What magic? Where magic? imposition of the occult. Pt1- Wilfrid.

Header – detail from Brit.Lib. MS Egerton 821 f.41v.
Two previous posts
  • Ending the “skies above/certain measures” series. (May 31, 2021)
  • New Voynich research (May 24, 2021)


wilfrid_voynich1Wilfrid Voynich dated the manuscript (Beinecke MS 408) to the latter part of the thirteenth century. He ascribed both composition and inscription for the whole of its content to Roger Bacon, and for no better reason than that he supposed the pictures were about ‘natural philosophy’ – however Wilfrid understood that phrase – and with better reason because the manuscript’s materials looked to him like something from the thirteenth century.

But in his talk of 1921, Wilfrid never said that the pictures resembled any from a book about magic, nor that Bacon had practiced or approved of magic, but only that Bacon had been accused of practicing ‘black magic’ when practicing ‘science’.

magic Bacon

from: Wilfrid M. Voynich, ‘A Preliminary Sketch of the History of the Roger Bacon Cipher Manuscript’, Transactions of the College of Physicians of Philadelphia, Vol. 43 (1921). p.415. 

Note Wilfrid’s saying ‘misrepresented’ – which is absolutely right. And of course, since the materials and form for the manuscript were not incompatible with productions from thirteenth century Europe, it never occurred to Wilfrid to look beyond it.

Wilfrid’s forward-driving, unchecked and associative style would set the pattern for the sort of logic that would, from 1912 onwards, be the typical style of quasi-historical theories about the manuscript.

You see how Wilfrid’s mind grasps at some impression of ‘nearest-fit’ for the images; this he then experiences as ‘recognition’ of something familiar despite being unable to read any, and from there he develops an irrational chain that runs .. “If then … then… then … then”…

One need look no further than O’Neill and the ‘sunflower’ theory to see how the pattern applies.

Impressions are all very well as a first phase of investigations and, within the compass of his/her own specialisation, a trained person’s impressions are often accurate, but experts routinely double- check impression against concrete examples and primary historical evidence. With a strongly self-critical eye!

Wilfrid’s ‘historical logic’ reads like someone who has misread a question in arithmetic and so argues that, ‘Given that 2+2=5. so then… and therefore.. and so probably… and therefore certainly…

We can ask questions of Wilfrid, but never answer them, such as – what exactly did he think “natural philosophy” meant? or ‘Did he ever have solid evidence to inform his ‘historical logic’? The only reference he mentions is a dictionary of sixteenth-century biographies.

Natural History and Natural Philosophy

In Latin Europe, until the twelfth century, ‘natural philosophy’ is closer to what we’d call natural history and comes down to the herbals, bestiaries and lapidaries and basic knowledge of the constellations – all of which might be taught as moralia. So when Wilfrid speaks of an encyclopaedic ‘variety’ of subjects, this could be what he meant. We’d call it a form of ‘natural history’. The first encyclopaedic work in the Latins tradition was Isidore of Seville‘s Etymologiae, compiled early in the 7th century AD but as we learn from such 12thC writers as Hugh of St.Victor, the encyclopaedic method existed as part of the art of memory before encyclopaedic writers such as Albertus, Bacon, Pierre d’Ailly or Peter Lombard began writing.

On Hugh of St.Victor and the art of memory, I recommend Mary Carruthers‘ works, beginning with

  • The Craft of Thought: Meditation, Rhetoric. and the Making of Images. 400–1200. Cambridge: Cambridge University Press. (1998).
  • The Book of Memory: A Study of Memory in Medieval Culture. New York: Cambridge University Press. 1990. (Second Ed. Cambridge: Cambridge University Press. 2008.) First edition was published in 1980.

From about the fourteenth century, and within university environments, ‘natural philosophy’ would gradually become little more than commentaries on Aristotle.

To argue, in the twenty-first century, that either sort of ‘natural philosophy’ informs the content of the Voynich manuscript one would have to address the fairly obvious objection that so far the Voynich manuscript has found no comparison in any copy, text or extant notebook from any fourteenth-century or early fifteenth-century university student or teacher.

re illustrated student notebooks in general. The closest comparison presented for the Voynich manuscript, in very general terms, is an illustrated notebook which was brought to notice by Marco Ponzi. Ponzi cites the manuscript as Pistoia Biblioteca Forteguerriana Manoscritti A 33 and describes it as made by a fifteen-year old named Sozomeno, under the tutelage of a teacher from San.Geminiano. The drawings are not closely similar, but are placed in the margins and they embody in allusive and associative ‘mnemonic’ form the content of the written text.

(I regret that Marco Ponzi does not publish for the public at large, and will permit or deny any given reader access, so there’s little point in offering a direct link to his essay in Viridis Green. I do recommend reading his work, though, if you can.)

On the shifting emphasis and definition of ‘natural philosophy’ in Latin European learning, and the divide between medieval and modern phases, see

  • ‘Natural philosophy, medieval’, Routledge Encyclopaedia of Philosophy. (entry by
    Edith Dudley Sylla). see also the special edition of Vivarium, Vol.35, No.2 (1977) ‘Roger Bacon and Aristotelianism’ especially
  • Jeremiah Hackett, ‘Bacon, Aristotle, and the Parisian Condemnations of 1270, 1277’ (pp.283-314).

For our needs, the key point is that ‘natural philosophy’ was never a euphemism for magic or occult practice, even if some attempted to gain weight by attributing their content to such figures as Solomon, Aristotle, or Hermes tresmegistus – and were later to include Roger Bacon’s name. Magicians’ borrowed plumes were something Bacon himself protested. The following comes Thorndike:

Thorndike on Bacon's naming falsely attributed texts

Roger Bacon detail from WellcomeIf Wilfrid had wanted to suggest that the Voynich manuscript’s content was about occult matter, he would hardly attribute it to Roger Bacon,

Because he believed it was by Bacon, he was at least consistent in saying only that Bacon had been accused of ‘black arts’ – but not that the manuscript included magic.

‘Natural history’ is not ‘natural magic’.  Nor was ‘natural philosophy’.

At the same time, Wilfrid did try to invert the normal logic of cause and effect, insinuating – not arguing – that because occult matters were (in his view) a late sixteenth-century pre-occupation in Prague, such matter might in some way be back-projected onto the manuscript which he, himself, insisted the autograph of an English Franciscan who had died in c.1220. It’s an outrageous bit of manipulation, but one which had continuing affect in the manuscript’s study.

We know, today, that samples from four folios in the top eleven quires returned a radio-carbon range of 1404-1438, so we can discard the ‘Bacon autograph‘ idea, and (of course) that back-projection of magic in Rudolfine Prague.

Rudolf’s great-great-great grandfather* might have been born when the manuscript was made.

*Frederick III. born 1415. Frederick III, Rudolf's great-great-great grandfather


Laying aside the inclusion of Bacon’s name in rote lists of ‘ancients’ in later magical works, Molland reports that..

.. our major legendary sources are reduced to essentially two. The first is a prose romance written probably in the late-sixteenth century and entitled ‘The Famous Historie of Fryer Bacon. Containing the Wonder full Things that he did in his Life: also the Manner of his Death; with the Lives and Deaths of the two Coniurers, Bungye and Vandermast. Very Pleasant and Delightfull to be Read’. This work, which I shall hereafter call the Famous Historie, formed the basis for Robert Greene’s play ‘The Honorable Historie of frier Bacon, and frier Bongay’, and the play contains no new legendary material of interest.

The second source is much earlier, but much shorter. It occurs in a recital of deeds of Franciscans written in Dubrovnik in 1384-85 by one Peter of Trau. In this Bacon is not explicitly spoken of as a magician, but as one who was more interested in performing experiments in real philosophy than in writing or teaching.

Nevertheless the deeds recounted are of a type that would later be termed magical. Both these accounts probably had a strong basis in oral tradition, and we may suspect that the uncertainties of orally transmitted stories formed the background to the volte-face made by the bibliographer John Bale.

In his Summarium of 1548 he [Bale] described Bacon as a ‘juggler and necromantic mage’ who was said to have performed great marvels at Oxford ‘not by the power of God but by the operation of evil spirits.’ But about ten years later, in his ‘Catalogue’, Bale wrote of Bacon, ‘He was possessed of incredible skill in mathematics, but devoid of necromancy, although many have slandered him with it”.

from: A.G. Molland, ‘Roger Bacon as Magician’, Traditio, Vol. 30 (1974), pp. 445-460

In sum: it looks as if the ‘occult content’ story is one of the few persistent legends that cannot be be attributed to the talk Wilfrid gave in  1921.

Instead, to discover its origin, we must turn to the talk delivered on the same occasion by Professor Romaine Newbold who, better informed about medieval history and more familiar with primary sources, associated Bacon’s ‘natural philosophy’ with Aristotle and experimental science.

What he might have thought or written had he first looked more critically at Wilfrid’s “Bacon-wrote-science-in-cipher” proposition, we’ll never know. His principal error was the same as came to infect study of the Voynich manuscript to the present day – he adopted his ‘givens’ without first subjecting them to rigorous cross-examination and imposed what he knew about his ‘given’ – a Roger Bacon ciphertext – onto the manuscript.

I’ll survey his paper in the next post.

The skies above Pt.2 ‘asteriskos’

In a media-savvy age, many readers will know that meaning is context-dependent.

“Everything’s fine. Just realised an ancient Egyptian deity was really  Macedonian. Me, in fact.”

This doesn’t mean we may haul the manuscript and its content into whatever context we find comfortable and then assert its meaning is whatever we, and ‘people like us’ find most agreeable.    My saying so may seem trite.. but don’t be fooled.

Hauling the past into a present social environment, to make easier the task of co-opting and re-interpreting it to suit self-and-friends has been a perennial activity, probably since human society began, and even more with images than with words.

Lewis Carroll’s  Humpty Dumpty offers an apt example:

When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.” “The question is,” said Alice, “whether you can make words mean so many different things.” “The question is,” said Humpty Dumpty, “which is to be master—that’s all.”

I begin with the diagrams entire though the usual habit has been to discuss the central emblems without reference to the rest.

Structure: format

There are present eleven diagrams of similar design in Quires 10-12.  Each diagram of the eleven occupies the equivalent of one page. By ‘page’ here I mean one face of the four in a standard bifolio of this manuscript.

All are accommodated within the range “folio 70v(part) to 73v” because Quires 10 and 11 are on lengths of vellum folded to suit the ordinary bifolios’ dimensions, and having a width equal to the ordinary bifolio’s length.

There is no indication of subsequent trimming for re-binding, and no record appears to exist of any ruling out or pricking for these or for the normal bifolios. Nor have I seen any reference to pressure-marks of the sort which might indicate use of frame and wire.

It takes little acquaintance with medieval manuscripts and their techniques of construction to realise that this absence of ruling out, or evidence of its erasure, is either an anomaly or (rather strangely) an omission from the Beinecke catalogue and from all other descriptions.

In addition, the corpus of Latin Christian manuscripts appears to contain nothing comparable to these ‘fold-outs’ – neither in their design or in the way the pages are folded..  Quire 11, in particular, uses a form of concertina fold, a type characteristic of small  sleeve calendars in Latin Europe and otherwise chiefly characteristic of Asian books in direct descent from palm-leaf books and Buddhist works on paper.

All eleven diagrams follow the same general scheme, viz:

(detail) f.70v (part).

A number of concentric circles presenting as wider and narrower bands, with the narrower occupied by script and the wider inhabited by discrete anthropoform figures, chiefly female, the majority on the ‘March’ folio being provided with a roughly cylindrical container and an element that immediately strikes the modern viewer as resembling a star held by a flexible cord, or a flower on a lax stem… or both at once. (I’ll come back to this).

Each diagram’s centre contains an emblem, none of which appears in any other diagram.  At some later time a different hand added to the centres the name of a month, some repeated but not exactly: they are not ‘replicated’.

I’d like to comment on this fact, for throughout the manuscript, and most unexpectedly with regard to great number of such anthropoform figures, which occur not only in these diagrams –  care has evidently been taken to distinguish each from all others, and this has been done with a degree of subtlety (or perhaps delicacy) which argues against the usual idea that the figures are just badly drawn.

(detail) folio 72r-ii

We do see one two emblems in which the creature has its ‘pair’,  but even there it is not exact: not a   ‘mirror image’and the painter (see the example shown right) has also emphasised that they cannot be confused for one another, or supposed to ‘replicate’ the other.

I consider this another item indicative of cultural mores mutual between the persons who first enunciated the ‘ladies’ folios, and the person who later added the month-names.  Of those month-names, some appear twice, yet each  is again distinguished by some small detail not jarring to the viewer, but again avoiding ‘replication’.

I’ll include bibliography for the ‘replication’ issue, in early medieval Cairo and in medieval Byzantium at the end of this series.  Anyone wanting the references sooner is welcome to email)


Structure (cont.)

Concentric circles are a near-universal convention for depicting the heavens and that technique informs many different systems employed for doing so, but diagrams of such structure have many other applications and need have no necessary reference to the heavens.  The eleven diagrams’ taking such form may, however, cause a modern viewer to feel an immediate (‘gut’) feeling the diagrams allude to stars, or astronomy or astrology and so over-ride consciousness of the gulf between subjective and objective certainty.

‘Asteriskos’ – questions unasked.

(detail) f.108r

The ‘flower-like’ stars, or ‘star-like’ flowers given most of the tiered figures has been another trigger for that reflexive reading, unchanged since the 1920s.

Appearance of a similar motif in the margins of Quire 20  naturally then raises questions of common subject-matter and  whether some direct connection was intended between the diagrams and this later section.  That is, whether the motif was intended to carry similar significance.

Quire 20 itself is not formed in the usual way of quires in Latin  European manuscripts. It is, or was, a septenion.

A quire of seven bifolios is rare in Europe but was fairly standard in the Arabic-speaking world a.  It was also seen in Irish manuscripts.

Update – 22 June 2021.  I should add that a septenion is not unknown in European works and I’ve noted one from the first half of the fifteenth century and which includes a letter to Poggio Bracciolini is in a manuscript that is (or was in 1905) at Harvard.  For an intense codicological analysis of that ms and its classical texts see
  • E. K. Rand, ‘A Harvard Manuscript of Ovid, Palladius and Tacitus’, The American Journal of Philology , Vol. 26, No. 3 (1905), pp. 291-329.

I made these points earlier:

  • D.N. O’Donovan,  ‘Expert Opinion: Myth vs Materials Science Pt3’, Voynichrevisionist, April 26th., 2019.


Again, certain Arabic texts are found in which a form of  flower-like “asterisk”  separates sentences.  I hope readers won’t mind that the illustrations come from the same 12thC treatise on theriac which I mentioned earlier in posts to Voynich Imagery (in a series entitled,  ‘Theriac:  rosetta stone?’)

The header (above) and the detail (below, left) are taken from that manuscript: .

As text-mark, the ‘asteriskos’ is attested from Hellenistic times, though we have no example of the earlier form.  In 3rdC AD Egypt, the works of the Christian philosopher, Origen, see him use and possibly invent the form for it which then passed in the context of authoritative Christian manuscripts to 6th-7thC Spain, and saw Origen’s version maintained within the Latin manuscript tradition where it appears as a vertical or a diagonal cross having dots set in the interstices.

The Latins’ ‘astericus’ did not have the form of a flower; nor what we might call  a ‘star-shape’. In short, it was never formed as are these motifs in the Voynich manuscript


Below is illustrated a detail from a 12thC  copy of Isidore’s Etymologies.  However, where he had included the Greek term in Greek letters, this copy romanised and then translates the Greek as ‘stella’ just as numerous other and later copies do, even while  keeping Origen’s and Isidore’s form for it all but unchanged.  In other cases (as the detail upper left), the motif served as virtual ornament and filler, while still expressing the asterisk’s significance, by showing that the half-filled line was not blank unintentionally; i.e. an intentional ‘omission’ of written text.

Significance in the Hellenistic and in the Latin traditions.

The oldest references to use of the ‘asteriskos’ are Hellenistic and show it marked places in a text where some item or passage had been duplicated. No examples remain.

Origen used the ‘asteriskos’  to mark points where he  re-inserted a passage from the Hebrew left untranslated by the Septuagint.  He did not replace the earlier text with his own revised version, but added his own below, marking both with his cross-shaped asterisk. In that way, the sign still signified ‘duplication’ but now, equally,  ‘omission’ – and the latter sense became the default in medieval Christian Europe.

There is another and deeper level of significance for the ‘asteriskos’ in Christian traditions. It is touched upon. lightly, by Isidore whose comment runs (in English translation):

“The asterisk is placed next to omissions, so that things which appear to be missing may be clarified through this mark, for star is called in Greek, ἀστήρ and the lLatin] term asteriscus is derived from this. ” 

  • Isidore of Seville, Etymologiae Bk I xxi.2  ‘De notis sententiarum’ (De Critical signs)

What Isidore  touched upon lightly, a later Christian authority, Jerome, would expand upon – and  “F.M.P”  has rightly drawn attention to the fact:

  • Anthony Grafton, Glenn W Most, Salvatore Settis, The Classical Tradition (2010).

The asterisk as  illuminating’ what was absent – light into the darkness – impressed the medieval mind. It will offer another and natural link to associations of ‘stella maris’.

But while I’m quite prepared to accept that the Voynich ‘star-flower’ (as signe de renvoi) might have been intended as  link-and-key to text in Quire 20, and even evoke intentionally that sense of   ‘lights in the darkness’ I cannot accept that the sort of men who knew how to employ their version of the ‘asteriskos’ in Christian texts – clerics by definition during most of the medieval period – would have ever created such figures as those ‘ladies’ set around the diagrams.

And of course, the Latin form for the asteriskos, though it had variations was always the dotted cross or X.  It came, in the Latin west, to serve as signe de renvoi, indicating a link between marginal text and main text, but adjacent, not separated by a number of quires.  A comparable practice is (so far as I’m aware) unattested before the introduction of printing, and even then not immediately.

On this see e.g.

for which example and other details I’m indebted to

  • Yin, ‘Asterisks in the Middle Ages’, medieval codes (August 5th., 2014)

I have found only one source where there is so much as a hint that any type of ‘flower’ form was used in any comparable way in a medieval text.

I owe that hint to:

  • Lori J. Walters, ‘The Rose as Sign: Diacritical Marks in the Tournai Rose [TOU]’. In: Revue belge de philologie et d’histoire, tome 83, fasc. 3, 2005. Langues et littératures modernes – Moderne taal en litterkunde. pp. 887-912;  doi : https://doi.org/10.3406/rbph.2005.4948   https://www.persee.fr/doc/rbph_0035-0818_2005_num_83_3_4948

The manuscript discussed by Walters dates to the 13thC, and is a revised version of the ‘Romance of the Rose’ devised by Gui de Mori in 1290AD.  It may be available to view if you’re in Tournai:

If anyone sees that manuscript, and whether its ‘rose’ mark has similar form to the Vms’, or similar purpose as the Latins’ asteriskos, I hope you’ll let me know.

  • Alan Scott, Origen and the Life of the Stars: A History of an Idea (Oxford Early Christian Studies)

Tentative conclusion (A)

Apart from that possible exception in TOU, the Latins did not have the linguistic frame [aster-asteriskos], nor did they have the requisite habits in marking texts,  to have invented such a form as the Voynich ‘flower-stars’ to serve in place of the traditional ‘asteriscus’.  Nor does there appear to exist any comparable use of their cross-shaped ‘astericus’ to link diagrams in one section of a manuscript with text separated from it, as is now the case in the VMS,  by several quires of other images and text, amounting to tens of pages.

The Latins’ habits in making manuscripts does not encourage the idea that,  if the diagrams’ star-flowers were meant as cue to Quire 20, that the materials’ first enunciation occurs in medieval western Christian Europe. The one possible clue to any exception is yet to be sighted.

Islamic works of the thirteenth century use a ‘flower’ form but it resembles a rose and not the Voynich manuscript’s spiky ‘aster-‘ whose drawing is consistent in the diagrams and  Quire 20, arguing a set habit.   However, the fact that the Voynich motifs have the same form in both sections doesn’t itself prove shared reference or common significance, here or in any posited  exemplars.

Something that I’m inclined to think might prove relevant, should anyone wish to take up the question for research, is that the motif in Quire 20 sees a conscious alteration of the centres, and this seems to me a possible echo of the way (as in the same 12thC ms cited earlier), ‘marginal’ passages of text were differentiated in Islamic manuscripts. I suggests consonant habits of mind.

What does not accord with the idea of origin for the Vms’ diagrams in works of the Islamic corpus is first, the ease with which the Voynich drawings equate flower-form, star-form and that hint of ‘light in the darkness’. It seems to me to imply familiarity with the Greek, and specifically if indirectly with the author of a ‘Theriac’ text, the Hellenistic poet Nicander who lived at just the time we first hear of Hellenistic use of the   ἀστερίσκος (‘little star’), whose form is unknown but which is attested in use, to mark duplication, by the 2ndC BC – precisely when Nicander lived.

It is also my opinion that much in the Voynich manuscript’s imagery points to first enunciation in the Hellenistic period, but I won’t elaborate on that except to refer to the form of the diagram’s unclothed female figures.


In his exhaustive paper on the history ot the Aster. Burgess quotes a fragment from Nicander, author of  ‘Theriaca’,

  • Edward Sandford Burgess, Studies in the History and Variations of Asters —Part I (etc.),  Memoirs of the Torrey Botanical Club, Vol. 10  (22 November 1902), pp. iii, v, vii-xii, 1-5, 7, 9-93, 95-447.

When the flood waters receded, all that was left around the two mortals was mud and slime. Astraea felt so sorry for them she wept, her tears. upon hitting the earth, turned to star-flowers (asteriskos). Sea=aster.

The ‘aster’ which Burgess, and all subsequent commentaries on this fragment of Nicander’s poetry have assumed meant is the purple-flowered Atticus, but I feel some doubt on this score, for Nicander calls it ‘luminous’ and there is a white, spiky-petalled ‘star-flower’ whose centres alternate in colour.  It is has been known over time as Aster pannonicus and  Aster tripolium, and its present description is  Tripolium pannonicum.  The oldest remaining version of the Greek legend of Astraea seems consciously to conflate the high- and the low dwelling types (see picture and caption, left).

Note: The plant we now call ‘Hellenium’  is arelated to the sunflower and native to the Americas.  The ancient ‘Helenium’ was Elecampagne, something known to European writers  by the early 17thC. as e.g. The General Practise of Physicke … Translated and Augmented by J. Mosan. B.L. (1605) p.817.


Nicander of Colophon, Aratus of Soli. Theriac, medicine and Stars.

coin of Soli in Cilicia.

Nicander was born in Claros, near Colophon on the Ionian coast of Asia Minor in 197 BC.  Though he may have known Aratus’ work well, or the material from which Aratus drew, Nicander’s  ‘Theriaca’ mentions stars only as markers of time; warning that this or that season was when a noxious creature was most numerous, most active or most highly venomous.  He died in young, in 170 BC.

Aratus had been born and died long before, being born in Soli in 325 BC and having died in Pella, of Greece in 240 BC.

Despite these disparities the two came to be often mentioned together, due to a witticism which commented on the fact that Aratus wrote about the stars but had no (practical) knowledge of them, while Nicander wrote on medicine and had no knowledge of pharmacy.

Cicero (de Oratore i.69) repeats the usual parallel but tries to a put a good spin on it – along the lines of ‘why shoudn’t someone learn only as much as they need for a present occasion?’

And as a last thought – the material copied to make the Voynich manuscript had to be copied because useful to someone; since the quires appear to have been bound by, or for a Latin, so presumably the information was perceived as useful to them.  It occurs to me that by translation from ‘aster” to ‘stella’ the sea-aster offered a natural association of ideas with the ‘stella maris’ which of course might mean the Pole star (which medieval Latins continued to associate as often with Cynosura as with Polaris), but it might also refer to the magnetised needle and surrounding compass (card).

Just a thought.


Status of the problem so far ….

The seemingly natural connection made by the first enunciator f these diagrams: ‘flower-star-textual asterisk’ implies, if intentional,  close familiarity with Greek.  .

The medieval Christian works, the ‘asterisks’ don’t take this form – certainly not before the 13thC  and even then there appears to be only one comparison known – the TOU.

On the other and each in their several ways suggest the aster- motif in both the diagrams and Quire 20 are likely to relate to the sea-aster whose range today is believed much reduced from what it was in the centuries BC, but still inhabits salt-marsh environments.

At the moment the three avenues offering strongest possibilities of responding to investigation are, in order of chronology  (1) Asia Minor, Hellenistic Greek. (2)  Arab-Persian c.12thC AD or (3) 13thC French culture ‘[TOU}.

If common significance could be proven between Quire 20 and the diagrams, the contextual range would be more limited – which is all to the good.

The first printed edition of NIcander’s medical work issued from the Aldine Press in 1499.  (In Greek). You can see a copy of that edition here,

.. but that’s how we begin.

Note – August 20th. I have corrected minor typos and improved the grammar of half a dozen sentences. The reader is asked either to excuse mis-typings which remain, or leave a comment below.