Edgar Thurston.

Castes and tribes of southern India (Volume 1) online

. (page 3 of 33)
Online LibraryEdgar ThurstonCastes and tribes of southern India (Volume 1) → online text (page 3 of 33)
Font size
QR-code for this ebook


Dravidian being separated from the Todas by reason of
their higher cephalic index, this index is, in the Todas,
actually higher than in some of the Dravidian peoples."
Accustomed as I was, in my wanderings among the
Tamil and Malayalam folk, to deal with heads in which
the dolichocephalic or sub-dolichocephalic type pre-
ponderates, I was amazed to find, in the course of an
expedition in the Bellary district (in the Canarese area),
that the question of the type of the Dravidian head was
not nearly so simple and straightforward as I had ima-
gined. My records of head measurements now include
a very large series taken in the plains in the Tulu,
Canarese, Telugu, Malayalam, and Tamil areas, and



Contemporary Science Series. f Madras Museum Bull., II, 3, 1899.



XXXV111



INTRODUCTION.



the measurements of a few Maratha (non-Dravidian)
classes settled in the Canarese country. In the fol-
lowing tabular statement, I have brought together, for
the purpose of comparison, the records of the head-
measurements of representative classes in each of these
areas :



Class.


Language.


t/3



U

en

**0 .


Cephalic Index.


<U

^3

!/! j

i.s

- c














Hi

* .2


4)








i.









rt





g


^w so






y


5


1


'c


cs






"


<


?


s


1 *


Sukun Sale


Marathi


3


S2-2


90-0


73-9


21


Suka Sale


Do.


30


Si-8


88-2


76-1


22


Vakkaliga


Canarese


So


817


93-8


72-5


27


Billava


Tulu


50


80-1


91'S


71-0


27


Rangari


Marathi


30


79-8


92-2


707


14


Agasa


Canarese


40


78-5


857


73 - 2


13


Bant


Tulu


40


78-0


91-2


70-8


12


Kapu ...


Telugu


49


78-0


87-6


71-6


16


Tota Balija


Do


39


78-0


86-0


73'3


10


Boya ...


Do


50


77 '9


89-2


70'S


14


Dasa Banajiga


Canarese


40


77-8


86-2


72-0


II


Ganiga


Do.


50


77-6


85-9


70'S


II


Golla


Telugu


60


77'5


89 "3


70-1


9


Kuruba


Canarese


50


77'3


83-9


69-6


10


Bestha


Telugu ...


60


77-1


85-1


70-5


9


Pallan


Tamil ...


50


75-9


87-0


70-1


6


Mukkuvan ...


Malayalam


40


75'i


83-5


68-6


2


Nayar


Do.


40


74 '4


81-9


70-0


I


Vellala


Tamil


40


74-1


81-1


67-9


2


Agamudaiyan


Do


40


74-0


80-9


667


I


Paraiyan


Do


40


73-6


78-3


64-8




Palli


Do


40


73 '0


80-0


64-4


I


Tiyan


Malayalam


40


73-0


78-9


68-6


...



The difference in the character of the cranium is
further brought out by the following tables, in which the



INTRODUCTION. XXXIX

details of the cephalic indices of typical classes in the
five linguistic areas under consideration are recorded :

(a) TULU. BILLAVA.





+ 4- ' Average.



83

84

85

86 4

87
88
89

90 4

9 1 +

(l>) CANARESE. VAKKALIGA.

73 +

74

75 + +

77 + +

79

80

81

82 ... ... ... Average.

83 <

84 <

85 <

86 <



xl INTRODUCTION.

87

88 +

89 4
90

92 4

93 *

94

(c) TELUGU. KAPU.

72 *

73

74 4> +

75 *4>

76

77

78 ^ ... ... ... ... Average.

79

80
81

82 +

83

84

85 +
86

87
88

(d) VELLALA. TAMIL.
68 ^
69
70

72

74 44 ... ... Average.

75
76

77
78

79
80
81



INTRODUCTION. xli

(<?) MALAYALAM. NAYAR.



7



72
73

74 + ... ... ... ... Average.

75

76
77
78

79 * +

80

81

82 4,

These tables not only bring out the difference in the
cephalic index of the classes selected as representative
of the different areas, but further show that there is a
greater constancy in the Tamil and Malayalam classes
than in the Tulus, Canarese and Telugus. The number
of individuals clustering round the average is conspicu-
ously greater in the two former than in the three
latter. I am not prepared to hazard any new theory
to account for the marked difference in the type of
cranium in the various areas under consideration, and
must content myself with the observation that, what-
ever may have been the influence which has brought
about the existing sub-brachycephalic or mesaticephalic
type in the northern areas, this influence has not
extended southward into the Tamil and Malayalam
countries, where Dravidian man remains dolicho- or
sub-dolichocephalic.

As an excellent example of constancy of type in the
cephalic index, I may cite, en passant, the following



xlii



INTRODUCTION.



results of measurement of the Todas, who inhabit the
plateau of the Nilgiri hills :

69 + +

70

71

72

74

75 + + + + + + + + +

77 +

78

79 4
80

81

I pass on to the consideration of the type of cranium
among various Brahman classes. In the following tables,
the results of measurement of representatives of Tulu,
Canarese, Marathi, Tamil and Malayalam Brahmans
are recorded :







3 <u


Cephalic Index.


Number






<*- '3




of times


Class.


Language.


o ^


bJD


3


|


index
was 80






li


rf
IH


3

"x


.1
'3


and






*


1


cS




over.


Shivalli


Tulu


30


80-4


96-4


69-4


17


Mandya


Canarese


50


80-2


88-2


69-8


31


Karnataka


Do.


60


78-4


89-5


69-8


19


Smarta (Desastha).


Marathi *


43


76-9


87-1


7i


9


Tamil (Madras city).


Tamil


40


76-s


84-


69


3


Nambutiri


Malayalam t




76-3






...


Pattar


Tamil J


25


74'5


81-4


69-1


2



* The cephalic indices of various Brahman classes in the Bombay Presidency,
supplied by Sir H. Risley, are as follows: Desastha, 76^9 ; Kokanasth, 77-3 ;
Sheni or Saraswat, 79 ; Nagar, 797.

f Measured by Mr. F. Fawcett.

J The Pattar Brahmans are Tamil Brahmans, settled in Malabar.



69 4.
70



INTRODUCTION. xliii

(a) TULU. SHIVALLI.



72 4

73 *
74

75
76
77
78

79

80 44 ... ... ... ... Average.

81 4-44

82 4-444

83

84 +0

85

86 4.

87

88 .

89 4

9

92

93
94

95
96 4

(b} CANARESE. KARNATAKA SMARTA.
70 4

72

73 ^4

74

75 +4>4>

77

79
D*



xliv INTRODUCTION.

80
81
82

83
84

85
86

87
88
89 *

(c) TAMIL. MADRAS CITY.

69 *

70 + +

72

73 + +

74 + + +

75 + + + +

76 Average.

77
78

79
80
81
82

83 4-

84

(d) TAMIL. PATTAR.
69
70

72

73

74 ... ... ... ... Average.

75

76 *

77

78 *

79 ^4

80 ^

81 *



INTRODUCTION. xlv

Taking the evidence of the figures, they demonstrate
that, like the other classes which have been analysed,
the Brahmans have a higher cephalic index, with a wider
range, in the northern than in the southern area.

There is a tradition that the Shivalli Brahmans of
the Tulu country came from Ahikshetra. As only males
migrated from their home, they were compelled to take
women from non- Brahman castes as wives. The ranks
are said to have been swelled by conversions from these
castes during the time of Sri Madhvacharya. The
Shivalli Brahmans are said to be referred to by the Bants
as Mathumaglu or Mathmalu (bride) in allusion to the
fact of their wives being taken from the Bant caste.
Besides the Shivallis, there are other Tulu Brahmans,
who are said to be recent converts. The Matti Brah-
mans were formerly considered low by the Shivallis,
and were not allowed to sit in the same line with the
Shivallis at meal time. They were only permitted to
sit in a cross line, separated from the Shivallis, though
in the same room. This was because the Matti Brah-
mans were supposed to be Mogers (fishing caste)
raised to Brahmanism by one Vathiraja Swami, a
Sanyasi. Having become Brahmans, they could not
carry on their hereditary occupation, and, to enable
them to earn a livelihood, the Sanyasi gave them
some brinjal (Solanum Melongena) seeds, and advised
them to cultivate the plant. From this fact, the variety
of brinjal, which is cultivated at Matti, is called
Vathiraja gulla. At the present day, the Matti Brah-
mans are on a par with the Shivalli Brahmans, and have
become disciples of the Sodhe mutt (religious institution)
at Udipi. In some of the popular accounts of Brah-
mans, which have been reduced to writing, it is stated
that, during the time of Mayura Varma of the Kadamba



xlvi INTRODUCTION.

dynasty,* some Andhra Brahmans were brought into
South Canara. As a sufficient number of Brahmans
were not available for the purpose of yagams (sacrifices),
these Andhra Brahmans selected a number of families
from the non- Brahman caste, made them Brahmans,
and chose exogamous sept names for them. Of these
names, Manoli (Cepkalandra Indica), Perala (Psidium
Guyava), Kudire (horse), and Ane (elephant) are
examples.

A character, with which I am very familiar, when
measuring the heads of all sorts and conditions of
natives of Southern India, is the absence of convexity
of the segment formed by the posterior portion of the
united parietal bones. The result of this absence of
convexity is that the back of the head, instead of forming
a curve gradually increasing from the top of the head
towards the occipital region, as in the European skull
figured in plate Ilia, forms a flattened area of consid-
erable length almost at right angles to the base of the
skull as in the " Hindu " skull represented in plate 111/5.
This character is shown in a marked degree in plate IV,
which represents a prosperous Linga Banajiga in the
Canarese country.

In discussing racial admixture, Quatrefages writes
as follows, f " Parfois on trouve encore quelques tribus
qui ont conserve plus on moins intacts tous les caracteres
de leur race. Les Coorumbas du Malwar [Malabar] et
du Coorg paraissent former un noyau plus considerable
encore, et avoir conserv6 dans les jungles de Wynaad
une independence a peu pres complete, et tous leurs



* According to the Brahman chronology, Mayura Varma reigned from 455
to 445 B.C., but his probable date was about 750 A.D. See Fleet, Dynasties of
the Kanarese Districts of the Bombay Presidency, 1882-86.

f Histoire generate des Races Humaines, 1889.




a. EUROPKAN SKULL.
/>. HINDU SKULL.



INTRODUCTION.



xlvii



caracteres ethnologiques." The purity of blood and
ethnological characters of various jungle tribes are
unhappily becoming lost as the result of contact meta-
morphosis from the opening up of the jungles for
planter's estates, and contact with more civilised tribes
and races, both brown and white. In illustration, I may
cite the Kanikars of Travancore, who till recently were
in the habit of sending all their women into the seclusion
of the jungle on the arrival of a stranger near their settle-
ments. This is now seldom done, and some Kanikars
have in modern times settled in the vicinity of towns,
and become domesticated. The primitive short, dark-
skinned and platyrhine type, though surviving, has
become changed, and many leptorhine or mesorhine
individuals above middle height are to be met with.
The following are the results of measurements of
Kanikars in the jungle, and at a village some miles from
Trivandrum, the capital of Travancore :






Stature cm.


Nasal Index.




Av.


Max.


Min.


Av.


Max.


Min.


Jungle


ISS'2


170-3


150-2


84-6


105


72-3


Domesticated


1587


170-4


148


81-2


90-5


70-8



Some jungle Chenchus, who inhabit the Nallamalai
hills in the Kurnool district, still exhibit the primitive
short stature and high nasal index, which are character-
istic of the unadulterated jungle tribes. But there is a
very conspicuous want of uniformity in their physical
characters, and many individuals are to be met with,
above middle height, or tall, with long narrow noses.
A case is recorded, in which a brick-maker married a
Chenchu girl. And I was told of a Boya man who



xlviii



INTRODUCTION.



had married into the tribe, and was living in a gudem
(Chenchu settlement).





Stature cm.


Nasal Index.






Av.


Max.


Min.


Av.


Max.


Min.






162-5


175


149-6


81-9


957


68-1





By the dolichocephalic type of cranium which has
persisted, and which the Chenchus possess in common
with various other jungle tribes, they are still, as shown
by the following table, at once differentiated from the
mesaticephalic dwellers in the plains near the foot of the
Nallamalais :






Cephalic
Index.


Number of
times the
index was 80
or over.


40 Chenchus ... ... ...


74*3


I




77'$


q


50 Boyas


77'Q


14




78-0


10




78-8


16




78-8


4




78-8


7




78-6


6




80'7


s









In a note on the jungle tribes, M. Louis Lapicque,*
who carried out anthropometric observations in Southern
India a few years ago, writes as follows. " Dans les
montagnes des Nilghirris et d'Anemale, situees au cceur



* Les Negres d'Asie, et la race Negre en general. Revue Scientifique, VI
July, 1906.




I.IXdA HAN A 1 1C. A.



INTRODUCTION.



xlix



de la contre'e dravidienne, on a signale" depuis longtemps
des petits sauvages cre"pus, qu'on a meme pense"
pouvoir, sur des documents insuffisants, identifier avec
les negritos. En realite", it n'existe pas dans ces
montagnes, ni probablement nulle part dans Flnde, un
te'moin de la race primitive comparable, comme purete,
aux Andamanais ni meme aux autres Negritos. Ce que
Ton trouve la, c'est simplement, mais c'est fort pre"cieux,
une population metisse qui continue au dela du Paria la
serie generate de 1'Inde. Au bord de la foret vierge ou
dans les collines partiellement defrichees, il y a des castes
demi-Parias, demi-sauvages. La hierachie sociale les
classe au-dessous du Paria ; on peut meme trouver des
groupes ou le facies negre, nettement dessine, est tout
a fait predominant. Ehbien, dans ces groupes, les
chevelures sont en general frisees, et on en observe
quelques-unes qu'on peut meme appeler crepues. On a
done le moyen de prolonger par 1'imagination la serie
des castes indiennes jusq'au type primitif qui etait (nous
n'avons plus qu'un pas a faire pour le reconstruire), un
Negre . . . Nous sommes arrives a reconstituer les
traits negres d'un type disparu en prolongeant une serie
graduee de metis. Par la meme methode nous pouvons
determiner theoriquement la forme du crane de ce type.
Avec une assez grande certitude, je crois pouvoir affirmer,
apres de nombreuses mesures systematiques, que le
negre primitif de 1'Inde etait sousdolichocephale avec
un indice voisin de 75 ou 76. Sa taille, plus difficile a
preciser, car les conditions de vie modifient ce caractere,
devait etre petite, plus haute pourtant que celle des
Andamanais. Quant au nom qu'il convient de lui
attribuer, la discussion des faits sociaux et linguistiques
sur lesquels est fondee la notion de dravidien permet
d'^tablir que ce negre etait anterieur aux dravidiens ;



1



INTRODUCTION.



il faut done 1'appeller Prddravidien, ou, si nous voulons
lui donner un nom qui ne soil pas relatif a une autre
population, on peut 1'appeler Negre Paria"

In support of M. Lapicque's statement that the
primitive inhabitant was dolichocephalic or sub-
dolichocephalic, I may produce the evidence of the
cephalic indices of the various jungle tribes which
I have examined in the Tamil, Malayalam, and Telugu
countries :

Cephalic Index.






Average.


Maximum.


Minimum.


Kadir


72'Q


80-0


6q - l


Irula, Chingleput
Kanikar ... ...


73'i

7V4


78-6

78 'Q


68-4
60 'I


Mala Vedan ... ...


73*4


80 -q


68-8


Panaiyan


74'O


81-1


60*4


Chenchu


74*3


So- f.


64"*


Sholaga
Paliyan ... ...


74-9

7V7


79'3
7Q'I


67-8

72 *Q


Irula, Nilgiris ... ... ...


7S-8


80 -Q


70 -8


Kurumba ... .., ...


76-"C


83-3


71-8











It is worthy of note that Haeckel defines the nose
of the Dravidian as a prominent and narrow organ. For
Risley has laid down * that, in the Dravidian type, the
nose is thick and broad, and the formula expressing the
proportionate dimension (nasal index) is higher than in
any known race, except the Negro ; and that the typical
Dravidian, as represented by the Male Paharia, has a
nose as broad in proportion to its length as the Negro,
while this feature in the Aryan group can fairly bear
comparison with the noses of sixty-eight Parisians,
measured by Topinard, which gave an average of 69 '4.



* Tribes and Castes of Bengal, 1891.



INTRODUCTION.



li



In this connection, I may record the statistics relating
to the nasal indices of various South Indian jungle
tribes :







Nasal Index.






Average.


Maximum.


Minimum.


Paniyan


Q^'I


108-6


72 -Q


Kadir ... ... ...


8q-8


1 1 ^'4


72'Q


Kurumba


86-1


III 'I


70-8


Sholaga


8">-i


IO7-7


72-8


Mala Vedan ...


84-0


IO2'6


7i'l


Irula, Nilgiris ...


84-0


IOO"


72*3


Kanikar ... ... ...


84-6


TO?'


72'2


Chenchu


Sl-Q


qe-7
957


68-1











In the following table, I have brought together, for
the purpose of comparison, the average stature and nasal
index of various Dravidian classes inhabiting the plains
of the Telugu, Tamil, Canarese, and Malayalam countries,
and jungle tribes :






Linguistic area.


Nasal
Index.


Stature.


Paniyan ...


Jungle tribe ... .


Qs'I


1^7-4.


Kadir ...


Do


89-8


IS7'7


Kurumba


Do
Do


86-1
8">'i


I57-9

ICQ/7


Irula, N'llgiris
Mala Vedan
Kanikar ...


Do
Do
Do


84-9
84-9
84-6


159-8
I54-2

ICC'2


Chenchu ... ...


Do.


Si-q


l62'S


Pallan


Tamil ... ... ...


8l'5


l64"?


Mukkuvan


Malayalam
Tamil . .. ...


81-
So-


163-1

i6vi


Palli


Do


77'Q


162-1;




Canarese . ..


76-1


165-8


Bestha


Telugu ... ...


7^*Q


l6V7


Tiyan


Malayalam . .


7*'


l6v7











Hi



INTRODUCTION.






Linguistic area.


Nasal
Index.


Stature.


Kuruba ...


Canarese


"7A'O


162*7


Boya


Telueu


lA'A




Tota Balija
Agasa ... ... ... ...


Do.
Canarese


74'4

7/1*7


163*9
162*4


Agamudaiyan
Golla


Tamil ...
Telugfu


74-2

74 I


165*8
:6v8


Vellala


Tamil . .


73*1


162*4


Vakkaliga
Dasa Banajiga


Canarese
Do.

Telugu


73'
72-8
72-8


167*2
165-3

ifiA'C




Malayalam


71*1


l65*2











This table demonstrates very clearly an unbroken
series ranging from the jungle men, short of stature and
platyrhine, to the leptorhine Nayars and other classes.

In plate V are figured a series of triangles represent-
ing (natural size) the maxima, minima, and average nasal
indices of Brahmans of Madras city (belonging to the
poorer classes), Tamil Paraiyans, and Paniyans. There
is obviously far less connection between the Brahman
minimum and the Paraiyan maximum than between the
Brahman and Paraiyan maxima and the Paniyan average ;
and the frequent occurrence of high nasal indices, result-
ing from short, broad noses, in many classes has to be
accounted for. Sir Alfred Lyall somewhere refers to
the gradual Brahmanising of the aboriginal non-Arayan,
or casteless tribes. "They pass," he writes, "into
Brahmanists by a natural upward transition, which leads
them to adopt the religion of the castes immediately
above them in the social scale of the composite population,
among which they settle down ; and we may reasonably
guess that this process has been working for centuries."
In the Madras Census Report, 1891, Mr. H. A. Stuart
states that " it has often been asserted, and is now the



PLATE V.




brahman.




Paraiyan.




Min.



Paniyan.

Average.

DIAGRAMS OF NOSES.



Max.



Hv INTRODUCTION.

general belief, that the Brahmans of the South are not
pure Aryans, but are a mixed Aryan and Dravidian
race. In the earliest times, the caste division was much
less rigid than now, and a person of another caste could
become a Brahman by attaining the Brahmanical standard
of knowledge, and assuming Brahmanical functions ; and,
when we see the Nambudiri Brahmans, even at the
present day, contracting alliances, informal though they
be, with the women of the country, it is not difficult to
believe that, on their first arrival, such unions were even
more common, and that the children born of them would
be recognised as Brahmans, though perhaps regarded as
an inferior class. However, those Brahmans, -in whose
veins mixed blood is supposed to run, are even to this
day regarded as lower in the social scale, and are not
allowed to mix freely with the pure Brahman community."
Popular traditions allude to wholesale conversions
of non- Brahmans into Brahmans. According to such
traditions, Rajas used to feed very large numbers of
Brahmans (a lakh of Brahmans) in expiation of some
sin, or to gain religious merit. To make up this large
number, non-Brahmans are said to have been made
Brahmans at the bidding of the Rajas. Here and there
are found a few sections of Brahmans, whom the more
orthodox Brahmans do not recognise as such, though
the ordinary members of the community regard them as
an inferior class of Brahmans. As an instance may be
cited the Marakas of the Mysore Province. Though it
is difficult to disprove the claim put forward by these
people, some demur to their being regarded as Brahmans.
Between a Brahman of high culture, with fair com-
plexion, and long, narrow nose on the one hand, and a
less highly civilised Brahman with dark skin and short
broad nose on the other, there is a vast difference, which



INTRODUCTION.



Iv



can only be reasonably explained on the assumption of
racial admixture ; and it is no insult to the higher
members of the Brahman community to trace, in their
more lowly brethren, the result of crossing with a dark-
skinned, and broad-nosed race of short stature. Whether
the jungle tribe are, as I believe, the microscopic rem-
nant of a pre-Dravidian people, or, as some hold, of
Dravidians driven by a conquering race to the seclu-
sion of the jungles, it is to the lasting influence of some
such broad-nosed ancestor that the high nasal index of
many of the inhabitants of Southern India must, it
seems to me, be attributed. Viewed in the light of this
remark, the connection between the following mixed
collection of individuals, all of very dark colour, short
of stature, and with nasal index exceeding 90, calls for
no explanation :






Stature.


Nasal
height.


Nasal
breadth.


Nasal
Index.




cm.


cm.


cm.




Vakkaliga


156


4'3


3'9


90-7


Moger


1 60


4-7


3*0








o


y


yU /


Saiyad Muhammadan ...


1 60


4'4


4


90-9


Kammalan


I54-4


4*4


4


90-9


Chakkiliyan


156-8


4'4


4


90-9


Vellala


-


4 "7


4 "2


o i *6




i ^4 o


1


5




Malaiyali


158-8


4


3-7


92-5


Konga Vellala


157


4'i


3-8


92-7


Pattar Brahman


157-6


4-2


3-9


92-9


Odde


I C,Q*6


4* 7






Smarta Brahman




O




3-9


95-1


Palli


I 57'8


4' T


3*n




Pallan ... .


1 55 '8


'
4*9.


y








*


4*2


IOO


Bestha


156-8


4 "I


4*7




Mukkuvan


150-8


6
4


6
4


IOO




ICfiM


4-7


4*7


TfVl


Tamil Paraiyan ...


1 60


J

4


6

4-2


IvAJ

l5



Ivi INTRODUCTION.

I pass on to a brief consideration of the languages
of Southern India. According to Mr. G. A. Grierson *
" the Dravidian family comprises all the principal langu-
ages of Southern India. The name Dravidian is a con-
ventional one. It is derived from the Sanskrit Dravida,
a word which is again probably derived from an older
Dramila, Damila, and is identical with the name of Tamil.
The name Dravidian is, accordingly, identical with
Tamulian, which name has formerly been used by
European writers as a common designation of the langu-
ages in question. The word Dravida forms part of the
denomination Andhra-Dravida-bhasha, the language of
the Andhras (i.e., Telugu), and Dravidas (i.e., Tamilians),
which Kumarila Bhatta (probably 7th Century A.D.)
employed to denote the Dravidian family. In India
Dravida has been used in more than one sense. Thus
the so-called five Dravidas are Telugu, Kanarese,
Marathi, Gujarati, and Tamil. In Europe, on the other
hand, Dravidian has long been the common denomination
of the whole family of languages to which Bishop Cald-
well applied it in his Comparative Grammar, and there
is no reason for abandoning the name which the founder
of Dravidian philology applied to this group of speeches."

The five principal languages are Tamil, Telugu,
Malayalam, Canarese, and Oriya. Of these, Oriya
belongs to the eastern group of the Indo-Aryan family,



Online LibraryEdgar ThurstonCastes and tribes of southern India (Volume 1) → online text (page 3 of 33)