S'mon said:
In "The Evolution of Civilisation" (pub 1921) which I am currently reading, it is, as far as I can tell. He uses "Aryan" to mean the white race(s) originating in the Caucasus region, the group(s) commonly referred to as Caucasian today.
That just goes to show you the strength of the assumed ties between Indo-European and the "Nordic" physical type that were assumed at the time, and the confusion that ensued because many authors didn't differentiate between the linguistic and assumed physical groups. In 1921, there was no general consensus that the "Aryans" originated in the Caucasus region, though--many people believed they originated in Northern Europe, others in the Himalayas, others that they were descendents of the Sumerians--there were as many theories as there were academics in the field. There were always other white race(s) that were non-"Aryan", ironically, including the Caucasian Georgians, Kartvelians, etc.
I've never heard usage like you're describing--that author using terms in a way that, in my experience, would have been controversial and oddball--or at the very least imprecise--even at the time he wrote.
Interestingly, the leading theory of Indo-European origins today puts them in the Pontic-Caspian steppes, which is at least adjacent to the Caucasus.
S'mon said:
When I said "textbooks" I probably erred, I didn't mean works aimed at a specialist academic audience, I meant popular textbooks aimed at the general public. The general public certainly does use "Caucasian", esp if they don't want to say "white".
OK, that makes a more sense to me then.

He was probably just speaking generally and slightly imprecisely. At the time, most of the "Caucasian" type folks were also "Aryan"--white people by and large have an Indo-European linguistic heritage. That's still true, it's just that nobody uses Aryan that way anymore with the exception of some "white pride" organizations that still cling to pre-WW2 anthropology, and a few other contries who still cling to terminology that's been outdated in most other Western research.
In a way it's too bad--Indo-European, while much more precise, is also much more cumbersome than Aryan. But the combination of Aryan only being "proved" as a common term amongst the Indo-Irandian branch of Indo-European and the stigma attached to the term thanks to the Nazis, there's no way that it'd still be used in the way it used to be.