[Robartes: ] There [in Cracow] I took up with a very handsome girl of the poorer classes and hired a couple of rooms in an old tumble down house. One night I was suddenly thrown out of bed and when I lit my tallow candle found that the bed — which had fallen at one side towards the head — had been propped up by a joint stool and by an old book bound in calf. I found in the morning that it was the Speculum Angelorum et hominis of Gyraldus printed at Cracow in 1594, a good many years before the first celebrated Cracow publications and in a much earlier style both as to the kind of woodcut and of type. It was indeed full of curious allegorical woodcuts, astronomical diagrams, where drawings of Noah’s Ark and the Tables of the Law were mixed up with Zodiacal signs and phases of the moon and geometrical diagrams where cones containing gyres sprang out of each other like strange vegetables. My beggar maid had found it she told me on the top shelf in an old cupboard in the wall where it had been left by the last tenant — an unfrocked priest who had joined a troup of gipsies and disappeared. I had made some progress with the Latin text when a quarrel with my beggar maid plunged me into wine and gloom once more. Then I turned violently from all sensual pleasures and decided to say my prayers at the Holy Sepulchre and from there I drifted to Damascus taking with me the book which more and more absorbed me. An attempt to reach Mecca ended in my finding myself in a remote Arab town thanks to a small medicine chest which I always carry I became first doctor and then a kind of steward to an Arab chief or petty king. One day our town — or rather village was visited by a tribe of Judwalis. There are several tribes of this strange sect who are known among other Arabs for the violent contrast of character amongst them, for one finds amongst them holy men and others extremely licentious. Fanatical on all matters of doctrine they seem tolerant of human frailty beyond any people I have ever met. One of them an old man celebrated for his wisdom whom I was treating for some complaint cried out in wonder on seeing the Astrological title page of my book. It contained, he said, certain doctrines of the Judwalis and from learning from him that they possessed an ancient book I resigned my post and wandered with his people for some years. The book which the old man showed me without difficulty is called       [space for Arabic name]       The way of souls between the moons and the suns, and is attributed to Kusta ben Luki the Christian philosopher and man of science at the court of Harun al-Rashid. Except in the case of the cones and gyres it was much more elaborate and much more detailed than the work of Gyraldus and from the prolonged study which I have since been able to make of both books I am convinced that it was derived like the work of Gyraldus itself from a lost Syriac original.
‘The Discoveries of Michael Robartes’, typescript dialogue of Robartes and Aherne, YVP 4 16-17;
c.f. AV A xvii-xx & AV B 38-41

Geraldus, Giraldus, Gyraldus?

Yeats’s spelling was rather weak and somewhat erratic; added to this, many mediaeval and Renaissance names appear in several versions, both vernacular and Latin, and vary hugely. Shakespeare famously spelt his own surname in a number of different ways, to which his contemporaries added further variants. The name of Yeats’s character was given as ‘Geraldus’ on his first appearance in print in the Preface to Michael Robartes and the Dancer in 1921 (VP 853), but as ‘Giraldus’ in AV A and then AV B, and this is followed by most critics. The other variant Yeats uses here, which appears in Laurie’s prospectus for subscribers and generally in the Script, ‘Gyraldus’, may be linked to a wordplay on ‘gyre’, though it appeared several months before the first use of the term ‘gyre’ in the Automatic Script.

Dulac's portrait of Yeats as Giraldus

Is there an historical Giraldus?

As discussed in the background to the fictions, Yeats first appears to have mentioned the mysterious writer, Giraldus, in a letter Augusta Gregory (4 January 1918; L 644), but he had obviously been considering him even earlier, since Dulac had already been asked for a portrait (see below). The name appears in the Automatic Script on 12 January 1918 in one of the answers, ‘Gyraldus primary // Arab anti’, already linked to a contrasting Arab perspective. The draft quoted above was used with minor changes in A Vision A, but certain details were removed. The Great Frontispiece of the Speculum Angelorum et Hominum identifies Giraldus as a man of Phase 18, like George Yeats: giving the figure George’s Phase and Yeats’s face was an elegant acknowledgement of the System’s parents, which was lost with this element of the fiction. Another element that was not explained in the published versions is where other, more complete, copies of the book might be, but in this draft Robartes states that, ‘I have no doubt that [this copy of Gyraldus’ Speculum] is unique and for no accidental reason. Its author must have repented of his purpose or have discovered that there were reasons he had not known of for silence’ (YVP 4 19), indicating a certain care in the fiction, which also harks back to the unique copy of the Liber inducens in Evangelium aeternum in the story ‘The Tables of the Law’ (Myth 296). This idea seems to have been part of Yeats’s thinking from the very first, though originally the book’s suppression seems to have been from the authorities rather than the author. In the very early ‘Untitled Mansucript’ (YVP 4 119-134), Yeats alludes to the fervour of the Counter-Reformation, while leaving Giraldus still in the shadows:
I found in the morning that it was the Speculum Angelorum et Hominis of Giraldis printed in 1594, a good many years before the first of the celebrate Crackow publications and in a much earlier style. This roused my curiosity more at first than the contents of the book & [I] made enquiries at all the booksellers & write to a number of librarians and book collectors. No one had ever heard of the book, a hitherto unknown book [till at last an authority on] the religeos history of the sixteenth century discovered an allusion in a controversial book of the time to the public burning of the works of a certain     [blank for name]     but of Giraldus him self he could learn nothing. It was now plain that in all probability I possessed the only volume of his work which some accident had saved.
(YVP 4 121)

Details here evoke the fate of such writers as Giordano Bruno, while the fortunes of Robartes’ unique copy of the Speculum recall the destruction and fortuitous escape of ‘divers Books in Manuscript, & Papers’ of the English magus, John Dee, which were found hidden in a cedar chest around 1662. They went unrecognised initially, leading a ‘Servant Maid to wast about one halfe of them under Pyes & other like uses’. The rest escaped the Great Fire of London, while ‘the Chest perished in the Flames’, until they finally found their way to Elias Ashmole.

The passage from the ‘Untitled Manuscript’ implies that Giraldus was writing in the sixteenth century, and in ‘The Discoveries of Michael Robartes’, Robartes refers to ‘Gyraldus writing in the midst of the Renaissance’ (YVP 4 39). However, elsewhere in the same typescript Yeats deliberately obscures even the period, so that although the book is ‘printed at Cracow in 1594’, Robartes comments that it is in ‘a much earlier style both as to the kind of woodcut and of type’ (YVP 4 16), and Yeats, in his instructions for the woodcut of the Great Wheel, told Edmund Dulac that he could ‘give the speculum what date you please’ (14 October 1923; L 700). A good number of books published in the sixteenth century had circulated in manuscript long before they were printed, and pictures in particular were reused between different works, not necessarily even by the same author, so that the reference to earlier styles may be meant to suggest that the blocks were being recycled. In short, Yeats sought to obfuscate almost every circumstance or detail of Giraldus as a person.

The possible identity of Yeats’s European ghost writer has nevertheless intrigued readers and researchers since G. R. S. Mead first reviewed A Vision A: ‘Take his mythical Giraldus. If he supposes that the famous Humanist of that name, Gregory of Ferasa, the friend of Picus de Mirandula, could supply sufficient camouflage for his purpose, he is greatly mistaken. . . . We have also a woodcut portrait of G. facing The Vision’s title-page, which will doubtless impress the unwary’ (The Quest 18:1, 96-97, October 1926; full text). Even before publication Edmund Dulac‚ who had just finished the offending pastiche portrait of Yeats as Giraldus, asked ‘Is your Giraldus, Giraldus Cambrensis?’ (15 February 1918; LTWBY 2 344). Evidently Frank Pearce Sturm also wondered, though Yeats appears to have given no clear answer when Sturm wrote to him in 1924 complaining that he could not trace Gyraldus; Sturm was clear that: ‘He is not Giraldus Cambrensis, but may be Gerard (or Gerald) of Cremona, who gave his life & great wealth to translation and had a go at everything from Aristotle to Kusta ben Luki’ (FPS 86).

These three figures remain the front contenders in critical writing.

  • Richard Ellmann follows Dulac’s idea, referring to the character as Giraldus Cambrensis in his account of the fictions in Yeats: The Man and the Masks (237).
  • Meanwhile, in their headnote to Dulac’s letter to Yeats asking if the model was Giraldus Cambrensis, Finneran, Harper and Murphy follow Sturm, stating that ‘the basis for Giraldus is not Giraldus Cambrensis but Gerard of Cremona, who translated Arabic scientific treatises into Latin’ (Letters to W. B. Yeats 344). They give no reasoning for the assertion, which is understandable in the space available, but Harper and Hood’s Critical Edition of Yeats’s ‘A Vision’ (1925) from the following year (1978) gives no positive identification of Giraldus at all, and it may be that Harper did not support the attribution.
  • Kathleen Raine, following ideas from Liam Miller, follows Mead, in seeing the Ferrara-born humanist scholar, Lilius Gregorius Gyraldus (also spelt Giraldus), as the primary model in ‘Giraldus’ (Yeats the Initiate).

1. Yeats certainly refers to Giraldus Cambrensis (c.1146-c.1223) in some of his earliest folkloric writing, such as his anthology, Fairy and Folk Tales of the Irish Peasantry (London: Walter Scott, 1888), commenting that Giraldus ‘found the people of the western islands a trifle paganish’ (Introduction), as well as in section prefaces (x, 47 and 148). Later references include a letter to the Speaker (1897; CL 2 100; L 285) and a note in Augusta Gregory’s Visions and Beliefs in the West of Ireland (written in 1914: Vol. 1, 91; Later Essays 278). His Topographica Hibernica (c.1188) and his Expugnatio Hibernica (c.1189), show an interest in Ireland, albeit a condescending and rather hostile one, describing the Irish as ‘cruel and blood-thirsty’, ‘a race sunk in vice’ (editors’ note, CL 2 100). Though he is probably the best-known Giraldus in the English-speaking world, his other works include handbooks for the instruction of the clergy and a prince, and he is not an obvious candidate for a symbolical and unorthodox work.

2. Gerardus Cremonensis, also Girardus, Gherard, Gurrardus and Giraldus (1114-87), seems far more appropriate. He was the pre-eminent figure in the early School of Toledo, which was rediscovered and perhaps reinvented by nineteenth-century scholars, where he worked as a translator (see Anthony Pym, ‘Toledo and All That’). Gerardus Cremonensis translated many works from Arabic into Latin (figures given include 66, 76, ‘about 80’ and 92) including Ptolemy’s Almagest, along with works on astronomy, dialectic, geometry, philosophy, physics, and other sciences. ‘Some of the works, however, with which he has been credited (including the Theoria or Theorica planetarum, and the versions of Avicenna’s Canon of Medicine . . . and of the Almansorius of Abu Bakr Razi) are probably due to a later Gerard, of the 13th century, also called Cremonensis but more precisely de Sabloneta (Sabbionetta). This writer undertook the task of interpreting to the Latin world some of the best work of Arabic physicians, and his translation of Avicenna is said to have been made by order of Frederic II’ (Encyclopaedia Britannica, 11th ed., 1910, Vol. 11, 764). A mythopoeic version of this last story has the elder Gerardus Cremonensis translating Avicenna at the behest of Frederick Barbarossa to make the dates appropriate, which is not dissimilar to Yeats’s cavalier treatment of the dates of Kusta ben Luka and Harun al-Rashid.

Gerard of Sabloneta was also the author of Geomantia Astronomica, but it is usually referred to as the work of Gerard of Cremona, without differentiating which one (an English version [PDF] is available here). Geomantia Astronomica, a work on geomancy, appears as an appendix in the Beringus brothers’ edition of Agrippa’s Latin works (a fictitious imprint), along with an unattributed work of Qusta ibn Luqa’s. The Yeatses later owned a copy of this edition of Agrippa (YL 24), and though it does not appear in the 1920s library list, they could have known it from libraries earlier. The same treatise also appeared in a more modern publication, The Principles of Astrological Geomancy by Franz Hartmann who gives a system of ‘Geomancy described by Cornelius Agrippa. It is called the Astronomical Geomancy of Gerhard of Cremona’, and Yeats was familiar with Hartmann’s writings, having used his digest of Boehme for The Works of William Blake.

Geomantia Astronomica utilises geomantic principles to draw up a pseudo-horoscope in order to answer a horary type of interrogation, and geomancy was part of the Golden Dawn’s required study, and the Golden Dawn methods described by Regardie are similar to Gerardus’ in their use of a horoscope format to place the geomantic figures. ‘Geomantia’ is a term first used in twelfth-century Spain to translate the Arabic name ’ilm al-raml (‘the science of sand’) and although random marks on a piece of paper have long been the most common method of arriving at the figures, the desert sands seem to have held an imaginative or symbolic fascination for Yeats. Warwick Gould points out that Yeats’s statement in Per Amica Silentia Lunae that, ‘like the Arab boy that became Vizier’ he has ‘taken stock of the desert sand and of the sayings of antiquity’ (Myth 343) is a rather cryptic allusion to geomancy in the Arabian Nights story of ‘King Wird Khan, his Women and his Wazirs’ (‘'A Lesson to the Circumspect'’). These desert sands are the precursors of those bearing the marks of the Dance of the Four Royal Persons, as played out by the followers of Yeats’s Kusta ben Luka, and ‘the emblems on the sand’ left by Kusta’s bride in ‘Desert Geometry or The Gift of Harun Al-Raschid’ (AV A 126, VP 468).

Taylor comments in his introduction to Frank Pearce Sturm, that Sturm ‘suggests most plausibly that the closest possible historical figure is Gerard of Cremona (Gerardus Sablonetanus) and not Giraldus Cambrensis. Again, the identification did not appear in the 1925 volume, and a garbled version is given in the second edition: ‘I had made a fruitless attempt to identify my Giraldus with Giraldus of Bologna.’ The desire to cloud the issue and identify his mythical, medieval persona . . . with Celtic Ireland as well as the Middle East is perfectly understandable, but the substitution of Bologna for Cremona can only be a blunder’ (FPS 58) which is probably correct; Yeats deliberately makes Robartes’ attempt at a positive identification fruitless, but also, misremembering himself, accidentally confuses the issue even further.

3. Lilius Gregorius Gyraldus or Giraldus (1479-1552), is one of the only suitably-named writers alive close to the date of publication that Yeats mentions (1594), though he deliberately muddies the issue by alluding to older traits in the style of the presentation. As well as Historia Deorum Gentilium ('A History of the Pagan Gods'), Lilius Gregorius Gyraldus wrote De Annis et Mensibus ('Concerning Years and Months'; Basel, 1541) which contains a discussion of the Annus Magnus that might have been of particular interest to Yeats. Kathleen Raine also suggests another possibility in a further Ferraran, Giambattista Giraldus (1504-1575), who edited the older Giraldus’ works, and is better known as the Cinthio whose Ecatommiti provided Shakespeare with the stories of Othello and Measure for Measure. This latter appears as ‘Giraldi’ in the entry preceding ‘Giraldus Cambrensis’ in Chambers Biographical Dictionary (YL 365).

It is also worth considering a few other possibilities:

4. One further Gerardus was actually publishing at the date Yeats chose, Gerardus Dorneus (Gerhard Dorn): reputed to be Paracelsus’ favourite pupil, he wrote on alchemical subjects, including the Emerald Tablet, and translated Paracelsus’ works from German into Latin, publishing between 1566 and 1602.

5. William H. O’Donnell notes another possibility closer in time and place to Yeats, William Allingham (see A Guide to the Prose Fiction, 144). A Donegal-born poet whose work influenced Yeats, he edited an anthology entitled Nightingale Valley in 1860 under the pseudonym of ‘Giraldus’ (perhaps to escape accusations of vanity for including six of his own lyrics in this number). Yeats was something of a proselytiser on Allingham’s behalf, and he wrote to Allingham’s widow in 1904: ‘I have the greatest posible admiration for Mr. Allingham’s poetry. I am sometimes inclined to believe that he was my own master in Irish verse, starting me in the way I have gone whether for good or evil’ (CL 3 683; L 446) and it is possible that he chose his pseudonym for a new master.

6. A more fanciful possibility could link the name with the Giralda, the famous campanile of Seville, converted from a minaret, and named after its weathervane (‘girar’- to turn), which marries the ideas of the gyre and the tower, as well as Arabic and Christian. Yeats did not visit the city until 1927, but could well have known of the tower, since, amongst other things, he had set Mosada in Andalucia.

Yeats appears to have avoided answering any questions about who his Giraldus was, and to have preferred that the character bring no external history, unlike Kusta ben Luka. So the probable answer to whether or not there is an historical Giraldus, is no, though it is probably also clear from the treatment above that, if one has to be chosen, I tend to prefer Gerard of Cremona. It is partly the symmetry of Qusta ibn Luqa’s involvement in the Abbasid school of translators, the ‘House of Wisdom’, that makes a European translator, working some two centuries later in the ‘School of Toledo’, seem most fitting. Even if he is appropriate, however, he may not have been the one on Yeats’s mind in 1917 when he first conjured with the name. Kathleen Raine is surely right to underline that the character is ‘one of [Yeats’s] dramatis personae, a fiction’, and that ‘a precise identification would situate the character in history not in the realm of imagination with which the artist is concerned. Besides, the imagined persona can combine characteristics to be found perhaps in no single individual. This is not to say that these figures, being the quintessences of history, are not carefully and exactly conceived’.

Speculum Angelorum et Hominum

The Speculum Angelorum et Hominum, ‘The Mirror of Angels and Men’, to give the work its final, most correct version, appears to have started in the earlier drafts as Speculum Angelorum et Hominis, ‘The Mirror of Angels and a Man’. Yeats probably thought that it could mean ‘Man’ in the sense of ‘Humanity’, but evidently learnt that this would not do. It was corrected to the plural ‘Hominum’, though a combination of Yeats’s often illegible handwriting and Wade’s habit of correcting Yeats’s spelling, means that the ‘Hominum’ in the letter to Augusta Gregory (4 January 1918; L 644) may or may not show an early correction. In the preface to Michael Robartes and the Dancer (Cuala, 1921), Yeats refers to ‘the Speculum Angelorum et Hominum of Geraldus’ (VP 853). At some stage, however, the form ‘Hominorum’ crept in, which is how the title appears in AV A (xvii ff.) and even ‘Homenorum’ on the Frontispiece. Yeats wrote to Frank Pearce Sturm in February 1926 that the ‘proofs were read & reread by me as I never read proofs before, then by a friend here & then by a professional proof reader in London. Yet the first day the book arrived I found under the frontispiece "Speculum Angelorum et Hominorum" instead of "et Hominum." I had written it down in haste, meant to look up the name & forgot. I have forgotten the little classics I knew forty years ago’ (FPS 93). He wondered in a postscript ‘if "hominorum" is too ungrammatical for dog Latin’, and was told that ‘However dog the Latin you can’t keep "Hominorum" because it isn't latin’, so that he could either admit the mistake or try for a diminutive such as ‘homunculorum’, which would have changed the meaning (FPS 94). G. R. S. Mead seems to have taken some schoolmasterly delight in twitting Yeats for the slip in October: ‘But "Homenorum" is a "howler" for which Smith Minor at a Preparatory School would receive condign punishment. Nor is this mitigated by Mr. Yeats’ "Hominorum" on p. xvii’ (The Quest 18:1, 97). It is ultimately a rather trivial detail.

Far more important are the implications of the title, since Speculum Angelorum et Hominum belongs to a strange assortment of works which use the term ‘speculum’, mirror, looking glass or prospective glass. It is probably too loose to be called a genre, yet Yeats has selected it with accuracy. Herbert Grabes, in Speculum, Mirror und Looking-Glass, lists over three hundred books published in England alone, between 1500 and 1700, which use the metaphor of the mirror in their title, almost 60 in the Latin form ‘Speculum’ (288-351). From the Middle Ages until the eighteenth century, when Latin titles became less common and a specialised medical usage began to take over, the description was attached to histories of instructive lives (both for imitation or warning), meditations on the human condition, examinations of alchemy and astrology, surveys of the state of the world, prognostications, ephemerides, Christian emblems, Rosicrucian tracts, Menippean satire and Cabbalistic symbolic synthesis. Most pertinent here are examples such as a life of St Francis by Brother Leo, Speculum Perfectionis, the Speculum Alchimiæ of Roger Bacon, a title also used by Arnaldus de Villa Nova (Arnau de Vilanova, Chaucer’s Arnold of the New Town), Francesco Giuntini’s Speculum Astrologiae (1573), William Vaughan’s Speculum Humanæ Condicionis (1598), the Rosicrucian Johann Valentin Andreae’s pseudonymous Menippus, sive dialogorum satyricorum centuria inanitatum nostratium speculum (1618), Daniel Mogling’s pseudonymous Rosicrucian manifesto Speculum Sophicum Rhodostauroticum (1618), Johannes Hintner’s anti-Rosicrucian Speculum Ambitionis (1620), the Jesuit Jacopus Masenius’s Speculum Imaginum Veritatis occultæ, exhibens Symbola, Hieroglyphica, Emblemata, Ænigmata (1650), Steffan Michelspacher’s Cabala, speculum artis et naturæ, in Alchymia (1654) and James Corss’s [Ouranoskopia], or, the Contemplation of the Heavens, in a perpetual Speculum, or general Prognostication for ever (1662; first word in Greek characters). Which of these is the closest to Giraldus’s Speculum is not immediately obvious, as in certain respects most of them touch on elements in A Vision, whether it is the array of exemplary lives, the esoteric geometry, the survey of world history or the emblematic gyres.

Yeats’s A Vision is not the Speculum of Giraldus, but the angels and men of Giraldus’s title are Yeats’s Daimons and men, and in many ways A Vision is the Daimonic mirror of human lives. It is analogous to the spirit vision of John Dee’s ‘black scrying stone’ (AV B 23), so that humanity is viewed with an abstracted, spiritualised and strange perspective, which Yeats also compares to the way in which artists such as Wyndham Lewis or Constantin Brancusi view reality (AV B 25). The speculum also embodies the duality of the System that Yeats proposes, where one form is inevitably mirrored in its opposite form in the duality of the tinctures, and the ideal world of the Principles is transferred or reflected into the phenomenal world of the Faculties through concave and convex mirrors (AV B 187). In an alternative formulation of the central ideas of A Vision, the ‘Seven Propositions’, Yeats also considers that timeless and spaceless ‘Spirits reflect themselves in time and space’ when they incarnate, so that the ‘horoscope is a set of geometrical relations between the Spirit’s reflection and the principal masses in the universe and defines that character’ (see Astrology and Fate): ‘Mirror on mirror mirrored is all the show’ (‘The Statues’, VP 610).



Return to Contents

Go to Site-map

Text and original images copyright © Neil Mann.