Messier 38, NGC 1912 and permanent object 170 are the same open cluster. Giovanni Battista Hodierna recorded it before 1654, Guillaume Le Gentil independently described it in 1749, and Charles Messier entered it in his catalogue in 1764. That history explains why the cluster carries both an NGC number and the older observer's famous M number. Source: [SEDS, Messier discovery table](https://messier.seds.org/xtra/history/dis-tab1.html).
An open cluster is a family of stars born from the same molecular cloud and still travelling together, though it will gradually disperse through internal encounters and the Galaxy's tides. M38 is commonly assigned an age near 300 million years. At that age its lower-mass members remain on the main sequence, where colour and brightness can be compared with unusually high precision.
The cluster seen in an eyepiece is not a clean membership list. Foreground and background Milky Way stars share the field, while faint members disappear below a given instrument's limit. Gaia's common parallax and proper-motion measurements let researchers reject many interlopers before interpreting the colour–magnitude diagram; visual pattern recognition cannot make the same separation.
Griggio and collaborators selected likely members with Gaia DR3 astrometry, combined Gaia and Sloan photometry, and found that the lower main sequence is wider than measurement errors alone predict. Their astrometric calculation gives 1,132 ± 2 parsecs before, and 1,183 ± 2 parsecs after, the Gaia parallax zero-point correction—roughly 3,860 light-years for the corrected value. Those small quoted errors are formal statistical values; calibration and stellar-model systematics still matter. Source: [Griggio et al., *Monthly Notices of the Royal Astronomical Society* 523, 5148](https://academic.oup.com/mnras/article/523/4/5148/7204150).
A broad main sequence can tempt an observer to claim multiple chemical populations, as in some massive clusters. The authors tested that possibility across several colour combinations. Differential reddening—slightly different amounts of foreground dust across the field—plus unresolved binary stars reproduced the observed width. Their data did not require an intrinsic spread in helium or metallicity. That is a constraint, not proof that every cluster member has exactly identical composition: it says the proposed spread is unnecessary at the precision and mass range examined.
The cluster's stored position is about right ascension 5h 29m and declination +35° 50′. It culminates near 81° from latitude 45° N but only about 19° from 35° S, before local horizon and refraction effects. These heights are original calculations from the repository coordinates.
For a practical route, identify Capella and Elnath, the bright southern corner of Auriga's usual outline. M38 lies inside the Charioteer, about 10.4° south of Capella and 7.2° north of Elnath from the stored positions. Use binoculars or a low-power finder to match the surrounding star field before increasing magnification. NGC 1931 is about 1.68° southeast of M38, but it is much smaller and fainter; do not mistake one for the other. The rendered accepted finder and reader link remain unchecked.
Dark-adapted binoculars can show M38 as a misty or grainy patch under a good sky. A small telescope at low power begins to resolve its scattered pattern; moderate aperture resolves many more members. No visual observation separates cluster members with Gaia precision, and no eyepiece view diagnoses the main-sequence width. Log the aperture, magnification and degree of resolution, while leaving the Moon and other changing weekly conditions to the calendar.
