Laying Down the Law

This naming business clearly threatens to get out of hand. But the complication is utterly necessary if really large numbers of objects are to be specified unambiguously. The 87GB and IRAS catalogs have around 55,000 and 350,000 entries, respectively. The USNO CCD Astrograph Catalog contains about 113 million stars, and the Hubble Guide Star Catalog II has 945,592,683 objects. The Sloan Digital Sky Survey now underway has so far gathered measurements on nearly 500 million stars and galaxies, each one demanding its own individual identity.

In an effort to manage the mammoth bookkeeping tasks of the future, the International Astronomical Union has urged astronomers to assign names within a single well-defined but fairly flexible system. New names are supposed to have two elements, an "origin" and a "sequence." Optionally a "specifier" may be added too.

For example, while deep-sky observing a few years ago I came upon what appears to be an unreported open cluster in Auriga near the asterism of the Kids. If I want to name it, I could do so the old-fashioned way: start a list named for myself and call the cluster Skiff 1. This would be appropriate if I were publishing a lengthy table of newly identified clusters. As long as I never publish a different kind of list and nobody else named Skiff does either, there would be no problem calling new clusters Skiff 1, 2, 3, and so on.

A better name for Skiff 1 might be BAS J0458.2+4301. "BAS" is the "origin;" it's my initials. (One or two letters wouldn't be acceptable; S is already commonly assigned to emission nebulae cataloged by Stewart Sharpless in 1959, and BS is widely used for stars in the Yale Bright Star Catalogue. BAS isn't taken, however, so I could claim it.) The "sequence" gives the position instead of just a serial number. The J indicates that the numbers following are right ascension and declination in the J2000 system, the precisely defined coordinate grid for epoch 2000.0. The numbers are given here to a precision of 0.1 minute of right ascension and one arcminute of declination; other degrees of precision are often used. In a similar way I could use B1950 coordinates, or maybe galactic latitude and longitude. (When older designations don't specify a coordinate epoch, it's assumed to be for 1950.0.)

I can add objects to my collection more or less endlessly as long as no more than one falls in the same square-arc-minute "bin" on the sky defined by the coordinates. (There are about 150 million such bins on the celestial sphere.) In a pinch, one could add a "specifier." An interacting pair of galaxies, for instance, might be called BAS J1234.5-3456 (SW) and BAS J1234.5-3456 (NE), indicating the southwestern and northeastern components of this fictitious object.

Although a scheme like this is necessary for the multitude of objects inhabiting the sky, I find it soulless and clinical. At a 1984 meeting of star-cluster specialists, Berkeley astronomer Ivan King reacted to one presentation by mocking the prospect of identifying well-known objects with long strings of digits. At one point he told the speaker, "I'm glad to see that you identified M49 as NGC 4472, because, although you are a man of the 18th century, I live in the 19th century myself and prefer NGC numbers." Out alone on a clear night with a telescope, those simpler names are far more attractive.