One of the most striking examples of borrowing from the Elementa Astronomica of Alfraganus occurs in the Paradiso, where Dante likens twenty-four spirits to as many brilliant stars.[96] He makes up the number by taking fifteen specially bright stars from different parts of the sky, and adding to them the stars of Ursa Major and two stars of Ursa Minor. The seven chief stars of Ursa Major are well known, but Beta and Gamma Ursæ Minoris are not conspicuous nor specially familiar; and why, if he takes fifteen unnamed, should he name any of the twenty-four? Turning to chapter 19 of Alfraganus’ book, we find that he follows Ptolemy in enumerating fifteen first-magnitude stars in different parts of the sky, and then he gives as examples of the second magnitude (i.e. next in brightness) Benet Naax and Alfarcatein, Arab constellations which correspond with the tail of Ursa Major, and Beta and Gamma of Ursa Minor.
There is no reason to suppose that Dante knew Arabic, but there were several versions of Alfraganus in Latin, and it is even possible to determine with some certainty which of them he used. Speaking of the movements of the planet Venus, he says that they may be found, summarized from the best demonstrations of the astrologers, in the Book of the Collection of the Stars:—
“Li quali, secondochè nel Libro dell’ Aggregazione delle Stelle epilogato si trova, dalla migliore dimostrazione degli astrologi ....” (Conv. II. vi. 133-136).
Although the usual name for the book of Alfraganus in Latin was Elementa Astronomica, the version for which mediæval students were indebted to the indefatigable translator Gerard of Cremona bears in the MSS. the title Alfragani liber de aggregationibus Scientiæ Stellarum et de principiis cœlestium motuum. It is this book, therefore, that Dante means, though he has translated the title somewhat inaccurately.[97]
There were several other books known to Dante from which he must have gleaned information about Ptolemy’s system, but Alfraganus we may regard as his standard reference book. Many versions of Alfraganus are still extant, and it is pleasant to feel that we can hold in our own hands Dante’s text-book on astronomy. The best known is the printed edition of Golius, published in Amsterdam in 1669: it is in both Latin and Arabic, and opening in the middle we may turn the pages backwards to see the beautiful Arabic letters of Alfraganus’ own language, which is written from right to left, or forwards to look at the Latin in which he became known to Dante.
Next in importance to Alfraganus among Dante’s authorities on astronomy was Aristotle. We have seen how overwhelming his authority became in natural science in the second half of the thirteenth century, and how for some time scholars failed to distinguish between the system of spheres devised by Eudoxus, which was the scheme upheld by Aristotle, and the epicycles and eccentrics of Ptolemy. Among all Aristotle’s admirers none was more devoted than Dante: for him Aristotle is not only “Il Filosofo,” the Philosopher par excellence, as he was generally called in that age; he is “quello glorioso filosofo al quale la Natura più aperse li suoi segreti,” his intellect was “quasi divino,” and his words are of “somma e altissima autoritade.”[98] Throughout Dante’s own writings his references to Aristotle are so frequent that Dr. Moore observes: “The amount and variety of Dante’s knowledge of the contents of the various works of Aristotle is nothing less than astonishing.”[99]
He does not, however, consider Aristotle infallible as regards the details of astronomy, for in these the philosopher was only following the observers and mathematicians;[100] but in all matters which depend upon first principles, in which Dante would include the form of the earth, Aristotle’s authority may not be called in question.[101] In cosmical physics, such as the doctrine of the four elements, and in meteorology, Dante follows “il mio Maestro”[102] implicitly, and it is also largely from the De Cælo that he gained his knowledge of early Greek speculations regarding the universe.
Dante knew little more of Greek than of Arabic. His occasional use of Greek words is enough to prove this—for example when he discusses the Pythagorean theory of “Antictona,” using the accusative as if it were a nominative.[103] In several passages he implies that the language was not known to him, and he distinctly states that he used translations of Aristotle’s works. Here again we are able to identify the versions he used. In Conv. II. xv. 59-73 he complains that it is impossible to know what Aristotle believed about the Milky Way, because in the “old translation” he is made to say one thing, and in the “new translation” quite another. From his quotations of these two conflicting opinions, we are able to deduce that the “old translation” was that which Michael Scot made from the Arabic, and the “new translation” that of Aquinas, which was made direct from the Greek.
He had access, therefore, to more than one Latin translation of Aristotle; and beside this, his quotations seem to have been often taken from commentators and compilers of his own times, especially Aquinas and Albertus Magnus. This was frequently the case with the Meteorologica of Aristotle, for the De Meteoris of Albertus Magnus seems to have been much used by Dante in place of Aristotle’s own work. Dante quotes it by name as Meteora in Conv. II. xiv. 169, and also in Conv. IV. xxiii. 125-126, where he mentions “Alberto” as the author.
These two, then, the Arab Alfraganus and the Greek Aristotle, were Dante’s chief authorities in astronomy, both in Latin translations.