Are you sure of your citation?

This piqued my curiosity, so I went to Ia, Q68, A1, but I don't find these words anywhere???

His citation was incorrect, but the words were correct:

https://aquinas.cc/la/en/~ST.I-II.Q54.A2.Rep2

Ad secundum dicendum quod terram esse rotundam per aliud medium demonstrat naturalis, et per aliud astrologus, astrologus enim hoc demonstrat per media mathematica, sicut per figuras eclipsium, vel per aliud huiusmodi; naturalis vero hoc demonstrat per medium naturale, sicut per motum gravium ad medium, vel per aliud huiusmodi. Tota autem virtus demonstrationis, quae est syllogismus faciens scire, ut dicitur in I Poster., dependet ex medio. Et ideo diversa media sunt sicut diversa principia activa, secundum quae habitus scientiarum diversificantur.

Reply Obj. 2: The physicist proves the earth to be round by one means, the astronomer by another: for the latter proves this by means of mathematics, e.g., by the shapes of eclipses, or something of the sort; while the former proves it by means of physics, e.g., by the movement of heavy bodies towards the center, and so forth. Now the whole force of a demonstration, which is a syllogism producing science, as stated in Poster. i, text. 5, depends on the mean. And consequently various means are as so many active principles, in respect of which the habits of science are distinguished.

https://en.wiktionary.org/wiki/rotundus#Latin

Etymology

Generally regarded as deriving from rotō (“turn, revolve”) or rota (“wheel”) +‎ -undus. First attested in the works of Cato the Elder(circa 200 BC).

It has also been suggested that the alternative form retundus, whence most of the Romance descendants derive, actually reflects the original Latin form (despite only being attested from the seventh century CE). If so, the first element would derive from an older *retō, from Proto-Indo-European *Hreth₂- (cf. Proto-Celtic *reteti), and the Classical rotundus would reflect later influence from rota(“wheel”).[1] Both theories regardless trace back to *Hreth₂-.

Pronunciation

• (Classical Latin) IPA(key): [rɔˈtʊn.dʊs]
• (modern Italianate Ecclesiastical) IPA(key): [roˈtun.dus]

Adjective
rotundus (feminine rotunda, neuter rotundum, comparative rotundior, superlative rotundissimus); first/second-declensionadjective

• round, circular
• spherical, rotund
• (figuratively) rounded, perfect
• (figuratively, of speech) polished, elegant

Latin?

You do know that a ball can be described BOTH as being round AND a sphere, right?

You do know that a ball can be described BOTH as being round AND a sphere, right?

... and, conversely, that round can refer to either a wheel (flat disc) or a sphere also?  Your point?

No, a sphere pertains to three dimensions. A three dimensional item always contains a two-dimensional "round" description. Not vice versa.

The context of use means everything.

No, the context is merely that which you decide to "read into" it, your personal "eisegesis".  I'm not sure what the babble above means.  What I said is that the word "rotundus" in Latin done not necessarily mean a sphere, but could be something two-dimensional.  St. Thomas simply uses the Latin "rotundus", which globers claim refers to a sphere, but that's not necessarily true, and there's no "context" in the citation above that requires that it be read as a "sphere" other than your own wishful thinking and confirmation bias, where you imagine that he used a word that means "sphere", which he did not.

No one has yet provided a citation of the Latin for the Commentary on Aristotle that someone translated as "sphere".

All spheres are round, but not all round things are spheres.  You get that, right?

A "yes" would have been sufficient.