Q: How do you tell if a globe is celestial?

By the presence of a graduated ecliptic, which is structural and survives wear better than any drawing. Named figures alone are not decisive: some terrestrial globes carry allegorical figures in the oceans, and some celestial globes are worn until only figures remain. Where the ecliptic is illegible the Institute records the sphere as indeterminate rather than guessing.

What is the difference between a celestial and a terrestrial globe?

Construction is the same: both are spheres carrying pasted gores on an inclined axis. The difference is that the sky is seen from inside and the earth from outside, so a celestial globe showing the constellations as they appear from the earth must be drawn mirrored. The Brennholt Institute records that choice as CV-a, viewed from outside, or CV-b, viewed from within.

Can a celestial globe be dated?

Less reliably than a terrestrial one. Constellations do not change, so there is no equivalent of the political geography that anchors a terrestrial date. What changes is the star catalogue used and the epoch to which positions are reduced, and that is stated on perhaps a third of examples: of six celestial globes in the Institute's 2025 sample, two carried an epoch.

Are celestial and terrestrial globes sold in pairs?

They were frequently made as pairs, and a pair is worth more than the two halves, which is why pairs are reassembled after the fact from unrelated spheres. The Institute records a pair as genuine only where sphere diameter, gore configuration and stand type all agree. In one 2025 case the two spheres differed by 19 mm in diameter and the listing did not mention it.

Celestial and terrestrial globes: what changes when the map is the sky

A celestial globe is built like a terrestrial one and read like its opposite. Both are spheres carrying pasted gores; both are mounted on an inclined axis; both are catalogued, sold and graded in the same trade. But the sky is seen from inside and the earth from outside, and a sphere that shows the sky correctly to an observer standing outside it must therefore be drawn mirrored.

The Institute grades and classifies both on the same scales — the six condition states and the seven stand types — with three departures set out below. No trade body has published a distinction between the two in its condition terminology, and sale catalogues routinely apply terrestrial language to celestial objects.

The three departures

Orientation is a fault class of its own: A terrestrial globe cannot be made inside out. A celestial globe can, and some were: the constellations are drawn either as seen from the earth or as seen from beyond the sphere, and the two are mirror images. The Institute records the choice as CV-a (viewed from outside, the conventional case) or CV-b (viewed from within). It is not a defect. It is a decision by the maker and it must be recorded, because a reader who assumes the wrong one reads every constellation reversed.
There is no cartouche convention: Celestial globes date badly. The political geography that anchors a terrestrial globe has no equivalent: the constellations do not change. What does change is the star catalogue used and the epoch to which positions are reduced, and that is stated on perhaps a third of examples. Of the 48 globes in the Institute's 2025 sample, 6 were celestial and only 2 carried an epoch.
Wear falls in different places: A terrestrial globe wears at the equator, where it is spun by hand. A celestial globe wears at the poles and along the ecliptic, because it is turned to bring a constellation into view rather than spun. Of the 6 celestial examples recorded, 4 showed their heaviest loss within 25 degrees of a pole, against 3 of the 42 terrestrial.

What does not change

The six condition states apply unaltered, since they describe paper, sphere and mount rather than content. So do the seven stand types and the gore configurations: a celestial sphere is gored exactly as a terrestrial one, in twelve segments of 30 degrees with two polar calottes in every example the Institute has recorded. The varnish states apply without qualification.

Pairs

Celestial and terrestrial globes were frequently made and sold as pairs, and the pair is worth more than the two halves. This creates the commonest problem in the category: pairs are reassembled after the fact from unrelated spheres. The Institute records a pair as a pair only where sphere diameter, gore configuration and stand type all agree, and states which of the three was used to accept or reject it.

Of the 6 celestial globes in the 2025 sample, 3 were catalogued as one of a pair. In one case the two spheres differed in diameter by 19 mm — a difference that no maker would produce and that the listing did not mention. The Institute recorded the pair as rejected on diameter and published the figure, which is its practice: a rejection without its measurement is an opinion.

A caution on identification

A sphere showing named figures rather than coastlines is not automatically celestial; some terrestrial globes carry allegorical figures in the oceans, and some celestial globes are so worn that only the figures survive. The Institute identifies on the presence of a graduated ecliptic, which is structural and survives wear better than any drawing, and where the ecliptic is illegible it records the sphere as indeterminate rather than guessing. One of the 6 was so recorded.