CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Müller, Friedrich. Lectures on the Science of Language. Second Series – T08

Lecture VIII.

Few philosophers have so clearly perceived the importance
of language in all the operations of the
human mind, few have so constantly insisted on the
necessity of watching the influence of words on
thought, as Locke in his Essay concerning Human
Understanding. Of the four books into which this
great work is divided, one, the third, is entirely devoted
to Words or Language in general. At the time
when Locke wrote, but little attention had been paid
to the philosophy of language, and the author, afraid
that he might seem to have given more prominence
to this subject than it deserved, thought it necessary
to defend himself against such a charge in the following
words: — ‘What I have here said concerning
words in this third book will possibly be thought by
some to be much more than what so slight a subject
required. I allow, it might be brought into a narrower
compass; but I was willing to stay my reader
on an argument that appears to me new, and a little
out of the way (I am sure it is one I thought not of
when I began to write); that by searching it to the
bottom, and turning it on every side, some part or
other might meet with every one's thoughts, and give
occasion to the most averse or negligent to reflect on
a general miscarriage, which, though of great consequence,
334is little taken notice of. When it is considered
what a pudder is made about essences, and
how much all sorts of knowledge, discourse, and conversation
are pestered and disordered by the careless
and confused use and application of words, it will,
perhaps, be thought worth while thoroughly to lay it
open. And I shall be pardoned if I have dwelt long
on an argument which I think, therefore, needs to be
inculcated; because the faults men are usually guilty
of in this kind are not only the greatest hindrances of
true knowledge, but are so well thought of as to pass
for it. Men would often see what a small pittance of
reason and truth, or possibly none at all, is mixed
with those huffing opinions they are swelled with,
if they would but look beyond fashionable sounds,
and observe what ideas are, or are not, comprehended
under those words with which they are so armed at
all points, and with which they so confidently lay
about them. I shall imagine I have done some service
to truth, peace, and learning, if, by an enlargement
on this subject, I can make men reflect on their own
use of language, and give them reason to suspect,
that since it is frequent for others, it may also be
possible for them, to have sometimes very good and
approved words in their mouths and writings, with
very uncertain, little, or no signification. And, therefore,
it is not unreasonable for them to be wary herein
themselves, and not to be unwilling to have these
examined by others.’ *1

And again, when summing up the results of his
inquiries, Locke says: ‘For since the things the mind
contemplates are none of them, besides itself, present
335to the understanding, it is necessary that something
else, as a sign or representation of the thing it considers,
should be present to it; and these are ideas.
And because the scene of ideas that make one man's
thoughts cannot be laid open to the immediate view
of another, nor laid up anywhere but in the memory
— a no very sure repository — therefore, to communicate
our thoughts to one another, as well as record
them for our own use, signs of our ideas are also
necessary. Those which men have found most convenient,
and therefore generally make use of, are
articulate sounds. The consideration, then, of. ideas and
words as the great instruments of knowledge, makes no
despicable part of their consideration, who would take
a view of human knowledge in the whole extent of it
And, perhaps, if they were distinctly weighed and duly
considered, they would afford us another sort of logic
and critic, than what we have been hitherto acquainted

But, although so strongly impressed with the importance
which language, as such, claims in the operations
of the understanding, Locke never perceived
that general ideas and words are inseparable, that
the one cannot exist without the other, and that an
arbitrary imposition of articulate sounds to signify
definite ideas, is an assumption unsupported by any
evidence. Locke never seems to have realized the
intricacies of the names-giving process, and though
he admits frequently the difficulty, nay, sometimes
the impossibility, of our handling any general ideas
without the outward signs of language, he never
questions for a moment the received theory that at
some time or other in the history of the world men
had accumulated a treasure of anonymous general
336conceptions, to which, when the time of intellectual
and social intercourse had arrived, they prudently
attached those phonetic labels which we call words.

The age in which Locke lived and wrote was not
partial to those inquiries into the early history of
mankind which have, during the last two generations,
engaged the attention of the most eminent philosophers.
Instead of gathering the fragments of the
primitive language, poetry, and religion, not only of
the Greeks and Romans, but of all the nations of the
world, and instead of trying to penetrate, as far as
possible, into the real and actual life of the fathers of
the human race, and thus to learn how both in our
thoughts and words we came to be what we are, the
great schools of philosophy in the 18th century were
satisfied with building up theories how language
might have sprung into life, how religion might have
been revealed or invented, how mythology might
have been put together by priests, or poets, or statesmen,
for the purposes of instruction, of amusement,
or of fraud. Such systems, though ingenious and
plausible, and still in full possession of many of our
handbooks of history and philosophy, will have to give
way to the spirit of what may be called the Historical
of the 19th century. The principles of these
two schools are diametrically opposed; the one begins
with theories without facts, the other with facts without
theories. The systems of Locke, Voltaire, and
Rousseau, and in later times of Comte, are plain, intelligible,
and perfectly rational; the facts collected by men
like Wolf, Niebuhr, F. Schlegel, W. von Humboldt, Bopp,
Burnouf, Grimm, Bunsen, and others, are fragmentary,
the inductions to which they point incomplete and
obscure, and opposed to many of our received ideas.
337Nevertheless, the study of the antiquity of man, the
Palæontology of the human mind, can never again be
allowed to become the playground of mere theorizers,
however bold and brilliant, but must henceforth be
cultivated in accordance with those principles that
have produced rich harvests in other fields of inductive
research. It is no want of respect for the
great men of former ages to say that they would have
written differently if they had lived in our days.
Locke, with the results of Comparative Philology
before him, would have cancelled, I believe, the whole
of his third book ‘On the Human Understanding;’
and even his zealous and ingenious pupil, Home Tooke,
would have given us a very different volume of
‘Diversions of Purley.’ But in spite of this, there
are no books which, with all their faults — nay, on
account of these very faults — are so instructive to
the student of language as Locke's Essay, and Home
Diversions; nay, there are many points bearing
on the later growth of language which they have
handled and cleared up with greater mastery than
even those who came after them.

Thus the fact that all words expressive of immaterial
conceptions are derived by metaphor from
words expressive of sensible ideas was for the first,
time clearly and definitely put forward by Locke, and
is now fully confirmed by the researches of comparative
philologists. All roots, i. e. all the material
elements of language, are expressive of sensuous impressions,
and of sensuous impressions only; and as
all words, even the most abstract and sublime, are
derived from roots, comparative philology fully endorses
the conclusions arrived at by Locke. This is
what Locke says (iii. 4, 3): —338

‘It may also lead us a little toward the original of
all our notions and knowledge, if we remark, how
great a dependence our words have on common
sensible ideas; and how those, which are made use of
to stand for actions and notions quite removed from
sense, have their rise from thence, and, from obvious
sensible ideas are transferred to more abstruse significations,
and made to stand for ideas that come not
under the cognizance of our senses: e. g. to imagine,
apprehend, comprehend, adhere, conceive, instil, disgust,
disturbance, tranquillity
, &c., are all words taken from
the operations of sensible things, and applied to certain
modes of thinking. Spirit, in its primary signification
is breath; angel, a messenger; and I doubt not, but
if we could trace them to their sources, we should find,
in all languages, the names which stand for things that
fall not under our senses, to have had their first rise
from sensible ideas
. By which we may give some
kind of guess, what kind of notions they were and
whence derived, which filled their minds, who were
the first beginners of languages; and how nature,
even in the naming of things, unawares suggested to
men the originals and principles of all their knowledge
; whilst, to give names, that might make known
to others any operations they felt in themselves, or
any other ideas that come not under their senses,
they were fain to borrow words from ordinary known
ideas of sensation, by that means to make others the
more easily to conceive those operations they experimented
in themselves, which made no outward
sensible appearances; and then, when they had got
known and agreed names, to signify these internal
operations of their own minds, they were sufficiently
furnished to make known by words all their other
339ideas, since they could consist of nothing but either
of outward sensible perceptions, or of the inward
operations of their minds about them; we having, as
has been proved, no ideas at all, but what originally
came either from sensible objects without, or what we
feel within ourselves from the inward workings of our
own spirits, of which we are conscious to ourselves

This passage, though somewhat involved and obscure,
is a classical passage, and has formed the
subject of many commentaries, both favourable and
unfavourable. Some of Locke's followers, particularly
Home Tooke, used the statement that all abstract
words had originally a material meaning, in order to
prove that all our knowledge was restricted to sensuous
knowledge; and such was the apparent cogency
of their arguments, that, to the present day, those
who are opposed to materialistic theories consider it
necessary to controvert the facts alleged by Locke
and Home Tooke, instead of examining the cogency
of the consequences that are supposed to flow from
them. Now the facts stated by Locke seem to be
above all doubt. Spiritus is certainly derived from a
verb spirare, which means to draw breath. The same
applies to animus. Animus, the mind, as Cicero says, *2
is so called from anima, air. The root is an, which in
Sanskrit means to blow, and which has given rise to
the Sanskrit and Greek words for wind, an-ila, and
án-emos. Thus the Greek thymós, the soul, comes
from thýein, to rush, to move violently, the Sanskrit
dhu to shake. From dhu we have in Sanskrit dhûli,
340dust, which comes from the same root, and dhûma,
smoke, the Latin furnus. In Greek, the same root
supplied thýella, storm-wind, and thymós, the soul, as
the seat of the passions. Plato guesses correctly
when he says (Crat. p. 419) that thymos, soul, is so
called ἀπὸ τῆς θύσεως καὶ ζέσεως τῆς ψυχῆς. To imagine
certainly meant in its original conception to make
pictures, to picture to ourselves; but even to picture
is far too mixed an idea to have been expressed by a
simple root. Imago, picture, stands for mimago, as
imitor for mimitor, the Greek miméomai, all from a
root , to measure, and therefore meaning originally
to measure again and again, to copy, to imitate. To
and to comprehend meant to grasp at a
thing and to grasp a thing together; to adhere to
one's opinions was literally to stick to one's opinions;
to conceive was to take and hold together; to instil
was to drop or pour in; to disgust was to create a
bad taste; to disturb was to throw into disorder; and
tranquillity was calmness and particularly the smoothness
of the sea.

Look at any words expressive of objects which
cannot fall under the immediate cognisance of the
senses, and you will not have much difficulty in testing
the truth of Locke's assertion that such words are
invariably derived from others which originally were
meant to express the objects of the senses.

I begin with a list of Kafir metaphors: —

tableau Words | Literal meaning | Figurative meaning | beta | beat | punish | dhlelana | to eat together | to be on terms of intercourse | fa | to be dying | to be sick | hlala | to sit | to dwell, live, continue341

tableau Words | Literal meaning | Figurative meaning | ihlati | bush | refuge | ingeala | flying-ant | uncommon dexterity | innewadi | kind of bulbous plant | book, glass | inja | dog | a dependant | kolwa | to be satisfied | to believe | lila | to cry | to mourn | mnandi | sweet | pleased, agreeable | gauka | to be snapped asunder | to be quite dead | umsila | tail | court messenger | zidhla | to eat oneself | to be proud | akasiboni | he does not see us | he is above noticing us | nikela indhlebe | give the ears | listen attentively | ukudhla ubomi | to eat life | to live | ukudhla umntu | to eat a person | to confiscate his property | ukumgekeza inkloko | to break his head | to weary one | ukunuka umntu | to smell a person | to accuse one of witchcraft *3

Tribulation, anxiety, is derived from tribulum, a
sledge used by the ancient Romans for rubbing out
the corn, consisting of a wooden platform, studded
underneath with sharp pieces of flint or with iron
teeth. 4 The similarity between the state of mind
that had to be expressed and the state of the grains
of corn shaken in a tribulum is evident, and so striking
that, if once used, it was not likely to be forgotten
again. This tribulum, again, is derived from the verb
terere, to rub or grind. Now suppose a man's mind
so oppressed with the weight of his former misdeeds
that he can hardly breathe, or look up, or resist the
pressure, but feels crushed and ground to dust within
himself, that man would describe his state of mind as
a state of contrition, which means ‘being ground to
pieces,’ from the same verb terere, to grind.342

The French penser, to think, is the Latin pensare,
which would mean to weigh, and lead us back to
pendere, to hang. ‘To be in suspense’ literally means
to be hung up, and swaying to and fro. ‘To suspend
judgment’ means to hang it up, to keep it from taking

Doubt, again, the Latin dubium, expresses literally
the position between two points, from duo, just as the
German Zweifel points back to zwei, two.

To believe is generally identified with the German
belieben, to be pleased with a thing, to approve of it;
the Latin libet, it pleases. But to believe, as well as the
German glauben, meant originally more than simply
to approve of a thing. Both words must be traced
back to the root lubh, which has retained its original
meaning in the Sanskrit lobha, desire, and the Latin
libido, violent, irresistible desire. The same root was
taken to express that irresistible passion of the soul,
which makes man break apparently through the
evidence of the senses and the laws of reason (credo
quia absurdum
), and drives him, by a power which
nothing can control, to embrace some truth which
alone can satisfy the natural cravings of his being.
This is belief in its truest sense, though it dwindles
down in the course of time to mean no more than to
suppose, or to be pleased, just as I love, which is derived
from the same root as to believe, comes to mean, I like.

Truth has been explained by Home Tooke as that
which a man troweth. This, however, would explain
very little. To trow is but a derivative verb, meaning
to make or hold a thing true. But what is true?
True is the Sanskrit dhruva *5 and means firm, solid
anything that will hold; from dhar, to hold.343

Another word for true in Sanskrit is satya, an
adjective formed from the participle present of the
auxiliary verb as, to be. Sat is the Latin ens, being;
from it satya, true, the Greek eteós, *6 the English
sooth. If I say that sat is the Latin ens, the similarity
may not seem very striking. Yet Latin ens clearly
stands for sens, which appears in præ-sens. The
nominative singular of sat is san, because in Sanskrit
you cannot have a word ending in ns. But the accusative
sing, is santam = sentem, the nom. plur. santas
= sentes; so that there can be no doubt as to the
identity of the two words in Sanskrit and Latin.

And how did language express what, if it were a
rational conception at all, would seem to be the most
immaterial of all conceptions — namely, nothing? It
was expressed in the only way in which it could be
expressed — namely, by the negation of, or the comparison
with, something real and tangible. It was
called in Sanskrit asat, that which is not being; in
Latin nihil, i. e. nihilum, 7 which stands for nifĭlum,
344i. e. ne-fîlum, and means ‘not a thread or shred.’ In
French rien is actually a mere corruption of rem, the
accusative of res, and retains its negative sense even
without the negative particle by which it was originally
preceded. Thus ne-pas is non-passum, not a
step; ne-point is non-punctum, not a point. The French
néant, Italian niente, are the Latin non ens. And
now observe for a moment how fables will grow up
under the charm of language. It was perfectly
correct to say, ‘I give you nothing,’ i. e. ‘I give you
not even a shred.’ Here we are speaking of a relative
nothing; in fact, we only deny something, or decline
to give something. It is likewise perfectly correct
to say, on stepping into an empty room, ‘There is
nothing here,’ meaning not that there is absolutely
nothing, but only that things which we expect to
find in a room are not there. But by dint of using
such phrases over and over again, a vague idea is
gradually formed in the mind of a Nothing, and
Nihil becomes the name of something positive and
real. People at a very early time began to talk of
the Nothing as if it were something; they talked
and trembled at the idea of annihilation — an idea
utterly inconceivable, except in the brain of a mad-man.
Annihilation, if it meant anything, could etymologically
— and in this case, we may add, logically
too — mean nothing but to be reduced to a something
which is not a shred — surely no very fearful state,
345considering that in strict logic it would comprehend
the whole realm of existence, exclusive only of what
is meant by shred. Yet what speculations, what
fears, what ravings, have sprung from this word Nihil
— a mere word, and nothing else! We see things
grow and decay, we witness the birth and death of
living things, but we never see anything lost or
annihilated. Now, what does not fall within the
cognizance of our senses, and what contradicts every
principle of our reasoning faculties, has no right to be
expressed in language. We may use the names of
material objects to express immaterial objects, if they
can be rationally conceived. We can conceive, for
instance, powers not within the ken of our senses, yet
endowed with a material reality. We can call them
spirits, literally breezes, though we understand perfectly
well that by spirits we mean something else
than mere breezes. We can call them ghosts, a name
connected with gust, yeast, gas, and other almost
imperceptible vapours. But a Nothing, an absolute
Nothing, that is neither visible, nor conceivable, not
imaginable, ought never to have found expression,
ought never to have been admitted into the dictionary
of rational beings.

Now, if we consider how people talk about the
Nothing, how poets make it the subject of the most
harrowing strains; how it has been, and still is, one of
the principal ingredients in most systems of philosophy
— nay, how it has been dragged into the domain
of religious thought, and, under the name of Nirvâna,
has become the highest goal of millions among the
followers of Buddha — we may perhaps, even at this
preliminary stage of our inquiries, begin to appreciate
the power of language over thought, and feel less
346surprise at the ancient nations for having allowed the
names of natural objects, the sky, the sun, the moon,
the dawn, and winds, to assume the character of
supernatural powers or divine personalities, or for
having offered worship and sacrifice to such abstract
names as Fate, Justice, or Victory. There is as much
mythology in our use of the word Nothing as in the
most absurd portions of the mythological phraseology
of India, Greece, and Rome: and if we ascribe the
former to a disease of language, the causes of which
we are able to explain, we shall have to admit that in
the latter, language has reached to an almost delirious
state, and has ceased to be what it was meant to be,
the expression of the impressions received through
the senses, or of the conceptions of a rational mind.

But to return to Locke's statement, that all names
of zmmaterial objects are derived from the names of
material objects. Many philosophers, as I remarked,
instead of grappling manfully with the conclusions
that are supposed to flow from Locke's observation,
have preferred to question the accuracy of his observation.

Victor Cousin, in his ‘Lectures on the History of
Philosophy during the Eighteenth Century,’ *8 endeavours
to controvert Locke's assertion by the following
process: — ‘I shall give you two words,’ he says, ‘and I
shall ask you to trace them back to primitive words
expressive of sensible ideas. Take the word je, I.
This word, at least in all languages known to me, is
not to be reduced, not to be decomposed, primitive;
and it expresses no sensible idea, it represents nothing
but the meaning which the mind attaches to it; it is
347a pure and true sign, without any reference to any
sensible idea. The word être, to be, is exactly in the
same case; it is primitive and altogether intellectual.
I know of no language in which the French verb être
is rendered by a corresponding word that expresses a
sensible idea; and therefore it is not true that all the
roots of language, in their last analysis, are signs of
sensible ideas.’

Now it must be admitted that the French je, which
is the Sanskrit aham, is a word of doubtful etymology.
It belongs to the earliest formations of Aryan
speech, and we need not wonder that even in Sanskrit
the materials out of which this pronoun was
formed should have disappeared. We can explain in
English such words as myself or your honour, but we
could not attempt, with the means supplied by English
alone, to analyse I, thou, and he. It is the same with the
Sanskrit aham, a word carried down by the stream of
language from such distant ages, that even the Vedas,
as compared with them, are but, as it were, of yesterday.
But though the etymology of aham is doubtful,
it has never been doubtful to any scholar that, like
all other words, it must have an etymology; that it
must be derived either from a predicative or from a
demonstrative root. Those who would derive aham
from a predicative root, have thought of the root ah,
to breathe, to speak. *9 Those who would derive it
348from a demonstrative root, refer us to the Vedic gha,
the later ha, this, used like the Greek hóde. How the
pronoun of the first person is expressed in Chinese
we saw in an earlier Lecture, and although such expressions
as ‘servant says,’ instead of ‘I say,’ may
seem to us modern and artificial, they are not so in
Chinese, and show at all events that even so colourless
an idea as I may meet with signs sufficiently pale
and faded to express it. *10

With regard to être, to be, the case is different.
Être 11 is the Latin esse, changed into essere and contracted.
The root, therefore, is as, which, in all the
Aryan languages, has supplied the material for the
auxiliary verb. Now even in Sanskrit, it is true,
this root as is completely divested of its material
character; it means to be, and nothing else. But
there is in Sanskrit a derivative of the root as,
namely, ásu, and in this asu, which means the vital
breath, the original meaning of the root as has been
preserved. As, in order to give rise to such a noun
as asu, must have meant to breathe, then to live, then
to exist, and it must have passed through all these
stages before it could have been used as the abstract
auxiliary verb which we find not only in Sanskrit
but in all Aryan languages. Unless this one derivative
asu, life, had been preserved in Sanskrit, it would
349have been impossible to guess the original material
meaning of the root as, to be; yet even then the
student of language would have been justified in
postulating such a meaning. And even in French,
though être may seem an entirely abstract word, the
imperfect j'étais, the participle été are clearly derived
from Latin stare, to stand, and show how easily so
definite an idea as to stand may dwindle down to the
abstract idea of being. If we look to other languages,
we shall find again and again the French verb être
rendered by corresponding words that expressed
originally a sensible idea. Our verb to be is derived
from Sanskrit bhû, which, as we learn from Greek
phýó, meant originally to grow. *12 I was is connected
with the Gothic visan, which means to dwell.

But though on this point the student of language
must side with Locke, and admit, without one single
exception, the material character of all words, nothing
can be more convincing than the manner in which
Victor Cousin disposes of the conclusions which some
philosophers, though certainly not Locke himself,
seem inclined to draw from such premises. ‘Further,’
he writes, ‘even if this were true, and absolutely
true, which is not the case, we could conclude no
more than this. Man is at first, by the action of all
his faculties, carried out of himself and toward the
external world; the phenomena of the external world
strike him first, and hence these phenomena receive
the first names. The first signs are borrowed from
sensible objects, and they are tinged to a certain extent
by their colours. When man afterwards turns
350back on himself, and lays hold more or less distinctly
of the intellectual phenomena which he had
always, though somewhat vaguely, perceived; if, then,
he wants to give expression to the new phenomena of
mind and soul, analogy leads him to connect the signs
he seeks with those he already possesses: for analogy
is the law of each growing or developed language.
Hence the metaphors to which our analysis traces back
most of the signs and names of the most abstract
moral ideas,’

Nothing can be truer than the caution thus given
by Cousin to those who would use Locke's observation
as an argument in favour of an one-sided sensualistic

Metaphor is one of the most powerful engines in
the construction of human speech, and without it we
can hardly imagine how any language could have
progressed beyond the simplest rudiments. Metaphor
generally means the transferring of a name from the
object to which it properly belongs to other objects
which strike the mind as in some way or other participating
in the peculiarities of the first object.
The mental process which gave to the root mar the
meaning of to propitiate was no other than this,
that men perceived some analogy between the smooth
surface produced by rubbing and polishing and the
smooth expression of countenance, the smoothness of
voice, and the calmness of looks produced even in
an enemy by kind and gentle words. Thus, when
we speak of a crane, we apply the name of a bird
to an engine. People were struck with some kind
of similarity between the long-legged bird picking
up his food with his long beak and their rude engines
for lifting weights. In Greek, too, géranos has both
351meanings. This is metaphor. Again, cutting remarks,
glowing words, fervent prayers, slashing articles, all
are metaphor. Spiritus in Latin meant originally
blowing, or wind. But when the principle of life
within man or animal had to be named, its outward
sign, namely, the breath of the mouth, was naturally
chosen to express it. Hence in Sanskrit asu, breath
and life; in Latin spiritus, breath and life. Again,
when it was perceived that there was something else
to be named, not the mere animal life, but that which
was supported by this animal life, the same word was
chosen, in the Modern Latin dialects, to express the
spiritual as opposed to the mere material or animal
element in man. All this is metaphor.

We read in the Veda, ii. 3, 4: *13 — ‘Who saw the
first-born when he who had no form (lit. bones) bore
him that had form? Where was the life (asuḥ), the
blood (asṛik), the self (âtmâ) of the earth? Who
went to ask this from any that knew it?’

Here breath, blood, self, are so many attempts at
expressing what we should call cause.

But let us now consider for a moment that what philosophers,
and particularly Locke, have pointed out as
a peculiarity of certain words, such as to apprehend, to
, to understand, to fathom, to imagine, spirit
and angel, must have been, in reality, a peculiarity of
a whole period in the early history of speech. No
advance was possible in the intellectual life of man
without metaphor. Most roots that have yet been discovered,
had originally a material meaning, and a meaning
so general and comprehensive 14 that they could
352easily be applied to many special objects. We meet
with roots meaning to strike, to shine, to creep, to
grow, to fall, but we never meet with primitive roots
expressive of states or actions that do not fall under
the cognisance of the senses, nor even with roots expressive
of such special acts as ‘raining, thundering,
hailing, sneezing, trying, helping,’ Yet Language has
been a very good housewife to her husband, the human
Mind; she has made very little go a long way. With
a very small store of such material roots as we just
mentioned, she has furnished decent clothing for the
numberless offspring of the Mind, leaving no idea, no
sentiment unprovided for, except, perhaps, the few
which, as we are told by some poets, are inexpressible.

Thus from roots meaning to shine, to be bright,
names were formed for sun, moon, stars, the eyes of
man, gold, silver, play, joy, happiness, love. With
roots meaning to strike, it was possible to name an
axe, the thunderbolt, a fist, a paralytic stroke, a striking
remark, and a stroke of business. From roots
meaning to go, names were derived for clouds, for ivy,
for creepers, serpents, cattle and chattel, moveable
and immoveable property. With a root meaning to
crumble, expressions were formed for sickness and
death, for evening and night, for old age and for the
fall of the year.

We must now endeavour to distinguish between
two kinds of metaphor, which I call radical and
poetical. I call it radical metaphor when a root which
means to shine is applied to form the names, not only
of the fire or the sun, bat of the spring of the year,
the morning light, the brightness of thought, or the
joyous outburst of hymns of praise. Ancient languages
353are brim full of such metaphors, and under
the microscope of the etymologist every word almost
discloses traces of its first metaphorical conception.

From this we must distinguish poetical metaphor,
namely, when a noun or verb, ready made and assigned
to one definite object or action, is transferred
poetically to another object or action. For instance,
when the rays of the sun are called the hands or
fingers of the sun, the noun which means hand or
finger existed ready made, and was, as such, transferred
poetically to the stretched out rays of the sun.
By the same process the clouds are called mountains,
the rain-clouds are spoken of as cows with heavy
udders, the thunder-cloud as a goat or as a goat-skin,
the sun as a horse, or as a bull, or as a giant bird, the
lightning as an arrow, or as a serpent.

What applies to nouns, applies likewise to verbs. A
verb such as ‘to give birth’ is used, for instance, of
the night producing, or, more correctly, preceding the
day, as well as of the day preceding the night. The
sun, under one name, is said to beget the dawn, because
the approach of daylight gives rise to the dawn;
under another name the sun is said to love the dawn,
because he follows her as a bridegroom follows after
his bride; and lastly, the sun is said to destroy the
dawn, because the dawn disappears as soon as the sun
has risen. From another point of view the dawn may
be said to give birth to the sun, because the sun seems
to spring from her lap; she may be said to die or disappear
after having given birth to her brilliant son,
because as soon as the sun is born, the dawn must
vanish. All these metaphors, however full of contradictions,
were perfectly intelligible to the ancient
poets, though to our modern understanding they are
354frequently riddles difficult to solve. We read in the
Rig-Veda (x. 189), *15 where the sunrise is described,
that the dawn comes near to the sun, and breathes her
last when the sun draws his first breath. The commentators
indulge in the most fanciful explanations
of this expression without suspecting the simple conception
of the poet, which after all is very natural.

Let us consider, then, that there was, necessarily
and really, a period in the history of our race when
all the thoughts that went beyond the narrow horizon
of our every-day life had to be expressed by means
of metaphors, and that these metaphors had not yet
become what they are to us, mere conventional and
traditional expressions, but were felt and understood
half in their original and half in their modified character.
We shall then perceive that such a period of
thought and speech must be marked by features very
different from those of any later age.

One of the first results would naturally be that
objects in themselves quite distinct, and originally
conceived as distinct by the human intellect, would
nevertheless receive the same name. If there was a
root meaning to shine forth, to revive, to gladden, that
root might be applied to the dawn, as the burst of
brightness after the dark night, to a spring of water,
gushing forth from the rock and gladdening the heart
of the traveller, and to the spring of the year, that
awakens the earth after the death-like rest of winter.
The spring of the year, the spring of water, the
dayspring, would thus go by the same name, they
would be what Aristotle calls homonymous or namesakes.
On the other hand, the same object might
strike the human mind in various ways. The sun
355might be called the warming and generating, but
likewise the scorching and killing; the sea might
be called the barrier as well as the bridge, and the
high-road of commerce; the clouds might be spoken
of as bright cows with heavy udders, or as dark
and roaring demons. Every day that dawns in the
morning might be called the twin of the night that
follows the day, or all the days of the year might be
called brothers, or so many head of cattle which are
driven to their heavenly pasture every morning, and
shut up in the dark stable of Augeias at night. In this
manner one and the same object would receive many
names, or would become, as the Stoics called it, polyonymous,
many-named — having many alias's. Now
it has always been pointed out as a peculiarity of what
we call ancient languages, that they have many words
for the same thing, these words being sometimes called
synonymes; and likewise, that their words have frequently
very numerous meanings. Yet what we call
ancient languages, such as the Sanskrit of the Vedas
or the Greek of Homer, are in reality very modern
languages; that is to say, they show clear traces of
having passed through many, many successive periods
of growth and decay, before they became what we know
them to be in the earliest literary documents of India
and Greece. What, then, must have been the state of
these languages in their earlier periods, before many
names, that might have been and were applied to
various objects, were restricted to one object, and
before each object, that might have been and was
called by various names, was reduced to one name!
Even in our days we confess that there is a great deal
in a name; how much more must that have been the
case during the primitive ages of man's childhood!356

The period in the history of language and thought
which I have thus endeavoured to describe as characterised
by what we may call two tendencies, the homonymous
and the polyonymous *16 I shall henceforth call
the mythic or mythological period, and I shall try to
show how much that has hitherto been a riddle in the
origin and spread of myths becomes intelligible if
considered in connection with the early phases through
which language and thought must necessarily pass.

Before I enter, however, on a fuller explanation of
my meaning, I think it right to guard from the beginning
against two mistakes, to which the name of
Mythic Period might possibly give rise. What I call
a period is not so in the strict sense of the word: it
has no fixed limits that could be laid down with
chronological accuracy. There is a time in the early
history of all nations in which the mythological character
predominates to such an extent that we may
speak of it as the mythological period, just as we
might call the age in which we live the age of discoveries.
But the tendencies which characterize the
mythological period, though they necessarily lose
much of that power with which, at one time, they
swayed every intellectual movement, continue to work
under different disguises in all ages, even in our
own, though perhaps the least given to metaphor,
poetry, and mythology.

Secondly, when I speak of a mythological period,
I do not use mythological in the restricted sense in
which it is generally used, namely, as being necessarily
connected with stories about gods, heroes, and
heroines. In the sense in which I use mythological, it
357is applicable to every sphere of thought and every
class of words, though, from reasons to be explained
hereafter, religious ideas are most liable to mythological
expression. Whenever any word, that was
at first used metaphorically, is used without a clear
conception of the steps that led from its original
to its metaphorical meaning, there is danger of mythology;
whenever those steps are forgotten and artificial
steps put in their places, we have mythology, or,
if I may say so, we have diseased language, whether
that language refers to religious or secular interests.
Why I use the term mythological in this wide sense,
a sense not justified by Greek or Roman usage, will
appear when we come to see how what is commonly
called mythology is but a part of a much more general
phase through which all language has at one time or
other to pass.

After these preliminary remarks, I now proceed to
examine some cases of what I called radical and
poetical metaphor.

Cases of radical metaphor, though numerous in
radical and agglutinative languages, are less frequent
in inflectional languages, such as Sanskrit, Greek,
and Latin. Nor is it difficult to account for this. It
was the very inconvenience caused by words which
failed to convey distinctly the intention of the
speaker that gave the impulse to that new phase of
life in language which we call inflectional. Because
it was felt to be important to distinguish between the
bright one, i. e. the sun, and the bright one, i. e. the day,
and the bright one, i. e. wealth, therefore the root vas,
to be bright, was modified by inflection, and broken
up into Vi-vas-vat, the sun, vas-ara, day, vas-u, wealth.
In a radical and in many an agglutinative language,
358the mere root vas would have been considered sufficient
to express, pro re natâ, any one of these meanings.
Yet inflectional languages, too, yield frequent instances
of radical metaphor, some of which, as we shall see,
have led to very ancient misunderstandings, and, in
course of time, to mythology.

There is, for instance, in Sanskrit, a root ark or
arch, which means to be bright; but, like most primitive
verbs, it is used both in a transitive and intransitive
sense, thus meaning both to be bright and to
make bright
. Only ‘to make bright’ meant more in
that ancient language than it means with us. To
make bright meant to cheer, to gladden, to celebrate,
to glorify, and it is constantly used in these different
senses by the ancient poets of the Veda. Now, by a
very simple and intelligible process, the meaning of
this root arch might be transferred to the sun, or the
moon, or the stars; all of them might be called arch
or ṛich without any change in the outward appearance
of the root. For all we know, ṛich, as a substantive,
may really have conveyed all these meanings during
the earliest period of the Aryan languages. But if we
look at the fully developed branches of that family of
speech, we find that in this, its simplest form, ṛich has
been divested of all meanings, except one; it only
means a song of praise, a hymn, that gladdens the heart
and brightens the countenance of the gods, or that
makes their power effulgent and manifest. *17 The other
meanings, however, which ṛich might have expressed
were not entirely given up; they ivere only rendered
359more definite by new and distinct grammatical modifications
of the same root. Thus, in order to express
light or ray, archí was formed, a masculine, and very
soon also a neuter, archís. Neither of these nouns is
ever used in the sense of praise which clings to ṛich;
they have only the sense of light and splendour.

Again, quite regularly, a new derivative was
formed, namely, arkáḥ, a masculine. This likewise
means light, or ray of light, but it has been fixed
upon as the proper name of the light of lights,
the sun. Arkáḥ, then, by a very natural metaphor,
became one of the many names of the sun; but by
another metaphor, which we explained before, arkah,
with exactly the same accent and gender, was also
used in the sense of hymn of praise. Now here we
have a clear case of radical metaphor in Sanskrit. It
was not the noun arkáḥ, in the sense of sun, that was,
by a bold flight of fancy, transferred to become the
name of a hymn of praise, nor vice versâ. The same
root arch, under exactly the same form, was bestowed
independently on two distinct conceptions. If the
reason of the independent bestowal of the same root
on these two distinct ideas, sun and hymn, was forgotten,
there was danger of mythology, and we actually
find in India that a myth sprang up, and that hymns
of praise were fabled to have proceeded from or to
have originally been revealed by the sun.

Our root arch offers us another instance of the same
kind of metaphor, but slightly differing from that just
examined. From ṛich in the sense of shining, it was
possible to form a derivative ṛíkta, in the sense of
lighted up, or bright. This form does not exist in
Sanskrit, but as kt in Sanskrit is liable to be changed
360into ks *18 we may recognise in ṛiksha the same derivative
of ṛich. Ṛiksha, in the sense of bright, has
become the name of the bear, so called either from his
bright eyes or from his brilliant tawny fur. 19 The
same name ṛiksha was given in Sanskrit to the stars,
the bright ones. It is used as a masculine and neuter
in the later Sanskrit, as a masculine only in the Veda.
In one passage of the Rig-Veda, i. 24, 10, we read
as follows: — ‘These stars fixed high above, which are
seen by night, whither did they go by day?’ The
commentator, it is curious to observe, is not satisfied
with this translation of ṛiksha in the sense of stars in
general, but appeals to the tradition of the Vâjasa
, in order to show that-the stars here called
ṛikshas are the same constellation which in later
Sanskrit is called ‘the Seven Rishis,’ or ‘the Seven
Sages,’ They are the stars that never seem to set
361during the night, and therefore the question whither
they went by day would be specially applicable to
them. Anyhow, the tradition is there, and the question
is whether it can be explained. Now, remember,
that the constellation here called the Ṛikshas, in the
sense of the bright ones, would be homonymous in
Sanskrit with the Bears. Remember also, that,
apparently without rhyme or reason, the same constellation
is called by Greeks and Romans the Bear,
in the singular, árktos and ursa. There may be some
similarity between that constellation and a waggon or
wain, but there is not a shadow of a likeness with a
bear. You will now perceive the influence of words
on thought, or the spontaneous growth of mythology.
The name ṛiksha was applied to the bear in the sense
of the bright fuscous animal, and in that sense it
became most popular in the later Sanskrit, and in
Greek and Latin. The same name, in the sense of
the bright ones, had been applied by the Vedic poets
to the stars in general, and more particularly to that
constellation which, in the northern parts of India, was
the most prominent. The etymological meaning of
ṛiksha, as simply the bright stars, was forgotten, the
popular meaning of ṛiksha, bear, was known to everybody.
And thus it happened that when the Greeks
had left their central home and settled in Europe, they
retained the name of Árktos for the same unchanging
stars, but not knowing why these stars had originally
received that name, they ceased to speak of them as
árktoi, or many bears, and spoke of them as the Bear,
the Great Bear, adding a bear-ward, the Arcturus
(oûros, ward), and in time even a Little Bear. Thus
the name of the Arctic regions rests on a misunderstanding
of a name framed thousands of years ago in
362Central Asia, and the surprise with which many a
thoughtful observer has looked at these seven bright
stars, wondering why they were ever called the bear,
is removed by a reference to the early annals of
human speech.

On the other hand, the Hindus also forgot the
original meaning of ṛiksha. It became a mere name,
apparently with two meanings, star and bear. In
India, however, the meaning of bear predominated,
and as ṛiksha became more and more the established
name of the animal, it lost in the same degree its connection
with the stars. So when, in later times, their
Seven Sages had become familiar to all under the
name of the Seven Ṛishis, the seven Ṛikshas, being
unattached, gradually drifted towards the Seven Ṛishis,
and many a fable sprang up as to the seven poets
dwelling in the seven stars. Such is the origin of a

The only doubtful point in the history of the myth
of the Great Bear is the uncertainty which attaches
to the exact etymological meaning of ṛiksha, bear.
We do not see why of all other animals the bear
should have been called the bright animal. *20 It is true
that the reason of many a name is beyond our reach,
and that we must frequently rest satisfied with the
fact that such a name is derived from such a root, and
therefore had originally such a meaning. The bear
was the king of beasts with many northern nations,
who did not know the lion; and it would be difficult
to say why the ancient Germans called him Goldfusz,
golden-footed. But even if the derivation of ṛiksha
363from arch were given up, the later chapters in the
history of the word would still remain the same. We
should have ṛiksha, star, derived from arch, to shine,
mixed up with ṛiksha, bear, derived from some other
root, such as, for instance, arś or ṛiś, to hurt; but the
reason why certain stars were afterwards conceived as
bears would not be affected by this. It should also
be stated that the bear is little known in the Veda.
In the two passages of the Rig-Veda where ṛiksha
occurs, it is explained by Sâyana, in the sense of hurtful
and of fire, not in that of bear. In the later
literature, however, ṛiksha, bear, is of very common

Another name of the Great Bear, or originally the
Seven Bears, or really the seven bright stars, is Septemtriones.
The two words which form the name are
occasionally used separately; for instance, ‘quas nostri
septem solid vocitare triones
.’ *21 Varro (L. L. vii. 73-75),
in a passage which is not very clear, tells us that
triones was the name by which, even at his time,
ploughmen used to call oxen when actually employed
for ploughing the earth. 22If we could quite depend
on the fact that oxen were ever called triones, we might
accept the explanation of Varro, and should have to admit
that at one time the seven stars were conceived as
seven oxen. But as a matter of fact, trio is never used
in this sense, except by Varro, for the purpose of an
etymology, nor are the seven stars ever again spoken
of as seven oxen, but only as ‘the oxen and the shaft,’
364boves et temo, a much more appropriate name. Boōtes,
too, the ploughman or cow-driver, given to the same
star which before we saw called Arcturus, or bear-keeper,
would only imply that the waggon (hámaxa)
was conceived as drawn by two or three oxen, but not
that all the seven stars were ever spoken of as oxen.
Though, in matters of this kind, it is impossible to
speak very positively, it seems not improbable that
the name triones, which certainly cannot be derived
from terra, may be an old name for star in general.
We saw that the stars in Sanskrit were called
star-as, the strewers of light; and the Latin stella is
but a contraction of sterula. The English star, the
German Stern, come from the same source. But besides
star, we find in Sanskrit another name for star,
namely, târâ, where the initial s of the root is lost.
Such a loss is by no means unfrequent, *23 and trio, in
Latin, might therefore represent an original strio,
star. The name strio, star, having become obsolete, like
ṛiksha, the Septentriones remained a mere traditional
name; and if, as Varro tells us, there was a vulgar
name for ox in Latin, namely, trio, which then would
have to be derived from tero, to pound, the peasants
speaking of the Septem triones, the seven stars, would
naturally imagine themselves speaking of seven oxen.

But as I doubt whether the seven stars ever suggested
by themselves the picture of seven animals,
whether bears or cows, I equally question whether the
seven were ever spoken of as temo, the shaft. Yarro
says they were called ‘boves et temo,’ ‘oxen and shaft,’
but not that they were called both oxen and shaft.
We can well imagine the four stars being taken for
365oxen, and the three for the shaft; or again, the four
stars being taken for the cart, one star for the shaft,
and two for the oxen; but no one, I think, could
ever have called the seven together the shaft. But
then it might be objected that temo, in Latin, means not
only shaft, but carriage, and should be taken as an
equivalent of hámaxa. This might be, only it has
never been shown that temo in Latin meant a carriage.
Varro, *24 no doubt, affirms that it was so, but we
have no further evidence. For if Juvenal says (Sat.
iv. 126), ‘De temone Britanno eoccidet Arviragus,’ this
really means from the shaft, because it was the custom
of the Britons to stand fighting on the shafts of
their chariots. 25 And in the other passages, 26 where
temo is supposed to mean car in general, it only means
our constellation, which can in no wise prove that temo
by itself ever had the meaning of car.

Temo stands for tegmo, and is derived from the root
taksh, which likewise yields tignum, a beam. In French,
too, le timon is never a carriage, but the shaft, the
German Deichsel, the Anglo-Saxon þixl or þisl, §27
366words which are themselves, in strict accordance with
Grimm's law, derived from the same root (tvaksh, or
taksh) as temo. The English team, on the contrary,
has no connection with temo or timon, but comes from
the Anglo-Saxon verb teon, to draw, the German
ziehen, the Gothic tiuhan, the Latin duco. It means
drawing, and a team of horses means literally a draught
of horses, a line of horses, ein Zug Pferde. The verb
teon, however, like the German ziehen, had likewise
the meaning of bringing up, or rearing; and as in
German ziehen, Zucht, and züchten, so in Anglo-Saxon
team was used in the sense of issue, progeny; teamian
(in English, for distinctness sake, spelt to teem) took
the sense of producing, propagating, and lastly of

According to the very nature of language, mythological
misunderstandings such as that which gave
rise to the stories of the Great Bear must be
more frequent in ancient than in modern dialects.
Nevertheless, the same mythological accidents will
happen even in modern French and English. To
speak of the seven bright stars, the Ṛikshas, as the
Bear, is no more than if in speaking of a walnut we
were to imagine that it had anything to do with a
wall. Walnut is the A. S. wealh-hnut, in German
Wälsche Nuss. Wälsch in German means originally
foreigner, barbarian, and was especially applied by
the Germans to the Italians. Hence Italy is to the
present day called Welschland in German. The
Saxon invaders gave the same name to the Celtic
inhabitants of the British Isles, who are called wealh
in Anglo-Saxon (plur. wealas). Hence the walnut
meant originally the foreign nut. In Lithuanian the
walnut goes by the name of the ‘Italian nut,’ in
367Russian by that of ‘Greek nut,’ *28 What Englishman,
in speaking of walnut, thinks that it means
foreign or Italian nut? But for the accident that
walnuts are no wall fruit, I have little doubt that by
this time schoolmasters would have insisted on spelling
the word with two l's, and that many a gardener would
have planted his walnut trees against the wall.

There is a soup called Palestine soup. It is made, I
believe, of artichokes called Jerusalem artichokes, but
the Jerusalem artichoke is so called from a mere misunderstanding.
The artichoke, being a kind of sunflower,
was called in Italian girasole, from the Latin
gyrus, circle, and sol, sun. Hence Jerusalem artichokes
and Palestine soups!

One other instance may here suffice, because we
shall have to return to this subject of modern mythology.
One of the seven wonders of the Dauphine
in France is la Tour sans venin, 29 the Tower without
poison, near Grenoble. It is said that poisonous
animals die as soon as they approach it. Though the
experiment has been tried, and has invariably failed,
yet the common people believe in the miraculous
power of the locality as much as ever. They appeal
to the name of la Tour sans venin, and all that the
more enlightened among them can be made to concede
is that the tower may have lost its miraculous character
in the present age, but that it certainly possessed
it in former days. The real name, however, of the tower
and of the chapel near it is San Verena or Saint Vrain.
This became san veneno, and at last sans venin.368

But we must return to ancient mythology. There
is a root in Sanskrit, ghar, which, like ark, means
to be bright and to make bright. *30 It was originally
used of the glittering of fat and ointment. This
earliest sense is preserved in passages of the Veda,
where the priest is said to brighten up the fire by
sprinkling butter on it. It never means sprinkling
in general, but always sprinkling with a bright fatty
substance (beglitzern). 31 From this root we haveghṛita,
the modern ghee, melted butter, and in general anything
fat (Schmalz), the fatness of the land and of the
clouds. Fat, however, means also bright, and hence
the dawn is called ghṛitápratikâ, bright-faced. Again,
the fire claims the same name, as well as ghṛitánirṇij,
with garments dripping with fat or with brilliant garments.
The horses of Agni or fire, too, are called ghṛitádpṛishṭhâk,
literally whose backs are covered with fat;
but, according to the commentator, well fed and shining.
The same horses are called vîtaprishṭha, with beautiful
backs, and ghṛitasnâh, bathed in fat, glittering, bedewed.
Other derivatives of this root ghar are ghṛiṇâ,
heat of the sun; in later Sanskrit ghṛiṇâ, warmth of
the heart or pity, but likewise heat or contempt.
Ghṛíṇi, too, means the burning heat of the sun.
Gharmá is heat in general, and may be used for anything
that is hot, the sun, the fire, warm milk, and
even the kettle. It is identical with Greek thermós,
and Latin formus, warm.

Instead of ghar we also find the root har, a slight
modification of the former, and having the same meaning.
369This root has given rise to several derivatives.
Two very well-known derivatives are hári and harít,
both meaning originally bright, resplendent. Now
let us remember that though occasionally both the
sun and the dawn are conceived by the Vedic poets
as themselves horses, *32 that is to say, as racers,
it became a more familiar conception of theirs to
speak of the sun and the dawn as drawn by horses.
These horses are very naturally called hdri, or harit,
bright and brilliant; and many similar names, such as
aruṇá, arushá, rohít, &c., 33 are applied to them, all expressive
of brightness of colour in its various shades.
After a time these adjectives became substantives.
Just as haríṇâ, from meaning bright brown, came to
mean the antelope, as we speak of a bay instead of a
bay horse, the Vedic poets spoke of the Harits as the
horses of the Sun and the Dawn, of the two Haris as
the horses of Indra, of the Rohits as the horses of
Agni or fire. After a time the etymological meaning
of these words was lost sight of, and hari and harit
became traditional names for the horses which either
represented the Dawn and the Sun, or were supposed
to be yoked to their chariots. When the Vedic poet
says, ‘The Sun has yoked the Harits for his course,’
what did that language originally mean? It meant
no more than what was manifest to every eye, namely,
that the bright rays of light which are seen at dawn
before sunrise, gathered in the east, rearing up to the
sky, and bounding forth in all directions with the
quickness of lightning, draw forth the light of the
sun, as horses draw the car of a warrior. But who
370can keep the reins of language? The bright ones,
the Harits, run away like horses, and very soon they
who were originally themselves the dawn, or the rays
of the Dawn, are recalled to be yoked as horses to
the car of the Dawn. Thus we read (Rv. vii. 75, 6),
‘The bright brilliant horses are seen bringing to us
the shining Dawn,’

If it be asked how it came to pass that rays of light
should be spoken of as horses, the most natural answer
would be that it was a poetical expression such as any
one might use. But if we watch the growth of language
and poetry, we find that many of the later
poetical expressions rest on the same metaphorical
principle which we considered before as so important
an agent in the original formation of nouns, and that
they were suggested to later poets by earlier poets,
i. e. by the framers of the very language which they
spoke. Thus in our case we can see that the same
name which was given to the flames of fire, namely,
vahni, was likewise used as a name for horse, vahni
being derived from a root vah, to carry along. There
are several other names which rays of light and horses
share in common, so that the idea of horse would
naturally ring through the mind whenever these names
for rays of light were touched. And here we are once
again in the midst of mythology; for all the fables of
Helios, the sun, and his horses, flow irresistibly from
this source.

But more than this. Remember that one of the
names given to the horses of the sun was Harit; remember
also that originally these horses of the sun
were intended for the rays of the dawn, or, if you like,
for the Dawn itself. In some passages the Dawn is
simply called aśvâ, the mare, originally the racing
371light. Even in the Veda, however, the Harits are
not always represented as mere horses, but assume
occasionally, like the Dawn, a more human aspect.
Thus, vii. 66, 15, they are called the Seven Sisters,
and in another passage (ix. 86, 37) they are represented
with beautiful wings. Let us now see whether
we can find any trace of these Harits or bright ones
in Greek mythology, which, like Sanskrit, is but
another dialect of the common Aryan mythology.
If their name exists at all in Greek, it could only be
under the form of Charis, Charites. The name, as
you know, exists, but what is its meaning? It never
means a horse. The name never passed through that
phase in the minds of the Greek poets which is so
familiar in the poetry of the Indian bards. It retained
its etymological meaning of lustrous brightness,
and became, as such, the name of the brightest brightness
of the sky, of the dawn. In Homer, Charis is
still used as one of the many names of Aphrodite, and,
like Aphrodite, she is called the wife of Hephæstos. *34
Aphrodite, the sea-born, was originally the dawn, the
most lovely of all the sights of nature, and hence very
naturally raised in the Greek mind to the rank of
goddess of beauty and love. As the dawn is called
in the Veda Duhitâ Divaḥ, the daughter of Dyaus,
372Charis, the dawn, is to the Greeks the daughter
of Zeus. One of the names of Aphrodite, Argynnis,
which the Greeks derived from a name of a sacred
place near the Cephissus, where Argynnis, the beloved
of Agamemnon had died, has been identified *35 with the
Sanskrit arjunî, the bright, the name of the dawn.
In progress of time the different names of the dawn
ceased to be understood, and Eos, Ushas, as the most
intelligible of them, became in Greece the chief representative
of the deity of the morning, drawn, as in
the Veda, by her bright horses. Aphrodite, the seaborn,
also called Enalia 36 and Pontia, became the
goddess of beauty and love, and was afterwards degraded
by an admixture of Syrian mythology. Charis,
on the contrary, was merged in the Charites, 37 who,
instead of being, as in India, the horses of the dawn,
were changed by an equally natural process into the
attendants of the bright gods, and particularly of
Aphrodite, whom ‘they wash at Paphos and anoint
with oil,’ §38 as if in remembrance of their descent from
the root ghar, which, as we saw, meant to anoint, to
render brilliant by oil.

It has been considered a fatal objection to the
history of the word Charis, as here given, that in Greek
373it would be impossible to separate Charis from other
words of a more general meaning. ‘What shall we do,
says Curtius, *39 with cháris, chará, chaírô, charízomai,
charíeis?’ Why, it would be extraordinary if such
words did not exist, if the root ghar had become
withered as soon as it had produced this one name of
Charis. These words which Curtius enumerates are
nothing but collateral offshoots of the same root which
produced the Harits in India and Charis in Greece.
One of the derivatives of the root har was carried off
by the stream of mythology, the others remained on
their native soil. Thus the root dyu or div gives rise
among others to the name of Zeus, in Sanskrit Dyaus,
but this is no reason why the same word should not
be used in the original sense of heaven, and produce
other nouns expressive of light, day, and similar
notions. The very word which in most Slavonic
languages appears in the sense of brightness, has in
Illyrian, under the form of zora, become the name of
the dawn. 40 Are we to suppose that Charis in Greek
meant first grace, beauty, and was then raised to the
rank of an abstract deity? It would be difficult to
find another such deity in Homer, originally a mere
abstract conception, 41 and yet made of such flesh
and bone as Charis, the wife of Hephæstos. Or shall
we suppose that Charis was first, for some reason
or other, the wife of Hephaestos, and that her name
afterwards dwindled down to mean splendour §42 or
charm in general; so that another goddess, Athene,
could be said to shower charis or charms upon a man?
374To this, too, I doubt whether any parallel could be
found in Homer, Everything, on the contrary, is clear
and natural, if we admit that from the root ghar or har,
to be fat, to be glittering, was derived, besides harit,
the bright horse of the sun in Sanskrit, and Charis,
the bright dawn in Greece, cháris meaning brightness
and fatness, then gladness and pleasantness in
general, according to a metaphor so common in ancient
language. It may seem strange to us that the charts,
that indescribable grace of Greek poetry and art,
should come from a root meaning to be fat, to be
greasy. Yet as fat and greasy infants grow into ‘airy,
fairy Lilians,’ so do words and ideas. The Psalmist
(cxxxiii. 2) does not shrink from even bolder metaphors.
‘Behold, how good and how pleasant (charíen)
it is for brethren to dwell together in unity!
It is like the precious ointment upon the head that
ran down upon the beard, even Aaron's beard: that
went down to the skirts of his garments,’ After the
Greek cháris had grown, and assumed the sense of
charm, such as it was conceived by the most highlycultivated
of races, no doubt it reacted on the mythological
Charis and Charites, and made them the embodiment
of all that the Greeks had learnt to call lovely
and graceful, so that in the end it is sometimes difficult
to say whether cháris is meant as an appellative
or as a mythological proper name. Yet though thus
converging in the later Greek, the starting-points of
the two words were clearly distinct — as distinct at least
as those of arka, sun, and arka, hymn of praise, which
we examined before, or as Dyaus, Zeus, a masculine,
and dyaus, a feminine, meaning heaven and day.
Which of the two is older, the appellative or the proper
name, Charis, the bright dawn, or cháris, loveliness,
375is a question which it is impossible to answer,
though Curtius declares in favour of the priority of
the appellative. This is by no means so certain as he
imagines. I fully agree with him when he says that
no etymology of any proper name can be satisfactory
which fails to explain the appellative nouns with
which it is connected; but the etymology of Charis
does not fail here. On the contrary, it lays bare the
deepest roots from which all its cognate offshoots can
be fully traced both in form and meaning, and it
can defy the closest criticism, both of the student of
comparative philology and of the lover of ancient
mythology. *43

In the cases which we have hitherto examined, a
mythological misunderstanding arose from the fact
that one and the same root was made to yield the
names of different conceptions; that after a time the
two names were supposed to be one and the same,
which led to the transference of the meaning of one to
the other. There was one point of similarity between
the bright bear and the bright stars to justify the
ancient framers of language in deriving from the
same root the names of both. But when the similarity
in quality was mistaken for identity in substance, mythology
became inevitable. The fact of the seven
bright stars being called Arktos, and being supposed
to mean the bear, I call mythology, and it is important
to observe that this myth has no connection whatever
with religious ideas or with the so-called gods of
antiquity. The legend of Kallisto, the beloved of Zeus,
and the mother of Arkas, has nothing to do with the
original naming of the stars. On the contrary, Kallisto
376was supposed to have been changed into the Arktos, or
the Great Bear, because she was the mother of Arkas,
that is to say, of the Arcadian or bear race, and her
name, or that of her son, reminded the Greeks of their
long-established name of the Northern constellation.
Here, then, we have mythology apart from religion,
we have a mythological misunderstanding very like
in character to those which we alluded to in ‘Palestine
soup’ and La Tour sans venin.

Let us now consider another class of metaphorical
expressions. The first class comprehended those
cases which owed their origin to the fact that two
substantially distinct conceptions received their name
from the same root, differently applied. The metaphor
had taken place simultaneously with the formation of
the words; the root itself and its meaning had been
modified in being adapted to the different conceptions
that waited to be named. This is radical metaphor.
If, on the contrary, we take such a word as star and
apply it to a flower; if we take the word ship and
apply it to a cloud, or wing and apply it to a sail; if
we call the sun horse, or the moon cow; or with verbs,
if we take such a verb as to die and apply it to the
setting sun, or if we read —

‘The moonlight clasps the earth,
And the sunbeams kiss the sea.’ *44

we have throughout poetical metaphors. These, too,
are of very frequent occurrence in the history of early
language and early thought. It was, for instance, a
very natural idea for people who watched the golden
beams of the sun playing as it were with the foliage
of the trees, to speak of these outstretched rays as
377hands or arms. Thus we see that in the Veda, *45
Savitar, one of the names of the sun, is called golden-handed.
Who would have thought that such a simple
metaphor could ever have caused any mythological
misunderstanding? Nevertheless, we find that the
commentators of the Veda see in the name golden-handed,
as applied to the sun, not the golden splendour
of his rays, but the gold which he carries in
his hands, and which he is ready to shower on his
pious worshippers. A kind of moral is drawn from
the old natural epithet, and people are encouraged to
worship the sun because he has gold in his hands to
bestow on his priests. We have a proverb in German,
Morgenstunde hat Gold im Munde,’ ‘Morning-hour
has gold in her mouth,’ which is intended to inculcate
the same lesson as,

‘Early to bed, and early to rise,
Makes a man healthy, and wealthy, and wise.’

But the origin of the German proverb is mythological.
It was the conception of the dawn as the golden light,
some similarity like that between aurum and aurora,
which suggested the proverbial or mythological expression
of the ‘golden-mouthed Dawn’ — for many
proverbs are chips of mythology. But to return to
the golden-handed Sun. He was not only turned
into a lesson, but he also grew into a respectable
myth. Whether people failed to see the natural
meaning of the golden-handed Sun, or whether they
would not see it, certain it is that the early theological
378treatises of the Brahmans *46 tell of the Sun as
having cut his hand at a sacrifice, and the priests
having replaced it by an artificial hand made of
gold. Nay, in later times the Sun, under the name
of Savitar, becomes himself a priest, and a legend is
told how at a sacrifice he cut off his hand, and how
the other priests made a golden hand for him.

All these myths and legends which we have hitherto
examined are clear enough; they are like fossils of the
most recent period, and their similarity with living
species is not to be mistaken. But if we dig somewhat
deeper, the similarity is less palpable, though
it may be traced by careful research. If the German
god Tyr, whom Grimm identifies with the Sanskrit
sun-god, 47 is spoken of as one-handed, it is because the
name of the golden-handed Sun had led to the conception
of the sun with one artificial hand, and afterwards,
by a strict logical conclusion, to a sun with but
one hand. Each nation invented its own story how
Savitar or Tyr came to lose their hands; and while
the priests of India imagined that Savitar hurt his
hand at a sacrifice, the sportsmen of the North told how
Tyr placed his hand, as a pledge, into the mouth of
the wolf, and how the wolf bit it off. Grimm compares
the legend of Tyr placing his hand, as a pledge, into
the mouth of the wolf, and thus losing it, with an
Indian legend of Sûrya or Savitar, the sun, laying
hold of a sacrificial animal and losing his hand by its
bite. This explanation is possible, but it wants confirmation,
particularly as the one-handed German god
Tyr has been accounted for in some other way. Tyr
379is the god of victory, as Wackernagel points out, and
as victory can only be on one side, the god of victory
might well have been thought of and spoken of as
himself one-handed. *48

It was a simple case of poetical metaphor if the
Greeks spoke of the stars as the eyes of the night.
But when they speak of Argos the all-seeing (Panóptēs),
and tell of his body being covered with eyes, we have
a clear case of mythology.

It is likewise perfectly intelligible when the poets
of the Veda speak of the Maruts or storms as singers.
This is no more than when poets speak of the music
of the winds; and in German such an expression as
‘The wind sings’ (der Wind singt) means no more
than the wind blows. But when the Maruts are called
not only singers, but musicians — nay, wise poets in the
Veda 49 — then again language has exceeded its proper
limits, and has landed us in the realm of fables.

Although the distinction between radical and
poetical metaphor is very essential, and helps us more
than anything else toward a clear perception of the
origin of fables, it must be admitted that there are
cases where it is difficult to carry out this distinction.
If modern poets call the clouds mountains, this is
clearly poetical metaphor; for mountain, by itself,
never means cloud. But when we see that in the
Veda the clouds are constantly called parvata, and
that parvata means, etymologically, knotty or rugged,
it is difficult to say positively whether in India the
clouds were called mountains by a simple poetical
metaphor, or whether both the clouds and the mountains
380were from the beginning conceived as full of
ruggedness and undulation, and thence called parvata.
The result, however, is the same, namely, mythology;
for if in the Veda it is said that the Maruts or storms
make the mountains to tremble (i. 39, 5), or pass
through the mountains (i. 116, 20), this, though
meaning originally that the storms made the clouds
shake, or passed through the clouds, came to mean, in
the eyes of later commentators, that the Maruts
actually shook the mountains or rent them asunder.

Appendix to lecture VIII.

Dr. Sonne, in several learned articles published in
‘Kuhn's Zeitschrift’ (x. 96, 161, 321, 401), has subjected
my conjecture as to the identity of harit and
cháris to the most searching criticism. On most points
I fully agree with him, as he will see from the more
complete statement of my views given in this Lecture;
and I feel most grateful to him for much additional
light which his exhaustive treatise has thrown on the
subject. We differ as to the original meaning of the
root ghar, which Dr. Sonne takes to be effusion or
shedding of light, while I ascribe to it the meaning of
glittering and fatness; yet we meet again in the
explanation of such words as ghṛiṇâ, pity; háras,
wrath; hṛiṇi, wrath; hṛiṇîte, he is angry (p. 100).
These meanings Dr. Sonne explains by a reference
to the Russian kraska, colour; krasnoĭ, red, beautiful;
krasa, beauty; krasnjetĭ, to blush; krasovatĭsja, to
rejoice. Dr. Sonne is certainly right in doubting the
381identity of chaírō and Sanskrit hṛish, the Latin horreo,
and in explaining chaírō as the Greek form of ghar,
to be bright and glad, conjugated according to the
fourth class. Whether the Sanskrit haryati, he desires,
is the Greek thélei, seems to me doubtful.

Why Dr. Sonne should prefer to identify cháris,
cháritos, with the Sanskrit hdri, rather than with harit,
he does not state. Is it on account of the accent?
I certainly think that there was a form cháris, corresponding
to hári, and I should derive from it the
accusative chárin, instead of chárita; also adjectives
like chárieis (hárivat). But I should certainly retain
the base which we have in hárit, in order to explain
such forms as cháris, cháritos. That chárit in Greek
ever passed through the same metamorphosis as the
Sanskrit harít, that it ever to a Greek mind conveyed
the meaning of horse, there is no evidence whatever.
Greek and Sanskrit myths, like Greek and Sanskrit
words, must be treated as co-ordinate, not as subordinate;
nor have I ever, as far as I recollect, referred
Greek myths or Greek words to Sanskrit as their
prototypes. What I said about the Charítes was very
little. On page 81 of my ‘Essay on Comparative
Mythology,’ I said: —

‘In other passages, however, they (the Harits) take
a more human form; and as the Dawn, which is sometimes
simply called aśvâ, the mare, is well known by
the name of the sister, these Harits also are called the
Seven Sisters (vii. 66, 15); and in one passage (ix.
86, 37) they appear as the Harits with beautiful
wings. After this I need hardly say that we have
here the prototype of the Grecian Charítes

If on any other occasion I had derived Greek from
Sanskrit myths, or, as Dr. Sonne expresses it, ethnic
382from ethnic myths, instead of deriving both from a
common Aryan or pro-ethnic source, my words might
have been liable to misapprehension. *50 But as they
stand in my essay, they were only intended to point
out that after tracing the Harits to their most primitive
source, and after showing how, starting from
thence, they entered on their mythological career in
India, we might discover there, in their earliest form,
the mould in which the myth of the Greek Charítes
was cast, while such epithets as ‘the sisters,’ and ‘with
beautiful wings,’ might indicate how conceptions that
remained sterile in Indian mythology, grew up under
a Grecian sky into those charming human forms which
we have all learned to admire in the Graces of Hellas.
That I had recognised the personal identity, if we
may say so, of the Greek Charis, the Aphrodite, the
Dawn, and the Sanskrit Ushas, the dawn, will be seen
from a short sentence towards the end of my essay,
p. 86: —

‘He (Eros) is the youngest of the gods, the son of
Zeus, the friend of the Charítes, also the son of the
chief Charis, Aphrodite, in whom we can hardly fail
to discover a female Eros (an Ushâ, dawn, instead of
an Agni aushasya)’.

Dr. Sonne will thus perceive that our roads, even
where they do not exactly coincide, run parallel, and
that we work in the same spirit and with the same
objects in view.383

1* Locke, On the Understanding, iii. 5, 16.

2* Cicero, Tuscul. i. 9, sub fin. Locke, Human Understanding,
iv. 3, 6, note (ed. London, 1836, p. 412). ‘Anima sit animus
ignisve nescio,’ &c.

3* Appleyard, l. c. p. 70.

4 See White, Latin-English Dictionary, s. v.

5* Kuhn's Zeitschrift, vii. 62.

6* See Pott, Etymologische Forschungen, ii. p. 364; Kern, in
Kuhn's Zeitschrift, viii. 400. It should be remembered that in
satya, the t belongs to the base, and that the derivative element
is not tya, Greek σιὸς, but ya. Whether εὸς represents the same
suffix as ya in Sanskrit may be doubtful. See, however, Bopp,
Vergleich. Gr. (2), § 109 a, 2 (p. 212); and § 956. Sattva in
Sanskrit means being and a being.

7 Cf. Kuhn, Zeitschrift, i. 544. Dietrich mentions similar
cases of shortening, such as cognĭtus and nótus, pejĕro and
jûro. Bopp has clearly given up the etymology of nihil, which
he proposed in the first edition of his Comparative Grammar,
as it is suppressed in the second. It is to be regretted that even
so careful a scholar as Mr. White, in his excellent Latin-English
, should still quote from the first edition only of Bopp's
work. As to h taking the place of f, we know that in Spanish
every Latin f is represented by h, e. g. hablar = fabulari, hijo =
filius, hierro = ferrum, hilo = filum. But in Latin itself these
two letters are frequently interchangeable. Instead of hircus, the
Sabines said fircus; instead of hædus, fædus; instead of harena,
farena. Nay, double forms are mentioned in Latin, such as hordeum
and fordeum; hostis and fostis; hariolus and fariolus.
See Corssen, Aussprache der Lateinischen Sprache, p. 46.

8* Paris, 1841. Vol. ii. p. 274.

9* I thought it possible, in my History of Sanskrit Literature,
p. 21, to connect ah-am with Sanskrit âha, I said, Greek ,
Latin ajo and nego, nay, with Gothic ahma (instead of agma),
spirit, but I do so no longer. Nor do I accept the opinion of
Benfey (Sanskrit Grammatik, § 773), who derives aham from the
pronominal root gha with a prosthetic a. It is a word which,
for the present, must remain without a genealogy.

10* Jean Paul, in his Levana, p. 32, says, ‘“I” is — excepting
God, the true I and true Thou at once — the highest and most
incomprehensible that can be uttered by language, or contemplated.
It is there all at once, as the whole realm of truth and
conscience, which, without “I,” is nothing. We must ascribe it
to God, as well as to unconscious beings, if we want to conceive
the being of the One and the existence of the others.’

11 Cf. Diez, Lexicon, s. v. essere.

12* See M. M.'s Essay on the Aryan and Aboriginal Languages
of India
, p. 344.

13* M. M., History of Sanskrit Literature, p. 20.

14 The specialization of general roots is more common than the
generalization of special roots, though both processes must be

15* See M. M., Die Todtenbestaltung der Brahmanen, p. xi.

16* Augustinus, De Civ. Dei, vii. 16. ‘Et aliquando unum deum
res plures, aliquando unam rem deos plures faciunt.’

17* The passage in the Vájasaneyi Sanhitā, 13, 39, ‘ṛiché tvâ
ruché tvâ,’ contains either an isolated remnant of the original
import of the root, preserved in a proverbial phrase, or it is an
etymological play.

18* Kuhn, in the Zeitschrift für die Wissenschaft der Sprache,
i. 155, was the first to point out the identity of Sk. ṛiksha and
Greek ἄρκτος in their mythological application. He proved that
ksh in Sanskrit represented an original kt, in takshan, carpenter,
Gr. τέκτων; in kshi, to dwell, κτίω; in vakshas, Lat. pectus.
Curtius, in his Grundzüge, added kshan, to kill, Gr. κταν;
Aufrecht (Kuhn's Zeitschrift, viii. 71), kshi, to kill, κτι; Leo
Meyer (v. 374), ksham, earth, Gr. χθών. To these may be added
kshi, to possess, κτάομαι; and perhaps kshu, to sneeze, πτύω, if it
stands for κτύω.

19 Grimm (D. W. s. v. Auge and Băr) compares ṛiksha, Bär,
not only with ἄρκτος, ursus, Lith. lokis (instead of olkis, orkis),
Irish art (instead of arct), but also with Old High-German elah,
which is not the bear but the elk, the alces described by Cæsar,
B. G. vi. 27. This alces, however, the Old High-German elah,
would agree better with ṛiśa or riśya, some kind of roebuck, mentioned
in the Veda (Rv. viii. 4. 10), with which Weber (K. Z.
vi. 320) has well compared ircus, the primitive form of hircus
(Quintil. i. 5, 20).

20* See, however, Welcker's remarks on the wolf in his Griechische
, p. 64.

21* Arat. in N. D. ii. 41, 105.

22 Triones enim boves appellantur a bubulcis etiam nunc
maxume quom arant terram; e quis ut dicti Talentes glebarii qui
facile proscindunt glebas, sic omnis qui terram arabant a terra
terriones, unde triones ut dicerentur e detrito.

23* See Kuhn, Zeitschrift, iv. 4 seq.

24* L. L. vii. 75. Temo dictus a tenendo, is enim continet
jugum. Et plaustrum appellatum, a parte totum, ut multa.

25 Cæs. B. G. iv. 33, v. 16.

26 Stat. Theb. i. 692. Sed jam temone supino Languet hyperboreæ
glacialis portitor Ursæ.

Stat. Theb. i. 370. Hyberno deprensus navita ponto, Cui neque
temo piger, neque amico sidere monstrat Luna vias.

Cic. N. D. ii. 42 (vertens Arati carmina) Arctophylax, vulgo
qui dicitur esse Bootes, Quod quasi temone adjunctam præ se
quatit Arcton.

Ovid, Met. x. 447. Interque triones Flexerat obliquo plaustrum
temone Bootes.

Lucan, lib. iv. v. 523. Flexoque Ursæ temone paverent.

Propert. iii. 5, 35. Cur serus versare boves et plaustra Bootes.

27§ In A. S. þisl is used as a name of tbe constellation of
Charles's Wain; like temo.

28* Pott, E. F. ii. 127. Itóliskas rĕssutys; Gréczkoĭ orjech.
The German Lamberts-nuss is nux Lombardica. Instead of
walnut we find welshnut, Philos. Transact. xviii. p. 819, and
walshnut in Gerarde's Herbal. In the Index to the Herbal
walnut is spelt with two l's, and classed with wallflower.

29 Brosses, Formation Mécanique des Langues, ii. 133.

30* Cf. Kuhn's Zeitschrift, i. 154, 566; iii. 346 (Schweizer),
iv. 354 (Pictet).

31 Rv. ii. 10, 4. ‘Jígbarmy agním havíshâ ghṛiténa,’ I anoint
or brighten up the fire with oblations of fat.

32* M. M.'s Essay on Comparative Mythology, p. 82. Böhtlingk
, Wörterbuch, s. v. aśva.

33 Cf. M. M.'s Essay on Comparative Mythology, pp. 81-83.

34* Il. xviii. 382:
τὴν δὲ ἴδε προμολοῦσα Χάρις λιταροκρήδεμνος
καλὴ τὴν ὢπυιε περικλυτός Ἀμφιγυήεις¤
In the Odyssey, the wife of Hephæstos is Aphrodite; and Nägelsbach,
not perceiving the synonymous character of the two names,
actually ascribed the passage in Od. viii. to another poet, because
the system of names in Homer, he says, is too firmly established
to allow of such variation. He likewise considers the marriage
of Hephæstos as purely allegorical. (Homerische Theologie, p. 114.)

35* Sonne, in Kuhn's Zeitschrift, x. 350. Rv. i. 49, 3. Arjuna,
a name of Indra, mentioned in the Brâhmaṇas, &c.

36 Cf. Ápyâ yóshâ, Rv. x. 10, 4; ápyâ yóshaṇâ, 11, 2.

37 Kuhn, Zeitschrift, i. 518, x. 125. The same change of one
deity into many took place in the case of the Moira, or fate. The
passages in Homer where more than one Moira are mentioned,
are considered as not genuine (Od. vii. 197, Il. xxiv. 49); but
Hesiod and the later poets are familiar with the plurality of the
Moiras. See Nägelsbach, Nachhomerische Theologie, p. 150.
Welcker, Griechische Götterlehre, p. 53.

38§ Od. vii. 364.

39* Curtius, G. E. i. 97.

40 Pictet, Origines, i. 155. Sonne, Kuhn's Zeitschrift, x. 354.

41 See Kuhn, Herabholung des Feuers, p. 17.

42§ Sonne, l. c. x. 355-6.

43* See Appendix at the end of this Lecture.

44* Cox, Tales of the Gods and Heroes, p. 55.

45* i. 22, 5, hiraṇyapâṇim ûtaye Savitâram upa hvaye.

i. 35, 9, hiraṇyapâṇih Savitâ vicharshaṇiḥ ubhe dyâvâpṛithivî
antar îyate.

i. 35, 10, hiraṇyahasta.

46* Kaushîtaki-brâhmaṇa, l. c. and Sâyaṇa.

47 Deutsche Mythologie, xlvii. p. 187.

48* Schweitzer Museum, i. 107.

49 Rv. i. 19, 4; 38, 15; 52, 15. Kuhn, Zeitschrift, i. 521.

50* I ought to mention, however, that Mr. Cox, in the Introduction
to his Tales of the Gods and Heroes, p. 67, has understood my
words in the same sense as Dr. Sonne. ‘The horses of the sun,’
he writes, ‘are called Harits; and in these we have the prototype
of the Greek Charites — an inverse transmutation, for while in the
other instances the human is changed into a brute personality, in
this the beasts are converted into maidens.’