Globes are sold by diameter, and the diameter quoted is frequently not the diameter of anything that can be measured. It may be the sphere, the sphere plus its varnish, the horizon ring, or the overall height of the object on its stand. No trade body has published a convention for what the stated size of a globe refers to. The Brennholt Institute of Applied Taxonomy therefore records three figures where catalogues record one.
Of the 48 globes recorded by the Institute from public sale catalogues between March and December 2025, 41 quoted a single unqualified figure. Where that figure could be checked against photographs and accompanying dimensions, 9 of the 41 corresponded to the swing diameter and not to the sphere. A buyer comparing two listings is, about one time in five, comparing two different quantities.
Diameters cluster, because the moulds and the printed gore sets are made in series. The clusters below account for 44 of the 48 in the sample. The Institute publishes them as observed frequencies and not as a set of standard sizes, since no body has ratified one.
| Sphere diameter | In inches, as commonly quoted | In sample | Typical use |
|---|---|---|---|
| 76–80 mm | 3 in | 3 | Desk ornament; often a pocket or pedestal globe |
| 100–115 mm | 4 in | 6 | Desk |
| 150–160 mm | 6 in | 9 | Desk, the commonest small size |
| 200–215 mm | 8 in | 11 | Desk and school |
| 300–310 mm | 12 in | 10 | School and library; the commonest size overall |
| 400–410 mm | 16 in | 3 | Library, floor-standing |
| 500 mm and above | 20 in and above | 2 | Institutional; usually GC-2 or GC-3 gore configuration |
The clusters fall on whole inches even in markets that have used millimetres for a century, because the gore sets and the sphere moulds outlive the units they were designed in. A sphere described as 300 mm is in almost every case a twelve-inch sphere at 304.8 mm, rounded down for the catalogue. The Institute records the measured figure and the nominal figure separately, and declines to reconcile them, since a rounded nominal is a commercial description and not a measurement.
The two are related but not determined. All of the spheres below 200 mm in the sample were GC-1, twelve gores. Above 400 mm, four of five were GC-2 or GC-3. Between those bounds both occur. The governing constraint is how far the paper will stretch at the waist of a gore, which is a property of the paper and not of the sphere, so the boundary moves with the stock used.
Nothing directly, and one thing indirectly. Larger spheres are heavier, are handled by the sphere rather than by the mount, and go out of round more often: of the seven globes in the sample at 400 mm or above, three were recorded out of round, against four of the remaining forty-one. Roundness enters the condition scale at G-4 and G-6, so size influences the grade without appearing in it. The Institute notes this as an observation on a small sample and not as a finding.