Presentation
of “Petri Nonii Salaciensis Opera” by Pedro Nunes  (this text in Portuguese)

João Filipe Queiró

Departamento de Matemática, Universidade de Coimbra, 2002.




Pedro Nunes (Alcácer do Sal, 1502 – Coimbra, 11 August 1578) is the most important Portuguese mathematician ever, and he is one of the great names of mathematics in Europe in the 16th century.

Many aspects of his life remain obscure to this day. We know the place and year of his birth, as well as the place and date of his death. We have some information on his activity, as a professor – in the university and at the royal court – and as a cosmographer. A few details concerning positions held and salaries received have also come to us, through documents written by others. And we have his works, from which we can extract an intellectual profile and draw some additional biographical conclusions.

All of this, which is much, is not enough to obtain a picture of the man. Information about his family, his youth and his formative years is lacking.[1] We hardly know of any European travels by him. There are practically no letters, to and from Pedro Nunes, and they have certainly existed and would today be documents of the utmost interest for the better understanding of the Portuguese mathematician’s life and work.[2] Even a portrait is lacking: there is no contemporary painting or engraving representing him.

In the absence of this information, we are left – and from a scientific perspective it is the most important thing – with Pedro Nunes’ writings. The works printed during his life have been identified, and it is unlikely, although not impossible, that others might surface. Also, no unpublished manuscripts have been found, with the exception of a single one, more than fifty years ago. But even about these works there is much we don’t know, in general, with respect to the circumstances surrounding publication: the choice of printer, the funding, the print run, distribution and sales, the reading and use by men of science, teachers, students.

The publication of Pedro Nunes’ printed works during his life is concentrated in two periods: 1537-1546 and 1566-1573. In the first period there is a work in Portuguese and several ones in Latin, and in the second period there are works in Latin and one in Spanish. We immediately notice the 20-year interval between the two periods. It is not easy to find explanations for it. In that respect, we may recall Pedro Nunes’ professional activity, as a university professor – in Coimbra from 1544 to 1562 – and as royal cosmographer since 1547. But there is another interesting aspect. The works printed in the first of those periods were all published in Portugal: the Tratado da Sphera, in 1537, in Lisbon; the De Crepusculis, in 1542, in Lisbon; the De erratis Orontii Finaei, in 1546, in Coimbra; the booklet Astronomici introductorii de spaera epitome, also from this period, has no indication about where it was printed, but it is probable that it was published in Coimbra or Lisbon. In the second period, the first two works, which are the most important (the other ones are re-editions), were published outside Portugal: Petri Nonii Salaciensis Opera, in 1566, in Basel, and Libro de Algebra, in 1567, in Antwerp. This fact, apart from possibly showing an interest in using quality or prestige printers, seems to correspond to a wish by Pedro Nunes for broader dissemination of these mature works. Both are rewritings and extensions of texts from the 30s: parts of the Tratado da Sphera, in the first case, and a lost manuscript in Portuguese, in the second.

More than before, Pedro Nunes addresses himself, with these two deeply meditated works, to the European scientific community. He chose Spanish for the Libro de Algebra, and published it in Antwerp, a Flemish city then under Spanish control. The Opera were published in Latin, in Basel.

The object of the present facsimile edition is this latter book. Of all of Pedro Nunes’ works, this one is possibly the least studied by historians of science. But it probably is one of the most interesting.

 

*

 

The Petri Nonii Salaciensis Opera consist of four texts, well identified in the front page, although not so well in the contents pages compiled by the author.

The first two of these texts may be considered Latin versions of the navigation treatises included in the 1537 Tratado da Sphera. But while the first, De duobus problematis circa nauigandi artem, is indeed, in large measure, a translation of the Tratado sobre certas dúvidas da navegação, in which, for the first time, constant-bearing navigation and great-circle navigation were clearly distinguished, the second one is a very different work, much longer and developed than the Tratado em defensam da carta de marear.

This extensive text,[3] entitled De regulis & instrumentis, ad uarias rerum tam maritimarum quam & coelestium apparentias deprehendendas, ex Mathematicis disciplinis, deals successively with the analysis and design of navigation charts, astronomical navigation and the laying out of rhumb-lines on a globe.

On the first subject, Pedro Nunes again discusses at length, as in 1537, the constant-latitude rectangular chart, in particular against the requisite that rhumb-lines be represented by straight lines. This is a typical contribution originating from navigational needs, which inform Nunes’ discussion more than the convenience of faithfully representing the globe. After pointing out the inevitable errors in a constant-latitude chart, with respect to rhumbs, coordinates of points on the globe, and distances between them, Pedro Nunes suggests the use of a solution of the type found in Ptolemy’s Geography (in turn inspired by Marinos of Tyre), where there appear several rectangular regional maps with varying proportions in their grids, according to the central parallel in each.[4] Pedro Nunes recommends that different charts should be made according to latitude, and adds that the extremal parallels in each chart should not be too distant from one another (“Extremos autem parallelos non admodum à se inuicem distare oportet”, pág. 25), to avoid excessive distortion. Furthermore, he says that in all charts the whole longitude of the globe should be included. (“Et ponenda est in omni tabula universa orbis longitudo”). Pedro Nunes thus proposes, differently from Ptolemy, a collection of horizontal strips covering the surface of the globe. Here we practically have already the so-called Mercator projection. It remains only, as Gomes Teixeira points out,[5] to join the narrow partial maps in a single chart, keeping their inner proportions and using the same scale everywhere for longitudes. According to Nordenskiöld,[6] it was in this way that Mercator proceeded in the design of his map, using ten-degree intervals between the parallels.   

The first author to publish tables allowing the construction of charts – with “increasing latitudes” – in which rhumb-lines are represented by straight lines was Edward Wright,[7] who says that the errors in the chart with constant latitudes were pointed out by several authors before him, “especially by Petrus Nonius, out of whom most part of the first Chapter of the Treatise following is almost worde for worde translated”. Wright used the Latin treatises of Pedro Nunes, and not the 1537 texts, as we can see from the beginning of his chapter II, in which he mentions “Petrus Nonius in his second booke of Geometricall observations, rules, and instruments”, which alludes to the title of the second treatise.

The next chapters of the De regulis & instrumentis deal with astronomical navigation, including several procedures to find geographical coordinates. In these chapters we find Pedro Nunes’ interesting commentaries to Copernicus, which were analysed in detail by Henrique Leitão.[8]

The last seven chapters of this second treatise are dedicated to the problem of laying out rhumb-lines on a globe. This problem is harder than the construction of the increasing-latitudes chart, although in modern mathematical terms one can see that the two are equivalent. Pedro Nunes presents a complicated procedure, consisting of the sequential resolution of several spherical triangles, to obtain points which are approximately on rhumb-lines on the sphere. This remarkable work by Pedro Nunes was studied in detail by Raymond d’Hollander.[9] At the end of his construction, on page 172, Pedro Nunes includes a table to present the results of the calculations for several rhumbs. This table is empty, and Pedro Nunes says on the previous page that its filling can be carried out by “studiosi adolescenti”. Here we see the mathematical spirit, after devising a procedure to solve a problem, showing no interest in the computations necessary to the application to practical situations. It should be recalled that both Gomes Teixeira and Raymond d’Hollander point out the inaccuracy of a criticism to Pedro Nunes’ construction by Simon Stevin, which unfortunately is repeated still today by those who have not read the original text of Pedro Nunes.

The two Latin treatises by Pedro Nunes on navigation were translated into Spanish by Juan Cedillo Diaz, in a manuscript kept at the Madrid National Library.[10]

The third text in the Opera is a short essay, entitled In Problema mechanicum Aristotelis de Motu nauigij ex remis, on the movement of boats with oars in the Mechanica of Aristotle. This essay was studied by Henrique Leitão.[11]

The fourth and last, In Theoricas Planetarum Georgii Purbachii annotationes aliquot, is an extended collection of annotations to the Theoricae Novae Planetarum by the 14th century Austrian mathematician Georg Peurbach.[12] These have been described as being among the most important annotations to the influential work of Peurbach.[13]

 

*

 

The present publication is being carried out for several reasons. Of course there is a wish to mark the 500th anniversary of Pedro Nunes’ birth. But there are deeper reasons. As is known, the content of Petri Nonii Salaciensis Opera was reproduced in a volume published in Coimbra in 1573, with another title – the two first treatises appearing under the heading De arte atque ratione navigandi – and containing other works as well. According to the guidelines of the publication of Pedro Nunes’ complete works – now resumed under the auspices of the Lisbon Academy of Sciences – it will be the text of this latter volume the one used in the future publication, since it is the last version published during the author’s lifetime. There does not seem to be much difference between the 1566 and the 1573 texts – apart from the title change, the correction of misprints and the redrawing of illustrations – but, whatever the case, it seems important to preserve the 1566 text, which probably was the version more widely known and read in Europe. In addition, the publication date itself is important. Already the 1537 treatises, in which Pedro Nunes studies and draws rhumb-lines for the first time, antedated Mercator’s 1541 globe, where those lines appear. Likewise, Nunes’ 1566 work precedes Mercator’s famous chart ad usum navigantium, which dates from 1569. It is also, for example, contemporary with several commentaries on Copernicus by German astronomers.[14]

 

*

 

The reasons that led Pedro Nunes to choose the Officina Henricpetrina, in Basel, to set and print the Petri Nonii Salaciensis Opera are not known. We can only conjecture that the choice took into account the prestige of that publisher, established in 1488, and owned by the Petri family until the 17th century. The publisher exists to this day in Basel, albeit with another name.

During the 16th century, the Basel publisher printed many works by important authors, in the humanities as well as in mathematics, science and medicine. As an example, we mention that, in the second half of the century, many works by Cardano were printed there, including an edition of the Ars Magna in 1570. It is also interesting to note that, in the same year the Petri Nonii Salaciensis Opera were published, the Officina Henricpetrina printed the second edition of Copernicus’ De revolutionibus orbium coelestium, containing the famous Narratio prima by Georg Rheticus.

The Petri Nonii Salaciensis Opera is a 304×205mm volume, with 308 pages plus 16 unnumbered pages. In the title-page we see the well-known emblem of the publisher, still in use today, illustrating a passage from the Book of Jeremiah.[15] The colophon is on page 307: “Basileae, Ex Officina Henricpetrina, Anno M. D. LXVI, Mense Septembri.”

It is to be remarked that the extant copies of this 1566 edition are not all identical. Some of them include four pages of errata compiled by the author to point out mistakes in the text and in the illustration captions.[16] This shows that Pedro Nunes had no opportunity to correct proofs of the text set by the Basel printer, but he had access to the printed pages in time to organize the errata, which were included only in part of the print run, since there exist copies of the Opera without them.[17] This interesting fact helps us understand what led to the Coimbra re-editions, seven years later, of the works contained in the Opera, with the errors corrected and the illustrations redrawn.[18]

The copy of the Petri Nonii Salaciensis Opera used for the present facsimile edition belongs to the Mathematics Library of Coimbra University. As can be seen from an inscription on the title page, it was once part of the great library of the Santa Cruz Monastery in Coimbra, closed and dispersed around 1830. There is a reference to it in the monumental catalogue of that library compiled by D. Pedro da Encarnação in the 18th century.[19] Later the volume belonged to Luciano Pereira da Silva, a Coimbra University professor, to whom we owe the elucidation of some important biobibliographical questions concerning Pedro Nunes. Luciano Pereira da Silva’s library now belongs to the School of Science and Technology of the University of Coimbra.

            The volume is in good condition, with the original parchment binding, and contains the above mentioned errata, which shows that it was part of the “complete” print run. It contains several handwritten notes and illustrations on the margins, presumably from the 16th century. There are handwritten additions to the errata, which may be from the author himself, in the same writing as the marginal notes. The corrections indicated in the errata are introduced throughout the text in blue ink. These corrections seem to date from the 20th century.

 

[1] The little we know can be found in Inquisition depositions by two grandchildren of Pedro Nunes in the 17th century.

[2] An explanation for this lack of documents lies in the sad fate of Pedro Nunes’ papers, lost by his uninterested heirs. In this respect see Joaquim de Carvalho, preface to Defensão do tratado da rumação do globo para a arte de navegar, Coimbra, 1952.

[3] It occupies pages 13 to 189 of Petri Nonii Salaciensis Opera.

[4] Mathematically, this corresponds to saying that, if the central parallel has latitude j, the longitude degrees on the grid should be multiplied by cos j, or, which is the same thing, the latitude degrees should be multiplied by sec j. Pedro Nunes gives several examples of situations of this type.

[5] Francisco Gomes Teixeira, Elogio Histórico de Pedro Nunes, Panegíricos e Conferências, Coimbra, 1925. The text is reproduced in his História das Matemáticas em Portugal, Lisbon, 1934.

[6] A. E. Nordenskiöld, Facsimile Atlas to the Early History of Cartography, Stockholm, 1889.

[7] Edward Wright, Certaine Errors in Navigation, London, 1599.

[8] Henrique Leitão, Uma nota sobre Pedro Nunes e Copérnico, Gazeta de Matemática, 143, p. 60-78, 2002.

[9] Raymond d’Hollander, Historique de la loxodromie, Mare Liberum, 1, p. 29-69, 1990.

[10] Los dos libros de la arte de nabegar, de Pº Núñez de Saá, traducidos de latin en Castellano, por el doctor Sñ. Cedillo Diaz.

[11] Henrique Leitão, O Comentário de Pedro Nunes à Navegação a Remos, Lisbon, Comissão Cultural de Marinha, 2002.

[12] Recall that a Portuguese translation of part of that work was included in the 1537 Tratado da Sphera.

[13] See, e. g., C. D. Hellman and N. M. Swerdlow, Georg Peurbach, Biographical Dictionary of Mathematicians, vol. 4, New York, 1991. In this article, when listing the translations of Peurbach’s work, no mention is made of Pedro Nunes’ 1537 translation.

[14] Henrique Leitão, Uma nota sobre Pedro Nunes e Copérnico.

[15] Is not my word like as a fire? and like a hammer that breaketh the rock in pieces?” (Jeremiah, 23:29)

[16] It is through an indication of Pedro Nunes in the errata that we know the year of his birth. On the first line of page 209, in the third proposition of the second annotation on the Sun, we read: “Exempli gratia, sit anno Domini 1592, quo ego natus sum (...)”. And in the errata: “Pa.209.li.I.1592.le.1502.”

[17] The inicial capital letter in the errata, the body of the four pages as well as the type in which they were set, are different from those used in the main text. The paper too is slightly different.

[18] António Mariz, the Coimbra University printer responsible for the 1573 edition, is very critical in his preface of the errors in the previous edition.

[19] See J. M. Teixeira de Carvalho, A Livraria do Mosteiro de Santa Cruz de Coimbra, Coimbra, 1921.