Novum Testamentum XXIII, 3 (1981)

RECONSTRUCTING THE OPHITE DIAGRAM

BY

A. J. WELBURN
New College, Oxford

The so-called 'Diagram of the Ophiani' known to us through

Origen (contra Celsum VI, 24-38) has for many years been cited
by scholars as the classic account of the Gnostic-initiatory ascent
through the spheres. Considerable attention has been lavished
on various of its features-in particular the formulae by which
the archons are to be vanquished-and the passage has several
times been translated. It would therefore seem regrettable if,
as G. BoRNKAMM concluded, it is impossible from the description
to gain a totally clear picture of the Diagram as a whole 1) . It is
true that the problem is aggravated by our having two descriptions,
one from Celsus and the other by Origen, of what may not be exactly
the same figure, confusingly compressed into a single passage 2 ).
A more or less acceptable division of the material was proposed
by R. M. GRANT in his 'split translation' 3) . But it is clear that
if the systems are not absolutely identical, the implied theology
of the Diagram matches quite acceptably the doctrines of the
Ophites reported by Irenaeus (Adv. Haer. I, 30), making it fairly
certain that Origen has rightly identified the group which used it 4).
I am also inclined to accept that our descriptions relate to two
copies of the same drawing, with only minor discrepancies between
them. Proceeding on similar assumptions, an attempt was made
(some considerable time ago now) by T. HoPFNER to reconstruct
the Diagram from Origen and certain parallel ideas from elsewhere 5) .

1) BoRNKAMM, art. 'Ophiten' in Pauly-Wissowa XVIII, i, 654- (657).

2) E. DE FAYE, Gnostiques etGnosticisme (Paris 1925), p. 359.
3) GRAN T, Gnosticism. An Anthology (London 1961), pp. 89-92.
4) Cf. R. MeL. WILSON, The Gnostic Problem (London 1958), p. 121.
6) HoPFNER, 'Das Diagramm der Ophiten' in Charisteria Alois Rzach
(Reichenberg 1930) pp. 86-98. He derived some of his ideas from LEISEGANG,
who in the third edition of his Die Gnosis (Freiburg 1941) revised his own
partial reconstruction in the light of HoFFNER's article--see his p. 170 n. 3·
I do not feel called upon to discuss the earlier efforts of J. MATTER in his
Histoire Critique du Gnosticisme (1843), or GIRAUD in his 'Dissertation on
262 A. J. WELBURN

His reconstruction was apparently accepted by H. CHADWICK,

who included it in his translation 6) ; nor have more recent scholars
questioned it 7 ) . It is certainly valuable as a pioneering attempt,
but it lacks what seem to me some of the essential features of a
'mystic diagram'. Such figures undertake to manifest to the
initiated the truths and interconnections of the spiritual worlds
through mathematical-geometrical proportions. Ideally, a kab­
balistic diagram (in the widest sense) shows an utterly interdepen­
dent system of structural ratios, where no single measurement
is left to chance. In contrast HOPFNER's version often shows a
disquieting arbitrariness : there is no necessary connection between
the sizes of the larger and smaller circles, and the latter are disposed
with an unpleasing asymmetry. Moreover his drawing does not
always fit the descriptive text very closely, and he plainly failed
to recognize allusions to a number of figures familiar to any reader
of the somewhat later kabbalistic works proper. It seems to me
worthwhile, therefore, to make another attempt at reproducing
the Diagram, especially since we may call upon some of the recently
recovered Sethian writings from Nag Hammadi for parallels and
confirmation at some points, notably when we come to discuss the
implicit system or theology behind its network of circles and lines.

I. The Planetary Angels and the lfayyoth ha-Kodesh

It will prove most convenient to begin at the beginning of
Origen's own description, and consider some of the details before
enquiring into the general shape of the Diagram.
He says (VI, 25): tv cJ> flv 8Locypocc:p� xu>tA(J)v, i.e. it included a
drawing of these circles, not that it consisted of them 8 ) . As HOPFNER
rightly saw, they make up only one area of the Diagram ; having
once seen this, I do not fully appreciate the motive behind his
additional 'planimetric' drawing in which the whole Diagram is
demonstrated revolving around the earth. There is no evidence
the Ophites' (1884), as their positive results were taken up into later attempts,
culminating in HoPFNER's complete reconstruction, along with some of
their errors.
6 ) CHADWICK, Origen: Contra Celsum (Cambridge 1953), pp. 337-9. He
also italicizes Celsus' material-d. HoFFNER, pp. 86-7 . For the text I have
used the GCS version, ed. KoETSCHAU (Leipzig 1899).
7) It is cited e.g. by W. FAUTH in his thorough study of 'Seth-Typhon,
Onoel, und der eselskopfige Sabaoth', Oriens Christianus 57 ( 1 973), 79-1 20
(97).
8 ) As it is taken in GRANT, Anthology p. 85.
 RECONSTRUCTING THE OPHITE DIAGRAM

that the Diagram existed in any such additional form, or that the
realms outside the cosmos were conceived to orbit in this way.
It is, I think, a remnant of the preconception that the entire
Diagram represents only the cosmic spheres and their rulers 9) .
It i s of course perfectly true that within the area of the Diagram
which stands for the cosmos there are circles, and these obviously
represent the planetary spheres.
We possess immediately two different lists of the seven rulers or
archons of the planets-one from Celsus, into which Origen has
inserted the actual names used by the Diagram he himself saw
(VI, 30) , and a different one from Origen alone, giving the Rufnamen
(HoPFNER) by which the initiated addressed them (VI, 31). These
lists are totally distinct. Since they undoubtedly refer alike to the
planetary spheres, however, we may list them together :
Celsus Name supplied by Origen Rujnamen, Planet
Lion -form Michael Horaios, Moon
Bull-form Souriel Ailoaios, Mercury
Snake-form Raphael Astaphaios, Venus
Eagle -form Gabriel Adonaios, Sun
Bear-form Thauthabaoth Sabaoth, Mars
Dog-form Erathaoth lao, Jupiter
Ass-form Thaphabaoth, Onoel Ialdabaoth, Saturn
or Thartharaoth

The planetary correspondences to Origen's series of Rujnamen

may be worked out with considerable certainty through comparison
with the parallel terminology of the Barbelognostics 10) . But
it will be noticed that I have reversed HOPFNER's order for the
Rufnamen against the names for Celsus' archons 11) . This is because
Origen apparently found it easier reading from the periphery
inwards when he came to transcribe the formulae for the several
spheres. Perhaps the Celsus-sequence was written in the lower

9 ) HOFFNER, p. 95; cf. LEISEGANG, opp . p. 160.

10) See my 'The Identity of the Archons in the Apocryphon Johannis',
Vig. Chr. 32 (1978), 241-54. GRANT, Gnosticism and Early Christianity
(London and New York 1966) gives an extremely confused and confusing
comparison of the two systems. His proposal that we think in terms of
planetary weekdays (p. 51) leads nowhere-though on this tack he was
anticipated by Celsus (VI, 22), who also fails to differentiate between the
Mithraic escape back through the time-cycle of the week and the Gnostic
planetary ascent. Cf. HoFFNER, p. 86.
11 ) HoFFNER, p. 88. I have also transposed the Sun to its natural place
in the sequence (this minor adjustment already in LEISEGANG, p. 172,
though he retains the basic and erroneous order).
 A. J. WELBURN

half of the cosmos with Saturn-Thaphabaoth at the bottom of

the list, and the Rufnamen in the top half, with Saturn at the
top of the list. (Origen, of course, nowhere claims to be giving
the sequences in the same planetary order). The formulae are
in themselves evidence against HOPFNER's view, since Horaios
is invoked as the one 'having authority over the first gate' , Ailoaios
over the second, etc. , in consistent order out to Ialdabaoth : the
numbers are not even thrown out by the accidental omission of
Adonaios, showing them to be an integral part of the formula
in each case. Yet HoPFNER might still be right if the initiates
were involved in a process of symbolic descent, so that the 'first
gate' meant the outermost sphere (Saturn). But it is hardly probable
that the Ophite mysteries were based on the idea of initiatory
descent 12). The longest of the invocations, too, that of Ialdabaoth,
makes better sense as the climax of the sequence rather than its
beginning. 'I pass by thy power free again', and the mention of
the Son and the Father hardly make any sense at all here except
on the supposition that the initiate is not entering Ialdabaoth's
sphere but leaving it, triumphantly reaching the freedom of the
divine realms beyond cosmic tyranny. It seems safe to conclude,
in other words, that Origen has transmitted the list in reverse
order, and that it and its formulae were understood by the Ophites
in terms of an initiatory ascent.
This obviously invalidates HoPFNER's correspondences for
Celsus' list of archons too 13). Michael must go to the lowest sphere
rather than to Saturn, Souriel to Mercury, Raphael to Venus,
Gabriel to the Sun, etc. HoPFNER based his system on the 'sympathy'
between the Lion-formed archon and Phainon-Saturn (VI, 31) 14),
arguing that Michael must therefore be the Saturn archon. Ial-

12) The Marcosians claimed that their Gnostic redemption 'leads them
down into the profundities of Bythos' (Irenaeus, Adv. Haer. I, 2 1, 2 ) . The
Nag Hammadi Valentinian Exposition speaks paradoxically of 'the descent
which is [the upward progression], that [is, our exodus] from the world
[into] the Aeon' (CG XI 41, 35-8) . In Jewish mysticism we soon come to the
'descenders to the Merkabah' , but the origin of this emphasis on descent
remains unexplained. See ScHOLEM, Major Trends (New York 1961) , pp. 46-7.
GRANT has connected Gnostic descents with Mesopotamian myth, Gnosticism
and Early Christianity, pp . 81 ff.
19 ) It could be argued that Celsus' lists may have been read in the same
direction as Origen's-but we shall shortly adduce a parallel that fixes
the lower four spheres, thus ruling this out.
14) HO FFNER, p. 88.
 RECONSTRUCTING THE OPHITE DIAGRAM

dabaoth is in fact frequently represented with a lions's head 15).

But this need not prove a stumbling-block to our scheme, since
Ialdabaoth has just before been invoked mysteriously as the one
'to whom power belongs as first and seventh' . This surely does not
mean simplistically that Saturn is the first sphere for those going
down and the seventh for those going up. Saturn appears to have
been associated by the Gnostics with the Jewish God, who figures
here as the chief of the archons (VI, 27 ; see Adv. Haer. I, 30, 10),
because the seventh day, the Sabbath or Saturday, was sacred to
him. No doubt, however, they were also aware that the seven-day
framework which made this possible testifies to a lunar calendar.
Yahweh is therefore to be seen in the broader sense as a Moon
deity 16). A more intriguingly precise connection between Moon
and Saturn, first and seventh, will also have conspired to produce
this conception, namely that the synodic month (reckoned, that is,
according to the complete cycle of phases of the Moon) has a
duration of 29t days, while Saturn completes a circuit through
the zodiac in very nearly 29! years 17 ) . It therefore made good
astronomical sense to connect the powers behind these two planets.
The Jewish God also fits the animal-mask of the seventh archon
found in Celsus, since he was represented polemically with an
ass's head. We may assume that the (obviously secondary) name
Onoel stands for the Greek �vo� with the Hebrew divine suffix -el
attached in the syncretistic manner 18 ) . Nothing really prevents us,
therefore, from assigning Michael to the Moon and Onoel to Saturn
as the Diagram's system appears to suggest; the gloss by Origen
on the 'sympathy' between Saturn and the lion-archon can then
be read precisely as an explanation of the 'first and seventh'
formula he has just quoted.
We are assisted in our further attempts at identification if we
next observe that the names of the first four archons (Michael,
Souriel, Raphael, Gabriel) are those of the traditional Jewish

16) B. GXRTNER, Theology of the Gospel of Thomas (London I96I) , pp.

I62 ff. conveniently collects some of this imagery.
18) Cf. s. PETREMENT, 'Le mythe des sept Archontes createurs', in u.
BIANCHI, Le Origini dello Gnosticismo (Leiden I967) , p. 481.
17) J. MEEKS, 'Archangel Periods', in Golden Blade 3I (I979) , I9-32 (29 f. ) .
Gnostic material in Irenaeus, Adv. Haer. I, I7, I.
18) Rightly rendered by HoPFNER, p. 87, as 'Eselsgott'; the possible
background in the cult-mysteries of the Egyptian god Seth is discussed
in FAUTH, art. cit., esp. pp. 99-IOO.
z66 A. J. WELBURN

archangels, and that their animal-forms are those of the lf.ayyoth

ha-Kodesh, the four 'Living Creatures' from Ezekiel's vision.
The only exception is that 'Snake' has crept in to replace the
more usual 'Man' , but in an Ophitic system this need not amaze
us too greatly 19) . These Creatures have no connections initially
with the seven planets, but rather with certain of the animal-signs
or zodiac. There they pick out the 'fixed cross' (one of the astrological
quadruplicities) :
Lion (Leo )
Michael
Bull (Taurus ) Eagle (Scorpio )
Souriel Gabriel
Man (Aquarius)
Raphael

That these signs were linked with the names of the archangels
in other Gnostic circles is perhaps confirmed by the fragment of
the system preserved from the book Baruch, to which HoPFNER
already referred 20 ) , where a zodiacal list of the 'Angels of Elohim'
is given. It includes two archangelic names, and if Baruch was
using the same terminology as the Ophites could be convincingly
reconstructed as follows : Michael (Leo), Amen (Virgo), Baruch
(Libra), Gabriel (Scorpio), Essadaios (Sagittarius), [ ] (Capricorn),
[Raphael] (Aquarius), [ ] (Pisces), [ ] (Aries), [Souriel] (Taurus).
Commencing in this way at Leo was in fact one of the standard
ways in Antiquity of cataloguing the zodiac. We may take it
that the feasibility of translating the Baruch-Ophite system into
good zodiacal terms is significant, therefore, and that in both the
four archangels were probably brought into connection with
the fixed signs 2 1) . There are other evidences of Gnostic interest in
the Ifayyoth elsewhere 22 ) .
19 ) On a similar motif in Baruch-Gnosis and its possible background
in the Isis-mysteries, cf. R. VAN DEN BROEK, 'The Shape of Edem according
to Justin the Gnostic' , Vig. Chr. 27 ( 1973) , 35-45·
20 ) HoPFNER, pp. 91 ff. Text in Hippolytus, Ref. V, 24-6.
21 ) At least it is unlikely that Baruch's Michael and Gabriel establish
a pattern which was not carried on in the missing part of the text. Certainly
not all archangel tetrads can be accomodated in this way: the traditional
Talmudic deployment will not fit at all (Num. R. II, 10). Compare, though,
the arrangement in the Qumran War Rule IX.
22 ) The Manichaean King of Darkness appears in Keph. 30 f. as a parody
of this angelic tetrad with his lion's face, his eagle-winged shoulders, his
dragon-belly (replacing the bull) and his fish's tail (replacing the land­
swimming snake for Aquarius) . Manichaean need for fivefold systems added
 RECONSTRUCTING THE OPHITE DIAGRAM

The Gospel of the Egyptians seems to follow to a certain extent

in the same tradition; but closer examination reveals a different
determining factor. We can recognize in the 'ministering angels' of
the Four Luminaries, whom I suspect are descendants of the
lfayyoth, a Gnostic version of the four archangels, here called:
Gamaliel, Gabriel, Samlo and Abrasax (CG III 52, 19 ff.). The
presence of Abrasax, whose name adds up by gematria to 365,
makes it plain that we have to do with the cycle of the year.
GRANT has already drawn attention to the quaternary 'Abraxas
Soter Bel lao' , who appear to be the rulers of the four seasons
elsewhere 2 3} . But he despairs rather prematurely of finding coher­
ence. In fact the list can be worked out quite straightforwardly:
the Luminaries come invariably in the order given by the Gospel
of the Egyptians here, and their 'ministers' may thus be read round
in the order of the seasons. Harmozel and Gamaliel ('benefit of
God') will have Spring; Oroiael and Gabriel will have Summer-an
Oriares ('light of the sun') figures in a different seasonal system
in Enoch 24) ; Davithe and Samlo (the equivalent in this system
of Soter?) will have Autumn; and Eleleth and Abrasax will have
Winter-naturally, since Abrasax refers numerically to the com­
pletion of the annual cycle at the winter solstice, when the death­
birth of the god Aion was celebrated. However, none of these
correspondences involves the symbolism of the Holy Living
Creatures. The archangels are connected directly with the seasons,
and these have only a secondary link with the signs making up the
fixed cross, which means simply that it is also spring when the
Sun is in Taurus, summer when the Sun is in Leo, and so on.
The system of the Gospel does not identify the archangel-ministers
with the signs, but correlates the independent zodiacal and
seasonal cycles. It is evidently requisite we understand that
archangels may be seen as fulfilling duties in the cosmic government
on more than one level. And when these various understandings

demon-hands and feet. The list appears in the Chinese literature, and lies
behind the confused report in Augustine, de haeresibus ad Quodvultdeum,
46, 2; and the more orderly de moribus Manichaeorum, q.
23) GRANT, Gnosticism and Early Christianity p. 46 {and see generally
pp. 42 ff.). I would not connect the Luminaries so directly with the archangels
-1 do not think, for instance, that Oroiael is 'obviously' the same as
(S) Ouriel. The tetrad cited by GRANT is restored from Hyginus, Fabulae
(ed. RosE, Leiden 1 933) CLXXXIII, 3 ·
24) GRANT, op. cit., p. 43·
268 A. J. WELBURN

of their functions overlap, it may well be that we will find certain

of their attributes transferred from their original sphere to a new
one.
This is an important conclusion as regards our understanding
of the Ophite Diagram. If in one context the archangels could
be seen as representing the seasons, while in another they are
assigned independently to zodiacal signs, then we might expect
to find them working in other ways too, and perhaps taking with
them their seasonal or zodiacal associations. It is in some such
way alone that I can explain the application of the four lfayyoth
ha-Kodesh to the first four planetary spheres, as indicated by
Celsus. For it is well-known that the archangels could on occasion
be multiplied to a planetary seven. HoPFNER cites one such Talmudic
scheme-though it is not one which throws light on our problem 26}.
We may tentatively propose as our solution that the four primary
archangels, who seem in Baruch and (indirectly} in The Gospel
of the Egyptians to have been linked with the lfayyoth or the
signs of the fixed cross, retained these associations when they
were in another context adopted into the sevenfold planetary list.
This might, of course, be accounted mere confusion unless we are
able to show that it was done with some consistency, in accordance
with some invariant principle. There is in fact a passage in Epi­
phanius (Panarion XXVI, ro, r} which furnishes some slight
evidence in the required direction. It is a highly confused and
distorted-and deliberately distorted-passage, so that it is not
even clear whether the heresiologist is continuing with his report
on the Phibionites or has passed to another group. He has success­
fully rendered the details useless for reconstruction of the original
doctrine, but merits our interest in mentioning 'in the third (heaven}
the archon Seth', while 'in the fourth . . . is Davides'. These are
names we now know to have been given not to planetary archons
but to the Luminaries : at least, Davides is clearly Daveithe, and
Oroiael is consistently said in the Coptic literature to be the abode
of the great Seth ( A pocryphon ]ohannis CG II g, II; Gospel of the
Egyptians CG III 65, r6}. Hence, unless Epiphanius has scrambled
his information completely, it appears that the names of the

25) HoPFNER, pp. 88-g n. II. On the seven heavens in Rabbinical teaching,
see STRACK-BILLERBECK, Kommentar III, p. 532 . A list of seven archangels
probably lies behind the passage in the long version of the Apocryphon
]ohannis (CG II 17, 29-32) which now tails off into nomina barbara.
 RECONSTRUCTING THE OPHITE DIAGRAM

Luminaries may have been linked by some Gnostics with the

lower four spheres. But this is very tenuous evidence. More striking
is the confirmation we receive from the canonical Apocalypse.
We are dealing there, it must be added, not with an ascent through
seven spheres, but with the revelatory opening of seven seals­
though it may plausibly be assumed that Gnostic readers could
have identified the two procedures. 'Seals' and 'sealing' had a
technical sense in the Mysteries, and indeed this symbolism is
featured in the system of the Ophites who used our Diagram
(VI, 27). In the Apocalypse these seals are not opened by the
initiate, however, but by the Lamb: he first appears 'in the midst
of the Throne and of the four Living Creatures' , and is also con­
nected with the 'Seven Spirits of God' (Rev. 5, 6). The lfayyoth
themselves had been introduced a little earlier (4, 7) :
The first Creature was like a Lion; the second Creature like a Calf; the
third Creature had a face like that of a Man; and the fourth Creature was
like a flying Eagle.

In this order, too, the Creatures participate somewhat obscurely

in the opening of the first four seals 26) . When the Lamb opens
the first seal, the first of the lfayyoth (i. e. the Lion-formed) suddenly
cries out with a voice of thunder, 'Come! ' At his invocation the
White Horse promptly appears. When the Lamb opens the second
seal, the second of the lfayyoth (i.e. the Bull-Calf) repeats the
command, and the Red Horse appears. At the opening of the
third seal, the third of the lfay yoth (i.e. the Man-faced) calls out,
summoning the Black Horse. And when the fourth seal is broken
the fourth of the lfayyoth (i. e. the Eagle) calls out 'Come!' and
the Green Horse appears to preserve the overall parallelism (Rev.
6, r-8). It can hardly escape our attention that the sequence of the
Creatures is that of the four lowest heavens according to Celsus'
description. The Ophites have further provided an analogous
series of animals (Bear, Dog, Ass) for the higher three spheres,
but the type of name alone would mark them off as distinct from
the classical four. Such a differentiation of the first four from the
following three is a characteristic of the structure of the Apo­
calypse 2 7) , and conjoined with the identical order of the lfayyoth
in the lower four spheres means either that the Ophites were
28} See A. FARRER, The Revelation of St. John the Divine (Oxford 1964),
pp. 97-8.
27) FARRER, pp. 8-9.
270 A. J. WELBURN

dependent on the canonical Apocalypse or, more probably, that

both were utilising a Jewish cosmological-apocalyptic schema.
Such a background link would tend to confirm not only our identifi­
cation of the order of Celsus• archons, but also the supposition that
the Luminaries seemingly placed by Epiphanius in the lower
heavens are indeed ultimately representatives of the four lfayyoth.
Their derivation from esoteric Judaism would not, of course,
rule out the influence of Iranian-Chaldaean ideas, as GRANT rightly
concluded, either at an earlier or a later (i.e. already Gnostic)
stage 28) .
Having identified the seven archons according to their various
names and shapes, and in some measure shown how in some
esoteric tradition the features of the Living Creatures had been
transferred to another area of thought, we pass to further questions.
Are we, for instance, able to specify the reason the Diagram gives
two distinct lists of archons in the first place ? A parallel instance
is possibly to be found in the Apocryphon of fohn which, as is
well-known 29) , gives two sets of names to the planetary rulers.
One series denotes the archons according to their own natures-and
these correspond closely with the Ophite Rujnamen. The other
series contains those 'given them according to the glory which
belongs to heaven for the destruction of the powers . .. Thus they
have two names' (CG II 12, 27 ff.). We might detect a similar
duality in Origen's formulae : they normally begin with an invoca­
tion to the archon, after which the initiate calls upon the power
of some contrary agent. At the fifth gate, for example, he confronts
the fivefold power of the archon, and conj ures him 'by a mightier
pentad, let me pass, since you see the symbol which your art
cannot attain' (VI, 31). Perhaps these other symbols were actually
depicted in the Diagram amongst the paraphernalia neither Celsus
nor Origen thought worthy of attention; perhaps too, they had
some connection with the alternative set of names inscribed among
the planetary spheres. We have already seen that some of them
were representatives of the Luminaries, who belong to the supra­
cosmic realms. A duality of powers in each heaven and permeating
the cosmos generally would also explain the remark in Irenaeus
(Adv. Haer. I, 30, g): the Ophites claimed that 'hebdomad' means
2 8) GRANT, op. cit., p. 46.
2 9) Having early been discussed by J. DoRESSE, Secret Books (New York
1960) , pp. 202 ff.; see in particular the magical text, p. 203.
 RECONSTRUCTING THE OPHITE DIAGRAM 271

'the seven stars which are called planets, and that the object­
ionable Snake has two names, Michael and Samael. ' It is tempting
here to invoke the Apocalypse once more, and relate this image
to that of Michael contending with the Dragon and casting him
down into the world (Rev. 12, 7 ; and cf. Adv. Haer. I, 30, 8) .
From the Gnostic standpoint these heavenly opponents would
then be seen remarkably fused into the positive and negative
aspects of a single state of existence 30) !
Beyond all the seven archons, we are told, was an all-circum ­
scribing circle ( ouv8ou�vCilv 8e U(jl' E:voc:; xuxt-ou), which is said to
have been named Leviathan or 'soul of the universe' . Beneath all
the heavens was written the name 'Behemoth' (�e:"Y)!Lwv) (VI, 25).
These complicating factors should deter us from prematurely
amending the text, which counts the circles as ten, to read 'seven',
as has been widely done since LIPSIUS on the grounds that there
are only seven planets 31). I have urged elsewhere that the 'ten'
spheres are only one point of contact with incipient kabbalistic
doctrines 32 ). At the same time it is obvious that only seven can
bear planetary names. To this problem we shall return in section IV.
We can deduce from the context that certain other details
recorded by Celsus-the -re:-rpocyCilvoc:; and the fiery sword, shown
as the diameter of a 'fiery circle' , the automatic gates, the multitudes
of the dying-also fell in the cosmos region of the Diagram (VI, 31;
34). But I do not believe we have enough evidence as to their
whereabouts or how they were represented to insert them into
our reconstruction. Certainly HoPFNER's reconstruction is here
almost whimsical : his 'square' has been awkwardly squashed into
a long oblong between the Leviathan-circle and a further circle
(supposed to be the 'fiery circle') added beyond by HoPFNER
himself. There is no support for this positioning in Origen. Moreover,
his 'fiery sword' is not a diameter at all-though the description
unmistakably insists that it should be. His basic notion of the
Paradise-square in the outer spaces of the cosmos has some

30) Although no such higher aspect is explicitly mentioned in Origen's

formulae, the recurring 'be with me' can hardly mean the archon in his
wicked nature/name.
31) LIPSIUS, •tl"ber die ophitischen Systeme' , Zeitschrift fur wissenschaftliche
Theologie 7 (1864), 37-57. Cf. CHADWICK, p. 340 n. 1; HoPFNER, p. 87;
FoERSTER, Gnosis I (Oxford 1972), p. 94 n. GRANT, Anthology p. 89 prints
both.
3 2 ) Art. cit., p. 245.
272 A. J. WELBURN

plausibility, though I do not actually think it correct. There

may be an allusion to 'squaring the circle', a persistent theme of
esoteric geometry, and the square may have been based on the
diameters of the seventh sphere or the Leviathan. More likely,
however, this is the ironic 'Paradise' of the Old Testament God
and the generated Adam which is described in several Coptic
works (Apocryphon ]ohannis CG II 21, r6 ff.; Hypostasis of the
Archons CG II 8 7, 23 ff.; Apocalypse of Adam CG V 64, 6 ff.; etc.).
This is a distinctly sublunar realm created by the archons (Hypo­
stasis 88, 24 ff.) for the Adam whom they have moulded from the
earth 33) . Moreover, in the Diagram we find that Horaeus-Moon
was addressed by the initiate as he who had passed 'beyond the
barrier of fire'. It may be that the Moon's orbit is the 'fiery circle',
above which sits enthroned its governing archon, and that Paradise
is the square constructed within the circle described. We should
have then a Paradise related to ancient views on the locale of
departed spirits, rather than to biblical geography. It is noteworthy
that the formula for Horaeus also contains references to the 'tree
of life' and 'innocence' (VI, 31). The 'fiery sword' may have been
the diameter defining the size of the square. But we cannot be
certain-though there is more evidence here than in HoPFNER.
More objectionable than the latter's guesswork, however, is his
surprisingly unscholarly importation of entities from a separate
Gnostic system: namely the trees of Baruch and Naas, the twelve
angels of Elohim and twelve of Edem from the book Baruch 34) .
I take it the reader will not be perplexed at their absence from my
reconstruction.

II. The Axe-Head; the Son and the Father

We turn now from the depths of the material cosmos to the
spiritual heights. According to Celsus the Ophians affirm 'that
33) Against this earthly location of Paradise stands Origin of the World
CG II 110, 2 ff. : but there the term has reverted to a positive meaning
through the influence of Manichaeism with its terra lucida. It envisages a
Paradise beyond the moon and sun.
34) HoFFNER, p. 94 (diagram) . The presence of Naas is not sufficient
to identify this work as Ophite, at least not as Ophite in the sense of the
exact group who devised our Diagram. As our angelic correspondences
hinted, these groups may have compatible systems-but this does not
warrant a promiscuous pooling of information in the reconstruction of the
Diagram's contents. The link with Baruch in this context goes back to
LIPSIUS, p. 39, though he does not conflate them.
 RECONSTRUCTING THE OPHITE DIAGRAM 273

among {!J.e-rcx�u) the supercelestial circles above {&vCil-repCil}, some

things are drawn, in particular two (circles), a larger and a smaller,
of the Son and of the Father' (VI, 38) . Origen adds that the labels
'Father' and 'Son' were actually inscribed on their diameters. Further
information is communicated rather obscurely in what follows :
And between the larger (circle ) , in which is the smaller, and another
( circle ) which consists of two circles, the outer yellow and the inner blue,
there is a partition (3�occppor.y!Lor.) shaped like a double-edged axe (n:ef..e:xoe�3e:i:
ax�!Lor."�).
HOPFNER quite rightly intuited that the 8�ocrppcx"(!LCX arises geo­
metrically out of the Diagram where its circles overlap. Yet his
' partition', whose shape is based on the interpretation advanced
by LIPSIUS 35) , would make a rather poor show if commandeered
as an axe. He had another, crowning intuition, which was that
three large circles make up the general morphology of the Diagram :
what he failed to do was bring these insights together in the right
way. Those who are at all conversant with later kabbalistic diagrams
will recognize instantly the figure which Origen is obviously
trying (with a great lack of perspicuity) to describe. It is one of the
constructions involving the celebrated vesica piscis-the fish-like
form constituted when two circles overlap and the centre of each
lies on the other's circumference. From this are derived a great
many of the 'mystical' properties of kabbalistic geometry. In
particular the vesica involves the ratio I: VJ, and so the relation
between simple unity and the irrational number which generates
out of itself the number 3· In the Diagram three circles are used,
giving two vesicae in the middle circle, whose remaining area
outside the vesicae then assumes the splendidly ceremonial shape
of the double-axe (see Fig. I) . For the sake of clarity I have blocked

Fig. I

86) LIPSIUS, p. 44·

18
274 A. J. WELBURN

in the axe-head shape, though it may not have been so blatantly

marked in the Diagram. Its name 'Partition' may well have been
written in along its narrow section, following the general pattern
in the Diagram. Attempts have been made to guess at the signifi­
cance of the axe-head 36) , and they may not be wholly wrong.
But it is the geometrical figures which generate it that have more
Gnostic significance. The presence of the vesica suggests there was
already a background of numerological speculations which emerge,
with a slightly altered character, in Marcus and the late Gnostic
literature such as Pistis Sophia. Other symbols, like the pentagram
already mentioned in the formula for Sabaoth, point in the same
direction 37 ) .
The centre circle necessary for the construction is described by
Origen only in its effects. The outer two which he does name are the
'larger' circles-the yellow one and that 'of the Father'. Our
reconstruction here is essentially very close to HoPFNER, whose
three circles have simply moved closer together. But it must again
be recorded that he strayed from exact consideration when he
came to more detailed reproduction. He showed the circle of the
Son, for instance, as a smaller concentric circle within that of the
Father. This would be plausible were it not that Origen says
expressly that their names were inscribed on the diameters: the
diameters of concentric circles either merge or cross, so that the
names on them would be confusing or messy. But there are other
ways a large circle can contain a smaller, and concentrism also
entails an arbitrariness in the relation between the two circles.
Whereas the intention was surely to show that the Son-circle
'proceeds' from the Father-circle by a mystic necessity. HoPFNER
gestures vaguely in this direction by joining the Son-circle to
one of the smaller ones-but there is still nothing to prevent
us altering both together without affecting the Diagram as a whole.
We can produce something with the required geometrical elegance,
while remaining in accord with Origen's words, if we suppose
instead that the large Father-circle was divided in half by a vertical
86) CHADWICK, p. 353 n. 2 has pointed to possible late-Hermetic symbolism.
LEGGE, Forerunners and Rivals of Christianity (Cambridge 191 5) II, p. 67
n. 3 suggested that it meant androgyny-and it is true that Ophitism knows
a 'male-female' Sophia (Adv. Haer. I, 30, 3).
37) On particularly the later developments, see (with caution) F. B.
·
BoND and T. S. LEE, Gematria (Oxford 1917) . They do not discuss the
numerology of our Diagram however.
 RECONSTRUCTING THE OPHITE DIAGRAM 2 75

diameter line, which forms a tangent on the large centre circle.

The name 'Father' could then have been written along the outer­
most side of this line as shown in the completed version (Fig. 3).
The Son-circle may be inserted inside the Father-circle, its diameter
being defined by the width of the vesica, and the name 'Son'
written on the diameter at right-angles to that of Father. We
shall adduce support from the probable theological import of the
Diagram in terms of the Ophite system later. Here we shall add
only that, as against HoPFNER's vertical disposition of the Diagram,
all we know of the Ophites (and much of what we know of Gnostics
in general) argues for a right-to-left orientation. In Irenaeus'
account the second Son (or Christ) is said to be 'on the right',
which is evidently also identified with the higher regions and the
direction of 'the Imperishable Aeon' (Adv. Haer. I, 30, 2); and the
Sophia or Prunicos is conversely called 'the Left' (I, 30, 3) 38) .

III. The Yellow Circle

Our efforts so far fall into two unconnected parts : the reconstruc­
tion of the sequence of cosmic circles within circles; and the super­
celestial circles of the Son and the Father, along with the yellow
(and the blue) circles. Our question is next, How are the two parts
related?
The crucial issue is the meaning of the yellow circle. What
domain of the universe does it represent? Here we may call upon
a piece of illuminating comparative material. In a neglected
article A. ALTMANN studied the evidence of Gnostic ideas in Talmu­
dic literature, and found a motif which plainly derives from some­
thing very like our Diagram 39 ) . 'An anonymous Baraita (b. Hag.
12a) , ' he says, 'defines Tohu (Gen. I, 2) as "a green line encom­
passing the whole world, from which darkness emanates." ' Tohu
wa bohu (traditionally rendered 'waste and void') describes the
dark, chaotic state of the as yet unformed world in Genesis. The
conception of a linear Tohu could further be j ustified by the
citation of Is. 34, II, which speaks of the 'line of tohu and the
plummet of bohu' : but there is actually more evidence that we

38 ) Also in VI, 27 the 'angels of light' probably stand on the right of

the dead soul, and the archontic angels on the left (parallels in Valentinian­
ism, e.g. Clement, Excerpta ex Theodoto 23, 3; 28; 34; 37; 47, 2).
39) ALTMANN, 'Gnostic Themes in Rabbinic Cosmology', in Essays in
Honour of ]. H. Hertz (London 1942) , p. 20.
 A. J. WELBURN

are here in the sphere of Gnostic teachings. The creation or emana­

tion of the cosmos was often understood in Gnosticism as beginning
from the outer spheres and proceeding inwards 40). The chaos
before creation would therefore be imagined lying at the periphery
of the present area of the cosmos, and this idea is almost certainly
detectable behind the name of the outermost sphere-Thaphabaoth.
GRANT conjectured that 'Thauthabaoth is a variant form of the
Tohu and Bohu which we meet in the Old Testament' 41) . And
Thaphabaoth is doubtless only an alternative attempt at the
decipherment of the same Hebrew words. Thoughts along these
lines seem to have induced M. JOEL to point 'to the Ophitic Diagram
. . . as the source of our Baraita. There the earth is surrounded
by seven circles of Archontes (Planetary Powers) and they in turn
are surrounded by a large yellow circle which bears the inscription
Leviathan' 42) . Now this is going just a little too fast. Certainly
the 'yellow' line could quite readily have become a 'green' line.
(If the Diagram of the Ophites was actually the source of the
notion, the copy seen by the anonymous author may simply
have faded since Origen saw his copy.) And certainly we can point
to signs of parallel speculations on Tohu and Bohu in the Diagram .
From this it seems safe to conclude that the yellow line must in
fact have encircled and contained the system of cosmic spheres.
Yet it would be surprising if Origen, having scrupulously told us
the positioning of the Son and the Father, neglected to mention
that the nearby yellow circle was the Leviathan he found so
disgusting a little while before (VI, 25) . An additional complication
is that there may really be ten circles to account for ; and if the
yellow circle were the Leviathan, with what could we identify
the smaller blue circle that must then lie somewhere inside the
cosmos?
For the moment therefore we must be content to place the
cosmos-section of the Diagram somewhere within the large yellow
circle ' encompassing the whole world' on the left of our reconstruc­
tion (Fig. r ) .
40 ) Another point of contact with Jewish mysticism and its doctrine
of 'contraction' (tsimtsum); see ScHOLEM, Majar Trends pp. 260 ff.
41 ) GRANT, Gnosticism and Early Christianity p. 47· FAUTH, p. 1 00 con­
jectured an alternative meaning for Thaphabaoth, deriving it from the
Aramaic for ausloschen.
42 ) ALTMANN, p. 20 citing M. JoEL, Blicke in die Religionsgeschichte I
(x88o) , p. 142.
 RECONSTRUCTING THE OPHITE DIAGRAM 277

IV. The Seven and the Ten

We cannot evade the problem any longer. Are there seven or ten
spheres in the Ophite cosmos ?
The problem arises because neither Origen nor Celsus is suffi­
ciently careful in delimiting the area of the Diagram occupied
by these spheres 43 ) . Basically, however, we are concerned with
the planets and one or two extras like Behemoth and Leviathan
(VI , 25 ) . The name of the latter was supposedly given to the
'soul of the universe' , and hence (despite Celsus' assertion) is
not properly the name of a circle at all. Origen says explicitly
that 'the author of that disgusting Diagram wrote this "Leviathan"
on the circle and on its centre, thus putting the name twice. '
So we must take Celsus' circle as the outer limit of Leviathan's
pervasive, all-ensouling presence, which is equally to be found at
the centre of the concentric system. This raises general doubts
about the exact relation of names to circles : the name Behemoth
is also found 'set in position after the bottom circle. ' Furthermore,
it would make sense to think that the long formulae for the initiate
on his ascent must have taken up virtually all the space between
the circles (unless the Diagram was extremely large ) . The formulae
themselves give the impression that the initiate sees the archon
and pronounces the words immediately after crossing the barrier­
circle of his orbit : Horaios, e.g. , has 'fearlessly passed beyond the
barrier of fire' when the initiate invokes him; he addresses Ialda­
baoth too as already 'having opened for the world the gate which
you by your aeon has closed, and go past your power as a free
man' ; and in the supracosmic formula also preserved by Origen
the initiate has already 'passed over the offence (rppoty(.L6c;; ) of wicked­
ness' (VI, 31) . It would appear, then, that the names were inscribed
between the spheres rather than on them 44 ) .
We should probably picture this part of the Diagram as follows.
In the centre, close to the earth, is inscribed the name of the world­
soul, Leviathan; then, below the first of the circles is written
Behemoth-this possibly means the sphere of the atmosphere
or air; the first circle may be a 'barrier of fire' above the air, but
at any rate between it and the second circle is written Horaios,
43) Though I was undoubtedly wrong elsewhere to suggest that Behemoth
+ 7 planets + Father + Son = 10 circles.
44) HoFFNER, p. 87 thus wrongly speaks of 'noch ein Kreis, der Beemon
heisst. ' Cf. LEISEGANG, p. 172 .
 A. J. WELBURN

and all the material relating to him ; proceeding outwards, between

the second and third circles is written Ailoaios, etc. ; and so we
proceed until, beyond the seventh circle, is written Ialdabaoth.
Encircling the whole is another, eighth circle, labelled Leviathan
to show that all the entities up to and including Ialdabaoth-Saturn
are ensouled by him . Thus we have a minimum of eight circles.
Add the yellow circle in which the whole is contained, we have
nine ; and the blue circle which is also contained in it besides the
cosmos brings the total to ten.
Ten is not a usual number among Gnostics. As an anticipation of
the extraordinary significance accorded to it in kabbalism, however,
it confirms that with the Ophites we are very close to Jewish
esoteric doctrines. G. ScHOLEM has added his voice on the subject,
announcing (without much explanation) that Leviathan is here
a genuine piece of Jewish lore, and pointing out other connections 45} .
From this, as well as from the parallels in the Apocalypse, we may
deduce that Ophite Gnosticism arose in an environment permeated
by esoteric Jewish teaching, which it took over and adapted to
the framework of dualistic gnosis. The system, once evolved,
no doubt exerted a reciprocal influence on the development of
'orthodox' kabbalah as we know it from its earliest documents
like the book Bahir (perhaps to be dated to the third century
in its original form) 46 ) .
Finally, can we conjecture the exact position of the eight cosmic
circles and the blue circle ? I believe we can. The yellow circle
is a foil on the left side of the Diagram to the Father-circle on the
right, and the demands of symmetry urge us to position the blue
circle to match the Son within the Father-i.e. within the vesica
on the opposite side of the axe-head. This leaves room for the
cosmos-system in the left half of the yellow circle. There is even
a hint from Celsus that the cosmos and the blue circle make sense
in this position. The last important entity in the Ophite system

45 ) See the several references to Ophitism in his On the Kabbalah and

its Symbolism (New York 1969) , pp . 160, 163-5; Major Trends pp . 360 f.
For the Gnostic Leviathan, see BoRNKAMM, Mythos und Legende in den
Thomasakten (Gottingen 1933) , pp. 28 f. Some rather vague Mandaean
parallels in ALTMANN, p. 2 1.
48 ) SCHOLEM, Das Buck Bahir (Leipzig 1923) . See Major Trends, pp. 74 f.
The third century would be the date of the earliest form of the present
work. On the early dating of these texts, see also his ]ewish Gnosticism
(New York 1960) , p. 8. It is in Bahir that we first hear of the ten sefiroth.
 RECONSTRUCTING THE OPHITE DIAGRAM 2 79

for which we have not yet accounted comes vividly before us in

his mention of 'the flowing power of a certain virgin Prunicos'
(VI, 34). As an aspect of Sophia she would fit very well in the
realm immediately beyond the cosmos in the direction of the
spiritual worlds-i.e. Ogdoad, which Origen tells us had a place
and a formula in the Diagram (VI, 31) . The symmetry of the
whole will be further enhanced when we draw in the tangent to the
Leviathan-circle and the blue circle, thus dividing the yellow circle
by way of analogy to the Father's diameter (see Fig. 3) . For after
describing the ten circles, Leviathan and Behemoth Celsus also
tells us that 'the Diagram is divided by a thick black line, and he
alledged that he had been told this was Gehenna, which is (the
equivalent of) Tartarus' (VI, 25) . HOFFNER and LEISEGANG placed
this line across the earth alone, a small circle in the centre of the
cosmos. But this is to be influenced too much by classical Greek
conceptions of Tartarus as a subterranean realm 47 ) . They failed
to adapt to the cosmic dimensions of the Gnostic underworld.
In the Gnostic literature known in Coptic, Tartarus almost invari­
ably means the whole area of darkness outside the pleroma;
and in it the whole cosmic system comes into being (Hypostasis
of the Archons CG II 95 , 12; cf. Origin of the World 102, 34; Book
of Thomas 142, 35 ff.; etc.) 48 ) . In this case, too, it becomes plain
that 'Tartarus' is not the name of the line but of the area to the
left of it; the thick black line simply defines its limit. Such a
placing of the line accords well with Celsus' assertion that the
Diagram is divided, intimating that it traversed a major part
of the Diagram's structure. Excellent sense can in fact be made
if the Tartarus-line does divide the large yellow circle diametrically :
we could then identify the 'cppoty[L6t; of wickedness' that the initiate
crosses after leaving behind the cosmic archons ; beyond it, in
the blue circle, lies Ogdoad (VI, 31) . This interpretation was
anticipated by LIPSIUS 49) . The motif of 'wickedness' gains a new
relevance if the 'fence' was actually the dividing-line of Gehenna!
Further support comes from the name Thartharaoth--one of the
protean alternatives for the outermost sphere. GRANT saw in

47) Only slight differences between HoPFNER, p. 93 and LEISEGANG, opp.

p. 160.
48 ) The correct interpretation was anticipated by LEGGE, II, p. 6g.
48) LIPSIUS, p. 42. On the phragmos see also BoRNKAMM, Mythos und
Legende p. 29.
280 A. J. WELBURN

this 'a Hebraized form of Tartarus' 50) ; though against him it

might be urged that it is just another corruption of Thaphabaoth
( Tohu wa bohu) . Perhaps what really happened is that the
=

copyist of the Diagram in some version contaminated the name

Thaphabaoth when his eye was drawn to the nearby 'Tartarus'
written along the dividing-line.

V. Sophia and Agape

Our final problem is the internal structure and the positioning
of the two 'small circles' described by Origen. We are given details
for only one of them, though we may assume from the context
that they formed a matching pair. They were labelled 'Love'
(&ycf7t1)) and 'Life' respectively. Origen says of them (VI, 38) :
On the second circle, which embraces and circumscribes two other circles
and a further rhomboid shape, (is inscribed) 'Providence of Sophia', and
inside their common sector 'Nature of Sophia' ; above their common sector
is a circle, in which is inscribed 'Gnosis' , and below is another, inscribed
'Insight' (o-Uve:aLc;) u ) .

This is hardly transparent. Yet the figure can be recognised

with considerable certainty from its later history. The two circles
and the rhomboid conspire to assure us that the vesica is once
again involved, since the intersection of the two interior circles
defines the proportions of the rhomboid. The figure is shown, without
any dependence on this Ophite material, by BoND and LEE 5 2) .
To suit our present system of Gnosticism all we need do is add the
circumscribing outer circle, marked 'Pronoia of Sophia'; the
inscription 'Physis of Sophia' can then be inserted in the central
vesica, and the rhomboid determined by this and the extremities
of the two inner circles will have its breadth and length in the
mystical ratio I : V3 (Fig. 2) . In all this HoPFNER just missed
the truth. HILGENFELD, following LIPSIUS and adding a visual
reconstruction, began the misunderstanding by placing the two
inner circles inside the rhombus : the rhombus itself hung arbitrarily
within the circumscribing circle, since it could not expand to
touch the circle at all four corners without becoming a straight­
forward square. The overlap of the circles was in his case arbitrary

60) GRANT, op. cit., p. 47·

61) The passage on the 'Nature of Sophia' is inexplicably omitted in
GRANT, Anthology p. 92 .
62) BoNo and LEE, Gematria p. 54·
 RECONSTRUCTING THE OPHITE DIAGRAM 281

Fig. 2

too, though this is easily remedied by insisting on a precise vesica 53) .

This slipshod geometry was taken over by HoPFNER and LEISE­
GANG 54) . In our reconstruction, contrastingly, there are firm
constraints on the proportions of all the figures when once the
diameter of the circumscribing circle is fixed-and I shall demon­
strate shortly that it in turn is determined by the Diagram as a
totality.
It seems that I also clash with HoPFNER in that by my reading
there are two extra circles to be inserted. Origen asserts that
the first lies above the vesica ('common sector') and the second
below it : I am unable to see how this could refer to the circles
actually possessing that common intersection, as he and his fore­
runners evidently thought 55) . The size of these circles (see Fig. 2)
is fixed by their touching the outer and the two interior circles.
There is room for one at the top and one at the bottom as Origen
says, and we may call them 'Gnosis' and ' Insight' respectively.
And with this our 'Life'-circle is complete. I shall add only that
whereas the centre inscription 'Nature' of Sophia presumably
names the area of the intersection, the 'Providence' of Sophia
is probably not the name of the circumscribing line but of the
63 ) HILGENFELD, Ketzergesckickte ( 1884) , p. 279. Cf. LIPSIUS, p. 45, though
he wrongly identifies the Life-sphere with that of the earlier mentioned
'square ' .
64) HoPFNER, p. 97 ; LEISEGANG, p. 1 70.
&& ) HoPFNER, lac. cit.
 A. J . WELBURN

whole area limited by it. Sophia's providence would thus extend

beyond her nature and also pervade Gnosis and Insight.
An analogous figure was probably constructed within the other
circle, that of 'Love'. But of that we know no more. The only
remaining problem, therefore, in our reconstruction of the Diagram
is the positioning of these 'Life' and 'Love' circles. In the first
place, Origen may be read as instructing us that the 'Love' circle
(and so I take it also the identical 'Life' circle) is smaller than
the large yellow circle (VI, 38) . In that case we may be fairly sure
we are dealing with an addition within the Diagram, that does not
drastically affect its general morphology. We may also elicit from
the text the information that the 'Love' circle was 'touching the
greater of the first two circles' 56) ,-which might refer either to
the yellow or (perhaps) the Father-circle. The difficulty is com­
pounded by an additional cross-reference to the axe-head, since
Origen appears to say that one circle (i.e. that of Love) was 'above'
it (&.vw·dpw ) , and the other (i.e. of Life) 'below' (xcx't'W't'epw ) . This
prompts the contradictory and surely less likely thought that
the new circles project out beyond the main three, above and
below. Nor does HoPFNER's solution advance us, since it results
in an ugly overlapping of the circle with the 8toccppcxy(Lcx. His small
circle is in contact not only with the large Father-circle, but with
the Son-circle too, and it seems to me unlikely that Origen would
have bothered to specify that it touched the larger circle, if it
actually touched both. As for the yellow circle-if that is the one
Origen meant-it cuts HoPFNER's circle randomly, without generat­
ing any definite or significant figure. His second circle is 'below' the
8toccp p cxy(Lcx only in the same, broad sense that most of his Diagram
is below it ; and surely Origen meant something a little more
helpful than that. The best (if not the only) way of solving the
problem is to appeal to LIPSius' suggestion that &vw't'epw and
xcx't'w't'epw are being used here in a loose 57 ) sense. They might
mean, rather than abovefbelow, 'in the upper part' f'in the lower
part' : in which case the circles so described could fulfil all the
requisite conditions by lying in the upper and lower heads of the
axe. This would precisely define their size ; they would then be in
contact with both of the previously mentioned large circles, amply
fitting the prescriptions of the text. In this way we are finally
66 ) See the version of CHADWICK, p. 353·
67) LIPSIUS, p. 45 ·
 RECONSTRUCTING THE OPHITE DIAGRAM

Fig. 3· The ophite diagram restored.

able to offer a complete reconstruction of the Ophite Diagram

(Fig. 3) . Some details remain a little speculative and uncertain,
but all fit the text as transmitted to us from Celsus and Origen,
and I shall argue below that in terms of the Ophite system every­
thing makes very good sense.
Most important is the strict interrelation of the circles and other
geometrical figures in our reconstruction. For in this, I take it,
lay the cogency of the mystic Diagram for the Gnostic mind.
Given the diameter of the Father-circle, the size of every other
circle and the length of every line is instantly determined. And
since the smallest circles are named 'Gnosis' and spiritual 'Insight' ,
referring t o the attainment of individual visionary illumination,
all the entities of the cosmos and the supracosmic realms could
be understood as stages in a chain of spiritual-mathematical
necessity stretching uninterruptedly from the individual spark
of light to the all-originating Father 58) .

VI. The Four and the Three

Irenaeus' presentation of Ophite doctrine ( Adv. Haer . I, 30,
I-IS) provides several near parallels in thought and terminology
with our Diagram. The heresiologist claims that this Gnostic
68 ) The implicitness of the entire spiritual-cosmological content of Gnosis
in the moment of self-knowledge is claimed, e.g. by the Book of Thomas
(CG II 1 38, 1 5 - 1 8) .
 A. J . WELBURN

group honoured several pre-cosmic entities, beginning with a

First Man, called Father, and a Second Man, called his Son ; these,
I doubt not, are the beings denoted by the Father- and Son-circles.
'Below these, ' he continues, 'is the Holy Spirit . . . they call it
the First Woman. ' And 'below the Spirit on high are the various
elements, water, darkness, abyss, chaos. . . . ' The Woman we
may presumably recognise in the central large circle, and the
dark Chaos in the yellow circle to the left. In this case the Woman
could appropriately be described as 'overfull and bubbling over
on the left side' (I, 30, 2) into the region of Chaos : her sphere
intersects that of the Father, and in the vesica there arises the Son
on the right ; and on the left where her circle overlaps the yellow
circle, the Woman becomes Prunicos, the being who wantonly
stirs the dark waters and takes a bodily form in them. These
relationships are very clearly represented in the structure of our
Diagram. Moreover, the Son is also able, since he is within the
Woman-sphere, to generate with her the Third Male, the Christ
whose place we shall discuss shortly 59} . This fits with the passage
extant in the Latin Interpretatio Vetus of Irenaeus at this point,
which adds that Prunicos was said to emerge from the left side
of the First Woman, and Christus (perhaps really, the Son) from
her right 60) . It might initially strike one as odd that the 'darkness'
should be represented by a yellow circle-but we must be wary
of proj ecting back modern naturalism 61) . The analogous 'green'
circle was understood by the author of the Baraita we cited to be
the one 'from which darkness emanates', and this once more
agrees well with our Diagram . Irenaeus even says that the material
of this realm 'encircled' Sophia-Prunicos (I, 30, 3) . The later,
cosmic phase of creation, however, is no longer under the Woman's
direction. It is effected by her son Ialdabaoth, 'who despises his
mother, inasmuch as he made sons and grandsons ( = archons)
without anyone's permission. ' This will take us to the far left of
the Diagram, beyond the arc of the Woman's circle containing
Prunicos, and so out of her sphere of power. From his physical
59) Cf. the 'Thrice Male' of the new Sethian literature, whose name seems
to be explained in various ways according to the system in which he appears .
60) See LEGGE, II, p. 46 and n. 1 .
61) I doubt whether CooK's idea (quoted in CHADWICK, p . 353 n . 2 ) that
the yellow and blue circles depict a 'burning' and a 'moist' zone of the
atmosphere could have been envisaged in Antiquity, or at any time before
the Renaissance.
 RECONSTRUCTING THE OPHITE DIAGRAM z8s

desire, Ialdabaoth proceeds to generate a son, and this turns out

to be the Snake so central to Ophite Gnosis : he comes in at
point very fitted to be the Leviathan pervading the cosmos in
our reconstruction ( I, 30, 5) 62) . The Snake produces six sons, to
form a counter-hebdomad to Ialdabaoth's planetary seven-perhaps
this is the myth which accounts for the dual names assigned to
each sphere.
Irenaeus' Ophite myth is complicated by the theme of the
repentance of Sophia, but this too may actually explain certain
features of the Diagram. In order to safeguard the trace of light
in her, Sophia is reported to have taken the following course :
She leapt up and was carried up to the height, and there she spread out
and became a covering, and made this visible heaven . . . which still has
the appearance of a watery body.
(I, 30, 4 )

If this refers to anything in our Diagram it must be the central

8�&cppocyf-C.Ot. Cosmologically this would be the actual concave
backdrop of the sky, against which from earth we see the heavenly
bodies borne in their circles over and around us. The idea of the
firmament as a dividing curtain between the cosmos and the
spiritual world is well-attested in Gnosticism 63 ) . In forming the
'visible heaven' , Sophia splits into two : her bodily part is left
behind as 'the woman from the woman' (i.e. the purely female,
natural part ) , but her higher nature shelters under the heaven
she has j ust made. The Ophites do not seem, like some later Gnos­
tics 64 ) , to have adopted one of the names for this repentant mother
to refer specifically to her one aspect or the other. But presumably
in terms of the Diagram her bodily part remained in the blue
Prunicos sphere, and, following the myth, her higher part might
well be represented in the shelter of the firmament-partition-where
we in fact placed the Life-circle. Inside this, it will be recalled,
is written ' Nature of Sophia' , evidently meaning her higher nature
62) Leviathan is an ophis in Is. 27, 1 LXX. The difficulty over Sophia
being the Snake in the view of some (Adv. Haer. I, 30, 1 5 ) may have arisen
because the Leviathan-Snake constitutes a sort of unofficial Ogdoad above
the seven spheres in our Diagram ; Irenaeus' Ophites seem to have rather
made Leviathan the seventh, as he generates only six sons below him .
6 3) Discussion in WILSON, Gnosis and the New Testament (Oxford z g68) ,
pp. 72 ff. If this is the significance of the diaphragma in our Diagram, the
closest parallels would seem to come from Basilides and certain Valentinians.
6 4) See Gospel of Philip log. 39 (ScHENKE), where Echamoth and Echmoth
are distinguished ; and the contrast between Enthymesis and Pathos in
the system of Ptolemaeus, etc .
z86 A. J. WELBURN

as she is there associated with Gnosis and VlSlonary Insight.

The repentance of Sophia was, of course, widely conceived as the
archetype of the Gnostic's transcending of the world. In the
8Loccppocy{Loc she is as yet in an intermediate condition. Though
rescued from the yellow sphere of chaos and creation, she is still
not in the spiritual sphere of the Father and the Son. The process
requires for its completion the intervention of Christ.
In the absence of any definite information it is tempting to
speculate on the place of the 'Third Male' in the Ophite Diagram­
unless he was, as has been supposed, a later addition to a pagan
system. And it is tempting to bring this question into relation
with the persistent lack of information about the 'Love' circle
which balances Sophia's 'Life' circle in the upper part of the
partition. In the Sethian Gospel of the Egyptians ocyoc1t'Y) is associated
with Seth and Jesus, the main soteriological figures ( CG III 53,
5 -6 ; 65 , 16-7 ) . In Irenaeus' Ophite system Christ is called the
'brother' of Sophia ; he also 'puts her on' when he descends, ap­
parently becoming temporarily one with her at least to the extent
of a syzygy or quasi-sexual union. Perhaps he was represented
by the Love-sphere which balances that of his sister-bride. For
indeed Love would be a finely appropriate power for him to rule.
But here no firm knowledge is available.
On the more definite side, however, the Diagram does suggest
a resolution of certain problems that arise in several related Gnostic
systems-the Sethians (whom Theodoret identifies with the
Ophites) , the Naassenes, Peratae, the Megale Apophasis. FoERSTER
has classed these as 'systems involving three principles', but notes
at the same time that in practice they cannot manage with these
alone 66) . The Naassenes, for example, have to appeal to an Esaldaios
(El Shaddai) , a fourth power who is evil (Hippolytus, Ref. V, 7, 30 ;
8, 5) . The Sethians (as reported by Hippolytus V, rg, I ff.) add
to the three pure roots which their system postulates a fourth
state of 'mixture' . The Peratae surreptitously made a fourth
principle from the hostile organization of the stars, and in the
Megale Apophasis a mysterious and inimical intransigence in the
world-process impedes the work of cyclical redemption. With
regard to the Sethians, the Nag Hammadi library has brought
us a version of the work which, with certain differences, had
apparently been used by Hippolytus for his resume. In the Para­
phrase of Shem we are told clearly that in the beginning there
86) FOERSTER, I, pp. 244 f.
 RECONSTRUCTING THE OPHITE DIAGRAM

were three 'great Powers' : a spiritual Light, a Darkness 'wrapped in

a chaotic fire' , and between them a Spirit, gentle and mild (CG VII
r, 23 - 2, ro) . These 'roots' agree extremely well with the three
circles of our Diagram, and they are further described as 'covering
each other, each with its power' (2, g-ro) , perhaps being visualised
in the manner of our overlapping circles. In their primordial
state each root is at rest, undisturbed within itself : but the created
world adds a fourth state where their Powers are mixed and
warring against each other.
In these several systems it was far from clear how four worlds
could derive legitimately from three principles-the sudden intro­
duction of a 'Fourth' or some undefined agent of mixture seemed
mere inconsistency. But the Diagram elegantly reconciles the
three and the four. There are three large circles : Father, Spirit,
Chaos. Yet from right to left we can easily count four domains
or stages in the cosmic process : the area of the Father alone ;
the area of the Son ; the area of Prunicos ; and finally that of Darkness
alone, in whose semi-circle is raised the starry system of the cosmos .
The intersection of three circles thus furnishes four worlds, and
we may see that the 'fourth principle' is not really an appeal to
something outside the system. It is simply the third or Dark
principle operating in isolation from the other two. In the Para­
phrase of Shem the material creation commences when the Dark
root tries of itself to assume the role of the Spirit (3, 7 ff.) . This
creates a 'confusion' and creative activity begins in the chaotic
world. In terms of our Diagram, it is this which disrupts symmetry
-which could be restored only beyond the reach of change in
the Father-circle, or more straightforwardly by the annihilation
of the Cosmos-structure in accordance with Gnostic eschatological
expectations. The latter alternative is, I take it, what the Gnostics
would have understood from the Diagram's lopsidedness. It
shows that a system with three principles can validly incorporate
a factor of intransigence through the partial independence of the
powers involved 66) .
66 ) HoFFNER, p. go, also treated the large circles of his reconstruction

as 'principles', and compared them with the Good-Elohim-Edem triad

in Baruch-Gnosis. But a Male-Male-Female structure shows less resemblance
to our Diagram than do the Ophite-Sethian-Naassene systems discussed
above. Cf. LEISEGANG, pp . 1 69-7 1 . The religious tradition which attributes
the impetus of creation to the Dark power is of course the Iranian, and
its rapprochement with Gnosticism culminates in the Manichaean myth.