Halley's Discovery of NGC 6231 and the Hazards of Early Star Nomenclature

A description and several fine photographs of NGC 6231 may be conveniently found in Burnham Robert Jr , Burnham's celestial handbook (3 vols, New York, 1978), iii, 1722–6.

See for example Jones Glyn Kenneth , The search for the nebulae (Chalfont St Giles, 1975), 78, no. 53. Under the heading “Nos in other Catalogues”, Glyn Jones references for NGC 6231 only Lac.II.13, referring to Lacaille's 1755 catalogue. Checking older sources, we find that John Herschel, in his “Catalogue of nebulae”, Philosophical transactions of the Royal Society of London, cliv (1864), 122, no. 4245, cites as authority only Dunlop, as Δ499. Turning to James Dunlop's nebulae catalogue, ibid., cxviii (1828), 139, no. 499, we are referred only to Bode, and consulting Bode's Allgemeine Beschreibung und Nachweisung der Gestirne (Berlin, 1801), 58, no. 150, we are informed that the cluster was observed by Lacaille. The trail stops at Lacaille. We shall discuss Lacaille's observation in more detail below.

Apparently the only authority to suggest that Halley observed NGC 6231 was Dreyer J. L. E. , who in his “New general catalogue”, Memoirs of the Royal Astronomical Society, xlix (1887–89), 171, no. 6231, included Halley along with Lac.II.13 and Δ499 under the heading of “other observer”. This single reference to Halley has either been overlooked or ignored, perhaps because Dreyer did not indicate where or when Halley might have observed the cluster.

I have consulted the following: Ptolemy , Almagestum (Venice, 1515), 84r, and idem (Venice, 1528), 79r; Halma Nicolas , Composition mathematique de Claude Ptolémée (2 vols, Paris, 1816), ii, 58–61; Baily Francis , “The catalogues of Ptolemy …”, Memoirs of the Royal Astronomical Society, xiii (1843), 44–45; Peters C. H. F. and Knobel E. B. , Ptolemy's catalogue of stars: A revision of the Almagest (Washington, 1915), 40, 63, 86, 105.

Identifying any single Ptolemaic star can be tricky, since the longitudes are much too small for the supposed epoch of a.d. 138, the latitudes suffer from a value for the obliquity that is about 10′ too large, and observational errors of 10–15′ are typical. See Pedersen Olaf , A survey of the Almagest (Acta historica scientiarum naturalium et medicinalium, xxx, Odense, 1974), 250–8. However, if one looks at groups of stars and considers only relative positions, most Ptolemaic stars can be reliably correlated with existing stars, since the positions of stars in the same region tend to err in the same direction by the same magnitude. For example, it is not comforting at first to realize that if Ptolemy's 12 Scorpii is our 26ε, then his value for the longitude was 63′ too small, and the latitude was 19′ too far north. However, when the rest of the identifications are made for the tail stars and the average difference turns out to be 69′ in longitude and 16′ in latitude, then the identifications as a group hold up. See Peters , Ptolemy's catalogue, 63, where he compares Ptolemy's positions for all the Scorpio stars with the true positions for epoch a.d. 100.

Peters was the first editor of Ptolemy to suggest that the latitude of 15 Scorpii should be 19°. It is strange then that he followed Baily and Halma in identifying 14 and 15 Scorpii as ζ¹ and ζ². If this were true, then 14 Scorpii would have a latitude error of 45′, while 15 Scorpii would be in error by only 16′. But if one equates 14 Scorpii with NGC 6231, which is about 30′ due north of ζ¹/ζ², then its latitude error would be reduced to 15′, giving it the same relative error as the surrounding stars. This seems rather convincing evidence that Ptolemy's 14 Scorpii was indeed NGC 6231. I have used the latitude errors calculated by Peters , Ptolemy's catalogue, 63.

Bayer Johann , Uranometria (Augsburg, 1603). Considering its importance as the first modern star atlas, it is surprising that the Uranometria has never been the subject of a worthy critical study. There is however much useful information in Warner Deborah J. , The sky explored: Celestial cartography 1500–1800 (Amsterdam, 1979), 18–19.

See Flamsteed's John introduction to his Atlas coelestis (London, 1729), where he takes Bayer to task for failing to make his constellation figures conform to Ptolemy's descriptions.

The Scorpio map is plate 29 in the Uranometria. The explanatory table, in the 1603 edition, is printed on the back of the map; in later editions the tables and maps were printed, and sold, separately.

Bayer's reliance on Tycho's catalogue is probably too well-known, because the dependence is generally exaggerated. Not only does Bayer include many northern stars that are not in Tycho's catalogue, but many of Tycho's catalogued stars are not inserted on the plates at the positions given by Tycho. The actual extent of the Uranometria's debt to Tycho is a matter that needs considerably more investigation.

Bayer did have a catalogue of the Scorpio tail stars available that was independent of Ptolemy, namely the one compiled by Pieter Keyzer in 1595–96. This catalogue included much more than just the twelve new southern constellations that Bayer depicted in plate 49 of his atlas. The Keyzer catalogue was not published separately, although parts of it later formed the basis for the catalogue of southern stars included in Kepler's Tabulae Rudolphinae, but the catalogue of Frederick de Houtman did get published, and since it has been established that Houtman's catalogue was copied from Keyzer, we may look there to see whether Keyzer observed the Scorpio tail region. And indeed we find that all the tail stars below ε were observed. Ignoring absolute declinations and considering only relative positions, we find listed: μ, then a star 3°50′ south, then another 36′ further south, and a fourth which is another degree south and 5° east. This list is perfectly consistent, and only consistent, with an identification of μ, NGC 6231, ζ, and η. Note that this catalogue is free of the latitude error made by Ptolemy for 15 Scorpii (ζ). Had Bayer used these positions, the later problems, and indeed this entire article, would never have developed. For a translation of the Houtman catalogue and a defence of its derivation from Keyzer's, see Knobel E. B. , “On Frederick de Houtman's catalogue of southern stars, and the origin of the southern constellations”, Monthly notices of the Royal Astronomical Society, lxxvii (1917), 414–32.

Ptolemy's value for the obliquity was 23°51′20′', which was about 10′ too large and responsible for 10′ errors in latitude for stars in the region of Scorpio. It was of course not known in the late Renaissance that Ptolemy's value was incorrect, but it was apparent that the obliquity had changed; whether the change was secular or periodic was much debated. Tycho's value was 23°31′30′', which was itself 2′ too large. See Pedersen , Survey, 95; Dreyer J. L. E. , Tycho Brahe (Edinburgh, 1890), 123.

By “true position” I mean the position the star would have had if the Ptolemaic latitude had been corrected to 19° and the change in the obliquity had been taken into account.

Schiller Julius , Coelum stellatum christianum (Augsburg, 1627). Schiller's atlas is usually treated as merely a quaint curiosity, but the Christian iconography aside, it contains a great deal of current astronomical information, and a detailed study would be most welcome. Warner , Sky explored, 229–32, contains an account which is unusually sympathetic and very informative.

Kepler Johann , Tabulae Rudolphinae (Ulm, 1627), 116. This is in the section of the catalogue which contains stars of the Second Class, i.e., those not observed by Tycho. Kepler first lists in roman type seven stars extracted from his own De stella nova, followed by ten stars, in italic type, from Ptolemy. The description of 14 Scorpii, the first entry in italic, is: “ex Ptol. In 3 spond. boreal. Cl. austr.”, which in less truncated language means that this was the northern star in the third segment according to Ptolemy, but the southern according to Clavius. Clavius was one of the first astronomers to attempt a critical edition of the Ptolemaic star catalogue, including a version with corrections in the many editions of his In sphaerum Ioannis de Sacro Bosco commentarius (3rd ed., Venice, 1596). Kepler gave the latitude of 14 Scorpii as 19° and 15 Scorpii as 18°20′, obtaining these figures by adding 20′ to the Ptolemaic values.

See Beg Ulugh , Tabulae long. ac lat. stellarum fixarum (Oxford, 1665), 88. This edition by Thomas Hyde was the first printing of Beg's complete star catalogue.

For an excellent account of Halley's compilation of his southern catalogue, see Armitage Angus , Edmond Halley (London, 1966), 24–36.

Halley Edmond , Catalogus stellarum australium (London, 1679), [2].

There can be little doubt about the identification; the position of Halley's nebula for 1950, using his sextant distances but modern coordinates for the determining stars, is α₁₉₅₀ = 16^h9·9′, δ₁₉₅₀ = −41°45′, a position which is right in the middle of NGC 6231.

Hevelius Johann , Prodromus astronomiae (Gdansk, 1690). It is not generally realized that Hevelius reprinted the entire Halley southern catalogue, for the year 1700, in the Prodromus, 309–20. Halley's 20 Scorpii is properly described as nebulous, p. 310. The map of Scorpio is Fig. II of the atlas, which, although included in the Prodromus, is sometimes referred to by its own title, Firmamentum Sobiescianum sive Uranographia.

The Sharp catalogue of southern stars is found in Flamsteed John , Historia coelestis (3 vols, London, 1725), iii, 77–83, continuing the pagination of Flamsteed's “Catalogus Britannicus”. Although the catalogue was based on Halley's observations, Sharp did not even mention Halley's name. The omission was quite intentional; see Crosthwait's letter to Sharp of 1722, in Baily Francis , An account of the Revd. John Flamsteed (London, 1835), 353, no. 259. Another unappreciated fact about Halley's southern catalogue that might be worth mentioning is that Halley printed the sextant observations right along with the star positions in his Catalogus of 1679. These were left out of the Hevelius and Sharp reprints, and were also omitted from the most widely used edition of Halley's catalogue, that compiled by Baily Francis , Memoirs of the Royal Astronomical Society, xiii (1843), 167–81.

Bevis John , Uranographia Britannica (London, c. 1750). See my forthcoming article, “John Bevis and his Uranographia (c. 1750)”, Proceedings of the American Philosophical Society (1981).

That Bevis used the Sharp reprint rather than the original Halley catalogue (or the Hevelius reprint) is apparent from the positions which Bevis gives for the Halley stars in his catalogue; they are simply the Sharp positions precessed to 1750, the epoch of the Bevis catalogue. Apparently only one copy of the Bevis star catalogue has survived; it is in the Library of the American Philosophical Society in Philadelphia. For complex reasons, Bevis's atlas and catalogue had slightly different epochs.

Jones Glyn , Search for the nebulae, 24.

Lacaille Abbé , “Table des ascensions droites et des déclinaisons apparentes des Étoiles australes”, Mémoires de l'Académie royale des Sciences for 1752 (Paris, 1756), 539–92. The Scorpio stars are given on p. 574. Lacaille's nebulae discoveries are discussed in Gingerich Owen , “Abbé Lacaille's list of clusters and nebulae”, Sky and telescope, xix (1960), 206–8, and Jones Glyn , Search for the nebulae, 44–49.