Sunday, May 16, 2010

Part 8: Islamic Contribution to Science

THE ARABIC ORIGIN OF THE SUMMA AND GEBER LATIN WORKS:
A REFUTATION OF BERTHELOT, RUSKA AND NEWMAN ON THE BASIS OF ARABIC SOURCES
AHMAD Y. AL-HASSAN

 ABSTRACT

This paper presents a reassessment of the Geber Problem on the basis of research into the extant Arabic works of Jābir ibn Ḥayyān and other Arabic works that incorporated his ideas. Part I discusses the hypotheses of Marcelin Berthelot which heralded the problem and texts from the Summa and Arabic sources are compared, and thus the Arabic identity of the Summa is confirmed.
Part II refutes the assumptions of Julius Ruska about a Latin author for part of the Liber Geberis De Investigatione Perfectionis Magisterii of the Riccardiana manuscript, and for the Summa. It follows that all the assumptions of William Newman that were built on Ruskas speculations, about a previously unknown compiler called Paul of Taranto as the author of the Summa are baseless. 

INTRODUCTION
Geber was the name ascribed to the author of a series of alchemical treatises which began to appear in the Latin West in the middle of the 13th century. These treatises included the Summa Perfectionis Magesterii; De Investigatione Perfectionis ; De Inventione Veritatis; Liber Fornacum and Testamentum, which were usually printed together between the 15th and 17th centuries. These were known until the nineteenth century to be translations of works originally written in Arabic by Jābir ibn Ḥayyān. His name, in the Latin form Geber, became widely celebrated. The Summa was so successful that, according to George Sarton, it became the main chemical textbook in medieval Europe, and its author was called the father and founder of chemistry by some Western historians.
This attribution to Jabir was not challenged until the end of the 19th century. In1893, Marcelin Berthelot claimed in his work La Chimie au Moyen Âge  that these treatises had been written by Latin authors who would have used Jābirs name in order to facilitate the diffusion of their own works. Berthelot was a noted scientist and a public figure, and as a high official he was most influential in France. He outlined his reason in writing his history of chemistry in the introduction of volume three. He said that it is necessary that we radically change the current ideas about the chemical knowledge of the Arabs, and on the influence exerted by this knowledge on the civilization and science of the West.
However, several eminent historians of chemistry and alchemy raised serious objections to Berthelots assumptions. The earliest appeared in 1905 by Henry E. Stapleton, while the largest and most consistent objections were raised by Eric J. Holmyard in a series of papers published between 1922 and 1928. James .R. Partington sided with Holmyard, whereas Lynn Thorndike would further question Berthelots accuracy and judgments.

Notwithstanding, in 1935, Julius Ruska attributed the authorship of a part of Liber Geberis De Investigatione Perfectionis Magisterii of the Riccardiana manuscript to a Latin author, who would be also the author of the Summa.

In 1986, William Newman adopted Ruskas assumptions and he attributed the Summa to a previously unknown writer by the name of Paul of Taranto.
In this way, although the Geber Problem is more than one century old and in spite of the definitive judgments by Holmyard and other scholars, the assumptions of Berthelot, Ruska and Newman are still adopted uncritically by Western historians of alchemy.
We have dealt with some of Berthelots assumptions elsewhere and will not be repeated here, thus in Part I of the present paper we make a brief summary of our refutation of these assumptions, and continue with a discussion of the remaining ones.
In Part II, we dispute Ruskas speculations in his study of the DIP and his unfounded assumption that a Latin author wrote part of it. We shall discuss also William Newmans assumptions which were built on Ruskas speculations and which culminated in the conjecture that an unknown compiler called Paul of Taranto was the author of the Summa.
In this way, we hope to contribute in bringing to light the deliberate errors on which the early history of Latin alchemy is built. 

REFUTATION OF MARCELIN BERTHELOTS ASSUMPTIONS
Berthelots main claims for Latin authors of Gebers works can be summarized in the following assumptions:
1. The treatises carrying Jābirs name were written by Latin authors, who attributed their work to Jābir due to his high standing in the West.
2. There are no Arabic originals for these same works.
3. The style in the Arabic works by Jābir is vague and allegoric.
4. The style of the Summa recalls the style of the Schoolmen.
5. The Summa is devoid of Muslim expressions, which are extravagant in the Arabic texts of Jābir.
6. The Summa contains an account of the arguments against transmutation, which is not existent in Arabic works.
7. The Arabic works of Jābir do not contain practical recipes for the preparation of materials.
8. The minor Latin works bearing Gebers name mention more modern materials, such as saltpeter, salt of tartar, rock alum and feather alum, as well as the preparation of nitric acid, which are absent in the Arabic works of Jābir.
9. The Arabic works do not mention the sulphur-mercury theory of the generation of metals, nor the three principles in metals sulphur, arsenic and mercury.
We shall now discuss these assumptions in the same order:

1.  Jābirs Hypothetical High Standing in the West
Before the translation of Arabic works into Latin, alchemy was unknown in the West. Robert of Chester finished in 1144 the first translation from Arabic of a book on alchemy Liber de Compositione Alchimiae. In the preface he states, Since what Alchymia is, and what its composition is, your Latin world does not yet know, I will explain in this present book. Between this and 1300, some major Arabic alchemical works were translated into Latin, including Tabula Smaragdina, Turba Philosophorum, The Secret of Creation of Bālīnās, De Perfecto Magisterio, attributed to Aristotle, De Aluminibus et Salibus and the Liber lumen luminum by al-Rāzī, parts of Kitāb al Sabʿīn (The Book of Seventy) by Jābir, and possibly De anima in arte alchimiae attributed to Ibn Sīnā (Avicenna).
The Book of Seventy that was partially translated by Gerard of Cremona in the 12th century, does not carry Jābirs name. Most Latin authors believed it was a work by al-Rāzī, and the actual author remained unknown until the end of the 19th century. Other than this, we do not know of any other work by Jābir which was translated into Latin before the middle of the 13th century.
The alchemical works of the thirteenth century that were written by Latin authors such as the works of Michael Scot (1175-1233) and of Vincent de Beauvais who wrote his speculum works between 1220 and 1244, quote numerous Arabic authors, but Jābir (Geber) is not among them. For Albertus Magnus, the only authority in alchemy was Ibn Sīnā, whereas Roger Bacon did not mention Geber (Jābir) also although he was acquainted with alchemy through Latin translations of Arabic works.
Therefore, as Jābir was not known in the West in the 13th century, there is no reason to suppose that any Latin author would attribute his work to him. On the other hand, according to Roger Bacons appraisal of the status of alchemy at the end of the century, it would be highly impossible for a Latin writer to compose such a considerable and mature corpus of alchemical knowledge: But there is another science which is about the generation of things from the elements [] of which we have nothing in the books of Aristotle; nor do natural philosophers know of these things, nor the whole Latin crowd of Latin writers.

Translator of Liber fornacum
Furthermore, there are frequent cross-references between the Summa and the Liber fornacum. It was possible to establish that the latter is a translation form the Arabic, and we currently know the name of the translator and the place and date of the translation. This fact is of utmost importance and it is sufficient in itself to demolish the assumptions of Latin authors for Jabirs Latin works. It is indeed bewildering why historians of chemistry and science kept silent about it.

2- Lack of Arabic Originals
We have surveyed all the extant dated Arabic MSS attributed to Jābir. The oldest ones (2%) do not go earlier than the 12th century. This is to say, all MSS by Jābir which preceded the 12th century have perished and, among them, most probably also the ones used by translators. All Arabic MSS were written on paper which deteriorates with the passage of time and the factors of the environment, and not on parchment which was the only writing material in the West before the advent of printing. 
On the other hand, we should remember that the Arabic originals of many significant Latin translations of Arabic scientific and philosophic works were also lost, surviving exclusively in Latin or Hebrew.

3- The Allegorical Style of Jābirs Arabic Works
Jābirs alchemical and chemical works may be classified in two groups. The first includes writings on the Art of alchemy, while the second consists of numerous treatises on practical alchemy and industrial chemistry. Berthelot selected for his analysis works belonging exclusively to the first group. This was already noticed by Holmyard: [Berthelot] deliberately wanted to underrate Jābir [], the choice of Jābirs works made by Berthelot is entirely misleading.
 
4- The Style of the Summa Recalls that of the Schoolmen
Jābir was a philosopher and according to al-Fihristhe wrote numerous works on philosophy. More recently, Paul Kraus was able to list 23 titles for Jabir on philosophy, among which several deal with logic. In several works of Jābir there are arguments where he describes two opposite points of view and employs logic to arrive at a right conclusion. Thus, Jābir was well versed in the tools later employed by the Schoolmen.

5. Muslim Expressions
According to Holmyard, It is here that Berthelots ignorance of Arabic led him astray. As a matter of fact, the Summa is full of Arabic phrases and turns of speech, and so are the other Latin works.
Our study of the Summa confirms Holmyards assertion. Indeed, it retained several Islamic expressions of praise to God, mostly of Quranic origin. Furthermore, there are also well known Arabic sayings. For instance in De Investigatione, Contraries set near each other are the more manifest وبضدها تتميز الاشياء; Haste is from the Devils side العجلة من الشيطان.

6. Arguments for and Against Transmutation
Debates regarding the validity of al- Ṣanʿa (The Art) and the possibility of the transmutation of base metals into gold began with the inception of Arabic alchemy itself. Throughout Jābirs works references are found to the need to defend the Art against those who denied it. Jābir systematically warned his readers to be aware of them and gave instructions on how to confront them. More specifically, he wrote two treatises devoted to the subject: Al-Burhān wa ithbāt al- Ṣanʿa (The Proof and the Verification of the Art) and Kitāb al-thiqa bi ṣiḥḥat al-ilm (The Book of Confidence in the Truth of Science).
After Jābir, the debate continued unabated. Al-Jāḥiẓ (ca.781-868) was not convinced of the validity of the Art and al-Kindī (ca.801-873) wrote Kitāb ibṭāl dawā al-muddaʿīn anʿat al-dhahab wa al fiḍḍa min ghayr maādinihā (A Refutation of Those Who Pretend to be Able to Win Gold and Silver Otherwise than from Ore). His contemporary, Ḥunayn ibn Isḥāq (808-873) was also opposed to alchemy. Their attacks were challenged by al- Rāzī (865-925) in his book Kitāb ithbāt al- ṣanʿa wa al-radd alā munkirīhā (Book of Confirmation of the Art and Refutation of Those Who Deny it). The debate continued well into the 14th century.

7. Recipes for the Preparation of Materials
Berthelot assumed that Jabirs works are devoid of recipes for the preparation of materials. A survey of 59 MSS by Jābir on practical alchemy shows the description of large umbers of recipes. There is a whole treatise of recipes which is Kitāb al-durra al-maknūna (The Book of the Hidden Pearl), which contains dozens of recipes on the colouring of glass, the manufacture of artificial pearls and improving their colour, and several other industrial products.  Also, Kitāb al- Khawāṣṣ al kabīr (The  Great Book of Properties), contains many chemical and industrial chemical  recipes on the manufacture and annealing of steel, the de-salination of sea and brackish water by ultra filtration,  the manufacture of  zunjufr (cinnabar) , the colouring of glass, the manufacture of pearls, several recipes on cosmetics (removing unwanted hair, dying of hair into yellow gold and dying the hands of maidens with various colours),  on varnishes and paints including waterproofing, making inks of various colours, and several other industrial products. The other books of Jābir contain also many recipes for the preparation of most of the chemical materials that were known, which include for example the making of salt of alkali (mil al-qa),  the refining of tin (raṣāṣ qalī)  of iron  and the other metals. 

8. Modern Materials
The large number of recipes described in the MSS declared above, mention all materials known to alchemists and chemists until the end of the Middle Ages. We have dealt with Berthelots assertion that saltpeter and nitric acid were first known after the 13th century elsewhere. There it was shown that saltpeter was known under various names since the beginnings of Arabic alchemy and chemistry, while several recipes for nitric acid are given in Jābirs Arabic works as well as in other Arabic treatises before the 13th century.

9. Theories of Alchemy in the Summa and in Arabic Works
Contrary to Berthelots views, the sulphur-mercury theory and the theory of three principles of metals sulphur, arsenic, and mercury arrived to the Latin West via Arabic translations. The sulphur-mercury theory was basic to Arabic alchemy. We shall discuss both theories as they were expounded in Arabic works and compare them with the texts of the Summa, together with other theories of Arabic alchemy.

a- The Two Exhalations Theory
In Arabic alchemy, smoke (dukhān) and vapour (bukhār) were considered to be the origin of metals and stones and were equated to sulphur and mercury. Although the smoke-vapour notion had started with Aristotle, the full account of their role in the generation of metals and the relation to the sulphur-mercury theory was first given in Bālīnās Kitāb sirr al khalīqah (Book of the Secret of Creation) or Kitāb al-ilal (The Book of Causes ). According to Paul Kraus, Jābir drew heavily from this source in his own works, including the two exhalations theory and the sulphur-mercury theory.
Bālīnās Book of the Secret of Creation was translated into Latin in the 12th century by Hugh of Santalla, who stayed in Tarazona from 1145 o 1151.
We compared the chapter dealing with the generation of metals in Françoise Hudrys edition of Hugo of Santallas Latin translation with the corresponding chapter in Ursula Weissers edition of Kitab sirr al khalīqah, and found them to be similar (Appendix 1). From this, it is evident that the two exhalations theory of the generation of metals and the sulphur-mercury theory were available in Latin since the middle of the 12th century and not at the end of the 13th century as Berthelot had claimed.
Vincent of Beauvais was acquainted with these theories. Lynn Thorndike asserts that in Speculum Doctrinale, Beauvais stated that:

 “…everything has an occult quality opposed to its natural one; that four spirits, mercury, sulphur, arsenic and sal ammoniac, and six metals, gold, silver, copper, tin, lead and iron are generated in the bowels of the earth; and that the metals are generated by mercury and sulphur.
For this reason Thorndike did not accept Berthelots assertion that these basic theories of alchemy were not known in the West until the Summa had appeared at the end of the 13th century. Although Thorndike did not question the authenticity of Beauvais statement, he was not sure about his source. Now, however, it is conceivable for it to be based on Hugo of Santallas Latin translation of the Book of the Secret of Creation

Comparison of the Exhalation Theory in Arabic alchemy and the Summa
The Arabic text for the exhalation theory and the text of the Summa, are reproduced in Appendix 2. An attentive reading of the Arabic and the Summa accounts shows them to be remarkably similar. Both assert that the metallic bodies cannot be generated from mercury and sulphur in their natural form (Summa) or in their coagulated form (Arabic). Both argue that natural sulphur and mercury cannot be found together in the same mine, but that each one is to be located in its own separate mine. For this reason, they should be used in the form of an earthy substance (Summa) or non-coagulated form (Arabic). Metallic bodies are, thus, formed from a double fume (Summa) or from vapour and smoke (Arabic).
This close resemblance of the Summas text to the Arabic one refutes Newmans assumption that the theory probably occurred first in the TP, from whence it was transferred to the Summa.  Indeed the TP’’s account itself is also taken from an Arabic origin.
It is quite obvious, therefore, that the account in the Summa for the exhalation theory is an Arabic one. This leads us to two corollaries, one regarding the corpuscular theory, and the other regarding the mercury alone theory.

b- The Corpuscular Theory
An assumed corpuscular theory was given great publicity by Newman and it was the main theme of at least one academic conference, the proceedings of which were lavishly published by Brill of Leiden. Newman thought that this theory was first propounded in the Summa and that it was a theory of Paul of Taranto. However, this so-called corpuscular theory in the Summa is nothing but the same two exhalations theory already discussed. Nonetheless, it is worthy to remind that this Aristotelian concept had been elaborated by Bālīnās in his Book of the Secret of Creation, which was one of the basic sources for Jābirian alchemy, and the basis of the sulphur-mercury theory.
To give special prominence to the alleged singularity of this theory, Newman chose the word corpscule to translate the Latin pars, in the stead of part as Russell had done. Nevertheless, the words pars, part and corpscule are translations of the same Arabic word juz.
Newman also attached particular significance to the degree of packing of the parts of a metal, as such packing affected its weight and its proximity to perfection. This same packing (talzīz or tarzīz) of the parts (ajzā, singular: juz’’) of a metallic body occurs frequently in Arabic alchemy within the context of the two exhalations theory. A small selection from Arabic texts is presented below in order to show how Newmans corpuscular theory is an old concept in Arabic alchemy.
Bālīnās
-         On gold: And it became heavy razīn because its parts entered into each other.
-         On mercury: It is heavy in weight and its parts entered into each other.
Jābir
-         On gold: Its parts entered into each other in an intermingling that cannot be separated and it works with them all.
-         On silver: To become gold, silver needs two things: the packing of its parts (tarzīz) and tinting.
Al- Jildakī
-         On metallic bodies in general: A condition for the removal of ailment from a metallic body is that its parts should be packed so that it acquires weightiness instead of lightness.

c- The Mercury Alone Theory
The emphasis on mercury, rather than sulphur, is based on old knowledge in Arabic alchemy. From a single sentence in the Summa, Newman assumed that this idea would have begun in the 13th-14th century. This sentence reads: And if you can perfect by Argentvive only you will be the Searcher out of a most precious Perfection; and of the Perfection of that which overcomes the Work of Nature.
This sentence appears in the Summas chapter on the nature of Venus or copper. The full paragraph reads:
 Hence it is manifest that those Bodies are of greater Perfection which contain more of Argentvive; but what contain less, of less Perfection. Therefore study in all your Works that Argentvive may excel in the Commixtion. And if you can perfect by Argentvive only you will be the Searcher out of a most precious Perfection; and of Perfection of that which overcomes the Work of Nature. For you may cleanse it most inwardly to which Mundification Nature cannot reach. But the Probation of this viz. that those Bodies which contain a greater Quantity of Argentvive are of greater Perfection is their easie Reception of Argentvive. For We see Bodies of Perfection amicably to embrace Argentvive.
This text is recommending mercury if you can. But in the Summa itself there are recipes prescribing other ingredients besides mercury. For instance, one recipe is for the solar medicine of the third order which transmutes silver into gold; here sulphur is the essential ingredient.
The importance of mercury as the matter of metals was repeatedly stated in the Arabic alchemical literature and it recurred in the Summa and also in the works of the fourteenth century Latin alchemists, and it is in conformity with the sulphur-mercury theory. Concerning the Arabic sources, the examples below will suffice:
Bālīnās
I say that the origin of all melting bodies is mercury () Mercury is the origin of melting bodies and it is the first one among them and they were formed from it.
Jābir
Mercury is the origin of melting bodies and it is their material and first object, like the sperm for animals or the seed for plants.
Jābirs comparison of mercury to sperm was repeated by Arnold of Villanova and John Dustin but does not occur in the Summa

d- The Sulphur Mercury Theory
and the Composition of Metals.
Berthelot assumed that the Arabic works of Jābir did not mention the sulphur-mercury theory.
Later, Newman assumed that a text in the TP on the differences in the constitution of metals is unique and is one of two main proofs for the relationship between the TP and the Summa. However, the account for differences in the composition of metals is part of the sulphur-mercury theory and is an essential concept in Arabic alchemy. Indeed, it is the basis on which the whole idea of transmutation is built. Gold was the perfect metal, followed by silver. The four remaining metals copper, iron, tin and lead were defective. The aim of alchemy was, precisely, to treat the defective metals in order to be brought back to the ideal composition of gold. Arabic alchemy texts give accounts of the differences among the metals in one form or another. The first account is found in the Book of the Secret of Creation of Bālīnās. Several other accounts are present in Jābirs works as well as in the works of other alchemists.
In the case of gold, the texts quoted below agree that mercury is its main constituent, while sulphur should be pure and non-combustible. Regarding other metals, the accounts by Jābir and the Summa are quite similar, with insignificant variations. Newman acknowledged that this part of alchemy was common knowledge in the 13th century. Nevertheless, he also believed that the Summa and the TP contained unique information regarding the fixedness (non volatility) and the indication of the amounts. A close look at the Arabic sources reveals that such information was not unique.
Jābir
Mercury is the origin of metals; it is their matter and their principal constituent.
 And we shall say also that all metallic bodies in their essences are mercury that was set (coagulated) by means of the sulphur of the mine that has risen to it with the vapours of the earth. And they (i.e. the bodies) have differed because of the differences in their properties; and their properties differed because of the differences in their sulphurs. The differences in their sulphurs are caused by the differences in their earths and in their positions in relation to the heat that reaches them from the sun as it oscillates in its orbit. And the finest of those sulphurs, the purest and the most temperate was the golden sulphur and for this reason mercury was coagulated with it firmly and temperately; and because of this temperance it resisted fire and it stood firm and fire was not able to burn it in the same way as it burns other bodies.

Ibn Sīnā
If the mercury be pure, and if it be commingled with and solidified by the virtue of white sulphur which neither induces combustion nor is impure, but on the contrary is more excellent than that prepared by the adepts, then the product is silver. If the sulphur besides being pure is even better than that just described, and whiter, and if in addition it possesses a tinctorial, fiery, subtle and non-combustive virtue, in short if it is superior to that which the adepts can prepare, it will solidify the mercury into gold.
Ikhwān al- Ṣafa
If mercury was pure and if sulphur was free from impurities and if their parts are comingled, and if their quantities were at the appropriate ratio then ibrīz gold will be formed after a very lengthy period of time.
Summa
Therefore, 'tis now clear from the precedent, that if clean, fixed, red, and clear sulphur fall upon the pure substance of argentvive (being it self not excelling, but of small quantity, and excelled) of it is created pure gold.
To conclude, it is clear that the constitution of metals according to the sulphur-mercury theory is the same in the Summa as it is in Arabic alchemy, from which it was derived.

e- The Theory of the Three Principles:
Mercury, Sulphur and Arsenic
One of Berthelots main hypotheses was that the theory of the three natural principles was not mentioned in the Arabic works. Newman stated similar views. This theory and the inclusion of arsenic as the third principle was Newmans second main argument to establish the TP as the source of the Summa: Let us now point out that the inclusion of arsenic among the metallic principles is not easily extracted from the Arabic sources that our texts may have used. He then concludes that The Summa and the TP share the unusual theory that arsenic must be included among the metallic principles: this further substantiates our view that dependence let us now say a direct dependence exists between the two texts.
Nevertheless, the three principles mercury, sulphur and arsenic are always grouped together in Arabic alchemical texts whenever spirits are discussed. This naturally also applies to Jābir. Arsenic was a major spirit, like sulphur, and there is extensive literature on its preparation and use in chemical operations. Thus, the conclusions of Berthelot and Newman must have been based on a lack of familiarity with Arabic sources and a very limited number of available Latin texts translated from Arabic. None knew Arabic; Berthelot relied on few texts of Jābir of the allegorical category translated for him, and Newman relied on a very small number of available Latin translations.
Jābir
And one of its principles is arsenic which has preparation, work and precious tincture; this is in addition to the high quality of this principle and its nobility.
 We have to believe also that sulphur is one of the spirits and it is necessary for the gold work; and arsenic is one of them and it is necessary for the silver work; and if arsenic is used in the gold work it will be deficient, and if sulphur is used in the silver work it will be deficient.
 
Summa
It now remains that we at present speak of arsenick. We say it is of a subtile matter, and like to sulphur; therefore it needs not be otherwise defined than sulphur. But it is diversified from sulphur in this, viz. because it is easily a tincture of whiteness, but of redness most difficultly: and sulphur, of whiteness most difficultly: but of redness easily.
f- The Three Orders of Medicines
In the Book of Seventy of Jābir, the concept of the three orders of medicines is mentioned in numerous chapters. The Summa contains complete texts  describing this concept that correspond to the texts of the Book of Seventy . Further, numerous chapters in the Summa are based on this concept also. We have discussed this topic elsewhere.

UNIQUE JĀBIR TRAITS IN THE SUMMA
Besides the discussions given above of Berthelots assumptions, we would like to close Part I of this essay by showing three unique traits of Jābirs writing, which distinguish his Arabic works. These same distinguishing features will be found again in the Summa and the other Geber Latin works.

a-  Our Volumes
Jābir wrote scores of books and treatises, for which he compiled three fihrists (indices). These are listed in the Fihrist of Ibn al-Nadīm. More recently, Paul Kraus devoted one full volume to catalogue the works of Jābir. Within these contexts, it is not surprising to find Jābir continually referring to his numerous other volumes or books. This referral became a characteristic feature of his style.
In each of the four Latin tracts, Geber also speaks of his other volumes. He declares that the Summa is the sum of what he had written in his other volumes. Certainly, those other volumes cannot possibly be the minor texts traditionally linked to the Summa. As we shall see below, Julius Ruska, followed by William R. Newman, assumed that the Riccardiana DIP was a source for the Summa. Ruska based his assumption on a single paragraph in the DIP which refers to the authors other volumes.
We conclude from all this that the expression our volumes does not apply to any of the above few Latin works. The expression our volumes, repeated in each of the Geber works, points out to an author who had written a large number of works on alchemy. Such an author cannot possibly be a pseudo-Geber, as we do not know of any 13th century Latin author who wrote so extensively on alchemy. Nor do we know of any Arabic or of any pre-Arabic author. The only known author who had composed scores of treatises and books on alchemy was Jābir, and his style is reflected also in the Latin works.

b-  The Principle of the Dispersion of Science
Paul Kraus affirms that one of the most characteristic traits in Jābirs works is his continual declaration of not having exposed the full truth in one place only, but that he had distributed the alchemical knowledge throughout his countless treatises. He constantly advises the student of the Art to collect and study his books. The Latin works of Geber also exhibit this same trait.
Jābir
Understand that we have compiled in this art many books in numerous topics and arranged them in different ways. Some were related to others and some were complete in themselves []. Each complete book is adequate on its own. As to those books that are related, each one needs the other, and no person can benefit by using them unless he gets hold of a complete collection (and) read them all and learn their purposes.
 
Geber
We declare that we have not treated of our science with a continued series of discourse, but have dispersed it in diverse chapters. And this was done; because, if it had been delivered in a continued series of speech, the just man, as well as him that is evil, might have usurped it unworthily. Therefore we have concealed it in places, where we more openly speak; yet not under an enigma, but in a plain discourse to the Artist.
 
c-  Jābirs Books of Sums
It is not rare to find similar declarations in Jābirs Arabic works, and in the Summa. The opening paragraph of the Summa is similar to the corresponding one in his Book of Seventy. Jābir distinguished between his larger and smaller books and in the preface to the former, sometimes he states that a larger book is a sum of the knowledge dispersed in the smaller ones.
Jābir
Since there appeared many books of ours on this Art that is called ḥikma (philosophy) which has no limit and is the ultimate of philosophy, it became unavoidable that we should put down a book that explains our previous abbreviated words; so we are explaining one word of a certain art by a hundred words of the same art . So that this volume <Book of Seventy> contains what was in our former and our later books.
 We have written before this book of ours several books dealing with such fundamentals like these, and all are dispersed. We have made this book of ours like the sum of those fundamentals, and arranged it in twenty parts.  

Summa
Our whole Science of chymistry, which, with a divers compilation, out of the books of the ancients, we have abbreviated in our volumes, we here reduce into one Sum. And what in other books written by us is diminished, that we have sufficiently made up, in the writing of this book of ours, and supplied the defect of them very briefly. And what was absconded by us in one part, which we have made manifest in the same part, in this our volume; that the compleatment of so excellent and noble a part of philosophy, may be apparent to the wise.
 
II.   JULIUS RUSKAS HYPOTHESIS ABOUT THE RICCARDIANA LIBER GEBERIS DE INVESTIGATIONE PERFECTIONIS  AND THE ASSUMPTION OF NEWMAN ABOUT PAUL OF TARANTO AS THE AUTHOR OF THE SUMMA

Ruskas and Newmans Assumptions

The following chart is a reworked copy of the one that was published by William Newman to summarize his assumptions regarding a pseudo author of the Summa perfectionis. We shall use it here also to summarize the assumptions of both Ruska and Newman.
In 1925, Ernst Darmstaedter discovered a codex in the Riccardiana Library of Florence (MS 933), containing Latin MSS devoted to Arabic alchemy. He affirmed that it corresponded to the end of the 13th century and it included the oldest MS of the Summa Perfectionis of Geber that I know, but among other things, the Liber Geberis de Investigatione Perfectionis . In this paper we shall designate the Riccardiana de Investigatione Perfectionis as DIP (number 5 in the diagram).
In that period, Julius Ruska was deeply involved in his study of al- Rāzī, and in 1935 he wrote an extensive paper in which he assumed that this newly discovered DIP (number 5) is a reworking of Liber Secretorum Bubacaris of the B.N. of Paris (number 3). He declared that the attribution of the DIP to Geber was erroneous only one example of the thoughtlessness with which ignorant writers and scribes put arbitrary names to alchemical treatises - and he decided to include it in his study of the Latin works of al- Rāzī. Ruska even suggested further that the last part of the DIP had been even written by a late Latin author (which comes under number 4) who would have also been the author of the Summa (number 6).  He considered that Liber Secretorum Bubacaris (number 3) was a reworking of  Liber Ebu Baccar er Raisy of Palermo (number 2), which was a translation of Kitāb al-asrār  of al- Rāzī (number 1).
In 1986, William R. Newman adopted all of Ruskas assumptions and based his work on them, and his main goal was to search for the unidentified Latin author that was imagined by Ruska. For this purpose, he conceived a maze of bewildering assumptions with abundance of Latin citations to conclude that a previously unknown Paul of Taranto, a compiler of a treatise with the title of Theorica and practica (number 4), was the author of both, the Riccardiana DIP (number 5) and the Summa  (number 6).
We shall prove in this part of our thesis that all the assumptions of Ruska and Newman are untenable and are without foundation. And since the Riccardiana DIP is pivotal in Ruskas and Newmans assumptions, the analysis and discussion of this treatise will be a major component in this paper   To do this we shall discuss all their assumptions as illustrated in the diagram, under the following main headings:

        The Latin MSS of Kitāb al-Asrār  of al- Rāzī: We shall prove here that the Palermo MS (Liber Ebu Baccar er Raisy no. 2) is not related to the Paris MS (Liber Secretorum de Voce Bubacaris- no. 3)
        Ruskas Assumption that the Riccardiana DIP  is an Edition of Liber Secretorum De Voce Bubacaris : We shall prove here that the DIP (no. 5) is not an edition of the Bubacaris (no. 3).
        Attribution to a Latin Author: We discuss here why Ruskas assumption of the attribution of part of the DIP and of the Summa to a Latin author (no. 4) is unsubstantiated.
        The Jābirian Paragraph of the DIP on which Ruska and Newman Based their Hypothesis of a Latin Author of the Summa: We prove here that the single paragraph on which Ruska had based his conjecture about a Latin author is simply a translation of one of Jabirs recognizable statements. This fact alone disproves the whole hypothesis of Ruska and Newman about the imaginary Latin author (no. 4).
        Arabic and Islamic Expressions in the DIP: We go a step further here to prove that the DIP as a whole is rich in Arabic and Islamic religious and non-religious expressions, including the part which Ruska had assumed to be written by a Latin author.
        Arabic Technical Terms in the DIP: We continue our proof that the DIP, including the part which Ruska had assumed to be written by a Latin author, is rich in Arabic technical terms which are not part of the usual terms that a Latin author will use in his writing.
        Jābir as the Likely Main Author of the DIP: We prove here that the major part of the DIP was written by Jābir, contrary to Ruskas assumptions. Our proofs are supported by the fact that the DIP refers to Jabirs Libro quietis (Kitāb al- rāḥa), and we discuss this book in some detail. 
        Further Examples from Newmans Assumptions: After we have proved that the whole structure shown in the following diagram is imaginary, we give few further examples from Newmans assumptions to demonstrate how they are without any foundation.
        The TP as a Compilation: we end our paper by giving some further examples to illustrate that thTP of Paul of Taranto is a mere compilation from translations of Arabic alchemy. This excludes any possibility for it to be a source for the DIP or the Summa as was assumed by Newman, (numbers 4, 5 and 6 in the diagram).
This list of topics should guide the reader in selecting what topic is of interest to him. If reading the whole paper is not possible we advice the reader to read: The Jābirian Paragraph of the DIP on which Ruska and Newman based their Hypothesis of a Latin Author of the Summa. This is a short text, but is of great significance.

Fig.  1 – The assumptions of Ruska and Newman
Ruska assumed that an unknown Latin author (4) wrote part of the DIP (5) and that he wrote also the Summa (6). Newman based all his work on Ruska’s assumptions and he imagined that the unknown Latin pseudo author is called Paul of Taranto (4) who wrote a treatise TP (4), the DIP (5) and the Summa(6).

The Latin MSS of Kitāb al-Asrār  of al- Rāzī
It is known, that Kitāb al- asrār (Book of Secrets) of al- Rāzī is a practical treatise on alchemy, devoid of theory. In the first two parts, it gives a classification of substances and a description of alchemical apparatus, and in the third and largest part, practical recipes. This work remained little known in the West throughout the centuries and the number of available Latin MSS is small.  Ruska examined six of them: 1) BN 6514; BN 7156, both from the 13th century; 2) Oxford Bodleian Digby 119, 14th century; 3) Cambridge Trinity College 1120, 15th century; 4) B.L. Sloane 1754, 14th century; 5) Palermo Codex Speciale 19, 14th century.
As the books of secrets on various subjects were very common, and in order to distinguish one from the other, the name of an author was customarily attached to the title. In this way, the first four copies, listed above, carried the name of Bubacar, while the latter two Sloane and Palermo carried a distorted name of al- Rāzī.
The importance of KA lies in its first two parts, on substances and apparatus, which in fact constitute a very small fraction of the whole work; and they were frequently quoted by compilers. On the other hand, the third part, on recipes, although it constitutes the major part, seems not to have been cited in its entirety, probably because a vast number of recipes, from both Arabic and Latin sources, were available to compilers This is the reason why the Latin versions of KA differ in the content and arrangement of the third part.
The four Bubacaris MSS have complete Latin translations of the first two parts. In 1927, Stapleton et al. published an English translation of the first two parts of KA, together with extracts from the third. An important feature of this translation is that it used both an Arabic MS and a Bubacaris MS (B.N. 6514), finding small differences between both.
In 1935, Ruska published a study on the Latin translations and re-workings of KA, and two year later, a German translation including all three parts. However he mistook KA that he had translated for a different work of al-Rāzī, which is Kitāb sirr al-asrār (The Book of the Secrets of Secrets).
While he described briefly the Bubacar MSS in his 1935 study, he gave excerpts from the Sloane and Palermo MSS. The translation of the Sloane MS was made by a Syrian priest in Antioch and retains more Arabic words than the others. Ruska did not give excerpts from the Palermo MS, but reproduced a listing published by Isodoro Carini, establishing his evaluation of this MS on Carinis headings of the recipes. Although a closer look at Carinis list shows that this MS is probably not a complete translation, Ruska concluded that not only it was, but that the others were reworkings of it.
Moreover, the six Latin MSS reported by Ruska can be differentiated into three types. The four Bubacaris manuscripts correspond to one type, while the Sloane and Palermo MSS are second and third types. The differences among all three types are noteworthy, thus the possibility of any one being a reworking of another seems remote.
We draw attention here also to a discrepancy in the historical dates of the six MSS. We do not know for sure which MSS preceded the other. If we consider the dates reported in Ruskas paper, we notice that the Palermo MS goes back to the 14th century, whereas the Bubacaris Paris MS goes back to the thirteenth. On the basis of these dates, the Palermo MS cannot be the source for the Bubacaris MS. This uncertainty about the dates of the MSS precludes the assumed relationship between them as hypothesized by Ruska in the foregoing diagram.

Ruskas Assumption that The Riccardiana DIP is an Edition of Liber Secretorum De Voce Bubacaris
The DIP is an extensive treatise. Written on 24 folios, comprising a total of 45,200 words, it is compiled from different sources. However, besides the initial few pages dealing with substances and equipment, Ruska could not find any Rāzian Latin texts matching the DIP. Moreover, his analysis showed that the share of Rāzian Latin texts amounted only to about 13% of the full MS. For this reason, he had to resort to the Arabic KA.
However, if Ruska wanted to prove that the DIP was a new edition of the Bubacaris, the Arabic text of the KA was not the right place to look for. And, if he was searching for the real sources of the DIP, he should not have limited himself to the Arabic  KA, but he should have surveyed the numerous extant works on Arabic alchemy, particularly Jābirs, since after all, the DIP does carry his Latin name.
But, since Ruska limited himself to KA, he had no other options but to compromise. Thus, sometimes he writes that the DIP text corresponds on the whole (im ganzen) to KA , while at other times he says that it is a rough approximation (annähernd). If he would have adhered to proper comparison rules, the share of materials taken from KA would diminish considerably. To illustrate: upon finding a difficulty in comparing the Arabic KA with the DIP, he explains: The DIP  text corresponds on the whole (im ganzen) to the first prescript in KAs chapter on the egg, however at the end it has a strong infringement or intrusion. The end seems to have been taken from a third source
Toiling in this way, to find in the Arabic KA similarities to the DIP, Ruska was able to add a further 27% to the share of Rāzian sources in the DIP, raising the total to about 40%. But the remaining 60% still had to be attributed.

Attribution to a Latin Author
As mentioned above, Ruska could not find in the Latin Bubacaris and the Arabic KA similarities justifying the hypothesis that the DIP was a reworking of the Latin Liber Secretorum Bubacaris. Thus, he had to admit that the compiler had have recourse to other sources. And he ventured to attribute the last part of the DIP to a Latin pseudo-author.
To substantiate his hypothesis, Ruska focused on the DIP section dealing with alums and salts (fols. 21r to 24r), which corresponds to the last paragraphs of Part IX and the full Part X, to conclude:
 What distinguishes these pieces from the largest part of the preceding compilations is obviously only due to their mature late Latin formulation. Here we do not have to deal with translations of Arabic writings, but with original Latin texts, which follow content-wise older models of the translation literature. In their style, however, they are quite Latin and do not reveal the spirit of the Arabic language.
However, in the first place, here Ruska made an assumption without presenting any evidence. Moreover, the description of the style of the text as mature late Latin is inaccurate. The text under analysis was transcribed in the last decades of the 13th century but the translation and compilation of the DIP were made obviously before transcribing the Riccardiana MS. We still have to keep in mind the state of the knowledge on alchemy in the Latin West by the middle of the 13th century, as portrayed by Roger Bacon.
The Latin literary style as a criterion to decide on the origin of this part of the DIP is not justifiable. It must be reminded that a translation may be literal or edited. In the latter case, a translator has his own understanding of a text, to then write it in another style. Thus, e.g. the translators of Toledo in the 12th century used the literal word-by-word style, but in the course of the 13th century, the translators became more knowledgeable and some began to edit their translations. The expression re-working often is used to denote the editing of a translated text. Nevertheless, an editor or a re-worker cannot be considered as the author of a text.
Finally, another genre is compilations, taken from several sources, as is the case of the DIP. Compilers may perform some editing. The work may carry the name of the compiler, and when it includes material from an important author it may bear his name. In the latter case, we have examples from both Arabic and Latin texts. Kitāb Ṣundūq al-ḥikma (Chest of Wisdom) is a compilation of chemical and alchemical recipes, and it is ascribed to Jābir. The Liber claritatis which is a compilation of chemical recipes translated from Arabic is ascribed to Geber, in the same way as the DIP which is ascribed to Geber also. Another example is the Artis Chemicae Principes or De anima in arte alchemia which is a compilation of Arabic chemical and alchemical recipes attributed to Avicenna (Ibn Sīnā).  These three Latin compilations of Arabic alchemy and chemistry (including the DIP) appeared in the thirteenth century at about the same period.
As a matter of fact, Ruska had based his hypothesis of a Latin author on a single paragraph in the DIP, which is a Jābirian one as will be now shown.

The Jābirian Paragraph of the DIP on which Ruska and Newman Based their Hypothesis of a Latin Author of the Summa
Ruska had based his hypothesis of a Latin author for part of the DIP and further, that this same author would have written the Summa on the following paragraph:
  De quorum nominibus, naturis et operationibus hic dispersa in diversis voluminibus posuimus capitula, et induimus opiniones diversas.
Alibi tamen cum Deo XE "cum Deo"  summam omnium, quae sparsim tradi­dimus, aggregabimus cum veritate probationis in summa una sermone brevi, in qua quidquid nostra volumina utile seu superfluum continent aut diminutum, hic per illam ibique per haec sanae mentis et diligentis indagationis artifex absque errore reperiet et perveniet ad desideratum perfectae artis actum et expectatum laboris effectum.
Et nos non collegimus Iob aliud multa ex antiquorum dictis et in voluminibus nostris ea multi­plicavimus, nisi ut ex illis eliceremus secretum eorum, et vitaremus errores, et ex eorum coniecturis nostri roboraremus perscrutationem sermonis via brevi et veritate perfecta, ad quam faciente glorioso et sublimi Deo, licet cum longi vigilia studii et magni laboris instantia usque quaquam pervenimus, et earn totam in libro qui Summa intitulabitur, non sub illorum scribemus aenigmate vel figuris, neque ita lucido trademus sermone, quin illum accidat necessario insipientes latere eosque subire errorem. Sed traditionum omnium assumentes arcanum ex his, quae perquisivimus, vidimus atque palpavimus. et certificati sumus cum experientia vera, tali sermone volente Deo explicabimus. Quod si se ad ea bonae mentis artifex exercita­verit, se totum [aut saltem partem] artis excelsae fructum Dei dono adinvenisse laetabitur.
In this paragraph the author makes three important declarations:
1-     He refers to his various volumes (diversis voluminibus).
2-     He has dispersed alchemical knowledge in these volumes.
3-     Therefore he will write a sum book (summa).
This is a Jabirs paragraph and we have already cited similar ones at the end of Part I under Jābirs Books of Sums when we discussed the Unique Jābir Traits.
We have shown that Jābir was the only author who had written numerous volumes throughout which he had dispersed knowledge and who wrote sums of this scattered knowledge.
Also in this same paragraph, the author employs four Islamic expressions of praise to God: cum Deo, glorioso et sublimi Deo, volente Deo and Dei dono. As it was discussed above also, Gebers works included this kind of Islamic expressions. Moreover, Geber was acknowledged among the Latin alchemists up to the 17th century as the author that most characteristically praised God, this being a sign of the Arabic origin of these works.
This paragraph is therefore, a translation from an Arabic Jābirian text. It ought to be very alarming for historians of science to realize that the whole hypothesis of Julius Ruska is based on his false interpretation of this single clause; and that the whole intricate structure of William R. Newman and his voluminous work concerning a pseudo Paul of Taranto and his imagined role in the history of Latin alchemy, is built on Julius Ruskas false understanding of just one Jābirian paragraph.
Arabic and Islamic Expressions in the DIP
The DIP as a whole, including the section selected by Ruska as a Latin original text, is rich in Arabic and Islamic expressions. The assumption that they were insertions by the author in order to imitate the Arabic style needs sound substantiation. It seems infinitely more probable that, inversely, they are an intrinsic part of the fabric of the DIP.
Latin translators used to purge the Arabic texts from conspicuously Muslim expressions, like the name of the Prophet and other explicit Islamic religious idioms. However, there are some Islamic expressions that can be applied to any religious belief, especially those that praise or glorify God. These expressions were sometimes kept in the translations.
Among the many Islamic expressions, one sentence reads: Benedictus igitur sit gloriosus et sublimis Deus qui nihil fecit regimine carens, meaning Therefore be blessed the glorious and sublime God, who made nothing which lacks order. This sentence is similar to one or two verses in the Qurān.
The DIP includes also 56 short Islamic religious expressions, such as cum deo, cum deo volente, and the like, distributed throughout its 24 folios 10 of which appearing in the section selected by Ruska as being written by a Latin author. In the Arabic alchemical literature, these kinds of expressions tend to occur at the end of recipes and they are translations of the Arabic inshāallāh (if God wills) and its other Arabic forms. The frequency of their use varied from one author to the other, but they are occurring in all of them. This is a typical Muslim phrase, derived from the Qurān. Its use is mandatory and is deeply rooted in Islamic culture.
The non-religious expression scias hoc appears 31 times 12 of them in the section selected by Ruska. It means know this, understand this and it is typical in Arabic texts whenever the author wanted to stress the importance of an idea or a prescription. The phrase et est de secretis occurs several times in the DIP, also being a typical expression in Arabic alchemy.
A quantitative analysis also serves to underscore the difference between Jābirs and al- Rāzīs texts. In the latters Arabic printed edition of KA, the expression if God wills occurs 4 times only, while know this occurs 7 times throughout all its 116 pages. This is in overt contrast to Jābirs Arabic texts. In his al-malāghim books devoted to the practical alchemy of amalgams-, between folios 2a and 36a there are 48 expressions of if God wills and 17 of know this.
God references by Latin authors
We have surveyed several alchemical treatises written by Latin authors and other works translated from Arabic from the 12th century on, looking for the word God and others signifying God, together with their qualifications. We found that the qualities attributed to God by Muslim alchemists were not used by Latin writers. That is to say, it is possible to distinguish a Latin from an Islamic author through the occurrences of the word God.
Latin authors had a particularly Christian style to refer to God. For instance, Arnald of Villanova, in Chymicall Treatise, mentions the word God devoid of the Islamic attributes. In this work, the term Holy Ghost appears more times than God while the latter is defined I say that the Father, Son and Holy Ghost are one, yet three. And still, The Word was a Spirit, and that Word the Spirit was with God, that is with himselfe, and God was that word, he himself was the Spirit, based on John 1:1. Thus, in this treatise we find a clear Christian tone, completely different from the Islamic.
In the Book of Quintessence, by John of Rupescissa (d. 1365), the author begins his essay in the name of the Holy Trinity, in opposition to the Quranic verse which opens an Arabic work like the Liber de Compositione Alchimiae. God is designated several times as our Lord God; the name Jesus Christ is used, and in no instance references to God include the Islamic attributes.
To quote one more example, in New Pearl of Great Price, written by Peter Bonus in the 14th century and edited by Janus Lacinius in the 16th, we find once again the same Christian style of references to God (without the Islamic attributes), besides the expressions: Christ, Jesus Christ, the Son of God, our Saviour Jesus Christ and our Lord Jesus Christ.
Arabic Technical Terms in the DIP
Many technical terms clearly betray the Arabic origin of the DIP. One instance is the units of weight. Latin translations of Arabic alchemical works generally used the libra, a translation of the Arabic word ral. One ral was usually equivalent to about 468 grams, although this value had regional variations. The Roman libra or pound was used throughout Europe, and was divided into 12 ounces (about 329 grams). However, many European merchants in the Middle Ages preferred to use a larger pound of 16 ounces (about 457 grams). As in the case of the ral, there were many standards simultaneously in use. For example, a 15th century Venetian libbra grossa, was equivalent to about 476 grams. Thus, the libra and the ral are nearly equivalent. Still the DIP compiler (or translator) used the Arabic ral. The word rotulum and its variants occur 55 times in the text.
Arabic texts also use the dirham, which is equivalent to about 3.6 grams. In the DIP, the dirham is translated as drachma, and occurs 54 times. In Spain, the drachma was a unit of mass in apothecaries, and it was equivalent to about 3.6 grams and in European countries in general it varied between 3.3 and 4 grams.. However, the drachma is not used in Latin alchemical works.
 Ocab corresponds to the Arabic uqāb, meaning eagle. This word is much used in Arabic alchemical works as a pseudonym of nushādir (sal-ammoniac). The word ocab appears 101 times in the DIP, while in other Latin translations or works written by Latin authors, the term employed was sal ammoniacum. Like this, there are many other terms in the DIP that kept closeness to their Arabic origin. A few examples are presented in Appendix 3.

Jābir as the Main Author of the DIP
If we accept Ruskas assumption that the Rāzīan contribution to the DIP amounts to 40 percent of the whole treatise, we are left with 60 percent that should be accounted for. We have demonstrated that the pivotal paragraph which Ruska imagined was written by a Latin author is a translation from Jābir and that the whole section which Ruska tried to assign to a Latin author is an Arabic translated text. Therefore we are left with one choice only which means that Jābir is the main author of the DIP. Let us elaborate.
The dominant figure in Arabic alchemy was Jābir ibn Ḥayyān. Many alchemical works bear his name, as it was pointed out above. On the other hand, al- Rāzī wrote a much smaller number of treatises, the most renowned of which is KA. For Arabic writers on alchemy, Jābir was the main authority. They quoted him systematically more often than al- Rāzī. For example, al- Jildakīs Nihāyat al- ṭalab a commentary to al-Irāqis treatise on the cultivation of gold mentions ten times one work only by al- Rāzī, whereas Jābir is mentioned 194 times and 42 works are cited.
This same huge disparity is found in the literature of Arabic alchemy and it is not difficult to explain. Al- Rāzī had an interest in alchemy in his youth, for a period of about ten years, to later devote himself to medicine. On the other hand, Jābir devoted his long life 90 years, according to al-Jildakī - mainly to alchemy.
We like to mention here that Jābir was the main source for al- Rāzī. This was demonstrated in detail by Stapleton.   Earlier, Abū Maslama al- Majrīṭī (d. 1008) had shown in his book Rutbat al- ḥakīm that al- Rāzī did not discuss any topic that was not discussed earlier by his teacher Jābir, and that Jābir had revealed facts that remained obscure to his student al- Rāzī. Another noted alchemist, al-Ṭughrāī (d. 1121) claimed in his book Mafātīḥ al-raḥma, that most of al- Rāzīs twelve books were copied. In recent times, Ruska acknowledged that the twelve books of al- Rāzī are influenced by the teachings of Jābir.
Thus, it cannot surprise us that a vast disparity is reflected in the alchemical works that were translated into Latin. The translators had a much larger choice from Jābirs works on both practical and mystical alchemy than from al- Rāzīs works. This helps to explain the existence of several Geber Latin works.
Therefore, the compiler of the Riccardiana DIP would have had two main Arabic sources: the lesser would be al- Rāzī, and the main one, Jābir. 

Libro quietis (Kitāb al- rāḥa)
A further argument for Jābir is found in a paragraph in fol. 6r of the DIP: Diximus superius in libro quietis utilia et non utilia, ubi diximus congelationes spirituum et coniunctiones corporum, et subtiliter diximus in operatione supradicti libri. Here the author states that he had talked above, in the libro quietis (Kitāb al-rāḥa), of the congelation of spirits, the union of bodies and their preparations.
Ruska had noticed this paragraph, but he was not able to find any mention to the libro quietis or the Kitāb al- rāḥa, neither in Bubacaris, nor in KA. Thus, he concluded that it must have referred to a different unknown source. However, as al- Rāzīs Kitāb al- rāḥa is not a book on alchemy, the libro quietis mentioned in the DIP must be Kitāb al- rāḥa of Jābir. This work by Jābir is missing, but we have significant information about it from al- Ṭughrāī  and al- Jildakī.
The importance of Kitāb al- rāḥa may be appraised from the following statement in al- Jildakīs Nihāyat al- ṭalab:
 And since to us was revealed everything concerned with this science we devoted this our book and K. ghāyat al-surūr and K. al-shams al munīr and K. al-taqrīb fī asrār al-tarkīb and K. sharḥ k. al- rāḥa of Jābir (Explanation of the Book of Rest) to important, useful and comprehensive practical discourses, which if mastered by the seeker of knowledge would enable him to grasp all the principles and doctrines of the Art.
From the paragraph in the DIP, we learn that the libro quietis discusses the coniunctiones corporum, that is, the union of bodies or their alloying. From al- Ṭughrāī , we learn that Kitāb al- rāḥa deals with the union (tazwij) of bodies. The word tazwij literally means intimate union and, in an alchemical sense, mixing or alloying. In other words, tazwij and coniunctiones convey similar meanings. Al- Ṭughrāī would also quote extensively from Bālīnās and Jābir. An instance of a long quotation taken from Kitāb al- rāḥa is the following:

 And Bālīnās was speaking in this chapter about the method of mixing and the union of the thin (rarefied) with the dense (thick). And his meanings are similar to what Jābir had said in Kitāb al- rāḥa, although the approaches are different; but the scientist perceives with Gods light and understands the relative relationship. Jābir ibn Ḥayyān said that no whiteness can take place, or redness, without the spirits and the spirits of bodies. And there is no way of differentiating and of bringing out the gentle spirit of the body except by the spirits of the spirits, This is because the spirits of bodies yearn for the spirits of spirits and seek them since all of them are spiritual and aeriform. Therefore if they are subjected to the heat of fire they fly and evaporate. So if the spirits are mixed with the spirits of bodies they cling to each other by an adherence that cannot separate.

At the end of this long citation, Al- Ṭughrāī comments, If Jābir in his book, which nobody has surpassed us in compiling, had given only this chapter it would have been sufficient, because it contains most of the principles that are needed in this Art.
Recapitulation the Bringing to an End of Ruskas and Newmans Assumptions
We have discussed until now in detail why the DIP cannot be considered as a re-working of the Bubacaris, why it was entirely compiled from sources of Arabic alchemy, mainly from Jābir, why the pivotal paragraph on which Ruska had based his hypothesis is translated from Jābir, and why no part of the DIP was written by a Latin author.
 Having refuted Ruskas speculations and since Newman had built his work on Ruskas hypothesis (see the diagram), it follows that all Newmans assumptions are without foundation.
With this conclusion, it is irrelevant to discuss the arguments given by Newman on the interdependence of the Bubacaris, the TP, the DIP and the Summa. There is no need for this anymore. Nevertheless, we shall give few examples.
Further Examples from Newmans Assumptions
We have given till now ample evidence about the Arabic identity of the Summa and the DIP and we have shown that the corpuscular theory and the mercury alone theory are of Arabic origin and were not the invention of the pseudo Paul of Taranto. We have shown also at the end of Part I that all alchemical theories in the Latin language came with the translation of Arabic works in the twelfth and thirteenth century.  Nevertheless we shall give now further samples of the kind of assumptions that Newman had used in his work.

The Bubacaris and the Summa
Newman would suggest that the Bubacaris was a source for the Summa. He had studied the two Latin translations of Arabic works available to him Gebers L. Misericordiae and De Re Tecta by pseudo-Avicenna and since he could not find in them anything comparable to the Summa, he concluded that it must have had the Bubacaris as a source.
This hypothesis would be a priori unlikely as these are very different works: the Bubacaris is a practical treatise, including very little theory, while the Summa is a theoretical work, with a minor content of practical alchemy. Thus, any potential similarities would be too fragile a ground to establish a dependence of either of them on the other. However, it is worthy to review one of the instances (about ceration) which Newman chose to base his assumption.

Ceration
This instance was deemed by Newman to be unique and would not occur in any other alchemical texts except the Summa and the Bubacaris. It concerns the materials that ought to be used as agents of ceration. He did not find such information in the two Arabic Latin translations that were available to him.
According to Newman, Bubacaris is using sulphur and arsenic in ceration, whereas the Summa is using mercury, sulphur and arsenic.  He says that the author of the Summa has divined the reason why the Bubacaris had used sulphur and arsenic only, so he added mercury.  
Newman thinks that ths alchemical knowledge is quite unique to the Summa, but its underpinning is found only in Bubacaris and he concludes that the Summa had used Bubacaris as a source
Summa

Putaverunt ideo aliqui cerationem debere ex oleis, liquidis, et aqueis fieri: sed erroneum est illud a principiis huius magisterii semotum penitus, et ex manifestis nature operibus reprobatum. Naturam enim non videmus in ipsis corporibus metallicis humiditatem cito terminabilem ad illorum fusionis et mollificationis necessitates posuisse.... In nullis autem  robus melius et possibilius et propinquius hec humiditas cerativa invenitur qualm in his-- videlicet sulphure et arsenico propinque-- propinquius autem in argento vivo et melius.
The Smma here is advocating the use of spirits only (sulphur, arsenic and mercury) for ceration and is opposing the use of oils, liquids and waters.
Bubacaris
Inceratio corporum sapientissimi philosophi cum sulheribus(!) et auripigmentis puris facere preceperunt quia commiscentur cum corporibus si coniunguntur et si cum ipsis fuerint. Elargant enim ea et diasolunt et faciunt currere (pro currere cod. legit cinerem). Et secundum quod multi dixerunt corpora incerantur cum salibus aut boracis et non intellexerunt quod pertinet incerationi et de salibus. Non est tamen inceratio nisi fuerit cum eis aut sulfur aut auripigmentum preparatum.
Bubacaris here is recommending using sulphur, arsenic, salts and boraces. This range of materials is not the same as the one in the Summa.
Jābir

Jābir discusses ceration in numerous books. We give here a selection only:
- Kitāb al-raḥma al-kabīr (The Great Book of Mercy) and Kitāb sharḥ Kitāb al- raḥma (The Book of Explanation of the Book of Mercy): Jābir advocates the use of mercury, sulphur, arsenic and sal-ammoniac.
- Kitāb muṣaḥḥaḥat iflātūn (The Book of iflātūns Corrections) : Mercury, sal-ammoniac and sharp waters.
- Kitāb al- uṣūl (Book of Principles) Sal-ammoniac solution
- Kitāb tadbīr al-arkān wa al- uṣūl (The Book of Treatment of Bases and Principles) : Sal-ammoniac solution.
- Kitāb al-tajrīd (The Book of Abstraction) : Sal-ammoniac solution.
- Kitāb al-riyāḍ (The Book of Gardens) : Water of eggs white with sal-ammoniac, borax and tinkār; also the fat of the horns of deer.
Comparing the Latin with the Arabic texts on ceration
Jābirs materials for ceration are: spirits (mercury, sulphur and arsenic), salts (mainly sal-ammoniac), borax, tinkār, sharp waters, water of the white of eggs and fat or oil from the horns of deer. These materials include all the reagents given in the Summa and the Bubacaris,
Let us remember that ceration is a basic step in a series of operations of Arabic alchemy, each one leading to the other, in order to obtain the elixir for metals such as gold. It begins by calcination, followed by ceration, solution and coagulation.
From the above comparison we conclude that the Summa and Bubacaris are not dependent on each other and that their ultimate sources are Arabic.
The TP and the DIP
Newman had assumed that the DIP had been based to some extent on the TP of Paul of Taranto, since some formulations are similar in both MSS. However, the fraction of such similar formulations is very small relative to the full contents of the DIP, representing a mere 3.3%. Thus, they cannot be taken as an argument for the dependence of the DIP on the TP, nor for Paul of Taranto as the author of the DIP.
This similarity is due to one of two causes: either the TP used the DIP as one of its sources, or both compilers used the same source. In addition to the very small fraction of similarities, we noticed that the compiler of the TP had cut down parts of some formulations and removed the typical Arabic Islamic expressions. The hypothesis that the Latin compiler of the DIP had purposefully introduced such Arabic and Islamic expressions is untenable.

Inter-dependence of the TP and the Summa
Under the title The Theorica et Practica and its relationship to the Summa Newman gave two main arguments to prove the inter-dependence of the Summa and the TP. One was about the composition of metals on the basis of the sulphur mercury theory, and the second was in connection with arsenic as one of the principles. These two assumptions were discussed above in Part I (see The sulphur mercury theory and the composition of metals; and also ‘‘The theory of the three principles: mercury, sulphur and arsenic, where we compared the Summa with the Arabic sources.  We proved there that this alchemical knowledge in both the Summa and the TP is a basic one in Arabic alchemy. The restriction of the search to a few Latin sources is the cause of this flawed hypothesis.

Differences between the TP and the Summa
Newman went further, and in order to prove that the Summa could not have been the source of the TP, he paid special attention to the differences between both texts. But in this way, what he did prove, indeed, was their lack of similarity.
The differences between the TP and the Summa are indisputable and there is no need to prove them. However, it is inconsistent to infer from the differences that the Summa is based on the TP. On the other hand, our discussion in Part I above clearly shows that all the chemical theories in the Summa are based on Jābirian alchemy.

 THE TP AS A COMPILATION
Additional Examples
It is not our aim to give in this paper a thorough discussion of the TP.  Its character as a compilation is stated in its colophon which says that it was compiled by Paul of Taranto. 
The theoretical part of the TP begins with a short article on what things and what kind of things this art takes as materials. The text gives several cover names (decknamen) for metals and their calxes. This reminds us of the tradition of Arabic texts. Moreover, in one page only we can find several Arabic terms mentioned in a distorted form such as, e.g. sodebeb (dhahab, gold); alkal (al-kuḥl, stibnite); anec (anūk, tin); kasdir (qaṣdīr, tin); sericon (zarīqūn, or sarīqūn, lead oxide); usurub (usrub, lead); saffron of iron is a translation of zafarān al- ḥadīd.
In the second article, dealing with the four principles (or spirits), mercury has alternative distorted Arabic names, such as azot, azet and zambac, from zibaq. It is also called servus fugitivus, which is the literal translation of the Arabic pseudonym of mercury, al-abd al-ābiq (the fugitive slave). Among the names of sulphur we find the Arabic kibrit. Arsenic is mentioned by its Arabic name zernech from zarnīkh. Sal-ammoniac is called almizedir and nischader, from nushādir, and capocab from uqāb (eagle).
The second part of the Practica, contains practical procedures and recipes easily identifiable as taken from translated Arabic works. We have examined the items and recipes that fall between folios 39V and 45R. There are 28 items from which 19 were identified by Newman to have beenre-written by the author of the TP from Latin translations of Arabic sources. These recipes are found in the original Arabic works also. 

Sal Alkali
An article under the heading how sal alkali should be made, describes how sal alkali, to make glass, is to be prepared from a herb. Newman remarks that this is the only recipe for sal alkali which I have found in an alchemical text that describes the preliminary roasting necessary for the production of potash. It probably reflects the authors own experience.
During the Middle Ages, and until the discovery of a manufacturing process for alkali, the chief source for this material for glass-making in Italy was an alkali imported from Syria, known as polverine, rochetta or allume catina. It is the ashes of a shrub called ushnān, which grows in the Syrian Desert, and belongs to the salsola soda family. The Venetians had established strong connections to Syrian glass-making since the 13th century. Allume catina became, thus, the most important single ingredient in the Venetian glass manufacture, establishing a regular and plentiful supply from their cotton trade with Syria in the 14th century. The use of Levantine ashes was mandatory for Venetians glass-makers, ensuring the superiority of their glass for centuries.
Therefore, the shrub in question did not grow then in Italy, and the compiler of the TP would not have been able to have a personal experience in this respect. Most probably, the recipe was compiled from an Arabic source. The description in the TP is found in Arabic and Persian texts on alchemy and on glazes and ceramics.
According to such one treatise, potash (qalī) was prepared by specialists (qallā). They worked on the edge of the desert and were renowned for the high quality of their product. The Salsola plants were collected while they were still not completely dry, and then burned in a slow smoldering fire, in a pit about one meter wide and two meters deep. The ashes were then calcined into blocks of ten pounds each.
Both al- Rāzī, in KA, and Jābir, in Kitāb ṣundūq al-ḥikma, give recipes to prepare the salt of al- qa  from this same material.

EPILOGUE
 The Geber Problem started with Berthelot in 1893. We have summarized most of his assumptions. In the 1920s, Holmyard gave substantial evidence against Berthelots hypotheses.
In 1935, Ruska removed the name of Geber from the Riccardiana DIP and considered it to be a reworking of the Bubacaris, with a part written by a Latin author, who would have also been the author of the Summa.
In 1986, Newman based his work on Ruskas conjecture, and attributed both DIP and Summa to Paul of Taranto.
Following Holmyard, once again we have disproved Berthelots hypotheses. And on comparing Arabic texts with the Summa, we were able to show that the Summa is based in its entirety on Arabic alchemy.
We gave in Part II substantial evidence which proves that the compiler of the DIP could not have been the author of any portion of it, and that it is rather a compilation from Latin translations of Arabic alchemy. The Liber Secretorum Bubacaris, or rather a different translation of al- Rāzīs Kitāb al-asrār, is only a minor source, whereas the major one is most probably Jābir (Geber). Therefore, the DIP is not a new edition of the Bubacaris, as Ruska had claimed. It also follows that the attribution of the DIP to Geber was not due to the scribes ignorance, but it was purposefully ascribed by the compiler, the only person who knew what sources he had employed.
Ruska had based his hypotheses regarding a Latin author of the Summa on a paragraph in the DIP which he assumed was written by a Latin author. We have proved that this paragraph is of Arabic origin and that it is a familiar one in Jābirs works.
Newman had built his work on Ruskas speculations. And since these were refuted, it follows that all Newmans assumptions about Paul of Taranto are unsubstantiated.
Finally, this paper as a whole gives ample evidence to prove that the hypotheses of Berthelot, Ruska and Newman are unfounded. The Summa is either a compilation from Arabic sources, mainly Jābir, or a complete translation of a missing Arabic treatise.
If this research is to serve a useful purpose, it should help to cast serious doubts on the reliability of the existing history of early Latin alchemy as written by Berthelot, Ruska and Newman who tried to divorce Latin alchemy from its Arabic origins.  This history cannot be written with fairness and impartiality without a thorough research into Arabic sources.

No comments: