iHtfllrsbw

(Citlbilf.

y? 7108 5-10

INTRODUCTION TO THE STUDY OF
INDIAN MUSIC

INTRODUCTION TO THE

STUDY OF INDIAN MUSIC

AN ATTEMPT TO RECONCILE MODERN HINDUSTANI
MUSIC WITH ANCIENT MUSICAL THEORY AND
TO PROPOUND AN ACCURATE AND COM-
PREHENSIVE METHOD OF TREATMENT
OF THE SUBJECT OF INDIAN
MUSICAL INTONATION

BY

E. CLEMENTS

OF THE INDIAN CIVIL SERVICE, BOMRAY PRESIDENCY

LONGMANS, GREEN, AND CO.

39 PATERNOSTER ROW, LONDON
NEW YORK, BOMBAY, AND CALCUTTA

1913

All rights reserved

FOREWORD

The time Is perhaps far distant when it will be possible to
write a connected history of Indian music, tracing its origins,
development, and old age. It is clear, however, that its golden
age — that period so short in the history of any art cycle, and
so prepotent in determining the modes of both art and life
for long subsequent periods — must lie far back from the
present. Not improbably, that golden age coincided with the
moment of greatest achievement in drama, Kalidasa, and for
the theory, Bharata. Long anterior to this, however, music
was a most highly cultivated — perhaps the most highly
cultivated — of Indian arts, and to the present day it has
remained the most continuously vital and most universally
appreciated art of India. Taking together what has been lost,
and what remains, music is, then, the most complete expression
of the soul or genius of the Indians — a mirror faithfully
reflecting their inner life. That English Orientalists and
educationists have so long ignored this music, is the measure
of their misunderstanding of India.

While it is true that, until modern times, music has
remained in the best sense one of the most popular of Indian
arts, it is also true, though with exceptions, that it has been
neglected and despised, for example, by Aurangzeb, as well
as by more modern puritans. But the music remained too
intimately associated with religion, with the drama, and with
life, whether courtly or popular, and was too faithfully
guarded by the traditions of the guilds for it to be possible
that it should die out altogether. There are to be found even
now, for the most part at the courts of Indian rajas, or in
specially musical towns like Lucknow, Tanjore, and Poona, a
few ustads who are artists of high, and even of supreme rank ;
but they belong to an order that is passing away.

vi FOREWORD

The neglect of centuries, as in so many analogous cases,
has proved less disastrous than the renewed patronage of a
few decades. The constant use of the tempered harmonium ;
the endeavour to adapt Indian modes to the purposes of tenth
or fifteenth-rate brass bands maintained by Indian rajas ; the
absence of any aesthetic element in modern Indian education ;
the mania for English accomplishments : all these causes have
actively contributed to the degeneration of Indian music. By
degeneration, I mean literally confusion, a running together,
and destruction of bounding-lines ; a process quite distinct
from any natural waning of vitality at the latter end of an
art cycle.

Now that life has changed, so that the old music, however
splendid, no longer expresses race-intention (we are no longer
united by such an intention), there are two considerations that
must weigh with us, when we think of Indian music ; to
maintain the memory of our past experience, as an interpre-
tation and inspiration and delight, and to clear the way for
new creators. For both these ends it is necessary to escape
from the confusion into which the theoretical part of Indian
music has unfortunately fallen. It is here, I think, that Mr.
Deval has done great service in applying a purely experi-
mental method to the analysis of the actual intonation of
thoroughly trustworthy hereditary musicians. Mr. Deval's
work, the results of which are published in his " Hindu
Musical Scale and the Twenty-two Srutees," deserves the
highest praise. It is true that Mr. Deval did not succeed
in his endeavour to improve his case by importing aid and
corroboration from scientific acoustics and Sanskrit philology ;
but I think that certain of his critics fall into more serious
error when they judge the results of his patient and invaluable
experimental work by weakness or inaccuracies in his method
of presentation.

The preparation of the sruti harmonium, and the presenta-
tion of the general results of Mr. Deval's work, combined with
a critical discussion of the theory of music according to
Bharata and Sarangdev in Mr. Clements' most interesting
book, mark, I think, an epoch in the scientific study of Indian
music. It will at any rate be possible for future writers, even
when they disagree with Mr. Clements, to say more clearly

FOREWORD vii

and definitely than heretofore, what they exactly mean ; and
still more important, for future recorders to make a nearer
approach to a true transcription of the Indian ragas. I cannot
but hope that Mr. Clements will himself extend his studies
in this direction. It may be a long time before we have as
full and as exact a knowledge of Indian music as we have of
Indian literature ; but if that time ever comes, it will, I am
sure, be acknowledged that the work of Mr. Deval and Mr.
Clements did much to clear the way for such a development
of knowledge.

I should like to say a word of warning with regard to the
sruti harmonium. This instrument is to be welcomed, in any
case, as infinitely preferable, from the standpoint of intonation,
to the tempered harmonium now in common use. It is a
valuable tool, and may be used for purposes of research, and
also for class teaching, where the instruction of large classes
(a process foreign to the Indian conception of educational
method) is unavoidable. Thus used, the sruti harmonium
will serve the ends of exact knowledge, and will not (as the
tempered harmonium now does) destroy the sensitiveness of
the Indian ear to those " hair's-breadth " distinctions which are
essential to a highly evolved art of pure melody. But, as I
think, no harmonium of any kind should ever be regarded as
a substitute for the tambura, because the quality of tone of the
tambura is so infinitely superior to that of the harmonium, to
say nothing of other aesthetic and social considerations ; above
all, the harmonium should never be used as an accompaniment
to the voice, leading or imitating note by note. This last,
even with the vina, would be foolish ; with a blatant instrument
like the harmonium, incapable, moreover, of any gliding from
note to note, it becomes repulsive.

Much the same argument applies to the use of a system
of notation ; for the purposes of exact knowledge — most
desirable as a means of escape from the present chaos — it is
very important that a suitable method of transcription should
be discovered. But the publication of Indian music in staff
notation, without warning that the scale is other than that
usually implied by that notation, tends to the destruction of
the character of that music in the same way as the use of a
tempered harmonium. It is for the purposes of science, of

viii FOREWORD

teaching, of the preservation of existing songs, and the
making of these accessible to Western students, that a notation
is now so necessary — above all, for the preservation of what is
so rapidly disappearing, and must soon be lost. But if it be
possible to maintain still, amidst the general popularization of
music in the modern and democratic sense, a tradition of
master-musicians in pupillary succession, as heretofore, then
for these it is far better that the method of oral transmission
should be maintained. No matter if the masters in different
parts of India do not all agree ; the very divergences of their
ragas may be an expression of local character. But it is not
for the sake of variety that I would preserve the system of
oral transmission ; but rather because this is the true method
of learning for an artist, because every singer so taught must
be in some degree a composer (he is taught, not merely to
repeat a given song, but to sing in a given mode and mood),
and because it is so great an advantage for the true musician
to need no external aid to memory, such as a printed score.
Indeed, I suppose that even if we succeed in recording the
greater part of Indian music as it still survives, the music
itself cannot persist as a part of everyday life unless it is thus
handed on as a sacred tradition.

In any case it is much that the existing music should be
recorded and analyzed for the student of whatever time or
country. The necessity of such a record in India need not be
dwelt upon ; but perhaps the most valuable result of the
growing interest in Indian music would be realized if the time
ever comes when, in the words of Captain Day, " the study of
the national music of the country will occupy, as it should, a
foremost place in all Indian schools," and certainly, also, in
the Universities. But I should also like to emphasize the
importance of this study for western musicians ; not only as
a means of better understanding the heart of India, but also
because it must be in the long run disadvantageous to ignore
one half of the world's experience in any art. If Indian
music is very different from European — and the fundamental
difference is less than at first sight appears — then all the more
reason for the Western musician to enlarge his outlook.
Perhaps even, in the words of M. Bourgault-Ducoudray,
Oriental music may "provide Western musicians with fresh

FOREWORD ix

resources of expression, and with colours hitherto unknown to
the palate of the musician." At least we may feel certain
that both for us, and for the Western student, the exact study
of the science of Indian music is a necessary process in the
interests of progress and interpretation.

It is then with gratitude that I have accepted Mr.
Clements' invitation to write a preface to his learned and
stimulating work ; in so doing I wish to specially commend
both the whole subject of Indian music, and this book, to the
notice of all Indians and Englishmen who have any voice in
determining modes of education in India.

ANANDA K. COOMARASWAMY.

London,
September i3, 1912.

CONTENTS

CHAPTER PAGE

Foreword v

Introduction xiii

I. The Srutis of Modern Hindustani Music ... i

II. The Staff Notation and the Srutis 15

III. The Staff Notation {continued) 37

IV. The Interpretation of the Ancient Text-Books 47

V. Concluding Observations 76

Appendix A, The Indian Srutis and the Tempered

Chromatic Scale 89

Appendix B. The Indian Harmonium 90

Appendix C. Transposition on the Indian Harmonium . 94

Appendix D. - 97

Appendix E 100

Index and Glossary 103

INTRODUCTION

The art of music in India has for centuries been neglected
and despised by the general public. That period is now
happily over, and an awakening of interest is everywhere
manifest ; educational institutions of recent birth are in a
flourishing condition, and there is a demand on all sides from
amateurs for musical tuition, while treatises, and new notations
and editions of ancient texts are continually being published.
It cannot be denied, however, that all this energy is in urgent
need of the guidance which a sound musical theory would
afford. Modern text-books may appear learned to the
uninitiated ; the historian will, however, frankly admit that,
since the days of the Sangit Ratnakar, Indian musical
systems have fallen into such confusion that no one has been
able to reconcile the teaching of that authoritative treatise
with later works on the subject, or with the practice or theory
of modern musicians. The art is also in grave danger of
being spoiled, as other Indian arts have in the past been
spoiled, by cheap imitation. Contact with the West has
resulted in a blend of Indian music with European intona-
tion, a combination in the highest degree inartistic and likely
to prove more harmful than the neglect of centuries.

Those who invent notations no better than others already
in existence, with an elaborate superfluity of new and
wonderful signs, in the hope of handing their names down
to posterity as inventors are friends of doubtful sincerity.
Those who use the staff-notation for the purpose, without
attempting to distinguish the special features of Indian
intonation, are encouraging the heresy that intonation is of
minor importance. Those teachers who promote the sale of
tempered harmoniums, and make use of them in the class-
room, are proving their own incapacity to guide the musical

xiv INTRODUCTION

renaissance of their country. The head of one institution
finds the tempered harmonium an excellent means of teach-
ing beginners " the scale." What scale, one may ask, for it
does not give a reasonable approximation to any Indian
scale } He admits that the " peti," as it is commonly called,
cannot render all the Ragas ; has he not, in his publications,
drawn attention to this defect ? When the beats given by the
fifths, fourths, and thirds of his instrument are show.n to him,
he says that he is aware of them, and considers them some-
what like the Indian embellishment known as "kampit."
When asked whether he follows the teaching of Sarangdev,
the author of the Sangit Ratnakar, he replies : " He is not
really an old authority ; we go back to the Sama Veda ; we
are of opinion that Sarangdev is wrong in many respects, and
we reckon our srutis downwards instead of upwards."^ To
go back to the Sama Veda is a happy inspiration, as that
work, so far as it touches the question of scales, deals in pure
generalities.

Europeans, on their part, are too ready to assume that
the Indian scales are artificial and capricious, and too prone
to ascribe to " quarter- tones " distinctions between intervals
with which they are not familiar, such as the difference
between the major-tone and the minor-tone. Intervals less
than a semitone are frequently employed in grace or embellish-
ment, but very seldom in scales. When they form part of a
scale, it is possible in many cases to regard them as con-
stituents of natural chords of the seventh, the tempered
equivalents of which are well known to the Western musician
as discords. It is also a prevalent idea that, in the study of
Indian music, intonation may be neglected as being of minor
importance. This view is, however, demonstrably wrong ;
the student who masters the subject of intonation will find
no difficulty in solving the remaining problems of Indian
musical theory.

The following pages deal with the subject of intonation,
principally as applied to the school of music known as the
Hindustani School. The author ventures to hope that when
once the rationale of the " kaishiki " or hair's-breadth distinc-
tions of the Sangit Ratnakar is grasped (and may the reader

' See p. 48.

INTRODUCTION xv

beware of following a certain well-known author, and calling

them ** aggravating ") writers and teachers will no longer be

found ignoring the difference between the ri of Bhup and

the ri of Deshkar, the ga and ni of Kafi, and the ga and ni

of Bahiravi Ac .'f is impossible to treat the subject of in-

without an appropriate musical notation,

will be devoted to the presentation of a

idian music may be accurately written.

r. Deval of Sangli, which have rendered

c treatment of the subject in hand, will

account given of the Indian harmonium

he accuracy of his conclusions has been

iges from the Natya Shastra of Bharata

^. nakar will be quoted and commented

^ ^ this new science of intonation. It will

>» 5ting conclusions suggest themselves as

^ gs of the Hindu scales. As a result of

I '^ )ns, tentative suggestions will be made

g assification of the Indian scales. For

are interested in European instruments

:hey may be adapted to perform Indian

on will be explained.

5 to acknowledge his indebtedness to

e of Malabar Hill, Bombay, the pub-

^it, and author of Hindustani Sangitachi

^orks, but for whose generous help in

and collating Hindu texts this volume

written.

E. C.

<

>^

H

o

<

.^ ^

1^/

'^

t-H

\f\

^

fO

W

<A

^ ^

1-4 '^

^ !/5

^.1

rf C>

" r^

U 0)

bo bo

rt <A

Ph Ph

xiv INTRODUCTION

renaissance of their country. The head of one institution
finds the tempered harmonium an excellent means of teach-
ing beginners " the scale." What scale, one may ask, for it
does not give a reasonable approximation to any Indian
scale } He admits that the " peti," as it is
cannot render all the Ragas ; has he not, in
drawn attention to this defect ? When the 1:
fifths, fourths, and thirds of his instrument i
he says that he is aware of them, and consi
what like the Indian embellishment kno)
When asked whether he follows the teachi
the author of the Sangit Ratnakar, he rep
really an old authority ; we go back to the
are of opinion that Sarangdev is wrong in m
we reckon our srutis downwards instead of
go back to the Sama Veda is a happy in
work, so far as it touches the question of scc
generalities.

Europeans, on their part, are too read]
the Indian scales are artificial and capriciou
to ascribe to " quarter-tones " distinctions 1
with which they are not familiar, such i
between the major-tone and the minor-ton
than a semitone are frequently employed in g
ment, but very seldom in scales. When tht
scale, it is possible in many cases to rega
stituents of natural chords of the sevent
equivalents of which are well known to the ^
as discords. It is also a prevalent idea tha
Indian music, intonation may be neglected <.
importance. This view is, however, demc
the student who masters the subject of int
no difficulty in solving the remaining pro
musical theory.

The following pages deal with the subj(
principally as applied to the school of mu;
Hindustani School. The author ventures tc
once the rationale of the " kaishiki " or hair's-breadth distinc-
tions of the Sangit Ratnakar is grasped (and may the reader

' See p. 48.

INTRODUCTION xv

beware of following a certain well-known author, and calling
them *' aggravating ") writers and teachers will no longer be
found ignoring the difference between the ri of Bhup and
the ri of Deshkar, the ga and ni of Kafi, and the ga and ni
of Bahiravi. As it is impossible to treat the subject of in-
tonation adequately without an appropriate musical notation,
part of this volume will be devoted to the presentation of a
method by which Indian music may be accurately written.
The researches of Mr. Deval of Sangli, which have rendered
possible the scientific treatment of the subject in hand, will
be described, and an account given of the Indian harmonium
by means of which the accuracy of his conclusions has been
demonstrated. Passages from the Natya Shastra of Bharata
and the Sangit Ratnakar will be quoted and commented
upon in the light of this new science of intonation. It will
be found that interesting conclusions suggest themselves as
to the early beginnings of the Hindu scales. As a result of
the author's conclusions, tentative suggestions will be made
as to the scientific classification of the Indian scales. For
the sake of those who are interested in European instruments
the manner in which they may be adapted to perform Indian
airs in correct intonation will be explained.

The author wishes to acknowledge his indebtedness to
Mr. V. N. Bhatkhande of Malabar Hill, Bombay, the pub-
lisher of Lakshya Sangit, and author of Hindustani Sangitachi
Paddhati, and other works, but for whose generous help in
the way of translating and collating Hindu texts this volume
would never have been written.

E. C.

September <), 19 12.

INTRODUCTION TO THE STUDY
OF INDIAN MUSIC

CHAPTER I

THE SRUTIS OF MODERN HINDUSTANI MUSIC

Mr. a. M. Chinnaswami Mudliar writes in his " Oriental
Music in European Notation : " " Considering -., „. , .
the prodigious number of nationalities and the and Karnatic
diversity of provincial dialects in existence schools of music,
throughout the length and breadth of the Indian Empire, it
should be no matter of astonishment if there be found any
number of heterogeneous systems, as well as incongruous
classifications in standard works forming the musical litera-
ture of the land. The primary distinction is into two classes,
Marga (celestial) and Deshi (terrestrial) ; the latter is now
broadly divided into Hindustani and Karnata, the former
representing the school established by Hanuma, and the
latter the much more ancient and authentic system introduced
by Narada, the inventor of all Arts and Sciences. It is clear,
however, that local tastes and methods of training have con-
siderably upset the theories originally propounded. ... Of
late the Hindustani element (which has itself much deterio-
rated owing to foreign admixture) has been ingrafted on the
Dravidian modes to an alarming extent."

Mr. Mudliar's opinions as to the respective merits of
" Dravidian " and Hindustani music, and on the subject of
deterioration, are interesting, but in view of the fact that
he accepts the tempered scale and the European tempered
notation as a medium of instruction, they need not command
acceptance. His volume is a praiseworthy effort and most
valuable to the student, but, owing to a deficiency in the

B

2 INTRODUCTION TO STUDY OF INDIAN MUSIC

critical faculty as regards musical intonation, Mr. Mudliar un-
fortunately encourages the worst kind of deterioration which
has ever affected Indian music. The present work deals with
Hindustani music only ; the author hopes to be able to show
that a great part of it is directly traceable to the systems set
forth in Bharata's Natya-Shastra of about the fifth century
A.D., and the Sangit Ratnakar of the thirteenth century.
These are the most closely reasoned and critically worded
of the early text-books. It is reputed that Sarangdev, the
author of the Sangit Ratnakar, was an inhabitant of Kashmir.
From internal evidence one would conclude that the music
he describes is that of Hindustan. However, the pandits of
Southern India endeavour to appropriate him to themselves.
The present writer hopes to show that it is only by doing
violence to his theory that it can be applied to Karnatic
music. Roughly speaking, Hindustani music may be said
to prevail in the north and west of India and the Deccan,
while Karnatic music is confined to the south and east.
Many scales are common to both, but the general spirit of
the two systems is apparent from the scales which are first
taught to beginners ; in the west, the scale is the same as
the just major scale of Europe,^ in the south it is a chro-
matic scale (known in Hindustani music as the scale of the
Raga Bhairava) with semitones between the first and second,
third and fourth, fifth and sixth, seventh and eighth degrees.

There are grounds for believing that the remote precursors
Indian music— °^ these two scales were pentatonic, one the
the Gramas, scale which has been found amongst almost

Jatis.and gas. ^|^ nations and which may be roughly indicated
thus — C, D, E, G, A, C, and the other the old Greek scale of
Olympus. Complete scales of seven notes were in existence
many centuries before Christ, and the notes bore the "tonic-
solfa " names sa ri, ga, ma, pa, dha, ni, abbreviations of shadj,
rishabh, gandhara, madhyam, pancham, dhaivat, nishad. Ancient
theory puts the development of scales and melodic forms in
the following order : first came the Gramas, which may be
regarded as collections of notes definitely related to one
another by musical intervals. Writers on the subject persist

' Sometimes the scale with the high sixth is used ; the practice in this respect
does not appear to be uniform.

THE SRUTIS OF MODERN HINDUSTANI MUSIC 3

in thinking of the Gramas as scales because each was named
after a particular note. This was not the mode of thought
followed by Bharata and Sarangdev. A grama might be
regarded as a string of notes ranging through three or four
octaves. To sing a scale out of the string, a starting-point
must be chosen. A scale of seven notes from the string was
called a MiircJiJiana of the note so chosen. This was the
second step. A scale which was to form the basis of a melody-
required something beyond a "lowest note," something to
establish its harmonic individuality. The Jdtis were elabo-
rated as the third step ; their character was largely determined
by the note chosen as the drone or pedal accompaniment ;
they also had a fixed final note, and Vadis and Samvadis ;
the latter being pairs of notes a fourth or a fifth apart which
determined the tonality or harmonic structure of the scale.
After the Jatis came the Grama-Ragas, which may be regarded
as generic melody-types, and their descendants the Rdgas of
modern India. The best singers in India, those whose art
has not been contaminated by the tempered harmonium,
prefer to sing to an accompaniment of the tambura alone, or
the tambura with drums. The tambura is a stringed instru-
ment of rich tone upon which a powerful drone is produced
consisting in almost all cases of the fundamental note, which
Europeans would be incHned at first to call the " tonic," and
its fifth. With such an accompaniment it would be next to
impossible for any singer with a sensitive ear to sing the first
five notes of the ordinary major-scale in anything but just
tuning, unless he departed therefrom of set purpose, for the
fifth upper partial tone of the bass is to be heard distinctly,
and the third upper partial of the fifth above. Wherefore the
use of the tambura is to be recommended for the singer of
Folk-songs also, unless a harmonized accompaniment in
natural tones is to be obtained. To proceed, the Indian
singer will always be able to state in what Raga or Ragas his
song is composed. The name of the Rdga connotes a scale
bearing a fixed relationship to the drone, with its harmonic
structure determined by a Vadi and Samvadi, a chief note
("ansha svara") occurring more frequently than others, a
lower limit described in terms of the Murchhana, occasionally
an upper limit also, certain characteristic turns of melody,

4 INTRODUCTION TO STUDY OF INDIAN MUSIC

recurring with frequency, certain rules regarding the employ-
ment of embellishments, and a stated time of the day for its
performance. It is a common practice, after singing an air in
a Raga, to improvise a series of free fantasia passages, each
returning in due course to a characteristic snatch of the
melody, only to wander off again in still more elaborate
variations. The whole performance must be " within the
Raga," that is, without transgression of the elaborate rules
governing its structure.

Until recently, with trifling exceptions, Ragas, with their
Scope of the rules and scales and compositions, were handed
present work. down from teacher to pupil without the aid of
any written record. It is not surprising, therefore, that a
good deal should have been lost, and that different singers
should be at sixes and sevens regarding the names and dis-
tinguishing features of the Ragas. Various investigators
have made compilations with the intention of making known
the scales used in various Ragas. These compilations are use-
less to the present writer, because they one and all ignore the
"hair's breadth" distinctions upon which correct intonation
depends. The present work does not essay to give a list of
Ragas or scales, but merely to point out a method by which
they may be classified. Where the reader may have any
criticism to offer regarding the scales given as examples, and
the names attached to them, he should remember that they
are taken, with one or two exceptions, from the repertoire of
one artist, Abdul Karim. The author is aware that in some
cases other singers employ other scales or sing the same
scales under other names. The reader should also under-
stand that, as an introduction to the study of Indian music,
this book does not go beyond the province of intonation.
Once a census of Indian Ragas is made, and once their
scales are classified into groups according to the principles of
correct intonation, their further subdivision according to
harmonic structure, as determined by the " Nyas Svara "
(final note), and Vadi and Samvadi, may be undertaken.
With the material at present available it is impossible to
make any suggestions as to the manner of proceeding with
such further subdivision.

From ancient times up to a comparatively modern date,

THE SRUTIS OF MODERN HINDUSTANI MUSIC 5

the notes or " svaras " which, strung together, constituted the
Gramas were called " shuddhl' which means xhe shuddh
" pure." In the modes or Jatis of each Grama andvikrit notes,
certain chromatic variations in the notes were used for the
sake of melodic effect. The notes so altered were called
" vikrit." The name "vikrit" was extended to any new
note obtained by shifting the frets of the solo instruments in
use, such as the vina and sitar. Various tuning devices
for a change of mode involving a shifting of the frets without
retuning the chanterelle or drone strings came to be employed.
A great deal of discussion has centred round these " shuddh "
and " vikrit " notes, but for the most part it has been infruc-
tuous because the looseness and inaccuracy of the term
" vikrit " has not been sufficiently grasped.

The ancients believed that if the octave were divided into
twenty-two roughly equal parts all the notes
in use could be obtained. They called these
small intervals " sriitis," and spoke of intervals of two srutis,
three srutis, and four srutis, and of raising or lowering a note
by one sruti or more. They believed that all the " shuddh "
and " vikrit " notes had srutis to themselves. The author
concludes, from a study of Bharata's Natya Shastra and the
Sangit Ratnakar of Sarangdev, that their system in reality
involved the use of 25 notes to the octave, and not 22 as
they imagined (see Appendix E).

Their theories were founded upon the system of tuning

described as confined to Hindustan in Captain -,. ,. z. •

^ The old tuning

Day s " Music of Southern India " (p. 109). method now
In that system the chanterelle strings were obsolete,
dhaivat, rishabh, and gandhara. This is a conclusion which
one cannot fail to draw from a careful study of their text-
books. Now, the modern system of tuning throughout India
has shadj as the principal drone, accompanied by pancham
or madhyam. Not only this, but shadj and pancham are
regarded as fixed notes which may never become " vikrit," or,
in other words, sharpened or flattened, and shadj has acquired
the privilege of being regarded as the basis of all scales. All
Jatis, therefore, start from Shadj, and all the scales of all the
Ragas.

It is clear, therefore, first that the modern srutis and the

6 INTRODUCTION TO STUDY OF INDIAN MUSIC

ancient srutis must differ in many cases, and, secondly, that

there are no longer strings of shuddh notes

minology no from which to construct Jatis and scales of

longer appro- Ragas, This latter conclusion, which there

priate. . ° . . , , , . .

IS no gainsaying, has escaped the notice of

all writers on the subject. In practice it has been recog-
nized by some of the Hindustani musicians, who accord-
ingly discard the word "shuddh" except in the case of
sa and pa, and call the other notes atikomal, komal,
madhya, or madhya-tivra, tivra, tartivra, according to their
position in the scale. The author follows their practice in
this respect.

Until Mr. K. B. Deval, a retired deputy collector, residing

at Sangli in the Southern Maratha Country,
researches and commenced his researches, the subject of Indian
the Indian intonation had baffled all inquirers. Many

books had been written by Indian gentlemen
and others with the laudable object of explaining what Indian
music was, but the European in India, who was interested in
the subject, found that there was a point beyond which he
could not go. Mr. Deval constructed a diachord consisting of
two wires of equal length stretched over a sounding board,
one wire being provided with a graduated scale and a
movable fret of the same height as itself. His method was
to tune both wires to the same pitch, that of the shadj of the
singer assisting him. He moved the fret of the wire which
had the graduated scale into the position which gave the
note which the singer had been asked to sing. A simple
calculation from the reading of the scale gave him the com-
parative vibration-number of the given note in relation to
shadj. He persevered for years at this investigation, deriving
assistance from many of the best singers that India could
produce. As regards most of the notes in use, his conclusions,
when referred back to ancient theory, may be summed up in
the statement that two srutis make a just semitone, three
srutis a minor-tone, and four srutis a major-tone. In respect
of these notes the accuracy of his conclusions can fairly be
said to be beyond controversy. The remaining notes belong
to certain irregular scales ; a knowledge of the melodic struc-
ture of the Ragas in which they are employed was called in

THE SRUTIS OF MODERN HINDUSTANI MUSIC 7

to assist the verification upon the diachord. Mr. Deval pub-
lished his conclusions in " The Hindu Musical Scale and the
Twenty-two Srutees," printed at the Arya Bhushan Press,
Poona. The next step was to order a harmonium tuned in
the twenty-two intervals which he termed the "twenty-two
srutees" of the "Hindu Musical Scale." This instrument
was designed by Mr. H. Keatley Moore, B.A., Mus. Bac,
who assisted the late Mr. A. J. Ellis in the latter's transla-
tion of Hemholtz' " Sensations of Tone." After some slight
modification in tuning and the arrangement of the keys,
the result of verification and experience, the instrument has
been patented (15548/11); the manufacturers are Messrs.
Moore & Moore of New Oxford St., London (the makers
of Ellis' Harmonical), and the agents in India are Messrs.
S. Rose & Co., Bombay. The author has, through Mr. Deval's
courtesy, and with the help of Abdul Karim and other singers,
been able to verify all the various scales mentioned in the
following pages upon this instrument.

The following table describes the twenty-four notes in
most frequent use, showing which of them are ^ cr'of on of
adopted in the Indian harmonium, and their terminology
relationship with the ancient srutis. ^^""^ adopted.

Comparative

The more common of the

European

equivalents,

Shadj being

F.

vibration

Numbers and names

Modern Srutis : in what

number, sa

of the Ancient

Ragas used. Moore's har-

Sign.

being taken for

Srutis.

monium notes numbered in

convenience of

brackets.

calculation to

be 240.

0. Kshobhiui

(0) Nishadkomal* (Kafi,
Khamaj)

LowE>

ni'fe

213^-

I. Tivra

(i) Nishad kaishik *
(Bahiravi)

Et^

nit?
nil^

216

2. Kumudvali

(2) Nishad tivra' (Kal-

E

225

3. Manda

yan)
(3) Nishad tarlivra
(Marva)

Ft^

nirj*

227I

4. Chhandovati

(4) Shadj »

F

s;i

240

5, Dayavati

(5) Rishabh atikomal
(" Septimal " Asavari)

Septimal

i-i [^

252

Note. — The names of Ragas are merely given as a help towards identification.
1 The modern Srutis, marked with a figure S are identical with the ancient
Srutis of the same serial number. This will be seen in a later chapter.

8 INTRODUCTION TO STUDY OF INDIAN MUSIC

Comparative

The more common of the

European

vibration

Numbers and names

Modern Srutis : in what

number, sa

of the Ancient

Ragas used. Moore's har-

ecjuivalents,

Shadj being

F.

Sign.

being taken for

Srutis.

monium notes numbered in

convenience of

brackets.

calculalion to

be 240.

6. Ranjani

(6) Rishabh komal '
(Bahiravi)

Gt?

ri|7

256

7. Raktika

(7) Rishabh niadhya •
(Kafi, Deshkar)

Low G

iH

266§

8. Raudri

(8) RishalDh tivra ^ (Kal-
yan, Asavari)

G

"tl

270

g. Krodha

(9) Gandhara komal '
(Kafi, Todi)

Low A !?■

ga'i^

284A

10. Vajrika

(10) Gandhara sadharan'
(Bahiravi)

At^

gal7

288

II. Prasarini

(11) Gandhara tivra'

A

300

(Kalyan) ^

12. Priti

(12) Gandhara tartivra
(Jayja>vanti)

High A

Bl? .

ga +

303f

Madhyam atikomal "

ma ]^

315

(Yaman-Kalyan)

Septimal

13. Marjani

(13) Madhyam komal'
(Kafi, Bahiravi, Bihag)

Low B !7

ma'i^

320

14. Kshiti

(14) Madhyam kaishik
(Dhani)

Bt7

mat?'
mailf

324

15. Rakta

(15) Madhyam tivra

B

337i

(Kalyan)

,

16. Sandipani

(16) Madhyam tartivra
(Marva)

C!7

ma^

341J

17. Alapini

(17) Pancham' (Kalyan,
Bahiravi)

c

pa

360

18. Madanli

(iS) Dhaivat atikomal
("septimal" Asavari,
Todi)

Septimal
Dt7

dha ]^

378

19. Rohini

(19) Dhaivat komal'
(Bahiravi)

D^

dhat?

384

20. Ramya

(20) Dhaivat madhya'
(Kafi, Bilaval, Marva)

Low D

dha +

400

21. Ugra

(21) Dhaivat tivra'
(Kalyan, Bihag)

D

dha if

405

Nishad atikomal

Septimal

E^

ni ]^

420

22. Kshobhini

(22) Nishad komal '
(Kafi, Khamaj)

Low E '7

ni't^

426 If

It is a lamentable fact that there is no uniform system of
naming the Indian notes. The difficulty which confronted the
author was to put distinctive names to srutis, which although

' See note i, previous page.

- If C on Moore's harmonium is taken as sfi, the stud marked Xf ('■<?• in key
pancham) gives atikomal ma. It was found impossible to inchide this note in
key F.

THE SRUTIS OF MODERN HINDUSTANI MUSIC 9

invariably distinguished in practice, are generally confounded
under the same name. Such is the case with numbers 7, 8,
and 20, 2 1 of the above. The name madhya or madhya-tivra
is used by Abdul Karim to distinguish the lower of the two
notes, and the practice being generally acceptable is here
adopted. With regard to the groups o, i and 9, 10 and 13, 14
more difficulty was experienced. The last (No. 14) has never
hitherto been differentiated in name from the note below.
The following table will show at a glance what justification
the author has for adopting the names above given.

Ancient
names.

Names ob-
taining in
Madras.

Names obtaining in Hindustan.

Names here

First Method.

Second Method.

adopted.

0

I

ni!?

Shuddh \
Kaishik /

Kaishik

Komal

(Atikomal
(Komal

Komal
Kaishik

9
10

gal?

Shuddh 1
Sadharan/

Sadharan

Komal

("Atikomal
(Komal

Komal
Sadharan

The confusion of nomenclature displayed in Southern
India and for the most part in Hindustan is a heritage from
the change of tuning discussed in later chapters. Atikomal
for ni 1s( and ga "fe will be felt to be a misnomer by any one
acquainted with the other atikomal notes. It is an absurdity
to call the principal notes of Raga Kafi or Sri-Raga by such
a name, notes which originally formed the basis of the Shuddh
Jatis. They are no more atikomal than ma "fe which is uni-
versally called either komal or shuddh. The author dis-
tinguishes ma ^ and mab on the analogy of the corre-
sponding srutis of nishad. The distinction between ga t? and
ga b is of precisely the same nature, and the name kaishik for
gab would be appropriate, were it not for the fact that the
note has for centuries been called sadharan ga.

It can hardly be disputed that the twenty-two notes
chosen as the basis of tuning of the Indian harmonium are
the commonest of the Indian notes. It must be remembered
that in calling them the twenty-two srutis of modern Indian
music, one is using the word sruti in a transferred sense, for
strictly speaking a sruti is an interval and not a note.

lo INTRODUCTION TO STUDY OF INDIAN MUSIC

Secondly, the orthodox ancient system regarded shadj (and
other notes also) as of fixed absolute pitch, the pitch of each
shuddh note being referable to the cry of some animal or the
note of a certain bird. Although some singers of the present
day follow the ancient theory in this respect and have their
own inflexible notion as to the proper pitch of shadj, the
usual custom is to take any pitch that is convenient to the
singer or performer. Thus the term sruti has been made to
depart from its original meaning in two directions.

It will be useful to tabulate the notes employed in the
Indian harmonium thus —

Name of note

as abbreviated

in singing.

Middle row of

studs, marked

" ak " and

"k."

Black keys.

Front row of

studs marked

"m."

White keys.

Back row of

studs marked

" tt."

sa

—

—

—

Shuddh —

Ri

Atikomal U

Komal !?■ Madhya «|.

Tivra tf —

Ga

Komal R

Sadharan I?

—

Tivra b 1 Tartivra Jf

Ma

Komal "fe^

Kaishik [?■

—

Tivra fa

Tartivra Jf

Pa

—

— - —

Shuddh —

Dha

Atikomal U

Komal !?■ Madhya ^j.

Tivra t;

—

Ni

Komal \

Kaishik !?■

—

Tivra j?

Tartivra J^

The rest of this chapter will be taken up with a brief
Brief explana- account of these twenty-two notes and their
tion of the nota- distribution among the commoner Indian
tion employed. ^^^^^^ -pj^j^ account is intended for the Euro-
pean reader who presumably has some acquaintance with the
theory and notation of Western music ; others are referred
to the more detailed explanation to be found in the next
chapter.

In the Indian harmonium, shadj has been given the pitch
of F (Philharmonic, 1896) ; apart from that circumstance, the
notation to be employed has absolutely no connection with
any key or pitch. Shadj, which may be defined as the bass
note of the drone in modern music, is represented as F for
convenience, and the two F's of the treble clef encompass

THE SRUTIS OF MODERN HINDUSTANI MUSIC ii

the middle register of all voices from soprano to bass. The
word " register " is here used in place of the Indian " saptak " ;
the Indian musician divides the compass of all human voices
into three "saptaks," or octaves, the middle one naturally
being the one most frequently in evidence. The signs used
above represent respectively — (i) [r a septimal flat, the true
harmonic seventh of the note on the black key next above ;
(2) "^ a low flat, derived in a descending series of perfect
fifths from sa ; thus ma "^ is a fifth below sa, ni "^ a fifth below
ma i^, and ga i^ a fifth below ni -^ ; (3) b a flat, ri b being a
major-third below ma % and the rest related to ri b by an
ascending series of perfect fifths ; (4) + a low natural, ri +
and dha+ being in tune with ma^ ; (5) jj a natural in just
tuning, the intervals being referred to the scale of pa (or C)
major ; (6) + a high natural in the case of ga +, which is a
fifth above dha tf, and respectively pab and sab in the other
two cases.

The first scale to consider is that of Raga Yaman —

Common Chords.

W

IC^I

1^=2^:

m^^^^^

Sa ri ga ma pa dha ni
Here, sa and pa are shuddh, and the rest tivra. As
usual, the sign ]l is omitted. In order to facilitate explana-
tion, the position on the Indian harmonium keyboard will be
considered ; that is to say, the above scale will be taken to
be actually from F to F. It will be noticed that the tuning
is just tuning in the key of C major. D, F and D, A are not
in tune ; C to D, F to G, A to B are major-tones ; D to E
and G to A minor-tones ; and E to F and B to C just semi-
tones. In the following scales (in all Indian scales, in fact)
the positions of sa, ri, ga, etc., on the lines and spaces are
exactly the same as above.

(Bahiravi.) (BihaE^-.)

Common Chords. (Kafi.)

:q:

!^3^EE^^^^g^

.^=|£i£i^S5^^g=t:

12 INTRODUCTION TO STUDY OF INDIAN MUSIC

The first (Bahiravi) is a mode of the scale just described.
E to E in the Yaman scale has the same intervals in the
same order. Here ri and dha and ma are komal, ga sadharan,
ni kaishik. The Bihag scale is that of Yaman with the flat
fourth. It differs from F major in just tuning in this respect,
that the sixth being high (or Pythagorean) the number of
common chords available is reduced to three. The Kafi
scale is the mode of its second ; in other words, corresponds
to the scale G to G in Bihag. The Kafi scale, therefore, has
only three common chords. The notes used in Kafi are ri
and dha madhya, ga, ma, ni komal.

The above are what might be termed " diatonic " scales ;
the first two belong to the Madhyam Grama, like the European
major-scale in just tuning, while the last two belong to the
Shadj Grama,^ It will be noticed that as many as sixteen
different srutis are comprised in them.

The following passage is also of a "diatonic " nature; it
introduces the high flat fourth (ma kaishik) : —

(Dhani.)

The scale which is given on the right is hexatonic.
Adding dha b (D b) makes it a mode of the Yaman scale,
corresponding to the mode of the sixth of A b major in just
tuning. This scale will have five common chords.

This finishes the notes comprised in the "diatonic" or
Grama-scales. The reader who is only acquainted with equal
temperament may find it hard to realize that the D in tune
with F, and the D in tune with G, to take an example, are
notes separated by an interval (§J), which any untrained ear
can appreciate and which can even be committed to memory
and reproduced vocally. The accidentals above, which
include a downward stroke (+, "jj^), in all cases mean a flatten-
ing by this interval, the interval which separates the minor-
tone from the major-tone.

' The use of the high sixth instead of the harmonic sixth in what may be loosely
termed the parent scale makes the difference between the two gramas. The
gramas are fully explained in the next chapter.

THE SRUTIS OF MODERN HINDUSTANI MUSIC 13

The difference between just tuning and temperament is
well illustrated, not only in the Indian Ragas, European Folk-
but in the treatment accorded to Folk-songs 30"S music,
by modern musicians. If it is not practicable to accompany
Folk-songs with just harmonies, one would think that it might
at least be advisable to write an accompaniment which could
be played upon an instrument in just tuning if such were
to be had. This, however, is not the course usually adopted.
The compiler of collections of Folk-songs, instead of ascer-
taining the chords in just tuning which are available to him,
trusts to a vague musical instinct, with results of which the
following is an example —

[iiiiis^B

^j^=^

7nJ'

In the Folk-song to which this is an introduction (No. 5
at p. 13, of " Trente Melodies Populaires de Grece et d'Orient,"
by M. Bourgault Ducoudray), the " final " is A and the
" dominant " D, that is to say, low D (D +), the note a fourth
above A. It is an error, in the chord marked {a), to make D
harmonize with G, unless just tuning is entirely disregarded.
To put the matter in another way, the D used in the G
chord is not in the mode. In the same way, in the next
number, at the thirty-fifth bar, a foreign chord is introduced.
Similar examples might be multiplied ; they are to be found
in many such collections.

The reader is referred to the next chapter for a full

explanation of the tartivra notes in connection », ,

•<1 4.U T f T>- T^/r- ^ J T • Explanation of

with the scales of Ragas Marva, and Jayjay- notation con-

vanti. Whether the view there suggested re- ^•""^'i'
garding the development of septimal harmonies from the
chromatic scales is correct, admits of considerable discussion.
Septimal intervals may be an addition for which the Hindu
people should be thankful to the Mahommedans. Harmoni-
cally they add considerable charm to the Indian Ragas, and

14 INTRODUCTION TO STUDY OF INDIAN MUSIC

this short sketch of the modern Indian srutis would be incom-
plete without some account of them. The following passage
from the Raga Yaman-Kalyan introduces an atikomal note
which it was found impossible to incorporate in the Indian
harmonium —

The eighth note is the harmonic seventh of C, forming the
interval I with the middle C, and 1,^^ with A. Those un-
acquainted with the natural intervals would be surprised at
the large interval separating this note from the ordinary B
flat {B i;^). As the Indian name atikomal implies, the note is
very flat indeed. The atikomal ri and dha of the following
scale, taken from a Raga which Abdul Karim call Asavari,
belong to the same class —

fafg^^-Egjg

-ifsz=m=i

An atikomal ga and an atikomal ni of similar description
are employed in some rare Ragas. A low ga, the very
slightest degree lower than ga tivra, making the interval |-
with ri b is also occasionally employed. The well-trained
Indian singer is thoroughly conversant with septimal intervals
and intones them with accuracy and without hesitation.

The number of irregular scales employed by different
singers is considerable, and modern Indian musicians, with
pardonable exaggeration, describe the srutis they sing as
" anant " or endless. In the following chapter will be found
a note regarding the extent to which it is possible to discuss
these rare scales and intervals in the present work.

CHAPTER II

THE STAFF NOTATION AND THE SRUTIS

The purpose of this chapter is to give an account, fuller than
the sketch given above, of the manner in which the various
notes or srutis in use in Indian Ragas may be represented in
an adaptation of the staff notation of Europe.

For the sake of those who are not thoroughly conversant
with the scientific side of the subject, it will be advisable to
explain as briefly as possible what is meant by the harmonic
series and by natural and tempered intervals.

The sensation of sound is due to vibrations of the air. A
musical note is distinguished by the regularity The measure-
of the vibrations of which it is composed, l^tervals^y''*'^^
Strictly speaking, a musical sound defined by ratios,
a certain number of vibrations per second should be denomi-
nated a *' tone." In order to avoid ambiguity, however, the
author uses the word " note " for the purpose. The original
meaning of " note " was the sign in notation by which a tone
was expressed ; in common parlance it has also the meaning
which is here assigned to it. The interval, or " distance " as
Bharata has it, between two notes is completely defined by
the ratio which their vibration-numbers bear to each other.
It follows that in order to obtain the sum of two intervals,
one must multiply their ratios ; and to obtain the interval
which constitutes the difference between them, one must
divide the bigger ratio by the less.

The notes of stretched strings and of many other musical
instruments are clothed in upper partial tones The Harmonic
or harmonics ; it is this circumstance which Series,
gives them their " timbre " or quality. It is a general law
that if the tension of a stretched string is constant, and
notes are sounded upon different lengths of it (this may be

i6 INTRODUCTION TO STUDY OF INDIAN MUSIC

done by means of the device adopted by Mr. Deval in his
diachord, namely, adjusting a movable fret of the same
height as the string at the required distance, pressing the
string lightly upon the fret, and plucking the string upon the
side adjusted to the given length), the vibration-numbers of
those notes will be inversely proportional to the length of
string required to produce them. Thus, if the whole string
gave a note of lOO vibrations a second, § of the string would
give a note of 150 vibrations a second. The upper partial
tones which enrich a note are thus explained : — The primary
note which the ear recognizes is that given by the vibrating
length of string as a whole. A plucked string, however,
vibrates not only in its whole length, but in nodal segments
corresponding to the fractions of the harmonic series, i, I, |,
1, I, etc. Hence, together with the primary note of, say, «
vibrations a second, are faintly sounded a long series of notes
of vibration-numbers represented by 2n, ^n, 4n, 5«, 6n, etc.
In the notes of the tambura, these upper partial tones are so
strong that the untrained ear may easily recognize them as
far as the fifth member of the series. The series of notes
formed by the generator and its upper partial tones is
commonly called the harmonic series. To the student of
the Indian scales a knowledge of it is essential, as all the
melodic intervals which the musicians of ancient Hindustan
thought beautiful and worth preserving are traceable ulti-
mately to its influence.

The staff-notation is a graphic method of writing music.
The Great Staff ^^ possesses a distinct advantage over any
and the three method which requires the eye to follow one
fndiS^vocal set of signs for melody (svara), and an entirely
music. distinct set for time (laya). The pitch of a

note in staff-notation is shown by its position {verticall)i) on
the staff, and the character of its accidental if any, its time
value by the sign used, and its place in the measure (avarta)
by its position {horizontally) in the bar. This chapter deals
only with the pitch of notes as shown by their vertical
position ; an uniform sign known as the minim (^) will be
used in every case.

The Indian system of vocal music allots three "Saptaks"
or octaves to the voice, each Saptak ranging from sa up to ni.

THE STAFF NOTATION AND THE SRUTIS 17

The lowest octave is called "mandra," the middle octave
" madhya," and the highest " tar." The Great Staff consists
of eleven lines, the middle one of which is thicker than
the others. This middle line represents the middle C of
the European keyboard. Other notes, above and below the
middle C, are represented by spaces and lines alternately.
Turning for a moment to the Indian harmonium, the shadj
of which is primarily intended to be the F of European music
(at the philharmonic pitch of 1896), the 22 srutis in the three
octaves may be thus represented on the Great Staff: —

Male Voice.
4 5 6 7 8 9 10 II 12 13 14 15 16 17

/I : \ 1 1 1 1 1— I I . ^- ^

tar.

madhys
mandra.

^:fefe^:^^:^&p=tp$p5pE^4pi^

1
ga ma pa

18 19 20 21 22(0) I 2 3 Female Voice (tar octave).
, , , . I I I I y--T^(<^ Sva

L ' ' r-E=|=lz:

sa ri ga ma pa dha ni sa
dha ni

These are the twenty-two more common srutis of modern
music, numbered according to the ancient system, in the
three " saptaks " of the male voice. Played on the Indian
harmonium at the pitch indicated by the Staff, they are correct
in point of pitch according to the notions of the singer Abdul
Karim. The soprano voice would be an octave higher accord-
ing to Indian ideas. The madhya and tar saptaks would be
the mandra and madhya saptaks respectively of the soprano.
The soprano tar octave would be written as indicated, either
using short or " leger " lines above the Staff, or using the
dotted lines and sign 8va, which mean "to be sung or
performed an octave higher."

C

iS INTRODUCTION TO STUDY OF INDIAN MUSIC

The next example shows the relative pitch of the first

„ .. . sixteen of the harmonic series of the lowest sa

The Harmonic , ^ ^ r^ , ^ ^ r ^ ^

Series in on the Great Staff. (As before brackets are

Notation. ^gg^ ^^ ^Y\o\v the three " saptaks.")

It

T^-

1=^

I

^-

4=-

12 3456 78 9 10 II 13 13 14 15 16

The 7th, nth, 13th and 14th members of the series are
not to be found among the keys of the Indian harmonium,
the 13th is quite unknown to Indian music, so also is the i ith,
unless, as stated in a later chapter, it formed part of the
Gandhara Grama, while the 7th and its octave the 14th are
the atikomal nishad mentioned above (p. 8).

It will be convenient to state at this point that it is
The use of the proposed to borrow from European music the
sign tl- practice of using the sign n only in cases

where the note of the denomination in question has, when
last used, borne a different accidental. For instance, the
fifth, ninth, tenth of the series are tivra, but the accidental is
dispensed with ; the fifteenth bears an accidental, as the note
immediately preceding it is also Nishad and bears the acci-
dental [r. Thus, sa and pa, being unchangeable, according to
the modern view, will never require an accidental, while the tivra
notes will only use their accidental tj; in the circumstances ex-
plained.^ In European music, the notes which require no acci-
dental ordinarily and make use of the sign u when necessary to
distinguish them from preceding notes, are the notes of the scale
of C major. The scale of C major in just tuning when played
from F to F is identical with the scale of the Raga Yaman,
which consists of sa and pa shuddh, ri, ga, ma, dha and ni tivra.

As the vibration-numbers of the notes of the harmonic
. . . series are proportionate to their serial numbers,

by the Harmonic the series gives at once the ratios of the
^^"^^- intervals employed in music. The intervals

given by the first ten numbers of the series are in order, the

> See p. 43.

THE STAFF NOTATION AND THE SRUTIS 19

octave "f, the fifth :;, the fourth f,^, the major third i^, the minor
third f;, the septimal third J.,^ the septimal tone 7, the major
tone ;^, the minor tone -Ij'. The 16th and 15th members give
the semitone [J-;. All Indian scales may be referred to some or
other of these intervals. It is not suggested that scales were
consciously built up upon such an elaborate foundation as the
first sixteen terms of the harmonic series. The intervals
given by the first five terms were sufficient for the construction
of the early Hindu scales. If the two fifths, sa to pa, pa to
ri, be added together and the interval sa to sa or an octave be
subtracted, the major tone sa to ri results. Thus § X ^ —- j = {!.
Subtracting the major tone sa to ri from the major third sa
to ga gives the minor tone ri to ga, J^f. The semitone || is
the difference between the fourth i and the major third f.
Another semitone, |^, which may be termed the Septimal
semitone, is obtained by subtracting the septimal tone ^ from
the minor third i'. Septimal intervals appear to have been
introduced at a comparatively modern period ; they pre-
suppose some kind of acquaintance with the seventh term of
the harmonic series.

In Appendix B will be found examples showing to what
extent the notes of the harmonic series may be The Notation
obtained from the Indian harmonium. The simplified,
reader who has access to such an instrument will be in-
terested to discover how harmoniously these notes blend
together. The examples are written in Pianoforte Score,
which consists of the Great Staff with the middle line
omitted. This method will be found to give more facility for
reading than the Great Staff; the omission of the central line
gives greater prominence to the position of the notes. The
Yaman Scale is here shown in the three octaves in Pianoforte
Score. The upper staff is known as the Treble and the
lower as the Bass ; the signs used to distinguish them are
called clefs.

^ The author adopts these names as they are simple and appropriate, in
preference to the cumbrous names at present used by European scientists.

20 INTRODUCTION TO STUDY OF INDIAN MUSIC

:^=^:

:^:

-^-

1^==^:

m.

.^_

:^==^

:?2=^

=^=:^:

i:

_j ^

For writing single melodies without accompaniment it will
be unnecessary to use both the treble and bass clefs as here
shown. The treble clef will be sufficient for the purpose, with
the aid of the following simple convention, which is ordinarily
followed in European vocal music. The melody will be
written for the female voice ; the male singer will sing an
octave lower. The following will represent the mandra,
madhya and tar octaves in the Yaman-scale for all voices.

tiir.

mandra.

madhya.

2^:^

If found more convenient, the lower and upper notes,
instead of being written upon leger lines, may be written in
the madhya octave, with the sign 8va... below or above
respectively.

So far, this discussion has reference to Indian music

The Notation in ^^^^^^^ ^o"" ^^^ Indian harmonium in the
reference to principal key of that instrument. That key

absolute pitch. (p philharmonic 1896 = Shadj) was chosen
for two reasons — first, because one of the most proficient of
Indian singers regards that as the correct absolute pitch of
shadj ; and secondly, because the key is the very one which
anyone in search of the simplest possible notation for Indian
music would unhesitatingly fix upon. For purely instrumental
music, F will do as well as any other pitch ; it will also suit a
large number of voices. Difficulty arises only when a singer
who finds F an inconvenient pitch wishes to accompany his
voice in unison on the Indian harmonium or to learn from that
instrument the correct intervals of any particular Riga. To
meet these cases. Appendix C has been compiled. From the

THE STAFF NOTATION AND THE SRUTIS 21

typical scales there given it will be easy to pick out the
correct notes of any Raga at approximately the pitch desired.
The question of notation with regard to keyed instruments
may be now put on one side, as, excepting the Indian har-
monium, no keyed instrument capable of playing Indian music
has yet been constructed.

For the vina, sitar and other Indian instruments, and for
the voice, it is not necessary that the notation shadj should
be of any specific pitch. The notation last described is by
far the best for two reasons. First, the Staff includes the
madhya saptak completely and conveniently. Secondly, as
modern Indian music allows no change in sa or pa, F sharp
or C sharp will never be employed ; a slight acquaintance
with the Indian system teaches one also that no other sharps
are in use ; consequently, the notation will dispense with the
sign jj, which is used for a sharp note.^

In the notation followed throughout this book the treble
clef will be used, the lowest space being appro- xhe Notation
priated to sa of the madhya saptak, and the finally chosen,
top line to the sa of the tar saptak. Except in connection
with the Indian harmonium, no element of absolute pitch is to
be associated with it.

A suitable notation having been found, the next step will
be to examine certain typical Indian scales Scales and
which between them furnish all the 22 notes Srutis defined,
set out above on the Great Staff. To avoid misapprehen-
sion, it will be advisable to premiss that a scale is a col-
lection of notes ranging from a given note to the note an
octave above and resembling a ladder in this respect, that
the array (" th^t ") of notes must afford a practical means of
ascent and descent. To describe the 22 srutis above given,
or the collection of Shuddh and Vikrit notes given in the
Sangit Ratnakar, as a "chromatic scale" is a misnomer.
The European chromatic scale of twelve semitones stands
upon a different footing, as chromatic passages taken from it
are employed in music. Then, again, it is an error to suppose
that the 12 srutis were ever pieced together by fitting small

' It is a common practice to put sa on the line appropriated to C. This
renders the use of the accidental J necessary for ma tivra. The scale of C is
also an awkward one for passages in the mandra octave.

22 INTRODUCTION TO STUDY OF INDIAN MUSIC

harmonic intervals between the fundamental note and its
octave. The ancient theorists of India had no knowledge of
harmonic ratios ; to them the interval termed a sruti appeared
to be the smallest interval which the human voice was capable
of singing ; consequently, they assumed roughly that their 22
srutis divided the octave into equal parts. Recent research
has, however, made it abundantly clear that the direct descen-
dants of the ancient intervals of four srutis, three srutis and
two srutis, are respectively, the major-tone Q), the minor- tone
(^) and the semitone (]-t;). One is therefore justified in
assuming that these were also the ancient intervals. It will
be found in later chapters that this postulate is of assistance
in unraveling the meaning of the old text-books. Now,
the sruti between four srutis and three srutis is §^,, while that
which lies between the latter interval and the semitone is |^.
So long as it is understood that the srutis were never equal,
no great object is gained by measuring them individually. It
was, however, a convenient terminology, and still is so, to
distinguish the major-tone, minor-tone and semitone by the
number of srutis they contain. The twenty-two notes adopted
in the Indian harmonium have this merit, that they conform
in this respect to ancient usage ; it will be found that every
major-tone obtainable from them consists of four srutis,^ every
minor-tone of three srutis, and every semitone of two srutis.

It is impossible to grasp the facts of Indian intonation
Th G " without a knowledge of the two principal

and Jatis of the Gramas, the Shadj Grama and the Madhyam
ancient system. Qj-ama, The Gandhara Grama is not in use,
and need not be discussed here ; but it is a mistake to
suppose that the other two Gramas are also to be regarded
merely as objects of antiquarian interest. Writers who
treat of the Greek modes and the Church modes, and
composers who make collections of Folk-songs, generally fail
to take adequate notice of this fundamental branch of in-
tonation. The same tendency is unfortunately at work in
India. The ancient Indian theory was based upon two
collections of notes, or " svaras," known as Gramas. They
differed in this respect, that the pancham was in one flatter

* There are two exceptions : the major-tone between srutis 3 and 6, and
srutis 16 and 19; but these intervals are rarely used, if at all.

THE STAFF NOTATION AND THE SRUTIS 23

by one sriiti. Until one of the notes was chosen as a
starting-point, they were Gramas and nothing more ; the
lowest note being chosen to sing from, a Murchhana named
after that note was produced. For a scale or mode which was
to become the basis of a musical composition, more was
required, namely, the harmonic relationship given by the
Vadi and Samvadi, and the stability given by a fixed initial
note (Graha) and final note (Nyas). The initial note appears
also to have been the drone or pedal note. The Vadi and
Samvadi corresponded very nearly to the final and dominant
of the Church modes ; they were invariably a fifth or a fourth
apart (never a third as in some of the Plagal modes), and, as
alternative Vadis and Samvadis were allotted to each Jati, it
is probable that a kind of modulation was allowed, this being
more permissible in a system which required a constant fixed
drone accompaniment. The simplest modes formed upon
the plan thus briefly described were called " Shuddh Jatis,"
There was one for every degree of the scale ; the " Shuddh
Jatis" were distinguished from others in having the initial and
final and chief note, or prevailing note of a composition, the same
as the " nama-svara," or that note which gave its name to the
Jati ; of the seven Shuddh Jatis, four belonged to the Shadj
Grama and three to the Madhyam Grama. As a knowledge
of the Shuddh Jatis is of great assistance to the student of the
Indian scales, they are here given, first in their primitive form,
and then in their modern position, based upon shadj ; it will
be observed that their names are derived from those of the
initial notes. The sruti-intervals between the different degrees
are shown in each case in figures.

24 INTRODUCTION TO STUDY OF INDIAN MUSIC

The SiiADj GrAma.
I. Dhaivati.

f^^^

2. Naishadi.

i^feg^^^^^^lSi

4-^4

W*

2 432443

(Bihag.)

fefg^s^^gli^Ji^i^

4 3

3. Shadji.

4 4 3

4 3 24432
(Kafi or SrI-Raga.)

^=2^5

sg^^-^jg^^i^gp^-^R

244

2 4

i

s

3244324

4. Arshabhi.^

S^

±^^s

:t:

t=^

i^^zfes^

443243 2443

The Madhvam Grama.
I. Gandhari. ,1^ (Yaman.)

4 3

ii^^^-plg^

±:

-I-

:^-^

^=^:

-:^-=-^-^-

--^~^.

:t=:

^ 4342432 4342432

2. Madhyama. ^^ %^ (Rageshvari.)

\J 1 A -7 ,1

4^Wi

*-' ■ 3 4
3. Panchama.

4324

4 3

ppiii^ii^^^^

—I

_^U^pz:^-

:^z:^z?^

:t:

t=:

:^

'^ 4 243243 4243243

The following quotation from Bharata (Natya Shastra,
Explanation of Ch. 28, Commentary on verse 25) may serve
accidentals and as a text upon which to hang an explanation
the two of the bewildering number of accidentals

Gramas. used : " Pa is lowered one sruti in the

Madhyam Grama. The interval passed over in raising or
lowering pa by one sruti is the measure of a sruti." Although

* If the text of the Sangit Ratnakar is followed literally, the scale here given
would be in the higher or tar oclave, the usual (inadhya) octave being an octave
lower.

THE STAFF NOTATION AND THE SRUTIS 25

the interval of two srutis is not double that of one sruti, nor
the interval of four srutis double that of two, although in fact
Bharata and his successors laboured under a misapprehension
in considering that the sruti was a unit by which intervals
could be exactly measured, yet the Indian musician had, and
still has, a definite idea of what he means when he says a note
is lowered or raised by a sruti. His idea of a sruti is ^J, the
difference between the major tone and minor tone. The
descending line used in the accidentals •[• and i;^ always implies
a descent by this interval of ^/,. To make this clear, the
following collection of major tones and minor tones is given : —
4 3 4 3

i

i^-z^-r^-W^

■^-^^^

=f^q

3

1 —

g^=^^g^^^g^g=gZ^j^=^^

:=]iz:=^=q

- _| — -1==— 1—- f^ — I- — ^

5^i

12^:

4 " 4 3

The figure 4 denotes a four-sruti interval, and 3 a three-
sruti interval. The tivra sign ti is used in the manner above
explained.

To enable the reader to distinguish with facility the
consonant intervals of 7 srutis (the major-third, 2) and 6
srutis (the minor-third, 2) from the discordant intervals of
8 srutis (ff;) and 5 srutis (.jf), the following collection of
intervals should prove useful : —
Major Thirds.

=g::fe^j?pjgi

Minor Thirds.

iz]-

g:^r^^E^E^=g

:ti=±:z:

Dissonances.

i:^---

^^^5grZ^=gEM=JgE

E^£^":5^:¥pEE^^e

:t:

tn::

26 INTRODUCTION TO STUDY OF INDIAN MUSIC

The Indian reader is probably familiar with the scales
of Ragas Bihag and Bilaval, although he may not perhaps
have realized that the dhaivat of the latter is one sruti lower ;
the notes of Bihag are sa and pa shuddh, ma komal, and the
rest tivra, while those of Bilaval have dhaivat madhya (+)
instead of tivra (n). These two scales are here given side
by side, first starting from shadj, and then in three trans-
positions from ma '^, dha b, and ni ^jj respectively. The
intervals are shown in srutis as before.
I. (a) Bihag. (^) Bilaval.

4 3
ib)

jsggpp^]

3 4

^i3^^g|p5^^^^

«-''i'^24'4 3 2 4 3 24'342

The scales marked {a) are all exactly the same ; so also
are those marked {b). The effect of lowering the sixth note
in the (b) class is to change the harmonic structure of the
scale. The sixth note, from being merely a relation (or in
harmony with) the second, is transformed into a relation of
the first and fourth notes of the scale. In Bilaval, for instance,
dhaivat harmonizes as a minor third below sa and a major
third above madhyam ; whereas in Bihag it is discordant with
both sa and ma, and a fifth above ri. Scale II. {b) introduces
the madhya rishabh, the note which is used with such telling
effect in Deshkar Raga. This note is in tune with ni ^ and
ma t?, whereby it is to be distinguished from ri tivra. Scale
III. {a) contains the high or "chadh" madhyam (ma b).
This note is used in Dhani Raga amongst others ; it is a rare
note in modern music, because the sa drone strongly attracts

THE STAFF NOTATION AND THE SRUTIS 27

the perfect fourth (ma isj). Scales IV. (a) and (d) are com-
posed of the shuddh notes of the ancient Shadj and Madhyam
Gramas. It will be obvious that the flattening of pa by one
sruti is precisely the same thing as the difference in harmonic
structure which distinguishes Bilaval from Bihag. Lest it
should be supposed that because Indian music does not employ
harmony in the Western manner, considerations drawn from
the study of harmonic intervals are of less importance than in
the West, the following scale passages which are divisible
into groups (as shown by brackets) of identical harmonic
structure, are taken from the scales under discussion. The
laws of harmony have been, in the East as in the West, the
chief determining factor in the evolution of musical scales.

Bihag Scale.

-,— i— -I "^^-^ ..^^^=^— ^-^^"^-

Bilaval Scale.

32 24 24 43 43

Returning, after this digression, to the Shuddh Jatis, the
first remark one is inclined to make is that ^j^^ ,^^^^ ^^^
it is singular that so few of the modern scales modern scales
are there. Three common scales, Bihag, Kafi, ^^o^^P^*^^ •
and Yaman, are comprised among the Jatis, and one rarer
one, Rageshvari. It maybe added as a possible explanation,
that from Ratnakar (Prakaran 4, verse 15) ^ it may perhaps
be inferred that the Madhyam Grama Jatis should be taken
from ma and not from sa, thus —

Ganclhari. .^^i.^feiS. Madhyama.

3fc^£E

±:

_^4^.

^^^^Eii

Panchama.

:^i?^:

^^i^=z^:

.f>fS^^

^^-^-

±:

Adding the sa drone would have the effect of compelling
the ear to refer the scale to shadj as a starting-point. This
gives for Gandhari and Panchama, the following familiar

1 See p. 58.

28 INTRODUCTION TO STUDY OF INDIAN MUSIC

scales ; but the Madhyama scale must be discarded, as modern
practice does not allow a flattened pa.

Gandhari (Bilaval). Panchama (Bahiravi).

:^^^g^

This method therefore gives five well-known scales in
place of three. Yaman and Kalyan may also have been
derived from the Shuddh Jati Naishadi by sharpening of
the fourth, that is to say, by the not uncommon change from
shuddh ga to antara ga, a change which will be fully
explained later on. From the form of Panchama first given,
the hexatonic scale of the uncommon Raga Dhani, may have
been derived. Dhaivati and Arshabhi must, however, be
rejected, the former because the fifth note is really komal
pancham and not to be tolerated nowadays in such a scale,
and the latter because the high fourth is not now used except
in the Raga Dhani and possibly in some rare derivatives of
the Raga Asavari. The secret of these dissimilarities and
anomalies will be explained more fully in later chapters. It
lies in a difference of tuning. In olden practice the Chante-
relle or drone strings were dha ^ and ri «f., and for convenience'
sake all Jatis were played from the pitch of one or the other
of these two notes. The modern practice is to make sa the
chief pedal note, and to add above it pa or ma ^ according to
the nature of the scale. Assuming that, in the majority of
ancient compositions, the scale was actually based upon ri •}.,
the drone would be the fundamental note of the scale together
with another note (dha •}.) a fourth below it ; in other words, it
would correspond to sa over pa, whereas the modern practice
is almost invariably to take the fifth as a drone, namely, pa
upon sd. If this radical change in the nature of the drone
accompaniment did actually take place, it is sufficient to fully ex-
plain the discrepancies between the ancient and modern scales.

The next important observation to be made is that in

modern music there are, strictly speaking, no

word "Grama" Gramas, although the difference between the

in relation to two Gramas is accurately exemplified in the

scales of Bihag and Biklval. The history of

the Indian scales may be divided into three periods, (i) the

THE STAFF NOTATION AND THE SRUTIS 29

Grama period, (2) the Transition period, (3) the Modern
period. In the first, all scales were played from the shuddh
notes of the Gramas ; in the second, in order to avoid
changing the drone and re-tuning the tambura and drums, all
scales were played from ri + as a basis, but as far as possible
the frets of the shuddh notes were employed, the wire being
tuned up or down to bring the required initial note (graha
svara) to the pitch of ri + ; in the third, that is at the present
time, all sense of the Gramas has died out, shadj is looked
upon as the starting-point of all scales, and the scales of the
Ragas are regarded, with certain exceptions, as independent
structures. Those exceptions, although of no great import-
ance, are highly interesting. Yaman and Bahiravi, as has
been seen, belong to the Madhyam Grama, while Bihag and
Kafi belong to the Shadj Grama. This inter-relationship is
exemplified in the common practice of playing these pairs of
Ragas respectively from the same set of frets. Thus I. a
below becomes I. b if the wire is tuned up by two srutis, and
W. a becomes II. ^^i if the wire is tuned down by four srutis.

I. (a)

ib)

^gEpE^zg^gigg^^^

i^t^ii

II. {a)

{b)

Indian music may be said, then, to have outgrown the
Gramas and the system of shuddh svaras. At the same time,
a knowledge of the theory of the Gramas is essential to the
student of Indian music, and the word Grama itself should
prove a useful substitute for the Western and less appropriate
term " diatonic," to designate scales which are derived from the
shuddh notes of ancient India. The scales which have been
discussed above may be referred to as the Grama scales.
They are distinguished from others by the fact that they
contain three major-tones, two minor-tones, and two semi-
tones.

In the music of Hindustan the Grama scales hold the
first place in point of number and popularity. Scales of less

30 INTRODUCTION TO STUDY OF INDIAN MUSIC

than seven notes, hexatonic or Shadava, i.e. of six notes,
Scales other ^"^ pentatonic or Odava, i.e. of five notes, may
than Grama be regarded as abridgments of the complete

scales to which they approximate. They will
most of them be classed among the Grama scales. Then
came a very important class consisting of scales which com-
prise three or four semitones in place of two. Such scales
are to be found also among the Folk-songs of Europe. They
may not inappropriately be designated Chromatic scales.
The following are examples of chromatic scales and transilient
(hexatonic or pentatonic) scales : —

(Bhairava.)

CHROMATIC.

(Deshi Todi.)

— — 4^— ^-^f^-f^-p^ i — '^— '

±^~.

±=zztz=£z£

(Purvi.)

^3^:^

:^=^i?^

5-r^z[^

(Multani.)

^i^^^^g

-A^^

(Bhup.)

TRANSILIENT.
(Deshkar.)

ggEJ^gg^Eg^^Eg^E^fg

(Hansadhvani.)

:?2i^l

(Malkans.) (Dhani.) (Marva.)

§

The names of the Ragas employing the scales given are
written in brackets. The Bhup scale may be regarded as an
abridgment of Bilaval, that of Deshkar as an abridgment
of Kafi with the sharpened or chromatic third, Hansadhvani
as derived from Bihag, and Malkans from Bahiravi. The
last two have an individuality of their own. Dhani is a
Grama-scale, while Marva is an irregular scale of a kind not
hitherto discussed in this chapter.

The affinity of the Marva scale to the Chromatic scales
Irregular above given is disguised by the Indian notation,

scales. which treats the fourth and sixth notes as

"tartivra," whereas, harmonically speaking, they are komal

THE STAFF NOTATION AND THE SRUTIS 31

notes. Tartivra ma, being a fourth above komal ri, should
really be called komal pa, and tartivra ni similarly komal sa.
Indian tradition is, however, bound up with a "solfa," or
" Sarigama " system which allots one of the names sa, ri, ga,
ma, etc., to the notes of all scales for the purposes of singing.
It has become part of that tradition in comparatively modern
times to restrict the names sa and pa to two fixed degrees
of the scale. The real character of Marva is made clear in
the following transposition (beginning the scale from ri instead
of sa) : —

Respecting the intervals (other than semitones of two
srutis) employed in the Chromatic scales, it will be found
that the sruti system as a method of comparing them breaks
down completely. The sruti system appears to have been
applied in its inception to the Grama scales only ; when used
in connection with other scales it leads to nothing but con-
fusion. The intervals of the Bhairava scale expressed in
ratios are —

16 ls> J n 0 J « 7_5 1 «
15» 64' loJ S» 15' 64> T5'

while those of Marva are —

J_6 7.") 2;-) 6 75 25 6 13 5
15' 64' 225» 64' 225» 12 8>

which approximate to

1 (I 7 S 7 S T_35
15' 6' 7' G' 7' 12 8'

It may be that in this indirect manner, Indian musicians
were led to incorporate septimal harmonies in their scales.
Certain it is that in modern Indian music septimal intervals
are employed with remarkable effect and in a considerable
number of Ragas. There is a form of Asavari which employs
atikomal ri in place of tivra ri and atikomal dha in place of
komal dha, a form of Todi which employs atikomal dha, and
many others which there is no need to mention "^ —

* Amongst them may be specially noticed the Raga Yaman-Kalyan, which
employs occasionally atikomal ma, followed by tivra ga (interval, -H/,). The scale
of the common form of Asavari is sa, ri |;|, ga |7, ma \, pa, dha [?, ni I?.

\2 INTRODUCTION TO STUDY OF INDIAN MUSIC

(Septimal form of Asavari.) (Todi.)

^^j^^^^-Jg^Ejz^S^Eg,^^:^:

Before leaving the Irregular scales, it will be of interest
to give the following short passage in the Riga which the
singer Abdul Karim calls Jayjayvanti —

T ^^"^"^^ 1 ' t-^i f— T 1 1 1 t-V 1 1 1 f— i

1 I t ^ — r |™'"i|" '■i|i '^

^^

This example anticipates some features of the notation which
are explained later on. To avoid misapprehension, the
accidentals have been repeated before each note. The
tartivra ni of Marva is really komal sa, as has been ex-
plained ; the tartivra ga of this Riga bears approximately
the same relationship to komal ma that the tartivra ni of
Mirva does to si, but at the same time it is used frequently
in conjunction with ri tivra and appears to be, harmonically
speaking, what is known as high A, i.e. sharper than ga tivra
by the interval gj. It is possible to regard the tartivra ni
of this Riga as " high E," instead of F b. The difference
between these two notes is so small as to be hardly per-
ceptible even to the trained ear of the Indian singer.

The reader may object that, as the above enumeration of

, , the Indian srutis is admittedly not exhaustive,
IrrcsTul&r SC3,16S
further dis- it is inadequate for the purpose in view. It

c"^^^^- is the writer's object, however, not to give

a complete list of the Rigas known to any individual

singer or singers, but to point out a method by which

the scales of Rigas may be reduced to writing and

classified. A slight consideration will show that all possible

Grima scales and Chromatic scales may be put together from

the twenty-two srutis accepted in the Indian harmonium.

The only ones omitted are those which depart from modern

usage in having a flattened pa. It is only with regard to

Irregular scales, therefore, that the above criticism has any

force, and in their case it should be remembered that, although

THE STAFF NOTATION AND THE SRUTIS 33

the exuberant fancy of Indian composers has led to an extra-
ordinary number and variety of scales, the Grama scales and
Chromatic scales greatly preponderate in point of universality
and popular favour. Again, the Irregular scales, so far as the
writer's experience goes, embody nothing new in principle,
beyond the diminishing of the semitone exemplified in
Jayjayvanti, and the use of the septimal intervals ^, ^ and f,V
It may be noted that the septimal diminished fifth Q) is
apparently unknown to Indian musicians as a melodic interval.
The possibilities opened up by these septimal harmonies are
completely summed up below .^ The figure 7 (above) signifies
that the natural (or tivra) note on which it is placed is flattened
in order to give a septimal interval.

^^-•^y^j^^i^^^El^gizg^^^

ri-

-i-;T

f^^^^^^^E^^^.

:*^

S 7

\^

fe^ij^pZr^^-f^-^^pai^szTlz^

Vr? ^^Vr^

Before concluding this chapter it will be convenient, now
that the nature of harmonic intervals has been ^.j^ . j.
discussed, to show in what way they differ equal tempera-
from the tempered tones and semitones of '"^"^*
Europe. The European system of equal temperament is a
modern creation ; it originated in a desire to have unfettered
liberty in the choice of keys and in modulating during the
course of a composition from one key to another. For this
purpose it was necessary that scales of exactly the same
pattern should be available in every key. This object was
attained by dividing the octave into twelve exactly equal
intervals termed semitones. Instruments with fixed keys like

' The progressions avoided involve either a flattening of sa or pa, or an
incursion into " the sharp keys," both of which would be contrary to the genius
of modern Indian music.

D

34 INTRODUCTION TO STUDY OF INDIAN MUSIC

the organ, piano, and various wind instruments were con-
structed and tuned accordingly, and this artificial system has
been assimilated to such an extent that singers and performers
upon stringed instruments without frets such as the violin
produce tempered music, in perfectly correct intonation,
without being aware in many cases that an element of arti-
fici-ality enters into their performances. Indian airs and
European Folk-songs, when sung without harmonized accom-
paniment in tempered notes, are exceedingly harsh ; the
addition of harmony tends to disguise the defective intonation,
because the mind which grasps the harmonies intended has the
capacity of correcti' g small errors. But the harmonizing of
Folk-songs and of Indian Ragas in tempered harmonies can
never be recommended, except by those whose acquaintance
with natural intervals is too slight for them fully to appreciate
their superiority.

In order to divide the octave (f) into two exactly equal

. , . ,.. intervals one must make use of the ratio
Analysis of the _

swlef*^^ > because this ratio when multiplied by itself

makes the ratio of the octave. Similarly, in order to divide
the octave into three equal parts the ratio required would be

. The tempered semitone, twelve of which make an

octave, has therefore the ratio . The same idea has been

developed further in order to afford a simpler means than
ratios of comparing intervals. Each semitone is divided in the
same manner into a hundred imaginary intervals called cents.
The cent is therefore the twelve-hundredth equal part of an
octave, and a hundred cents make a tempered semitone, and
two hundred cents a tempered tone. With the aid of
logarithms it is a simple matter to turn ratios into cents ; in
this way it is discovered that the major-tone contains 204, the
minor-tone 182, and the semitone 112. The practical advan-
tage gained by this method is that intervals may be graphically
compared (as in Appendix A), and the addition or subtraction
of intervals simply means the addition or subtraction of the
numbers of cents they contain, whereas it is necessary in the
case of ratios to multiply and divide.

THE STAFF NOTATION AND THE SRUTIS 35

The demerits of the tempered scale as a vehicle for the
expression of Indian musical thought may be Eaual temoera-
summed up as follows : — ment unsuited to

1. The distinction between the interval of Indian music.

4 srutis and that of 3 is obliterated. Bihag is confused with
Bilaval, and Bhup with Deshkar, while it is impossible to give
any idea of such Ragas as Kafi, Bhimpalashi, Rageshvari.

2. The atikomal and tartivra notes cannot be expressed.

3. All intervals are out of tune.

The minor-third is 300 cents instead of 316; the major-
third 400 instead of 386 ; the fourth 500 instead of 498 ; and
the fifth 700 instead of 702. A high degree of skill or training
is not required to detect the errors in the thirds ; they are
flagrant ; in fact, the minor-third is much nearer the dis-
sonance of the third in Kafi (sa to ga ^, 294 cents), and the
major-third to the interval from ma ^ to dha Q, another
dissonance, 408 cents.

One can only conclude that Indian writers who openly

advocate the use of tempered instruments are ^.

r 1 • . , ^ . The ordinary

unaware of their utter madequacy to give any Western

idea of Indian intonation. A word of warning notation equally

unsuitable,
appears to be needed by others, who, although

not in favour of tempered music, are ensnared by Western

notation. They should remember that Western notation,

without drastic changes such as those here recommended, is

as detrimental to their music as the tempered harmonium.

It is a tempered notation. The extent to which it confuses

the Indian intervals may be seen from the following : —

600

800

900

36 INTRODUCTION TO STUDY OF INDIAN MUSIC

Each of the tempered notes shown below takes the place
of two ; the extent of error involved in the compromise is
evident from the figures of cents given. The keys of
Western music with their key-signatures are part and parcel
of the tempered system, and are not suited to the Indian
Ragas. Some of the latter might be written under a key-
signature adapted from those in use in Europe, but the
majority could not, and nothing could be gained thereby. A
great deal of Mr. Mudliar's book (alluded to in the previous
chapter) is taken up with a lengthy description of the keys
and key-signatures of European music. The information is
of no value except to those who can take pleasure in hearing
an approximation to Indian airs played upon the tempered
harmonium.

CHAPTER III

THE STAFF NOTATION {continued)

It is now necessary to explain how Indian music may be
represented accurately and in detail by the Staff Notation. In
doing this the author will have to go over ground which has
already been traversed by other writers. Where he has struck
out a line of his own, his object has been to secure the maxi-
mum of simplicity with the minimum of innovation.

As already explained, sound is represented by notes.
Silence is shown by what are called "rests." The duration of
silence (vishranti) or sound (svara) is indicated by the shape
of the rests or notes used.

Note.

IMI

Corresponding
rest.

English name.

Breve
Semibreve

Minim

Crotchet

Quaver

Semiquaver

Demi-semiquaver

Scml-demi-semiquaver

Indian equiva-
lent in matras.

38 INTRODUCTION TO STUDY OF INDIAN MUSIC

There are three degrees of speed in Indian music : vilamhit,
or slow, madhya, or moderate, and driita, or fast. Whatever
the speed is, each measure (avarta) consists of a fixed number
of units of duration called matras. The mdtra in madhya
time will be represented by the note known as the crotchet
(^'). In slow time it will be the minim (^J), and in fast time
the quaver (^^). The relative values of notes and rests and
their equivalents in matras of madhya time are as in the
preceding page.

Smaller notes and rests still may be obtained by adding
more hooks to the stem.

A dot after a note makes it half as long again. The use
of dots with rests is not recommended. Two dots make a
note I 4- ^ + i or i| the length. Thus —

Just as rests are combined to make up the value required,
notes may be combined where it is necessary or more con-
venient to do so in order to make the phrasing or grouping
of notes clearer. In this case the tie or bitid is used. Two
notes of the same pitch joined by a tie are played as one note
of the combined value of the two, thus : J 0=2 matras ;

2), matras ; ^ 4^

I V matras ;

.^ =

= il

matras ; •*

^ =

I matra.

THE STAFF NOTATION 39

The groHpi)ig of notes is a guide to the accent, the first
note of the group being generally more accented than the fol-
lowing ones. It is also, when used with a figure and slur, a
means of showing the division of the matra (or other larger or
smaller unit of duration) into a number of equal parts which
is not a multiple of two. Grouping is effected by joining the

hooks of notes, thus J^iJ • Grouping by slurs may extend

over any passage of melody ; it means that the passage covered
by the slur is to be played legato, i.e. smoothly, as one phrase.
Grouping to show a division of the time-unit into equal parts is
effected in the following manner : —

(i) J J J means 3 notes equal in duration to 2, i.e. each

v,^ I matra.
/2\ n = two notes of \ matra each.

I I I
( ^) #1 ^ •• = three notes of 1 matra each.

(4) ^Ts S = four notes of \ matra each.
(c) tfSJdd = five notes of \ matra each.

(6) ^TTm ^ ^ = six notes of \ matra each.

(7) |=|sq=p:)=B = seven notes of \ matra each.

(8) JJJ^^^Jt^ = eight notes of \ matra each.

(9) J ^ J J J J ^J^ — "i"^ notes of J, matra each.

In No. 6 the accent comes on the first, third, and fifth
notes of the group. If two accents only are desired, the
grouping is effected thus : —

{6) dt^4f €^SS Here each group totals \ matra.

It will be noticed that the same number of hooks is retained from
one multiple of two in the dividend of the fraction, which shows what

40 INTRODUCTION TO STUDY OF INDIAN MUSIC

proportion each note bears to the unit, to the next multiple of two. One
hook is used for half-matras and third matras ; two hooks for fourths,
fifths, sixths, sevenths ; three hooks for eighths, ninths, etc.

Grouping by Slurs. — To avoid an unnecessary multiplicity
of small rests, it is laid down that the last note of a phrase
grouped by a slur is to be played as if of half the length shown,
the remaining half being a rest. This rule applies where the
note is a hooked one ; in the case of longer notes, it is better
to show the exact amount of rest desired. The curved line
exactly similar to a tie used in the groupings below is desig-
nated a slur. It may be used of any length to show the
phrasing or legato grouping of a melody.

1 1 i/t\iiii 1 I I , 1 '/'A

2 2 ZyiJ'S 'g TS 'S 3 3 2 A 2 A\i}

Below each note is written its matra-value, rests being shown
in brackets. The manner in which the four demi-semiquavers
are grouped with the three following quavers should be noted.
A more cumbrous way of writing this passage would be —

It is clear that the last note of a phrase must be cut off or
detached from what follows ; hence the use of a slur implies a
short rest. The rule above stated as to the time-value of such
rests is founded on the practice of Indian singers.

Sometimes notes which do not come at the end of a legato
phrase are played in a detached manner. This is called
staccato. There are three degrees of staccato, shown by (i) a
slur and dots above or beneath the notes, (2) by dots alone,
(3) by dashes. The effect of each is here shown in the same
manner as above by writing the "laya" or matra-values in
fractions.

4 (i) 4 (i) 2 (2) 2 (2) vsd 5)ts(t'5) TsdyiVdlj) s (s) s (») I (s) s (s)

THE STAFF NOTATION

41

Emphasis or accent is sometimes shown by grouping, as
above stated ; it also follows from the nature of the measure,
a subject to be dealt with ; occasionally a special emphasis is
given to a particular note apart from that which comes natu-
rally from the measure or the phrasing. Emphasis of this kind
is denoted by the signs A or r> placed above or below the note
in question. When very strong emphasis is required the letters
Sf are used.

Music set to time is divided into measures (a-vartas), which
are marked off by upright lines or dars. The measure consists
of a stated number of matras with accents occurring at regular
intervals. Accents are : strong (sam), medium (tali), or weak
(khal). The strong accent follows immediately after the bar ;
the medium accent follows the dotted bar ; and the weak
accent follows the dotted half-bar. The character of the
measure may be shown by a time signature. A very good
form of signature is one adopted by Mr. H. P. Krishnarao, of
Mysore ("First Steps in Hindu Music," Weekes & Co., Lon-
don). Above are noted the matras upon which the chief
accents fall, and below the number of matras in a measure

and their time-value. Thus -^^ means that there are eight

O 1

matras represented by crotchets in the measure, and that the
accent falls on the ist, 5th, and 7th. The addition of sub-
sidiary dotted bars is useful, as they show at a glance the
phrasing of the melody. The measures in most frequent use
are —

I. Adital, trital or tintal. ^-'^^- (Panjabi is slower).

=nn"

ziz*r^^dzit^*z*=l

=q=

^^^

II. Chautal. '--^ — (Ekkatal differs as regards drumming).

* The dha is ^ like the preceding one. See the paragraph below on " Acci-
dentals."

42 INTRODUCTION TO STUDY OF INDIAN MUSIC
III. Zhampa. -^'^

-I-

-M±

IV. Surphakta.

1.5.7

^^^^m^m^^^^s^

V. Dhamar.

I. 6. II

'$^-

^^EW

:j«i:^P

:te::pr|ip==ir

;t=:

l^-T-

=^^4=i=

:*=^^:

VI. Ada Chautal

I-3-7-II

l»l^-,

VII, Tevra.

T.4.6

j^i-glj^^:i^si|iSgi^il^^j;

VIII. MattaTal. ''•^•^•^•7-^.

9r

g

=!>»zz

IX. Saviri. I -5 -9. 13. 14.

Si

=P=P=^

-T^ha— I T=4^-

F*<«==

X. Zhumbra. — ^^^

^igS^=f=E^-^^g^£^|E^^^i|

XI. Dipchandi. ''^'
14^

rrlpzerizmfz^ii

^|E^p3^Ei=Ur±=z|=P:

THE STAFF NOTATION

43

XII. Dhumali.* L^^i^hli

bzi^zzTT^czTi^irzi

XIII. Dadra

i^^£liii=3ii^s

^^^^^^

XIV. Kerava.

ilgi«

55Sii^

-^^

The sign placed before a note, which shows which sruti of
that svara is to be used, is called an accidental. Where the
same sruti occurs more than once in an avarta, or measure,
the accidental need be used once only. The object of this
rule is to avoid an unnecessary number of accidentals.

Thus, in Example IV., first complete bar, there are three komal
madhyams, the accidental "fe being used for the first only. If the second,
however, had been tivra, and the third again komal, it would have been
necessary to place the sign tj before the second, and R before the third.
It will be seen that the rule only applies where the sruti is not changed.

It is necessary to adopt terms arid signs which will show
the intensity or degree of loudness or softness of a composi-
tion. The following terms are in general use in European
music ; they represent fixed degrees of intensity : —

Sign.

Term.

Rendering.

PP

pianissimo

very soft

P

piano

soft

Vlp

mezzo-piano

moderately soft

m/

mezzo-forte

moderately loud

f

forte

loud

ff

fortissimo

very loud

A gradual increase of intensity is shown by the
sign -=mi!II and a gradual decrease by IIIIir=-

* Dhumali is not differentiated from Tintal except in a certain kind of
solemn music.

44 INTRODUCTION TO STUDY OF INDIAN MUSIC

A dotible-bar

is used to mark the end of a com-

position or of any section thereof, such as the " astai " or
•' antara." It may occur at the end of or at any place in a
measure or " avarta." When used with the sign /tn, which

is known as the pause, thus i=:Bzi:, it signifies the end of

the whole composition. A double-bar with dots —

is a direction that the composition is to be repeated up to
that point, either {a) from the beginning, or {b) where the

double-bar in this form ^Rl^ has been previously en-
countered, from that point. Double-bars should be so arranged
that the integrity of the measures is preserved. In the follow-
ing example, the first and third bars together form in the
repeat one whole bar of eight matras.

Astai.

w

:^

q^=p:

4=1=4

E^^Ei^Ei

X

©

! i I -

:il=i):

ws^^^^

Occasionally the last bar before the repeat sign is changed
after the repetition, so as to lead to the next section. This
is shown by the use of the numerals i and 2. The portion
marked i is to be performed the first time, and at each repeti-
tion except the last. The part marked 2 is to be performed
at the last repetition in substitution for i, which is omitted.
For instance, the last portion of the above example might be
written as follows : —

Antara.

m^^^^^^^m^

In the next example there is no change.

Antara.

THE STAFF NOTATION 45

It is convenient to write such words as astai and antara at
the beginning of the appropriate sections. Such directions
as " repeat astai once or twice," when used with a double bar,
will be easily understood and will save writing preceding
sections again in full.

If it is required to repeat a composition from the beginning
or from a certain point and this cannot be conveniently indi-
cated by a double bar and dots, the directions " repeat from
beginning " and " repeat from % " may be used respectively,
the sign S being placed over the point from which the repeat
is to commence.

EmbellisJiments or ornaments are much used in Indian
music. With the exception of the mend and ghasit, they
consist chiefly of short notes the time of which is taken out
of that of the note which they adorn. These notes are some-
times " diatonic," that is to say, taken from the scale of the
Raga, and sometimes what are generally spoken of as quarter-
tones. The author makes no attempt to classify them or
define them ; it is sufficient for his purpose to describe those
which require special signs to express them. Others may be
represented by the actual notes and in the actual time by
means of the signs and notes already described.

The mend is a glide from one note to another. So also is
the ghasit, but they differ in this, that whereas the ghasit
proceeds from a grace-note and is quick and somewhat
violent, the mend passes over all intermediate sounds gently
and is sometimes allowed to dwell for the briefest possible
moment on the diatonic notes. Their signs are —

Mend. Ghasit.

=1:

l^g^

Thok is an ornament applied to the first note in a mend.
It consists in attacking the note forcibly, blending it at the
same time with the "quarter-tone " above. It is primarily an
ornament obtained by plucking a string with emphasis while
the left hand almost simultaneously stretches the string to
one side. The sign here adopted is taken from the sign of
emphasis (>).

46 INTRODUCTION TO STUDY OF INDIAN MUSIC

The khatka may be represented by grace-notes ; no
special sign is needed. Grace-notes are minute in point of
time ; they take their time out of that allotted to the note to
which they are attached, and are represented by small notes.
A single grace-note is usually cut through with a line as
shown, in order to distinguish it from a peculiar kind of note
known in European music as the "appoggiatura."

. Khatka.

Two kinds of trill are in common use, gamak and bhelava.
The first is somewhat like a " mordent " with quarter-tones ;
it may be represented thus : —

Bhelava comes after a mend, and consists of a slow trill
with the next higher or lower sruti. It resembles the vibrato.

The ordinary trill of Europe is performed in combination
with the next note above in the scale, and is denoted by the
sign tr.

The writer's acknowledgments are due to Abdul Karim,
some of whose melodies ^ have been used or mutilated to
serve as examples, and to Yeshvantrao Dinkar Bhramanalkar,
who has published in a kind of " tonic-solfa " notation several
of Abdul Karim's melodies.^

* "Sangit Svaraprakash." Sriramtatva-Prakash Printing Press, Belgaum.

• "Jain Bhajanamrit Padyavali." Printed at the Arya Bhushan Press, Poona.

CHAPTER IV

THE INTERPRETATION OF THE ANCIENT TEXT-BOOKS

This chapter will include annotated translations of the most
important parts of the Natya Shastra of Bharata, which is
generally allotted to the fifth century A.D., and of the
thirteenth-century text-book, the Sangit Ratnakar of
Sarangdev. The writer knows of no detailed exposition of
Indian musical theory in any treatise except Bharata's Natya
Shastra, earlier than the Sangit Ratnakar. Regarding later
works, it may be said without fear of contradiction that the
Ratnakar has been consistently misunderstood by all suc-
ceeding authors.

To understand the two works named one must grasp two
important facts, namely —

(i) That the system of tuning upon which they are based
is that mentioned at p. 109 of Captain Day's "Music of
Southern India." The chanterelle strings in such instruments
as the vina and sitar were dha "f (the bass note), and ri ^ and
ga k The drone used was almost invariably dha + and ri •{.
combined. These were probably the notes given by the
tambura. The wire upon which the melody was played
was ri +.

(2) The sruti was not an exact unit of measurement.
Two srutis made a just semitone, three a minor-tone, and
four a major-tone.

The evidence to establish the first of the facts mentioned
is as follows : —

(i) Although the present writer has not been able to
come across a vina-player who tunes his instrument in the
method named, it is clear from Captain Day that the method
survived until modern times in the Hindustani school of
music.

48 INTRODUCTION TO STUDY OF INDIAN MUSIC

(2) The tuning method of Sarangdev could not have been
the same as that now in vogue. Sa is omitted in several of
the hexatonic and pentatonic scales mentioned by him.

(3) The process of Grama-Sadharan discussed below is
almost sufficient in itself to prove that two of the open strings
were tuned to ri •}« and dha •}..

(4) This is not, perhaps, a very strong inference, but the
name shadj itself — born of the sixth — shows that shadj was
not always the fundamental note of music as it is now, and is
compatible with the assignment of a more prominent position
to dhaivat, for shadj is six srutis removed from dhaivat.

(5) It will be found that it is quite unnecessary to make
any verbal alterations or to do violence to the natural meaning
of the text, if the two facts mentioned are accepted as a basis
of interpretation. The general method adopted by modern
Indian writers in dealing with the Ratnakar, is either to quote
passages which obviously have not the remotest application to
the modern system, without any attempt at comment or
explanation, or to distort the meaning so as to make it
applicable. For instance, Rajah S.M. Tagore takes the scale
of Bihag as his shuddh scale (unconscious of the circumstance
that Indian music has outgrown its " shuddh svaras "), and
endeavours to fit it into the scale which Sarangdev calls
Shadj i. The two scales are here shown side by side, with the
sruti intervals below in figures —

n Shadji. , Bihag. ,

3^E^

^^js^^^^" I I

3244324 4324432

Rajah S.M, Tagore's argument is apparently as follows :
" Our scale of Bihag must be the same as that of Shadji. If
we take the srutis of sa which are four in number to be those
above it instead of those which separate it from ni, the srutis
allotted to each note work out the same. Hence it is reason-
able to conclude that the srutis of a note are really those
which separate it from the note above, and not from the note
below. If there are any passages in the Sangit Ratnakar
which are opposed to this view, there must be some error in

the text."

(6) There is a very significant passage in the Sangit

INTERPRETATION OF ANCIENT TEXT-BOOKS 49

Ratnakar which shows that even in his time there were rival
schools in the matter of tuning. In Prakaran IV., verses 14,
1 5 are to the following effect : " Some say that the last six
Murchhanas of the Shadj Grama are got by putting each of
the notes ni, dha, pa, ma, ga, ri, respectively on sa. The
same principle may be adopted in the Madhyam Grama.
That is to say, in place of the original first Murchhana, ma,
pa, dha, ni, sa, ri, ga, ma, is to be read ga, ma, pa, dha, ni, sa,
ri, ga, and so on."

This is the same as the modern system except that now-
adays the Gramas are not kept distinct, and the difference of
pitch from sa to ma is not used to distinguish one from the
other.

Bharata: Natya Shastra (Ch. 28). Text and

Coiiimentary.

(Translated from the Bombay Edition) — Kavya Mala series.
No. 42. Nirnaya Sagar Press.

Verse 22. — There are seven svaras : shadj, rishabh,
gandhara, madhyam, pancham, dhaivat, and nishad.

Verse 23. — They are of four kinds according to the
number of srutis between them : Vadi, Samvadi, Anuvadi,
Vivadi.

Commentary. — The "Ansha" svara is the same as the
Vadi. Svaras between which there is a distance of nine or
thirteen srutis are Samvadi to each other (examples are
here given).

Verse 24. — In the Madhyam Grama, pa and ri are
Samvadi ; in the Shadj Grama, sa and pa.

Commentary. — Those are Vivadi between which there is a
distance of twenty srutis, e.g. ga and ri, ni and dha. The
Vadi, Samvadi and Vivadi thus established, the rest are called
Anuvadi. (A list of the Anuvadi intervals is given.) The
speaking note is called Vadi ; the note which clashes (with it)
Vivadi ; the note which converses (with it) Samvadi ; the note
which increases the beauty of the Raga, Anuvadi. These
svaras, when played, suffer slight modifications owing to
differences in the wires and keyboard.

Note. — The chief note of a composition (ansha svara) according

E

50 INTRODUCTION TO STUDY OF INDIAN MUSIC

to this passage, is the Vadi. Its Samvadi is either at a distance of a
perfect fourth or fifth. The note at an interval of an octave less a
semitone is Vivadi, and the rest of the notes in the scale Anuvadi. The
classification has an element of harmony in it, but regard is had to the
melodic rather than harmonic point of view. The reader will remember
that the notes here dealt with are the following : —

33^

]±i

m^f^-

^^^.

,^'^

I Shadj Grama. Madhyam Grama.

The double-stemmed notes are the Vadi and Samvadi.

It would be more in consonance with Indian ideas to represent the
Gramas in the form of a closed circle ; as has already been pointed
out, they are not regarded as scales.

Grama and the
Each is said to comprise twenty-two

The Gramas.

There are two Gramas, the Shadj
Madhyam Grama,
srutis.

Verse 25. — In the Shadj Grama the srutis are arranged in
the order following : 3, 2, 4, 4, 3, 2, 4,

Commeiitai-y. — Pa is lowered one sruti in the Madhyam
Grama. The interval passed over in raising or lowering pa
by one sruti is the measure of a sruti.

Note. — This passage, and the difference in harmonic structure
between the Gramas, have been fully explained.* The ascending or
descending line made use of in the accidentals means a difiference of
one sruti or |^, except in two cases, ma -^ and ni f^, which should really
be called pa t? and sa t?. From the passage " each is said to comprise
twenty-two srutis," it might be argued with some show of reason that
the existence of twenty-two srutis need not be taken to imply twenty-two

» Ch. II.

INTERPRETATION OF ANCIENT TEXT-BOOKS 51

individual notes. The answer to this argument is found in the fact that
individual names were allotted to the whole number of srutis.

Verse 26. — Sa is of 4 srutis, ri of 3, ga of 2, ma of 4.

Verse 27. — Pa of 4, dha of 3, ni of 2. This is the Shadj
Grama.

Verse 28. — In the Madhyam Grama, ma is of 4, pa of 3,
dha of 4.

Verse 29. — Ni of 2, sa of 4, ri of 3, ga of 2.

MURCHHANAS.

There are 14 Murchhanas in the two Gramas.

Verse 30. — The first is Uttarmandra, and starts from sa ;
the second, Rajani from ni ; the third, Uttarayata from dha ;
the fourth, Shuddhshadja from pa ; the fifth Matsarikriti
from ma.

Verse 31. — The sixth is Asvakranta from ga ; the seventh
Abhirudgata from ri. Such are the Murchhanas of the Shadj
Grama.

Verse 32. — In the Madhyam Grama the Murchhanas are :
Sauviri, Harinasva, Kalopanata, Shuddhmadhya, Margi,
Pauravi, Hrishyaka.

Verse 33. — Such are the Murchhanas of the Madhyam
Grama. The starting notes are ma, ga, ri, sa, ni, dha, pa.

These Murchhanas are of four descriptions: (i) those
which are Sampurna {i.e. complete with seven notes) ; (2) (3)
the Shadava and Odava {i.e. hexatonic and pentatonic) ; (4)
those which take the Sadharan (common) notes.

Verse 34. — When all seven notes are taken in order, the
Murchhana is called Sampurna. When five or six notes are
taken it becomes Odava or Shadava.

Verse 35. — Sadharan Murchhanas are of two kinds : those
which take the Kakali Sadharan, and those which take the
Antara Sadharan.

Coninientary. — Each Murchhana is produced in two ways.
Thus, in the Shadj Grama, when ga, in the Murchhana or
Grama, is raised two srutis and made into dha. (So also in
the Madhyam Grama, Murchhanas assume two forms, (i)
when the dha is lowered, (2) when the ni is raised.) This
is calling the same thing by another name, because the sruti-
intervals are not changed. The interval between pa and dha

52 INTRODUCTION TO STUDY OF INDIAN MUSIC

is four srutis. Even so, raising gandhara gives an interval of
four srutis. The remaining svaras likewise become madhya-
madi {i.e. relations of madhyam).

Note. — (i) Murchhana is a high-sounding name connoting h'ttle.
In Sarangdev's time, as the examples given by him clearly show, the name
of the Murchhana merely indicated the lower compass of a composition.
It would perhaps have been more logical to treat the subject of
"Sadharan" under the head of Jatis. However, Sarangdev also deals
with it under the section relating to Murchhanas.

(2) The Kakali Sadharan was ni t| ; the Antara Sadharan ga tf-
They appear to have been employed as alternative notes to the shuddh
ones for melodic effect.

(3) The commentary below verse 35 has hitherto proved a stumbling-
block to students. The brackets have been added by the present writer.
They enclose a sentence which is to be separated from the rest of the
text, as it is merely a digression by way of analogy. Ga is raised two
srutis, that is to ga tj, and becomes " dha." That means that the scale
arrived at is at once realized as in the Madhyam Grama, if ga is called
dha.

^^^^^^wr^"^

S^^

It:

S

'ma' 'pa' 'dha' 'ni' 'sa' 'ri' 'ga' 'ma' ma pa dha ni sa ri ga ma

The scale arrived at by change of ga is that on the left ; the one
on the right is the first Murchhana of the Madhyam Grama. It will
be seen that, except in point of pitch, the two scales are identical.

Again, lowering the dha in the Madhyam Grama gives —

fe^^i^

_i,=Jfeei^

w

which is the first Murchhana of the Shadj Grama transposed a fourth
higher. If the reader will work out the ratios for himself, he will find that
this is so. The inclined line in all cases means lowered by fj. Raising
the ni of the Madhyam Grama gives

^

^^

i^Efe

It:

which, transposed, is the following : —

-A

-^l

it=^3^;

:^^_^

m

INTERPRETATION OF ANCIENT TEXT-BOOKS 53

This is an interesting scale ; tivra ma must be what is known in
Europe as an auxiliary note, that is auxiliary to pa, from which it would
not be separated. The reader may ask, with respect to the first two
of the scales above, why the lowering of pa should not have been con-
sidered sufficient for the change of Gramas without the use of dha komal
and antara ga. The writer thinks that the practice originated partly
from convenience of pitch, and partly to suit the drone of the open
strings. The question will be more fully discussed in treating of the
Ratnakar Jatis. [The first scale gives the Jati Panchami when ri "f (an
open string) is taken as the starting-point.]

Nothing would be gained by transcribing further passages from
Bharata. His Jatis are the same as Sarangdev's.

THE SANGIT RATNAKAR.
{Ananddshrama Series^ Poona).
SVARADHYAYA (PRAKARAN 3).

Verses 3 to 1 1 are taken up with an attempt to explain
the phenomena of sound.

Verse 12. — Take two vinas with 22 wires each and tune as
follows : Let the first wire give the lowest possible note, the
next a note a little higher and so on, so that between the
notes given by any two adjacent wires a third note is
impossible.

Verses 1$ to 16. — These successive notes are the srutis.
Sa will stand on the fourth wire, being a svara of four srutis ;
ri will be on the third wire counting from the fifth ; ga, which
has only two srutis, will fall on the second, counting from the
eighth ; ma, being of four srutis, on the fourth, counting from
the tenth ; pa on the fourth, counting from the fourteenth ;
dha on the third after pa ; ni on the second after dha ; so
ni will fall on the twenty-second sruti.

Note. — Verses 24 to 38 are taken up with the various names of the
svaras and srutis. These names are to be found on p. 77, below. As
regards dipta, ayata, etc., the commentator, Kallinath, says, "It is said
that this classification is based upon the effects which the srutis are
supposed to produce."

Verse 40. — Thus are produced the seven notes of the
mandra saptak (lowest octave or chest register) ; the seven

54 INTRODUCTION TO STUDY OF INDIAN MUSIC

notes of the madhya (middle octave or throat register) ; and
the seven notes of the tar saptak (highest octave or head
register). These seven notes are called the shuddh svaras.

Note. — The notes are those of the Shadj Grama (p. 50).

Ve?-se 41. — When the shuddh svaras change their places
among the srutis, they become vikrit svaras. There are 12
vikrits.

Verses 42, 43. — Sa has two vikrit positions — Chyut and
Achyut. In both it is of two srutis. It becomes Chyut when
ni takes its first sruti and changes to kaishik ni. Becoming
Chyut, it leaves its unused (fourth) sruti to ri ; thereby ri
becomes vikrit because its distance from sa is increased.

Sa becomes Achyut (not fallen or unmoved) when ni is
raised two srutis and made kakali ni.

So, when ni becomes a svara of three srutis, by taking one
from sa,sa is lowered one sruti, and becomes Chyut ; when ni
becomes a svara of four srutis, by taking two from sa, sa
becomes Achyut.

Verse 44. — Sadharan and Antara are the two vikrit states
of g^ ; Sadharan of three srutis and Antara of four srutis.

Verse 45. — Ma, like sa, has two vikrit conditions, Chyut
and Achyut. When ga takes its first sruti, ma, in order to
stand at a distance of two srutis from this Sadharan ga,
becomes Chyut ; again, when ga is raised to the second sruti
of ma, becoming Antara, ma stands in its original place,
but is called Achyut. It has lost two srutis, and so has become
vikrit.

Verse 46. — Pa also has two vikrits. First, when it becomes
a shuddh svara of the Madhyam Grama ; secondly, it will
become Chyut or Kaishik, when ma, the preceding note, is
Chyut. It must fall one sruti to keep four srutis from Chyut
ma. When pa becomes Kaishik, its last sruti goes to dha and
makes dha vikrit.

Verse 47. — When ni takes the first sruti of sa, it is
Kaishik ; and when it takes two, it is Kakali. These are
the two vikrits of ni.

Such are the twelve vikrits. The total of shuddh and
vikrit svaras is nineteen.

Note. — The above classification of vikiits is inexact and confused.

INTERPRETATION OF ANCIENT TEXT-BOOKS 55

It was probably in vogue among musicians, and Sarangdev took it as he
found it. Out of the twelve vikrits, seven, namely, chyut sa, ma and
pa, antara and sadharan ga, kaishik and kakali ni, are new ; the rest are
old svaras with new names. When chyut sa, ma and pa are made use of,
the scale assumes the form here given : —

g

w*^ bog a, .G ^ f

These changes (it is important to notice that the only notes which
escape flattening or raising are ri and dha) are known as Gfdma-
SAdhdran. The real sadharan ga and kaishik ni, corresponding to the
modern Karnatic notes of those names, are to be seen in the Gandhdra
Grama (see below). Here ga and ni become of three srutis, i.e. a minor-
tone removed from ri and dha respectively. They are ga •)« ^nd ni •{. as
shown, and not gal?. and nit?. The confusion of nomenclature is due to
the theory of the equality of the srutis. The inclined lines all mean a
flattening by |J. If one is removed from each note of the scale, that of
Bihag results, of which, as has been pointed out, Kafi is the mode of the
second. In other words, the above collection of notes taken from ri "I. to
ri "j. gives the Jati known as Shadji (to be explained later). Grama-
Sadharan was a tuning device adopted to enable the performer to play
Shadji to a ri -f drone.

The lowering of pa in the Madhyam Grama has already been ex-
plained. The manner in which antara ga and kakali ni are used is
dealt with under the head " Sddharan." Achyut is merely a fanciful
epithet.

Verse 48. — (Gives the names of animals which utter the
shuddh notes.)

Note.— This is interesting merely as showing that there was some
idea, as there is to-day also, of absolute pitch.

Verses 49, 50. — Svaras are divided into four classes —
Vadi, Samvadi, Anuvadi, Vivadi. The note which occurs
most frequently in a song is the Vadi.

Verse 51. — Samvadi svaras are those between which are
eight or twelve srutis.

Ni and ga are Vivadi (discordant) with all others ; in
other words, ni and ga are respectively Vivadi with ri and
dha ; ri and dha are respectively Vivadi with ga and ni.

56 INTRODUCTION TO STUDY OF INDIAN MUSIC

Svaras which are neither Vadi, Samvadi, nor Vivadi are
called Anuvadi.

Note. — Ni and ri shuddh are not Vivadi ; perhaps the author had in
mind ni kakaH. Even then his opinion conflicts with Bharata. It will
be noted that to have eight or twelve srutis between, means intervals of
nine and thirteen srutis respectively.

There is nothing important in verses 52 to 60.

The Chapter on Gramas and Murchhanas
(Prakaran 4).

Verse i. — A Grama is a collection of svaras. It is the
foundation upon which Murchhanas are built. There are only
two Gramas in this world. The chief Grama is the Shadj
Grama.

Verse 2. — The other Grama is the Madhyam Grama.
When pancham stands on its proper fourth sruti, that is the
Shadj Grama arrangement of svaras.

Verse 3. — When pancham is lowered one sruti and stands
on its third sruti, that is the Madhyan Grama. To express
it in another way, in the Shadj Grama, dhaivat is of three
srutis, and in the Madhyam Grama, it is of four srutis.

Verse 4. — When gandhara takes one sruti from ri, and one
from ma, that is the Gandhara Grama. In this arrangement
dha also takes a sruti from pa ; and ni takes one from dha and
one from sa.

Verse 5. — Narada called this arrangement the Gandhara
Grama. This Grama is practised by celestial musicians.

Note. — The Gdiidhdra Grdnia has always presented difficulties to
the student, and has always proved an attractive problem in spite of the
fact that it was obsolete in Sarangdev's time. " Ga takes one sruti from
ri " means that ri is lowered from ri "f to ri !?. At the same time ga takes
one from ma, and so rises from ga \ to gat?. Dha takes one from pa,
making pa " chyut." Behaving like ga, ni sends dha down to dha f? and
itself rises to nit?.

:=j^g*^±g:

^^V^^^

=t:

The difficulty of the problem attaches to the chyut pa, which divides
the six-sruti interval between ma and dha into two intervals of three srutis.

INTERPRETATION OF ANCIENT TEXT-BOOKS 57

Now, two minor-tones -\p- x \y are greater than a minor-third (5 or six
srutis). Here, then, is another practical example of the inequality of the
srutis. If pa is taken to be a minor-tone or three srutis below dha 1?, the
interval separating it from ma ^ will be ||, a difficult interval to sing.
This may account for the disappearance of the Gandhara Grama. The
differences in harmonic structure between the three Gramas will be
apparent from the following : —

First Murchhana at pitch sa.

Shadj Gn\ma. Madhyam Grama.

Ei3^

^^z

e^s^eeeB

Gandhara Grama.

i^li^s^SS

First Murchhana at pitch ri.

^=m-

=p=^

iEgE^^=^^^

i^tusi

:^=:^2=^:

:t=t:

--!-

:^z^^=^=g

:?2=^;

t=t

Second Murchhana at pitch sCi.

F=i^

— I-

~^t=^-^

^=?2IZ^

--^~m

l=t::

t=t

-^-^z

:^=^=^:

:^=^

t=t

izz:

-^-^-^-r^-F

i

2:^

Common Chords.

m

s

m

:^g=

r

The only possible alternative solution is to regard chyut pa as the
eleventh upper partial of ri |7. This would give a beautiful scale with the
intervals : ga to ma ^Q ; ma to pa jj ; pa to dha -}f ; dha to ni » ; ni to
sa iji ; sa to ri i-J ; ri to ga ;!:;.

Verse 6. — The chief Grama is the Shadj Grama for two
reasons : (i) It is the first of the series. (2) It has more
Samvadis. The Madhyam Grama is also important because
that svara is never omitted.

58 INTRODUCTION TO STUDY OF INDIAN MUSIC

Note. — The reasoning of this verse is not very convincing. The last
sentence refers to " Shuddh Tans '' (see below).

Verse g. — A Murchhana consists of the ascent and descent
of the seven svaras in their natural order. Each Grama has
seven Murchhanas.

Verses lo to 12 name the Murchhanas. The names are
those given by Bharata (p. 51).

VtTse 13. — The first Murchhana of the Shadj Grama starts
from sa of the middle register. The others start respectively
from ni, dha, pa, ma, ga, ri in the lower octave. The first of
the Madhyam Grama starts from ma of the middle octave.
The remaining six begin from ga, ri, sa, ni, dha, pa (the last
three of the lower octave).

Verse 14. — Some say that the last six Mtirchhajias of the
Shadj Grdina are got by pjitting each of tfie notes ni, dha, pa, ma,
ga, ri respectively oji sd.

Verse 15. — The same principle may be adopted in the
Madhyam Grdina. That is to say, i?i place of the original fj-st
Mjirchliana, ma, pa, dha, ?ii, sd, ri, ga, ma, is to be read ga, ma,
pa, dha, ni, sd, ri, ga, attd so on.

Note. — This is the method now universally accepted, with this varia-
tion, that Ragas of both Gramas commence their scales from sa. It is,
however, not the method treated of in Ratnakar. The reader should
remember that these discussions refer more particularly to Jatis than to
Murchhanas. (See p. 52.)

Verse 17. — Murchhanas are divided into four classes : —
(i) Those which take the shuddh svaras.

(2) Those which take kakali ni and the rest shuddh.

(3) Those which take antara ga and the rest shuddh.

(4) Those which take antara ga, kakali ni, and the rest
shuddh.

In this way will be obtained 14 X 4, or 56 Murchhanas
in all.

Note. — This classification has the fault of overlapping. To show
that this is so, the scales are here given : —

INTERPRETATION OF ANCIENT TEXT-BOOKS 59

Shadj Grama
2nd

'• I 3rd.

---^—t

z^.

^z:^^:

p— p-

Madhyam Grama.

^-^-

1 ^

4th.

— i— I — t d

:^:*f^fzEl

:t

^4^^^m

1?fO

The third variety of the Shadj Grama is the Madhyam Grama pure
and simple, transposed a fourth below. The fourth variety is the third
of the Madhyam Grama transposed a fourth below.

The writer suspects that verse 17 is probably merely inserted in order
to lead up to one of those arithmetical calculations which Sarangdev so
delights in.

The Chapter on Shuddh Tans (Prakaran 5).

Verse 27. — When Shuddh Murchhanas are made Shadava
or Odava they become Shuddh Tans.

Note. — Sarangdev distinguishes such Tans from Kut-Tans. The
latter are simply permutations of a given number of svaras, such as sd ri
ga, sd ga ri, ri ga sa, etc.

The rest of the chapter is taken up with a calculation of the number
of possible combinations of six and five svaras in the two Gramas.

The Chapter on Sadharan (Prakaran 6).

Note. — Bharata, after counting up the Tans, tackled this subject also.
In verse 36, he says : " It is chilly in the shade, but in the sun one per-
spires ; spring is coming, but winter has not yet departed." This is the
Sadharan of the seasons. "In music there are two Sadharans — Svara-
sadharan and Jati-sadharan. In Svara-sadharan, ni is raised two srutis ;
it is called kakali, but is not shadj. The word Sadharan is appropriate,
as it stands between ni and sa. Gandhara also becomes SadhS,ran, and
is called antara ga, and not ma. It stands midway between ga and ma."

6o INTRODUCTION TO STUDY OF INDIAN MUSIC

After defining Jati-sadharan, the author returns to the subject and pro-
ceeds—" Svara-sadharan is assigned to the two Gramas. There is
Shadj-sadharan (kakaU ni) in the Shadj Grama, and Madhyam-sadharan
(antara ga) in the Madhyam Grama."

Ratnakar : Verse i. — Sadharan is of two kinds: Svara-
sadharan and Jati-sadharan. Svara-sadharan is of four kinds :
Kakali-sadharan, Antara-sadharan, Shadj-sadharan, Madhyam-
sadharan.

Verse 2. — Kakali is the Sadharan of sa and ni. It is
common to both ; hence the name Sadharan.

Verse 3. — Similarly, antara is the Sadharan of ga and
ma. The singer should sing sa, and then take kakali and
dhaivat in order.

Verse 4. — Likewise, after singing ma, he should take antara
and rishabh, or he may take sa and kakali and then return
to sa.

Verse 5. — After that, if any of the other svaras is to be
omitted, the singer should, omitting it, go to the next. Let
the same thing be understood with the antara.

Verse 6. — Antara and kakali are always sparsely used.

Note. — These notes were therefore auxihary or passing notes ; they
had originally no connection with the chromatic scales. This circum-
stance is of the highest importance in examining the Grama-ragas and
Jatis.

Verse 7. — When ni takes the first sruti of sa and ri takes
its last sruti, Shadj-sadharan occurs.

Verse 8. — In the same way, in the Madhyam Grama, when
ga takes the first sruti of ma and pa takes the last sruti of ma,
Madhyam-sadharan occurs.

Note. — This process is discussed at p. 55. " In the Madhyam
Grama " is merely a way of expressing the fact that pa is chyut. Divid-
ing Grima-sadharan into two parts appears to be merely pedantic.

Verse 9. — These two Sadharans are also called Kaishik
(of a hair's breadth) because of their tenuity. They are also
called Grama-sadharan.

Verse 10. — Jati-sadharan occurs when two Jatis of the
same Grama have an ansha-svara in common. Examples
are to be found in the Ragas themselves.

INTERPRETATION OF ANCIENT TEXT-BOOKS 6i

The Chapter on J axis (Prakaran 7).

Verse i. — Shuddh Jatis are seven in number; their
names, taken from those of the seven principal svaras, are :
Shadji, Arshabhi, Gandhari, Madhyama, Panchami, Dhaivati,
Naishadi.

Verse 2. — A shuddh Jati must satisfy the following five
tests : The Nyas Svara (the final) must be the Nama-Svara
(that note from which the Jati takes its name). So also must
the Graha-Svara (the initial note ; also, in all probability,
either the upper or lower note of the chord forming the pedal
accompaniment or drone) ; so also the Apanyas (final of a
middle cadence, may be rendered *' medial "), and the Ansha
Svara (the chief or prevailing note ; according to Bharata,
the same as the Vadi). The Jati must be Sampurna (complete
with seven notes), and the final must never come in the Tar
(higher) octave.

Verse ^. — When a Jati takes a note other than the Nama-
Svara as medial, initial, or ansha it becomes a vikrit Jati.
But even in this case the last rule as to the final must be
strictly observed.

Note. — According to Bharata, the Ansha Svara is the same as the
Vadi. Judging by modern practice and from the Sarigama examples
given by Sarangdev, one can be certain that the Nyas or final is not
invariably part of the drone. The Graha Svara is more likely to be one
of the drone notes, but it is a curious fact that theory paid no great atten-
tion to the relation of the drone to the scale.

It will not be out of place at this stage to describe in brief the Church
Modes of Europe. Each was a Sampurna scale. It was determined by
three factors— the lowest note (in India, the note which gives its name to
the Murchhana), the final (the Nyas), and the dominant (that note round
which the melody centred). Modes were either Authentic or Plagal. In
the former, the lowest note and the final were the same ; in the latter, the
final was a fourth above the lowest note. The dominant of an Authentic
mode was, with certain necessary exceptions, the fifth of the scale ; the
dominant of a Plagal mode was a third or fourth above the final.
Generally speaking, final and dominant corresponded to Vadi and
Samvadi ; as they were not invariably a fourth or fifth apart, but some-
times a third or sixth, the modes gave more harmonic variety than the
Shuddh Jatis. These modes are still employed for special effects. They
are as follows :—

62 INTRODUCTION TO STUDY OF INDIAN MUSIC

Authentic modes.

Plagal modes.

Verses 4, 5, 7 are taken up with a calculation of the number of per-
mutations and combinations of vikrit Jatis. This is a kind of exercise in
which musical pandits seem to revel. The total arrived at is 153.

In verse 6 it is laid down that no Jati of less than five notes is allowed.
Thus there are three kinds in respect of the number of svaras— Sampurna,
Shadava, and Odava.

Verse 8 states that there are only eleven vikrit Jatis in use. Verses 9
to 20 give their names and other particulars (see below).

Verse 21. — Sages such as Bharata advise the use of
Svara-sadharan {i.e. the use of antara ga or kakali ni) in the
case of Jatis Madhyama, Panchami, and Shadj-ma-dhyama.

Verse 22. — More especially so when the Vadi notes are sa,
ma, and pa. Others such as Kambal and Asvatar recommend
the use of the antara and kakali in Jatis where ga and ni are
weak. According to them this rule is to be observed not only
in Jatis, but in Ragas of all kinds.

Verse 23. — When in the Shadj-madhyama Jati ni and ga
are Vadi, there will be no Svara-sadharan. Moreover, it takes
place in vikrit and not in shuddh Jatis.

Note. — The above rules are intelligible and of considerable importance.

* Final and dominant are shown by asterisks. The accidentals, introduced
by the writer, show that here also there are two Gramas. This fact is not
as widely recognised as it should be.

INTERPRETATION OF ANCIENT TEXT-BOOKS 63

Verses 24 /o 27 detail the number of Vadi or Ansha notes
in the 7 shuddh and 1 1 vikrit Jatis. Sarangdev's observations
on the subject of Vadi and Samvadi appear to be somewhat
deficient in discrimination. One may hazard a conjecture
that as a Jati may take more than one pair of Vadi and
Samvadi, its tonality may change during a composition, or its
derivative Ragas may be distinguished one from another by
the Vadi and Samvadi.

Verses 28, 29. — There are 1 3 criteria by which to deter-
mine a Jati : (i) Graha, (2) Vadi, (3) Nyas, (4) Tar {i.e. the
higher compass, or top note), (5) Mandra (the lowest note, the
one which gives its name to the Murchhana of the Jati), (6)
Apanyas, (7), (8) Sannyas, Vinyas (two varieties of medial,
the latter occurring at the end of the first line, and the former
at the end of the first division of the song), (9) Bahulatva
(strength or frequency of a note), (10) Alpatva (weakness or
infrequency), (11) Shadavatva (having six notes), (12)
Odavatva (having five notes), (13) Antarmarg (the relations of
the Vadi with the other notes).

Sarangdev proceeds to describe in detail (i) the Shuddh
Jatis, (2) the Vikrit Jatis, (3) the Grama-ragas, which are
generic Ragas themselves derived from the Jatis.

So far two clear indications have been met with of the
practice of varying the pitch of the different scales, the first in
the commentary on v. 35 of Ch. 28 of Bharata's Natya Shastra,
and the second in Sarangdev's description of Grama-Sadharan.
The first reduces the Nama-svara of Panchami to ri •[•, and the
second that of Shadji to the same note. It is not unreason-
able to assume that in the dhaivat system of tuning, the drone
consisted of ri + above dha + (the latter being clearly, from
Captain Day's account, the bass note), and that the Nama-
Svara of all the Shuddh Jatis was tuned down to either ri •{»
or dha +, so as to avoid a readjustment of the chanterelle
strings every time the Jati was changed. The keyboard
string or speaking-string was in its normal state tuned to ri -f.
and the frets arranged accordingly. A readjustment of the
frets (Grama-Sadharan) was necessary in order that Shadji
should be played from the pitch of ri +. Arshabhi required no
change of frets or tuning. Gandhari could be obtained by the
simple device of tuning the keyboard wire down by two srutis,

64 INTRODUCTION TO STUDY OF INDIAN MUSIC

leaving the frets unchanged. The intervals would be those of
the Jati, but ga would be at the pitch (ri +) required for the
drone.^ In the case of Madhyama there are two alternatives.
The ga b string may have been brought into requisition for
the drone, giving ga fa above dha +. In this case tuning
down the keyboard wire by a semitone would bring Madhyama
to the required pitch. The second alternative would be to
play Madhyama without change over a ri •{. drone ; this
was possible, as ri +, dha + and ma ^ harmonize together.
Panchami required the use of antara ga, as explained in the
notes below Bharata. Dhaivati, like Arshabhi, required no
change ; Naishadi, on a keyboard wire flattened by a semitone,
would take the dha + drone. It is possible that Dhaivati
and Naishadi were played from the ri + and ga ^ frets
respectively (the latter with a flattened wire), and brought into
line with the other Jatis as regards pitch. The only difficulty
in the way of this supposition is that it would require a
lowering of the dha fret in order to give the intervals of the
two Jatis. This flattening of dha is mentioned by Bharata,
but not by Sarangdev. One would expect to find a reference
to it in the Ratnakar if the method were in vogue. On the
whole there is more reason to conclude that Naishadi and
Dhaivati took the pitch of dha +. Although, curiously enough,
Sarangdev does not enter into the question of the manner in
which the drone strings were adjusted to the Jatis, or the
Jatis to the drone, there can be no doubt that the method
employed was something like the one outlined here. That
the Jatis were adjusted to the drone, and not the reverse,
is proved by the existence of Grama-Sadharan ; no other
explanation is conceivable. The manner of adjustment must
have been something like that now suggested, first because it
is in conformity with the text of Ratnakar and requires no
new " vikrit svaras," and secondly because similar devices are
in use at the present day.

The Jatis.

Material particulars only are here transcribed. Those
omitted have reference to the rhythm or the time and mode

' This kind of practice survives to the present day (see p. 29).

INTERPRETATION OF ANCIENT TEXT-BOOKS 65

of performance. In small print will be found a short dis-
cussion of derivative Grama-ragas or Ragas. Then will
follow in each case three scales, the first representing the
scale of the Jati with shuddh notes (the Murchhana being
indicated by a separate note), the second representing the
scale with the new tuning mentioned in Prakaran 4, verses
14 and 15, the drone of Shadj tuning being indicated by
double stems to the notes, the third showing what notes were
actually produced by the various devices employed under the
old tuning to adjust the Jitis to the drone of dha + or dha +
combined with ri + (or perhaps, in the case of Madhyama,
dha -f and ga).

Shadji. — Vadi may be sa, ga, ma, pa, dha. Nyas s^ ;
Apanyas ga and pa ; Murchhana dha of Shadj Grama ; ga
frequent. When Sampurna, some make ni kakali. When
Shadava, omit ni. Sa, ga and sa, dha often occur in pairs.
When ga is Vadi, ni will not be omitted. Example : Varati-
Raga.

Note. — R%a-vibodh, Svaramela Kalanidhi, Sangit Sarimrit, all give
varieties of the Hindustani Todi scale for Varati or Varali. The modern
Varadi of Hindustan is sung to the Purvi scale. The present writer can-
not explain this discrepancy ; he can only point out that Southern music
has always, so far as is known, been steeped in chromatic influences,
from which Bharata and Ratnakar were comparatively free. The modern
Varadi may have come to Hindustan from the South.

Old Style. New Style (the same).

Grdma Sadharan.

^-^'^ t^-'-^^-=?p:^

^r^ V- y ^rM

Note. — None of the Grima-rdgas are derived direct from this
Jati.

Arshabhi. — Vadi may be ni, ri, dha. Nyas ri ; Apanyas,
the Vidis ; Murchhana pa, of the Shadj Grama. Pa in-
frequent ; Shadava omits sa ; Odava omits s^, pa. Ga and
ni will be in close relation with the other svaras (this

F

66 INTRODUCTION TO STUDY OF INDIAN MUSIC

simply means that they will be frequent). Examples : Deshi,
Madhukari.

Note. — Raga-vibodh puts Deshi in our Purvi scale. Svaramela
Kalanidhi does the same with Ardra-Deshi. Saramrit puts Shuddh-Deshi
in the Kafi scale, and Ardra-Deshi in the Bhairava scale. In Hindustan,
Deshi is sung either in the Kafi or Asavari scale. Madhukari is an un-
common Raga ; it is mentioned by one Telugu writer.

Old Style.

New Style.

^^^^^^^^^i

^^ig^^

f

1r^

4^^=-^

Note. — Taking Prakaran IV., verses 14 and 15 literally, the scale of
the " Madhya Saptak " would be an octave lower,

Grdma-Rdgas. — The Grama-raga Pancham-shadava (No. 22 of
Sarangdev's list) is derived from Dhaivati and Arshabhi. It appears to
consist of the shuddh notes with a ri drone. Derived from it is the Raga
Gujri which Raga-vibodh, Chaturdandi, Svaramela, Sarimrit put in our
Bhairava scale. The modern Hindustani Gujri is in Todi. Sarangdev
says that Pancham-shadava takes ni kakaU. That may only be as a
passing note. From the example given, it is probable that ni is shuddh.
The example would be inharmonious if sung in Bhairava, as the pro-
gression ma ni and ni ma is frequent and prominent. Moreover, pa and
sa are omitted in the example given.

Grama-raga Revagupta (No. 26) is also derived from this Jati (com-
bined with Madhyama). Deshi (discussed above) is the descendant of
Revagupta. The scale which best suits the example given is the
Madhyam Grama with dha komal (it will be remembered that Bharata
mentions this variety), and ri as drone. In fact, no other scale appears
to fit in with this example. It is curious that if one starts Arshabhi from
the Murchhana pa, one arrives at this very scale. Starting from the
drone, the scale is as follows : —

m

^

-^.

s^=w=^

=^P=^=

or, new style (putting pa I? instead of the current but irrational ma +) —

5^E^^g3ppE^^

A Raga resembling Bhairava may quite well have originated in this
Grama-Raga.

INTERPRETATION OF ANCIENT TEXT-BOOKS 67

Gandhari. — Any note but ri or dha may be Vadi ; Nyas
ga ; Apanyas sa and pa ; Murchhana dha of the Madhyam
Grama. When Shadava omit ri, when Odava, ri and dha.
When pa is V^di, it is not Shadava ; and when ni, sa, ma, pa
are Vadi, it will not be Odava. The Nyas will pair with one
of the five Vadis. When Sampurna, ri and dha will come
together. Examples : Gandharapancham, Deshi, Velavali.

Note. — Madhyam-Grama, No. 17 of the Grama-ragas, is compounded
of this Jati with Madhyama and Panchami. It takes the kakali ni.
(That must mean occasionally.) Madhyamadi is derived from it. That
Raga is even now sung in the Kafi scale, and all the pandits agree in
putting Madhyam-Grama under Kafi. The shuddh scale of Ratnakar
also suits the example given. Thus there can hardly be any doubt
that the notes of this Grama-raga are the shuddh svaras of the Madhyam
Grama.

From Gandhara-pancham is derived Deshakhya, which the Southern
pandits treat as chromatic. This Grama-raga begins with sa, " takes the
kakali," and the scales of Yaman or Bihag suit it admirably.

Velavali : Chaturdandiprakasha, Saramrit, Svaramela put this Raga
under Kafi, and Raga-vibodh under our Bilaval (the common major
scale).

Old Style. ^ New Style.

^=-^

Actual tuning.

The second scale transposed to Sa would be Yaman ;
taken as a Murchhana from sa, it would be Bilaval. The
third is arrived at by tuning the speaking string one semitone
lower, so that the ga fret gives ri. The 3rd, 4th, and 7th notes
are respectively srutis 14, 18, 5 of those in the list on p. 'j'j.

Madhyama. — Vadi may be any note except ga and ni.
Nyas ma ; the Vadis will be Apanyas ; Murchhana ri of
Madhyam-Grama ; sa and ma frequent ; ga less frequent.
When Shadava, ma is omitted ; when Odava, ni and ga.
Examples : Shuddh-shadava, Deshi and Andholi.

Note. — Grama-ragas 17 (Madhyama -grama, mentioned under Jdti
Gandhari), 19 (Shuddh-shadava), 21 (Bhinn-pancham), 26 (Revagupta

68 INTRODUCTION TO STUDY OF INDIAN MUSIC

discussed under Arshabhi), 28 (Kukubh), are derived from this Jati. No.
19 is said to be a " vikrit " form of Madhyama. Ma is Vadi, Nyas, Graha,
Murchhana. It takes antara and kakali. From it are supposed to have
originated Todi and Bangal. The former is our Bahiravi ; the Southern
pandits mention a Shuddh-Bangal with the Kafi scale, Chaturdandi and
Saramrit notice a Karnat-Bangal in the Bhairava scale, while Raga-vibodh
and Svaramela assign Bhairava scale to Bangal, and a chromatic scale
called Karnat to Shuddh-Bangal. From the example given, there is no
reason whatever to assume the scale to be other than that described in un-
mistakable terms by Sarangdev. It is Bilaval starting from madhyam with
the low sixth {i.e. in this case, ri "f.). Bhinn-pancham is the parent of
Varati, which is discussed under Shadji. The Sarigama given is
clearly in the shuddh Madhyam GrS.ma scale, and the drone must be
Madhyam.

In Kukubh, dha is Vadi and everything else, except Nyas, which
position is assigned to pa. Descended from it is Asavari, which Raga-
vibodh puts in our Bhairava scale. Bahiravi with a shadj drone suits
best the Sarigama given. As the Grama-raga is partly produced from
Dhaivati, it may have taken the form of Bahiravi.

Deshi has been described (under Arshabhi). Andholi is Khamaj
(Saramrit), or Kafi (Chaturdandi and Svaramela).

r

:t==±

The Jati may also have been performed with the shuddh
notes over the dha •{. and ri •}» drone. It takes chyut pa even
with the new tuning. The time when chyut pa was abandoned
must coincide with the time when the practice of beginning
Madhyam Grama scales from ma instead of sa fell into disuse.

Panchami. — Vadis ri and pa ; Nyas pa ; Apanyas ri, pa,
ni ; Murchhana ri of Madhyam Grama ; sa, ga, ma in-
frequent. When it is Sampurna, the leap from ga to ni is
used. When Sh^dava, omit ga ; when Odava, ga, ni. When
ri is Vadi, the Odava form is not used. Ri and ma will come
together. Examples : Shuddh-pancham, Deshi, Andholi.

Note. — Grama-ragas 17, 21, 28 described above, are partly derived
from Panchami. Shuddh-pancham is No. 30. It is said to be composed
in part of Madhyama and in part of Panchami. Pa is Vadi, Graha and
Nyis, and also the Murchhana. Shuddh-pancham is the progenitor of

INTERPRETATION OF ANCIENT TEXT-BOOKS 69

Dakshinatya, Andholi, Malhari, and Malhar ; also of Karnat-Gaud,
Deshval-Gaud, Turushk-Gaud, Dravid-Gaud. Sarangdev, in his chapter
on Instrumentation, states that Deshval-Gaud is also called Kedar-Gaud,
and Turushk-Gaud Malava-Gaud. Svaramala gives for the former a
Bilaval scale without ga and with two madhyams. Chaturdandi and
Saramrit introduce a komal nishad. The latter (Malava-Gaud) is by
common consent the Hindustani Bhairava. The Sarigama given for No.
30 appears to be in Shadj tuning, and the scale to be like that of
Deshkar, but Sampurna (complete with seven notes). Deshi and Andholi
have been discussed.

|=g^^S|gfe

fe^g^l^,

:r;-^p-'r-r~r=t:

I

w

^^Em^^:

^^E^EEE3

4/s__

The third of these scales is the Deshval Gaud or Kedar
Gaud of Chaturdandi. With ni kakali, it is Shuddh-
pancham. These are interesting circumstances, as they
support the writer's arguments.

Dhaivati. — Vidi ri, dha ; Nyas dha ; Apanyas ri, ma,
dha ; Murchhana ri of Shadj Gr^ma ; in ascending the scale,
sa and pa are omitted ; when Shadava, pa is dropped ; when
Odava, sa and pa. Examples : Shuddh-Kaishik, Deshi.

Note. — 22 and 28 described respectively under Arshabhi and
Madhyama are partly derived from this Jati. So is 24, Takka by
name. Takka is described as follows : Sa is Vadi, Graha, Nyas and
Murchhana ; takes antara and kakali ; pa infrequent. It is the parent
scale of Gaud or Gaul and Kolahal. All the pandits agree in putting
Gaud under the Hindustani Bhairava (their Malava-Gaud) scale. The
Sarigama given is in shadj tuning ; ri is omitted ; it would be most unin-
teresting and could not by any stretch of the imagination be referred as
it is to Dhaivati unless dha is taken to be komal. If that svara is komal,
the scale becomes that of Bhairava.

Shuddh-Kaishik above appears to be an error for Takka- Kaishik, No.
31. The latter is produced from the Dhaivati and Madhyama Jatis ;
Murchhana, Vadi, Graha, Nyas all dhaivat ; takes antara and kakali.
The example given is probably in dhaivat tuning, and is therefore
as described. Transposed to shadj it would be the Hindustani
Asavari.

70 INTRODUCTION TO STUDY OF INDIAN MUSIC

i

^^fe=^^

^^=^^

^BEfe^^^

Naishadi. — Vadi ni, ri, ga ; Nyas ni ; Apanyas the
Vadis ; Murchhana ga of the Shadj Grama ; ni, ri, ga fre-
quent ; when Shadava, omits pa ; when Odava, pa and sa.
Examples : Shuddh-Sadharit, Deshi, Velavali.

Note. — Shuddh-Sadharit is described as follows : Sa is Vadi, Graha
and Murchhana ; ma is Nyas ; ni and ga are infrequent. It appears to
have the shuddh svaras of the Kafi scale, with shadj as drone.

Old Style.

New Style.

*> Murchhana p-a.

=.-^=^=^-q

t:

Murchhana ga.
Actual tuning.

--?^-

d=^^-p=e

:q:

:^=i^

-4=^-^-

:^=g^-i^-

The speaking string is tuned down one
semitone, so that the ni fret gives dha. The
6th note is Sruti 1 5 of the list on p. yj.

For convenience of reference the suggested scales of the
Grama-ragas mentioned in the above discussion are here
given —

(i) Shuddh-Sadharit. (17) Madhyam-Grama.

-4- ,^.,^^-^4^^^^

^g^^^f^^^g^i^^g^g^

(19) Shuddh-Shadava.

.1^r^

I—
(21) Bhinn- Pancham,

^H \-^'^^=^ %Ji'^ rJ^(^^^\ I _J_]

(22) Pancham-Shudava.

(sa and pa are omitted.) (23) Takka.

-I -T— ^^— ,s4^-r 1- 1-

INTERPRETATION OF ANCIENT TEXT-BOOKS 71

(26) Revagupta. ^, (27) Gandhara-Pancham.

(28) Kukubh.

(30) Shuddh-Pancham.
-I-

i^i^^^H^^^S^

(31) Takka-Kaishik

r

— I-

1^-

There are eleven composite Jatis made up as follows
(Ratnakar on Jatis, verses 9 to 16) : —

Serial
No.

Name.

Component Jatis.

8

Shadj-Kaishiki

Shadji, Gandhari

10

Shadj-Madhyama

Shadji, Madhyama

16

Gandhara- Panch ami

Gandhari, Panchami

17

Andhri

Gandhari, Arshabhi

9

Shadjodichyava

Shadji, Gandhari, Dhaivati

15

Kamarivi

Naishadi, Panchami, Arshabhi

18

Nandayanti

Gandhari, Panchami, Arshabhi

II

Gandharodichyava

Gandhari, Dhaivati, Shadji, Madhyama

H

Madhyamodictiyava

Gandhari, Dhaivati, Panchami, Madhyama

12

Rakta-Gandhari

Gandhari, Naishadi, Panchami, Madhyama

13

Kaishiki

Shadji, Gandhari, Panchami, Madhyama, Naishadi

Nos. 8 and 13 of the list are both called Kaishiki, and
both have Shadj and Gandhari in their composition. The
writer, from the examples given, is of opinion that the scale
used was the following : —

i^^^^^^^"¥^i^

As Played.

It is an interesting fact that the scale which suggests itself
in the case of Takka-kaishik (Grama-ra.ga 31) has similar
intervals. Kaishiki means " of a hair's breadth," and seems to
have been applied to a scale differing from the shuddh scales
by a hair's breadth. Here ri, ga, dha, and ni are all one sruti

72 INTRODUCTION TO STUDY OF INDIAN MUSIC

only higher than those notes in Shadji. Shadjodichyava
appears to be in the Bilaval scale ; the others are very Hkely
in the shuddh Gramas. The commentator, Kallinath by
name, puts all the Jatis in the shuddh Gramas ; there is
nothing, however, to show that he applied his mind to the
difficulties of the subject.

The thirty Grima-ragas of Ratnakar appear to be
generic Ragas, from which the songs and Ragas in use were
derived. No object would be gained at this stage in hazard-
ing conjectures as to the constitution of those which have not
already been discussed.

Recapitulation.

The orthodox system of tuning in ancient times was in
dhaivat madhya. This was the original dhaivat. It is that
dhaivat which is in tune with shadj, that is, at an interval of
six srutis from it. Shadj was so-called {i.e. born of the sixth)
from this circumstance. The drone strings were dha -f, ri +
and ga ; the speaking-string, upon which the air was played,
was ri +• Just as the standard scales of modern Hindustan
may be taken to be Bihag with the high dhaivat, and Bilaval
with the low or madhya dhaivat, so the standard scales of
Bharata and Ratnakar were from ni ^ to ni "fej with pa shuddh
(in which antara ga corresponded to the tivra madhyam of
scales of the Kalyan type), and from ni t? to ni t? with pa
chyut. The modern " dha " was represented by pa. This is
the secret of the much-vexed question of the Gramas. The
first scale comprised the notes and intervals of the Shadj
Grama, and the second those of the Madhyam Grama. Just
as in the Church Modes of Europe and the Modes of Ancient
Greece, new and interesting combinations were obtained by
varying the starting-point of the scale, so in India, first the
Murchhanas, and then the Jatis came into existence. It was
the existence of the drone in Indian music which led to the
evolution of the Jatis. Once they were evolved, the word
Murchhana came to mean nothing more than the lowest note
permissible in any Jati or Raga. The compass of a com-
position was determined by its Murchhana ; its Jdti was
determined by the drone and by various factors somewhat akin

INTERPRETATION OF ANCIENT TEXT-BOOKS 73

to the final and dominant of the Church Modes. There were
seven shuddh Jatis, one for each shuddh note. It was found
that the Shadj Grama was best suited to four of them, and the
Madhyam Grama to the other three. Each required a drone
corresponding to its chief note (Nama-svara). In practice it
was found more convenient to vary the pitch of the scale than
that of the drone ; various devices were invented for bringing
the Nama-svara to the pitch of one of the drone strings.
The reader's attention has already been drawn to the fact
that present practice tunes the ma string ^ a semitone higher,
so that the ni fret in the Yaman scale gives sa, in order to
play Bahiravi without difficulty. Similarly, Kafi is got from
the Bihag scale by tuning down a whole tone, and taking ri
to be sa. In Sarangdev's time, the speaking-wire (ri +) was
tuned down a semitone for Jatis Gandhari, Madhyama,
Naishadi ; Panchami was obtained with the aid of antara ga ;
Arshabhi and Dhaivati required no special device ; and
Shadji necessitated the use of Grama-Sadharan, for, other-
wise, the wire would have had to be tuned either a whole tone
higher or a minor-third lower.^ Tightening a string by a
whole tone is quite a different matter from lowering it by that
interval. Possibly it was this Grama-Sadharan (in which the
sa, ma and pa frets were shifted down one sruti, and the ga
and ni frets shifted up), which led to the general adoption of
shadj tuning. It is a curious fact that Minappa, a Madras
writer, puts shadj on the sruti reserved for chyut shadj.
Among the vikrit or irregular Jatis, the most interesting is
Kaishiki. This appears to have differed from Shadji in this,
that, in the process of Sadharan, sa and ma were achyut (or
unchanged), while ga moved up to antara and ni to kakali.
The scale was called Kaishiki because it differed by a hair's
breadth from Shadji. The musicians of the present day in
Western India call a scale somewhat like the first (but with
komal dha) Asavari (some say Jivanpuri) ; Shadji corresponds
to Kafi. The Jatis in their turn led to the Grama-ragas, and
the Grama-ragas to the Ragas in common use. Even in

' That is, the stiing upon which the melody is played.

" Sarangdev and Bharata are completely silent on the question of the drone.
They take it for granted that the reader knows. These statements are therefore
inferences ; but none the less valid.

74 INTRODUCTION TO STUDY OF INDIAN MUSIC

Sarangdev's time, the first two seem to have existed in
theory only. Ragas were divided into two classes, marga
and deshi, which may be rendered classical and modern. In
all probability, these names had some reference to the two
schools of tuning. At the present day (although dhaivat
tuning is mentioned by Captain Day) shadj tuning appears to
have driven its rival off the field ; this may then perhaps be
regarded as a victory for the deshi musician. The Sangit
Ratnakar was written in the thirteenth century. From
internal evidence one may conclude that the contest between
the two systems was then in progress. As regards the
Gandhara Grama, it may be mentioned that it was admittedly
obsolete at the time of Ratnakar ; the description therein
given of the scale is intelligible, and shows that it was one
difficult of execution but capable of considerable development.
One of its Jatis would be represented nowadays by Bihag
with the note which the present writer calls kaishik ma in
place of komal ma ; another by Kafi with sadharan ga in
place of komal ga.

The consequences of the change of tuning from dhaivat
to shadj may be here briefly outlined. Whereas, in the
orthodox system, roughly speaking, all compositions were
performed in the key of ri +, with a drone consisting of
dha "f (below) and ri •{> (above), in the modern system the
universal key is shadj, and the drone pa or ma, according
to circumstances, above sa. There appears to have been
an intermediate stage in which the key for compositions in
the Madhyam Grama was ma % with the drone ma above
si. The effect while that stage lasted was, as regards the
Madhyam Grama, precisely the same as in dhaivat tuning.^
With the disappearance of that stage went the note known
as chyut pa, and it is a fundamental characteristic of the
modern system that sa and pa are fixed and can never become
vikrit or chromatically altered. They correspond in that
respect to the ancient ri + and dha +, which never really
became vikrit as the examples of actual tuning for the various
Jatis and Gramas given in Appendix E will show. In a word,
it may be said that the chief change to be associated with the
substitution of tuning in shadj for tuning in dhaivat was that

' Except, apparently, compositions in Jati Madhyama.

INTERPRETATION OF ANCIENT TEXT-BOOKS 75

sa and pa took the place of dha 4, and ri +, and the chord
of the drone was turned upside down. This must have led
to changes of intonation and possibly of tonality. Signs of
such changes are not wanting ; as an example may be quoted
the modern Bahiravi scale with the harmonic fourth which
has taken the place of the Jati Arshabhi.

Arshabhi. (Bahiravi.)

■ I, •

^^^i^±^zz::p^.

Drone. Drone.

The lower note of the drone is more powerful and more
reinforced in the tambura than the upper note of the drone.
It is difficult to sing the high fourth in the scale of Bahiravi.
The change of tuning led to a deplorable confusion of nomen-
clature ; this and the organic changes just described may
account for the difficulties encountered in any attempt to
trace the modern Ragas back to the Grama-ragas of the
Sangit Ratnakar.

CHAPTER V

CONCLUDING OBSERVATIONS

In this chapter will be included (i) speculations concerning
the origin of the Indian scales ; (2) some remarks regarding
Sarangdev's immediate successors ; (3) a suggestion as to the
most convenient method of classifying the Indian Ragas ; (4)
a discussion of the manner in which Indian airs may be per-
formed in correct intonation upon instruments of European
fashion.

As a result of the above discussion of the actual orthodox
I ferences r methods of tuning for the performance of the

garding ancient Jatis, it now becomes possible to reconstruct

belirrwn fr?m° ^^^ ^^^^ °^ ^^"^^^ employed by Sarangdev. It
the Sanskrit is abundantly proved that he did not clearly
Class-names. distinguish between all of them. Appendix E
shows at a glance how Grama-Sadharan, the tuning down of
Gandhari Jati, etc., by two srutis, and the shuddh notes of
the three Gramas and of the scale known as Kaishiki, together
furnished twenty-two srutis according to Sarangdev's theory
of the equality of the srutis. When one goes further and
ascertains from the laws of harmony the exact positions of
these srutis, one finds that in three cases, two srutis which are
near to one another but not identical are treated as one.

In the following list, the exact position of the ancient srutis
is given, together with the Sanskrit individual names and the
class- names mentioned in the Sangit Ratnakar (Prakaran 3,
verses 24-38). As remarked in an earlier chapter, the term
" sruti " was primarily applied to an interval.

CONCLUDING OBSERVATIONS

n

The Srutis of the Sangit Ratnakar.

Class-names.

Comparative
vibration

Serial

Individual

Descriplion.

number.

names.

Original
svaras.

Later
additions.

inimbers, sa
being 240.

O

Kshobhini

madhya

ni \ (shuddh)

2i3i

^

\xi\^ (kaishik)

Ini •f' (part of Grama-Sad-

216

I

Tivra

dipta

222f

haran) '

2

Kumudvati

ayata

ni h (kakali)
sa.^ (chyut sa) ^

225

3

Manda

mridu

23757

4

Chhandovati

madhya

sa (shuddh)

240

5

Dayavati

karuna

sa J (major-third above
dha +) '

250

6

Ranjani

madhya

rit?

256

7

Raktika

mridu

ri + (shuddh)

266g

8

Raudri

dipta

ga \ (shuddh)

270

9

Krodha

ayata

284-5-

lO

Vajrika

dipta

|ga|? (sadharan)

\ga •], (Grama-Sadharan) '

288
2965T

II

Prasarini

ayata

ga n (antara)

300

12

Priti

mridu

ma *|« (chyut ma) ^

316^1

13

Marjani

madhya

ma \ (shuddh)

320

H

Kshiti

mridu

ma 4. (major-third from
ri+)'

333i

IS

Rakta

madhya

ma

, (Gandhara Grama) '
. (Madhyam Grama) '

337*

i6

Sandipani

ayata

/pa*
Ipa.

3455-
3555

17

Alapini

karuna

pa (shuddh)

360

i8

Madanti

karuna

pa J (major-third from

gat;)^

375

19

Rohini

ayata

dhat^

384

20

Ramya

madhya

dha 4, (shuddh)

400

21

Ugra

dipta

dhajj

405

22

Kshobhini

madhya

ni \ (shuddh)

426S

* These notes are no longer in use.

In the columns headed Class-names are collected the
names regarding which the remark is made : " It is said that
this classification is based upon the effects which the srutis
are supposed to produce." They may be translated : " mridu,"
soft ; " madhya," intermediate, that is to say, intermediate
between soft and pathetic ; " karuna," pathetic ; " dipta,"
brilliant ; " ayata," stretched. The notes have been divided
into original svaras and later additions on the following
principles: (i) It is assumed that the term "ayata" was
primarily applied to a note for which no fret was in use, and

7^ INTRODUCTION TO STUDY OF INDIAN MUSIC

which was therefore obtained in the manner familiar to anyone
acquainted with Indian methods — that is, stretching the wire to
one side. The whole of the ayata class are therefore rejected.
(2) The notes of the Gandhara Grama, and of the Kaishiki
scale, and those used in Grama-Sadharan, are likewise regarded
as later accretions to the Hindu musical system. (3) The
" madhya " notes go in pairs, thus —

dha^

ma|j
ma'R

It is possible that shadj tuning (which, although not con-
sidered orthodox by Sarangdev, may have been invented
centuries before his time), may have introduced the ri b and
ma t], and that they may have been named " madhya " on the
analogy of ni "fcj. At the present day, the first fret on the
open wires sa and ma gives ri b and ma b, respectively. The

intervals — ^ and — —^ are not the same : the same fret with
sa ma "jj

the help of a slant or a curve is made to do service for both,

the sa and ma wires lying side by side. Of the six notes

given in the diagram, it seems reasonable to retain ni "J;j,

dha ^, sa, and ma % The choice is based on the assumption

that dhaivat tuning was the original system.

The notes which the above considerations lead one to

regard as the earliest foundation of the Hindu musical system

fall naturally into two scales : —

i

i^g^^

f

^^^

l^__CJ.

:t

\4

t«i

The reader will notice that the Class-names, so far as
these scales are concerned, may be truly said to denote " the
effect which the srutis are supposed to produce." The pro-
priety of the classification would be made still more apparent
by the addition of a drone in dha + . These scales apparently
take one back to a very remote period in the history of the
Aryan race. The Class-names may even be earlier than the

CONCLUDING OBSERVATIONS 79

solfa names, shadj, rishabh, etc. ; they are at any rate older than
the name Gandhara. As regards the derivation of the solfa
names : — (i) Nishad is generally connected with the Sanskrit
root which means " to sit" ; it may be looked upon as a kind
of final or tonic. (2) Shadj is " born of the sixth," that is,
developed at an interval of six srutis from dhaivat.
(3) Madhyam is the "middle" note, as near as may be, in
the second of the two scales given ; similarly ma 4.. is the
middle note of the first scale. (4) Pancham is the "fifth"
note of the second scale. (5) The derivation of dhaivat,
rishabh, gandhara is more open to controversy. The first
scale is the same as the archaic scale of Olympus, well-known
to Greek scholars ; the second is one of which traces are to
be found in the remains of all known bygone civilizations (see
Carl Engel's " Music of the most Ancient Nations "). The
first appears to contain the germ of the Karnatic system,
which Madras scholars attribute to Narada, while the second
is undoubtedly the forerunner of the Shadj Grama and the
musical system which forms the basis of the treatises of
Bharata and Sarangdev.

These conclusions, it may be mentioned parenthetically,
run counter to the hypothesis of the develop- a rival

ment of scales from " clusters of quarter-tones." hypothesis.
Regarding them as rival theories, one's judgment between
them should be based on evidence rather than authority,
because no theorist, however remote the age in which he
lived, was privileged to witness the actual evolution of any
musical scale. Writers who deal with the subject of modes
generally regard them as built up individually from tetra-
chords, and it is a favourite theory that the earliest tetrachord
was evolved from a cluster of quarter-tones or semitones,
followed by a leap to the note a fourth above the starting
point. In course of time, the theory runs, the cluster became
reduced to two notes a semitone apart, leaving a scale of the
same form as that of Olympus. The wide intervals between
the second and third and fifth and sixth degrees of the scales
were then bridged over in various ways, leaving a scale con-
sisting of two similar tetrachords. To the historian of Karnatic
music may be left the solution of the question as to what sup-
port the theory derives from what is known of the chromatic

8o INTRODUCTION TO STUDY OF INDIAN MUSIC

scales of India. As regards the Grama scales the writer is of
opinion that the evidence so far is entirely adverse. The
subject, however, can hardly be said to be ripe for discussion
until the question of tonality in Indian music has been
thoroughly investigated. As stated in an earlier chapter,
correct intonation must form the groundwork of such an
inquiry, and it is with intonation alone that this volume pro-
fesses to deal. This much may here be said respecting
tonality, that the degree of tonality, which is indispensable to
the tempered scales of Europe, is not to be looked for in dia-
lectical scales built up from natural intervals. The tonality
of an Indian Raga may be said to be determined by the
factors enumerated by Bharata and Sarangdev, to wit, the
Vadi and Samvadi, the Ansha Svara (prevailing note),
the initial, medial and final (Graha Svara, Apanyas and
Nyas).

The consequences of the triumph of shadj tuning over its

_ , rival have been discussed. There is no hint

Errors of

Sarangdev's in Kalinath s commentary on the Ratnakar,
successors. published two centuries later, regarding the

progress made by shadj tuning. No addition was made to
the literature of Hindustani music until Parijat's time (the
seventeenth century). Meanwhile, two treatises had appeared
in the South, the Svaramela Kalanidhi of Rama Amatya, of
about 1550, and the Raga Vibodh of Somnath, of 1609.
These two writers made collections of the Ragas of Southern
India. Their tuning was in shadj. They knew of the Rat-
nakar, and looked upon it as a work of great authority, but
they appear to have been entirely ignorant of the fact that it
was based upon a different system of tuning from their own.
They accepted Sarangdev's theories, and assumed without
hesitation that his shuddh notes were the same as their own.
What their own were is clear, first from the Raga Mukhari,
which takes the shuddh notes, and secondly from modern
treatises, such as that of Mr. A. M. C. Mudliar. In tempered
notation he represents the " shuddh " scale as follows : —

~f' I I l~i — I — 7".~^^^l^^^^=^^^=l

CONCLUDING OBSERVATIONS

8i

He states that the double-flat notes are "as near as
possible" D and A natural. The scale may possibly be
either of the following ; it needs verifying : —

^E^^S^

i^:z^4?25^^

=T

t^BE^J?^?EE=E

te-?2-

— 1==3

sa ri ga ma pa dha ni sa. sa ri ga ma pa dha ri si.

How this collection of notes came to be regarded as
" shuddh " is an interesting problem. Counting up the
srutis from " ni," the author of Raga Vibodh said his sa was
on the fourth, the ri on the seventh, the ga on the ninth, and
so on, copying from Ratnakar, and never imagining that they
had another scale in Hindustan. It is obvious that a de-
plorable confusion of names was bound to ensue, especially
when, in due course, an incursion was made into the region of
" vikrits." Rima Amatya, copying the chyut svaras of Grama
Sidhiran, spoke of " chyut madhyam ga," " chyut pancham
ma," and " chyut shadj ni." Somnath contracted these names
to mridu ma, mridu pa, mridu sa. The actual notes called by
these names appear to have been the antara ga of Ratnakar
(ga h), ma tivra, and the kakali ni (ni b). Somnath describes
the two following scales, regarding the identity of which there
is no doubt whatever, in these terms : —

I. Raga Todi (the Hindustani Bahiravi).

Somnath's notes.

ri

shuddh.

sadharan.

ma, pa
shuddh.

dha
shuddh.

ni
kaishik.

Actual notes with )
accidentals /

rif

gat? '

ma \ pa

dha^

nil?

2. Sri Raga (

the Hindustani Kafi).

Somnath's notes.

ri
tivra.

sadharan.

ma, pa
shuddh.

dha
tivra.

ni

kaishik.

Actual notes with \
accidentals J

" +

ga'k

ma \, pa

dha 4,

ni'k

No further proof of the disorder introduced into Indian

G

82 INTRODUCTION TO STUDY OF INDIAN MUSIC

musical terminology by these authors need be adduced. This

state of affairs has unhappily prevailed up to the present

day. When Parijat was written, shadj tuning prevailed.

The author showed more perspicacity than the Karnatic

writers, appropriating the shuddh notes of the Ratnakar to

the Kafi scale, and designating the Madras " shuddh " notes

" purva," but no scientific terminology could be expected of

any writer who adhered to Sarangdev's system. Nowadays,

among the practical musicians of Western India, the Sangit

Ratnakar is looked upon as belonging to a bygone age

although no one is able to say what it is which makes its

theories inapplicable to modern practice. Professional

musicians have constructed their own systems ; needless to

say, they differ widely one from another.

A considerable amount of space in the earlier chapters of

_. ,.,., e this book has been devoted to the Indian
The utility of i ■ i • j .

the Indian harmonmm. The reader may be mclmed to

harmonium. ^^^ what useful purpose such an instrument

is likely to serve. The tempered harmonium is at present
enjoying a great and increasing popularity. It is employed
either to furnish a pedal sa and pa as drone accompaniment,
or to follow the voice in a varying degree of approximation
in a kind of discordant scramble. For the drone, Moore's
harmonium gives more scope in pitch, and far better results,
as the fifths are in tune. The latter kind of accompaniment
has nothing to recommend it according to Western taste;
whatever virtue it may appear to possess may be obtained
with the addition of accuracy from Moore's harmonium, but
subject to a certain limitation with respect to key. There are
possibilities in the way of harmony, but for the present the
great object which the inventors have in view is to educate
public taste and preserve the Indian scales. For the class-
room the instrument should prove indispensable, and amateurs,
who can afford to purchase an instrument, will find it of great
service in enlarging their acquaintance with the Indian
Ragas. The value of the instrument for scientific research
cannot be overestimated ; without its help it is no exaggera-
tion to say that the tabulation and systematic classification
of the Ragas is likely to remain impossible of achieve-
ment.

CONCLUDING OBSERVATIONS S3

In the method recommended for such an inquiry, as
already stated, the first step would be to classify Suggestions as
the Ragas according to the scales they employ, JJJjn^of'tlS^^'^^
in groups arranged so as to distinguish the Ragas.
important features of Indian intonation. The main groups
will be (i) Grama scales, (2) Chromatic scales, (3) Irregular
scales, (4) Mixed scales. Until a comprehensive list of Ragas
with their scales and distinguishing characteristics is made,
one can only throw out rough suggestions as to the sub-
heads into which each group should be divided. With regard
to the Grama scales, the author is not in favour of the system
adopted by European writers in treating of similar dialectical
scales, nor is the method familiar or likely to be acceptable
to Indian musicians. The system alluded to is that of dis-
secting scales into tetrachords (i.e. groups of four notes) and
naming modes after their Greek prototypes. This method is
not really in accordance with any natural process in the
evolution of scales, and in some cases is calculated to mislead
one as to the origin of Indian scales. Neither does it suffi-
ciently accentuate differences of intonation, which are the life
and soul of music. The method may serve some purpose in
further classification of the Ragas included under each sub-
head. But at present one cannot do better than follow the
principles adopted by Mr. Bhatkhande in his "Hindustani
Sangitachi Paddhati." Mr. Bhatkhande takes the scales of
certain Ragas as typical " thats " or " arrays " of notes, and
allows a certain amount of latitude in respect of chromatic
change among the scales assembled under each "that." His
"thats" are (i) Kalyan, (2) Bilaval, (3) Khamaj, (4) Kafi,
(5) Asavari, (6) Bahiravi, (7) Bhairava, (8) Todi, (9) Purvi,
(10) Marva. Under the first head he includes such scales
as Yaman, and scales which, like Kalyan itself, take two
madhyams, tivra and komal. Under the second he includes
some scales with the high sixth, and some which take two
nishads, komal or kaishik and tivra. With a little more
attention to correct intonation, his classification will be the
most convenient, and is likely to prove the most acceptable
to the Indian musician. The following are suggested as
typical Grama scales : the black notes show what chromatic
changes are permissible. Each type, speaking generally,

84 INTRODUCTION TO STUDY OF INDIAN MUSIC

includes scales which (i) in part use the secondary notes in
preference to the corresponding primary ones, or (2) in whole
or in part make use of both primary and secondary as
alternatives, or (3) use the primary notes exclusively. The
primary notes are given first, and the corresponding secondary
notes are placed immediately after them. The word "primary"
is intended to convey the sense of original, and " secondary "
that of altered, or " vikrit," to use the ancient term.^

Shadj Grama.
I. The Kalyan or Bihag type. II. The Kafi type.

iggjgg^ggS^^BdsJd^^^gg

W

III. The Dhani type.

IV. The Khamaj type.

-§^^^^^^S^^^^^

V. The Bilaval type.

Madhyam Grama.

VI. The Bahiravi type.

i^^gggl^i^g^gSl

VII. The Asavari type.

It is necessary to add the following explanations.

Types IV. and V., and types VI. and VII., to a certain
extent, overlap ; it will never be doubtful, however, in practice
to which group any particular scale belongs. The main
difference between these kindred groups is one of tonality.
No scale resembling types VI. or VII., which includes dhab
and ni t) without the occasional use of ni b, should be included
under those types. Such scales should be put in one of the
groups of Chromatic Scales. An interesting scale which
makes use of all the nine notes comprised in Scale VII. is

» In European music, the same sort of distinction is expressed by the words
"diatonic" and "chromatic." These terms, however, are not altogetlier
appropriate here.

CONCLUDING OBSERVATIONS 85

that of Raga Lachari. Some singers restrict the name
" Asavari" to the Riga, which takes the septimal notes ri ^
and dha u, and call the scale comprising the primary notes
of VII. Jivanpuri. Others call the septimal Raga '* Ramkali
Asavari." The latter variety should be classed among the
Irregular Ragas.

The following grouping is suggested for the Chromatic
Scales : —

Chromatic Genus.
VIII. The Bhairava type. IX. The Deshi Todi or Kandratype.

t=±

:t==:t:

The Purvi type. XI. The Multani type.

Irregular scales will include such divergent types as
Yaraan-Kalyan and Marva. Until a list of them is prepared,
it is impossible to put forward any definite plan for their
classification. It may be possible to regard them as off-
.shoots of the typical Grama-scales, or Chromatic Scales.
Raga-scales, which are obviously compounded of two or more
dissimilar types, will come under the head of Mixed Scales.
The number of Ragas possessing mixed scales will be found
to be considerable.

The above suggestions are meant to be of a tentative
nature. A fuller acquaintance with the Indian Ragas than
the author possesses may reveal defects in the grouping
suggested, or lead to the formation of other groups. The
main object of the above classification is to accentuate such
differences as those between dha t] and dha +» ri D and ri •}., ga b
and ga "^, and so on. To exemplify this feature of the proposed
method, the scales of Bhup and Deshkar may be quoted.
They are as follows : —

Bhup. Deshkar.

j^:

^^^E^E^^^Ejg

86 INTRODUCTION TO STUDY OF INDIAN MUSIC

How to adapt
Western keyed
instruments to
Indian music.

They will be included, the former under Group V. and the
latter under Group II.

The last subject which the writer proposes to discuss is
the feasibility of adapting Western keyed
instruments, such as the instruments used in
military bands, for the performance of Indian
airs in correct intonation. The prevalent
impression appears to be that it is impossible to render
more than one or two Ragas in just tuning if one is confined
to twelve notes in the octave. Those who take a fancy to
European instruments argue therefrom that one is driven to
have recourse to the tempered scale. The following typical
scales have been drawn up in order to correct this mis-
apprehension, and to show that the twelve notes of the
European octave may be so tuned as to give a very great
scope for the performance of Indian airs. It must be under-
stood that purely instrumental music is contemplated where
the question of pitch is not an essential consideration.

Tuning.

i

:^g=y

The notation is that followed throughout this volume.
The naturals are in just tuning in key C ; B ^ is a fourth
above F ; G b is a major third beiow B ^, and the remaining
flats are obtained from G b by an ascending series of fifths.
The concords available are —

The following typical scales may be correctly rendered : —

CONCLUDING OBSERVATIONS

I. (Kalyan or Bihag.) II. (Kafi.)

87

T

T^=^

^-f^

gE^a^^^^

:^g^=g:

t=it

1 — r-

1

III. (Dhani.)
^_r 1 1 1_

(Deshkar.)

i

--^=^

:^— ^— P-

-.2-:^-.

l^

r=s2

IV. (Khamaj.)

V. (Bilaval.)

SE^i^^^

I^Z^ill^

:?2:

:^==^:

-■^-=^

(Bhup.)

I
VI. (Bahiravi.)

i

^^=^2=^:

I

f

:^:

:^=2^:

:?2:

3Eg5^3^EEp^

:?^.

VII. (Asavari.)

VIII. (Bhairava.)

IX. (Kandra.)

^-^^

^E§Efe:^_E&^^:

^^-T^-^P

^z^^_^^=^

^^=^5=^

X. (Purvi and Vasant.) ^

^«i^^ggg^^i^j£gig

XI. (Multini.)

Irregular: (Marva.)

~l-

^S=?2=^:

^,^s-t^-

SE^Ee=e:

t:

:^=?^

r

It would not be impossible to obtain a complete band of
instruments tuned in the manner indicated. The drone is
shown in the same way as in previous chapters ; four sets of
instruments of percussion would seem to be required. As
regards pitch, it would probably be found convenient to
substitute E b for F.

Indian music opens up a new world to the student of
harmony. Where every chord is a little out of tune, as in
equal temperament, the harshest discords may be tolerated ;
in just tuning, common chords are so pure that their admixture
with the simplest discords must be managed with care and
restraint.

' Vasant takes both mas ; Purvi takes ma □ only.

^

Q

2;

^s

ti]

w

tiJ

u

K/l

W

3

Kn

CL,

b

>—t

S

O

Q

en

<

>

P-.

Ph

Q

<

<

O)

S

o

H

c;

13

Di

Cfi

^

<i

1— 1

Q

^

t-H

w

K

H

<§^

n c 1

00

=2-«^

2 -a

CO J

OS-

OS.

8^

rT

vo H

ON ^

■<*■ S3

00 es

00^

O rt

~a

.„ 1 1 1

I

1

1=

1 ■' 1

N_C.

00. 2

<D O

'b >

pq

•^ 2 .s

>-. tn o f3
►^ .S

•^ ^

o,

^

S

CI

-o

m

« 43

r£3

*-*

.S.°
^ S
^5

U)

PJ

" o

-s §

V rj

° N

.„00

APPENDIX B

THE INDIAN HARMONIUM

The keyboard of the Indian harmonium is constructed as
follows : —

The white keys consist of sa and pa shuddh and the rest tivra.
The five black notes are : ri komal, ga sadharan, ma kaishik, dha
komal, ni kaishik. The notation sign for all the white notes is t^ and
for all the black notes t?. The remaining srutis are sounded by
pressing upon brass studs. The front studs marked " m " are the
madhya notes ri and dha (+). The stud marked " tj" is for atikomal
madhyam ( [r) in pancham key {i.e. for voices which find pa a more
convenient starting point than sa). The middle row of studs includes
the notes which are one sruti less than the black notes. They are
marked " ak " in the case of the septimal notes atikomal ri and dha,
and '* k " in the case of komal ga, ma, and ni. The principal one,
ma komal, is for convenience placed in advance of the others. The
signs for them are ]^ and '^ respectively. The back row of studs
consist of the tartivra notes ga, ma, ni. They are marked " tt," and
their sign is .+.

The harmonium is tuned from pa, the octave above the middle C
of the great staff. The method of tuning will be clear from the

APPENDIX B

91

following chords which give an harmonious combined sound without
sensible beats. Beats consist of a periodical increase and decrease
in the volume of sound. When notes which should form harmony
are not in tune, beats are heard ; the further the notes are from true
concord, the quicker the beats, until they become so rapid that the
mind, unable to follow them, receives the sensation of roughness or
harshness. When the notes are very near to their correct pitch,
beats may become so slow that what is heard is a roughness recurring
at regular intervals of time. The reader who has access to an Indian
harmonium may study the question of beats by sounding together ri
tivra and ri madhya, first in the lowest octave and then in each
succeeding octave. He will notice that a discordant interval in one
octave gives exactly double the number of beats which the same
interval gives in the octave below. If he plays the nearest equivalents
to the concords here shown on a tempered harmonium, he will find
that, with the exception of the octave, all intervals give beats more
or less rapid.

i

w^

:(ate

S:

a^F^^^

w^

?2:

r-

i

=^a=t

^i

=^

ife^g:

m

-^^^ ^gj"

s=^

The " pitch C " is the C an octave above the middle C. This
note is tuned from a tuning fork to 522 vibrations a second. This is
the London Philharmonic pitch as settled in the year 1896, and for
all practical purposes the French diapason normale at a temperature
of 60° Fahrenheit. Thus^ the sa of the madhya octave (man's voice)
is of 174 vibrations.

Three of the scales given in the text are graphically shown below
as played on the Indian harmonium. To assist the beginner, the
best method of fingering is shown by figures, i stands for the thumb,
5 for the little finger, and the other numbers for the remaining fingers
in order.

92

APPENDIX B

1-3
<
>

►J

5

(jjj uJ fmj ^y

II II

®

(ak

®i©

Pi.
<<

ak ( k

©y©

ak) ( k

©

ak

@M@

The scope afforded by the Indian harmonium for demonstrating
the nature of the harmonic series is clear from the following. The
terms of the series given above each of the generators are numbered.

APPENDIX B

93

i

4568

9 10

¥

r-

:^

' ' 'i^^J%g^

7 5i|,^

:^:

-^-r— h-

i^it;

-s:^-

--^-

2 3

I 2

i

_ o 9 10

s^-=^r'^^

,&P2^

3 4 5

;Ji::i:

ztziz:

^:^

-f^^

9 10

r-

:^-

z£^:

$

45678

-f^-

i^—^

1-^

2 3

APPENDIX C

TRANSPOSITION ON THE INDIAN HARMONIUM

(i) The Yaman Scale.

.^-n

fe^g^

^^3

^E^&=e^_E^Ee

-^r^-

i^^=

^=^-^

^^E^^=m

m^^

$s

:;g:^^=F

^p%7-:

(2) The Bilaval Scale.

--T^^A

a

^-n

12:^

^^fe^-^

*^:

:g-r^-p-

Lzfei:

^^^-^^

-^-^^

4^-

:?2:

^=g^^^

i

:^=f=:

q=^:

ig^J^ig:!^

(3) The Bahiravi Scale.

^g^^gg^Sgigj^^ggggJEJ

— ZZ^^fe^^

^g

?^

-fe^-^^

:^^=?2:

— ^=^z&^

=g=^^^=^

t^i^^^^S^

trt^

::i=:^:

--^--

:t:

^

APPENDIX C

95

(4) The Bhairava Scale.

^^EM^mm^^

^=5g-

^~<z^

^^TT^'^T

S

WSi^

I^-to-

^E^g--^=F£gE:

(5) The Kafi Scale.

^Eg^M^^^Eg^gg

1^=^:

:?2=^:

-^-

±1:

^^^^efe

S^^S^iP^^B

=iS^w=^

1 — I —

^^i^i^g^gi^^^

(6) The Marva Scale.

gjgg^g^^gg

+- \—:=t-^<5> P 1 -4^-

:^=fe:^:

-^m

±1

t=^t=:

1^

5Eg^giagg^gfe=g^^g^gg

(a) This note is approximate only.

* The Raga Yaman-Kalyan makes use of the Septimal 4th (ma [^). In this
key the exact note is given by the F stud marked (xj).

(I)) The progressions bracketed are the septimal ones to which the scale
approximates. See p. 31.

Note. — The resources of the instrument for transposing would
be enormously increased if the key marked XJ were tuned one sruti
higher than shadj, the key which it pierces. It would then be
"kaishik ma" in key pancham instead of " atikomal ma" as at
present. The Raga Yaman-Kalyan which takes two ma's, tivra and

96 APPENDIX C

atikomal, could be rendered in two keys, namely dha [? and ga l7.
At present it can only be correctly rendered in one (dhat? being
taken as shadj). This change in tuning is easily effected. The
instrument would thereby lose something in the way of harmonic
possibilities.

APPENDIX D

SArigaais of Jatis and Grama-Ragas discussed in the Text, These
are taken from Ratnakar. A dot means a note in the low octave,
a dash one in the high octave.

Shadj-Kaishiki (8),

sa sa ma' pa* ga ri ma ga ma ma

ma ma ma ma sa* sa* sa* sa*

dha dha pa pa dha dha ri ri ma

ri ri ni* ni' ni* ni* ni* ni*

dha dha pa dha ni ma ma pa pa

dha dha pa dha ni dha dha pa pa

sa sa sa sa sa sa sa sa

dha dha pa dha dha ni dha dha dha

sa sa sa ri ga sa ri ga dha dha

ma ,dha pa pa dha dha ni ni

ri ri ga sa sa* sa* sa* ga*

dha* ri" sa* ri" sa* ri" ri* sa* sa* sa*

sa sa ri ri sa ri ri sa sa sa

ma ma ma ma ni dha pa dha ma ma

ni ni pa pa ma pa pa ma pa dha n ga

ga ga ga ga ga ga ga ga

Shadjodichyava (9).

sa

sa

sa

sa

ma*

ma-

ga-

ga-

ga

ma

pa

ma

ga

ma

ma

dha

sa
dha

sa
ni

ma

sa

sa

pa •
dha

pa
ni

ni
pa

dha
ma

ga'

sa

sa

sa

sa

sa

sa

ga*

dha

dha

pa

dha

pa

ni

dha

dha

sa
ni

ga-
dha

ga*
pa

ga"

dha

ga*
pa

ga-
dha

sa
dha

sa
dha

sa'

sa'

ma

ga

pa

pa

ni

dha

dha

ni

sa'

sa'

dha

ni

pa

ma

ga-'

sa

sa

sa

sa

sa

sa

ga-

dha

dha

pa

dha

ma'

ma'

ma'

ma'

H

98 APPENDIX D

Kaishiki (13),

pa

dha ni

pa

dha ni

ga

ga

ga

ga

pa

pa

ma

ni dha

ni dha

pa

pa

pa

dha

ni

sa'

sa

ri

ri

ri

ri

sa

sa

sa

ri

ga

ma

ma

ma

ma*

dha-

ni*

dha*

ma"

dha-

ma*

pa*

ga

ri

sa

dha ni

ri

ri

ri

ri

ga

ri

sa

sa

dha

dha

ma

ma

ga

ga

ga

ma

ma

ni, dha, ni

ni

ni

ga

ga

ni

ni

sa

ga

ga

ga

ga'

ga

ni'

ni'

ni' dha'

pa

pa'

pa,

ma'

pa'

ma'

Pi^'

pa'

pa'

ma'

ma

sa'

ma

ga'

ni dha ni

ni

ni'

ma'

ga'

Grama-Raga Shuddh-Sadharit (i).

sa
dha

sa
dha

dha
ni

ni
ni

pa
ri'

pa
ri*

ri

dha

pa
ma

pa
dha

pa
sa

dha
sa

ni

sa

dha

dha

sa

dha

sa

ri

ri

dha

ga
ma

pa
dha

pa
ma

pa

sa

pa
sa

pa

dha

ni dha

pa

ma

pa

Shuddh-Shadava (ig).

Takka (23).

ni

dha

Pf

sa

ni

sa

si

ga

ri

ga

ri

sa

ma*

ma*

ga-

pa
sa

ma

si

ga
ri

ga

ri

ga

pa pa

pa pa

pa ma

sa sa

ga sa

pa pa

ma*

ma*

dha*

dha*

sa

dha

ni

pa

dha

ni*

ma*

ma*

ma'

ri

ma*

ri

dha*

ni*

sa*

sa*

ga

riga

dha

dha

sa

dha

sa

ma ga

ma*

ma*

ma'

ma'

ma ga

ri

S?

ma

ma

ma

pa ma

ga

ri

ga

sa*

sa*

ma*

ma*

ma*

ma*

ni

dha*

ni

dha*

sa*

sa*

sa*

sa

ga

ri

ri

ga

ma*

ma'

ma'

ma*

sa

sa

dha

dha

ma

ma

ma

ma

s4

sa ni

dha

sa

sa

sa

sa

sa

sa

sa

ga

ga

sa

ma

ga

ma

dha

sa

ni dha

fa

sa

sa

sa

sa

dha

ni

sa

ga

ma

dha

ma

ga

sa

sa

dha

ni

sa ni

dha

dha

dha

sa

sa

pa

ni

ma

ga

ma

gf

ga

ga

dha

ni

sa

sa

sa

sa

Revagupta (26).

pa ni

ri

ri

sa

ga' pa pa ni dha

ri

ni

ma ga

ma

ga

ri

ni

sa

ni

pa

pa

ni

ga
sa

ri

ni

sa
sa

ma

ma

ma

APPENDIX D 99

Gandhara-Pancham (27).

sa

ni

sa

ga

sa

ga

ga

ga

ma

pa

ma

pa

ga

ga

ga

ga

ga

pa

sa

ga

ga

ga

ga

gani

ni

pa

ma

pa ma

ga

ga

ga

ga ^

ga

ga

ga

ga ni

ni

ni

ni

ni sa

ni

pa

ma

pa ma

ga

ga

ga

ga

ma

pa

s^

ga

ga

ga

ma

gani

ni

pa

ma

pa ma

ga

ga

ga

ga

KUKUEH (28).

dha

dha

S3

sa

dha

dha

ri

ri

dha

dha

dha

dha

pa

dha

pa

ma

ri

ri

ma

ma

pa

dha

pa

ma

pa

dha

pa

ma

ma

ma

ma

ma

ri

ri

ma

ma

dha

dha

pa

ma

pa

ma

pa

pa

dha

dha

pa

ma

pa

dha

pa

ma

sa

ri

sa

ri

ga

sa

pa

pa

pa

pa

pa

pa

Shuddh-Pancham (30).

sS'

sa-

sa-

sa-

n'

ri'

ga'

sa-

ma

2?

pa ma

ga

ri'

ri'

ri'

ri'

ma*

sa-

sa'

sa-

ri'

ri'

ga-

sa-

ma

ga-

pa ma

ga

ri'

ri'

ri'

ri'

ri'

ri-

ma'

ma*

pa

pa

dha

ma

ma

dha

sa*

sa'

ni

dha

pa

pa

dha-

ni'

ri'

ma'

ri'

ma-

pa

pa

dha

ma*

dha

ni'

pa

pa

pa

pa

Takka-Kaishik (31).

dha

dha

dha

dha

dha

ma

pa

pa

dha

dha

ri

ga

sa'

sa'

ri

ga

dha-

dha-

ma"

dha

dha'

dha'

dha'

dha

ma*

dha-

ma'

dha'

dha-

dha'

dha-

dha'

dha-

dha-

sa'

sa'

ga

ri

ma

ma

ri

ri

ma

ma

dha

ri

ma

ma

dha

dha

dha

dha

dha

dha

dha

sa

dha

pa"

ma

dha

dha

dha

dha

dha

APPENDIX E

O »- CM CO

1088 /I 200

1200

1088 /1200

APPENDIX E loi

The table shows at a glance the twenty-two srutis of Ratnakar,
and how they were used in the Gramas and Jatis. The octave is
shown divided into twenty-two equal parts in order to expose
graphically the fallacy of Bharata's theories. Shadji is first given
with the " shuddh svaras " (pure notes) and then as usually played
from ri with Grama-Sadharan, In the Kaishik scale, ri, ga, dha, ni
are a hair's breadth (one sruti or 22 cents) higher. That scale also is
shown as played from ri. Gandhari and Naishadi were played by
lowering the speaking-wire by a semitone, the former actually starting
from ri and the latter from dhaivat, although the frets used were
those of ga and ni respectively. The Gandhara-Grama is shown,
first from ga sadharan as laid down in Ratnakar, and secondly as
played from ri. The sign g distinguishes the " Graha Svara " or
initial note. The object of arranging the Jatis, etc., so as to begin
from ri or dha was to obtain the appropriate drone on the tambura,
which was tuned in ancient times to those notes. Naishadi and
Dhaivati may also have been played from ri like the others using the
scale noted by Bharata in which dha is komal, and the rest of the
svaras shuddh, and in the case of Naishadi lowering the speaking-
wire a semitone.

The fallacy underlying the theory of the equality of the srutis is
demonstrated by the numbers given. They are calculated on the
basis that a one-sruti interval is 22 cents, two srutis 112, three srutis
182, and four srutis 204. The 3 sruti interval of the Gandhara
Grama is 134 as explained in the text. It will be seen that the
ancient system required 25 srutis, and not 22, three of them being
confounded with their neighbours.

INDEX AND GLOSSARY

Cents. A means of measuring intervals, 34.
Deval. Mr. Deval's researches, 6.

Drone. Indian music is played with a drone or pedal accompaniment. The
" tambura " is the stringed instrument used for the purpose to accompany the
voice. Latterly, theatrical companies have introduced the practice of hold-
ing down two notes on the harmonium. Solo instruments have " chanterelle "
strings, which add a drone to the melody, 3, 47, 63, 73.
Grama. The material from which scales and modes are obtained. The chief
element of its connotation was a certain order of intervals. Fixity of pitch
of the notes forming the intervals was originally another element. See pages
2, 22, so, 56, 72.
Grama-Raga, 3, 63, 72.

Harmonium, the Indian, 7, 10, 20, 82, and Appendices B, C.
Instruments, tuning of keyed, to adapt them to Indian Music, 86.
Jati, the nearest equivalent to the European " Mode," 3, 23, 53, 61 ; Shuddh

Jatis, 23, 61 ; Vikrit Jatis, 61, 71.
Kaishik, 9, 54, 60.
Modes, the Church, 61.

A Murchhana was a Scale beginning with the lowest note permissible in any given
melody-type and ending on the note on octave above. The scale of a Jati
is best rendered by taking the "Graha-Svara " as the starting-point. In the
author's opinion the " Graha-svara " (initial note of a composition) formed
part of the drone. The difference between the Authentic and Plagal Modes of
Europe (page 62) would, in Indian phraseology, be partly explained by the
difference of Murchhana. See pages 3, 51, 58. The modern Indian method
is to sing all scales from shadj as a starting-point. The word Murchhana
has fallen into disuse and is frequently misunderstood.
Notation, Chapters II and III ; Accidental, 43 ; the Great Staff, 16 ; Piano-
forte Score, 19 ; Indian Measures, 41 ; embellishments, 45,
Pitch of Indian Music, 20, 55.
Raga. Explained, 3. Various modern Hindustani Ragas are mentioned on

pages 11-14, 24-32, 84-87.
Raga-Vibodh, 80.
Sadharan, 55-59.

Saptaks, the three. Music written for any voice is confined to three octaves or
"saptaks," the "tar saptak " (the highest octave), the "madhya saptak "
(the middle octave), the " mandra saptak " (the lowest octave), 16.
Scale. Explained, 21 ; earliest scales, 78 ; Sampurna, Shadava, Odava, 30, 51 ;
the tempered scale, 33-36 ; the Madras shuddh scale, 81 ; Grama scales, 29 ;
Chromatic scales, 30.

I04 INDEX AND GLOSSARY

Septimal intervals, 13, 19, 31-33.

Shuddh and Vikrit, 5, 54, 61.

Sound, theory of, 15-18 ; upper partial tones, 16, 18, Appendix B ; beats, 91.

Srutis, the twenty-two, 5, 6, 53, Appendix E. The srutis are primarily the
intervals into which the octave is divided by the sum-total of the notes in
use. The ancients held the erroneous opinion that these intervals were
equal. Their system was, however, a convenient one for distinguishing
between the major-tone (four srutis), the minor-tone (three srutis), and
semitone (two srutis). By an interval of "one sruti," they understood the
" comma " (fl), 24.

Svara. A note of a scale, or more properly a " tone," 15 ; names of the Indian
"svaras," 2.

Svaramela Kalanidhi, 80.

Tan, 59.

Text-books, confusion of existing, 48, 74, 81.

Tuning, ancient method of, 47, 63-72.
Vadi, Samvadi, 49, 55, 63.

THE END

PRINTED BY WILLIAM CLOWES AND SONS, LIMITED, LONDON AND BECCLES.

1

DATE DUE

IMVI9

ffHifk

I

'TO

'

i

GAYLORD

PRINTED IN U S A

i

I

imm

3 5002 00322 8652

Clements, Ernest

Introduction to the study of Indian musi

ML 33a .C6

Clemen-ts, Ernest., 1673-

Introduc-tion t.Q -the st.udy of
Indian music

''.-'.\-->\

:-«Sft

"A':x!(i\

;v vyA^; " ■;vv:;)^'^;>f J^>^K^v;«-:^^^