Orbis Terrarum #118 TITLE: Orbis Terrarum DATE: A.D. 20 AUTHOR: Marcus Vipsanius Agrippa DESCRIPTION: The profound difference between the Roman and the Greek mind is illustrated with peculiar clarity in their maps. The Romans were indifferent to mathematical geography, with its system of latitudes and longitudes, its astronomical measurements, and its problem of projections. What they wanted was a practical map to be used for military and administrative purposes. Disregarding the elaborate projections of the Greeks, they reverted to the old disk map of the Ionian geographers as being better adapted to their purposes. Within this round frame the Roman cartographers placed the Orbis Terrarum, the circuit of the world. There are only scanty records of Roman maps of the Republic. The earliest of which we hear, the Sardinia map of 174 B.C., clearly had a strong pictorial element. But there is some evidence that, as we should expect from a land-based and, at that time, well advanced agricultural people, subsequent mapping development before Julius Caesar was dominated by land survey; the earliest recorded Roman survey map is as early as 167-164 B.C. If land survey did play such an important part, then these plans, being based on centuriation requirements and therefore square or rectangular, may have influenced the shape of smaller-scale maps. This shape was also one that suited the Roman habit of placing a large map on a wall of a temple or colonnade. Varro (116-27 B.C.) in his De re rustica, published in 37 B.C., introduces the speakers meeting at the temple of Mother Earth [Tellus] as they look at Italiam pictam [Italy painted]. The context shows that he must be talking about a map, since he makes the philosopher among his group start with Eratosthenes’ division of the world into North and South. This leads him on to the advantages of the northern half from the point of view of agriculture. The speakers compare Italy with Asia Minor, a country on similar latitudes where Greeks had experience of farming. After this they discuss in more detail the regions of Italy. As a visual aid to this discussion, the temple map will have been envisaged as particularly helpful. But whether it was only intended to be imagined by readers or was actually illustrated in the book is not clear. The same applies to possible cartographic illustration of Varro’s Antiquitates rerum humanarum et divinarum, of which Books VII-XIII dealt with Italy. But at least we know that he was keen on illustration, since his Hebdomades vel de imaginibus, a biographical work in fifteen books, was illustrated with as many as seven hundred portraits. Since we are told that this work was widely circulated, some scholars have wondered whether Varro used some mechanical means of duplicating his miniatures; but educated slaves were plentiful, and we should almost certainly have heard about any such device if it had existed. Although copies of Agrippa’s map were taken to all of the great cities of the Roman Empire, not a single copy has survived. The reconstructions shown here are based upon data in the medieval world maps that were, in turn, derived from Roman originals, plus textual descriptions by classical geographers such as Strabo, Pomponius Mela and Pliny. The original was made at the command of Agrippa’s father-in-law, the Emperor Augustus (27 B.C. - A.D. 14), and completed in A.D. 20. The map was presumably developed from the Roman road itineraries, and may have been circular in 1 Orbis Terrarum #118 shape, thus differing from the Roman Peutinger Table (#120). Note that most scholars, however, believe that due to its placement on the column in a portico or stoa open to the public, the Porticus Vipsani, it was probably rectangular, not circular. Shown here are three continents in more or less symmetrical arrangements with Asia in the east at the top of the map (hence the term orientation). The emphasis upon Rome is reflected in the stubby form of Italy, which made it possible to show the Italian provinces on an enlarged scale. Moreover, about four-fifths of the area of the map is devoted to the Roman Empire alone. India, Seres [China], and Scythia and Sarmatia [Russia] are reduced to small outlying regions on the periphery, thus taking on some features similar to the egocentric maps of the Chinese. A “reconstruction” of the Orbis Terrarum, from Raisz Note that most scholars, however, believe that due to its placement on the column in a portico or stoa open to the public, the Porticus Vipsani, it was probably rectangular, not circular. 2 Orbis Terrarum #118 The only reported Roman world map before Agrippa’s was the one that Julius Caesar commissioned but never lived to see completed. We are told by late Roman and medieval sources that he employed four Greeks, who started work on the map in 44 B.C. These were no doubt freedmen, of whom there were large numbers in Rome, including many skilled artisans. The four regions of the world are not self-explanatory, but what Caesar seems to have meant is as follows: the East (by the cartographer Nicodemus), included everything to the east of Asia Minor; the West (by Didymus), included Europe except Greece, Macedonia and Thrace; the North (by Theodotus), included Greece, Macedonia, Thrace and Asia Minor; and the South (by Polyclitus), included all of Africa. If Romans were planning this, they would place the northern section much further west, whereas the cartographers were Greeks, and they followed a tradition which originated in Rhodes or Alexandria. We may speculate whether this map was flat and circular, even though such a shape might have been considered ‘unscientific’ and poorly adapted to the shape of the known world. That is the form of the Hereford world map (Book IIB, #226), which seriously distorts the relative positions and sizes of areas of the world in a way we should not imagine Julius Caesar and his technicians would have. A late Roman geographical manual gives totals of geographical features in this lost map with recording names, but even the totals, on examination, turn out to be unreliable. Agrippa’s map was compiled to further Roman imperial expansion. M. Vipsanius Agrippa (64/63-12 B.C.) was one of the earliest supporters of the young Octavian in his fight to establish himself as Julius Caesar’s heir. He first became prominent as governor of Gaul, where he improved the road system and put down a rebellion in Aquitania. He pacified the area near Cologne (later founded as a Roman colony) by settling the Ubii at their request on the west bank of the Rhine. In 37 B.C. he was a consul and built Octavian a fleet which enabled him the following year to defeat Sextus Pompeius in Sicily; Agrippa as admiral of this fleet used a new type of grapnel devised by him. His greatest victory was in 31 B.C. when off Actium, near Preveza in western Greece, Octavian and he defeated Anthony and Cleopatra. He was one of the main helpers of Octavian when in 27 B.C. the latter was invested with special powers and the title Augustus. In 23 B.C. Augustus, as he was ill, handed his signet-ring to Agrippa, thus indicating him as acting emperor. The same year Agrippa was given charge of all the eastern parts of the Empire, with headquarters at Mitylene. In 21 B.C. he returned to Rome and married Augustus’ daughter Julia. After he had put down the Cantabri of northern Spain in 19 B.C., he returned to Rome more permanently and was given additional favors. From 17/16 to 13 B.C. he was pacifying the eastern provinces, and in 12 B.C. went to Pannonia, but died shortly after his return. Augustus had a practical interest in sponsoring the new map of the inhabited world entrusted to Agrippa. On the re-establishment of peace after the civil wars, he was determined on the one hand to found new colonies to provide land for discharged veterans, on the other hand to build up a new image of Rome as the benevolent head of a vast empire. Mapping enabled him to carry out these objectives and to perfect a task begun by Julius Caesar. It became, among other things, a useful tool in the propaganda of imperial Rome. Agrippa was an obvious choice as composer of such a map, being a naval man who had traveled widely and had an interest in the technical side. He must have had plans drawn, and may even have devised and used large-scale maps to help him with the conversion of Lake Avemus and the Lacus Lucrinus into naval ports. 3 Orbis Terrarum #118 Incomplete at Agrippa’s death in 12 B.C., Augustus himself completed the world map. It was erected in Rome on the wall of a portico named after Agrippa, which extended along the east side of the Via Lata [modern Via del Corso]. This portico, of which fragments have been found near Via del Tritone, was usually called Porticus Vipsania, but may have been the same as the one that Martial calls Porticus Europæ, probably from a painting of Europa on its walls. The building of this colonnade was under taken by Agrippa’s sister Vipsania Polla. The date at which the building was started is not known, but it was still incomplete in 7 B.C. Whether the map was painted or engraved on the wall we do not know. The theory that it was circular is in conflict with a shape that would suit a colonnade wall. Some scholars believe that the map is even likely to have been rectangular, probably with North rather than South at the top. The chief ancient writer who refers to Agrippa’s map is the elder Pliny, who frequently quotes Agrippa by name; though whether in most cases his source is the map or the commentary is hard to say. Pliny’s most specific reference to the map is where he records that the length of Bætica, the southern Spanish province, given as 475 Roman miles and its width as 258 Roman miles, whereas the width could still be correct, depending on how it was calculated. Pliny continues: “Who would believe that Agrippa, a very careful man who took great pains over his work, should, when he was going to set up the map to be looked at by the people of Rome, have made this mistake, and how could Augustus have accepted it? For it was Augustus who, when Agrippa’s sister had begun building the portico, carried through the scheme from the intention and notes [commentarii] of M. Agrippa.” In point of fact Augustus may have delegated the detailed checking to one of his freedmen, such as his librarian C. Iulius Hyginus. Certain phases in Pliny lead one to suppose that they came from a commentary, not a map. Thus Agrippa is said to have written that the whole coast of the Caspian from the Casus River consists of very high cliffs, which prevent landing for 425 miles. If the commentary had not been continuous, but had merely served as supplementary notes where required, there is a possibility that by Pliny’s time, some eighty years later, it might have gone out of circulation. Two late geographical writings, the Divisio orbis and the Dimensuratio provinciarum (commonly abbreviated to Divisio and Dimensuratio) may be thought to come from Agrippa, because they show similarities with Pliny’s figures. There are, however, cases, e.g. the combined measurements of Macedonia, Thrace and the Hellespont, that agree with Pliny on areas where he does not name Agrippa but may nevertheless in fact have been using him. We may treat as secondary sources Orosius, Historiae adversum paganos, and the Irish geographical writer Dicuil (AD 825). Orosius seems to have read, and followed fairly closely both Agrippa and Pliny, as well as early writers from Eratosthenes onwards. Dicuil tells us that he followed Pliny except where he had reason to believe that Pliny was wrong. It is also claimed that Strabo (#115) obtained his figures for Italy, Corsica, Sardinia and Sicily from Agrippa. His source was clearly one commissioned by Romans, not Greeks, as his figures for those areas are in miles, not stades. But Strabo never names this source, referring only to the chorographer. Such a word certainly ties up with Divisio I: “The world is divided up into three parts, named Europe, Asia, Libya or Africa. Augustus was the first to show it [the world] by chorography.” Evidently there is a slight difference of 4 Orbis Terrarum #118 meaning between this and Ptolemy’s definition, by which chorography refers to regional mapping. Although the term chorographia literally means “regional topography”, it seems to include fairly detailed cartography of the known world. The Agrippa map probably did not, in the absence of any mention, use any system of latitude and longitude. It no doubt inherited a system of regional shapes from Eratosthenes (#112). It is, as one might expect, more accurate in well-known than less-known parts, and more accurate for land than for sea areas. From the quotations given by Dilke, there would appear to be a general tendency by Agrippa to underestimate land distances in Gaul, Germany and in the Far East, and to overestimate sea distances. If West Africa is any guide, in areas where distances were not well established, they were probably entered only very selectively. What purpose was served by giving a width for the long strip from the Black Sea to the Baltic Sea is not clear. For a more complete assessment of what Agrippa wrote or ordered to be put on his map, we may again turn to passages where Pliny quotes him specifically as reference. These include both land and sea measurements, though the most common are lengths and breadths of provinces or groups of provinces. In this context, length normally means the greater of the two measurements. The fact that for continental measurements it also usually means west-east or north-west/south-east is largely coincidental. Although the words used are longitudo and latitudo, they have no connection with longitudinal and latitudinal degree divisions. O.A.W. Dilke provides a detailed discussion of Agrippa’s measurements using quotes from the elder Pliny’s Natural History. It is a pity that Pliny, who seems to be chiefly interested in measurements, gives us so little other information about Agrippa’s map. For a general description, however, of what is meant by chorography we may turn to Strabo ii.5.17 (as mentioned above, Strabo nowhere names Agrippa as his source): It is the sea above all which shapes and defines the land, fashioning gulfs, oceans and straits, and likewise isthmuses, peninsulas and promontories. But rivers and mountains too help with this. It is through such features that continents, nations, favorable sites of cities, and other refinements have been conceived, features of which a regional [chorographic] map is full; one also finds a quantity of islands scattered over the seas and along the coasts. Clearly Agrippa’s map had many of the above features, but whether it also contained main roads is uncertain. But on the credit side, Agrippa’s map, sponsored by Augustus, was obviously an improvement on that of Julius Caesar on which it is likely to have been based. The fact that such an insignificant and distant place as Charax was named on the map shows the detail that it embodied. Moreover it seems to have been the first Latin map to be accompanied by notes or commentary. Romans going to colonies, particularly outside Italy, could obtain information about the location or characteristics of a particular place. Also the full extent of the Roman Empire could be seen at a glance. Certain medieval maps, including the Hereford and Ebstorf world maps (see monographs #224 and #226 in Book IIB) are now believed to have been derived from the Orbis Terraum of Agrippa, and point to the existence of a series of maps, now lost, that carried the traditions of Roman cartography into Christian Europe. The small T-O maps (see Book IIA, #205) so popular in later Roman times may, themselves, have been derived from reductions of the Agrippa map (these were the ubiquitous type of diagrammatic 5 Orbis Terrarum #118 map of the world inserted in many geographical treatises of the later Roman/early Medieval period). The following is an excerpt from an article by J.J. Tierney. The map of Agrippa was not the first map to be set up in a public place in ancient Rome. Livy tells us (XLI, 28, 8) that the consul Ti Sempronius Gracchus in the year 174 B.C. set up a map of Sardinia in the temple of Mater Matuta in memory of his campaign in the island. Varro, in the following century, tells us of a map of Italy that was painted on the wall of the temple of Tellus. We do not know the occasion of this dedication, but since it was meant to celebrate a victory it may been intended for the geographical instruction of the Roman public. The map of Agrippa, however, was set up, not in a sacred place, but in a portico or stoa open to the public, the Porticus Vipsania. It was not a map of a part of the Empire, not even a map of the Empire as a whole, but rather a map of the whole known world, of which the Roman Empire was merely one part. We know, therefore, that when Agrippa died in 12 B.C. he had not as yet begun this particular one of his many buildings. His sister, Vipsania Polla, began the work, and we know from Dio Cassius that it was still unfinished in the year 7 B.C. It was completed by Augustus at a later date, probably within the following few years. We know that the campus Agrippae was in the campus Markus on the east side of the Via Flaminia and that it was bordered towards the street on the west by the Porticus Vipsania. Remains of the portico are stated to have been found opposite the Piazza Colonna on the Corso at about the position of the column of Marcus Aurelius and further north. They are said to allow the conclusion that it had the same dimensions and construction as the adjacent porticus saeptorum, whose dimensions were 1,500 ft. in length by 200 ft in breadth. The Porticus Vipsania was, therefore, an enormous colonnade and it follows that the map with which we are concerned was only one decorative item among the many that adorned it. The largest imaginable dimensions of such a map would be a rectangle of say 30 ft. by 60 ft. and even this would only cover one-third of one of the short walls of the portico (note that the only known “reconstructions” of Agrippa’s map use a circular design). The physical appearance and construction of the map have long been matters of dispute. The general history of ancient cartography and our knowledge of Roman buildings in the Augustan period would appear to be our surest guides. There can be no reasonable doubt that the map was a rectangular one with the east-west measurements running horizontally and the north-south measurements running vertically. In regard to the materials of construction I think we have to choose between the painted type of wall- map mentioned by Varro and the construction of marble slabs that is used in the forma urbis Romae of two centuries later, of which considerable parts are extant. In view of the widespread use of marble facings that characterizes the age of Augustus, the marble slab method appears more probable. Inscriptions might either be painted or engraved and painted, on this marble surface. The Greek geographer Strabo was in Rome in 7 B.C., or shortly afterwards, and he several times mentions “the chorographer,” “the chorography” and “the chorographic map”. It is now the common view of scholars that these passages refer to Agrippa and to his map. Though Strabo does not mention Agrippa’s name here he is probably merely being tactful with regard to the Emperor who, presumably, took a large part in the completion of the portico with its map. A serious point of disagreement among scholars has been whether the commentarii of Agrippa mentioned by Pliny were published at the time of construction of 6 Orbis Terrarum #118 the portico. The German philologist and historian Detlev Detlefsen always clung steadfastly the view that there was no such publication and that the inscriptions on the map itself provided all the geographical information that was available to later times under the name of Agrippa. One must grant Detlefsen that in Pliny’s main reference there is talk only of a map and the commentarii are merely the basis of the map. Detlefsen, as against the view of Partsch, effectively quoted the passage of the younger Pliny, on the 160 volumes of his uncle’s commentarii, which he describes as electorum… commentarios, opisthographos quidem et minutissime scriptos, annotated excerpts, written on the back in a minute hand. Riese and Partsch had argued that certain references to Agrippa in Pliny, in particular the reference to the inaccessibility of part of the coast of the Caspian Sea and also that to the Punic origins of the coastal towns of Baetica refer more naturally to a published work than to the map in the Porticus Vipsania. The volumes of commentary referred to by the younger Pliny were not published, but were clearly digested to the point where little further work was needed to prepare them for publication, and the same situation may well be accepted for the commentarii of Agrippa. Another element in this problem that demands some explanation is the origin of the two later works the Demensuratio provinciarum and the Divisio orbis terrarum which are both derived from Agrippa, probably through a common source. Detlefsen had explained their origin by assuming the production of smaller hand-copies of Agrippa’s map, their smallness then making a written text desirable. Partsch on the contrary, had assumed an original publication, contemporary with the original map, of a Tabellenwerk, that is, a series of tabulated lists. It is, of course, possible to imagine that tabulated lists were put up as an adjunct to the map at the short ends, but the references to Spain and the Caspian seem somewhat out of place even here, and the balance of probability on this problem seems to lie, although rather precariously in favor of a contemporary, or nearly contemporary, publication of at least a selection of Agrippa’s material comprising something more than mere lists of names and figures. Apart from the information supplied by Pliny, our chief evidence for the reconstruction of the map is provided by the two works already mentioned, the Demensuratio provinciarum and the Divisio orbis terrarum. The Demensuratio was first published in Milan in A.D. 1475, and, according to Paul Schnabel (Text und Karten des Ptolemãus, 1938) the thirteen manuscripts of the 15th and 16th centuries and a further manuscript of the 13th century all derive from the ninth-century codex in the library of Merton College, Oxford. The other work, the Divisio orbis terrarum was first edited by Schweder in A.D. 1876 from a Vatican manuscript of the 13th century. The text, however, had long previously been known from its reproduction in the first five chapters of the De Mensura Orbis Terrae, published by the Irish scholar Dicuil in A.D. 825. Dicuil worked and wrote probably at the Frankish palace at Aix-la-Chapelle in the time of Charlemagne and Louis the Pious. Dicuil, in his preface, promises to give the measurement of the provinces made by the envoys of the Emperor Theodosius, and at the end of chapter five he quotes twelve verses of these envoys in which they describe their procedure. The lines themselves, however, show that Dicuil largely misconceived the situation. What they show is rather that on the orders of Theodosius two members of his household composed a map of the world, one written and the other a painting. They completed the work in a few months. We do not know the size of this map that has perished, or whether its descent from the map of Agrippa was through a series of hand-copies as Detlefsen supposed. But it did quite clearly derive from that map, whether in the map- 7 Orbis Terrarum #118 form or in a written form, with its list of seas, mountains, rivers, harbors, gulfs and cities. The twelve lines were inscribed on this map and also on an obviously contemporary written version thereof, and it is this written version that has been preserved for us both by Dicuil and in the various manuscripts of the Divisio orbis terrarum, whereas the map has perished. The date of the making of the map was probably the fifteenth consulate of Theodosius II, that is, A.D. 435, not the fifteenth year of his reign, as understood by Dicuil. According to Tierney, Detlefsen regarded these two works as derived from small-scale copies of Agrippa’s map. Small discrepancies were to be explained by differences in the copies of the map used by each. The classical scholar Alfred Klotz, however, in his articles on the map has shown that a number of correspondences between the two works as against Pliny point rather to a common source from which both works are derived. Detlefsen’s view that both works were the transformation of actual maps into a written record had the advantage that the differences in the order of the material in the two works was of little consequence, the map giving merely the visual impact, and the writer “being free to begin his description at whatever point on the map he preferred”. In actual fact Pliny and the Divisio both begin their description from the straits of Gibraltar, moving east, while the Demensuratio, on the contrary, begins with India moves west. On Klotz’s view that both works derive from a common written source this major divergence becomes a problem to he explained, but Klotz can offer no explanation. Klotz, however, believed that be could determine the original succession of countries and groups of countries as treated in the published work of Agrippa by criteria. One of these was the direction shown in the order of naming several particular countries where several are included in the same section, or the direction shown in the list of the boundaries of the section. Thus he says that the order Cevennes-Jura for the northern boundary of Narbonese Gaul shows motion from west to east, and again the list Macedonia, Hellespont, left side it the Black Sea shows the same movement. The second criterion is that the use (the alleged use) of the term longitudo for a north-south direction or for any direction other than the canonical one of east-west, shows us the direction of Agrippa’s order in treating of the geography. Tierney does not believe that either of these criteria can show us the order of treatment in the original publication, presuming, that is, that there was an original publication. We are prone to forget that all ancient geographers were necessarily map-minded, and even when the map was not before their physical eye, it was before their mental eye. The order of countries within a section would, I think, very much depend on the momentary motions or aberrations of that mental eye. Tierney believes however, that the west to east movement supposed by Klotz is, in fact, correct, but not for his reasons. The west to east movement in geographical description seems to have been normal in the geographers from the time of the sixth- century Massaliot periplus [the Latinization of the Greek word περίπλους (periplous, contracted from periploos), literally “a sailing-around”] of which part is preserved in the Ora Maritima of Festus Rufus Avienus, right down to the map of Ptolemy. Greek cartography, like Greek writing, ran from left to right and perhaps the former practice was promoted by the fact that the western boundary of Europe was well known, at least from the time of Pytheas (ca. fourth century B.C.), whereas the eastern boundary of Asia was always unknown and fluctuating. Our idea of the detail of the map of Agrippa must be based on a study of the references in Strabo, Pliny, the Divisio and the Demensuratio. Strabo used Agrippa only 8 Orbis Terrarum #118 for Italy and the neighboring islands, so that our chief evidence comes from the other three sources. Occasionally there is confirmation from the chapter on geography in Orosius. There are twenty-four sections in the Divisio and thirty in the Demensuratio, the difference being mainly due to the absence from the Divisio of the sections on the islands of the Mediterranean and the Atlantic. Detlefsen believed that the island sections were later added to the Demensuratio, but according to Tierney there can be little doubt that he is wrong in this and that Klotz is right in thinking that these sections have rather fallen out of the Divisio. The elements provided in each section are rather monotonously the same. First come the boundaries of the province given in the constant order east, west, north, south. Following on this and connected there with come the longitude and latitude, in that order, expressed in Roman miles with Roman numerals. The boundaries are marked by the natural features, usually the mountains, rivers, deserts and oceans, only occasionally by towns or other features. The Demensuratio on one occasion gives the fauna and flora of Eastern India, which it calls the land of pepper, elephants, snakes, sphinxes and parrots. These sections are largely identical with passages in Pliny’s geographical books (Books III to VI), and show that many passages in Pliny are taken from Agrippa beyond those where he is actually named. There are also numerous discrepancies. The numerals are much more corrupt than those in Pliny, and there is usually a presumption, therefore, that Pliny’s figures preserve a better version of Agrippa. The consensus of the views of modern scholars on Agrippa’s map, is that it represents a conscientious attempt to give a credible version of the geography of the known world. It relies on the general scheme of the Greek maps that had been current since the time of Eratosthenes and Hipparchus, and attempts to rectify them, particularly in Western Europe, with recent information derived from the Roman itineraries and route-books. But this consensus is not quite complete and therefore I now turn to consider the view of Agrippa’s map put forward by Professor Paul Schnabel in his article in Philologus of 1935. We have to thank Professor Schnabel for providing in the same article a new critical text of both the Demensuratio and the Divisio. Schnabel, while expressing appreciation of the earlier work of Alfred Klotz, yet criticizes Klotz severely on two grounds. Firstly, Klotz has not discussed the possible use of Agrippa in Ptolemy’s Geography, and secondly, and much more fundamentally, he has not recognized the scientific importance of the world-map of Agrippa as a link between Eratosthenes and Hipparchus on the one hand and Marinus and Ptolemy on the other, but has merely repeated traditional views dating from the end of the 19th century. These views stated that Agrippa’s work was constructed on the basis of Roman itinerary measurements and took no note of the scientific results of the astronomical geography of the Greeks. Schnabel does not himself take up the general question of the use of Agrippa by Marinus and Ptolemy. He sets out, at first, to prove that Agrippa’s map possessed a network of lines of longitude and latitude. Schnabel here refers to the last chapter of the geographical books of Pliny’s Natural History, that is, Book VI, cap. 39, sections 211-219, where Pliny mentions evidently as a work of supererogation, “the subtle Greek invention” of parallels of latitude, showing the areas of equal shadows and the relationship of day and night Pliny then gives seven parallels, running at intervals between Alexandria and the mouth of the Dnieper, with longest days running from fourteen to fifteen hours. For the sixth of 9 Orbis Terrarum #118 these parallels he gives a slight correction due to the Publius Nigidius Figulus (ca. 98-45 B.C., a scholar of the Late Roman Republic). Pliny then adds “from later students” five more parallels, three of them, those of the Don, of Britain, and of Thule, running north of the original seven, and two, those of Meroe and Syene, running south of them. Of these parallels Schnabel tries to establish that at least two are due to Agrippa, to wit, the first of the new parallels passing through the Don, and the seventh of the old parallels passing through the mouth of the Dnieper. He argues that the Dnieper was used as a line of demarcation between Sarmatia to the east and Dacia to the west only on the map of Agrippa, and neither earlier nor later, and that therefore the parallel from the Don through the Dnieper must derive from that map. This argument, however, is unsound for a number of reasons, of which the most obvious in this context is that whereas the Dnieper is given by our sources as the west boundary of Sarmatia, it is never given as the east boundary of Dacia. For this east boundary our three sources unanimously give instead the deserts of Sarmatia. Schnabel continues with the negative argument that the Don parallel cannot belong to the school of Hipparchus. Hipparchus had made the Don parallel the seventeen-hour parallel, corresponding to 54° N latitude, whereas Pliny here puts it at sixteen hours or 48° 30" N latitude, which is nearly correct. Therefore, Schnabel argues, the incorrect latitude of Hipparchus was corrected by Agrippa who had experience of the Black Sea in his later years. It follows further, says Schnabel, that Pliny’s seventh parallel, that of fifteen hours, belongs equally to Agrippa. Here again Pliny’s figure is nearly correct while that of Hipparchus is 2° too high. Both these arguments of Schnabel must however, also be rejected. In the first place the heading of Pliny’s chapter on the parallels, cap. 39, section 211, refers obviously to all that follows as far as the end of Book VI and shows that the complete passage is taken from Greek sources. His proximate source, moreover, he names in section 217, according to his usual custom, as Nigidius Figulus. Detlefsen pointed this out in 1909, and Kroll (after Honigmann) throws further light on the subject in his article on Nigidius in R. E. XVII, 200-212 (1936). Pliny’s seven klimata are a piece of astrological geography and derive through Nigidius from Serapion of Antioch, who was probably a pupil of Hipparchus, or if not was a student of his work. Nigidius was a notorious student of the occult and his astrological geography was contained in a work apparently entitled de terris. This work seems to have included his commentary on the sphaera Graecanica describing the Greek constellations and his sphaera barbarica on the non-Greek constellations. Nigidius’ “barbaric sphere” was derived from the like-named work of Asclepiades of Myrlea. The detailed extension of the Greek parallels into the Roman west is apparently due to Nigidius. We may speculate as to whether Pliny’s phrase regarding the “careful later students,” does not refer to Nigidius himself. This appears improbable. The parallels of Meroe and Thule were both equally useless in astrological geography although they may well have been mentioned by Nigidius, as they are by Strabo simply by way of clearing the decks. Pliny’s final phrase about these scholars adding half an hour to all parallels denotes rather the astronomer than the astrologer. What is new in Pliny’s parallels may be referred to the Greek astronomers of the age of Hipparchus or the two or three generations after him. Many of these men’s names are known, but very little more is known of most of them. But there is no vestige of probability or proof that Agrippa made new gnomonic readings to correct Hipparchus. As Detlefsen put it: 1 0