tradesman with his soroban would easily outstrip a rapid and accurate Western accountant
even with his adding machine.
An exciting contest between the Japanese abacus and the electric calculating machine
was held in Tokyo on November 12, 1946, under the sponsorship of the U. S. Army
newspaper, the Stars and Stripes. In reporting the contest, the Stars and Stripes
remarked: “The machine age took a step backward yesterday at the Ernie Pyle Theater as
the abacus, centuries old, dealt defeat to the most up-to-date electric machine now being
used by the United States Government. The abacus victory was decisive.”
The Nippon Times reported the contest as follows: “Civilization, on the threshold of the
atomic age, tottered Monday afternoon as the 2,000-year-old abacus beat the electric
calculating machine in adding, subtracting, dividing and a problem including all three with
multiplication thrown in, according to UP. Only in multiplication alone did the machine
triumph...”
The American representative of the calculating machine was Pvt. Thomas Nathan Wood
of the 240th Finance Disbursing Section of General MacArthur’s headquarters, who had
been selected in an arithmetic contest as the most expert operator of the electric
calculator in Japan. The Japanese representative was Mr. Kiyoshi Matsuzaki, a champion
operator of the abacus in the Savings Bureau of the Ministry of Postal Administration.
As may be seen from the results tabulated on the following page, the abacus scored a
total of 4 points as against 1 point for the electric calculator. Such results should convince
even the most skeptical that, at least so far as addition and subtraction are concerned, the
abacus possesses an indisputable advantage over the calculating machine. Its advantages
in the fields of multiplication and division, however, were not so decisively demonstrated:
4
RESULTS OF CONTEST
MATSUZAKI (Abacus) vs. WOOD (Electric Calculator).
Type of Problem Name 1st Heat 2nd Heat 3rd Heat Score
Addition: 50 number Matsuzaki 1m 14.8s 1m 16s -

each consisting of Victor Victor 1
from 3 to 6 digits
Wood 2m 0.2s 1m 53s -
Defeated Defeated 0
Subtraction: 1m 0.4s 1m 0.8s 1m 0s
5 problems, with Matsuzaki All correct. 4 correct. All correct. 1
minuends and Victor No decision Victor
subtrahends of from
6 to 8 digits each 1m 30s 1m 36s 1m 22s
Wood All correct. 4 correct. 4 correct. 0
Defeated No decision Defeated
Multiplication: 1m 44.6s 1m 19s 2m 14.4s

problems, each Matsuzaki 4 correct. All correct. 3 correct. 0
containing 5to 12 Defeated Victor Defeated
digits in multiplier
and multiplicand 2m 22s 1m 20s 1m 53.6s
Wood 4 correct. All correct. 4 correct. 1
Victor Defeated Victor
Division: 1m 36.6s 1m 23.4s 1m 21s

5 problems, each Matsuzaki All correct. 4 correct. All correct. 1
containing 5 to 12 Victor Defeated Victor
digits in divisor
and dividend 1m 48s 1m 19s 1m 26.6s
Wood All correct. All correct. 4 correct. 0
Defeated Victor Defeated
Composite problem:
1 problem in addition 1m 21s
of 30 6-digit numbers; Matsuzaki All correct. - - 1
3 problems in Victor
subtraction, each with
two 6-digit numbers; 3
problems in
multiplication, each
with two figures 1m 26.6s
containing a total of 5 Wood 4 correct. - - 0
to 12 digits; 3 problems Defeated
in division, each with
two figures containing
a total of 5 to 12 digits
Total Score Matsuzaki - - - 4

Wood - - - 1
5
For reliable information on the comparative merits of the abacus and the calculating
machine, we can do nothing better then turn to the Abacus Committee of the Japan
Chamber of Commerce and Industry, which has made minute investigations concerning the
potentialities of the Japanese abacus. The Committee has acted as judge of the
semi-annual examination for abacus operators’ licenses since the examinations were in-
itiated in 1931, such licenses being divided into three classes, according to the
manipulators’ efficiency.
The Committee says: “In a contest in addition and subtraction, a first-grade abacus
operator can easily defeat the best operator of an electric machine, solving problems
twice as fast as the latter, no matter how many digits the numbers contain. If the numbers
do not contain over six digits, the abacus manipulator can halve the time of the operation
by relying in part upon mental calculation (a system peculiar to the abacus, to be
described hereafter). In multiplication and division the first-grade abacus operator can
maintain some margin of advantage over the electric calculator so long as the problem
does not contain more than a total of about ten digits in multiplicand and multiplier or in
divisor and quotient. The abacus and the electric machine are on a par in a problem which
contains a total of ten to twelve digits. With each additional digit in a problem, the
advantage of the electric calculating machine increases.”
A similar view is held by Mr. Kiyoshi Matsuzaki, who made the following remark
concerning the contest described in preceding pages: “In addition and subtraction even
the third-grade abacus worker can hold his own against the electric calculating machine.
In multiplication and division the first-grade abacus worker may have a good chance to win
over the calculating machine, provided the problem does not have more than a total of ten
digits in multiplicand and multiplier or in divisor and quotient. I felt nervous at the contest
and made more mistakes than I might have done otherwise. My opponent may have felt the
same, though. A good first-grade abacus worker ought to be able to make a better showing
when he is at ease.”
As examples of the proficiency required of the abacus operator, it will be of interest to

cite a few problems used in the examination for abacus operators’ licences.
A. ADDITION AND SUBTRACTION

No. 1 2 3 4 5
1 ¥ 6 393 082.74 ¥40 693 718.52 ¥ 160 384.72 ¥ 730.49 ¥ 352 719.48
2 269.31 52 687.09 83 479 051.26 6 089 547.31 84 936.20
3 541 793.60 7 180 592.43 -21 479.50 463 195.28 92 460 385.71
4 82 706 314.95 1 715.38 9 058 627.13 97 820.56 -718 024.36
5 72 940.18 63 847 529.10 -3 780.29 3 985 271.04 45 178.62
6 3 014 725.86 26 073.94 27 915.64 10 476 825.93 8 327 605.94
7 98 156 .02 309 861.75 40 715 368.92 54 613.78 -19 062.53
8 15 726 408.39 8 714 905.26 86 203.41 218 769.45 -4 085 237.61
9 970 285.13 346.17 -504 189.76 3 428.01 25 963 180.47
10 45 963.78 295 130.86 -6 037 512.89 82 605 917.34 70 941.28
11 6 831 750.24 94 038 726.51 924.35 61 853.20 -6 798.05
12 64 371.59 69 052.74 763 815.04 250 376.19 -50 824 361.79
13 249 168.07 150 938.42 -20 849 136.57 3 576 904.82 953.16
14 70 593 826.41 43 281.65 4 102 653.98 49 021.67 3 107 425.89
15 4 352.80 7 916 403.28 95 467.83 57 316 482.90 639 507.14
計
6
B. MULTIPLICATION C. DIVISION
No. No.
1 759.843 x 57.941 = 1 4 768 788 098 / 14 593 =
2 302.162 x 83.602 = 2 971 837 849 / 51.682 =
3 967.408 x 70.589 = 3 47 408 509 168 / 49.201 =
4 20.359 x 628.134 = 5 0.3481095257 / 0.06457 =
5 84.2697 x 9.4076 = 5 66 014 150 202 / 92 378 =
6 135.941 x 46.295 = 6 3 657.6146092 / 80 914 =
7 0.4271805 x 0.2513 = 7 166.4719833 / 0.6702 =
8 669.378 x 0.31908 = 8 0. 4537275087 / 7.38609 =
9 0.914053 x 68.087 = 9 328 399.09042 / 35.746 =
10 587.216 x 17.452 = 10 24 484 596 290 / 28.135 =
D. MENTAL CALCULATION
No. 1 2 3 4 5
1 ¥ 74.68 ¥ 3.46 ¥ 52.31 ¥ 8.09 ¥ 90.47
2 2.98 97.98 30.64 2.41 3.51
3 50.41 6.05 —9.28 56.37 —76.29
4 83.72 2.13 14.75 1.52 —1.83
5 1.35 50.79 8.39 70.86 54.02
6 6.84 8.21 —45.05 3.94 8.35
7 95.01 19.64 6.17 96.70 29.14
8 3.27 78.30 —2.83 6.28 —6.50
9 65.10 4.56 7.14 37.19 47.26
10 4.92 82.07 91.26 48.05 1.83
計
To receive a first-grade license an applicant must be able to work problems similar to

the foregoing with 80 per cent accuracy within a time limit of five minutes each for the
first three groups—A, B and C—and one minute for the fourth.
The problem in mental calculation requires a few words of explanation as the method of
solving it depends directly upon a knowledge of the use of the abacus, and being an
integral part of abacus technique, it is entirely different from any Western method of
calculation. This abacus method of mental arithmetic is described in some detail in
Chapter VIII. Suffice is to say here that the method consists in mentally visualizing an
abacus and working the problem out by standard methods on the imaginary instrument.
The process is easier than it sounds and accounts for the incredible and almost mystifying
peaks of efficiency attained by masters of abacus operation.
To cite but one example of proficiency in this type of mental arithmetic, on May 28,
1952, during the Sixth All-Japan Abacus Contest, held in Tokyo, a master abacus operator,
Mr. Yoshio Kojima, gave a demonstration of his skill in mental arithmetic.
In one minute and 18.4 seconds he gave correct answer to 50 division problems, each of
which contained five to seven digits in its dividend and divisor. Next, in a twinkling of 13.6
seconds he added 10 numbers of ten digits each. Thus he set two remarkable records -and
all with no aid other than the mentally visualized abacus! This means that he could have
added mentally the fifteen numbers given in one of the columns of Problem A, page 6, in
7
one-fourth of the given time limit of one minute, and in one-eighth of the time required by
the best operator of an electric calculating machine.
A. ADDITION AND SUBTRACTION

No. 1 2 3 4 5
1 ¥ 71 896 ¥ 93 502 ¥ 130 745 ¥ 60 374 ¥ 9 180
2 306 425 8 164 59 280 875 126 25 634
3 839 802 635 4 968 23 601 418 275
4 50 178 378 102 7 284 54 361
5 2 941 25 910 701 539 932 7 903
6 567 308 -8 756 48 075 -506 849 86 215
7 762 650 481 6 714 -39 256 903 587
8 82 037 362 429 1 023 643
9 3 694 71 049 602 893 485 71 852
10 470 589 2 913 27 564 187 683 408
11 24 310 134 795 4 931 -96 178 640 729
12 165 -19 247 93 270 -4 517 210
13 5 742 -804 315 687 -760 92 674
14 904 213 -65 798 90 312 248 951 5 096
15 68 951 746 083 856 19 074 837 921
計
B. MULTIPLICATION C. DIVISION
No. No.
1 6 742 x 358 = 1 435 633 / 921 =
2 2 681 x 609 = 2 315 56 / 805 =
3 5 093 x 176 = 3 18.998 / 236 =
4 0.825 x 94.12 = 4 63.162 / 0.087 =
5 3 310 x 803 = 5 223.792 / 394 =
6 9 478 x 0.645 = 6 400.026 / 418 =
7 76 506 x 5.2 = 7 180 096 / 64 =
8 193.4 x 4.18 = 8 0.105118 / 0.753 =
9 0.4052 x 0.267 = 9 104 249 / 1.709 =
10 9 718 x 703 = 10 0.21918 / 5.62 =
To receive a third-grade license an applicant must be able to work problems similar to

those on page 8 with 70 per cent accuracy within a time limit of five minutes for each
group.
The primary advantage of the abacus is its incredible speed resulting from the
mechanization or simplification of calculation, by means of which the answer to a given
problem forms itself naturally or mechanically on the board, thus reducing mental labor to
a minimum. The theoretical explanation of this mechanization of calculation is given in
Chapter IV (see Note 3 to Example 9, Note 3 to Example 10, and Notes 3, 4 and 5 to
Example 20).
Another big advantage of the abacus is its extremely moderate price, ranging generally
between 25¢ and $2.50 or $3.50 to quote prices in U.S. dollar equivalents. How many times
more does the ordinary calculating machine cost, to say nothing of the gleaming electric
machines which abound in Western business houses?
8
Among many other merits of the abacus one should not overlook its handy construction,
its portability, and the ease of its operational methods, which are nothing more than
simplifications of the four processes of arithmetic.
The most peculiar advantage of the abacus is that a problem in addition and subtraction
is worked out from left to right instead of from right to left as is the case with written
arithmetic, and thus harmonizes perfectly with the normal way of reading and writing
numbers. In this way a number can be added or subtracted while it is being given. For
example, if the first number in the problem is 753, the operator can enter 7 on the abacus
the instant he hears or sees “seven hundred,” and then proceed on next to the 5 and finally
the 3, whereas in written calculation he would usually have to wait until all figures were
given and then start calculating backward from the 3 of 753.
The one admitted disadvantage of the abacus is that the instrument produces only a
final result without preserving any record of intermediate steps. If any error is made, the
whole calculation must be carried through again from beginning to end. But this seeming
disadvantage is more than offset by the rapidity and accuracy which the abacus makes
possible. And it is the result that counts.
The chief factor which discredits the abacus in Western eyes is the length of time and
practice required to become a skilled operator. Certainly the abacus requires much more
practice than the calculating machine. But this apparent disadvantage is not so great an
obstacle as it is generally thought to be. Some experience and practice with this simple but
highly scientific instrument will convince the reader that this Western idea is largely a
prejudice. A few weeks of practice for an hour each day with proper procedures will give
anyone sufficient skill to turn to the abacus instead of pencil and paper for arithmetical
computation.
According to the Abacus Committee, average students, who begin their practice while
in their teens, should be able to pass the examination for third-grade licenses after half a
year of daily practice of one hour, and bright students or students with a mathematical
bent after only three months. Generally speaking, another half year of practice will enable
a third-grade abacus operator to obtain a second-grade license; and one more full year
should make him a first-grade operator. As is generally the case with any other art or
accomplishment, it is best to start practicing under right guidance when young. Those who
take up their study of the abacus after they are out of their teens are never able to pass
the first-grade examination, but it is definitely possible for them to attain to the third rank,
and occasionally even to the second.
In recent years the abacus has enjoyed an amazing increase in popularity in Japan. For
example, in the last seven years since the war’s end, the successful examinees for the
first-grade license have totaled approximately 5 700, those for the secondgrade 28 000,
and those for the third-grade 217 000, making a total of 250 700. This figure for the last
seven years is almost five times that of seventeen years immediately preceding the war.
The abacus has even found its way into the curriculum of all grade schools as one of the
elements of arithmetic, and there are now numerous abacus schools to meet the needs of
those preparing to go into business. In short, the abacus has become such a popular
favorite that it is to be found in practically every household.
How are we to account for the sudden spurt in the popularity of the old-fashioned
9
abacus, here in the middle of the mechanized twentieth century? Undoubtedly the
principal explanation lies in the fact that its operational methods have recently been
markedly simplified and improved. As will be explained in Chapter VI, the old method of
division required the memorization of a difficult division table, and was the chief factor
which alienated the average, non-commercial Japanese from the abacus. Once this
difficulty was overcome by the introduction of the newer method of division -so much
simpler and, in a sense, so much more accurate that it marked a milestone in the im-
provement of abacus technique- the instrument rapidly attained the universal popularity
which it now enjoys.
But how account for the almost exclusive use of the abacus in offices and firms which
could well afford electric calculating machines? Let statistics give the answer. According
to figures compiled by the Abacus Committee, in the conduct of an average business the
four types of arithmetical calculation occur in about the following proportions: addition
seventy percent, subtraction five percent, multiplication twenty percent, and division five
percent. As previously mentioned, the abacus can add and subtract faster than the electric
calculating machine. As for problems in multiplication and division, those which contain
more than a total of ten digits in their multiplicand and multiplier or in their divisor and
quotient are exceptional. This means that a good operator can work out most
multiplication and division problems as fast as or even faster on an abacus than on an
electric calculation machine, to say nothing of the much slower non-electric machine.
Hence, it is not surprising that the abacus is used almost exclusively in all Japanese
commercial establishments, from the tiny store to the giant corporation. To give exact
figures, in Japanese business ninety-two percent of all calculation is done on the abacus,
five percent on calculating machines (mostly of the manually operated variety), and the
remaining three percent by means of calculating tables, slide rules or written and mental
arithmetic.
At the present time various experiments are being undertaken to improve still further
the operational technique of the abacus. But this little handbook will introduce only the
best of the established methods and verified theories, essential for learning to operate the
abacus with good understanding and rapidity. Once the basic rules have been mastered,
the secret of acquiring skill in abacus operation lies in constant daily practice.
Problems involving the extraction of roots can also be solved on the abacus with great
rapidity. But the extraction of roots, which is rarely used in everyday and business
calculation, is outside the scope of this book.
In concluding our comparison of the abacus and the calculating machine, we shall not go
so far as to make the rash assertion that the abacus is worthy of immediate adoption by
Western countries, where calculating machines are readily available. But we do feel
justified in saying that the soroban is at least worthy of study and consideration by
Westerners and that by introducing to the West this unique and practical example of
Eastern science we will be repaying a modicum of our heavy debt of gratitude for the great
amount of Western civilization which we have adopted here in Japan.
10
II – BRIEF HISTORY OF THE ABACUS
The imperfect numerical notation and the scarcity of suitable writing materials in
ancient times are presumed to have given rise to the need for devices of mechanical
calculation. While the definite origin of the abacus is obscure, there is some reason for
believing that its earliest form was a reckoning table covered with sand or fine dust, in
which figures were drawn with a stylus, to be erased with the finger when necessary. The
English word abacus is etymologically derived from the Greek abax, meaning a reckoning
table covered with dust, which in turn comes from a
Semitic word meaning dust or a reckoning table covered
with dust or sand. In time this sanddust abacus gave place
to a ruled table upon which counters or disks were
arranged on lines to indicate numbers. Various forms of this
line abacus were in common use in Europe until the
opening of the seventeenth century. In rather remote
times, a third form of abacus appeared in certain parts of
the world. Instead of lines on which loose counters were
laid, the table had movable counters sliding up and down
grooves.
All three types of abacuses were found at some time or other in ancient Rome —the dust
abacus, the line abacus, and the grooved abacus. Out of this last type yet a fourth form of
the abacus was developed—one with beads sliding on rods fixed in a frame. This form, the
bead or rod abacus, with which calculations can be made much more quickly than on paper,
is still used in China, Japan, and other parts of the world. In Europe, after the introduction
of Arabic numerals, instrumental arithmetic ceased to make much progress and finally
gave way altogether to the graphical as the supply of writing materials became gradually
abundant.
As for the Orient, a form of the counting-rod abacus, called ch’eou in China and sangi in
Japan, had been used since ancient times as a means of calculation. The Chinese abacus
itself seems, according to the best evidence, to have originated in Central or Western Asia.
There is a sixth-century Chinese reference to an abacus on which counters were rolled in
grooves. The description of this ancient Chinese abacus and the known intercourse
between East and West give us good reason to believe that the Chinese abacus was
suggested by the Roman. The Chinese write in vertical columns from above downwards. If
they ever are compelled to write in a horizontal line, they write from right to left. But the
abacus is worked from left to right. This is another indication that the abacus was not
indigenous to China. The present Chinese bead abacus, which is generally called suan-pan
(arithmetic board) in Mandarin and soo-pan in the southern dialect, was a later
development, probably appearing in the twelfth century, and did not come into common
use till the fourteenth century. It is only natural that the people of the Orient, having
retained a system of numerical notation unsuited for calculation, should have developed
the abacus to a high degree, and its continuous universal use even after the introduction of
Arabic numerals is eloquent testimony to the great efficiency achieved in its development.
The Japanese word for abacus, soroban, is probably the Japanese rendering of the
Chinese suan-pan. Although the soroban did not come into popular use in Japan until the
seventeenth century, there is no doubt that it must have been known to Japanese
merchants at least a couple of centuries earlier. In any case, once this convenient
instrument of calculation became widely known in Japan, it was studied extensively and
11
intensively by many mathematicians, including Seki Kowa (1640 — 1708), who discovered a
native calculus independent of the Newtonian theory. As a result of all this study, the form
and operational methods of the abacus have undergone one improvement after another.
Like the present-day Chinese suan-pan, the soroban long had two beads above the beam
and five below. But toward the close of the nineteenth century it was simplified by
reducing the two beads above the beam to one, and finally around 1920 it acquired its
present shape by omitting yet another bead, reducing those below the beam from five to
four.
Thus the present form of the soroban is a crystallization of labor and ingenuity in the
field of Oriental mathematics and science. We feel sure that the soroban, enjoying
widespread use in this mechanical age on account of its distinct advantages over the
lightning calculating machine, will continue to be used in the coming atomic age as well.
12
III – BASIC PRINCIPLES OF CALCULATION

The abacus is a simple instrument for performing rapid arithmetical calculation. It
consists of an oblong wooden frame or board holding a number of vertically arranged rods,
on which wooden beads, balls, or counters slide up and down. A beam running across the
board divides the rods into two sections: upper and lower. The most common type of
abacus in Japan has twenty-one bamboo rods, and is about twelve inches long by two
inches wide. But larger types with twenty-seven or thirty-one rods, and smaller ones with
seventeen or thirteen rods, are also used. As described
at the conclusion of the preceding chapter, and as may
be seen in the accompanying illustration, the number of
beads per rod has been progressively reduced, in the
interests of simplicity and ease of operation, from seven
to six, and finally to five. Until recently an abacus with
five beads in the lower section of each rod was in
general use. But this type of abacus has now been largely
replaced by a one with four beads on each rod below the
beam.
The abacus is based on the decimal system. For convenience in calculation the beam is
marked with a unit point at every third rod. These unit points serve to indicate the decimal
point and other units of decimal measure. For example,
select any rod near the center of the board which is marked
with a unit point, and call this the unit rod of the problem.
Then the first rod to its left is the tens' rod, the second is
the hundreds’ rod, the third rod (marked with another unit
point) is the thousands’ rod, etc. On the other hand, the
first rod to the right of the unit rod is the tenths’, the
second is the hundredths’, the third (likewise marked with
another point) is the thousandths’ rod, etc.
Each of the four beads on the lower section of a rod has the value of 1, while the bead
on the upper section of a rod has the value of 5. Each of the 1-unit beads below the beam
obtains its value when it is moved up toward the beam, and loses its value when it is moved
back down to its former position. On the other hand, each of the 5-unit beads above the
beam obtains its value when it is moved down to the beam and loses its value when it is
moved back up.
The beads in Figure 1, using the third unit point from the right to designate the unit rod,
represent the number 1 345, while Figure 2 shows 46 709.
Before using the abacus, make sure that all the beads are in the neutral position
representing zero. This is done by moving up all 5-unit beads and moving down all 1-unit
beads. In clearing the abacus for use, hold the left end with your left middle finger on its
upper edge and your left thumb on its lower edge, and move all beads down by slanting the
upper edge toward your body. After leveling the abacus again, raise all 5-unit beads by
moving the right index finger from left to right along the upper edge of the beam.
When calculating on the abacus, use two fingers: the right index finger and thumb.
Some operators use only the index finger, but experiments show that it is more efficient to
use the thumb as well. Nearly all experts use two fingers. Use the index linger to move
13
5-units beads up and down and to move 1-unit beads down, while using the thumb only to
move 1-unit beads up. For instance, to place the figure 7 on the abacus with only the index
finger requires two successive motions— first move down a 5-unit bead, and then move up
two 1-unit beads—whereas these motions can be performed simultaneously with two
fingers, with a corresponding increase in efficiency.
Moreover, in our everyday actions we commonly employ two or more fingers, say in
picking up something or in holding a pen, and the hand is so made that the index finger
almost always requires the assistance of the thumb. This accounts for the proven fact that,
in the long run, it is much less tiring to operate the abacus with two fingers than with but
one.
Experiments also show that the index linger can move beads down more quickly and
accurately than the thumb, while on the other hand the thumb can move beads up with
greater speed, force, and accuracy than the index finger.
The best and quickest way to acquire skill in abacus manipulation is to use the index
finger and thumb in strict accord with the prescribed rules for bead manipulation. The
correct finger movements will be indicated in detail for a number of problems in the next
chapter. They should be carefully heeded and practiced many times until you can flick
your two fingers as nimbly and effortlessly as the fingers of a pianist glide over the keys in
executing a sonata.
Another important secret for acquiring rapid skill in abacus calculation is always to keep
your fingers close to the beads. Never raise your fingers high from the beads nor put them
deep between the beads. Glide the beads up and down by touching their ridges just slightly
with the tips of your fingers.
The guiding principles for the movement of beads, as followed hereafter, may be
summarized thus:
General Rules for Moving Beads
1. Move down a 5-unit bead and move up one or more 1-unit beads as the same time.
(See Example 5, next chapter.)
2. First move down one or more 1-unit beads,
and then move up a 5-unit bead. (See
Example 6.)
3. In quick succession first move down a 5-unit
bead, and then one or more 1-unit beads.
(See Examples 7 and 9.)
4. In quick succession first move up one or
more 1-unit beads, and then a 5-unit head.
(See Examples 8 and 9.)
5. In addition, after finishing operation on the
unit rod, move up a 1-unit bead on the tens’
rod. (See Examples 11, 13, 15, 17 and 19.)
6. In subtraction, after subtracting a 1-unit
bead from the tens’ rod, operate on the unit
rod. (See Examples 12, 14, 16, 18 and 20.)
14
When working with the abacus, sit up straight at a desk. A good posture will have much
to do with the speed and accuracy of your calculations.
Finally, in studying the illustrations which accompany the examples given throughout
the rest of the book, the following key should be kept in mind:
Key to Illustrations
1. A white bead is one which is in its original position and has no numerical
value.
2. A striped bead is one which has just been moved, thereby having either
obtained or lost its numerical value.
3. A black bead is one which obtained numerical value in a previous step.
4. ↓ indicates that beads are to be moved down with the index finger.
5. ↑ indicates that beads are lo be moved up with the index finger.
6. indicates that beads are to be moved up with the thumb.
7. Figures in parentheses accompanying the foregoing signs indicate the order in which
beads are to be moved.
15
IV. ADDITION AND SUBTRACTION

There are four principal arithmetical calculations on the abacus: addition, subtraction,
multiplication, and division. Of these, addition and subtraction are basic processes, for
unless you know how to add and subtract on the abacus, you cannot multiply or divide. In
our daily life and business accounts, addition is used far more frequently than the other
processes and is most important of all.
The central part of the abacus is generally used for addition and subtraction. However,
when many large numbers are to be added, the first number is set at the right side of the
abacus, because the working extends to the left. In any case a one digit number and the
last digit of a larger number should always be set on a unit rod, that is, on a rod marked
with a unit point.
1. Adding and Subtracting One-Digit Numbers
Example 1. 1+2=3
Step 1: Set the number 1 by moving up one 1-unit bead with the thumb (Fig.
5). See that you set 1 on a unit rod marked with a unit point.
Step 2: Add 2 to 1 by moving up, on the same rod, two more 1-unit beads,
using the thumb (Fig. 6).
Note 1 : This example illustrates the procedure used in adding one or more 1-unit beads.
The problems to which this procedure applies are:
1+1 2+1 3+1 5+1 6+1 7+1 8+1

1+2 2+2 5+2 6+2 7+2
1+3 5+3 6+3
5+4
Note 2 : Hereafter, such phrases as “move up one 1-unit bead,” “move down three
1-unit beads,” etc., will be shortened to “set 1,” “move down 3,” etc.
Example 2. 3-2=1
Step 1: Set 3 with the thumb (Fig. 7).
Step 2: Subtract 2 by moving down two 1-unit beads with the index finger
(Fig. 8).
Note: This example illustrates the procedure used in subtracting one or more 1-unit
beads. The problems to which this procedure applies are:
2—1 3—1 4—1 6—1 7—1 8—1 9—1

3—2 4—2 7—2 8—2 9—2
4—3 8—3 9—3
9—4
16
Example 3. 2+5=7
Step 1: Set 2 (Fig. 9).
Step 2: Move down 5 with the index finger (Fig. 10).
Note: This example illustrates the procedure used in adding a 5-unit bead. The problems
to which this procedure applies are:
1+5 2+5 3+5 4+5
Example 4. 7-5=2
Step 2: Move up 5 with the index finger (Fig. 12).
Note: This example shows the procedure used in subtracting a 5-unit bead. The
problems to which this procedure are:
5—5 6—5 8—5 9—5
Example 5. 2+6=8
Step 2: Move down 5 with the index finger and move up 1 with the thumb
at the same time (Fig. 14).
Note : This example illustrates the procedure used in adding both a 5-unit bead and one
or more 1-unit beads. The problems to which this procedure applies are:
1+6 2+6 3+6

1+7 2+7
1+8
Example 6. 8—6=2
Step 2: After moving down 1 with the index finger, move up 5 with the
same finger (Fig. 16).
Note 1: This example shows how to subtract both one or more 1-unit beads and a 5-unit
bead. The problems to which this procedure applies are:
17
9—9 9—8 9—7 9—6

8—8 8—7 8—6
7—7 7—6
6—6
Note 2: The above-mentioned procedure is preferable to moving up 5 first and 1 next

with the index finger. As is explained in Note 2 of Example 15, in some cases the latter
procedure makes it difficult to move the fingers nimbly, e. g., the addition of 4 to 9.
Example 7. 4+1=5
Step 2: Move down 5 first and 4 next in close succession with the
index finger (Fig. 18).
Note 1: This example illustrates the procedure used in setting 5 when the addition of
two numbers makes 5. The problems to which this procedure applies are:
4+1 3+2 2+3 1+4
Note 2: Fig. 19 illustrates the incorrect way to perform Step 2. Note that the correct
way (Fig. 18) requires but a single continuing down stroke of the finger whereas the
incorrect way requires three separate movements: (1) move down 4, (2) move the finger
back up, and (3) move down 5, resulting in a loss of time and effort.
Example 8. 5 —1 = 4
Step 2: First move up 4 with the thumb, and then move up 5 with
the index finger in close succession (Fig. 21). Flick the thumb and the
index finger with the idea of performing the two motions at the same
time.
Note 1 : This example illustrates the procedure of subtracting a number from 5. The
problems to which this procedure applies are:
5—1 5—2 5—3 5—4

6—2 6—3 6—4
7—3 7—4
8—4
Note 2: As explained in Step 2 of this example, when 1 is subtracted from 5, four 1-unit
beads and one 5-unit bead should be pushed up at the same time. But if the beginner finds
it hard to make the two motions at the same time, he may perform each separately by first
pushing up four 1-unit beads, and then a 5-unit bead (Fig. 21).
Fig. 22 illustrates the incorrect way to perform Step 2. Note that the correct way
requires but a single continuous up stroke of the thumb and the index finger, whereas the
incorrect way requires three separate movements. This means that your finger or at least
18
your attention has to travel further, resulting in a loss of time and effort.
Example 9. 4 + 3 = 7
Step 2: First move down 5, and then 2 in close succession with the index
finger (Fig. 24).
Note 1 : This example illustrates the procedure used in adding a 5-unit bead and
subtracting one or more 1-unit beads. The problems to which this procedure applies:
4+1 4+2 4+3 4+4

3+2 3+3 3+4
2+3 2+4
1+4
Note 2: Since three 1-unit beads cannot be added to the four 1-unit beads, 5 is added
and 2 is subtracted to offset the excess. This operation may be represented in the form of
the equation: 4+3=4+(5—2)=7
Note 3: When working the foregoing example, do not think: Since 3 plus 4 equals 7, I
must form 7 on the board. Instead, simply remember that 2 is the complementary digit
with which 3 makes 5, and by flicking down 5 and 2 in rapid succession, allow the sum 7 to
form itself naturally on the board. There are only two groups of complementary digits for
5: 3 and 2, and 4 and 1. Operation by means of complementary digits is much simpler and
less liable to error than the ordinary mode of calculation. For further explanation of
calculation by means of complementary digits, see Note 3 to Example 20.
Example 10. 7-3=4
Step 2: In close succession, first move up 2 with the thumb, and then
move up the 5 with the index finger with the idea of performing the two
motions at the same time (Fig. 26).
Note 1: This example illustrates the procedure for adding one or more 1-unit beads and
subtracting one 5-unit bead. Problems to which applicable:
5—1 5—2 5—3 5—4

6—2 6—3 6—4
7—3 7—4
8—4
Note 2: Since three 1-unit beads cannot be subtracted from the two 1-unit beads, 2 is
added and 5 is subtracted. This operation may be represented by the equation:
7-3=7+2-5=4
Note 3: When working this example, do not think: 3 from 7 leaves 4, so 4 must be formed
on the board. Instead, simply remember that 2 is the complementary digit with which 3
makes 5, and allow the result to form itself naturally on the board. (See Note 3 to Example
19
20.)
Example 11. 3 + 7 = 10
Step 1: Set 3 on a unit rod, which we shall call B (Fig. 27).
Step 2: In close succession, first move down then 3 on B with the index
finger, and then move up 1 on the tens’ rod, here called A, with the thumb
(Fig. 28). Flick the index finger and the thumb in a twisting manner so that
you may perform the two motions at the same time.
Note 1: This example shows how to set 10 when it is the sum of two digits. This
procedure requires the subtraction of one or more 1-unit beads. Problems to which
applicable:
1+9 2+8 3+7 4+6
Note 2: In working out this example, do not move up 1 on the tens’ rod until you have
moved down 3 on the unit rod B. If you follow this incorrect procedure, you will never
improve in bead calculation. For the theoretical reasons for the advantages of the correct
procedure, see Note 5 (The Order of Operation) to Example 20.
Example 12. 10 — 7 = 3
Step 1: Set 10. This is done by simply moving up one bead on the tens’
rod A (Fig. 29).
Step 2: First remove the 10 by moving down the 1 on the tens’ rod A
with the index finger, and then move up 3 on the unit rod B with the thumb
(Fig. 30).
Note 1: This example shows how to subtract 10 and add one or more 1-unit beads.
Problems to which such procedure applies:
10—9 10—8 10—7 10—6
Note 2: In working out this example, be sure to move down 1 on the tens’ rod A before
moving up 3 on the unit rod B. For the theoretical reasons for the advantages of the correct
procedure, see Note 5 (The Order of Operation) to Example 20.
Example 13. 6 + 4 = 10
Step 1: Set 6 on the unit rod B.
Step 2: Move down 1 on B with the index finger, then move up 5 on B

with the same finger, and finally move up 1 on the tens’ rod A with the
thumb. Work the index finger and the thumb with the idea of
performing the last two motions at the same time.
Note: This example shows how to form the sum 10 when it is made by the addition of
two digits. This procedure, requiring the subtraction of both one or more 1-unit beads and
a 5-unit bead, applies to the problems:
20
6+4 7+3 8+2 9+1
Example 14. 10 — 4 = 6
Step 1: Set 1 on the tens’ rod A.
Step 2: First move down the 1 on A with the index finger, then move
down 5 and move up 1 on B at the same time.
Note: This example shows how to subtract 10 and add both a 5-unit bead and one or
more 1-unit beads. Applicable problems:
10—4 10—3 10—2 10—1
Example 15. 9 + 4 = 13
Step 2: Move down 1 on B with the index finger, then move up 5 on B

with the same finger and finally move up 1 on A with the thumb. Work the
index finger and the thumb with the idea of performing the last two
motions at the same time.
Note 1 : This example shows how to add 10 after subtracting one or more 1-unit beads
and a 5-unit bead. Applicable problems:
9+1 9+2 9+3 9+4

8+2 8+3 8+4
7+3 7+4
6+4
Note 2: Do not reverse motions 1 and 2 of Step 2. If you do, your operation will slow
down. Because after moving up 5, you will find it hard to move down 1 on B and move up
1 on A at the same time in the manner of twisting your fingers, although this latter
procedure is workable in some cases, for example, in adding 4 to 6 or 7. This is the main
reason why experts, in working out Example 6 ( 8 — 6 = 2 ), disfavor the procedure of
moving up 5 first, and moving down 1 next.
Example 16. 13 — 4 = 9
Step 1: Set 13 on AB.
Step 2: After moving down the 1 on A, move down 5 and move up 1 on B

at the same time.
Note: This example shows how to add both a 5-unit bead and one or more 1-unit bead
after subtracting 10. Applicable problems.
11—2 11—3 11—4

12—3 12—4
13—4
21
Example 17. 6 + 6 = 12
Step 2: Move up 1 on B (Motion 1), move up 5 on B (Motion 2), and

move up 1 on A (Motion 3). Work the two fingers with the idea of
performing the first two motions at the same time.
Note 1: This example shows how to add 1 to the tens’ rod after adding one or more
1-unit beads and subtracting a 5-unit bead on the unit rod. Applicable problems:
5+6 6+6 7+6 8+6

5+7 6+7 7+7
5+8 6+8
5+9
Note 2: It is possible to perform the last two motions of Step 2 above

at the same time after completing the first motion. But this procedure
should not be followed, as it does not work in some cases. For example,
the problem, “46 + 6 = 52” (Figs. 41 and 42) or “96 + 6 = 102,” can be
worked in no other way than that indicated.
Example 18. 12 — 6 = 6
Step 2: After moving down the 1 on A, move down 5 and 1 on B in

succession.
Note: This example shows how to add a 5-unit bead and subtract one or more 1-unit
beads after subtracting 10. Applicable problems:
11—6
12—6 12—7
13—6 13—7 13—8
14—6 14—7 14—8 14—9
Example 19. 9 + 7 = 16
Step 2: Move down 3 on B, and move up 1 on A. Perform the two

motions mechanically at the same time as if twisting the thumb and the
index finger.
Note 1: This example shows how to subtract one or more 1-unit beads from the unit rod
and add Applicable problems:
22
2+9 3+9 4+9 6+9 7+9 8+9 9+9

3+8 4+8 7+8 8+8 9+8
4+7 8+7 9+7
9+6
Note 2: Since 7 cannot be added to the 9 on the unit rod B, 3, the complementary digit of
7 for 10, is subtracted and 10 is added. This operation may be represented by the
equation: 9 + 7 = 9 - 3 + 10 = 16
Example 20. 16 - 7 = 9
Step 2: Move down the 1 on A, and move up 3 on B.
Note 1: When setting a two-digit number on the board, as in Step 1, always set the tens’
digit first.
Note 2: This example shows how to subtract 10 and add one or more 1-unit beads.
Applicable problems:
15—6 15—7 15—8 15—9

16—7 16—8 16—9
17—8 17—9
18—9
Note 3: Since 7 cannot be subtracted from the 6 on the unit rod, 10 is subtracted from
rod A, and 3, the complementary digit of 7 for 10, is added. The basis for this operation
may be represented by the equation: 16 - 7 = 16 - 10 + 3 = 9
Note 4: Mechanization of Operation.
The fundamental principle which makes abacus operation simple and speedy is
mechanization. To give a theoretical explanation, the mechanical operation of the abacus
is designed to minimize your mental labor and limit it to the unit rod, without carrying it to
the tens’ rod, by means of the complementary digits for 10 and 5, and to let the result
form itself mechanically and naturally on the board.
To give an example, in adding 7 to 9, the student accustomed to the Western mode of

calculation will probably form 16 on the board as a result of mental calculation to the
effect that 9 and 7 is 16. But such procedure is in every way inferior to the
above-mentioned mechanical one. Not only does this Western method require mental
exertion and time but it is liable to cause perplexity and errors.
When a problem of addition and subtraction is worked on the board, the procedure is
very simple. Addition and subtraction, which involve two rods, are simplified by means of
a complementary digit, that is, the digit necessary to give the sum 10 when added to a
given digit. For instance, suppose we have to add 7 on a rod where there is 9; then we think
or say, “7 and 3 is 10,” and subtract 3 from the rod in question, and add 1 to next rod on
the left. When we have to subtract 7 from 16, we think or say, “7 from 10 leaves 3,” and
subtract 1 from the next rod on the left, and add 3 to the rod in question. This means that
23
10 is always reduced to 1, and added or subtracted on the tens’ rod. Therefore, after
recalling the complementary digit, the operator has simply to perform either of the two
mechanical operations: subtracting the complementary digit and adding 1 on the tens’ rod
(in addition) or subtracting 1 on the tens’ rod and adding the complementary digit (in
subtraction). The result then will naturally form on the board. No matter how many digits
may be contained in the numbers to be added or subtracted, the entire operation is
performed by applying this mechanical method to each digit in turn.
The same mechanical method applies to the operations which require the analysis of 5
(see Examples 7 to 10). Suppose we have to add 3 to 4; then we merely think of the
complementary digit of 3 for 5 (that is, the digit necessary to give the sum 5 when added
to 3), and we move down the 5-unit digit and two 1-unit digits on the rod in question. Then
the result will naturally appear on the board. Any attempt to calculate the answer
mentally will retard the operation.
10 has only five groups of complementary digits: 9 and 1, 8 and 2, 7 and 3, 6 and 4, and
5 and 5, while 5 has only two: 4 and 1, and 3 and 2. Accordingly, the use of the mechanized
method requires no more mental effort than that of remembering one of the elements of
each of these very few pairs of complementary digits. This is the fundamental reason
which makes calculation by means of the complementary digit much simpler and speedier
and less liable to error than the ordinary way of mental or written calculation.
The following examples will show how much more laborious the ordinary calculation is.
In written calculation we proceed from right to left. For instance in the problem 99 + 88 +
77 + 66, we first add the unit digits, thinking “9 + 8 = 17, 17 + 7 = 24 and 24 + 6 = 30.” Next
we add the 30 to the 90 of 99 and work on. In the problem 567— 89, we cannot subtract the
9 from 7, so borrowing 10 from the 6 in the tens’ place, we get 8. Next proceeding to the
tens’ place we again find that we cannot subtract the 8 from the remaining 5 of the minu-
end, so we borrow 1 from the remaining 5 in the hundreds’ place, and we get 7 in the tens’
place and the answer 478. These processes involve laborious mental exertion.
On the other hand, all calculations on the abacus proceed from left to right, that is,
from the highest to the lowest digit. This accords with our natural customary practice of
naming or remembering all numbers from the highest to the lowest digit. Therefore, to set
numbers on the board is to calculate numbers.
In conclusion, incredibly speedy abacus operation is mainly attributable to three

reasons: mechanical operation by means of the complementary digit, left to right
operation, and the previously explained dozen rules of rational or scientific bead
manipulation. These are the reasons why, no matter how rapidly numbers may be
mentioned, as long as they are given distinctly, the skilled abacus operator can add and
subtract without any error, irrespective of how many digits the numbers may contain.
Note 5: The Order of Operation.
When addition involves two rods, as in the example 9 + 7=16, he sure to subtract 3 from
the unit rod, and next add 10 in the form of 1 to the tens’ rod. Thus 9 + 7 = 9 — 3 +10 = 16.
The idea of 7 in the terms of the complementary digit is “ 7 = 10—3.” So you would be
tempted to add 10 first and subtract 3 next. But as already pointed out, such a procedure
which involves unnecessary shifts of attention between the unit rod and the tens’ rod,
should not be followed. Because in adding 7 to 9, you will naturally first observe the unit
24
rod to add 7. Now subtract 3 from the unit rod, and then, proceeding to the tens’ rod, add
1. This procedure requires only two shifts of attention and operation, On the other hand, if
you did not subtract 3 first, you would have to come back to the unit rod to subtract 3 after
adding 1 to the tens’ rod. This inferior procedure would delay your operation.
The advantage of the correct procedure becomes even clearer in

some problems containing more than one digit. For instance, in
adding 7 to 996, note the decided advantage of forming 3 and
proceeding mechanically straight to the left to clear the tens’ and
hundreds’ rods of their 9 and to set 1 on the thousands’, as shown
in Figs. 49 and 50.
When subtraction involves two rods, as in the example 15—7 = 8, be sure to subtract 10
in the form of 1 from the tens’ rod, and next add 3 to the unit rod. Thus:
15 - 7 = 15 - 10 + 3 = 8.
In subtracting 7, naturally you will first look at the unit rod; then you will see that it is
impossible to subtract 7 from 6 and that you must borrow 10 from the tens’ rod. At this in-
stant subtract 10 in the form of 1, and then add 3, i. e., the complementary digit of 7 for
10, to the unit rod. So following this natural order of attention, first subtract the 10 and
then add the 3. If you were to reverse this order, you would have to shift your attention
back again to the tens’ rod to subtract 10 after adding 3. Thus this wrong procedure would
cause needless shifting of attention and delay operation. Note that failure to use the
complementary digit would necessitate the less efficient method of mental calculation.
The correct procedure is especially advantageous in some problems containing more

than one digit, e.g., the problem 1 000—1 = 999.
Following the correct procedure, in this problem we can proceed mechanically straight
from left to right (Fig. 52), while the incorrect procedure (Fig. 53) involves the loss of time
and labor.
When setting numbers of two or more digits, set the tens’ first. Also, when adding or
subtracting numbers of two or more digits, add or subtract beginning with the
highest-place digit. This is another fundamental rule which will produce efficiency. As
previously explained, when a number is named or given, beginning with the highest digit,
it can be mentally remembered or set and calculated much more naturally and easily on
the board than beginning with the lowest digit. This method is opposite to that of written
calculation, which is started back. ward with the last digit after a number has been given.
2. Adding and Subtracting Two-Digit Numbers
When setting two-digit numbers, set the tens’ first. Also when adding and subtracting
two-digit numbers, add and subtract the tens’ first, On the abacus always operate from
left to right. This is a fundamental rule based on efficiency. The efficiency of this rule is
especially true in the calculation of large numbers, as in Examples 26 and 27. As explained
25
in the introduction, since a number is named or given, beginning with the highest digit, it
can be mentally remembered or set and calculated much more naturally and easily on the
board, beginning with the highest-place digit than with the lowest place. Written
calculation is started backward with the last digit after a number has been given, whereas
on the abacus a number is calculated while it is being given, in other words, to set a
number is to calculate it.
Example 21. 14 + 25 = 39
Step 1: Set 14 on AB, with the 4 appearing on the unit rod B (Fig. 54).
Step 2: Add the 2 of 25 to the 1 on A with the thumb. This gives 3 on A and 34 on AB (Fig.
55).
Step 3: Add the remaining 5 of 25 to the 4 on B with the forefinger. This gives 9 on B.
The answer is 39 (Fig. 56).
The same procedure can be expressed diagrammatically as seen above.
Example 22. 45 + 27 = 72
Step 1: Set 45 on AB (Fig. 57).
Step 2: Add the 2 of 27 to the 4 on A. This

gives 6 on A and 65 on AB (Fig. 58).
Step 3: Add the remaining 7 of 27 to the 5 on B. This gives 2 on B and 7 on A, as 1 is

carried to the 6 on A. The answer is 72 (Fig. 59).
Example 23. 79 — 23 = 56
Step 2: Subtract the 2 of 23 from the 7 on A.

This leaves 5 on A and 59 on AB (Fig. 61).
Step 3: Subtract the remaining 3 of 23 from the 9 on B. This leaves 6 on B. The answer
is 56 (Fig. 62).
Example 24. 83 — 49 = 34
Step 2: Subtract the 4 of 49 from the 8 on A.

This leaves 4 on A and 43 on AB (Fig. 64).
26
Step 3: Subtract the remaining 9 of 49 from the 3 on B. As you cannot subtract 9 from 3,
borrow 1 from A This leaves 3 on A and enables you to subtract 9 from 10 on B. Add, to the
3 on B, the remainder 1 of 9 from 10, and you get 4 on B. The answer is 34 (Fig. 65).
3. Adding and Subtracting Numbers of Over Two Digits
The methods used in adding or subtracting numbers containing three or more digits are
the same as those just described in the case of two-digit numbers. Two problems each in
addition and subtraction will suffice to make this clear.
Example 25. 456 + 789 = 1 245
Step 1: Set 456 on BCD (Fig. 66).
Step 2: Add the 7 of 789 to the 4 on B. This gives you 11 on AB and 1 156 on ABCD (Fig.
67).
Step 3: Add the 8 of the remaining 89 to the 5 on C. This gives you 23 on BC and 1 236 on
ABCD (Fig. 68).
Step 4: Add the remaining 9 to the 6 on D. This gives you 45 on CD. The answer is 1 245.
(Fig. 69).
Example 26. 3 179 + 5 876 = 9 055
Step 1: Set 3 179 on ABCD (Fig. 70).
Step 2: Add the 5 of 5 876 to the 3 on A. This gives you 8 on A and 8 179 on ABCD.
Step 3: Add the 8 of the remaining 876 to the 1 on B. This gives you 9 on B and 8 979 on
ABCD (Fig. 72).
Step 4: Add the 7 of the remaining 76 to the 7 on C. This gives you 904 on ABC and 9 049
on ABCD (Fig. 73).
Step 5: Add the remaining 6 to the 9 on D. This gives you 55 on CD. The answer is 9 055
(Fig. 74).
27
Example 27. 623 — 375 = 248
Step 1: Set 623 on ABC (Fig. 75).
Step 2: Subtract the 3 of 375 from the 6 on A. This leaves 3 on A and 323 on ABC (Fig.
76).
Step 3: Subtract the 7 of the remaining 75 from the 2 on B by borrowing 1 from the 3 on
A. This leaves 25 on AB and 253 on ABC (Fig. 77).
Step 4: Subtract the remaining 5 from the 3 on C by borrowing 1 from the 5 on B. This
leaves 48 on BC. The answer is 248 (Fig. 78).
Example 28. 6 342 — 2 547 = 3 795
Step 1: Set 6 342 on ABCD, with 6 on the thousands’ rod and 2 on the ones’ (Fig. 79).
Step 2: Subtract the 2 of 2 547 from the 6 on A. This leaves 4 on A and 4 342 on ABCD (Fig.
80).
Step 3: Subtract the 5 of the remaining 547 from the 3 on B by borrowing 1 from the 4 on
A. This leaves 38 on AB and 3 842 on ABCD (Fig. 81).
Step 4: Subtract the 4 of the remaining 47 from the 4 on C. This leaves 0 on C on 3 802
on ABCD (Fig. 82).
Step 5: Subtract the remaining 7 from the 2 on D by borrowing 1 from the 8 on B. This
leaves 795 on BCD. The answer is 3 795 (Fig. 83).
4. Exercises
Probably the most convenient way of dealing with problems containing

a long file of numbers, such as the following, is to use the top edge of the
abacus as a marker. For example, in problem 1 on page 52, first place the
top edge of the abacus immediately under the first number, 24, and form
it on the board (Fig. 84); then move the abacus down until the next
number, 20, appears directly under the beads in use, and add that number on the beads
(Fig. 85); and continue in this fashion to the end of the problem.
28
Abacus calculation is also greatly facilitated by having someone call out the successive
numbers. Numbers should be read distinctly and quickly. For example, the number 123 456
789 should be given as “one, two, three million; four, five, six thousand; seven, eight,
nine.”
The best exercise for attaining skill is to add 123 456 789 fine times. If your sum is
correct, it will be 1 111 111 101. Again, add 789 fine times on the rods GHI, with I as the
unit rod; next add 456 on DEF fine times; finally add 123 on ABC nine times, and you will
get the same sum. Subtract 123 456 789 from 1 111 111 101 nine times, and you will end
with zero. These three exercises involve every procedure used in problems of addition and
subtraction. A third-grade abacus operator can work each of the three exercises in one
minute, and a first-grade operator in thirty seconds.
I. Adding and Subtracting Two-Digit Numbers

(1) (2) (3) (4) (5) (6) (7) (8)
24 55 22 11 33 55 66 55
20 44 66 44 77 55 44 77
—33 —55 —88 —33 —88 —55 —22 —80
—11 —22 77 44 99 54 33 69
22 55 —66 —22 —88 35 —44 —76
12 —11 55 33 77 —55 55 88
—23 —55 —66 —44 —99 75 —33 —66
11 11 0 33 11 164 99 67
II. Adding and Subtracting Three-Digit Numbers

(1) (2) (3) (4) (5)
222 345 561 621 158
665 762 259 946 782
778 473 846 —255 —345
555 528 667 428 566
335 981 445 564 444
778 811 778 —392 —657
222 176 289 734 216
889 634 265 855 774
443 367 778 —628 —889
223 189 665 —476 —677
5 110 5 266 5 553 2 397 322
III. Adding and Subtracting Four-Digit Numbers

(1) (2) (3) (4) (5)
3,627 9,105 2,456 7,081 6,924
1,508 2,746 8,193 5,469 8,570
9,472 1,809 5,647 —2,505 1,439
6,345 5,321 7,038 3,748 —3,268
8,160 4,684 9,825 4,917 —7,015
2,079 3,263 3,741 —6,803 9,847
4,384 5,162 6,580 —6,294 5,192
7,819 7,038 1,269 1,372 2,603
5,623 8,574 4,001 9,620 —3,786
1,950 9,970 9,372 8,135 4,051
50, 967 57,672 58, 122 24,740 24, 557
29
V. MULTIPLICATION
There are several methods of multiplication on the abacus. The one introduced in the
following pages is a recent method which is generally considered the best and is now the
standard method taught in grade schools. In describing the method, standard terminology
will be used. Thus, for example, in the problem 5 x 2 = 10, 5 will be called the multiplicand,
2 the multiplier, and 10 the product.
It is customary to set the multiplicand at the central part of the abacus and the
multiplier to the left, leaving two or three rods unused between the two numbers, just
enough to separate them clearly but not too widely. The decision of the Abacus Committee
in favor of two unused rods will be followed in our problems here.
The method of multiplication used here gives the product immediately to the right of
the multiplicand. There is a less favored method which gives the first figure of the product
immediately to the left.
Two main reasons can be given for setting the multiplicand on the right and the
multiplier on the left. Since the abacus is operated with the right hand, a reverse order of
setting the two figures would cause the multiplier to be hidden by the hand much of the
time. Moreover, in case the multiplier is a large number, too much space would have to be
left between it and the multiplicand. Otherwise the product, which is produced at the
right of the multiplicand would extend right into the multiplier.
Although the use of a unit rod marked with a unit point does not have as much bearing
in problems of multiplication and division as in addition and subtraction, it does facilitate
calculation in many ways. Therefore, in the following problems the unit figure of the
multiplicand is set on a unit rod. In the case of the multiplier, however, so long as it is not
a fractional number, the unit rod is disregarded and the unit figure is set on the third rod
to the left of the multiplicand.
As for the order of setting the multiplicand and the multiplier, since the unit figure of
the multiplicand must be set on a unit rod, it is advisable for the beginner to set the
multiplicand ahead of the multiplier. It should be noted, however, that experts can locate
both the multiplicand and the multiplier at a glance. So they very often set the multiplier
ahead of the multiplicand, thus saving the time required in shifting the hand back to the
left after setting the multiplicand. Even more frequently experts set only the multiplicand
on the board not even troubling to set the multiplier.
1. Multiplying by One-Digit Numbers
Example 1. 4x2=8
Step 1: Set the multiplicand 4 on the

unit rod D and the multiplier 2 on rod A,
thus leaving two vacant rods between
the numbers as in Fig. 86
Step 2: Multiply the multiplicand 4 by the mu1tiplier 2. Set the product 8 on F, the
second rod to the right of the multiplicand, and clear rod D of its 4. Fig. 87 shows the result
30
of Step 2.
Note: The accompanying diagram shows another way to illustrate the same problem.
Here the two figures in the row designated Step 1 indicate that the multiplier 2 and the
multiplicand 4 have been set on A and D respectively. The two figures in the row
designated Result show the multiplier 2 remaining on A and the product 8 which has been
set on F as the result of the multiplication.
All the following examples will be illustrated in the two ways shown above.
The reasons for clearing off the multiplicand after its multiplication will be given at the
end the section on multiplication by two-digit numbers.
Example 2. 8 x 6 = 48
Step 1: Set 8 on the unit rod D and 6

on A, leaving two vacant rods (Fig. 88).
Step 2: Multiplying 8 by 6, set the

product 48 on EF, and clear D of its 8. In
this step, the first rod to the right of the multiplicand, designated E, is the tens’ rod of the
product 48 (Fig. 89).
Note: Some experts say it is desirable to clear away the multiplicand before setting the
product. For instance, in the above example, they say that the product 48 should be set on
EF after clearing D of its 8. This method has the merit of saving the time of shifting the
hand back to the left to clear off the multiplicand after setting the product. But the
Abacus Committee frowns upon this procedure, saying that, especially for beginners, it is
apt to cause confusion in that the multiplicand must be carried in the memory after it has
been cleared away from the board.
Example 3. 24 x 7 = 168
Step 1: Set 24 on DE, with E as

the unit rod, and set 7 on A (Fig.
90).
Step 2: Multiplying the 4 in 24 by

7, set the product 28 on FG, and
clear E of its 4 (Fig. 91).
Step 3: Multiplying the

remaining 2 in 24 by 7, set the
product 14 on EF, thereby adding this new product to the 28 on FG, and clear E of its 2.
This makes a total of 168 on EFG, which is the answer (Fig. 92).
Note 1: The reason for setting the product 14 on the rods EF, which are one place higher
than FG, is obvious. When adding 14, do not take the trouble of thinking that this product
is 140 in actual value and that therefore this must be set on EF. Instead just mechanically
set the 1 in 14 on E and add the 4 in 14 to the previous 2 on F, and let the result form itself
automatically.
31
Note 2: In Step 2, F is the tens’ rod of the product 28, while in Step 3, E is the tens’ rod
of the product 14. In each step of multiplication, the first rod to the right of that figure in
the multiplicand which is multiplied is the tens’ rod of the product.
Note 3: When there are two digits in the multiplicand, first multiply the last digit by the
multiplier and then the first digit.
2. Multiplying by Two-Digit Numbers
Example 4. 8 x 17 = 136
Step 1: Set 8 on the unit rod E

and 17 on AB (Fig. 93).
Step 2: Multiplying the 8 by

the 1 in 17, set the product 8 on
G (Fig. 94).
Step 3: Multiplying the 8 by the 7 in 17, set the product 56 on

GH, and clear E of its 8. Since you already have 8 on G, you get, on
FGH, a total of 136, which is the answer (Fig. 95).
Note: When there are two digits in the multiplier, first multiply the multiplicand by the
first digit of the multiplier and next by the last digit of the multiplier.
Example 5. 46 x 23 = 1,058
Step 1: Set 46 on EF, with E as the unit

rod, and set 23 on AB (Fig. 96).
Step 2: Multiplying the 6 in 46 by the 2

in 23, set the product 12 on GH (Fig. 97).
Step 3: Multiplying the same 6 in 46 by

the 3 in 23, set the product 18 on HI and
clear F of its 6. Since you have 12 on GH,
you get a total of 138 on GH (Fig. 98). Re-
member that each time the same digit in
the multiplicand is multiplied by one digit
after another in the multiplier, the value
of the product is reduced by one rod or
place.
Step 4: Multiplying the 4 in 46 by the 2 in 23, set the product 8 on G. This makes a total
of 938 on GH (Fig. 99).
Step 5: Multiplying the same 4 in 46 by the 3 in 23, set the product 12 on GH and clear
E of its 4. This leaves the answer 1 058 on FGHI (Fig. 100).
32
Note: In case both the multiplier and the multiplicand have two digits, (1) multiply the last
digit of the multiplicand by the first digit of the multiplier; (2) multiply the same digit of
the multiplicand by the last digit of the multiplier; (3) multiply the first digit of the
multiplicand by the first digit of the multiplier; and (4) multiply the same first digit of the
multiplicand by the last digit of the multiplier. This is the fundamental rule of
multiplication.
Example 6. 97 x 48 = 4 656
Step 1: Set 97 on EF, with F as the unit

rod, and set 48 on AB (Fig. 101).
Step 2: Multiplying the 7 in 97 by the 4

in 48, set the product 28 on GH (Fig.
102).
Step 3: Multiplying the same 7 in 97 by

the 8 in 48, set the product 56 on HI, and
clear F of its 7. Since you have 28 on GH,
you get a total of 336 on GHI (Fig. 103).
Step 4: Multiplying the remaining 9 in

97 by the 4 in 48, set the product 36 on
FG. This makes a total of 3 936 on FGHI
(Fig. 104).
Step 5: Multiplying the same 9 in 97 by the 8 in 48, set the product 72 on GH, and clear
E of its 9. This gives you, on FGHI, a total of 4 656, which is the answer (Fig. 105).
Note: The preceding examples will have indicated the desirability of clearing off each
digit in the multiplicand after its multiplication by all the digits in the multiplier. If you did
not do so, you would be greatly inconvenienced in operation. This is especially the case
when the multiplicand is a large number. First, you would often find it hard to tell which of
the digits in the multiplicand you had multiplied by all the digits in the multiplier. Second,
this incorrect procedure would necessitate the removal of the multiplier further to the
right beyond the product of the correct procedure by as many digits as there are in the
multiplicand.
3. Multiplying by Numbers of Over Two-Digits
No matter how many digits the multiplier may have, the principle of multiplication is
the same as that of multiplying by two-digit numbers. You have only to see that you do not
mistake the order of multiplication and the rods on which to set products.
Example 7. 37 x 432 = 15 984
Step 1: Set 37 on FG, with G as the

unit rod, and set 432 on ABC (Fig.
106).

33
the 4 in 432, set the product 28 on HI

(Fig. 107).
Step 3: Multiplying the same 7 in

37 by the 3 in 432, set the product 21
on IJ. This makes a total of 301 on HIJ
(Fig. 108).

on JK, and clear G of its 7. This makes
a total of 3,024 on HIJK (Fig. 109).

the 4 in 432, set the product 12 on
GH. This makes a total of 15,024 on
GHIJK (Fig. 110).
Step 6: Multiplying the 3 in 37 by the 3 in 432, set the product 9 on 1. This makes a total
of 15,924 on GHIJK (Fig. 111).
Step 7: Multiplying the 3 in 37 by the 2 in 432, set the product 6 on J, and clear F of its
3. This makes, on GHIJK, a total of 15,984, which is the answer (Fig. 112).
Example 8. 78 x 503 = 39 234
Step 1: Set 78 on FG, with G as the

unit rod, and set 503 on ABC (Fig.
113).

the 5 in 508, set the 4 of the product
40 on H (Fig. 114).

on JK and clear G of the 8. This makes
a total of 4,024 on HIJK. In setting
this product skip rod I as the second
figure of the multiplier 503 is zero. In
other words, the product must be set
on JK instead of on IJ (Fig. 115).
Step 4: Multiplying the 7 in 78 by the 5 in 503, set the product 35 on GH. This makes a
total of 39 024 on GHIJK (Fig. 116).
Step 5: Multiplying the same 7 in 78 by the 3 in 503, set the product 21 on IJ instead of
HI, as the second figure of 503 is zero, and clear F of its 7. This leaves, on GHIJK, a total of
39 234, which is the answer (Fig. 117).
34
4. Exercises
Group I
1. 34 x 4 = 136 11. 21 x 23 = 483
2. 23 x 5 =115 12. 12 x 32 = 384
3. 12 x 4 = 48 13. 21 x 43 = 903
4. 82 x 3 = 96 14. 12 x 56 = 672
5. 21 x 5 = 105 15. 31 x 64 = 1 984
6. 33 x 45 = 1 485 16. 43 x 56 = 2 408
7. 52 x 56 = 2 912 17. 32 x 64 = 2 048
8. 23 x 65 = 1 495 18. 53 x 76 = 4 028
9. 53 x 75 = 3 975 19. 23 x 83 = 1 909
10. 25 x 85 = 2 125 20. 35 x 96 = 8 860
Group II
1. 112 x 23 = 2 576 11. 1 023 x 34 = 34 782
2. 123 x 35 = 4 305 12. 3 243 x 45 = 145 935
3. 212 x 46 = 9 752 13. 4 352 x 58 = 252 416
4. 845 x 57 = 19 665 14. 5 624 x 67 = 376 808
5. 423 x 64 = 27 072 15. 6 712 x 78 = 523 536
6. 513 x 76 = 38 9S8 16. 132 x 334 = 44 088
7. 607 x 87 = 52 809 17. 234 x 456 = 106 704
8. 452 x 85 = 38 420 18. 431 x 467 = 201 277
9. 631 x 95 = 59 945 19. 546 x 686 = 374 556
10. 608 x 97 = 58 976 20. 756 x 879 = 664 524
35
VI. DIVISION
There are two fundamental methods of division on the abacus. The older method,
though still favored by some, has fallen out of general use since about 1930 because it
requires the memorization of a special division table. The newer method, which is the
easier to learn because it uses the multiplication instead of the division table, is the
standard one now taught in grade schools, and will be introduced in the following pages.
Strictly speaking, it is not new, as it has long been used, but only in a very limited use until
around 1930 when it was improved and publicized. Standard terminology will be used in
describing the method. For example, in the problem 50 / 5 = 10, 50 is the dividend, 5 the
divisor, and 10 the quotient.
It is customary to set the dividend a little to the right of the central part of the abacus
and the divisor at the left. The two numbers are generally separated by three or four
unused rods. As the Abacus Committee favors leaving four unused rods between the two
numbers, the following examples will adhere to that practice.
The method of division used here gives the first digit of the quotient between the
dividend and divisor. Two main reasons can be given for setting the dividend on the right
and the divisor on the left. One is that since the abacus is operated with the right hand,
the reverse order of setting the two numbers would cause the multiplier to be hidden by
the hand much of the time, as in the case of multiplication. The other is that in case the
dividend is indivisible by the divisor, the reverse order would cause the quotient to extend
right into the divisor.
In division, as in multiplication, the use of the unit rod is not too essential, but does
facilitate calculation in many ways. Therefore, the unit figure of the dividend is always set
on a unit rod. When the divisor is a whole number, however, we shall disregard the unit rod,
and simply set the divisor in such a way that its last digit is located on the fifth rod to the
left of the dividend.
As for the order of setting the dividend and divisor, since the last digit of the dividend
must be set on a unit rod, it is advisable for the beginner to set the dividend before setting
the divisor. As is the case with multiplication, however, experts often reverse the
procedure, setting the divisor first or not at all.
1. Dividing by One-Digit Numbers
Example 1. 8/ 2 = 4
Step 1: Set the dividend 8 on rod F

and the divisor 2 on rod A, with four
vacant rods between the two numbers.
Make sure that F is a unit rod marked
with a unit point (Fig. 118).
Step 2: Mentally divide 8 by 2 ( 8 / 2 = 4 ); set the quotient 4 on D, the second rod to the
left of the dividend; and clear F of its 8. Fig. 119 and the row of figures designated Result
in the diagram show the result of this step.
36
Example 2. 837+ 3 = 279
Step 1: Set 837 on the rods FGH, with H

as the unit rod, and set 3 on A (Fig. 120).
Step 2: Compare the 3 with the 8 in 837.

3 goes into 8 twice with 2 left over. Set the
quotient figure 2 on D, the second rod to
the left of 8 in 837. Next multiply the divisor 3 by this quotient figure
2, and subtract the product 6 from the 8 on F. This leaves 2 on F (Fig.
121).
Step 3: Compare the 3 with 23 on FG. The 2 on F is the remainder

left over as a result of the previous step. 3 goes into 23 seven times
with 2 left over. Set 7 as the quotient figure on E. Next multiply the
divisor 3 by this 7, and subtract the product 21 from the 23 on FG.
This leaves 2 on G (Fig. 122).
Step 4: Compare the 3 with 27 on GH.

The 2 on G is the remainder left over as a
result of the second step. 3 goes into 27
nine times. Set the quotient figure 9 on F.
Next multiply the 3 by this 9, and subtract
the product 27 from the 27 on GH. This
clears GH and leaves the answer 279 on DEF (Fig. 123).
Note: Answers to problems in division can be easily checked by multiplication. Thus, to

check the foregoing answer, simply multiply the quotient 279 on DEF by the divisor 3, that
is, the number you originally divided by, and you will get the product 837 on FGH, i. e., the
same rods on which you had 837 as the dividend. By this checking the student will see that
the position of the quotient in division is that of the multiplicand in multiplication, and
that the position of the dividend in division is that of the product in multiplication.
Therefore, we may say that the methods of multiplication and division introduced in this
book form the counterpart of each other.
Example 3. 6 013 / 7 = 859
Step 1: Set 6,013 on FGHL with 1 as

the unit rod, and set 7 on A (Fig. 124).
Step 2: Compare the divisor 7 with

the 6 in 6,013. 7 will not go into 6. So
compare the 7 with the 60 in 6,013. 7
goes into 60 eight times. In this case
set the quotient figure 8 on E. the first
rod to the left of the first digit of the
dividend. Next multiply the divisor 7 by
this 8, and subtract the product 56
from the 60 on FG. This leaves 4 on G
(Fig. 125).
Step 3: Compare the 7 with 41 on GH 7 goes into 41 five times. Set the quotient figure 5 on
37
F. Next multiply the 7 by this 5, and subtract the product 35 from the 41 on GH. This leaves
6 on H (Fig. 126).
Step 4: Compare the 7 with 63 remaining

on HI. 7 goes into 63 nine times. Set the
quotient figure 9 on G. Next multiply the 7
by this 9, and subtract the product 63 from
the 63 remaining on HI. This clears HI, and
leaves the answer 859 on EFG (Fig. 127).
Note: When the divisor is larger than the first digit of the dividend, compare it with the
first two digits of the dividend. In this case set the quotient figure on the first rod to the
left of the first digit of the dividend. The chief merit of this procedure is that, in checking,
the quotient multiplied by the divisor gives the product on the very rods on which the
dividend was located previous to its division.
This procedure is the same as the principle of graphic division. In dividing 36 by 2, you
write the quotient figure 1 above the 3 in 36. But in dividing 36 by 4, you write the quotient
figure 9 above the 6 in 36. On the abacus board the quotient figure cannot be put above
the dividend. So in dividing 36 by 2, the first quotient figure 1 is set on the second rod to
the left of 36, while in dividing 36 by 4, the quotient figure 9 is set on the first rod to the
left of 36.
2. Dividing by Two-Digit Numbers
Example 4. 552 / 23 = 24
Step 1: Set 552 on GHI, with 1 as the

unit rod, and set 23 on AB (Fig. 128).
Step 2: Compare the 2 in 23 with the 5

in 552. 2 goes into 5 two times. Set the
quotient figure 2 on E, the second rod to
the left of the 5 in 552. Next multiply the
2 in 23 by this quotient figure 2, and
subtract the product 4 from the 5 on G.
This leaves 1 on G (Fig. 129).
Step 3: Now multiply the 3 in 23 by the

same quotient figure 2, and subtract the
product 6 from 15 on GH. This leaves 9 on
H (Fig. 130).

on H. 2 goes into 9 four times. Set the
quotient figure 4 on F. Next multiply the 2 in 23 by this quotient figure 4, and subtract the
product 8 from the 9 on H. This leaves 1 on H (Fig. 131).
Step 5: Multiply the 3 in 23 by the same 4, and subtract the product 12 from the 12
remaining on HI. This clears HI and leaves the answer 24 on EF (Fig. 132).
38
Example 5. 6 308 / 83 = 76
Step 1: Set 6 308 on GHIJ, with J as the

unit rod, and set 83 on AB (Fig. 133).

in 6,308. 8 will not go into 6. So
compare the 8 with the 63 in 6,308. 8
goes into 63 seven times. Set the
quotient figure 7 on F, the first rod to
the left of the 6 in 6,308. Next multiply
the 8 in 83 by this 7 in the quotient, and
subtract the product 56 from the 63 on
GH. This leaves 7 on H (Fig. 134).
Step 3: Multiply the 3 in 83 by the same quotient figure 7, and subtract the product 21
from 70 on HI. This leaves 49 on HI (Fig. 135).
Step 4: Compare the 8 in 83 with

the 49 on HI. 8 goes into 49 six times.
Set the quotient figure 6 on G. Next
multiply the 8 in 83 by this 6, and
HI. This leaves 1 on I (Fig. 136).
Step 5: Multiply the 3 in 83 by the same quotient figure 6, and subtract the product 18
from 18 on IJ. This clears IJ. and leaves the answer 76 on FG (Fig. 137).
Note: In case the divisor is a two-digit number, do not take the trouble of comparing its
two digits with the first two or three digits of the dividend to work out the correct quotient
figure mentally. Simply compare the first digit of the divisor with that of the dividend.
When the first digit of the divisor is larger than that of the dividend, compare it with the
first two digits of the dividend. In case quotient figures tried are incorrect, correct them
by the methods shown in Examples 6, 7, and 8 instead of perplexing yourself with mental
arithmetic. Thus make the most of the chief advantage of the abacus, the complete
mechanical process which minimizes mental labor, and experience will enable you to find
correct quotient figures at a glance.
Example 6. 4,698 / 54 = 87
This example shows how the process

of division must be revised when too
large a quotient figure has been used.
Step 1: Set 4 698 on GHIJ, with J as

the unit rod, and set 54 on AB (Fig.
138).
Step 2: The 5 in 54 will not go into the 4 in 4 698. So compare the 5 with the 46 in 4 698.
5 goes into 46 nine times. Now suppose you have tried 9 as the quotient figure instead of
the correct 8 and have set it on F. Then you will multiply the 5 in 54 by 9, and subtract the
product 45 from the 46 on GH. This leaves 1 on H. Next multiplying the 4 in 54 by the same
9, you will find that the product 36 is larger than the 19 remaining on HI and that you ought
39
to have tried a quotient figure one less than 9 (Fig. 139).
Step 3: To revise the incorrect quotient figure 9 to 8,

subtract 1 from the 9 on F, and you get the new quotient
figure 8 on F. Next multiply the 5 in 54 by 1, i. e., the
difference between the quotient figures 9 and 8, and add
the product 5 to the 1 on H. Now you have 6 on H (Fig.
140).
Step 4: Multiply the 4 in 54 by the new quotient figure 8,

and subtract the product 32 from 69 on HI. This leaves 37
on HI (Fig. 141).

the 37 on HI. 5 goes into 37 seven times.
So set the quotient figure 7 on G. Next
multiply the 5 in 54 by this 7, and
HI. This leaves 2 on I (Fig. 142).
Step 6: Multiply the 4 in 54 by the

same quotient figure 7, and subtract
the product 28 from 28 remaining on IJ.
This clears IJ and leaves the answer 87
on FG (Fig. 143).
Example 7. 1 666 / 17 = 98
This example shows how a problem of division is worked when the first digit of both
divisor and dividend are the same.

the unit rod, and set 17 on AB (Fig. 144).
Step 2: When the first digit of the

divisor and the dividend are the same,
as in this example, compare the second
digits of the two numbers. In such a
situation, if the second digit of the
dividend is smaller than that of the
divisor, try 9 as the quotient figure. If 9 is
too large, try 8 as in Step 5 of this
example. If 8 is still too large, go on
trying a quotient figure one less till the
correct one is found. In such a case 9 is
the figure likeliest to be correct.
Now try 9 as the quotient figure and set it on F, the first rod to the left of the first digit of
the dividend. Next multiply the 1 in 17 by this 9 and subtract the product 9 from 16 on GH.
This leaves 7 on H (Fig. 145).
Step 3: Multiply the 7 in 17 by this same 9, and subtract the product 63 from 76 on HI.
This leaves 13 on HI (Fig. 146).
40
Step 4: The 1 in 17 and the 1 remaining on H are the same. So compare the 7 in 17 and
the 3 remaining on I. 3 is smaller than 7. So try 9 as the quotient figure and set it on G. Now
multiply the 1 in 17 by this 9 and subtract the product 9 from the 13 on HI. This leaves 4 on
1. Next, multiplying the 7 in17 by this same 9, you will see that the product 63 is larger
than 46 remaining on IJ. So you will find that you ought to have tried 8 as the quotient
figure (Fig. 147).
Step 5: To revise the incorrect quotient figure 9 to 8, subtract 1

from the 9 on G. Next you must revise the division in Step 4. So
multiply the 1 in 17 by 1, the difference between the 9 and 8, and
add the product 1 to the 4 remaining on I. Then you get 5 on I (Fig.
148).
Step 6: Multiply the 7 in 17 by the new

quotient figure 8 and subtract the
product 56 from 56 on IJ. This clears IJ
and leaves the answer 98 on FG (Fig.
149).
Note: In cases where the first digits of both the divisor and the dividend are the same, if
the second digit of the dividend is larger than that of the divisor, set 1 as the quotient
figure on the second rod to the left of the first digit of the dividend. An instance is given in
Example 9.
Example 8. 7 644 / 84 = 91
This example is to show how division

is to be revised when the quotient figure
tried is too small.

the unit rod, and set 84 on AB (Fig. 150).
Step 2: The 8 in 84 will not go into the 7 in 7 644. So compare

the 8 with the 76 in 7 644. 8 goes into 76 nine times. So you ought
to try 9 as the quotient figure. But suppose by mistake you have
tried 8 as the quotient figure instead of the correct 9 and have set
it on F. Then you will multiply the 8 in 84 by 8 and subtract the
product 64 from the 76 on GH. This will leave 12 on GH (Fig. 151).
Step 3: Multiplying the 4 in 84 by the same quotient figure 8,

you will subtract the product 32 from 124 on GHI. Then you will
find that the remainder 92 is larger than 84 and that you ought to
have tried 9, i.e., a quotient figure one more than 8 (Fig. 152).
Step 4: To revise the incorrect quotient figure 8 to 9, add 1 to

the quotient figure 8 on F. Next multiply the divisor 84 by 1, i. e.,
the difference between the two quotient figures, 8 and 9, and
subtract the product 84 from the 92 on HI. This leaves 8 on I (Fig.
153).
41
Step 5: The 8 in 84 and the 8

remaining on I are the same. So compare
the 4 in 84 with the 4 remaining on and
you can see that they are also the same.
Therefore, set the quotient figure 1 on G.
Now, multiplying the 8 in 84 by 1,
subtract the product 8 from the 8 on I. Next multiplying the 4 in 84 by the same 1, subtract
the product 4 from the 4 on J. This clears IJ and leaves the quotient 91 on FG (Fig. 154).
Example 9. 3 978 / 234 = 17
Step 1: Set 3,978 on HIJK, with K

as the unit rod, and set 234 on ABC
(Fig. 155).

the 3 in 3 978. 2 goes into 3 one time.
Set the quotient figure 1 on F, the
second rod to the left of the 3 in
3978. Now multiply the 2 in 234 by
this quotient figure 1, and subtract
the product 2 from the 3 on H. This
leaves 1 on H (Fig. 156).
Step 3: Multiply the 3 in 234 by the same quotient figure 1, and

subtract the product 3 from 9 on 1. This leaves 6 on 1 and 167 on
HIJ (Fig. 157).
Step 4: Multiply the 4 in 234 by the same quotient figure 1, and

subtract the product 4 from 7 on J. This leaves 3 on J and 1638 on
HIJK (Fig. 158).
Step 5: Compare the 2 in 234 with the 16 remaining on HI. 2

goes into 16 eight times. Suppose you have tried 8 as the quotient
figure instead of the correct 7 and have set it on G. Then you will multiply the 2 in 234
by 8, and subtract the product 16 from the 16 on HI. This clears HI. Next, multiplying the
3 in 234 by the same 8, you will find that the product 24 is larger than 3 remaining on J, and
that you ought to have tried a quotient figure one less than 8 (Fig. 159).
Step 6: To revise the incorrect

quotient figure 8 to 7, subtract 1 from
the 8 on G, and you get the new
quotient figure 7 on G. Next multiply
the 2 in 234 by 1, i. e., the difference
between the quotient figures 8 and 7,
and set the product 2 on I. Now you
have 2 on I and 234 on IJK (Fig. 160).
Step 7: Multiply the 3 in 234 by

the new quotient figure 7, and subtract the product 21 from 23 on IJ. This leaves 2 on 3 and
28 on JK (Fig. 161).
Step 8: Next multiply the 4 in 234 by the same new quotient figure 7, and subtract the
product 28 from 28 on JK. This clears JK and leaves the answer 17 on FG (Fig. 162).
42
Example 10. 7,061 / 307 = 23
Step 1: Set 7 061 on HIJK, with K as

the unit rod, and set 307 on ABC,
leaving as always four vacant rods
between the two numbers (Fig. 163).
Step 2: Comparing the 3 in 307 and

the 7 in 7,061 you can see that 3 goes
into 7 two times. Set the quotient
figure 2 on F. Next multiply the 3 in 307
by this 2, and subtract the product 6
from the 7 on H. This leaves 1 on H (Fig.
164).
Step 3: Multiply the 7 in 307 by the

same quotient figure 2, and setting
the product 14 on IJ, subtract it from
106 on HIJ. This leaves 92 on IJ. Since
the second digit in 307 is zero, see
that you set the product 14 on IJ
instead of HI (Fig. 165).
Step 4: The 3 in 307 goes into the 9 on I three times. So set the quotient figure 3 on G.
Next multiply the 3 in 307 by this quotient figure 3 and subtract the product 9 from the 9
on 1. This leaves 21 on JK (Fig. 166).
Step 5: Multiply the 7 in 307 by the same quotient figure 3 and subtract the product 21
from the 21 on JK. This clears JK and leaves the answer 23 on FG (Fig. 167).
4. Exercises
Group I
1. 24 / 2 = 12 11. 132 / 12 = 11
2. 36 / 3 = 12 12. 441 / 21 = 21
3. 115 / 5 = 23 18. 1 495 / 65 = 23
4. 204 / 6 = 34 14. 2 451 / 57 = 43
5. 357 / 7 = 51 15. 4 293 / 81 = 53
6. 424 / 4 = 106 16. 9 384 / 92 = 102
7. 4 008 / 8 = 501 17. 5 134 / 34 =151
8. 7 470 / 9 = 830 18. 4 635 / 45 = 103
9. 5 202 / 2 = 2 601 19. 15 990 / 78 = 205
10. 2 804 / 4 = 701 20. 84 056 / 14= 6 004
Group II
1. 8 296 / 68 = 122 11. 6 342 / 453 = 14
2. 4 270 / 14 = 305 12. 9 728 / 304 = 32
3. 11 100 / 75 = 148 13. 38 920 / 695 = 56
4. 7 560 / 28 = 270 14. 46 113 / 809 = 57
5. 24 957 / 47 = 531 15. 26 460 / 147 =180
6. 42 024 / 51 = 824 16. 215 940 / 236 = 915
7. 48 052 / 82 = 586 17. 178 712 / 502 = 356
8. 87 608 / 94 = 932 18. 459 780 / 970 = 474
9. 21 245 / 35 = 607 19. 874 038 / 418 = 2 091
10. 73 264 / 76 = 964 20. 690 988 / 761 = 908
43
VII. DECIMALS
In addition and subtraction the unit point serves as the mark of a decimal point, and the
calculation of decimal problems is quite the same as that of whole numbers.
However, in multiplication and division, you cannot easily find the unit rod of the
product and that of the quotient unless you know two rules covering the position of the
decimal point of the product and two others covering the position of the decimal point of
the quotient. These four rules may be best explained and illustrated in paired counterparts.
The first pair of rules applies to whole or mixed-decimal numbers, and the second to
decimal fractions.
Rule A. When the multiplier is a whole or a mixed decimal, the unit rod of the product
moves to the right of that of the multiplicand by as many rods plus one as there are
whole digits in the multiplier.
Rule B. When the divisor is a whole or a mixed decimal number, the unit rod of the
quotient moves to the left of the unit rod of the dividend by as many rods plus one
as there are whole digits in the divisor.
Rule C. When the multiplier is a decimal fraction whose first significant figure is in the
tens place, the last digit of the product is formed on the first rod to the right of the
last digit of the multiplicand. Call this the basic rod. Then, each time the value of
this multiplier is reduced by one place, the last digit of the product shifts by one
rod to the left of this basic rod.
Rule D. When the divisor is a decimal fraction whose first significant figure is in the tens
place, the last digit of the quotient is formed on the first rod to the left of the last
digit of the dividend. Call this the basic rod. Then, each time the value of this
divisor is reduced by one place, the last digit of the quotient shifts by one rod to
the right of this basic rod.
Example 1. (A) 4 x 2 = 8 (B) 8 / 2 = 4
As seen in the first diagram above, showing the position of the multiplier (Rod A), the
multiplicand (Rod D), and the product (Rod F), when the multiplier is a one-digit number,
the unit rod of the product moves by two rods to the right of that of the multiplicand. In
other words, the last digit of the product is formed on the second rod to the right of that
of the multiplicand.
As seen in the second diagram above, showing the position of the divisor (Rod A), the
dividend (Rod F), and the quotient (Rod D), when the divisor is a one-digit number, the unit
rod of the quotient moves by two rods to the left of that of the dividend. In other words,
the last digit of the quotient is formed on the second rod to the left of that of the dividend.
Example 2. (A) 25 x 15 = 375 (B) 375 / 15 = 25
The first diagram below shows that when the multiplier is a two-digit number, the last
44
digit of the product is formed on the third rod to the right of that of the multiplicand.
The second diagram below shows that when the divisor is a two-digit number, the last
digit of the quotient is formed on the third rod to the left of that of the dividend.
Example 3. (A) 405 X 123 = 49 815 (B) 49 815 / 123 = 405
The first diagram above shows that when the multiplier is a three-digit number, the last
digit of the product is formed on the fourth rod to the right of that of the multiplicand.
The second diagram shows that when the divisor is a three. digit number the last digit of
the quotient is formed on the fourth rod to the left of that of the dividend.
Note on Example 3 (B): In case the dividend is separated from the divisor with four
vacant rods, as in this example, the quotient product is clearly distinguishable from the
divisor, since two vacant rods are left between them, thus:
But if the dividend were separated from the divisor with only three vacant rods, the
quotient produced would be hardly distinguishable from the quotient, since only one
vacant rod would be left between them thus:
From this example the reader will see that in case the second figure of the quotient is a
cipher, the quotient is hardly distinguishable from the divisor. Thus it is always preferable
to set the divisor on the fifth instead of the fourth rod to the left of the dividend.
Example 4. (A) 34 x 1.2 = 40.8 (B) 40.8 / 1.2 = 34
45
Observe the first diagram above, and you will find that when the multiplier is a mixed
number, the last whole digit of the product moves to the right of that of the multiplicand
by as many rods plus one as there are whole digits in the multiplier.
Observe the second diagram, and you will find that when the divisor is a mixed number,
the last whole digit of the quotient moves to the left of that of the dividend by as many
rods plus one as there are whole digits in the divisor.
Example 5. (C) 32 x 0.4 =12.8 (C’) 98 x 0.32 = 31.36

(D) 12.8 / 0.4 = 32 (D’) 31.36 / 0.32 = 98
Diagrams C and C’ above show that when the divisor is a decimal fraction whose first
significant figure is in the tenth place, the last whole digit of the product is formed on the
first rod to the right of that of the multiplicand.
Diagrams D and D’, above show that when the divisor is a decimal fraction whose first
significant figure is in the tenth place, the last whole digit of the quotient is formed on the
first rod to the left of that of the dividend.
Example 6. (C) 32 x 0.04 = 1.28 (C’) 98 x 0.032 = 3.136

(D) 1.28 / 0.04 = 32 (D’) 3.136 / 0.032 = 98
Diagrams C and C’ above show that when the multiplier is a decimal fraction whose first
significant figure is in the hundredth place, the last whole digit of the product is formed on
the very rod on which that of the multiplicand is located.
Diagrams D and D’ above show that when the divisor is a decimal fraction whose first
significant figure is in the hundredth place, the last whole digit of the quotient is formed
on the very rod on which that of the dividend is located.
Example7. (C) 32 x 0.004 = 0.128 (C’) 98 x 0.0032 = 0.3136

(D) 0.128 / 0.004 = 32 (D’) 0.3136 / 0.0032 = 98
46
Diagrams C and C’ above show that when the multiplier is a decimal fraction whose first
significant figure is in the thousandth place, the last whole digit of the product is formed
on the first rod to the left of the last whole digit of the multiplicand.
Diagrams D and D’ show that when the divisor is a decimal fraction whose first
significant figure is in the thousandth place, the last whole digit of the quotient is formed
on the first rod to the right of the last whole digit of the dividend.
47
VIII. MENTAL CALCULATION

All abacus experts can calculate mentally with miraculous rapidity. On an average they
are twice as quick in mental calculation as on the abacus. It is possible for anyone to attain
astonishing rapidity in such mental calculation by proper practice. The secret lies in
applying abacus calculation to mental arithmetic by visualizing abacus manipulation.
Here are the vital points:
1. For example, in adding 24 to 76, close your eyes and visualize the beads of an abacus
set to 76. Then mentally add 24 onto the beads, aiding your visualization of the
abacus by flicking the index finger and thumb of your right hand as if really
calculating on an abacus.
2. When adding a series of numbers, say, 24 + 76 + 62 + 50, aid your memory by folding
one of your left fingers each time the sum has come up to 100.
3. At first, practice the addition of numbers of two or more digits which come up to a
round sum, for example, 76 + 24, or 222 + 555 + 223, and the like.
4. Remember, practicing a few minutes at a time for many days is worth more than
practicing hours on a single day.
48
EXERCISES
Constant daily practice is essential if one is to become proficient in the use of the
abacus. The following exercises, prepared and arranged in accordance with the most
up-to-date methods, have been kindly furnished by Professor Miyokichi Ban, an
outstanding abacus authority. They will provide a good beginning for the serious student,
who can then find more problems in any ordinary arithmetic book. Also note that problems
in multiplication and division may be used as problems in addition and subtraction
respectively.
The exercises are arranged so that a student can measure his progress against the
yardstick of the Japanese licensing system, the required standard of proficiency for the
particular grade being given at the beginning of each group. The possessor of a first,
second or third grade license, as awarded by the Abacus Committee, is officially qualified
for employment in a public corporation or business house. Licenses for the lower grades
are given on the basis of unofficial examinations conducted by numerous private abacus
schools.
The exercises are also chosen to give the maximum of variety to the problems, with
each digit from zero to nine receiving equal attention - an essential requirement for im-
provement in abacus operation. The system followed is that just initiated by the Central
Abacus Committee after long and careful research.
I - Sixth Grade Operator
Group A
(1 set per minute, or entire group with 70% accuracy in 10 minutes.)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
528 967 482 106 815 360 769 241 654 185
160 239 251 543 302 829 420 952 516 730
427 650 147 928 491 213 195 309 740 698
951 108 598 710 852 308 —513 756 495 246
719 243 120 954 —169 497 —854 487 536 809
452 758 —643 329 —401 932 508 360 —785 953
106 491 —839 267 —958 589 274 617 —320 721
843 536 304 695 576 690 421 508 —197 370
385 702 987 514 740 147 963 873 201 617
690 871 439 870 183 674 —307 420 873 164
724 460 -671 308 —235 850 —631 196 124 902
381 629 —305 796 —673 201 —175 689 —482 596
203 984 —526 632 340 765 286 204 —968 480
579 315 760 807 927 148 840 795 319 342
634 870 215 481 264 756 392 138 203 875
7 782 8 823 1 319 8 940 3 054 7 959 2 588 7 545 1 909 8 688
49
Group B
(1 set per minute or entire group with 70% accuracy in 10 minutes.)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
728 36 627 52 271 13 806 4 105 936 3 094

631 87 75 718 6 104 624 54 78 21 65
4 089 52 238 60 42 51 —716 29 6 084 413
50 705 —94 42 35 379 —81 712 —18 817
92 68 —426 8 096 487 785 439 67 —745 23
175 96 —81 621 93 41 5 021 834 —50 71
47 349 7 513 97 312 2 095 48 203 62 938
904 2 138 59 481 —78 36 27 40 329 26
72 510 60 305 —5 083 984 —78 526 —1 203 7 041
593 74 —3 041 574 20 63 —3 605 81 —459 86
61 421 -52 85 961 542 —953 9 036 48 259
86 907 139 19 —854 807 130 17 76 95
8 260 53 807 73 —69 1 068 69 395 597 508
354 619 40 9 038 —705 70 92 58 30 672
13 8 042 968 264 96 92 247 649 817 40
16 155 14 157 6 832 20 520 1 632 7 650 1 500 16 830 6 525 14 148
Group C Group D
(70% accuracy, 5 minutes.) (70% accuracy, 5 minutes.)
(1) 187 x 53 = 9 911 (1) 1 725 x 51 = 87 975
(2) 245 x 21 = 5 145 (2) 2 698 x 24 = 64 752
(3) 309 x 19 = 5 871 (3) 3 980 x 30 = 119 400
(4) 408 x 38 = 15 504 (4) 4 509 x 65 = 293 085
(5) 561 x 60 = 33 660 (5) 5 062 x 73 = 369 526
(6) 620 x 42 = 26 040 (6) 6 874 x 68 = 467 432
(7) 716 x 90 = 64 440 (7) 7 431 x 80 = 594 480
(8) 832 x 57 = 47 424 (8) 8 146 x 12 = 97 752
(9) 954 x 74 = 70 596 (9) 9 357 x 49 = 458 493
(10) 973 x 86 = 83 678 (10) 8 230 x 97 = 798 310
Group E Group F
(1) 960 / 24 = 40 (1) ¥ 986 / 34 = ¥ 29
(2) 810 / 45 = 18 (2) ¥ 855 / 19 = ¥ 45
(3) 7 505 / 79 = 95 (3) ¥ 7 280 / 80 = ¥ 91
(4) 6 640 / 80 = 83 (4) ¥ 6 240 / 78 = ¥ 80
(5) 5 920 / 16 = 37 (5) ¥ 5 092 / 67 = ¥ 76
(6) 4 080 / 68 = 60 (6) ¥ 4 128 / 96 = ¥ 43
(7) 3 127 / 53 = 59 (7) ¥ 390 / 13 = ¥ 130
(8) 2 160 / 30 = 72 (8) ¥ 2 320 / 40 = ¥ 58
(9) 1 152 / 72 = 16 (9) ¥ 1 550 / 25 = ¥ 62
(10) 2 184 / 91 = 24 (10) ¥ 884 / 52 = ¥ 17
50
II - Fifth Grade Operator
Group A
(70% accuracy, 10 minutes.)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
425 619 502 7 245 167 895 3 471 912 237 498
839 153 698 461 9 035 7 043 584 463 4 093 6 309
5 302 762 2 013 956 716 509 628 9 324 308 146
791 8 523 147 179 3 540 2 918 309 647 7 410 285
514 —478 9 684 317 295 451 5 672 1 509 —984 5 037
1 283 —694 5 726 420 138 734 491 854 —2 536 708
960 —7 081 409 8 096 869 1 086 —236 6 082 —841 3 572
2 048 377 971 —543 327 427 —7 018 165 965 219
683 1 049 3 056 —835 5 609 308 4 763 978 8 074 867
4 067 812 843 2 684 952 872 517 240 129 421
794 9 235 329 708 4 786 164 902 2 086 715 1 653
3 176 —504 760 1 032 128 6 217 8 059 731 —5 268 820
952 —6 380 4 215 —319 8 073 9 620 —126 3 108 —607 9 046
805 426 837 —908 409 563 —340 597 259 714
671 905 158 —6 527 241 35 —895 375 136 935
23 310 7 724 30 348 12 966 35 285 31 842 16 781 28 071 12 090 31 230
Group B
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
3 128 504 196 837 6 059 463 958 7 342 217 592
940 6 142 2 501 7 590 3 240 8 594 149 3 180 1 048 403
8 437 879 385 2 601 918 3 027 4 807 453 9 731 1 780
1 056 —5 263 6 210 9 082 136 —7 380 758 9 508 5 109 6 814
9 582 —198 1 479 —649 2 847 —2 759 3 146 4 261 8 925 509
865 7 920 8 937 —3 078 9 325 —902 6 291 937 —2 346 7 921
5 297 691 7 068 —429 1 706 6 178 530 8 679 —697 2 078
7 809 4 702 927 8 563 695 4 813 2 085 518 7 163 8 637
254 9 087 4 253 930 8 014 632 5 239 6 024 3 508 981
2 375 6 815 5 406 —358 4 587 9 264 7 621 2 187 4 386 3 146
718 3 960 8 045 —1 876 273 850 1 375 9 840 —6 029 4 065
6 042 —1 536 712 —6 145 7 208 1 096 8 407 5 706 —275 8 254
163 —8 473 9 634 7 201 462 —5 709 4 362 695 —5 314 672
4 301 —348 843 5 914 3 159 —641 9 024 1 039 430 5 890
1 649 2 057 3 521 4 726 5 341 8 175 613 726 852 9 763
52 116 26 939 60 117 34 909 53 970 25 701 55 365 61 095 26708 61 005
51
Group C
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
359 9 535 604 413 2 190 4 157 516 8 594 740 1 507
7 569 174 2 895 390 647 823 7 082 250 6 294 960
408 812 731 706 574 496 395 913 103 8 023
163 3 720 1 048 6 054 1 697 273 279 —341 857 649
914 —647 269 318 481 5 082 153 —7 082 5 019 752
792 —1 093 817 8 249 156 761 6 042 361 431 836
5 021 356 4 580 —634 3 078 —845 821 2 473 520 6 395
8 630 2 680 932 —267 423 —3 978 3 150 749 648 548
325 906 126 —9835 310 —109 264 605 3 976 125
206 185 658 142 962 634 4 987 128 837 409
127 —263 9 053 5 061 835 2 014 730 9 086 2 719 287
4 381 —8 472 316 975 5 289 361 674 —597 301 9 168
948 —598 470 —7 529 206 —580 968 —1 632 4 658 314
6 057 704 3 749 —807 4 058 —6 792 401 —805 285 7 431
874 419 527 182 739 905 5 893 467 961 270
36 774 8 418 26 775 3 418 21 645 3 202 32 355 13 169 28 349 37 674
Group D Group E
(70 % accuracy, 5 minutes.) (70% accuracy, 5 minutes.)
(1) 942 x 495 = 466 290 (1) 848 x 276 = 96 048

(2) 839 x 457 = 383 423 (2) 854 x 965 = 824 110
(3) 723 x 980 = 708 540 (3) 902 x 804 = 725 208
(4) 680 x 134 = 91 120 (4) 627 x 108 = 67 716
(5) 508 x 268 = 136 144 (5) 105 x 519 = 54 495
(6) 417 x 873 = 364 041 (6) 570 x 843 = 480 510
(7) 396 x 629 = 249 084 (7) 489 x 751 = 367 239
(8) 204 x 316 = 64 464 (8) 613 x 397 = 243 361
(9) 165 x 501 = 82 665 (9) 236 x 632 = 149 152
(10) 751 x 702 = 527 202 (10) 791 x 420 = 332 220
Group F Group G
(70% accuracy 5 minutes.) (70% accuracy 5 minutes.)
(1) 9 724 / 26 = 374 (1) 63 207 / 90 = 7 023

(2) 8 151 / 13 = 627 (2) 533 484 / 87 = 6 132
(3) 7 739 / 71 = 109 (3) 420 616 / 74 = 5 684
(4) 62 560 / 80 = 782 (4) 113 148 / 63 = 1 796
(5) 5 556 / 12 = 463 (5) 278 772 / 52 = 5 361
(6) 49 572 / 54 = 918 (6) 85 075 / 41 = 2 075
(7) 30 150 / 67 = 450 (7) 366 873 / 39 = 9 407
(8) 23 128 / 98 = 236 (8) 98 504 / 28 = 3 518
(9) 17 535 / 35 = 501 (9) 72 885 / 15 = 4 859
(10) 43 855 / 49 = 895 (10) 214 240 / 26 = 8 240
52
III - Fourth Grade Operator
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
6 374 4 561 3 458 9 526 1 459 7 201 2 951 5 482 8 954 6 753
5 021 9 753 2 983 4 198 3 146 6 759 9 160 8 035 4 710 1 832
7 913 3 670 4 120 8 973 4 723 1 093 3 294 4 973 3 986 8 094
9 265 —1 256 6 309 7 269 2 368 2 346 8 643 5 106 5 603 2 869
4 537 —5 904 5 092 9 085 6 912 8 712 —4 712 3 683 9 215 7 150
5 084 8 329 2 148 —6 450 5 279 6 507 —1 035 1 290 5 087 3 908
8 762 2 048 5 871 —8 317 2 905 3 874 —5 368 9 624 1 693 9 201
7 190 6 827 1 397 —3 802 7 816 4 158 7 214 2 067 6 854 4 127
3 856 —7 895 7 539 1 236 3 047 8 527 6 870 6 541 2 401 9 382
4 280 —3 047 2 710 7 084 8 102 4 096 5 907 3 712 —6 049 5 240
8 152 —2 739 6 087 6 735 5 680 3 648 —7 582 6 154 —8 731 8 476
2 409 6 180 8 264 —4 391 9 834 5 930 —4 326 7 298 —3 278 6 395
6 371 5 912 4 675 —5 140 5 091 2 489 8 059 4 870 7 142 7 564
1 948 4 106 9 561 2 604 6 470 1 360 9 781 7 951 9 320 5 013
9 603 1 438 3 406 5 172 7 538 9 125 6 403 8 309 7 562 4 671
90 765 31 983 73 620 33 782 80 370 75 825 45 259 85 095 54 469 90 675
Group B
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
2 453 5 906 3 629 8 431 9 504 6 712 72 934 1 846 4 165 8 679
5 192 98 710 41 568 7 846 1 029 8 950 1 653 30 782 1 594 15 032
67 941 4 825 9 205 20 753 85 934 4 167 3 817 4 935 3 872 6 985
36 029 5 174 80 453 4 501 2 318 60 843 9 051 8 071 51 943 2 803
1 683 9 631 —2 687 59 327 —7 840 6 679 1 329 6 329 92 381 41 697
4 808 3 712 —31 405 7 698 —43 126 1 703 —53 682 95 874 —6 705 9 306
8 574 20 893 8 716 3 086 —2 735 29 078 —8 306 3 105 —4 372 5 821
3 129 8 074 4 932 64 279 3 052 8 634 —2 148 5 763 —87 156 4 068
48 531 2 689 —17 043 1 058 9 607 39 581 80 597 78 491 9 034 37 459
4 067 60 243 —5 890 6 912 17 269 7 824 5 431 40 982 13 507 2 516
73 215 87 069 —9 134 2 769 8 573 2 307 14 970 2 417 5 420 70 128
2 750 1 574 4 716 90 635 —54 618 40 592 —7 026 7 650 7 263 8 749
10 896 6 458 76 352 5 804 —6 481 91 486 —25 469 63 208 —20 489 3 290
7 208 53 961 2 879 8 140 70 396 6 251 4 205 9 016 —8 016 9 347
9 645 2 307 5 021 1 972 4 125 8 095 6 748 2 569 2 698 60 175
285 621 371 226 171 312 323 211 97 007 316 902 104 104 361 038 65 139 286 055
53
Group C
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
2 701 58 976 378 683 8 472 4 583 75 219 839 13 597 9 218
342 4 109 917 50 746 601 13 659 327 523 763 40 137
60 179 249 4 586 —1 597 7 340 732 139 19 768 27 458 572
9 630 80 651 1 469 —265 45 981 403 —6 875 5 017 89 162 1 065
71 524 1 590 70 854 619 790 29 581 —906 240 430 56 387
8 495 —468 3 042 7 301 13 267 690 50 864 6 754 2 049 941
50 723 —5 107 793 25 478 6 154 32 071 8 542 30 482 —98 625 259
268 —67 328 2 138 9 032 24 896 947 41 038 1 937 —506 7 308
956 782 96 205 —359 974 8 765 —751 684 —4 317 32 596
814 8 245 7 621 —40 168 32 058 1 296 —60 213 93 521 5 231 483
5 209 92 361 5 839 —8 274 613 95 820 —2 496 8 075 70 318 60 725
43 167 873 19 057 63 921 50 789 7 415 9 547 42 816 —2 679 864
1 083 70 934 415 34 580 9 523 178 37 081 7 109 —985 24 139
548 —3 012 80 246 904 862 80 264 320 24 365 6 140 8 074
86 937 —654 68 120 2 817 1 035 6 304 4 693 690 804 1 690
342 576 237 201 361 180 145 418 203 355 282 708 156 524 242 820 108 840 244 458
Group D Group E
(1) 92 854 x 84 = 7 799 736 (1) 1 375 x 562 = 772 750

(2) 86 213 x 59 = 5 086 567 (2) 2 610 x 148 = 386 280
(3) 73 041 x 90 = 6 573 690 (3) 3 784 x 625 = 2 365 000
(4) 60 378 x 16 = 966 048 (4) 4 208 x 201 = 845 808
(5) 51 762 x 27 = 1 397 574 (5) 5 429 x 874 = 4 744 946
(6) 47 609 x 70 = 3 382 630 (6) 6 057 x 903 = 5 469 471
(7) 30 427 x 32 = 973 664 (7) 7 906 x 417 = 3 296 802
(8) 29 185 x 63 = 1 838 655 (8) 8 591 x 730 = 6 271 430
(9) 18 596 x 45 = 836 820 (9) 9 832 x 986 = 9 694 352
(10) 54 930 x 81 = 4 449 330 (10) 4 163 x 359 = 1 494 517
Group F Group G
(1) 9 108 x 379 = 3 451 932 (1) 2 647 x 3 740 = 9 899 780
(2) 8 240 x 568 = 4 680 320 (2) 3 068 x 2 698 = 8 277 464
(3) 7 894 x 740 = 5 841 560 (3) 9 854 x 7 219 = 71 136 026
(4) 6 372 x 953 = 6 072 516 (4) 1 370 x 4 805 = 6 582 850
(5) 5 423 x 182 = 986 986 (5) 8 401 x 6 457 = 54 245 257
(6) 4 617 x 194 = 895 698 (6) 4 936 x 9 523 = 47 005 528
(7) 3 581 x 807 = 2 889 867 (7) 6 125 x 5 184 = 31 752 000
(8) 2 056 x 625 = 1 285 000 (8) 2 519 x 8 306 = 20 922 814
(9) 1 905 x 401 = 763 905 (9) 7 093 x 1 962 = 13 916 466
(10) 3 769 x 236 = 889 484 (10) 5 782 x 3 071 = 17 756 522
54
Group H Group I
(70% accuracy 5 minutes.) (70% accuracy 5 minutes)
(1) 379 428 / 42 = 9 034 (1) 94 235 / 401 = 235

(2) 706 860 / 85 = 8 316 (2) 87 040 / 256 = 340
(3) 235 662 / 31 = 7 602 (3) 752 128 / 832 = 904
(4) 658 145 / 97 = 6 785 (4) 64 220 / 380 = 169
(5) 406 164 / 68 = 5 973 (5) 548 784 / 927 = 592
(6) 87 362 / 19 = 4 598 (6) 431 748 / 603 = 716
(7) 115 710 / 30 = 3 857 (7) 385 746 / 478 = 807
(8) 55 637 / 23 = 2 419 (8) 219 537 / 519 = 423
(9) 94 240 / 76 = 1 240 (9) 107 912 / 164 = 658
(10) 325 134 / 54 = 6 021 (10) 620 895 / 795 = 781
Group J Group K
(1) 636 768 / 792 = 804 (1) 553 149 / 6 829 = 81

(2) 35 984 / 208 = 173 (2) 273 375 / 3 645 = 75
(3) 194 394 / 537 = 362 (3) 97 351 / 1 453 = 67
(4) 403 425 / 815 = 495 (4) 761 664 / 7 934 = 96
(5) 86 163 / 373 = 231 (5) 81 018 / 4 501 = 18
(6) 531 340 / 620 = 857 (6) 84 942 / 2 178 = 39
(7) 932 832 / 984 = 948 (7) 424 901 / 8 017 = 53
(8) 286 090 / 469 = 610 (9) 120 048 / 5 002 = 24
(9) 363 726 / 501 = 726 (9) 611 640 / 6 796 = 90
(10) 74 314 / 146 = 509 (10) 394 044 / 9 382 = 42
IV - Third Grade Operators
Group A
(1) (2) (3) (4) (5)
8 127 526 105 942 41 306 28 640

659 4 192 835 7 962 135
17 492 60 271 94 516 95 641 86 029
961 037 358 604 62 481 529 401 286
5 208 —963 83 672 890 375 514
638 125 —71 850 1 450 —6 813 37 269
80 734 —5 397 238 107 —380 276 2 478
9 270 409 715 396 784 903 851
25 816 842 740 138 978 250 18 394
401 369 —17 438 253 3 795 549 076
756 —732 609 57 048 12 047 3 702
36 594 90 386 6 729 604 518 153
520 943 6 127 78 915 —421 695 718
481 28 459 406 329 —53 608 4 367
74 308 813 045 9 067 —29 134 70 925
2 780 919 943 910 1 885 878 2 164 955 2 802 537
55
(6) (7) (8) (9) (10)
350 624 79 328 60 382 952 14 538

93 041 8 653 847 38 207 751
—68 729 184 705 2 415 706 394 26 374
—134 31 894 531 2 460 652 096
791 560 936 891 270 875 8 725
45 287 507 269 63 092 410 936 479 163
2 759 421 —726 9 283 680
59 641 60 584 —308 619 65 148 13 849
—6 358 245 139 —15 974 931 054 804 975
—804 293 372 704 368 729 5 812
—15 807 4 715 —3 051 54 071 30 427
983 693 807 —57 249 13 567 209
7 410 16 042 425 698 6 182 90 412
362 2 610 9 703 207 648 587 936
201 876 80 597 56 184 89 513 3 160
658 222 1 917 072 1 828 871 2 537 019 2 719 107
Group B
(70% accuracy 10 minutes. Calculate problems 1-10
to the nearest thousandth; 11-20 to the nearest dollar.)
(1) 4 097 x 238 = 975 086 (11) $ 2 594 x 376 = $ 975 344
(2) 5 638 x 149 = 840 062 (12) $ 4 608 x 0.189 = $ 871
(3) 14 902 x 52 = 774 904 (13) $ 7 832 x 897 = $7 025 304
(4) 7 105 x 0.098 = 696.29 (14) $ 94 120 x 6.4 = $ 602 368
(5) 9 674 x 603 = 5 833 422 (15) $ 8 029 x 738 = $5 925 402
(6) 63.25 x 7.64 = 483.23 (16) $ 975 x 45.12 = $ 43 992
(7) 853 x 4.017 = 3 426 501 (17) $ 5 176 x 0.625 = $ 3 235
(8) 0.3081 x 0.926 = 0.285 (18) $ 3 061 x 903 = $2 764 083
(9) 2 984 x 351 = 1 047 384 (19) $ 6 843 x 201 = $1 375 443
(10) 0.2176 x 87.5 = 19.04 (20) $ 6 549 x 643 = $4 211 007
Group C
(70% accuracy 10 minutes. Calculate problems 1—10
to the nearest thousandth; 11—20 to the nearest dollar.)
(1) 937 015 / 965 = 971 (11) $ 99 / 0.368 = $ 269

(2) 0.08988 / 6.42 = 0.014 (12) $ 83 619 / 27 = $ 3 097
(3) 0.070654 / 0.136 = 0.520 (13) $ 71 967 / 149 = $ 483
(4) 63 366 / 708 = 89.5 (14) $ 649 612 / 7 061 = $ 92
(5) 55.426 / 214 = 0.259 (15) $ 560 / 0.875 = $ 640
(6) 415 473 / 591 = 703 (16) $ 415 693 / 593 = $ 701
(7) 315.333 / 45.9 = 6.87 (17) $ 33 154 / 60.5 = $ 548
(8) 280 932 / 82 = 3 426 (18) $ 2 485 / 2.84 = $ 875
(9) 17.316 / 0.037 = 468 (19) $ 122 563 / 901 = $ 136
(10) 241 893 / 7 803 = 31 (20) $ 54 / 0.432 = $ 125
56
Advanced Abacus Japanese Theory and Practice, by Takashi Kojima
ADVANCED ABACUS
Japanese
Theory and Practice
by
TAKASHI KOJIMA
CHARLES E. TUTTLE COMPANY

TOKYO - JAPAN
First edition, 1963
1
TABLE OF CONTENTS
Foreword.......…………................................................... 2
Author’s preface......……………………………............................. 3
I. Further remarks on abacus history……...........................… 5
II. Negative answers from subtraction……….........................… 12
III. Other methods of multiplication…….……..........................… 17
IV. Other methods of division..………………..........................…… 32
V. More about decimals................................................... 40
VI. Calculations involving more than one unit of measurement…. 43
VII. Extracting square roots................................................ 53
VIII. More exercises…………………………………................................. 59
FOREWORD
Mr. Kojima’s second book on the abacus gives important information on the further
practical use of the abacus and on the principies of its use in business. I believe that his
complete explanation of operational methods and their theoretical basis will be of
especial help to those foreign students who have no guide or instructor except books.
Aside from its immense utility in business and everyday calculation, the abacus is a
far more effective instrument for teaching arithmetic in blind schools than is braille.
Moreover, if introduced into ordinary schools, it will prove an excellent time-saver in
arithmetic instruction. Half of the problems in arithmetic textbooks are calculation
problems and the other half can be reduced to calculation problems by some
mathematical reasoning. Consequently, those arithmetic hours allotted for the teaching
of abacus operation, by improving the mental arithmetic of students, will enable them to
calculate much faster than with pencil and paper, thus creating additional time for a
more advanced study of arithmetic.
As Chairman of the Committee of the International Association of Abacus Operators

of the Japan Chamber of Commerce and Industry, I have been most pleased to assist Mr.
Kojima by making available the findings of recent teclinical and theoretical studies and
by revising his manuscript in the light of all the latest information.
Yoemon Yamazaki
Professor of Economics, Nihon University
Vice-President, All-Japan Federation of Abacus Operators
Chairman, Committee of the Int. Assn. of Abacus Operators
2
AUTHOR’S PREFACE
This book has been written as a sequel to my earlier work The Japanese Abacus: Its
Use and Theory. In this volume I shall both expand the explanation of sorne of the basic
abacus operations given in the first volume and discuss new operations and new ways of
doing the basic operations.
In the first chapter I shall discuss the Oriental history of the abacus in greater
detail than was done in my first book. In the second chapter I shall deal with the new
prob1em of negative numbers or negative answers resulting from the subtraction of a
larger number from a smaller one. The third and fourth chapters will respectively
consider other methods of multiplication and division than the ones explained in my first
book. Chapter five is an expanded discussion of decimals in multiplication and division.
Chapter six is a practical discussion of how to handle calculations involving more than
one unit of measurement (sud as inches and feet or pounds and ounces). Chapter seven
introduces a method of extracting square roots on the abacus. Finally, chapter eight
provides more exercises for the reader to practice on. Through it he will be able to
measure his abacus ability by taking actual examinations given to Japanese applicants for
proficiency grades eight to one.
I would like to include here sorne noteworthy statistics concerning the recent
license examinations. Of the successful examinees for the third-grade license, about 70%
were between thirteen and eighteen years old, 5% were nineteen and over, and 25 %
were twelve and under (including 0.3 % who were nine and under). Of the successful
examinees for the second-grade license, about 91 % were between thirteen and eighteen,
2 % were nineteen and over, and 7% were between ten and twelve. Of the successful
examinees for the first-grade license, about 87% were between thirteen aud eighteen,
12% were nineteen and over, and 1 % were under twelve. These figures do not completely
teil the story because among those who pass the first-grade examination every year are
some who have already passed it, but either want the practice or a higher score. This
explains why the percentage of persons nineteen and over who passed the first-grade
examination is larger than the percentage of the same group who passed the second-
grade examination.
An interesting conclusion can be drawn from these statistics—it is rather difficult

for persons over nineteen and under twelve to pass the first- and second-grade
examinations.
The ratio of boys to girls who passed these examinations is also worthy of mention.
Of the successful examinees for the third- and second-grade licenses, about 60% were
girls and 40% were boys, while the reverse was true of the firstgrade examination.
In Japan the abacus is definitely a practical skill. It has found its way into the
curriculum of all japanese grade schools as a fundamental part of arithmetic. Many senior
commercial high schools require all students to pass at least the third-grade examination.
There have also been many abacus schools established to meet the needs of those
preparing to go into business. And, as I hope I am demonstrating in these two books,
there is a good reason why the abacus can be found in practically every Japanese
household.
3
I am greatly indebted to several authorities who kindly furnished me with valuable

information and suggestions. My most grateful acknowledgments are due to Professor
Yoemon Yamazaki of Nihon University. He is the ViceChairman of the Abacus Research
Institute, and Advisor to the Central Committee of the Federation of Abacus Workers
(hereafter referred to simply as the Abacus Committee). These organizations, being
under the sponsorship of the Japan Chamber of Commerce and Industry, are far and away
the largest and most important of all abacus organizations in Japan.
I also extend my most sincere gratitude to Professor Miyokichi Ban, of the above-
mentioned Abacus Committee, who was kind enough not only to furnish this book with a
great many exercises especially prepared and arranged for the sake of the foreign
student but also to read the book in proof and give me many valuable suggestions.
I also must express my sincere thanks to Mr. Shinji Ishikawa, President of the Japan
Association of Abacus Calculation, who spared himself no trouble in reading the
manuscript and the proof and furnishing much invaluable up-to-date information.
Grateful acknowledgments are also due to Mr. Hisao Suzuki for his information on
the history of the abacus and to Mr. Zenji Arai for valuable suggestions on the uses of the
abacus.
I also wish to express my sincere thanks to Mr. Yataro Nagata, Chief of the Abacus
Operators’ License Examinations Section in the Japan Chamber of Commerce and Industry,
for information on the national examinations for abacus operators’ licenses and for
permission to reprint in this book the problems presented in the 1959 National License
examinations.
Last, but not least, I must thank Mr. William R. Whitney and the editorial staff of
the Charles E. Tuttle Company for their valuable suggestions and improvements in both
the manuscript and the proof stages.
TAKASHI KOJIMA
4
I – FURTHER REMARKS ON ABACUS HISTORY

The ancient Chinese books on mathematics which have been preserved furnish
hardly any information on the abacus. Accordingly, nothing definite is known about its
origin. The only reliable account of the origin of the Oriental abacus is in a book entitled
Mathematical Treatises by the Ancients compiled by Hsu Yo toward the close of the Later
Han dynasty (A.D. 25—220) at the beginning of the third century and annotated by Chen
Luan in the sixth century. This book gives sorne information about various reckoning
devices of those days and was one of the Ten Books on Mathematics (Suan-hwei-shi-chu)
which were included among the textbooks to be read for government service
examinations in China and Japan for many centuries.
Chen Luan in his note gives the following description of the calculating device:
“The abacus is divided into three

sections. In the uppermost and lowest
section, idle counters are kept. In the middle
section designating the places of numbers,
calculation is performed. Each column in the
middle section may have five counters, one
uppermost five-unit counter and four
differently colored one-unit counters.”
The above figure represents the abacus as pictured in accordance with the
foregoing description. The board represents the number 37 295.
The extent to which the counting board was used may be toid by Hsu Yo’s poetical
description of the board. The verse, which is highly figurative and difficult to decipher,
may read: “It controls the four seasons, and coordinates the three orders, heaven, earth,
and man.” This means that it was used in astronomical or calendar calculations, in
geodetic surveys, and in calculations concerning human affairs.
The reader will notice a close similarity between this original Oriental abacus and
the Roman grooved abacus, except for the difference that counters were laid down in the
former while they were moved along the grooves in the latter. Because of this and other
evidence, many leading Japanese historians of mathematics and the abacus have
advanced the theory that the above-mentioned prototype of the abacus was the result of
the introduction into the East of the Roman grooved abacus.
The following corroborative pieces of evidence in favor of this theory are cited in
the latest works by Prof. Yoemon Yamazaki and Prof. Hisao Suzuki of Nihon University.
(1) The original Chinese abacus has a striking resemblance in construction to the
Roman grooved abacus, as is evident in the foregoing quotation from Hsu Yo’s book, e.g.,
four one-unit counters and one five-unit counter in each column.
(2) The method of operation of the ancient Chinese abacus was remarkably similar
to the ancient Roman method.
5
In ancient China, multiplication and division were performed by the repetition of

addition and subtraction:
MULTIPLICATION:
Procedure A: 23 x 5 = (23 x 2) + (23 x 2) + 23 = 115 (Ans.)
Procedure B: 23 x 5 = 23 + 23 + 23 + 23 + 23 = 115 (Ans.)
DIVISION:
Procedure A: 115 + 23 = 115 — 23 — 46 — 46 = 0 (Ans.5)
Procedure B: 115 + 23 = 115 — 23 — 23 — 23 — 23 — 23 = 0 (Ans. 5)
In the case of multiplication, each time 23 or 46 was added, 1 or 2 was added to

the factor on the left of the board. In the case of division, each time 23 or 46 was
subtracted, 1 or 2 was added to the quotient on the left of the board. It is obvious that
anyone could easily learn and perform these simple primitive operations.
(3) Traces of reckoning by 5’s may be found in the Chinese pictorial representation
of reckoning-block calculation as in the Roman numerals, as:
six: VI (5 + 1) seven: VII (5 + 2)

eight: VIII (5 +3) four: IV (5 — 1)
(4) Trade was carried on between China and Rome. Chinese historical documents
written in the Han dynasty (206 B.C.-A.D. 220) furnish descriptions of two land routes,
called silk roads, connecting the two great empires.
Inasmuch as even in olden days valuable products or devices made in one country
were transmitted to others with astonishing rapidity, the above facts may well
substantiate this theory.
Among the dozen other reckoning devices mentioned in this book are the reckoning
boards pictured below. These boards are presumed to date back to the days of the Chou
dynasty, which ended in 249 B.C.
(The number on the board is 23 957)
(When yellow counters were used, the squares in

each column represented 1, 2, 3, and 4 respectively.
When blue ones were used, they represented 5, 6, 7, 8,
and 9 respectively. The black balls in the figure stand for
blue counters. The number on the board represents
3581.)
6
These and other reckoning devices are believed to have gone out of use as the
previously mentioned abacus developed and gained popularity.
Now two questions present themselves. One is why this ancient abacus developed
in the East to become such an efficient calculating machine. The other is why this de-
velopment did not also take place in the West.
The reasons lay perhaps in the systems of calculation and the numerical
nomenclature which were used in the East and West. They differ significantly. In ancient
China and Japan numbers were named, written, and set on a calculating board from left
to right, from the highest denomination to the lowest. Thus the introduction of the
abacus to China provided the Chinese with an ideal tool in terms of their method of
naming and using numbers. This compatibility and normal inventiveness caused the
primitive abacus to be developed into its modern form during the long development of
the Chinese civilization.
The chief calculating devices which are known to have been used in China from
before 1 000 B.C. to the days when the abacus came into wide use are reckoning blocks
called ch’eou in China and sangi in Japan and slender bamboo sticks called chanchu in
China and zeichiku in Japan. The former device continued to be used in the East for
calculation until not many years ago, and the latter device, which was more awkward,
was largely replaced by the former for calculating purposes and is presently used only by
fortunetellers for purposes of divination.
Until the introduction of Western mathematics, mathematicians in China and

Japan utilized reckoning-block calculation, which had not only been developed to the
point of performing basic arithmetic operations but was also used to solve quadratic,
cubic, and even simultaneous equations. It is presumed that they did not think it worth
while to concern themselves with the other reckoning devices, including the abacus,
which was, in their eyes, an inferior calculator barely capable of performing
multiplication and division by means of the primitive cumulative method of addition and
subtraction. Probably another reason which alienated mathematicians from these
reckoning devices was that these instruments gave only the result of calculation, and
were incapable of showing either the process of calculation or the original problem.
In ancient times China was primarily a nomadic and agricultural country, and
business in those days liad little need of instruments of rapid calculation. Anyway a
millennium after the Han dynasty there was no record of the abacus. During the dozen
centuries beginning with its first mention in the Han dynasty until its development, this
primitive calculator remained in the background.
However, with the gradual rise of commerce and industry, the need for rapid
calculation grew. The modern, highly efficient abacus, which probably appeared late in
the Sung dynasty (906—1279), came into common use in the fourteenth century. The
great rise and prosperity of free commerce and industry during the Ming dynasty (1368 -
1636) are presumed to have promoted the use and development of the abacus. A number
of books on mathematics brought out in those days give descriptions of the modern
Chinese abacus and give accounts of the modern methods of abacus operation, including
those of multiplication and division.
7
Bamboo, indigenous in the East, has furnished an abundant source of ideal

material for an efficient and inexpensive abacus. Since the Ming period, on account of its
remarkable efficiency, low price, and handiness, the abacus has been the favorite
instrument of calculation in the East.
The Chinese abacus of the Ming period had two five-unit counters and five one-unit
ones on each rod. The primitive abacus was changed into the present Chinese form to
suit the convenience of figuring up the Chinese weights not based on the decimal system.
The weights were also important for conversion of currency. Another cogent reason why
the Chinese abacus has two five-unit counters on each rod is that a rod with two five-unit
counters is more convenient to abacus operation by means of the Chinese method of
multiplication and also by means of the older method of division which uses a special
division table.
In Europe, the line abacus or counting board appeared first in France about the
beginning of the thirteenth century and rapidly became popular. From the fourteenth to
the seventeenth century the practice of this manual arithmetic was universal in business
and in households, as well as in the departments of government. Its immense popularity
may well be illustrated by the following pleasantly expressed stanza attributed to
Brébeuf as it is quoted in Francis Pierrepont Barnard’s Casting Counter and the Counting-
Board.
Les courtisans sont des jetons;

Leur valeur dépend de leur place;
Dans la faveur, des millions;
Et des zéros dans la disgrâce.
The same book also quotes the phrase, “Faux comme un jeton,” which arose from
the practice of gilding or plating jettons and passing them as money, or creating a
deceptive impression.
The number 2 376 would be expressed by jettons on the

line abacus or counting board as in Fig. 4.
(Each line upwards is ten times the value of that below it.
Each space is five times as much as the line next below it. In
addition, the process began at the units, and in subtraction at the
higher digits.)
However, in Europe the une abacus failed to develop into the efficient rod abacus,
and gradually gave way to the cipher system of greater efficiency, until it was given the
coup de grâce by the French Revolution, which enforced the nation-wide ciphering
system. One of the major causes for this result is presumed to be found in the fact that
before the introduction of Arabic numerals European countries used diverse systems of
numerical notation-duodecimal, binary, sexagesimal, etc. The division of daytime into
twelve hours and that of one hour into sixty minutes, etc. may be mentioned as vestiges
of these numerical systems. The rod abacus can never be worked with efficiency on these
numerical scales. Another remote cause may be traced back to the way in which the
Arabs, who introduced the cipher system into Europe, named their numbers. The Semites,
including the Arabs, named their numbers beginning at the units, although they wrote
8
from right to left. Thus for instance, in Arabic, one hundred and twenty-five was called
five and twenty and one hundred, the result appearing as 125, as Prof. Cargill G. Knott
says in his treatise on the abacus. This is believed to be the primary reason why the Arabs,
who achieved remarkable development in mathematics in the medieval ages, made use
of their Arabic numerals without recourse to the less efficient lime abacus or other
calculating devices. Nor could the rod abacus have been used with efficiency by a race
which used such a system of naming their numbers. The early Indians, who are credited
with the invention of the cipher, spoke like the Arabs although they wrote from left to
right. The Chinese named their numbers beginning with the largest denomination,
although they wrote from top to bottom, proceeding from right to left.
Now the second question before us is: What causes prevented the adoption of a
cipher system in China and Japan? The Chinese and the Japanese write in vertical
columns from top to bottom, while the cipher system is worked from left to right.
However, this is not considered the primary cause, for in the remote past coeval with the
origination of Chinese characters, the Chinese carried out their calculation by means of
reckoning blocks working left to right. However, this reckoning-block calculation was
cumbersome and was no more fit for rapid operation than the Western line abacus.
A couple of examples of the arrangement of reckoning blocks are given below. The
numbers 123 and 5 078 are represented:
In the units, hundreds, and other odd places, the numbers up to five are each
represented by the corresponding number of vertical strokes, and the numbers from six
to nine are each represented by the addition of the requisite number of strokes below a
five-unit horizontal line. In the tens, thousands, and other even places, the numbers up
to five are each represented by the corresponding number of horizontal strokes, and the
numbers from six to fine are each represented by the addition of the requisite number of
strokes aboye a five-unit horizontal line.
The Chinese numerical notation, which was probably the pictorial representation
of reckoning-block calculation, was of far less practical use in calculation than reckoning
blocks. Accordingly, mathematical calculation was generally performed with reckoning
blocks and later also with the abacus. In remote antiquity, probably more than 2 000
years back, reckoning blocks were arranged differently for calculation. In those days the
numbers in units and hundreds places were represented by horizontal blocks instead of
vertical ones, and numbers in tens and thousands places were represented by vertical
blocks. Thus the Chinese numerals, 一 (1), 二(2), 三 (3), and 百 (100), comprised of
horizontal strokes, are pictographs, representing horizontally arranged blocks (- = ≡), and
the numerals, 十(10), 廿(20), 卅(30), and 千 (1 000), comprised of vertical strokes, are
pictographic imitations of blocks arranged vertically.
What caused the change in the arrangement of reckoning blocks is a knotty

problem, to which no satisfactory solution has been offered. However, some scholars
conjecture that because of the great importance of divination in early China the
9
arrangement of reckoning blocks might probably have been influenced by the method of
arranging reckoning sticks for divining purposes.
Probably the major cause which prevented the replacement of the cumbersome
numerical notation by the cipher system was the development of the abacus, which could
meet everyday public needs in business and household calculation. Reckoning-block
calculation, which was applied to the primitive abacus, made a remarkable development,
and by the Ming dynasty the abacus had become a far more efficient computer than the
cipher system.
In the days of feudal government, learning was mostly of classics and was the
exclusive heritage of certain officials and limited circles of scholars. Mathematics was
studied only by the few who were initiated into this mystery of learning, and many of
them formed exclusive esoteric sects of hereditary transmission to preserve their
patrimonial positions or living. Under these social conditions it was none of their concern
to teach or popularize their secrets of mathematics. The enlightened scholars who were
favored with exceptional opportunities to study Western science and mathematics may
have been aware of the superiority of Arabic numerals to the cumbersome Oriental
numerical notation. But these intellectuals must have been too few and far between and
their outcry to initiate the reform too feeble to arouse public attention.
Among the other important causes may be mentioned the want of free
international trade and communication, the virtual isolation of Eastern countries from
the West, and the consequent lack of understanding of international situations and
national prejudice against foreign culture, and among the rest, the conservatism of
human nature. The Chinese officialdom was so prominently conservative that it would
firmly have resisted any attempts at such reforms or improvements in the hoary customs
or timehonored classics of national veneration, many of which had been included among
textbooks for government service examinations during the long Chinese historical period
extending over twenty centuries.
In Japan it was not until several years after the 1868 political revolution, which
overthrew the shogunate (government by the supreme feudal ruler), that the progressive
modem government, awakened to the progress of the world, enacted the compulsory
education law, including in the curriculum the cipher system, without which the effective
teaching of modern mathematics to the public is impossible.
Now the Japanese word for abacus, soro ban, is probably the Japanese rendering
of the Chinese suan-pan, (soo-pan in the southern dialect or sur-pan in Manchuria). The
soroban in Japan did not come into common use until the seventeenth century. However,
the historical fact that beginning with the seventh century, there were at times as many
as 2 000 Japanese students studying at the then Chinese capital in Chang-an, now called
Si-an, furnishes us with reliable evidence that the abacus was introduced into Japan at a
far earlier date, although the oldest documentary evidence of the Japanese abacus does
not date further back than the sixteenth century.
In any case, once this convenient instrument of calculation gained popularity in

Japan, it was studied extensively and intensively by many mathematicians including Seki
Kowa (1640—1709), who discovered a native calculus independent of the Newtonian
theory. As a result, the form and methods of operating the abacus have undergone one
improvement after another. For a long time in Japan two kinds of abacus were used
10
concurrently until the 1868 political revolution: the Chinese-style one with two five-unit
counters and five one-unit counters and the older Japanese-style one with one five-unit
counter and five oneunit counters. After the time of the revolution, the Chinese-style
abacus went completely out of use. Finally since around 1940, the older-style Japanese
abacus has largely been replaced by the present more advanced and efficient one with
one five-unit counter and four one-unit counters.
BIBLIOGRAPHY
Barnard, Francis Pierrepont: The Casting-Counter and the Counting-Board,

Oxford Press.
Comprehensive Dictionary of Abacus Calculation, A. Akatsuki Publishing
Company, Tokyo (in Japanese).
Ishikawa, Shinji: Abacus Calculation: Its Theory and Technique, Abacus
Research Society, Tokyo (in Japanese).
Knott, Cargill G.: “The Abacus in Its Historic and Scientific Aspects,”
Transactions of The Asiatic Society of Japan, Vol. XIV, 1886.
Mikami, Yoshio: “The Characteristics of Chinese Mathematics,” Journal of
Oriental Studies (Toyo Gakuho), Vol. XV, No. 4, 1926 (in Japanese).
Yamazaki, Yoemon: A Collection of Eastern and Western Literature on the
Abacus, two volumes, Morikita Publishing Company, Tokyo (in Japanese).
Yamazaki, Yoemon: The Origin of the Chinese Abacus.
Yamazaki, Yoemon; Suzuki, Hisao; and Toyo, Sei-ichi: A History of Abacus
Calculation, Morikita Publishing Company, Tokyo (in Japanese).
11
II - NEGATIVE ANSWERS FROM SUBTRACTION

Subtraction of larger numbers from smaller ones is performed by means of
complementary numbers.
For this purpose it is necessary to be able to read

complementary numbers on the board at a glance.
For example, the complementary digit of 6 with respect to

10 is 4. However, the board shows 3 in the position of the
complementary digits (Fig. 5).
The complementary numbers of 23 and 457 are 77 and 543 respectively. However, on
the board they appear as 76 and 542 (Figs. 6 and 7). Therefore, on the board, the true
complementary number can be obtained by adding one to the complementary digit on
the last rod.
The following problem involving complements can be solved without making any
mental or written calculations.
PROBLEM: A customer made a purchase of 7 dollars 68 cents and gave a clerk a ten-
dollar bill. How much change should the customer receive?
If the clerk simply sets 768 on the abacus, the answer, 2 dollars 32 cents,
will naturally appear on the board of its own accord in the form of the
complementary number (Fig. 8).
Here are sorne examples of abacus calculation by means of complementary digits.
EXAMPLE 1: 2 - 9 = -7
STEP 1: Set 2 on B (Fig. 9).
STEP 2: Since you cannot subtract 9

from 2, you must borrow 10 from the tens
rod A, and subtract 9 from 12. This gives
you 3 on B. However, since you borrowed
10 previously, you must return it. In other words, as 10 was added to the 3 on rod B, you
must subtract 10 from it. You can do this very simply by setting minus 10 on rod B in your
mind. Then the difference between the 3 and the minus 10 on rod B, i.e. minus 7, will
mechanically appear in the form of the complementary number. This is the answer (Fig.
10). Note that the advantage of the abacus operation is to work out and change the
difference between 2 and minus 9, i.e. minus 7, into that between 3 and minus 10 and to
show the difference in the clearer and more obvious form.
NOTES: (a) In subtracting 9 from 2, do not make themental calculation of subtracting 9

from 12 but add, to the 2 on B, the complementary number 1, with respect to 10, with
the idea that you are subtracting 9 from 10. Nor should you take the trouble of setting 1
on rod A except for practice. (b) In calculation by complementary numbers, it is
important for you to remember that the counters or beads which have been moved next
12
to the beam always indicate positive numbers, while their complements represent
negative numbers.
EXAMPLE 2: 15 - 84 = -69
STEP 1: Set 15 on BC (Fig. 11).
STEP 2: Borrow 100 from the hundreds rod A, and subtract 84

from 115.
This gives you 31 on BC, and the answer 69 mechanically appears on the board in the
form of the complementary number of 31 for 100 (Fig. 12).
NOTE: The above process can be analyzed as follows:
EXAMPLE 3: 29 - 76 + 94 = 47
STEP 1: Set 29 on BC (Fig. 13).
STEP 2: Borrow 100 from A, and subtract

76 from 129. This gives you 53 on BC (Fig.
14).
STEP 3: Add 94 to the 53 on BC. This gives you 147 on ABC (Fig. 15).
STEP 4: As you borrowed 100 previously, you must return it. So remove the 1 on A. The
result is 47 on BC (Fig. 16).
NOTE: In step 3, the result 147 which you got is larger than the 100 which you
borrowed. This shows that the result is positive. Accordingly, the result is 47 on BC.
EXAMPLE 4: 628 - 936 + 864 = 556
STEP 1: Set 628 on BCD (Fig. 17).
STEP 2: Borrow 1 000 from rod A,

and subtract 936 from 1 628. This
13
gives you 692 on BCD (Fig. 18).
STEP 3: Add 864 to the 692 on BCD. This gives you 1 556 on ABCD (Fig. 19).
STEP 4: Remove the 1 on A which you borrowed in step 2. The answer is 556 (Fig. 20).
EXAMPLE 5: 60 - 84 - 96 = -120
In making successive subtractions, if it is not sufficient to borrow 100 from the first rod
to the left, borrow 1,000 from the second rod to the left at the start of your calculations.
We shall call this Method A.
The alternative is to borrow 100 from the first rod to the left in making the first
subtraction and later borrow 900 when you need another 100. We shall call this Method B.
Although Method A is more efficient in that you only have to borrow once, you may not
be able to anticipate your later need for more. In case you haven’t borrowed enough at
the beginning you have to resort to Method B.
METHOD A:
STEP 1: Set 60 on CD (Fig. 21).
STEP 2: As it will not be sufficient to borrow

100, borrow 1 000 from rod A, and subtract 84
from 1 060. This gives you 976 on BOD. In
borrowing 1 000, set 9 on B and, leaving the 9
intact there, shift 100 to B and C and then subtract 84 from 160. In this way you will
automatically get the answer, minus 24, on CD in the form of the complementary number
of 976 with respect to 1 000 (Fig. 22).
STEP 3: Next subtract 96 from the 976 on BCD. This leaves 880 on BCD. The answer is
minus 120, which appears in the form of the complementary number on the board (Fig.
23).
NOTE: This minus 120 is the difference

between the 1 000 which you borrowed and
the 880 which you have on the board.
14
METHOD B:
STEP 2: Borrow 100 from rod B,

and subtract 84 from the 60 on CD,
treating the 60 as 160. This gives you
76 on CD. This number is really
minus 24 (Fig. 25).
STEP 3: You cannot subtract 96 from minus 24. So set 9 on B (Fig. 26).
NOTE: Setting 9 on B now produces the same result as having borrowed 1 000 from rod
A in the beginning since you borrowed 100 previously from rod B.
STEP 4: Now subtract 96 from the 976 on BCD. This leaves 880 on BCD. The answer is
minus 120, which appears in the form of the complementary number (Fig. 27).
NOTE: When you have to borrow again, after borrowing 100, never borrow 100 but 900.
If you borrow 100 twice, you cannot get the complementary number for 200 on the board,
which slows down your operation.
You will find this in the following steps, in which the above Example 5 is worked by
this wrong method.
STEP 2: Borrow 100 from rod B, and subtract 84

from the 60 on CD, treating the 60 as 160. This
gives you 76 on CD (Fig. 29).
STEP 3: Borrow 100 again from B and subtract 96 from 176, and you get 80 on CD (Fig.
30).
Since you borrowed 200, now you must return 200. But as the board does not show the
complementary number of 80 for 200, you must work out the answer by subtracting 80
from 200. Thus this method is rather awkward.
EXAMPLE 6: 55 - 67 - 4 297 = -4 309
STEP 1: Set 55 on DE (Fig. 31).

15
STEP 2: Borrow 100 from C, and subtract 67 from 55, treating the 55 as 155. This gives
you 88 on DE. The result is minus 12 (Fig. 32).
STEP 3: As you cannot subtract 4,297 from the 88 on DE, you must borrow 10 000 from
rod A. So set 9 on both C and B. You must note that setting 9 on both C and B produces
the same result as borrowing 10 000 from A, for the reason that you have previously
borrowed 100 from C (Fig. 33).
STEP 4: Now subtract 4 297 from the 9 988 on BCDE. This gives you 5 691 on BCDE. The
answer is the complementary number of 5 691 with respect to minus 10 000, i.e. minus
4309, which appears on the board (Fig. 34).
Exercises
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
167 283 309 434 512 608 274 713 82 903
280 406 472 159 296 152 460 264 39 -75
349 137 563 618 -947 591 -958 380 65 -829
265 459 128 -703 613 235 -215 521 -721 -6473
420 261 -796 -841 308 -764 309 -849 -9604 146
593 -825 204 397 247 318 132 -706 76 2704
178 -570 -485 204 105 -973 341 -693 148 92
-837 -968 -601 -972 430 -896 -604 -942 -85 -98650
-672 309 815 120 -867 409 -897 -874 -629 -3518
-905 147 937 584 -698 240 -786 150 -4053 637
-159 -361 1 546 0 -1 -80 1 944 2 036 -14 682 105 063
16
III - OTHER METHODS OF MULTIPLICATION

There are several methods of multiplication that can be used on the abacus. A
comparative study of these methods may be interesting. I hope that it will lead the
reader to the conclusion that the standard method of multiplication as described in my
first book is the best. In addition to discussing other basic methods of multiplication I
shall introduce some methods of simplified multiplication.
1. The Variant of the Standard Multiplication
There is a variant method of the standard multiplication, which was popular in Japan
around 1930, but which was largely replaced in less than ten years by the standard
method of multiplication. The variant is still favored by quite a few experts including
entrants in abacus contests, because it is a little faster than the standard method. Here
is an example.
EXAMPLE 1: 56 x 49 = 2 744
STEP 1: Set the multiplicand 56 on EF

and the multiplier 49 on AB (Fig. 35).
STEP 2: Multiplying the 4 on A by the 6

on F, set the product 24 on FG after
clearing F of the 6. This gives you 24 on FG
(Fig. 36).
STEP 3: Multiplying the 9 on B by the

same 6 which you remember was on F, set
the product 54 on GH. Since you had 24 on
FG, you get a total product of 294 on FGH
(Fig. 37).
STEP 4: Multiplying the 4 on A by the

5 on E, set the product 20 on EF after
clearing E of the 5. This gives you a total
of 2 294 on EFGH (Fig. 38).

same 5 which you remember was on E,
set the product 45 on FG. Since you had 2 294 on EFGH, you get, on EFGH, a total of 2
744, which is the answer (Fig. 39).
Advantages and Disadvantages
ADVANTAGE:
This method is a little faster than the standard method, because the distance between
the multiplier and the product has been reduced by one rod.
DISADVANTAGES:
(1) The product of this multiplication does not form in the position of the dividend of
the standard division, as is the case with the standard multiplication and division.
17
It should be noted that this method produces its product in the position of the dividend
of the older method of division. In other words, it forms the counterpart of the older
method of division, the performance of which requires the use of its own special division
table. With the decay of the older method of division, this multiplication method has also
given way to the standard method of multiplication.
(2) The beginner may find this method a little less easy to follow than the standard
one, as he has to remember each digit in the multiplicand after it is removed.
2. Multiplication Starting with the Final Digits of the Multiplier and the
Multiplicand
The order of multiplication in this method is identical with that of written

multiplication. Accordingly, the student of the abacus will find it interesting and useful
to compare this method with the standard method. This arithmetic method is the oldest
multiplication method here introduced. It was used extensively in Japan until around
1930, when it was largely replaced by the variant of the standard multiplication method
later developed, and which, in turn, was replaced by the standard method itself.
This method may be broken down into two variants. One of them, which formed the
counterpart of the older method of division, used to be popular in Japan. We shall call
this Variant B. But today, when the older method of division has fallen out of favor, the
variant of this method which I shall call Variant A makes a counterpart of the standard
method of division and is free from the incidental details of operation that complicate
Variant B.
Variant A
EXAMPLE 1: 78 x 89 = 6 942
STEP 1: Set 78 on EF and 89 on AB (Fig.

40).
STEP 2: Multiplying the 9 on B by the 8

on F, set the product 72 on HL (Fig. 41).
STEP 3: Multiplying the 8 on A by the

same 8 on F, set the product 64 on GH,
and clear F of the 8. This gives you a
total product of 712 on GHI (Fig. 42).
STEP 4: Multiplying the 9 on B by

the 7 on E, add the product 63 to the
71 on GH. This makes a total product
of 1 342 on FGHI (Fig. 43).
STEP 5: Multiplying the 8 on A by

the same 7 on E, add the product 56 to the 13 on FC, and clear E of its 7. This gives you,
on FGHI, a total product of 6 942, which is the answer (Fig. 44).
18
NOTE: The procedure of this Variant A forms the unit rod of the product on the rod to
the right of the unit rod of the multiplicand by as many rods plus one as there are digits
in the multiplier.
Variant B
The disadvantages of Variant B may be illustrated best by the same example, since this
method can become complicated when both multiplier and multiplicand have large digits.
But the complications offer no serious obstacles to experienced operators. Hence, until
recently, Variant B has been preferred to Variant A primarily because it is faster and also
because it is the counterpart of the older method of division.
EXAMPLE 2: 78 x 89 = 6 942
STEP 1: Set 78 on EF and 89 on AB (Fig. 45).
STEP 2: Multiplying the 9 on B by the 8 on F,

set the product 72 on GH (Fig. 46).
NOTE: The reader can see that Variant B

sets the product one rod closer to the
multiplicand and thus is slightly speedier
than Variant B.
STEP 3: Multiplying the 8 on A by the 8 on F, set the product 64 on FG after clearing F

of its 8. This gives you a total product of 712 on FGH (Fig. 47).
STEP 4: Multiplying the 9 on B by the 7

on E, set the product 63 on FC. This
makes a total of 7 342 on EFGH. In this
procedure do not add 1 to the 7 on E, but
remember to add it in the next step (Fig.
48).
STEP 5: Multiplying the 8 on A by the 7 on E, set the product 56 on EF after removing

the 7 on E. In this step do not forget the 1 which must be added to E. This leaves you
with a total of 6 942 on EFCH, which is the answer (Fig. 49).
NOTE: The 1 to be added in step 4 must be remembered till step 5. This kind of
situation occurs especially when both the multiplier and the multiplicand are large
numbers.
EXAMPLE 3: 78 x 456 = 35 568
STEP 1: Set the multiplicand 78 on

FG and the multiplier 456 on ABC (Fig.
50).
STEP 2: Multiplying the 6 on C by

the 8 on G set the product 48 on IJ
19
(Fig. 51).

the same 8 on set the product 40 on
HL. This makes a total of 448 on HIJ
(Fig. 52).

the same 8 on G, set the product 32
on GH, and clear G of its 8. This
makes a total of 3 648 on GHIJ (Fig.
53).

the 7 on F, set the product 42 on HI.
This makes a total of 4 068 on GHIJ
(Fig. 54).

the same 7 on F, set the product 35
on GH. This makes a total of 7 568 on
GHIJ (Fig. 55).
STEP 7: Multiplying the 4 on A by the same 7 on F, set the product 28 on FG, and clear
F of the 7. This gives you, on FGHIJ, a total of 35 568, which is the answer (Fig. 56).
ADVANTAGES:
(1) Variant A of this method forms the counterpart of the standard method of division,
that is, it forms its product in the position of the dividend of the latter method.
(2) Since the order of multiplication is identical with that of written multiplication,
the beginner may find it easier to learn.
DISADVANTAGES:
This method is not commonly used for the following reasons.
(1) The operator has to take the trouble of counting the digits in the multiplier. As
was previously explained, Variant A requires that the unit rod of the first product be
separated from the unit rod of the multiplicand by as many rods plus one as there are
digits in the multiplier, while Variant B requires that the unit rod of the first product be
separated from the last rod of the multiplicand by as many rods as there are digits in the
multiplier. The greater the distance, the greater the inconvenience.
(2) Right-to-left operation makes this method a little slower than the other methods.
(3) Variant B of this method is often complicated by the kind of inconvenience which
arose in Example 2 of this section.
20
Why Is Right-to-Left Operation Slower than Left-to-Right Operation?
The reasons for the slowness of right-to-left operation of this method may be seen
through the analytical comparison of this method with the standard one.
The procedure of multiplication illustrated in Example 2 sets the products as given

below.
The above shows that in this operation the

finger has to travel a total of 9 rods. In step 2,
the finger travels 1 rod from G to H. In step 3,
it travels 3 rods from H back to F and forward
to G. In step 4, it travels 2 rods from G back to
F and forward again to G. In step 5, it travels 3
rods from G back to E and forward to F.
If the same problem, 78 x 89, is performed

by the standard method of multiplication, the
finger has to travel no more than 7 rods, as
indicated in the figure below.
The figure shows that, in step 2, the finger

travels for 1 rod from G to H. In step 3, it
travels for 1 rod from H to I. In step 4, it
travels for 4 rods from I back to F and forward
to G. And in step 5, it traveis for 1 rod from G
to H.
Furthermore, the right-to-left operation of this method may be said to run counter to
the efficient left-to-right operation of the standard method.
3. Multiplication Beginning with the Highest Digits of the Multiplier and

Multiplicand
It is also possible to do multiplication beginning with the highest digits of the

multiplier and the multiplicand. For this method it is necessary to separate the multiplier
and the multiplicand by as many rods plus two as there are digits in the multiplier.
Otherwise the product will extend into the multiplicand, and operation will become im-
possible.
EXAMPLE 1: 43 x 72 = 3 096
STEP 1: Set the multiplicand 43 on GH and

the multiplier 72 on AB, with four vacant
rods between them (Fig. 57).
STEP 2: Multiplying the 7 on A by the 4 on

G, set the product 28 on EF (Fig. 58).
21
STEP 3: Multiplying the 2 on B by the same 4 on set the product 8 on G after clearing G
of the 4, that is, the highest digit of the multiplier. This makes a total product of 288 on
EFG (Fig. 59).

on H, set the product 21 on FG. This makes
a total product of 309 on EFG (Fig. 60).

same 3 on set the product 6 on H after
clearing H of the 3, that is, the second digit
of the multiplier. This gives you, on EFGH,
a total product of 3 096, which is the
answer (Fig. 61).
ADVANTAGES:
(1) When the multiplier is a whole number, the unit digit of the product forms on the
unit rod of the multiplicand, so the necessity of searching for the unit rod of the product
is eliminated. However, this rule does not hold when the multiplier ends in one or more
zeros. The product moves to the right of the multiplicand by as many rods as there are
zeros at the end of the multiplier.
(2) As the operation starts by multiplying the first digits of the multiplier and
multiplicand, it is convenient for approximations.
DISADVANTAGES:
This method is not used very much for the following reasons.
(1) It necessitates counting the digits of the multiplier. The multiplier must be
separated from the multiplicand by as many rods plus two as there are digits in the mul-
tiplier. When the multiplier is long, this method becomes rather awkward.
(2) It is inconvenient for the calculation of compound numbers not based on the
decimal system. Try using this method on a problem involving hours, minutes, and
seconds, for example.
(3) It has no corresponding method of division. In other words, there is no method of

division where the quotient has the position of the multiplicand in this method of
multiplication.
4. The Chinese Method of Multiplication
The following example introduces the method of multiplication widely practiced in

China.
22
EXAMPLE 1: 345 x 67 = 23 115
STEP 1: Set the multiplicand 67 on FG

and the multiplier 345 on ABC (Fig. 62).

7 on G, set the product 28 on HL (Fig.
63).

the same 7 on set the product 35 on IJ.
This gives you a total of 315 on HIJ
(Fig. 64).

the same 7 on G, set the product 21 on
GH after clearing G of its 7. This gives
you a total product of 2 415 on GHIJ
(Fig. 65).
STEP 5: Now multiplying the 4 on

B by the 6 on set the product 24 on
GH. This gives you a total of 4 815 on
GHIJ (Fig. 66).
STEP 6: Multiplying the 5 on C by the same 6 on F, set the product 30 on HI. This gives
you a total of 5 115 on GHIJ (Fig. 67).
STEP 7: Finally, multiplying the 3 on A by the same 6 on F, set the product 18 on FG,
after clearing F of its 6. This gives you, on FGHIJ, a total product of 23 115, which is the
answer (Fig. 68).
The reader will see that this method of multiplication forms its product in the same
position as does the variant of the standard method of multiplication introduced at the
beginning of this chapter.
NOTES: (a) When this method of multiplication is followed, it often facilitates

calculation for the rods to have two five-unit counters when both the multiplier and
multiplicand have large digits. This is one of the reasons why the Chinese abacus has two
five-unit counters on each rod. (b) Also notice that in China the multiplicand is set on the
left and the multiplier is set on the right.
ADVANTAGES:
In this example, suppose you had used the variant of the standard method of
multiplication. Then, in step 2, rod G would have been cleared of the 7 before the
product was set on GH. However, in this method, the 7 on G remains on the board until
step 4, when all of the operations involving that 7 have been completed.
23
DISADVANTAGES:
In this example, the additions of the products proceed from left to right till step 3. But
in step 4, the hand must shift back to the left to add the product. Since the operation
proceeds in this way, it is somewhat slower than the standard method and its variant.
5. The Elimination of the Final Digit of a Multiplier Ending in One
Multiplication can be facilitated by leaving out of calculation the final digit one of the
multiplier. The Japanese technical term for this kind of multiplication may be translated
as the final-digit-one elimination multiplication.
This multiplication starts its calculation with the highest digits of the multiplier and
the multiplicand. The multiplier and the multiplicand must be separated by as many rods
plus one or two as there are digits in the multiplier; otherwise the product will extend
into the multiplier.
EXAMPLE: 74 x 31 = 2 294
STEP 1: Set the multiplicand 74 on GH and the 3 of the multiplier 31 on B (Fig. 69).
NOTES: (a) The trouble of setting the

final digit one of the multiplier is spared,
because it is not used in this operation. (b) A
beginner using this method is liable to make
errors.
STEP 2: Multiplying the 3 on B by the 7 on G, set the product 21 on EF, and you get 2
174 on EFGH (Fig. 70).
NOTES: (a) In this method of multiplication, the multiplicand 74 is supposed to have

been multiplied by the 1 of the multiplier 31, and is left intact on the board, and the
products made by multiplying the 3 of 31, i.e. 30, by the two digits of the multiplicand
each are added to the multiplicand 74. (b) When the multiplier is a whole number, the
unit digit of the product forms on the
unit rod of the multiplicand. (c) In step 2,
the 3 on B is really 30 and the 7 on G is
really 70. Accordingly, their product 21 is
set on EF, because it is really 2 100.
STEP 3: Multiplying the same 3 on B by the 4 on H, add the product 12 to the 17 on FG.
This gives you, on EFGH, a total of 2 294, which is the answer (Fig. 71).
NOTE: In this step, the 3 on B is really 30, but the 4 on H is a real 4. Accordingly, their
product 12 is set on FG, because it is really 120.
This multiplication can be performed by subtracting 1 from a multiplier whose last

digit is 2 or a larger digit, although this procedure does not improve calculations. An
example will be given below.
24
EXAMPLE 2: 83 x 42 = 3 486
STEP 1: Set the multiplicand 83 on GH and

the multiplier 41 on AB (Fig. 72).
NOTE: In this multiplication, the

multiplicand is left intact on the board, and
the operation performed can be expressed as
83 + [83 x (42—1)].
STEP 2: Multiplying the 4 on A by the 8 on

G, set the product 32 on EF. This gives you 3
283 on EFGH (Fig. 73).
STEP 3: Next, multiplying the 1 on B by

the same 8 on G, set the product 8 on G.
This gives you 3 363 on EFGH (Fig. 74).
STEP 4: Now, multiplying the 4 on B by

the 3 on set the product 12 on FG. This
gives you a total of 3 483 on EFGH (Fig.
75).
STEP 5: Finally, multiplying the 1 on B by the same 3 on H, set the product 3 on H.

And you get, on EFGH, a total of 3 486, which is the answer (Fig. 76).
ADVANTAGE:
This method simplifies calculation by reducing the number of the digits in the
multiplier by one.
DISADVANTAGES:
(1) This method necessitates counting the digits in the multiplier so as to separate the
multiplier from the multiplicand by as many rods plus one or two as there are digits in
the multiplier.
(2) The value of this method is limited to the case in which the final digit of the
multiplier is one.
6. The Elimination of the Initial Digit of o Multiplier Beginning with One
Multiplication can also be simplified by leaving out the initial digit one of the
multiplier. The Japanese technical term for this multiplication may be translated as the
initial-digit-one elimination multiplication.
25
EXAMPLE 1: 23 x 12 = 276
STEP 1: Set the multiplicand 23 on EF and the 2 of the multiplier 12 on B, the third
rod to the left of rod E (Fig. 77).
NOTES: (a) There is no objection to setting the whole multiplier

12 on AB. However, experts who use this method do not customarily
set the first digit 1 because it is not used in calculation, and they can
remember that the 2 here is the 12 of the multiplier. However, this
method of multiplication is apt to give the beginner trouble. (b) In
this multiplication, 23 is supposed to have been multiplied by the 10
of the multiplier 12 and is regarded as 230, with G as its unit rod. The product is made by
multiplying the 2 of 12 by each digit of 23 and adding the result to the 230 on EFG.
STEP 2: Multiplying
the 2 on B by the 3 on F,
set the product 6 on G.
This makes a total of 236
on EFG (Fig. 78).
NOTE: Since both the 2 on B and the 3 on F are in the unit place, the product 6 must
be set on G, the unit rod of the product.
STEP 3: Next multiplying the same 2 on B by the 2 on E, add the product 4 to the 3 on
F. This gives you, on EFG, a total of 276, which is the answer (Fig. 79).
NOTES: (a) In this step, since

the 2 on B is in the unit place and
the 2 on E is in the tens place, the
product 4 must be set on F, the
tens rod of the product. (b)
Operation must start with the last
digit of the multiplicand. Otherwise, the product will extend into the multiplicand and
render calculation impossible. (c) This method simplifies calculation especially when the
second digit of the multiplier is small, i.e., when the multiplier is a number such as 11,
107, 1 008, etc.
EXAMPLE 2: 75 x 103 = 7 725
STEP 1: Set the multiplicand 75 on DE and the 3 of the multiplier 103 on A (Fig. 80).
NOTE: In this problem, the multiplicand 75 is supposed to have been multiplied by the
100 of the multiplier 103. Accordingly, the 75 on DE is regarded as the product 7 500 on
DEFG, with G as the unit rod, F as the tens rod, and E as the hundreds rod.
STEP 2: Multiplying the 3 on A by the 5 on E, set the product

15 on FG. This makes a total of 7 5 15 on DEFG (Fig. 81).
STEP 3: Next multiplying the 3 on A by the 7 on add the

product 21 to the 51 on EF. This gives you, on EFGH, a total of 7
725, which is the answer (Fig. 82).
26
NOTE: Since the 7 on D is in the

tens place, and the 3 on A in the unit
place, the product 21 is really 210 and
must be set on EF.
EXAMPLE 3: 78 x 13 = 1 014
This example is to show that when the second digit in the multiplier is not a small one,
this method becomes complicated.
STEP 1: Set the multiplicand 78 on DE and the 3 of the

multiplier 13 on A (Fig. 83).
NOTE: As the multiplier is a two-digit number, F becomes the

unit rod of the product.
STEP 2: Multiplying the 3 on A by the 8 on E, set the product 24 on EF. This makes a
total of 804 on DEF (Fig. 84).
NOTE: However, you must remember till the next step that the digit which was
originally on D was not 8 but 7.
STEP 3: Multiplying the 3 on A by the 7 which you remember was on D, set the product
21 on DE. This gives you, on CDEF, a total of 1 014, which is the answer (Fig. 85).
EXAMPLE 4: An article was bought for $250 and sold at a gain of 6.8%. Find the selling
price.
Since the selling price is determined by adding the cost and the profit, it can be found
by the following multiplication: $250 x (1 + 0.068) $250 x 1.068 = $267
There are two methods for working

this multiplication. One is the standard
method of multiplication, and the other
the method of eliminating the initial
digit one of the multiplier. The latter
operation is shown below.
STEP 1: Set the dividend 250 on EFG, with G as the unit rod, and set the 68 of the
divisor 1.068 on AB (Fig. 86).
NOTE: When the multiplier is a mixed decimal, it is generally advisable to set its unit
figure on the unit rod. But experts usually do not bother to, since the digit 1 is not used
in calculation. As long as you remember that 68 stands for 1.068, it matters little to set
the unit digit of the multiplier on the unit rod.
STEP 2: Since the multiplier is a four-digit number, suppose that 250 has been
multiplied by 1 000, producing 250 000 on EFGHIJ, with G as the unit rod (Fig. 87). Now
multiplying the 6 on A by the 5 on F, set the product 30 on GH (Fig. 87).
27
NOTE: As both the 6 on A and the 5 on F are in the tens place, the product 30 must be
set on GH.
STEP 3: Next multiplying the 8 on

B by the same 5 on F, set the product
40 on HL. This gives you 34 on GH
(Fig. 88).
STEP 4: Next multiplying the 6 on A by the 2 on E, add the product 12 to the 53 on FG.
This gives you 654 on FGH (Fig. 89).
STEP 5: Finally multiplying the 8

on B by the same 2 on E, add the
product 16 to the 54 on GH. This
gives you, on EFG, a total of 267,
which is the answer (Fig. 90).
NOTE: It should be noted that when the multiplier is a mixed decimal whose one
whole figure is in the unit place, the unit rod of the multiplicand becomes that of the
product. The reason for this is that the product may be regarded as having been
multiplied by one. Accordingly, in this multiplication, it is quite easy to locate the unit
rod of the product.
Rules for Finding the Unit Rod of the Product
RULE 1: When the multiplier is a mixed decimal whose whole figure one is in the unit
place—i.e., 1—the unit rod of the quotient forms on that of the multiplicand.
RULE 2: Each time the value of this multiplier is raised by one place, the unit rod of
the product shifts by one rod to the right of that of the multiplicand, and each time the
value of this muitiplier is reduced by one place, the unit rod of the product shifts by one
rod to the left of that of the multiplicand.
Rule (1) means that when the multiplier is 1.05 or 1.023, the unit rod of the product
forms on that of the multiplicand.
forms on the first rod to the right of that of the multiplicand, and that when the
multiplier is 0.105 or 0.1023, the unit rod of the product forms on the first rod to the left
of that of the multiplicand.
ADVANTAGES:
(1) This method of multiplication simplifies calculation especially when the second
digit of the multiplier is zero, as in 103, 109, 1.082, etc.
When the multiplier is a number whose second digit is small, such as 11, 114, 125,
1.078, etc., calculation is often simplified. However, this situation somehmes requires a
figure or figures which have to be carried over to the rod next on the left.
28
(2) As shown in Example 4, when the multiplier is a mixed decimal whose whole figure
one is in the unit place, the multiplicand itself is used as the product. In other words, the
unit digit of the product forms on the unit rod of the multiplicand. This enables the
operator to locate easily the unit rod of the product and to simplify calculations.
Because of these two advantages, this method of multiplication is extensively used in

the calculation of percentages in business.
DISADVANTAGE:
When the multiplier is a number whose second digit is large, like 17, 189, etc., this
method becomes awkward. It may involve remembering so many digits that calculation
becomes extremely difficult. This is especially true when the multiplicand also is a
number with many large digits.
7. Multiplication by Complementary Numbers
Multiplication can often be simplified by using the complement of a multiplier. For

instance, take the problem 26 x 98. The 26 is multiplied by 100, becoming 2 600. From
this 26 x 2, 52 is subtracted. In this way the answer 2 548 is obtained. The 2 is, of course,
the complement of 98 with respect to 100. This method of computation is better than the
ordinary method of multiplication when the multiplier is a number a little smaller than
100 or 1 000, etc., such as 97, 996, etc.
EXAMPLE 1: 26 x 98 = 2 548
STEP 1: Set the multiplicand 26

on DE, and on A, set 2, the
complement of 98 with respect to
100 (Fig. 91). When the multiplier is
a two-figure number, the
multiplicand is regarded as having
been multiplied by 100. So in this problem the unit rod of the product shifts to G.
STEP 2: Multiplying the 2 on A by the 6 on E, subtract the

product 12, which is to be set on FG, from the 6 on E. This leaves
2 588 on DEFG (Fig. 92).
STEP 3: Multiplying the 2 on A by the 2 on D, subtract the

product 4 from the 8 on F. This leaves 2 548 on DEFG (Fig. 93), the
answer.
EXAMPLE 2: A manufacturer wishes to purchase a piece of machinery priced at $7 250

on which there is a discount of 4% for cash. Find the price he pays in cash. The answer
can be found by the following simplified multiplication.
7 250 x 0.96
= 7 250 x (1—0.04)
= 7 250 — (7 250 x 0.04)
= 6 960
29
STEP 1: Set the multiplicand 7 250 on DEFG, and on A, set the 4 of 0.04, which is the
complement of 0.96 with respect to 1 (Fig. 94).
NOTE: On the board you have this problem: 7 250 — (7 250 x 0.04). In this problem,
the actual multiplier is 0.96, i.e., a decimal with its first significant digit in the tenths
place. So you may consider that the answer is obtained by keeping the multiplicand as it
is and subtracting from it 7 250 x 0.04. You
should note that when the multiplier is a
decimal with its first significant digit in the
tenths place, the unit rod of the product
remains that of the multiplicand.

on F, subtract the product 20, which is to
be set on GH, from the 5 on F. This leaves
7 248 on DEFG (Fig. 95).
NOTE: As the 5 on F is in the tens place

of the dividend, the product 20 must be
set on GH.
STEP 3: Multiplying the 4 on A by the 2 on E, subtract the product 8 from the 8 on G.

This leaves 724 on DEF (Fig. 96).
STEP 4: Multiplying the 4 on A by the 7 on D, subtract the product 28 from the 724 on
DEF. This leaves 6 960 on DEFG (Fig. 97). The answer is $6 960.
Rules for Finding the Unit Rod of the Product
RULE 1: When the multiplier is a decimal whose first significant digit is in the tenths
place, the unit rod of the multiplicand remains that of the product.
RULE 2: Each time the value of this multiplier is raised by one digit, the unit rod of
the product shifts by one rod to the right of that of the multiplicand, and each time the
value of this multiplier is reduced by one digit, the unit rod of the product shifts by one
rod to the left.
forms on that of the multiplicand.
forms on the first rod to the right of that of the multiplicand, and that when the
multiplier is 0.095 or 0.0934, the unit rod of the product forms on the first rod to the left
of that of the multiplicand.
30
ADVANTAGES:
(1) The reader can easily see in Example 2 that this method is easier than actually
multiplying 7 250 x 0.96.
(2) As stated in Rule 1 for finding the unit rod of the product, when the multiplier is a
mixed decimal whose first significant digit is in the tenths place, the unit rod of the
multiplicand remains that of the product. This saves the operator the trouble of finding
the unit rod of the product. This is a real advantage.
For these reasons this method of multiplication is extensively employed in the

calculation of percentages in business.
DISADVANTAGES:
This method has no particular disadvantages. However, it should be noted that its
utility is limited to problems in which the multipliers are numbers a little smaller than
100, 1 000, etc. or numbers whose first digit is 9, such as 98, 987, 0.96, etc.
Exercises
(1) 24 x 88 = 2 112 (4) 169 x 987 = 166 803

(2) 37 x 98 = 3 626 (5) 408 x 996 = 406 368
(3) 851 x 99 = 84 249 (6) 3 600 x 0.92 = 3 312
31
IV - OTHER METHODS OF DIVISION

There is an older method of division which is as fast as the standard method of division
that I explained in my first book. But this method is no faster and it requires the
memorization of a complicated division table. Although this method is still used by the
older generation, it is generally becoming obsolete. Since this method is so complicated
and since it is hardly used today, I will not spend time explaining it. Instead I shall
present some methods of division which are popular for business calculations.
Historically speaking, however, the “standard method of division” is older than what I
have called here the older method of division. In ancient China the “standard method of
division” was used by means of reckoning blocks. However, the “older method of
division,” because of its convenience in the calculation of weights not based on the
decimal system, became more popular than the “standard method of division.” These
weights were extremely important in Chinese life because they were used in the con-
version of currency. When the abacus was introduced to Japan, it was the “older method
of division” which accompanied it. The currency in Japan also was not based on the
decimal system, so that although the “standard method of division” was strongly
advocated by some mathematicians, it never won real public support. It was not until the
new decimal-based currency of the Meiji era replaced the old Japanese currency that the
“standard method of division” actually became standard.
1. The Elimination of the Initial Digit of a Divisor Beginning with One
There is a method which simplifies division by leaving out of calculation the initial
digit, one, of the divisor. The Japanese technical term for this division may be translated
as the initial-digit-one elimination division.
This division is particularly useful when the divisor is a number a little larger than 100,
1000 etc., such as 103, 1 014, etc. It is a counterpart of the multiplication method in
which the initial digit one of the multiplier is left out of calculation.
Let us take a very simple example. If you divide 306 by 102, you get the quotient 3. In
this case, since the first digit of the divisor 102 is 1, the first digit 3 of the dividend 306
may be used as the quotient. This principie is applied to this method of division on the
abacus, On the abacus board, in each step of the division, the first digit of the dividend is
used as the trial quotient, and thus the necessity of setting the quotient and that of
removing the first digit of the dividend is eliminated, and calculation is considerably
simplified and accelerated. Example 1 shows how this division, “306 / 102,” is actually
done on the board.
EXAMPLE 1: 306 / 102 = 3
STEP 1: Set the

dividend 306 on FGH
with H as the unit rod,
and set the divisor 102
on ABC (Fig. 98).
32
STEP 2: Suppose that you have divided the 306 on FGH by the 100 of the 102 on ABC
and have got the 3 on F as the quotient figure. Multiply the 2 on C by the 3 on F, and
subtract the product 6 from the 6 on H. This clears the board of the dividend, and leaves
the quotient 3, the answer, on F (Fig. 99).
NOTE: On the board, 306 is divided by 100. But since the actual divisor is 102, the
product obtained by the multiplication of 2 (i.e., the difference between 100 and 102)
and 3 (i.e., the quotient) must be subtracted from 306. On the board, the procedures of
setting the first quotient figure 3 and of subtracting the first digit 3 of the dividend are
saved or eliminated.
EXAMPLE 2: 2 575 / 103 = 25
STEP 1: Set the dividend 2 575 on DEFG and the 3 of 103 on A (Fig. 100).
NOTE: There is no objection to setting the whole divisor 103 on the board. However,
experts do not, be-cause this digit is not actually used in calculation, and so long as they
remember that the 3 on A stands for 103 it is not important to set the whole divisor on
the board.
STEP 2: As the divisor 103 is a three-digit

number, suppose that you have divided the 257
of 2 575 by the 100 of 103 and have tried the 2
on D as the quotient figure. Multiplying the 3 on
A by the 2 on D, subtract the product 6 from the
7 on F. This leaves, on EFG, 515 as the remainder of the dividend (Fig. 101).
NOTES: (a) As the second figure of the divisor 103 is zero, you must set the product 6
on F, skipping over E. (b) In this step, as the 2 on D is used as the first quotient figure,
the procedures of setting the first quotient figure 2 and of subtracting the first digit 2 of
the dividend 257 are eliminated.
STEP 3: Next, suppose that you have divided the 515 on EFG by 100 and have got the 5
on E as the second quotient figure. Multiplying the 3 on A by the 5 on E, subtract the
product 15 from the 15 on FG. This clears the board of the remaining dividend, and
leaves, on DEF, 25 as the quotient (Fig. 102).
As the dividend was practically divided by

100, the unit rod of the quotient moves to
the second rod to the left of that of the
dividend. Therefore, E becomes the unit rod
of the quotient. The answer is 25.
NOTE: In this step, as the 5 on E is used as the second quotient figure, the procedures
of setting the second quotient figure and of subtracting the first digit 5 of the dividend
515 are eliminated.
EXAMPLE 3: A merchant’s sales increased 0.8% in the second month. The sales for the
second month amounted to $3 780. How much were the sales the first month?
33
This problem can be worked out by the following division: $3 780 / 1.008 = $3 750
STEP 1: Set the dividend 3 780 on

DEFG and the 8 of the divisor 1.008
on A (Fig. 103).
NOTE: You must remember that 8

stands for 1.008.
STEP 2: As the divisor is a four-digit number, suppose that you have divided the 3 780
on DEFG by 1000 and have got the 3 on D as the first quotient figure. Multiplying the 8 on
A by the 3 on D, subtract the product 24 from the 80 on FG. This leaves, on EFG, 756 as
the remainder of the dividend (Fig. 104).
NOTE: As the second and third figures of 1.008 are zero, you must set the product 24
on FG and not on EF.
STEP 3: Suppose that you have divided

the 7 560 on EFGH by 1,000 and have got
the 7 on E as the second quotient figure.
Multiplying the 8 on A by the 7 on E,
FGH. This leaves, on FGH, 504 as the remainder of the dividend (Fig. 105).
NOTE: Again, be sure to set the product 56 on GH.
STEP 4: Suppose that you have divided the 5 040 on FGHI by 1 000, with the 5 on F as
the third quotient figure. Multiplying the 8 on A by the 5 on F, subtract the product 40
from the 40 on HI. This clears the board of the dividend, and leaves 375 as the quotient
on DEF (Fig. 106).
As the divisor 1.008 is a mixed decimal, whose whole figure

one is in the unit place, the unit rod of the quotient is
identical with that of the dividend. The answer is $3 750.
When the divisor is a number whose second digit is not zero,

such as 125, 137, etc., it often happens that a figure a little
smaller than the first digit of the dividend must be tried as a
quotient figure. In such cases this method is less efficient.
Here is an example showing how the division should be worked in such a case.
EXAMPLE 4: A man withdrew his savings from an account after they had earned 12%
interest. If the amount he withdrew was $5,3 76, what was his original investment?
This problem can be worked by the

following division: 5 376 / 1.12 = 4 800
STEP 1: Set the dividend 5 376 on EFGH

and the last two digits 12 of the divisor
1.12 on AB (Fig. 107).
34
NOTE: When a divisor is a mixed decimal, it is advisable for the beginner to set its unit
digit on a unit rod.
STEP 2: As the divisor is a three-digit number, sup-pose that you have divided the 537
on EFG by 100, with the 5 on E as the first quotient figure, and found that 5 is too large.
It is because if you multiply the 1 on A by the 5 on E, the product 5 which is to be set on
F is larger than the 3 on F. So try 4 as the first trial quotient figure in your mind. Then 1
is left on E as the first figure of the remaining dividend. Multiply the 1 on A by this 4
which is mentally on E, and you will see that you can subtract the product 4 from the 13
which remains mentally on EF. Now subtract the 4 from the 13 by setting 4 on E and
adding 6 to the 3 on F. This gives you 49 on EF, with the 4 on E as the first quotient figure.
Of course, in this particular problem if the

operator had looked at the problem carefully,
he would have tried 4 the first time. An
experienced operator can see at a glance that
4 is the proper first number of the quotient.
STEP 3: Now multiply the 2 on B by the 4 on E, and subtract the product 8 from the 97
on FG. This leaves, on FGH, 896 as the remaining dividend (Fig. 109).
STEP 4: Next suppose that, in dividing the 896 on FGH by 100,

you have tried the 8 on F as the second quotient figure.
Multiplying the 1 on A by the 8 on F, subtract the product 8
from the 9 on G. This leaves, on GH, 16 as the remaining
subtrahend (Fig. 110).
STEP 5: Finally multiplying the 2 on B by the 8 on F, subtract

the product 16 from the 16 on GH. This clears the board of the
subtrahend and leaves 48 as the quotient on EF (Fig. 111). The
answer is $4 800.
NOTE: As the divisor 1.12 is a mixed decimal whose whole

figure one is in the unit place, the unit rod of the dividend
remains that of the quotient.
Rules for Finding the Unit Rod of the Quotient
RULE 1: When the divisor is a mixed decimal whose whole digit one is in the unit place,
the unit rod of the quotient forms on that of the dividend.
RULE 2: Each time the value of this divisor is raised by one digit, the unit rod of the
quotient shifts one rod to the left of that of the dividend, and each time the value of this
divisor is reduced by one digit, the unit rod of the quotient shifts one rod to the right of
that of the dividend.
For instance, when the divisor is 1.05 or 1.023, the unit rod of the quotient forms on
that of the dividend. When the divisor is 10.5 or 10.23, the unit rod of the quotient forms
on the first rod to the left of that of the dividend. When the divisor is 0.105 or 0.1023,
the unit rod of the quotient forms on the first rod to the right of that of the dividend.
35
This rule may be restated as follows:
RULE 2: The unit rod of the quotient forms on the rod to the left of that of the
dividend by as many rods minus one as there are whole digits in the divisor.
The unit rod of the quotient forms on the rod immediately to the right of that of the
dividend by as many rods plus one as there are zeros before the first significant decimal
figure.
The advantages and disadvantages of this method of division are nearly identical with
those of the method of multiplication which leaves the initial digit one of the multiplier
out of calculation (see chapter 3, section 6).
ADVANTAGES:
(1) This method of division simplifies calculation when the second digit of the divisor is
zero, as in 105, 1.08, etc. When the divisor is a number whose second digit is small, such
as 13 or 1.12, etc., often this method simplifies calculation, although the sort of
difficulties illustrated in Example 4 may arise.
(2) As shown in Examples 3 and 4, when the divisor is a mixed decimal whose whole
digit one is in the units place, the unit rod of the quotient is identical with that of the
dividend. This is a great advantage to the abacus operator, as he has no trouble in finding
the unit rod of the quotient.
For these reasons this method of division is employed extensively in the calculation of
percentages in business, that is, for finding the cost of goods from the selling price and
the percent of profit.
DISADVANTAGES:
(1) When the divisor is a number whose second digit is large, such as 19, 183, etc., this
method is awkward.
(2) When the divisor is a number whose second and successive digits are large, this
method usually is much too complicated.
2. Division by Complementary Numbers
When the divisor is a number a little smaller than 100, 1 000, etc., such as 98, 997,
etc., division can be accelerated by using the complement of the divisor with respect to
100, 1000, etc. Let us take two simple examples. 98 goes into 196 twice. So you can
divide 196 by 100 with 2 as the quotient if you add, to 196, 4 which is obtained by
multiplying 2, the difference between 100 and 98, and 2, the quotient. The 4 is the same
number you have subtracted in excess. In the same way you can divide 294 by 100 with 3
as the quotient if you add, to 294, 6 which is obtained by multiplying the 2 and 3, the
36
quotient. The following two rules will help you find the quotient of this division easily,
especially if the first digit of the divisor is 9.
RULE 1: The quotient can generally be found by adding 1 to the first digit of the
dividend when the quotient is a one-digit number, as in the aboye examples, and when
the final digit of the quotient is to be obtained.
RULE 2: In case the quotient is a more-than-one-digit number, the first digit of each
dividend is generally used as the trial quotient figure, except in finding the final digit of
the quotient.
EXAMPLE 1: 882 / 98 = 9
STEP 1: Set the dividend 882 on

DEF with F as the unit rod and set,
on A, 2, which is the complementary
number of 98 with respect to 100. It
is important to note that although
100 is the divisor, only the complementary number of the real divisor with respect to 100
is set on the board.
STEP 2: Since the quotient is a one-digit number, you can apply the above-mentioned
Rule 1 to this problem. Add 1 to the first digit 8 of the dividend 882 and you get 9. If 9 is
the correct quotient, the following two equations should be found equal.
Equation A: 882 + (2 x 9) = 900

Equation B: 100 x 9 = 900
Since the above two equations are equal, 9 is the correct quotient. In equation A, 882
is the dividend, 2 is the complementary number, and 9 is the quotient. In equation B, 100
is the divisor.
Now, multiplying the 2 on A by the trial quotient figure 9, add the product 18 to the 82
on EF. This gives you 900 on DEF. Since the divisor 98 is a two-whole-digit number, the
unit rod of the quotient is formed on D. The answer is 9 on D (Fig. 113).
EXAMPLE 2: 8 160 / 96 = 85
STEP 1: Set the dividend 8 160 on

DEFG, and set on A, 4, which is the
complement of 96 with respect to
100 (Fig. 114). Since the dividend is
a two-digit number, 100 is used as
the divisor.
STEP 2: Divide the 816 on DEF by 100, with the 8 on D as the

first trial quotient figure. In this step, apply Rule 2. Now,
multiplying the 4 on A by the 8 on D, add the product 32 to the 16
on EF. This gives you 480 on EFG (Fig. 115).
37
STEP 3: In this final step, apply Rule 1 and try 5, which is larger than the 4 on E by 1,
as the next trial quotient figure. Multiplying the 4 on A by this 5, add the product 20 to
the 80 on FG, and you get 500 on FGH. This number is equal to the 500 obtained by
multiplying the divisor 100 by the trial quotient 5. As the divisor is a three-whole-digit
number, the unit rod of the quotient is formed on the second rod to the left of that of
the dividend. The answer is 85 on DE (Fig. 116).
EXAMPLE 3: 31 428 / 0.97 = 32 400
STEP 1: Set the dividend 31 428 on DEFGH,

and set on A, 3, the complement of 97 with
respect to 100 (Fig. 117).
STEP 2: Divide the 314 on DEF by 100,

with the 3 on D as the first quotient figure.
Multiplying the 3 on A by the 3 on D, add
the product 9 to the 4 on F. This leaves, on
EFGH, 2 328 as the remainder of the
dividend (Fig. 118).
STEP 3: Now divide the 232 on EFG by 100, with the 2 on E as the second quotient
figure. Multiplying the 3 on A by the 2 on E, add the product 6 to the 2 on G. This leaves
388 on FGH as the remainder of the dividend (Fig. 119).
STEP 4: In this final step, apply Rule 1 and try 4, which is larger than the 3 on F by 1,
as the final trial quotient figure. Add, to the 388 on FGH, the
product 12 obtained by multiplying the 3 on A by this 4, and you
get 400 on F. This product is the same as that obtained by
multiplying the divisor 100 by this same 4. This shows that 4 is
the correct third quotient figure. In this division, the divisor is a
decimal fraction whose first significant digit is in the tenths’
place. Therefore the unit rod of the quotient is identical with
that of the dividend. The answer is 32 400 (Fig. 120).
Rules for Finding the Unit Rod of the Quotient
RULE 1: When the divisor is a decimal fraction whose first significant digit is in the
tenths place, the unit rod of the dividend remains that of the quotient.
RULE 2: Each time the value of this divisor is raised by one place, the unit rod of the
quotient shifts by one rod to the left of that of the dividend, and each time the value of
this divisor is reduced by one place, the unit rod of the quotient shifts by one rod to the
right of that of the dividend.
For instance, when the divisor is 0.95, the unit rod of the quotient forms on that of the
dividend. When the divisor is 9.5, the unit rod of the quotient forms on the first rod to
the left of that of the dividend. When the divisor is 0.095, the unit rod of the quotient
forms on the first rod to the right of that of the dividend.
Rule 2 may be restated as follows:
38
RULE 2: The unit rod of the quotient forms on the rod immediately to the left of that
of the dividend by as many rods as there are whole digits in the divisor.
The unit rod of the quotient forms on the rod immediately to the right of that of the
dividend by as many rods as there are zeros before the first significant decimal figure.
The advantages and disadvantages of this method of division are nearly identical with
those of the method of multiplication by means of complementary numbers.
ADVANTAGES:
(1) This method of division simplifies calculations when the divisor is a number a little
smaller than 100, 1 000, 1, or 0.01, etc., or especially numbers whose first significant
digit is 9, such as 98, 997, 0.96, or 0.0094, etc.
(2) As described in Rule 1 for finding the unit rod of the quotient, when the divisor is a
decimal fraction whose first significant digit is in the tenths place, the unit rod of the
quotient forms on that of the dividend. This is a great advantage to the abacus operator,
as he has no trouble in finding the unit rod of the quotient.
Because of these advantages, this method is extensively used in business for the
calculation of percentages. For example, it is useful in finding the regular price from the
discount price and the discount rate.
DISADVANTAGE:
Its utility is limited to problems in which the divisor is a number just a little smaller
than 100, 1 000, etc. When the divisor is a number whose first significant digit is smaller
than 9—such as 75, 0.82, etc., Rules 1 and 2 given for finding trial quotient figures do not
often apply. In such cases, this method of division is awkward, and the standard method
of division is more efficient.
Exercises
(1) 2 116 / 92 = 23 (4) 382 305 / 993 = 385

(2) 4 048 / 88 = 46 (5) 705 042 / 897 = 786
(3) 13 832 / 988 = 14 (6) 1 645 / 0.94 = 1 750
39
V - MORE ABOUT DECIMALS

In my first book I included a chapter on decimals. I tried to show in that discussion the
Japanese method of handling decimal fractions of various sorts. However, an astute
reader may have seen two interesting things about my previous presentation. One thing
was simply an oversight on my part, namely that I gaye very few clues as to how such a
multiplication problem as 0.2 x 0.4 = 0.08 would be done on the abacus, i.e. a problem
having decimals both in the multiplicand and the multiplier. Also I neglected such division
problems as 1.2 / 3 = 0.4, i.e., a problem having decimals in the quotient.
The other thing that a Western reader in particular might have found strange is that I
presented the following three problems as if they were different and each required a
different kind of treatment: 31.36 / 0.32 = 98, 3.136 / 0.032 = 98, and 0.3136 / 0.0032 =
98. The reason this might seem strange is that when an Occidental sets up any of these
three problems on paper for long division, he immediately reduces them all to the same
problem, namely 3 136 / 32 = 98. These are treated as distinct problems in Japan
because of the abacus tradition of avoiding any mental effort and following the une of
least resistance. The Western reader may not be satisfied with this solution and may
want to work out his own compromise with the Japanese method. I think, however, that
after a beginner has practiced this method enough so that it becomes mechanical, he will
consider it better than any system which requires a non-mechanical, mental effort—
better than the Western method of recognizing the essential simplicity of some problems.
Now let me turn to my original problem. In my first book, I explained how to do the
following problems:
MULTIPLICATION DIVISION
4x2=8 8/2=4
25 x 15 = 375 375 / 15 = 25
405 x 123 = 49 815 49 815 / 123 = 405
34 x 1.2 = 40.8 40.8 / 1.2 = 34
32 x 0.4 = 12.8 12.8 / 0.4 = 32
98 x 0.32 = 31.36 31.36 / 0.32 = 98
32 x 0.04 = 1.28 1.28 / 0.04 = 32
98 x 0.032 = 3.136 3.136 / 0.032 = 98
32 x 0.004 = 0.128 0.128 / 0.004 = 32
98 x 0.0032 = 0.3136 0.3136 / 0.0032 = 98
The following five examples should explain how to do any other normal multiplication
and division problems.
EXAMPLE 1: (A) 18.7 x 53 = 991.1

(B) 98.6 / 34 = 2.9
(A) shows that when the multiplier

is a two-digit whole number, the unit
digit of the product is formed on the
third rod to the right of that of the
multiplicand, i.e. on rod J.
40
(B) shows that when the divisor is a

two-digit whole number, the unit digit
of the quotient is formed on the third
rod to the left of that of the dividend,
i.e. on rod G.
EXAMPLE 2: (A) 3.4 x 1.2 = 4.08

(B) 6.45 / 7.5 = 0.86

is a mixed number with only one
whole digit in it, the unit rod of the
product is formed on the second rod
to the right of that of the
multiplicand, i.e. on rod I.
mixed number with one whole digit,
the unit rod of the quotient forms on
the second rod to the left of that of
the dividend, E.
EXAMPLE 3: (A) 31 x 31.9 = 988.9

(B) 0.9429 / 72.5 = 0.013

is a mixed number with two whole
digits, the unit rod of the product is
formed on the third rod to the right
of that of the multiplicand, i.e. rod K.
(B) shows that when the divisor is
a mixed number with two whole
digits, the unit rod of the quotient is
formed on the third rod to the left of
that of the dividend, i.e. rod E.
EXAMPLE 4: (A) 0.922 x 0.65 = 0.5993

(B) 0.5796 / 0.28 = 2.07
(A) shows that when the divisor is a

decimal fraction with its first
significant digit in the tenths place,
the unit digit of the product is formed
on the first rod to the right of that of
the multiplicand, i.e. on rod H.
decimal fraction with its flrst
significant digit in the tenths place,
the unit digit of the quotient is formed
on the first rod to the left of that of
the dividend, i.e. on rod F.
41
EXAMPLE 5: (A) 0.0289 x 0.00741 = 0.000214149

(B) 0.0374644 / 0.00409 = 9.16

is a decimal fraction with its first
significant digit in the thousandths
place, the unit rod of the product is
formed on the first rod to the left of
that of the multiplicand, i.e. on rod F.
decimal fraction with its first
significant digit in the thousandths
place, the unit digit of the quotient is
formed on the first rod to the right of
that of the dividend, i.e., on rod K.
42
VI - CALCULATIONS INVOLVING MORE THAN ONE UNIT OF

MEASUREMENT
1. Setting Compound Numbers on the Board
In setting a compound number, a measurement expressed in more than one unit, on

the board, the part with the highest denomination is first set on a unit rod, and all the
other parts are set so that they are separated by at least one vacant rod. Below are
examples of how a compound number should be set on the board.
EXAMPLE 1: £15 14 s. 9 d.
EXAMPLE 2: 12 hr. 5 min. 20 sec.
In Example 1, 14 s. is separated from £15 by one vacant rod, while 9 d. is separated

from 14 s. by two vacant rods (Fig. 121), because during calculations the number of
pence may require the use of two digits and thus two rods. (There are 12 pence in a
shilling and 20 shillings in a pound.)
In Example 2, 5 min. is separated from 12 hr. by two vacant rods, because during
calculations we may have as many as 59 minutes and thereby need two rods for
calculations and one for separation. The 20 seconds is separated from 5 min. by one
vacant rod, because no additional rods will be necessary (Fig. 122).
It should be noted that the aboye arrangements of the compound numbers are the
most convenient not only for addition and subtraction but also for multiplication and
division. For further details, see Example 3.
EXAMPLE 3: 7 yd. 1 ft. 9 in.
In Example 3, 1 ft. is separated from 7 yd. by one vacant rod, while 9 in. is separated
from 1 ft. by two vacant rods, because during calculations we may have as many as 11
inches and thereby two rods (Fig. 123).
When only addition and subtraction are to be performed, each unit of measurement is
often given a separate unit rod, e.g., 1 ft. would be set on G and 9 in. on J. But when
multiplication and division are also to be performed, it is more convenient to set the
units as in the above examples.
43
2. Converting Compound Numbers
EXAMPLE 1: Find the value of £37 10 s. in dollars if one pound equals $4.03.
This problem can be simplified as follows:
10 s.= £0.5 (NOTE: 10 s. / 20 s. = 0.5)

£37 10 s. = £37.5
£37.5 x 4.03 = $151.125
STEP 1: Set the multiplicand 37.5 on FGH, with G as

the unit rod. Set the multiplier 4.03 on ABC, leaving two
rods vacant (Fig. 124).
NOTE: It is not really necessary to set the unit figure

of the multiplier on a unit rod.
STEP 2: Multiplying 4 on A by 5 on H, set product 20 on

IJ (Fig. 125).
STEP 3: Multiplying the 3 on C by the same 5 on H, set

the product 15 on KL, and clear H of the 5. This makes a
total of 2 015 on IJKL (Fig. 126).
NOTE: As the second figure of the multiplier is zero,

set the product 15 on KL, skipping JK.
STEP 4: Next, multiplying the 4 on A by the 7 on G, set

the product 28 on HI. This makes a total of 30 015 on
HIJKL (Fig. 127).
STEP 5: Multiplying the 3 on C by the same 7 on G, set

the product 21 on JK, and clear G of the 7. This makes a
total of 30 225 on HIJKL (Fig. 128).
STEP 6: Next, multiplying the 4 on A by the 3 on F, set

the product 12 on GH. This makes a total of 150 225 on
GHIJKL (Fig. 129).
STEP 7: Finally multiplying the 3 on C by the same 3 on

F, set the product 9 on J, and clear F of the 3. This gives
you a total of 151 125 on GHIJK (Fig. 130).
As the multiplier is a mixed decimal whose whole

figure is in the unit place, the unit rod of the product
forms on I, the second rod to the right of G, the unit rod
of the multiplicand. The answer is $151.125.
44
EXAMPLE 2: Find the value of the following compound number in terms of pence (12
pence = 1 shilling; 20 shillings = 1 pound). £12 16 s. 8 d.
Written calculations:
On the abacus, the operation is

performed much in the same way as in
written calculation. First £12 is
multiplied by 20 and the resulting 240 s.
is added to 16 s. Next the sum 256 s. is
multiplied by 12 and the product 3 072 d.
is added to 8 d. The answer is 3 080 d.
However, on the abacus, both the multiplication and addition are performed
simultaneously, and calculations are much faster.
STEP 1: Set 12 on AB, with B as the unit

rod, 16 on DE, with E as the unit rod, and 8
on H, with H as the unit rod (Fig. 131).
STEP 2: First multiply the 2 on B by 20,

add the product 40 to the 16 on DE, and
clear B of the 2. This makes a total of 56 on
DE. Next multiplying the 1 on A by 20, set
the product 20 on CD. and clear A of the 1.
This gives you a total of 256 on CDE (Fig.
132).
NOTE: In this procedure multiply the 2 on B by 20 first and the 1 on A by 20 next. If

you multiply the digit on A first, the product may extend to rod B. This will happen if the
digit on A is large. If, for example, the number on A were 8, setting the product 16 on BC
would be rather confusing.
STEP 3: Now the 256 on CDE must be multiplied by 12 and added

to the 8 on H. First multiply the 6 on E by 10, and set the product
60 on GH. This makes a total of 68 on GH. Next multiplying the
same 6 on E by 2, add the product 12 to the 68 on GH, and clear E
of the 6. This gives you a total of 80 on GH (Fig. 133).
NOTE: Be sure to dispose of the 6 of 256 first, next the 5, and finally the 2.
STEP 4: Multiplying the 5 on D by 10, set the product 50 on

FG. This makes a total of 580 on FGH. Next, multiplying the
same 5 on D by 2, add the product 10 to the 58 on FG, and
clear D of the 5. This gives you a total of 680 on FGH (Fig. 134).
STEP 5: Multiplying the 2 remaining on C by 10, set the

product 20 on EF. This makes a total of 2 680 on EFGH.
Finally, multiplying the same 2 on C by 2, add the product 4

to the 6 on F, and clear C of the 2. This gives you a total of 3
080 on EFGH (Fig. 135). The answer is 3 080 d.
45
NOTE: Experts usually save the trouble of setting such simple multipliers as 20 and 12
on the board.
EXAMPLE 3: Find the value of 6 190 d. in terms of pounds and shillings.
Written calculation:
On the abacus, the operation is performed

in the same way as in written calculation.
First 6 190 is divided by 12, next the quotient
515 is divided by 20.
STEP 1: Set 6 190 on IJKL, with L as the unit rod, and

set 12 on CD. See that there are four vacant rods between
the two numbers (Fig. 136).
STEP 2: Comparing the divisor 12 and the 61 in 6 190,

set the quotient figure 5 on G. Now multiply the 1 on C by
this 5, and subtract the product 5 from the 6 on I. This
leaves 1 190 on IJKL. Next multiply the 2 on D by the same
5 on and subtract the product 10 from the 11 on IJ. This
leaves 190 on JKL (Fig. 137).
STEP 3: Comparing the divisor 12 and the 19 on JK, set

the second quotient figure 1 on H. Now multiply the 1 on C
by the 1 on H, and subtract the product 1 from the 1 on J.
This leaves 90 on KL. Next multiply the 2 on D by the same
1 on H, and subtract the product 2 from the 9 on K. This
leaves 70 on KL (Fig. 138).
STEP 4: Comparing the divisor 12 and the 70 on KL, set

the third quotient figure 5 on 1. Now multiply the 1 on C
by this 5 on 1, and subtract the product 5 from the 7 on K.
This leaves 20 on KL. Next multiply the 2 on D by the same
5 on I, and subtract the product 10 from the 20 on KL. This
leaves 10 on KL (Fig. 139).
STEP 5: Now leaving the 10 d. on KL as it is, you must

divide the 515 s. on GHI by 20 to convert it into pounds.
Clear CD of its 12, and set the divisor 20 on AB, leaving
four vacant rods between it and the 515 on GHI (Fig. 140).
STEP 6: Now comparing the 2 on A and the 5 on G, set

the quotient 2 on E. Multiplying the 2 on A by this 2 on E,
subtract the product 4 from the 5 on G. This leaves 115 on
GHI (Fig. 141).
46
STEP 7: Next compare the 2 on A and the 11 on GH and

set the quotient figure 5 on F. Multiplying the 2 on A by
this 5 on F, subtract the product 10 from the 11 on GH.
This leaves 15 s. on HI. The answer is £25 15 s. 10 d. (Fig.
142).
NOTE: Why are the pounds, shillings, and pence figures

set as they are in Examples 2 and 3?
The close arrangement of these figures is the most

efficient arrangement for performing the four arithmetic
operations. Any closer arrangement would cause confusion.
This arrangement affords another special advantage in

that the standard methods of multiplication and division
can be used most efficiently.
Examine step 2 in Example 2, where pounds are con-verted to shillings and you can see
that the standard method of multiplication forms the unit digit of the product on the unit
rod of the shillings figure on the board.
Examine steps 2, 3, and 4 in Example 3, where the pence are converted to shillings,
and you can see that the standard method of division forms the unit digit of the quotient
on the unit rod of the shillings figure on the board. Any other arrangement of the pounds,
shillings, and pence figures would require the use of methods of multiplication and
division other than the standard methods and would make calculations less efficient.
3. Adding and Subtracting Compound Numbers
EXAMPLE 1:
There are two ways of adding measures.

One is to add each compound number
from higher to lower denominations, and
the other is to add up column by column
from lower to higlier denominations as in written calculation. The former method is
generally used when both addition and subtraction are to be performed, and it must
always be followed when successive compound numbers are dictated to the operator.
The latter is considered to be a little faster if it can be used.
METHOD A: Adding from higher to lower denominations:
STEP 1: Set 21 yd. on AB, with B as the

unit rod. Set 1 ft. on D and 8 in. on G (Fig.
143). Refer to Fig. 123 at the beginning of
this chapter.
STEP 2: Add 15 to the 21 on AB. This

makes a total of 36 on AB. Now add 2 to
the 1 on D. This makes a total of 3 on D.
47
Next add 7 to the 8 on G. This makes a total of 15 on FG (Fig. 144).
STEP 3: Add 29 to the 36 on AB.

This makes a total of 65 on AB. Next
add 2 to the 3 on D. This makes a
total of 5 on D. Finally add 11 to the
15 on FG. This makes a total of 26
on FG (Fig. 145).
STEP 4: Now divide the 26 on FG by 12, and add the quotient 2 to the 5 on D, leaving
the remainder 2 on G. This makes a total of 7 on D (Fig. 146).
NOTE: In dividing the 26 on FG, experienced operators generally do not set the divisor
12 on the board.
STEP 5: Finally divide the 7 on D by 3, and

add the quotient 2 to the 65 on AB, leaving the
remainder 1 on D. This makes a total of 67 on AB.
The answer is 67 yd. 1 ft. 2 in. (Fig. 147).
METHOD B: Adding from lower to higher denominations:
STEP 1: Set 8 on unit rod G. Add

7 and 11 to the 8 on G, and you get
a total of 26 on FG (Fig. 148).
STEP 2: Divide the 26 on FG by

12, and set the quotient 2 on D,
which is the proper rod on which to
set the quotient. This gives you 2 on
D and leaves 2 on G (Fig. 149).
STEP 3: Add 1, 2, and 2

successively to the 2 on D. This
makes a total of 7 on D (Fig. 150).
STEP 4: Divide the 7 on D by 3, and set the quotient 2 on B, which is the proper rod on
which to set the quotient. This gives you 2 on B and leaves 1 on D (Fig. 151).
STEP 5: Finally add 21, 15, and 29

successively to the 2 on B. This makes a total of
67 on AB. The answer is 67 yd. 1 ft. 2 in. on AB,
D, and G respectively (Fig. 152).
NOTE: Why are the yards, feet, and inches figures set as they are in this example?
The great advantage of this arrangement of these figures is that the standard methods
of multiplication and division can be used. Notice the use of the standard method of
division in steps 4 and 5 of Method A, and in steps 2 and 4 of Method B in this example.
For further details see Examples 2 and 3, section 2 of this chapter.
48
EXAMPLE 2: Subtract 12 days 17 hours 48 minutes from 25 days 12 hours 43 minutes.
STEP 1: Set 25 on AB, with B as the

unit rod, set 12 on DE, with E as the unit
rod, and set 43 on GH, with H as the unit
rod (Fig. 153).
STEP 2: Subtract 12 from the 25 on AB.

This leaves 13 on AB (Fig. 154).

from the 12 on DE, borrow 1 from B, and subtracting 17 from 24, add 7 to the 12 on DE.
This makes 19 on DE (Fig. 155).
NOTE: In borrowing from B, you may

add 24 to the 12 on DE, and subtract 17
from the sum 36. This gives you the same
result 19.

from the 43 on GH, borrow 1 from the 9
on E, and subtracting 48 from 60, add the
result 12 to the 43 on GH. This makes 55
on GH. Now you have 12 on AB, 18 on DE,
and 55 On GH. The answer is 12 days 18
hours 55 minutes (Fig. 156).
EXAMPLE 3:
In the above example, the second and

fourth compound numbers are to be added,
and the third to be subtracted.
When both additions and subtractions are to be performed, calculations generally

move from higher to lower denominations on each row. But they can also move down
each column—completing each column before moving to the higlier denomination.
METHOD A: Calculating from higher to lower denominations:
STEP 1: Set £95 on BC, with C as the unit

rod, set 7 s. on unit rod F, and set 3 d. on
unit rod I (Fig. 157).
STEP 2: Add 12 to the 95 on BC,

producing 107 on ABC. Add 8 to the 7 on F,
yielding 15 on EF. Add 5 to the 3 on I
producing 8 on I (Fig. 158).
STEP 3: Subtract 7 from the 107 on ABC,

leaving 100 on ABC. As you cannot subtract
49
19 from the 15 on EF, borrow 1 from rod C. Subtracting 19 from 20, add 1 to the 15 on EF.
This gives you 16 on EF and 99 on BC (Fig. 159).
STEP 4: As you cannot subtract 10 from the 8 on I, borrow 1 from the 16 on EF, and
subtracting 10 from 12, add the result 2 to the 8 on I. This gives you 10 on HI and 15 on
EF (Fig. 160).
STEP 5: Now add 26 to the 99 on BC, and you get 125 on ABC.
Add 2 to the 15 on EF, leaving 17 on EF. Add 7 to the 10 on HI
yielding 17 on HI. Finally subtract 12 from this 17 and carry 1 to
the 17 on EF. This gives you 5 on I and 18 on EF. The answer is
£125 18 s. 5 d. (Fig. 161).
METHOD B: Calculating from lower to higher denominations:
An alternative is to calculate column by column from lower to higher denominations.
First do the pence column on HI by means of complementary numbers, yielding 5 d. on

I.
Next do the shillings column on EF by means of complementary numbers, and you get
minus 2 s. on F. So borrow £1 from L95 to rectify this minus quantity, and you get 18 s.
on EF and £94 left on BC.
In subtracting 19 s. from 15 s., you may borrow L1 instead of calculating by means of

complementary numbers.
Finally setting L94 on BC, calculate the pound column, which leaves £ 125 on ABC.
This gives you the same result, £125, 18 s., and 5 d. on ABC, EF, and I respectively.
4. Multiplying Compound Numbers
EXAMPLE 1: Multiply the following compound number by 18: £4 2s. 6d.
On the abacus the computation is made mostly in

the same way as in the following written calculation.
STEP 1: Set 4 on unit rod E, and 2 and 6 on H and K

respectively. Set the multiplier 18 on AB (Fig. 162).
NOTE: Experts, however, often do not actually set

the multiplier.
50
STEP 2: First multiply the 6 on K by 18, and you get

the product 108 on LMN. Now clear K of its 6 (Fig. 163).
STEP 3: Now divide the 108 on LMN by 12 and you

get the quotient 9 on K, having cleared LMN of the
108. Here you should remember the divisor 12
without setting it on the board (Fig. 164).
STEP 4: Next you must multiply the 2 on H by 18,

and add the product 36 to the 9 on K. In this
procedure, first multiply the 1 on A by the 2 on H,
then set the product 2 on J. This gives you 29 on JK.
Next multiply the 8 on B by the same 2 on H; add the
product 16 to the 29 on JK, and clear H of its 2. This
gives you a total of 45 on JK (Fig. 165).
STEP 5: Now divide the 45 on JK by 20, and you get the quotient 2 on H, with 5 left
over on K (Fig. 166).
STEP 6: Finally you must multiply the 4 on E by 18, and add the product 72 to the 2 on
H. In this process, first multiply the 1 on A by the 4 on E, and set the product 4 on G. This
gives you 42 on GH. Next multiply the 8 on B by the same 4 on E, add the product 32 to
the 42 on GH, and clear E of its 4. This gives you a total of 74 on GH. The answer is £74 5
s. (Fig. 167).
NOTE: An alternative method for working this problem is to convert the numerical
figures of this problem to pence, and then to multiply the result by 18. At the end of the
problem the product is reconverted to pounds, shillings, and pence. This method is often
used when the multiplier is a decimal fraction.
5. Dividing Compound Numbers
EXAMPLE 1: Divide the following compound number by 14: £243 2s. 8d.
On the abacus the division is performed

much the same way as in the following
written calculation:
STEP 1: Set 243 on GHI, with I as the unit

rod, and set 2 on unit rod L and 8 on unit rod
O. The divisor 14 may be set either on AB or
on BC, though preferably on the former with
four vacant rods between B and G (Fig. 168).
STEP 2: Divide the 243 on GHI by 14, and

you get the quotient 17 on EF and a
remainder of 5 on I (Fig. 169).
STEP 3: Multiply the 5 on I by 20, and add the product 100 to the 2 on L, clearing the 5
on I, and you get 102 on JKL (Fig. 170).
51
STEP 4: Divide the 102 on JKL by 14, leaving a

quotient of 7 on I and a remainder of 4 on L (Fig.
171).
STEP 5: Multiply the 4 on L by 12, and add the

product 48 to the 8 on O, clearing the 4 on L, and
you get 56 on NO (Fig. 172).
STEP 6: Divide the 56 on NO by 14, and you get

the quotient 4 on L with no remainder. The answer is
£17 7 s. 4 d. (Fig. 173).
NOTES: (a) An alternative method for this

problem is to convert everything to pence and then
to divide the result by 14. This result is converted to
pounds, shillings, and pence. When the figure of the
highest denomination is smaller than the divisor,
division cannot be performed unless it is converted
into that of the lower denomination and makes a
figure larger than the divisor. (b) On the board,
multiplication is performed simultaneously with
addition, and division simultaneously with
subtraction. This means that to multiply numbers is
to add numbers and that to divide numbers is to
subtract numbers. Accordingly, multiplication and
division are all the faster on the board than on paper.
52
VII – EXTRACTING SQUARE ROOTS

There are several methods of finding both the square and cube roots of numbers on
the abacus. They are all adaptations of algebraic methods.
Since the methods of finding cube roots are complicated, and are beyond the scope of
the interest and need of ordinary abacus operators, this book only concerns itself with
three typical methods of extracting square roots.
The first is the exact algebraic method. The second is considered the best by sorne
authoritative experts. The third is the most representative traditional method and is the
most widely accepted by modern abacus operators.
The method of finding the two-figure square root of a number is founded upon the
method of finding the square of an ordinary binomial (or two-membered) expression, (a +
b). The number is analyzed into (a2 + 2ab + b2) as in algebra, and the square root is
obtained.
Let us take the number 625 as an example of finding a two-figure square root.
In all the three methods introduced here, the reader will see that the first step
consists in subtracting the square of a or 20 (= a2 or 400), the next of subtracting 2ab or 2
x 20 x 5 (=200), and the third of subtracting the square of b or 5 (= b2 or 25)
As a refresher, the standard written method of extracting square roots is as follows:
625 is divided into pairs of digits beginning from the decimal point and
moving right and left (thus 18 967.103 is divided 1’89’67.10'30). Then the
largest perfect square which matches the first set of numbers is selected—4
is the largest perfect square smaller than or equal to 6.2—and the square
root of 4, 2, is written both above the 6 and to the side. The product, 4, is
written below the 6. The 4 is subtracted from 6 and the next pair of numbers is brought
down—making 225. The number which is above the square root sign, 2, is doubled and
brought down to the side—here the 4 of the 45 at the side. Then a number is selected, 5,
such that when placed with 4 to make a two-digit number, 45, and that number is
multiplied by the original number selected, 45 x 5, the result is the largest number less
than or equal to the “quotient,” which at this point is 225.
The following analysis will also help the reader to find the square root of 625 on the
abacus. Note thatin the series of algebraic equations given here, a is in substance 20 and
b is 5.
(a+b)2
= a2 + 2ab + b2
= 202 + (2 x 20 x 5) + 52
= 400 + 200 + 25
= 625
Both the abacus method and this written method are based on algebraic equations. In
a simple problem (one involving a two-digit answer) both methods of extracting a square
53
root can be described by the binomial equation (a+b)2 - a2 - 2ab - b2 = 0. In a slightly

more complicated problem (one involving a three-digit answer) both can be described by
the equation (a + b + c)2 - a2 - 2ab - b2 - 2ac - 2bc - c2 = 0. To understand this compare
the examples below, which show the standard written method (left) and the abacus
method (right).
In a trinomial problem this is what happens. At left is the written method and at right
the abacus method.
This means that the numbers handled in the Japanese abacus method as opposed to
the Western written method are smaller and simpler, on the other hand, the number of
manipulations required is larger. Nevertheless, once an operator becomes accustomed to
using this Japanese method he can obviously do a problem faster than he would have
been able to on paper.
EXAMPLE 1: Find the square root of 625.
METHOD A:
STEP 1: Set 625 on EFG, with G as

the unit rod (Fig. 174). Mark off the
digits in sets of twos. The largest
perfect square less than 6 is 4.
STEP 2: Set, on C, 2, the root of the 4. Square this 2, and subtract the product 4 from
the 6 on E. This leaves 225 on EFG (Fig. 175).
NOTE: The 2 on C corresponds to the a of the expression (a2 + 2ab + b2).
STEP 3: Double the 2 on C, and set the product 4 on A (Fig. 176).
NOTE: Now the 4 on A corresponds with the 2a of the algebraic expression.
STEP 4: The 4 on A divides into the 22 on EF five times. Set the second trial quotient
figure 5 on D (Fig. 177).
54
NOTE: The 5 on D corresponds with the b of the algebraic expression.
STEP 5: Multiplying the 4 on A

by the 5 on D, subtract the
product 20 from the 22 on EF. This
leaves 25 on FG (Fig. 178).
NOTE: (4 x 5) corresponds with the 2ab of the expression.
STEP 6: Square the 5 on D, and subtract the product 25 from the 25 on FG. This clears
the board of 625, and shows that the 5 on D is the correct second quotient figure. The
answer is 25 on CD (Fig. 179).
NOTE: 52 or 25 corresponds with the b2 of the expression. Also note that this method
corresponds more strictly to the written algebraic method than the other methods
introduced below.
METHOD B:
STEP 1: Set 625 on CDE, with E

as the unit rod (Fig. 180). Mark off
the digits in sets of twos. The
largest square less than 6 is 4.
STEP 2: Set, on A, 2, the root of the 4. Square this 2, and subtract the product 4 from
the 6 on C. This leaves 225 on CDE (Fig. 181).
STEP 3: Double the 2 on A, making 4 on A (Fig. 182).
STEP 4: The 4 on A goes into the 22 on CD five times. Set the second trial quotient
figure 5 on B (Fig. 183).
NOTE: The 5 on B corresponds with the b of the expression.
STEP 5: Multiply the 4 on A by the 5 on

B, and subtract the product 20 from the
22 on CD. This leaves 25 on DE (Fig. 184).
STEP 6: Square the 5 on B, and subtract

the product 25 from the 25 on DE, and the
board is cleared of the dividend (Fig. 185).
STEP 7: Finally halve the 4 on A into its original digit 2. Our square root is 25 on AB
(Fig. 186).
METHOD C:
STEP 1: Set 625 on CDE, with E as the unit rod (Fig. 187). Mark off the digits in sets of
twos. The largest square less than 6 is 4.
55
STEP 2: Set, on A, 2, the root of the 4. Square this 2, and subtract the product 4 from
the 6 on C. This leaves 225 on CDE (Fig. 188).
NOTE: The 225 on CDE corresponds to the (2ab +b2) of the algebraic expression.
STEP 3: Multiply the 225 on CDE

by 0.5, and set the product 112.5 on
CDEF (Fig. 189).
NOTES: (a) If you find it hard to

do this division mentally, you may first set the product 112.5 on FGHI and later set it on
CDEF. (b) The 112.5 on CDEF corresponds to (ab + b2/ 2) of the algebraic expression.
STEP 4: The 2 on A divides into

the 11 on CD five times. Set the
second trial quotient figure 5 on B
(Fig. 190).
NOTE: The 5 on B corresponds

with the b of the expression.
STEP 5: Multiply the 2 on A by the 5 on B, and subtract the product 10 from the 11 on
CD. This leaves 12.5 on DEF (Fig. 191).
NOTE: The 12.5 on DEF corresponds to b2 / 2 of the expression.
STEP 6: Square the 5 on B mentally, and you get 25. Next either divide this 25 by 2 or
multiply it by 0.5, and you get 12.5. If you find it hard to do this calculation mentally you
can do it on the board.
Now subtract this 12.5 from the 12.5 on DEF. This clears the board of the 625, and
leaves the answer 25 on AB (Fig. 192).
EXAMPLE 2: Find the square root of 4 489 (Method B).
STEP 1: Set 4 489 on DEFG, with

G as the unit rod (Fig. 193). Mark off
the digits in sets of twos. In the first
pair 44, the highest square root is 6.
STEP 2: Set, on B, 6, which is the root of the 44 on DE. Square this 6, and subtract the
product 36 from the 44 on DE. This leaves 889 on EFG (Fig. 194).
STEP 3: Double the 6 on B into 12 on AB (Fig. 195).
STEP 4: The 12 on AB divides into the 88 On EF seven times. Set the second trial
divisor figure 7 on C (Fig. 196).
NOTE: The 7 on C corresponds with the b of the expression.
56
STEP 5: Now multiplying the 12

on AB by the 7 on C, subtract the
product 84 from the 88 on EF. This
leaves 49 on FG (Fig. 197).
NOTE: The 12 on AB multiplied by 7 corresponds with 2ab or (120 x 7).
STEP 6: Next square the 7 on C, and subtract the product 49

from the 49 On FG. This clears the board of the remainder of the
number or the b2 of the expression (Fig. 198).
STEP 7: Finally halve the 12 on AB into its original 6 on B. Our

square root is 67 on BC (Fig. 199).
EXAMPLE 3: Find the square root of 119 716.
The method of finding the three-digit square root of a number is founded upon the
algebraic method of finding the square of an ordinary trinomial (or three-membered)
expression, (a + b + c). A number is analyzed as in algebra, into (a2 + 2ab + b2 + 2ac + 2bc
+ c2) and the square root is found.
STEP 1: Set 119 716 on CDEFGH, and mark off the digits in sets of twos. In the first
pair 11, the highest square root is 3 (Fig. 200).
STEP 2: Set, on A, 3, the first trial figure

in the root. Square the 3, and subtract the
product 9 from the 11 on CD. This leaves 29
716 on DEFGH (Fig. 201).
NOTE: 3 corresponds to the a of the algebraic expression (a2 + 2ab + b2 + 2ac + 2bc +
2
c ).
STEP 3: Double the 3 on A making 6 on A (Fig. 202).
STEP 4: Figuring how many times the 6

on A goes into the 29 on DE, we find that 4
is correct. So we set 4 on B as the second
trial figure in the root (Fig. 203).
NOTE: The 4 on B corresponds to the b of the algebraic expression.
STEP 5: Multiply the 6 on A by the 4 on

13, and subtract the product 24 from the
29 on DE. This leaves 5 716 on EFGH (Fig.
204).
NOTE: (4 x 6) or 24 corresponds to the 2ab of the algebraic expression.
STEP 6: Now square the 4 on 13, and subtract the product 16 from the 57 on EF. This
leaves 4 116 on EFGH (Fig. 205).
57
NOTE: The square of 4 corresponds to the b2 of the algebraic expression.
STEP 7: Double the 4 on B, making 8 (Fig. 206).
NOTE: When the second trial divisor

figure is 5 or larger, move the first divisor
figure to the first rod to the left, and set
double the second trial quotient figure on
AB.
STEP 8: Figuring how many times the 6 on A goes into the 41 on EF, we find that 6 is
the proper figure. Set 6 on C as the third trial figure in the root (Fig. 207).
NOTE: The 6 on C corresponds to c of the expression.
STEP 9: Multiply the 6 on A by the 6 on C,

and subtract the product 36 from the 41 on EF.
This leaves 516 on FGH (Fig. 208).
NOTE: 6 x 6 corresponds to the 2ac of the

algebraic expression.
STEP 10: Multiply the 8 on B by the same 6 on C, and subtract the product 48 from the
51 on FG. This leaves you 36 on GH (Fig. 209).
NOTE: (8 on B x 6 on C) corresponds to the 2bc of the algebraic expression.
STEP 11: Square the 6 on C, and subtract

the product 36 from the 36 on GH. This clears
the board of the remainder of the number,
which corresponds to the c2 of the algebraic
expression (Fig. 210).
STEP 12: Now halve the 6 on A and the 8 on B into the original 3 and 4 respectively.
Our square root is 346 on ABC (Fig. 211).
NOTE: The 68 which was on AB corresponded to the 2ab of the algebraic expression.
58
VIII – MORE EXERCISES

The exercises for the first, second, and third grades were actually used in recent
national abacus license examinations and are here reprinted through the courtesy of the
Japan Chamber of Commerce and Industry.
1. Eighth-Grade Operator
Group A
(1 set per minute, or entire group with 70% accuracy in 10 minutes)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
62 35 71 20 86 14 90 57 82 49
15 70 45 38 92 20 26 69 54 80
40 29 36 92 50 97 -83 32 -73 21
71 86 50 13 29 65 59 70 -38 68
84 43 -92 56 -74 28 42 31 10 32
26 58 -27 90 51 73 -15 46 29 14
97 10 83 84 63 41 -60 18 76 95
80 64 19 47 -30 56 47 94 -61 70
39 17 -60 75 -48 30 31 25 40 53
53 92 84 61 17 89 78 80 95 76
567 504 209 576 236 513 215 522 214 558
Group B Group C
(1) 45 x 8 = 360 (1) 72 / 2 = 36
(2) 73 x 4 = 292 (2) 135 / 5 = 27
(3) 52 x 2 = 104 (3) 246 / 3 = 82
(4) 38 x 7 = 266 (4) 126 / 9 = 14
(5) 94 x 3 = 282 (5) 413 / 7 = 59
(6) 15 x 6 = 90 (6) 300 / 4 = 75
(7) 86 x 9 = 774 (7) 744 / 8 = 93
(8) 67 x 5 = 335 (8) 288 / 6 = 48
(9) 21 x 8 = 168 (9) 183 / 3 = 61
(10) 79 x 6 = 474 (10) 435 / 5 = 87
(11) 134 x 3 = 402 (11) 8 829 / 9 = 981
(12) 509 x 2 = 1 018 (12) 420 / 4 = 105
(13) 876 x 8 = 7 008 (13) 2 634 / 6 = 439
(14) 623 x 4 = 2 492 (14) 752 / 2 = 376
(15) 258 x 9 = 2 322 (15) 4 336 / 8 = 542
(16) 480 x 6 = 2 880 (16) 4 050 / 5 = 810
(17) 942 x 5 = 4 710 (17) 1 906 / 2 = 953
(18) 715 x 7 = 5 005 (18) 2 124 / 3 = 708
(19) 301 x 4 = 1 204 (19) 1 068 / 4 = 267
(20) 697 x 3 = 2 091 (20) 4 368 / 7 = 624
59
2. Seventh-Grade Operator
Assortment 1
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
45 63 87 31 56 94 26 79 70 12
78 95 60 19 27 70 98 16 45 60
30 59 31 40 68 19 -20 84 61 85
59 -80 48 95 90 51 -57 40 13 -32
61 -43 24 -87 31 35 41 62 90 -78
96 78 12 -28 47 60 93 97 24 45
20 51 39 30 82 89 -65 23 57 90
83 -27 50 72 43 46 10 95 38 76
52 82 67 16 10 18 74 60 47 -53
17 90 29 -57 24 42 58 35 96 10
93 -64 15 20 93 37 -89 21 80 34
40 -37 98 65 51 68 -13 74 12 -89
62 10 54 -43 80 27 30 83 35 -71
81 46 70 -64 69 50 76 18 26 94
74 12 36 98 75 23 42 50 89 67
891 335 720 207 846 729 304 837 783 250
Group B Group C
(1) 7 832 x 6 = 46 992 (1) 402 / 2 = 201
(2) 4 096 x 7 = 28 672 (2) 6 776 / 7 = 968
(3) 6 348 x 8 = 50 784 (3) 3 282 / 6 = 547
(4) 1 573 x 4 = 6 292 (4) 2 430 / 3 = 810
(5) 8 105 x 9 = 72 945 (5) 5 536 / 8 = 692
(6) 2 619 x 3 = 7 857 (6) 1 380 / 4 = 345
(7) 5 980 x 6 = 35 880 (7) 3 180 / 6 = 530
(8) 3 467 x 5 = 17 335 (8) 7 101 / 9 = 789
(9) 9 024 x 9 = 81 216 (9) 3 408 / 8 = 426
(10) 2 751 x 2 = 5 502 (10) 865 / 5 = 173
(11) 3 647 x 6 = 21 882 (11) 3 896 / 4 = 974
(12) 9 502 x 4 = 38 008 (12) 966 / 7 = 138
(13) 6 138 x 3 = 18 414 (13) 1 184 / 2 = 592
(14) 8 420 x 9 = 75 780 (14) 4 830 / 6 = 805
(15) 4 975 x 2 = 9 950 (15) 807 / 3 = 269
(16) 1 769 x 8 = 14 152 (16) 2 250 / 5 = 450
(17) 7 081 x 4 = 28 324 (17) 2 468 / 4 = 617
(18) 2 153 x 2 = 4 306 (18) 682 / 2 = 341
(19) 5 304 x 5 = 26 520 (19) 5 648 / 8 = 706
(20) 8 296 x 7 = 58 072 (20) 7 407 / 9 = 823
60
Assortment 2
Group A
(70% accuracy, 10 minutes)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
61 47 21 86 34 53 95 78 16 41
72 16 90 47 68 61 59 40 83 68
30 28 64 91 -50 42 10 61 -24 90
84 60 -39 20 -37 90 -46 35 -70 42
12 59 -70 35 64 57 -89 84 25 37
89 32 94 58 15 32 71 27 93 56
93 75 18 24 40 18 20 95 41 27
10 39 56 10 -83 25 64 50 -67 10
56 40 -82 73 29 70 -23 32 80 96
34 18 25 69 71 64 87 19 35 53
59 83 43 26 97 86 30 94 78 39
67 71 80 49 -21 39 15 87 -46 28
75 50 -67 18 -60 80 -78 23 -59 70
20 94 -71 30 58 74 -62 16 90 14
48 26 53 75 92 91 34 60 12 85
810 738 215 711 317 882 187 801 287 756
Group B Group C
(1) 2 604 x 3 = 7 812 (1) 4 077 / 9 = 453
(2) 5 219 x 7 = 36 533 (2) 2 667 / 7 = 381
(3) 8 045 x 5 = 4 0 225 (3) 1 612 / 2 = 806
(4) 1 573 x 6 = 9438 (4) 2 810 / 5 = 562
(5) 3 987 x 2 = 7 974 (5) 882 / 3 = 294
(6) 6 820 x 4 = 27 280 (6) 5 096 / 7 = 728
(7) 9 751 x 9 = 87 759 (7) 680 / 4 = 170
(8) 7 406 x 8 = 59 248 (8) 957 / 3 = 319
(9) 4 132 x 3 = 12 396 (9) 7 240 / 8 = 905
(10) 9 368 x 7 = 65 576 (10) 3 882 / 6 = 647
(11) 6 972 x 6 = 41 832 (11) 5 481 / 7 = 783
(12) 3 285 x 3 = 9 855 (12) 2 100 / 6 = 350
(13) 5 049 x 2 = 10 098 (13) 6 246 / 9 = 694
(14) 1 856 x 4 = 7 424 (14) 1 195 / 5 = 239
(15) 8 391 x 8 = 67 128 (15) 3 488 / 4 = 872
(16) 4 703 x 9 = 42 327 (16) 435 / 3 = 145
(17) 2 168 x 4 = 8 672 (17) 1 054 / 2 = 527
(18) 9 027 x 3 = 27 081 (18) 900 / 5 = 180
(19) 7 430 x 7 = 52 010 (19) 7 744 / 8 = 968
(20) 6 514 x 5 = 32 570 (20) 832 / 2 = 416
61
3. Sixth-Grade Operator
Assortment 1
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
35 64 9 017 781 29 83 459 6 082 54 207
409 5 013 928 350 67 962 104 94 592 18
852 937 241 19 8 024 38 -76 350 9 364 173
14 790 -58 28 71 5 093 -395 71 -108 14
3 267 28 -703 974 -438 126 7 201 259 -46 3 059
198 576 342 6 831 -92 40 68 37 15 86
41 102 96 402 684 57 890 25 687 542
620 58 430 56 305 814 24 68 30 69
93 4 681 73 98 586 71 16 1 704 879 250
8 065 25 16 34 90 2 708 -73 219 13 67
47 872 65 2 097 -1 632 345 -248 73 205 938
971 34 -809 516 -741 96 -3 091 854 -36 8 401
53 609 -7 521 45 -53 607 67 380 -4 071 97
702 91 -86 27 910 54 582 69 -728 45
86 43 54 603 57 219 35 146 92 623
15 453 13 923 2 085 12 861 7 867 11 313 5 563 10 431 6 942 14 589
Group B Group C
(1) 46 x 754 = 34 684 (1) 1288 / 56 = 23
(2) 703 x 13 = 9 139 (2) 1568 / 98 = 16
(3) 28 x 972 = 27 216 (3) 1554 / 21 = 74
(4) 159 x 69 = 10 971 (4) 1505 / 35 = 43
(5) 82 x 407 = 33 374 (5) 4030 / 62 = 65
(6) 514 x 18 = 9 252 (6) 4324 / 47 = 92
(7) 37 x 230 = 8 510 (7) 2573 / 83 = 31
(8) 950 x 51 = 48 450 (8) 1102 / 19 = 58
(9) 64 x 805 = 51 520 (9) 1406 / 74 = 19
(10) 821 x 36 = 29 556 (10) 8 004 / 92 = 87
(11) 798 x 84 = 67 032 (11) 516 / 43 = 12
(12) 45 x 289 = 13 005 (12) 918 / 27 = 34
(13) 172 x 52 = 8 944 (13) 5 472 / 72 = 76
(14) 630 x 65 = 40 950 (14) 1 302 / 14 = 93
(15) 84 x 906 = 76 104 (15) 3 936 / 89 = 41
(16) 209 x 47 = 9 823 (16) 4 930 / 58 = 85
(17) 51 x 743 = 37 893 (17) 5 963 / 89 = 67
(18) 306 x 19 = 5 814 (18) 899 / 31 = 29
(19) 97 x 308 = 29 876 (19) 3 640 / 65 = 56
(20) 63 x 261 = 16 443 (20) 2 592 / 54 = 48
62
Assortment 2
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
586 41 873 268 184 725 92 308 657 468
93 705 345 981 510 86 743 96 14 274
610 539 62 705 61 97 10 473 280 59
139 285 204 -53 647 830 451 89 906 412
304 129 583 -178 79 15 -809 21 45 160
78 -450 691 94 803 957 -65 140 -12 358
42 -683 78 360 52 401 586 897 -138 907
175 -42 306 412 923 648 31 205 -879 13
762 801 452 -29 490 312 708 34 923 76
407 93 18 -506 735 60 -342 516 498 802
863 217 790 -635 96 502 -74 925 60 35
51 -964 29 840 318 243 -623 650 327 791
920 -78 145 27 574 89 290 874 -705 623
98 360 67 491 26 173 968 731 -46 580
245 76 910 73 208 694 157 62 531 49
5 373 1 029 5 553 2 850 5 706 5 832 2 123 6 021 2 461 5 607
Group B Group C
(1) 403 x 73 = 29 419 (1) 888 / 74 = 12
(2) 26 x 208 = 5 408 (2) 782 / 23 = 34
(3) 589 x 74 = 43 586 (3) 3 953 / 59 = 67
(4) 81 x 861 = 69 741 (4) 476 / 17 = 28
(5) 170 x 49 = 8 330 (5) 7 990 / 85 = 94
(6) 68 x 617 = 41 956 (6) 2 576 / 46 = 56
(7) 945 x 14 = 13 230 (7) 6 643 / 91 = 73
(8) 307 x 32 = 9 824 (8) 1 558 / 38 = 41
(9) 69 x 506 = 34 914 (9) 5 518 / 62 = 89
(10) 72 x 985 = 70 920 (10) 945 / 27 = 35
(11) 34 x 194 = 6 596 (11) 4 335 / 85 = 51
(12) 512 x 83 = 42 496 (12) 3 570 / 42 = 85
(13) 49 x 508 = 24 892 (13) 882 / 14 = 63
(14) 301 x 27 = 8 127 (14) 2 652 / 68 = 39
(15) 758 x 91 = 68 978 (15) 7 068 / 93 = 76
(16) 93 x 609 = 56 637 (16) 5 586 / 57 = 98
(17) 206 x 35 = 7 210 (17) 783 / 29 = 27
(18) 87 x 402 = 34 974 (18) 994 / 71 = 14
(19) 65 x 253 = 16 445 (19) 1 512 / 36 = 42
(20) 124 x 76 = 9 424 (20) 684 / 19 = 36
63
4. Fifth-Grade Operator
Assortment 1
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
576 8 294 603 3 918 702 423 186 307 9 684 237
2 089 435 481 706 9 318 357 534 1 273 509 7 815
961 3 108 894 327 529 4 108 129 918 763 592
846 912 5 021 986 437 689 9 705 134 -2 157 346
417 593 372 8 031 -654 1 594 671 8 059 -406 954
6 752 237 9 140 795 -231 380 830 462 -813 5 081
408 1 689 975 641 -3 048 516 -3 284 391 6 045 478
7 813 860 256 4 823 197 7 634 -157 4 907 526 643
359 7 046 -417 590 6 851 290 5 021 746 270 9 075
9 210 572 -8 306 254 960 825 789 598 3 481 380
783 904 421 1 085 -893 2 067 -493 7 610 934 162
5 034 745 6 359 143 -2 786 942 -7 260 852 -317 6 239
192 8 061 -738 562 402 8 759 -948 625 -8 059 170
624 753 -2 069 476 1 975 301 6 402 583 298 601
305 621 -847 7 209 064 176 365 2 064 172 4 928
36 369 34 830 12 045 30 546 18 823 29 061 12 490 29 529 10 930 37 701
Group B Group C
(1) 601 x 138 = 82 938 (1) 9 180 / 108 = 85
(2) 247 x 375 = 92 625 (2) 8 232 / 42 = 196
(3) 5 630 x 87 = 489 810 (3) 50 934 / 653 = 78
(4) 415 x 746 = 309 590 (4) 49 282 / 82 = 601
(5) 879 x 213 = 187 227 (5) 15 141 / 309 = 49
(6) 358 x 601 = 215 158 (6) 28 917 / 51 = 567
(7) 926 x 950 = 879 700 (7) 66 822 / 74 = 903
(8) 104 x 594 = 61 776 (8) 26 236 / 937 = 28
(9) 72 x 8 062 = 580 464 (9) 7 550 / 25 = 302
(10) 983 x 429 = 421 707 (10) 30 784 / 416 = 74
(11) 459 x 946 = 434 214 (11) 14 196 / 39 = 364
(12) 703 x 217 = 152 551 (12) 8 333 / 641 = 13
(13) 198 x 683 = 135 234 (13) 5 508 / 204 = 27
(14) 942 x 825 = 777 150 (14) 70 590 / 78 = 905
(15) 61 x 3 901 = 237 961 (15) 45 568 / 512 = 89
(16) 205 x 472 = 96 760 (16) 51 085 / 85 = 601
(17) 862 x 730 = 629 260 (17) 9 280 / 16 = 580
(18) 3 807 x 69 = 262 683 (18) 65 016 / 903 = 72
(19) 174 x 504 = 87 696 (19) 22 372 / 47 = 476
(20) 536 x 158 = 84 688 (20) 31 850 / 325 = 98
64
Assortment 2
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
1 063 274 319 4 657 530 6 841 805 956 728 4 059
542 8 029 753 208 3 714 265 154 7 209 156 236
951 367 8 406 431 862 1 079 7 061 567 4 809 941
674 3 510 837 -1 629 495 682 142 498 237 1 308
8 502 781 264 -370 7 023 236 869 3 015 416 179
349 -608 1 472 593 971 8 903 237 947 6 038 897
175 -2 153 950 8 042 839 579 9 740 572 345 2 610
928 -467 3 029 416 9 506 713 -392 1 086 -293 798
6 730 892 145 359 648 2 054 -128 621 -9 507 562
481 9 046 581 6 283 187 928 -5 093 835 184 8 205
7 209 431 936 107 2 368 435 817 2 690 451 386
856 584 7 620 -925 240 197 6 284 723 7 069 463
293 -5 670 814 -768 571 4 780 356 134 -273 5 017
4 018 -239 6 095 -5 890 124 305 -3 479 408 -690 423
367 915 728 174 6 059 146 -605 8 341 -5 182 574
33 138 15 692 32 949 11 688 34 137 28 143 16 768 28 602 4 488 26 658
Group B Group C
(1) 723 x 815 = 589 245 (1) 51 874 / 74 = 701
(2) 38 x 4 079 = 155 002 (2) 45 301 / 509 = 89
(3) 901 x 687 = 618 987 (3) 14 118 / 26 = 543
(4) 467 x 942 = 439 914 (4) 42 151 / 691 = 61
(5) 235 x 351 = 82 485 (5) 7 585 / 37 = 205
(6) 6 702 x 14 = 93 828 (6) 8 676 / 482 = 18
(7) 149 x 530 = 78 970 (7) 29 056 / 908 = 32
(8) 584 x 603 = 352 152 (8) 3 302 / 13 = 254
(9) 816 x 296 = 241 536 (9) 37 060 / 85 = 436
(10) 950 x 728 = 691 600 (10) 68 482 / 706 = 97
(11) 257 x 498 = 127 986 (11) 11 856 / 624 = 19
(12) 183 x 532 = 97 356 (12) 40 356 / 57 = 708
(13) 49 x 6 705 = 328 545 (13) 52 136 / 931 = 56
(14) 802 x 103 = 82 606 (14) 33 615 / 405 = 83
(15) 716 x 789 = 564 924 (15) 59 512 / 86 = 692
(16) 364 x 217 = 78 988 (16) 2 856 / 14 = 204
(17) 9 408 x 31 = 261 648 (17) 26 274 / 302 = 87
(18) 521 x 856 = 445 976 (18) 8 428 / 28 = 301
(19) 390 x 264 = 102 960 (19) 77 104 / 79 = 976
(20) 675 x 940 = 634 500 (20) 6 885 / 153 = 45
65
Assortment 3
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
741 2 908 456 1 063 601 397 8 509 516 907 371
6 102 847 321 759 2 738 206 971 4 089 132 653
819 758 9 604 3 941 573 421 128 934 8 021 492
4 057 3 416 259 528 697 1 035 613 5 129 346 6 104
938 673 512 709 528 647 2 760 761 238 589
692 -105 7 364 247 7 089 531 597 830 -176 4 068
8 720 -369 947 5 482 316 785 381 6 208 -2 759 815
143 -5 280 396 620 -8 172 4 190 435 792 -681 7 920
9 567 721 4 018 135 -450 928 -7 043 831 3 460 9 173
285 954 130 8 062 849 6 273 -816 627 594 258
306 -1 470 685 594 9 403 806 564 7 283 -815 304
928 -536 8 207 317 -961 5 469 9 308 345 -5 092 167
5 034 182 793 486 -245 714 -726 590 4 978 5 049
476 6 093 278 901 -3 016 852 -4 952 1 054 703 726
315 942 1 850 6 873 254 3 098 -240 476 564 832
39 123 9 734 35 820 30 717 10 204 26 352 10 489 30 465 10 420 37 521
Group B Group C
(1) 415 x 948 = 393 420 (1) 55 647 / 81 = 687
(2) 952 x 852 = 811 104 (2) 56 870 / 605 = 94
(3) 381 x 176 = 67 056 (3) 7 599 / 149 = 51
(4) 64 x 6 307 = 403 648 (4) 36 868 / 52 = 709
(5) 890 x 219 = 194 910 (5) 2 808 / 216 = 13
(6) 167 x 563 = 94 021 (6) 67 200 / 75 = 896
(7) 523 x 490 = 256 270 (7) 8 289 / 307 = 27
(8) 7 048 x 75 = 528 600 (8) 41 495 / 43 = 965
(9) 279 x 184 = 79 236 (9) 28 952 / 94 = 308
(10) 306 x 301 = 92 106 (10) 5 376 / 128 = 42
(11) 675 x 369 = 249 075 (11) 18 666 / 306 = 61
(12) 1 948 x 54 = 105 192 (12) 7 605 / 15 = 507
(13) 403 x 890 = 358 670 (13) 30 848 / 64 = 482
(14) 21 x 4 701 = 98 721 (14) 7 884 / 219 = 36
(15) 906 x 548 = 496 488 (15) 72 071 / 743 = 97
(16) 587 x 132 = 77 484 (16) 82 156 / 92 = 893
(17) 862 x 607 = 523 234 (17) 34 983 / 507 = 69
(18) 359 x 286 = 102 674 (18) 57 024 / 81 = 704
(19) 730 x 913 = 666 490 (19) 8 136 / 452 = 18
(20) 124 x 725 = 89 900 (20) 7 790 / 38 = 205
66
5. Fourth-Grade Operator
Assortment 1
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
629 7 084 13 472 8 563 698 40 615 293 5 471 90 386 3 257
3 167 653 4 081 795 5 907 123 6 748 186 419 60 382
82 036 3 471 958 6 041 654 7 850 913 42 305 3 504 935
578 49 680 65 412 359 3 761 584 1 095 -627 927 78 106
9 247 -8 163 7 504 7 261 80 342 1 609 78 932 -6 741 4 036 -679
15 402 -794 963 50 178 531 90 476 689 20 567 71 285 -1 538
139 162 80 721 98 432 4 270 6 398 17 504 492 598 -82 041
8 576 27 905 3 852 -7 510 65 492 213 9 827 91 058 32 960 467
30 951 8 741 139 -42 087 789 27 064 451 -9 683 4 187 50 219
745 91 036 26 598 -623 10 928 947 82 076 -539 651 -4 983
2 810 56 327 9 605 39 401 5 163 54 801 1 463 -31 870 14 706 -759
71 684 209 437 896 73 095 1 326 60 852 5 324 3 542 27 130
493 -852 70 216 -8 304 817 37 892 3 174 908 80 269 6 814
4 508 -30 915 8 647 -712 1204 935 52 630 3 216 723 4 096
60 392 -2 548 390 26 945 42 386 8 752 405 80 947 5 871 425
291357 201996 292995 179653 296037 279585 317052 201014 314064 141831
Group B Group C
(1) 4 321 x 607 = 2 622 847 (1) 130 977 / 567 = 231
(2) 9 108 x 938 = 8 543 304 (2) 440 700 / 975 = 452
(3) 7 685 x 452 = 3 473 620 (3) 336 226 / 341 = 986
(4) 2 934 x 120 = 352 080 (4) 81 624 / 456 = 179
(5) 5 017 x 543 = 2 724 231 (5) 622 800 / 720 = 865
(6) 8 652 x 789 = 6 826 428 (6) 67 402 / 134 = 503
(7) 3 476 x 204 = 709 104 (7) 502 854 / 802 = 627
(8) 1 809 x 671 = 1 213 839 (8) 90 746 / 289 = 314
(9) 6 523 x 316 = 2 061 268 (9) 294 240 / 613 = 480
(10) 7 940 x 895 = 7 106 300 (10) 724 584 / 908 = 798
(11) 2 463 x 271 = 667 473 (11) 821 215 / 851 = 965
(12) 6 289 x 613 = 3 855 157 (12) 99 144 / 408 = 243
(13) 3 045 x 492 = 1 498 140 (13) 493 425 / 675 = 731
(14) 8 251 x 507 = 4 183 257 (14) 77 089 / 127 = 607
(15) 7 836 x 324 = 2 538 864 (15) 31 968 / 296 = 108
(16) 5 970 x 749 = 4 471 530 (16) 248 992 / 502 = 496
(17) 9 012 x 830 = 7 479 960 (17) 84 940 / 310 = 274
(18) 4 758 x 186 = 884 988 (18) 567 207 / 963 = 589
(19) 8 617 x 905 = 7 798 385 (19) 136 090 / 439 = 310
(20) 1 394 x 658 = 917 252 (20) 667 968 / 784 = 852
67
Assortment 2
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
4 068 371 93 416 8 506 74 289 1 903 6 081 25 937 813 56 209
975 30 142 6 904 75 243 315 246 943 3 216 51 402 392
43 106 5 489 21 586 759 1 082 36 152 3 257 64 108 2 769 60 528
1 279 901 7 843 60 138 471 580 87 109 967 80 624 137
50 813 6 492 397 -7 865 16 208 29 365 548 -7 091 6 157 7 630
497 72 054 6 052 -971 4 539 5 078 20 816 -459 47 098 -293
7160 -523 87 931 1 084 293 8 914 9 457 620 834 -8 145
9 582 -64 370 728 32 416 90 614 57 681 872 30 415 3 571 -94 028
251 -7 865 240 -9 302 5 367 409 5 139 732 18 290 3 684
86 304 218 5 072 -615 86 970 1 853 726 -54 803 419 10 957
629 45 806 53 419 -48 127 9 156 60 729 43 065 -1 278 2 365 479
20 374 -9 762 805 793 304 498 2 493 -564 942 -4 851
8 735 -689 70 693 6 540 762 7 031 76 340 8 925 6 087 -713
281 10 935 2 581 924 8 045 247 629 90 183 356 5 046
95 346 8 317 164 20 398 57 823 46 372 15 084 7 846 79 503 81 762
329400 97516 357831 139921 356238 257058 272259 168754 301230 118794
Group B Group C
(1) 9 634 x 426 = 4 104 084 (1) 346 800 / 850 = 408
(2) 2 761 x 593 = 1 637 273 (2) 64 390 / 274 = 235
(3) 4 508 x 180 = 811 440 (3) 419 152 / 536 = 782
(4) 7 985 x 867 = 6 922 995 (4) 89 027 / 701 = 127
(5) 3 120 x 305 = 951 600 (5) 80 850 / 165 = 490
(6) 6 842 x 751 = 5 138 342 (6) 250 158 / 482 = 519
(7) 1 076 x 918 = 987 768 (7) 130 472 / 347 = 376
(8) 8 453 x 472 = 3 989 816 (8) 586 112 / 608 = 964
(9) 5 917 x 609 = 3 603 453 (9) 787 319 / 923 = 853
(10) 3 209 x 234 = 750 906 (10) 95 559 / 159 = 601
(11) 6 013 x 801 = 4 816 413 (11) 368 193 / 623 = 591
(12) 3 497 x 487 = 1 703 039 (12) 26 520 / 195 = 136
(13) 9 084 x 712 = 6 467 808 (13) 216 832 / 308 = 704
(14) 4 526 x 206 = 932 356 (14) 603 880 / 974 = 620
(15) 8 761 x 658 = 5 764 738 (15) 140 322 / 546 = 257
(16) 2 905 x 923 = 2 681 315 (16) 666 462 / 831 = 802
(17) 5 430 x 165 = 895 950 (17) 88 665 / 257 = 345
(18) 1 258 x 594 = 747 252 (18) 402 047 / 409 = 983
(19) 3 872 x 370 = 1 432 640 (19) 326 876 / 782 = 418
(20) 7 619 x 439 = 3 344 741 (20) 469 090 / 610 = 769
68
Assortment 3
Group A
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
2 819 314 80 593 4 295 63 017 7 694 579 18 024 9 072 798
15 038 2 639 921 50 973 4 361 853 4 106 9 132 265 34 019
946 708 1 650 391 25 703 4 081 95 248 389 21 403 5 628
71 405 4 512 789 -8 406 892 725 134 65 708 561 60 937
571 98 056 6 432 -829 7 210 28 139 3 079 -8 421 70 134 146
2 795 67 321 958 37 146 982 4 506 18 620 -968 8 372 -4 503
346 -485 29 076 1 065 80 124 13 472 493 90 614 45 031 -654
49 630 -7 249 513 94 752 6 873 956 7 851 2 837 483 -52 890
7 512 -30 967 71 408 218 58 907 2 780 89 315 -756 2 917 7 652
86 057 8 293 4 817 -15 630 764 91 352 407 -36 571 596 21 345
983 51 420 43 160 -728 9 658 6 041 1 986 -4 720 13 609 761
6 108 -137 5 249 -6 203 435 467 50 732 945 4 928 -1 803
729 -6 905 763 871 42 059 36 908 276 51 092 86 745 -472
20 463 584 67 024 60 534 913 219 6 023 7 653 680 97 081
8 234 10 876 2 385 9 487 1 546 70 583 42 568 304 7 859 8 293
273636 198980 315738 227936 303444 268776 321417 195262 272655 176338
Group B Group C
(1) 1 975 x 419 = 827 525 (1) 93 063 / 463 = 201
(2) 4 702 x 165 = 775 830 (2) 653 650 / 850 = 769
(3) 6 819 x 382 = 2 604 858 (3) 45 310 / 197 = 230
(4) 8 431 x 674 = 5 682 494 (4) 141 360 / 304 = 465
(5) 3 587 x 503 = 1 804 261 (5) 546 336 / 672 = 813
(6) 1 064 x 926 = 985 264 (6) 894 132 / 918 = 974
(7) 5 293 x 708 = 3 747 444 (7) 102 910 / 205 = 502
(8) 7 648 x 853 = 6 523 744 (8) 72 954 / 386 = 189
(9) 9 320 x 497 = 4 632 040 (9) 337 608 / 521 = 648
(10) 2 056 x 201 = 413 256 (10) 267 393 / 749 = 357
(11) 3 674 x 159 = 584 166 (11) 213 852 / 502 = 426
(12) 7 831 x 840 = 6 578 040 (12) 74 889 / 471 = 159
(13) 9 502 x 765 = 7 269 030 (13) 756 286 / 943 = 802
(14) 1 425 x 238 = 339 150 (14) 199 080 / 316 = 630
(15) 2 368 x 406 = 961 408 (15) 694 848 / 704 = 987
(16) 5 093 x 687 = 3 498 891 (16) 97 552 / 268 = 364
(17) 8 956 x 924 = 8 275 344 (17) 498 261 / 921 = 541
(18) 2 709 x 173 = 468 657 (18) 84 525 / 805 = 105
(19) 6 147 x 302 = 1 856 394 (19) 524 286 / 657 = 798
(20) 4 081 x 591 = 2 411 871 (20) 37 674 / 138 = 273
69
The National Examination for the Third-Grade Abacus License
Held under the auspices of the Japan Chamber of Commerce and Industry
Group A
(Entire group of ten sets with 70% accuracy in ten minutes)
(1) (2) (3) (4) (5)

20 139 954 638 7 524 37 258
826 360 728 780 241 301 492 1 630
432 501 46 170 2 516 59 710 849 075
96 317 8 397 37 690 834 26 914
4 620 27 816 -8 215 6 185 -604 132
81 453 819 605 -496 507 720 943 -98 206
708 319 3 547 -21 365 62 397 715
5 492 64 052 619 958 3 854
975 573 84 035 36 270 179 038
67 154 945 610 7 283 917 846 43 927
849 560 8 324 209 714 513 -879
2 738 21 496 -95 480 6 905 -50 186
683 952 -347 80 673 -2 937
58 027 102 873 57 981 654 812 564
907 614 38 109 430 692 28 041 624 105
3 236 418 2 449 206 989 495 2 885 103 1 010 740
(6) (7) (8) (9) (10)

406 725 6 208 872 5 930 18 357
157 37 461 46 237 20 718 9 041
54 860 801 537 695 018 463 802 378 260
3 746 972 79 561 549 56 834
760 912 -42 796 6 409 -87 056 672
538 -194 083 84 357 -109 867 29 138
28 301 -8 615 360 172 97 385 710 953
804 139 23 590 483 8 476 5 280
6 852 314 51 290 714 429
19 026 260 475 3 625 689 502 69 704
497 7 132 976 108 -54 923 804 519
96 813 -56 084 741 -361 235
570 289 -392 29 083 -1 274 1 627
31 794 904 875 4 539 246 013 48 706
2 435 69 581 508 412 93 251 963 415
2 787 084 1 810 175 2 846 907 1 372 859 3 097 170
70
Group B Group C
(70 % accuracy, 10 minutes. Calculate (70 % accuracy, 10 minutes. Calculate
problems 1-10 to the nearest thousandth; problems 1-10 to the nearest thousandth;
11-20 to the nearest yen.) 11-20 to the nearest yen.)
(1) 2 931 x 819 = 2 400 489 (1) 172 894 / 631 = 274
(2) 6 052 x 235 = 1 422 220 (2) 0.20592 / 3.96 = 0.052
(3) 84.67 x 90.7 = 7 679.569 (3) 361 444 / 829 = 436
(4) 9 234 x 572 = 5 281 848 (4) 76.296 / 408 = 0.187
(5) 70 496 x 86 = 6 062 656 (5) 50.735 / 0.073 = 695
(6) 1 507 x 0.461 = 694.727 (6) 186 296 / 584 = 319
(7) 218 x 1 059 = 230 862 (7) 212.795 / 26.5 = 8.03
(8) 0.3875 x 6.48 = 2.511 (8) 0.47693 / 0.917 = 0.520
(9) 5 689 x 3.24 = 18 432.36 (9) 639 184 / 7 024 = 91
(10) 0.4103 x 0.073 = 0.030 (10) 11 526 / 15 = 768.4
(11) ¥6 379 x 108 = ¥688 932 (11) ¥324 563 / 463 = ¥701
(12) ¥2 014 x 0.429 = ¥864 (12) ¥185 / 0.625 = ¥296
(13) ¥1 493 x 963 = ¥1 437 759 (13) ¥646 919 / 7 109 = ¥91
(14) ¥9 584 x 0.875 = ¥8 386 (14) ¥455 / 0.547 = ¥832
(15) ¥765 x 504.8 = ¥386 172 (15) ¥341 105 / 85 = ¥4 013
(16) ¥2 301 x 637 = ¥1 465 737 (16) ¥66 / 0.176 = ¥375
(17) ¥40 678 x 91 = ¥3 701 698 (17) ¥78 196 / 90.4 = ¥865
(18) ¥5 896 x 20.5 = ¥120 868 (18) ¥282 048 / 312 = ¥904
(19) ¥8 125 x 0.736 = ¥5 980 (19) ¥1 820 / 2.08 = ¥875
(20) ¥3 207 x 214 = ¥686 298 (20) ¥602 196 / 938 = ¥642
71
The National Examination for the Second-Grade Abacus License
Group A
(1) (2) (3) (4) (5)

1 289 307 483 971 3 029 176 62 501 478 974 362
675 243 81 602 459 15 849 729 360 4 265 103
90 428 165 5 871 304 436 250 3 968 251 58 130 279
17 436 -90 562 20 754 318 91 087 86 341
7 360 294 -264 918 8 109 567 -46 083 529 679 854
6 548 -6 738 095 3 291 -857 901 1 502 463
874 310 10 947 283 97 846 5 340 782 7 318
48 039 152 306 947 521 738 19 648 80 264 597
2 157 896 54 732 70 983 462 654 270 392 086
93 405 7 086 529 1 240 573 -2 593 18 759
526 781 -32 169 840 612 984 -9 145 806 2 840 915
3 185 069 -7 615 54 601 -36 714 76 051 432
43 972 9 520 736 9 072 345 20 794 136 419 687
54 702 631 64 187 865 913 1 283 097 23 590
951 820 815 023 65 378 024 476 835 3 705 241
210 352 029 77 482 141 181 175 937 39 732 401 229 362 027
(6) (7) (8) (9) (10)

537 641 312 650 8 029 137 86 374 80 625 749
7 241 930 1 287 094 79 630 821 10 342 569 79 853
35 082 469 64 238 948 675 493 720 5 986 072
697 508 40 175 982 52 149 6 205 917 395 846
-4 803 621 3 674 -2 407 536 8 253 10 793
-28 159 5 214 803 -163 904 274 068 9 753 261
-90 154 682 793 165 60 245 817 31 679 148 507
763 240 83 620 941 74 396 7 052 841 73 062 918
17 563 19 576 5 720 681 54 189 302 2 814 730
2 058 374 6 048 239 819 570 924 516 2 684
-386 019 852 497 -6 438 8 760 193 490 172
-9 785 41 056 -31 057 294 17 245 36 405
10 964 827 9 536 718 -98 763 536 980 41 205 396
8 435 719 27 418 305 4 502 698 3 875 421 6 897 013
70 296 907 523 381 052 90 641 835 541 628
-29 512 699 176 296 461 126 671 061 183 440 913 222 051 027
72
Group B Group C
(1) 60 937 x 2 154 = 131 258 298 (1) 8 485 074 / 2 754 = 3 081
(2) 42 618 x 7 309 = 311 494 962 (2) 44 080 524 / 5 436 = 8 109
(3) 804 752 x 962 = 774 171 424 (3) 0.2059086 / 0.7068 = 0.2913
(4) 18 509 x 0.5817 = 10 766.6853 (4) 52 622 371 / 61 403 = 857
(5) 93.125 x 6.048 = 563.22 (5) 0.6793975 / 9.371 = 0.0725
(6) 76 084 x 8 539 = 649 681 276 (6) 2 984 785 / 482 = 6 192.5
(7) 0.20436 x 470.6 = 96.1718 (7) 1 709.6483 / 3 592 = 0.4760
(8) 8 741 x 39 281 = 343 355 221 (8) 5.74902 / 0.0615 = 93.48
(9) 31 597 x 1.475 = 46 605 575 (9) 12 426 698 / 1 987 = 6 254
(10) 0.52963 x 0.0623 = 0.0330 (10) 610.42124 / 82.09 = 7.436
(11) ¥47 512 x 8 063 = ¥383 089 256 (11) ¥11 393 544 / 3 918 = ¥2 908
(12) ¥90 386 x 0.7421 = ¥67 075 (12) ¥641 / 0.0784 = ¥8 176
(13) ¥2 975 x 508.96 = ¥1514156 (13) ¥12 945 709 / 287 = ¥45 107
(14) ¥76 128 x 0.4375 = ¥33 306 (14) ¥2 187 / 0.4306 = ¥5 079
(15) ¥38 604 x 2 869 = ¥110 754 876 (15) ¥52 899 / 5.496 = ¥9 625
(16) ¥65 093 x 0.0942 = ¥6 132 (16) ¥6 593 836 / 1 637 = ¥4 028
(17) ¥17 249 x 3 517 = ¥60 664 733 (17) ¥1 131 / 0.8125 = ¥1 392
(18) ¥504 831 x 658 = ¥332 178 798 (18) ¥59 415 129 / 92 403 = ¥643
(19) ¥49 625 x 1.024 = ¥50.816 (19) ¥360 804 / 70.25 = ¥5 136
(20) ¥81 307 x 7 391 = ¥600 940 037 (20) ¥51 693 213 / 6 591 = ¥7 843
Group D
Mental Calculation
(80 % accuracy, 2 minutes)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
58 130 61 987 83 704 32 501 23 436
716 97 389 19 206 16 284 48 741 28
65 681 24 870 45 860 91 17 308 510
901 458 85 21 731 29 105 930 49 67
45 76 150 36 52 487 53 376 85 940
234 802 93 603 417 72 314 29 132 79
82 37 704 546 180 918 790 862 64 12
320 26 417 89 94 63 85 75 250 386
79 940 52 405 39 590 67 409 96 807
143 59 236 72 625 35 426 68 71 95
2 643 3 306 2 211 3 648 2 472 3 774 2 247 3 315 2 319 3 360
73
The National Examination for the First-Grade Abacus License
Group A
(1) (2) (3) (4) (5)

52 089 461 432 186 075 9 135 842 4 170 326 958 247 013 589
7 153 820 8 145 903 267 304 827 196 9 852 036 5 928 376
3 045 861 297 9 247 183 96 705 12 493 870 36 740 218
179 625 408 -62 098 541 2 036 754 981 -736 981 425 27 190
43 162 -762 309 1 408 273 -572 908 1 029 846 753
81 904 753 -5 280 931 764 83 927 654 1 630 857 3 507 461
9 286 041 3 679 452 702 149 836 6 058 249 731 574 268 109
7 306 154 289 24 805 691 5 368 071 93 175 648 61 930 548
8 479 503 901 326 785 6 520 743 189 567 903 412 459 862
297 580 364 4 270 516 49 386 527 -2 647 103 8 702 164 539
165 937 71 058 934 610 945 -3 901 728 564 6 382 047
64 327 810 -358 712 640 9 017 254 368 -28 157 496 418 079 635
4 710 698 325 -6 450 839 168 032 579 845 069 273 92 543 718
5 732 649 1 097 534 628 2 571 460 24 019 1 605 923
923 517 086 83 917 50 492 813 4 615 380 7 350 291 864
16 692 619 905 4 981 140 355 18 952 760 439 7 093 253 688 18 530 789 832
(6) (7) (8) (9) (10)

2 709 658 341 7 421 630 921 376 048 6 194 873 314 726 905
5 182 069 65 049 6 347 208 195 804 725 361 2 645 718
890 246 753 4 086 739 215 -70 153 264 16 249 59 832 067
-731 482 615 908 743 -694 503 30 954 187 8 031 276 549
-46 029 378 27 851 409 162 980 375 1 054 678 932 694 783
-9 017 486 523 9 372 586 3 015 467 928 7 240 156 3 501 624
3 592 710 7 251 094 863 7 935 681 201 536 498 968 135 470
109 264 538 36 281 597 84 029 135 871 250 25 840 391
72 503 694 548 617 320 3 748 256 76 109 382 6 401 279 835
65 871 3 946 075 -456 102 987 5 192 783 604 7 453 102
-1 874 305 8 905 173 246 -8 561 740 8 459 736 46 087 921
-380 516 924 639 182 -2 609 845 317 420 368 915 172 943 058
5 024 387 196 4 280 951 27 439 69 547 023 5 318 960
61 740 935 92 314 708 7 890 251 9 043 182 567 69 287
8 961 247 170 523 864 93 014 862 5 073 829 7 890 421 356
-561 035 258 21 760 230 438 7 498 320 359 16 921 742 562 23 930 227 026
74
Group B Group C
(1) 432 159 x 68 194 = 29 470 650 846 (1) 9 373 655 025 / 48 237 = 194 325
(2) 751 083 x 95 036 = 71 379 923 988 (2) 59 530 739 332 / 70 652 = 842 591
(3) 628 407 x 23 871 = 15 000 703 497 (3) 0.1271344371 / 2.14863 = 0.05917
(4) 0.97815 x 80.6253 = 78.86364 (4) 52 274 789 792 / 56 104 = 931 748
(5) 360 798 x 0.17409 = 62 811.32382 (5) 40.87907525 / 0.13807 = 296.075
(6) 584 162 x 75 382 = 44 035 299 884 (6) 330 085.21078 / 69 018 = 4.78260
(7) 23.6054 x 40.617 = 958.78053 (7) 7 534 346 391 / 35 769 = 210 639
(8) 8 079 231 x 3 925 = 31 710 981 675 (8) 0.4164695782 / 0.07391 = 5.63482
(9) 127 496 x 5.4178 = 690 747.8288 (9) 80 233 503 322 / 9 254 = 8 670 143
(10) 0.603945 x 0.02649 = 0.01600 (10) 624.88295605 / 84.925 = 7.35806
(11) ¥576 183 x 42 817 = ¥24 670 427 511 (11) ¥3 931 366 928 / 15 896 = ¥247 318
(12) ¥862 054 x 0.06759 = ¥58 266 (12) ¥635 615 / 0.98013 = ¥648 501
(13) ¥3 915 478 x 5 603 = ¥21 938 423 234 (13) ¥44 731 890 / 63.125 = ¥708 624
(14) ¥740 625 x 0.84512 = ¥625 917 (14) ¥7 643 415 038 / 37 142 = ¥205 789
(15) ¥184 397 x 739.24 = ¥136 313 638 (15) ¥254 071 068 / 296.71 = ¥856 294
(16) ¥902 871 x 0.28196 = ¥254 574 (16) ¥46 113 / 0.06459 = ¥713 934
(17) ¥451 063 x 92 031 = ¥41 511 778 953 (17) ¥52 348 910 242 / 58 067 = ¥901 526
(18) ¥273 509 x 54 768 = ¥14 979 540 912 (18) ¥37 108 174 071 / 4 057 = ¥9 146 703
(19) ¥68 192 x 360.875 = ¥24 608 788 (19) ¥174 669 / 0.83424 = ¥209 375
(20) ¥390 246 x 13 904 = ¥5 425 980 384 (20) ¥39 176 554 843 / 729 803 = ¥53 681
Group D
Mental Calculation
(80 % accuracy, 2 minutes)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
321 5 680 759 6 014 495 8 720 147 3 098 254 9 413
2 038 173 9 370 397 8 150 479 6 250 917 8 190 126
716 3 527 164 4 680 -243 6 093 863 1 340 -436 2 780
1 542 8 601 -527 251 -3 961 158 -235 654 -5 021 837
653 912 2 843 7 406 518 7 826 4 012 2 801 617 586
9 704 485 682 5 937 2 074 613 9 581 579 4 203 7 105
285 7 034 4 105 873 1 632 2 760 729 465 3 568 954
5 410 298 -3 524 695 -925 594 -1 936 7 602 -937 4 270
869 6 940 -913 1 208 784 3 408 -348 8 937 784 6 098
4 937 796 8 061 982 6 307 951 5 704 286 1 592 369
26 475 34 446 21 020 28 443 14 831 31 602 24 767 26 679 12 814 32 538
75

Japanese Abacus

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Japanese Abacus

Caricato da

Copyright:

Formati disponibili

The japanese abacus, its use and theory, by Takashi Kojima

CHARLES E. TUTTLE COMPANY

First edition, April 1954

valuable suggestions and a large part of the information in Chapter I. He is the

Foreword by Yoemon Yamazaki…………………..2

I - ABACUS VERSUS ELECTRIC CALCULATOR

MATSUZAKI (Abacus) vs. WOOD (Electric Calculator).

Addition: 50 number Matsuzaki 1m 14.8s 1m 16s -

Multiplication: 1m 44.6s 1m 19s 2m 14.4s

Division: 1m 36.6s 1m 23.4s 1m 21s

Total Score Matsuzaki - - - 4

As examples of the proficiency required of the abacus operator, it will be of interest to

A. ADDITION AND SUBTRACTION

To receive a first-grade license an applicant must be able to work problems similar to

A. ADDITION AND SUBTRACTION

To receive a third-grade license an applicant must be able to work problems similar to

II – BRIEF HISTORY OF THE ABACUS

III – BASIC PRINCIPLES OF CALCULATION

General Rules for Moving Beads

3. A black bead is one which obtained numerical value in a previous step.

5. ↑ indicates that beads are lo be moved up with the index finger.

6. indicates that beads are to be moved up with the thumb.

IV. ADDITION AND SUBTRACTION

1. Adding and Subtracting One-Digit Numbers

1+1 2+1 3+1 5+1 6+1 7+1 8+1

Step 1: Set 3 with the thumb (Fig. 7).

2—1 3—1 4—1 6—1 7—1 8—1 9—1

Step 1: Set 2 (Fig. 9).

Step 2: Move down 5 with the index finger (Fig. 10).

1+5 2+5 3+5 4+5

Step 1: Set 7 (Fig. 11).

Step 2: Move up 5 with the index finger (Fig. 12).

5—5 6—5 8—5 9—5

Step 1: Set 2 (Fig. 13).

1+6 2+6 3+6

Step 1: Set 8 (Fig. 15).

9—9 9—8 9—7 9—6

Note 2: The above-mentioned procedure is preferable to moving up 5 first and 1 next

Step 1: Set 4 (Fig. 17).

4+1 3+2 2+3 1+4

Step 1: Set 5 (Fig. 20).

5—1 5—2 5—3 5—4

Step 1: Set 4 (Fig. 23).

4+1 4+2 4+3 4+4

Example 10. 7-3=4

Step 1: Set 7 (Fig. 25).

5—1 5—2 5—3 5—4

Step 1: Set 3 on a unit rod, which we shall call B (Fig. 27).

1+9 2+8 3+7 4+6

10—9 10—8 10—7 10—6

Step 1: Set 6 on the unit rod B.

Step 2: Move down 1 on B with the index finger, then move up 5 on B

6+4 7+3 8+2 9+1

Step 1: Set 1 on the tens’ rod A.

10—4 10—3 10—2 10—1

Step 1: Set 9 on the unit rod B.

Step 2: Move down 1 on B with the index finger, then move up 5 on B

9+1 9+2 9+3 9+4

Step 1: Set 13 on AB.

Step 2: After moving down the 1 on A, move down 5 and move up 1 on B

11—2 11—3 11—4

Step 1: Set 6 on the unit rod B.

Step 2: Move up 1 on B (Motion 1), move up 5 on B (Motion 2), and