What Is a Star Atlas? Getting From a Catalogue to Something Usable

Mapping a sphere onto flat pages raises the same difficulties as terrestrial cartography and a few of its own. A projection must be chosen and every one distorts, so charts generally use several different projections for different sky regions or accept distortion towards the edges. Depth must be decided, meaning how faint a star to include, and that choice determines whether a chart matches what a viewer sees with the naked eye, with binoculars or with a telescope, and a chart showing far more stars than the instrument reveals is as useless as one showing too few. Symbol conventions must encode magnitude by dot size and object type by shape, consistently enough to be read in red light at night. Labelling must identify enough objects to orient the user without obscuring the field. Overlap between pages matters enormously, since a target falling exactly on a page edge is the commonest practical frustration. And the whole thing must work upside down and reversed, since telescopes invert.

A small number of works defined the form and several are objects of substantial artistic interest:

• •Bayer's Uranometria of 1603, which introduced the system of labelling stars by Greek letter within each constellation and which is still in use
• •Hevelius's atlas of 1690, published posthumously by his wife Elisabeth, who was herself an observer and completed the work
• •Flamsteed's Atlas Coelestis of 1729, which used a different projection convention and whose numbering system for stars remains standard
• •Bode's Uranographia of 1801, the most elaborate of the great engraved atlases and the last major one to depict constellation figures in full
• •Argelander's Bonner Durchmusterung in the 1850s and 1860s, which abandoned artistic figures for a systematic survey and marks the shift from atlas as artwork to atlas as data
• •Norton's Star Atlas from 1910, which established the format of a practical observing atlas and ran through many editions across a century

Early atlases drew the constellation figures as pictures, which was beautiful and increasingly impractical. The figures are inconsistent between atlases, since each artist drew them differently, and stars sit ambiguously within them so that the same star could be assigned to different figures by different cartographers. The boundaries were undefined until fixed internationally in 1930, after which a constellation became a region rather than a picture and the figures became decorative. Modern charts draw simple stick lines connecting the principal stars, which is a convention rather than anything official, and different publishers connect them differently. The shift tracks a change in what an atlas is for, since a chart used to identify a mythological figure needs the drawing while one used to locate a faint galaxy needs accurate positions and nothing else. Both continue to be published, with decorative reproductions of the great engraved atlases selling steadily alongside the working charts.

Planetarium software and phone applications have displaced printed atlases for most purposes and the change is not purely a loss. Software shows the sky for any time and location, orients automatically, zooms continuously, filters by object type and magnitude, and connects directly to telescope mounts that will slew to a chosen target. Databases behind them contain vastly more objects than any printed volume. Augmented reality applications identify what a phone is pointed at, which is genuinely useful for beginners. What is lost is worth naming, since a printed chart shows a whole region at once in a way a screen does not, it does not destroy dark adaptation, it does not require power, and using one builds a mental map of the sky that automated pointing bypasses entirely. Experienced observers frequently use both, and the practice of star hopping, meaning navigating from a bright star to a faint target through recognised patterns, remains a taught skill precisely because it produces that mental map.

Projection, depth, symbols and page overlap all determine whether a chart can actually be used, and one showing far more stars than the instrument reveals is as bad as one showing too few. Bayer's Greek letters from 1603 and Flamsteed's numbers from 1729 are both still standard. Constellation figures were drawn inconsistently until boundaries were fixed in 1930, after which they became decorative.