Press of Isaac Friedenwald,

32 S. Paca St.



The ritual literature of the Atharva-Veda, like that of the other Vedas, has attached to itself certain paricistas, or supplements.
Of these, the thirty-fifth, according to the best accessible MS, is the Asurl-Kalpa, an abhicara, or witchcraft practice, containing rites to be used in connection with the asurl-plant. The question as to what this plant was will be discussed below. The use of the word kalpa for such a text is explained by a passage in the Atharvanlya-Paddhati, which states, on the authority of Uparvarsa,' that in addition to the five AV. kalpas—K a u 9 i k a, Vaitana, Naksatra, ^anti, and A n g i r a s a—which are called prz^/z ' inspired,' there are certain other kalpas which are to be
considered as smrti ' handed down by tradition.' ^
Three MSS have been consulted in preparing this paper. Two of them are copies ofjhe parifistas of the AV. ; the third is a commentary to the Asuri-Kalpa. All three are loans to Dr. Bloomfield from the British Government in India, Just here I may say that I am greatly indebted to Dr. Bloomfield for the use of these MSS, for the encouragement and assistance which he has given me, and for his kindness in looking over my work. The MSS are as follows:
A, large sheets of light yellow paper, bound in book form, written lengthwise in a large clear hand and with considerable
care. It is a modern copy.
B, narrow sheets of light blue paper, bound in book form, written lengthwise, text fuller in places than the preceding, but in a poor hand and with numerous errors. It must be a very recent copy. Both of these MSS are numbered 23.
S (Scholiast), much older than either of the preceding, single sheets of light brown paper grown dark at the edges, written lengthwise as the other MSS, but in a very poor, though large,

1. A mimansa (purva-) teacher. See Life and Essays of H. T. Colebrooke, Vol. II, pp. 319-49.
2.  Cf. J. A. O.S. XI 377, Bloomfield, On the Position of the Vaitana- Sutra in the Literature of the Atharva-Veda.

Page 2

hand. It contains three sections or chapters, Part first (ScXxos ib-6a') contains all the practices in brief form, and occupies about one-fourth of the MS. Part second (folios Sa'-jb") treats only of the externals of the principal rite, and occupies scarcely one-tenth of the MS. Part third is an elaborate commentary on what has preceded ; but in its present condition deals with only about two thirds of the practices, since the MS lacks some folios at the close. 1
This MS is numbered 120; but is also marked p (pattrdni) 18; sam. {saihvat) 1880-81 ; and, on the last folio, written across the end on the margin, 347. From the appearance of the MS it might be as old as one hundred and fifty years; and since sam. 1880-81 probably has reference to Kielhorn's Report 2, the MS may be as old as it looks. It contains about 200 clokas.
At the beginning of part third it names Mahadeva as the speaker, 3 who introduces his commentary (see p. 11, note 19) by saying :
' It [the mantra] is not to be uttered (performed) without teachers; by the precept of a teacher this magic power (success) [comes into being]. Accordingly in a single final commentary the Asurl-[rite] should succeed,' —
vhid gurun akariavyam gJirtivdkyena siddhtdam  (cod. sldhi-), ekdntiniatikdmadhye (cod. ekdnte-) sddhayeta tad dsiiri.

2. Mahadeva 5 is spoken of as the seer of the divine asurl-text 6, and as becomes a rsi he speaks of the Gayatri, Tristubh, and Anustubh metres {gdyatrHnstubamsh(pchandah'), after which he proceeds to give full instructions concerning the rites.
1. Part first seems to be in fact a version of the paricista, fuller than the text and differing from it in some passages, but still essentially the same.
The chief points of difference have been noted as readings of S. They have been put into gloka form where the MS seemed to warrant it. Readings froms parts second and third are so marked.
2 See p. 5, foot-note i. The MS is catalogued on p. 58.
3 The paricistas as a whole are in the form of dialogues. Cf. Weber, History of Indian Literature, p. 153.
4 The conjectural reading siddhidam requires a regular fem. noun to be regarded as neu.; but for this text it may be allowable, since the MS departs widely in places from all rules of grammar, and also treats siddhi as a neu. in other passages.
The comma and period (, and .) have been used in all Sanskrit passages as the simplest means of transliterating the two Sanskrit marks of punctuation (I and II).
5 An epithet of Rudra or Civa, also of Visnu and the name of various persons.
It is an appropriate title, " Great-Lord," for the teacher of such a text,
6 asya cryasurimantrasya (cod. criastu-) mahadeva rsih.

Page 3

Apart from its subject-matter S possesses no little interest, because it contains abundant evidence of the character of the people having to do with its rites. It is exceedingly corrupt, as a few examples may suffice to show. The common writing for sapta is satpa ; for asuri, asiiri ; for suksma, suksma; for curna, curna, etc: juhiyat and juhiyata are used for juhuyat; mryate for mriyate, etc.: rdayam is found for hrdayam; bhimantritenia for abhi- (beginning of a sentence); karaye for yet; titha for tithir, etc.: little or no attention is paid to samdhi: the confusion of sibilants, 1 s for q and vice versa, is exceedingly common: and other curious freaks in spelling occur, notably the use of cy for c {cyurna for curna, and muncyati for muncati), which is of some interest from a phonetic standpoint, and the writing of the word vacikartukama in eight different ways, while using it but twelve times, with a mistake of some kind in every single instance. 2
The errors are doubtless due in part to later copyists ; but, from the present state of corruption, it may be safe to infer that the original MS was bad at the start ; for it seems hardly possible that the scribes should be guilty of all the errors which it contains, even if the present MS is the result of several successive transcriptions.
The nature of the mistakes stamps the writer at once as an ignorant and perhaps degraded person. It is about such a document as might be expected to be written in English by some Voodoo doctor among the blacks of the South. Numerous repetitions serve to light up otherwise hopeless passages, and when the brief outlines of the paricista are combined with the commentary the whole practice becomes clear. No two of the MSS exactly agree in the order in which the different forms of the rite are treated, and B has a passage not found in either A or S. Fortunately the paricista is mostly written in clokas, which is of great service in determining the true reading. 3 In style the paricista is somewhat like the sutras, being terse and technical in its forms of expression, and consisting mostly of what may

1 Cf. Proc. a. O. S., May, 1886. Introduction to the Study of the Old-Indian Sibilants; by Prof. Bloomfield and Dr. Edward H. Spieker.
2 It may be said in addition that there is hardly a sentence in the entire MS in which there are not mistakes in the case-forms, the most common being the use of a stem-form for an ace.

3  In the text, where a MS reading is of no importance, it has been thought best to omit it; so, in the quotations from S the MS reading has been omitted where the emendation is obvious, where the same mistake is repeated several times, and, in a few instances, where MS evidence warrants the change; on the other hand, where it has been thought best to do so, the passage has been quoted verbatim.

Page 4

be called rules ; the commentary is, of course, more like an ordinary text.
In this paper the attempt has been made not only to present a correct version of the paricista, so far as the material at hand would allow, but also to reproduce to some extent the scholiast by citing, mostly from the first division, such passages, with the text, as bear on the same part of the rite, and by incorporating into the commentary accompanying the translation such other passages as throw light upon those already cited, or give an idea of additional matters not treated of in the text at all. In this way most of the salient points of S have been preserved without, at the same time,
copying its tiresome minuteness of detail and unending repetitions- not that the commentary is of so much importance in itself, for, as has been shown, it represents the work of a person of little intelligence apparently, certainly of small acquirements; but that the picture of the whole might be as complete as possible. The practice of witchcraft forms a dark chapter in the history of mankind, and anything that throws light upon the attitude of mind in which its devotees have practiced their curious rites is not to be despised. The "meditations" of S may not be without their suggestions
to those who care to read between the lines, and the whole practice is a curious bit of evidence of the power of superstition over the human mind.
While the Asuri-Kalpa has proved a rich field for emendation, and has afforded some opportunity for conjecture, it has not been altogether unfruitful in new material, as the following list will show.


Denominative Verb: pistaya, to grind up, make into meal.
Nouns (members of compounds): nastika [nasti], destruction.
ravi, 1, a tree or plant of some kind, sadi (not in a comp.), a collection of six. sruca (?) [sruc],. sacrifice-ladle.
Adjectives: pretaka [preta] belonging to a dead [man].
Possibly (?) jigaisa, desiring to conquer.
Particles: klnn, ksaum, and crim. 2
Analogical Vocative: duhite [duhitar] O daughter.
New Meanings or Uses: sureevari (compound stem), asuri (plant and probably also goddess). So laksmi, apparently and
possibly cri. caturtham (?), fourthly (as adverb).

1 See page 25, foot-note 4.
2 Evidently from cri 'beauty, welfare. 1 These words are used as part of a muttered spell, and have, therefore, no particular meaning.


Nouns: aprajatva, childlessness, utkarana, overcoming (?).
Adjectives: daksinakarnika, having its point (ear) to the south, devija, goddess-born, raktavasasa, having a reddish
garment, vacyaga, subdued. Possibly pratyamukha, facing.
Neuters as Adverbs: dinatrayam, at the three parts of the day (A. M., M., and P. M.) dinastakam, at the eight parts
(watches) of the day. Possibly (?) saptahanam, at the seventh dawn.


aparajaya, invincibleness. karmakarika (fern, of adj. -raka), deed-performer, nagendra, a plant, probably Betel. vacikartukama, the desire to render submissive. Possibly also surati, a plant of some kind.
A few words have as yet baffled all attempts at a solution. They will be mentioned as they occur.
That the Asuri-Kalpa must at one time have occupied a position of some importance appears from the fact that it is mentioned, according to Weber, Ind. Stud. XIII 415, under the name Asuriyah Kalpah in the Mahabhasya IV i, 19, Varttikam f. 19b. In this connection it may be added that the conjecture offered by Professor Bloomfield (J. A. O. S. XI 378) : ''panicakalpi is probably not to be understood (with Weber, Ind. Stud. XIII 455) as one studying five different kalpas, i. e. crauta-sutras, but means an Atharvavedin who is familiar with these five kalpas," i. e. the five belonging to the AV., has recently been confirmed by the discovery, made by the same scholar, of the word pancakalpi {stem -in) used in the colophon of a Kauc. MS 1 to mean the writer of a Kauc. MS. In connection with pancakalpah, says Weber (loc. cit.), the Mahabhas ya (Vartt. 3f. 67a) mentions the words kalpasutrah, paracarakalpikah, and matrkalpikah.
This last word Weber does not attempt to define, but says of it: " Letzleres Wort ist in der vorliegenden Beziehung unklar."
In the Kaucika-Sutra, 8, 24, is mentioned a gana of hymns (AV. II 2, VI III, and VIII 6) under the title matrnamani, the
object of which is the preventing or removing of evil; and Atharva-Paricista 2
34, 4, mentions the same gana with the

1 No. 86. Report on the Search for Sanskrit MSS in the Bombay Presidency, 1880-81, by F. Kielhorn.
2 A. No. 32, S No. 34. The latter numbering makes the Asuri-Kalpa No. 37 ; for each MS gives between it and the
Ganama1a two other paricistas-

addition of AV. IV 20, under the same name. 1 It also adds, iti matrgana. 3 As kalpasutrah means one familiar with the Kalpasutras, and paragaraka/pikah seems to have been used of a person who had studied the Paracara-Kalpa, 3 it is safe to infer that the word matrkalpikah meant one who was familiar with or made use of the Matr-Kalpa, and such a text may yet be found. If it ever appears, Professor Bloomfield conjectures that it will prove to be a ritual for the use of a priest in connection with this Matrgana. The presence of these words in the Mahabhasya, which contains many Atharvanic words not found elsewhere, cited as they are without explanation, goes to show that they were all well understood by the people of Patanijali's time, and therefore referred to rites and practices so familiar to the Hindoos that the mere name was sufficient to make the reader understand the author's meaning. As they are all Atharvanic,
and the word Asuri-Kalpah is also Atharvanic, jhere can be no doubt that the Asuri-Kalpah and the Asuilyah Ka1pah are essentially the same, though the text may have suffered some changes at the hands of later authorities on the
uses of a s u r i, and it is evident that the parisista must have had considerable currency among those who made use of Atharvan rites. Additional evidence of the familiarity of the Hindoos with such practices is to be found in the Laws of Manu (XI 63). where the practice of witchcraft {abhicara) and of magic with roots (mulakarman) is mentioned in a list of secondary crimes (upapataka). This reference also makes clear the fact that such practices are old; for they must have been well established when the Manava-Dharmacastra took its present shape, and go back, therefore, in all probability, some hundreds of years before our era. On the other hand, it must be said that the MSS bear
marks of a late origin. S mentions the Hindoo trinity (brahmavisnuhara), contains the Buddhistical word hevara, uses the gen. for the loc. and ins., etc.; and all the MSS contain forms (transfers to the a-declension, etc.) due to analogy and not cited in any of the dictionaries, besides exhibiting in the subject-matter certain

the Mahabhiseka and the Anuloma-K alpa. B does not number the latter or the Asuri-Kalpa, but has after the Mahabhiseka what is evidently a corruption for JJ. The Peters. Lex., with A, makes the Anuloma-Kalpa No. 34. The numbering of B has been taken to correspond to Dr. Bloomfield's edition of the Kauc.
    1 Cf. Weber, Omina et Portenta, pp. 350-53-
    2 Not in A. or B but see Bloomfield, Kaug. S, 24, note 5.
    3 Cf. Weber, Ind. Stud. XIII 445.

tendencies which are recognized as modern. They are mentioned below.
The word asuri is the fern, of an adj. from asura "spirit, demon," and therefore means primarily, " belonging to, or having to do with, spirits or demons." Under the form asuri, the Peters. Lex. gives the meaning, schwarzer Senf, Sinapis ramosa Roxb. 1 From the evidence of the MSS, asuri must be a plant with a pungent leaf, and must bear fruit (phala) and flowers; moreover, a religious meditation (dhyana) of S, which can hardly refer to anything else, speaks of the
"bright four-sided granter of wishes"; then of the same as "reddish," "blue-colored," "having a sword in the hand," "having a hook in the hand," " having a ' redstone' in the hand," etc. All these expressions are based upon characteristics of the plant, as will appear below. In describing the oblation the paricista says: 'The wise man should make meal of rajika' (rajikam pistayed budhah), while S in the same passage speaks of asuri as made into meal. The word rajika, in fact, occurs in S only in part third, never in connection with asuri, and always where the latter might be expected. The same is true of the word rajasarsapa, for example, -
vidhane purvavat karmapratimam rajasarsapaih, purvavat karayen nyasam, chedayet purvavad api.
' In [his] preparation, as before, [one should cause] an image for the rite [to be made] with black mustard seeds. As in the former case, he should cause the [limb]-placing ceremony to be performed ; he should cause [the image] to be chopped also as before.' The word rajika, which was left untranslated above, is the common name for the Black Mustard of India. This plant has bright yellow flowers, and bears small dark seeds contained in a pod which is tipped by a long, straight, flattened, and seedless beak. 2 In all members of the Mustard Family, the pungency pervades the entire plant. 3
There can be no doubt that this was the plant actually used, and it is plain that the ignorant and superstitious devotee saw a goddess in the plant itself, 4 and found,

1 Wm. Roxburg, Flora Indie a, Semapore, 1832.
2 Hooker, Flora of British India, I 157. The Black Mustard of Europe, which is closely related, is described as having smooth erect pods which are somewhat four-sided and tipped with a sword-shaped style. They contain small dark brown or nearly black seeds. The Black Mustard of the U. S. is similar.
3 Gray, Introduction to Structural and Systematic Botany, and Vegetable Physiology. 1873, p. 389 f.
4 Cf. the frequent similar personifications of the AV.

perhaps, in the effect of the seeds upon his palate an evidence of
her supernatural power.' The " red-stone " (j-udhird) mentioned
above, and defined by the dictionaries as a certain red stone, not
a ruby, here plainly means the seeds in the pod of the asurlplant,
while the pod itself is probably the "sword," and possibly
also the " hook."
The chief object to be attained was the subduing of another to
one's will, or the destruction of an enemy. The use of the hymns
of the AV. for the latter purpose is sanctioned by the Laws of
Manu (XI 33) : 'With the thought 'one should utter (perform)
the hymns of the Atharva-Veda,' [let him be] without hesitation
the ' word ' is the Brahman's weapon, you know, with it the twiceborn
should smite [his] enemies,'

gruiir atharvdiigirasili kiirydd ity avicdrayan,
vdk gastraih vdi brdhnianasya iena hanydd arm dvijdh. jj.
The other practice, as has been stated, is pronounced criminal by
the same authority. The rite itself is briefly as follows : after
certain introductory ceremonies, the person grinds up mustard
into meal, with which he makes an image representing the person
whom he desires to overcome or destroy. Having muttered certain
spells to give efficiency to the rite, he chops up the image,
anoints it with ghee (melted butter), curds, or some similar substance,
and finally burns it in a " sacred-fire-pot." The idea that
an image thus destroyed accomplishes the destruction of the
person represented, or at least does him serious harm, still survives
i n India, and it can be duplicated in almost any country in which
witchcraft has been practiced. The Samavidhana-Brahmana
contains a similar practice, in which an image of dough is roasted
so as to cause the moisture to exude, and it is then cut to pieces
and eaten by the sorcerer. An image of wax has been largely
used in various countries, the life of the enemy represented having
been supposed to waste away as the wax gradually melted over
a slow fire. This process was known to the Greeks, to the
Romans, to the Germans, and even to the Chaldeans." A vari-

1 This may also account for the name, since at the time when these practices
originated the Hindoos were both very superstitious and extremely unscientific
in all matters pertaining to natural phenomena, and they would, therefore,
quite naturally assign the pungency of the plant to some spirit or demon.
^Cf.Theocr. Idyll II 28, Hon Epod. XVII 76 ; Grimm, Deutsche Mythologie,
1047 ff. ; Lenormant, Chaldean Magic, p. 5, foot-note i, and p. 63;
Burnell, Samavidhana-Brahmana, Vol. I, Introd. p. xxv, and see p. 26,
foot-note I, end.

A variation of the same performance is to fill the image with pins, attach
a hated name to it, and set it away to melt or dry up according to
the material used. This is said to be still practiced in some parts
of America, England and the Continent.' It is reported ^that a
practice of this kind, i. e. the making of an effigy to be used for
his destruction by means of sorcery, was tried on Henry VI of
England ; and early in the present century a similar trick was
used against the Nizam of the Deccan.^ Among the Indians of
our own country, the Ojibway sorcerers were supposed to be able
to transfer a disease from one person to another by a somewhat
similar process. They were accustomed to make, for the patient
who paid them, a small wooden image representing his enemy ;
then, piercing the heart of this image, they put in small powders,
and pretended by this means, with the help of certain incantations,
to accomplish the desired end.' The fact that an image has been
so universally used in witchcraft practices is no more remarkable
than the fact that all nations have made use of images to represent
their gods in religious worship, and the two things may both
be referred to some law of the human mind by which similar
conditions produce similar results. There is no discoverable connection
between the Ojibway's wooden image and the Hindoo's
effigy of dough other than the mere fact that each is the outcome
of a desire to injure, and nature teaches them both to think of
what is practically the same expedient.
The minor practices of the Asurl-Kalpa, which are designed
either to work harm to an enemy or good to the practitioner, will
be found in their turn below. They seem to indicate a desire on
the part of the author to furnish a short cut to power and to some
of the more important blessings which were supposed to be gained
by the sacrifices prescribed by the Brahmanas; indeed, the
practices of the Asurl-Kalpa, as a whole, seem to show a disposition
to supplant certain religious forms by simpler magical rites,
while endeavoring at the same time to obtain powers for harm
which religious practices either left in the hands of the educated
Brahmans or did not bestow at all. It must be added, however,
that the belief in the efficacy of repetition, so conspicuous in the
modern " prayer-mills " of Thibet, is here plainly to be seen. In
the Asurl-Kalpa, as in all other Indian witchcraft practices,
there is, of course, an underlying stratum of skepticism ; but the

1 Conway, Demonology and Devil-Lore, Vol. I, p. 272.
2 Lyall, Asiatic Studies, p. 88.
3 Dorman, Origin of Primitive Superstitions, p. 361.

great power of the priests is tacitly recognized by the care
enjoined upon one who undertakes to subdue a Brahman. The
practices for obtaining blessings are confined to the latter part of
the pari fist a,' and, from their general character, seem like an
extension of the original practices, perhaps for the purpose of
giving additional currency or respectability to the whole; they
may possibly be regarded as a further indication that the Asurl-
K a 1 p a, however ancient its main practices may be, is, in its present
shape, comparatively modern.
At the present time in America, the interest felt in witchcraft is
shown by our surprisingly large and growing literature on the
subject.^ In India the interest felt is of a different nature, but it
is none the less strong. To the Hindoo the subject is a living one,
and while the native literature referring to magic and superstition
has always been great, at present, especially in the vernacular
dialects, it is enormous, and forms the favorite reading of the
people.^ So great is its hold upon the natives that Lyall says of
it :* " It is probable that in no other time or country has witchcraft
ever been so comfortably practiced as it is now in India under
British rule" ;' again, " in India everyone believes in witchcraft as
a fact"; and just below, " In every village of Central India they
keep a hereditary servant whose profession it is to ward off
impending hailstorms by incantations, by consulting the motion of
water in certain pots, and by dancing about with a sword."
Beside this may be placed the statement of Conway,* that there
are 84,000 charms to produce evil made use of in Ceylon at the
present time. In so far as it throws light on the past history of
such practices, the work on the Asurl-Kalpa may not have
been in vain.
'Both MSS recognize a division of the practices into groups

A. into two,
as shown by the figures (/ and 2), and JB apparently into three ; for it has a two
{2) where A has one (/), and what may be a one (/) in the passage which it
alone contains. It lacks the number at the end. The divisions of A have
been marked in Roman numerals, since it has been thought best to number
the 9 1 okas, although the MSS do not do so. The practices of the second
division are all of the same general nature.
-See Poole's Index, third edition, iS32, under the headings Witchcraft,
Demonology, Magic, etc.
•^ Burnell, Sam a vidh ana-Brahman a, I, p. xxv.
* Asiatic Studies, 1882, p. 96.
^"Of course the witch is punished when he takes to poisoning or pure
swindling " (1 o c. c i t.)
'Demonology and Devil-Lore, I 274.

A Witchcraft Practice of the Atharva-Veda -

Hindu Section

Main Library