The Vedanta Sutras, commentary by Sankaracharya (SBE34), tr. by George Thibaut  at sacred-texts.com
22. The âkâsa, i.e. ether (is Brahman) on account of characteristic marks (of the latter being mentioned).
In the Khândogya (I, 9) the following passage is met with, 'What is the origin of this world?' 'Ether,' he replied. 'For all these beings take their rise from the ether only, and return into the ether. Ether is greater than these, ether is their rest.'--Here the following doubt arises. Does the word 'ether' denote the highest Brahman or the elemental ether?--Whence the doubt?--Because the word is seen to be used in both senses. Its use in the sense of 'elemental ether' is well established in ordinary as well as in Vedic speech;
and, on the other hand, we see that it is sometimes used to denote Brahman, viz. in cases where we ascertain, either from some complementary sentence or from the fact of special qualities being mentioned, that Brahman is meant. So, for instance, Taitt. Up. II, 7, 'If that bliss existed not in the ether;' and Kh. Up. VIII, 14, 'That which is called ether is the revealer of all forms and names; that within which forms and names are 1 that is Brahman.' Hence the doubt.--Which sense is then to be adopted in our case?--The sense of elemental ether, the pûrvapakshin replies; because this sense belongs to the word more commonly, and therefore presents itself to the mind more readily. The word 'ether' cannot be taken in both senses equally, because that would involve a (faulty) attribution of several meanings to one and the same word. Hence the term 'ether' applies to Brahman in a secondary (metaphorical) sense only; on account of Brahman being in many of its attributes, such as all pervadingness and the like, similar to ether. The rule is, that when the primary sense of a word is possible, the word must not be taken in a secondary sense. And in the passage under discussion only the primary sense of the word 'ether' is admissible. Should it be objected that, if we refer the passage under discussion to the elemental ether, a complementary passage ('for all these beings take their rise from the ether only, &c.') cannot be satisfactorily accounted for; we reply that the elemental ether also may be represented as a cause, viz. of air, fire, &c. in due succession. For we read in Scripture (Taitt. Up. II, 1), 'From that Self sprang ether, from ether air, from air fire, and so on.' The qualities also of being greater and of being a place of rest may be ascribed to the elemental ether, if we consider its relations to all other beings. Therefore we conclude that the word 'ether' here denotes the elemental ether.
To this we reply as follows:--The word ether must here be taken to denote Brahman, on account of characteristic marks of the latter being mentioned. For the sentence,
[paragraph continues] 'All these beings take their rise from the ether only,' clearly indicates the highest Brahman, since all Vedânta-texts agree in definitely declaring that all beings spring from the highest Brahman.--But, the opponent may say, we have shown that the elemental ether also may be represented as the cause, viz. of air, fire, and the other elements in due succession.--We admit this. But still there remains the difficulty, that, unless we understand the word to apply to the fundamental cause of all, viz. Brahman, the affirmation contained in the word 'only' and the qualification expressed by the word 'all' (in 'all beings') would be out of place. Moreover, the clause, 'They return into the ether,' again points to Brahman, and so likewise the phrase, 'Ether is greater than these, ether is their rest;' for absolute superiority in point of greatness Scripture attributes to the highest Self only; cp. Kh. Up. III, 14, 3, 'Greater than the earth, greater than the sky, greater than heaven, greater than all these worlds.' The quality of being a place of rest likewise agrees best with the highest Brahman, on account of its being the highest cause. This is confirmed by the following scriptural passage: 'Knowledge and bliss is Brahman, it is the rest of him who gives gifts' (Bri. Up. III, 9, 28). Moreover, Gaivali finding fault with the doctrine of Sâlâvatya, on account of (his sâman) having an end (Kh. Up. I, 8, 8), and wishing to proclaim something that has no end chooses the ether, and then, having identified the ether with the Udgîtha, concludes, 'He is the Udgîtha greater than great; he is without end.' Now this endlessness is a characteristic mark of Brahman. To the remark that the sense of 'elemental ether' presents itself to the mind more readily, because it is the better established sense of the word âkâsa, we reply, that, although it may present itself to the mind first, yet it is not to be accepted, because we see that qualities of Brahman are mentioned in the complementary sentences. That the word âkâsa is also used to denote Brahman has been shown already; cp. such passages as, 'Ether is the revealer of all names and forms.' We see, moreover, that various synonyma of âkâsa are employed to denote Brahman. So, for instance, Rik Samh.
[paragraph continues] I, 164, 39, 'In which the Vedas are 1, in the Imperishable one (i. e. Brahman), the highest, the ether (vyoman), on which all gods have their seat.' And Taitt. Up. III, 6, 'This is the knowledge of Bhrigu and Varuna, founded on the highest ether (vyoman).' And again, 'Om, ka is Brahman, ether (kha) is Brahman' (Kh. Up. IV, 10, 5), and 'the old ether' (Bri. Up. V, 1) 2. And other similar passages. On account of the force of the complementary passage we are justified in deciding that the word 'ether,' although occurring in the beginning of the passage, refers to Brahman. The case is analogous to that of the sentence, 'Agni (lit. the fire) studies a chapter,' where the word agni, although occurring in the beginning, is at once seen to denote a boy 3. It is therefore settled that the word 'ether' denotes Brahman.
82:1 Or else 'that which is within forms and names.'
84:1 Viz. as intimating it. Thus Ân. Gi. and Go. Ân. against the accent of rikáh. Sâyana explains rikáh as genitive.
84:2 Omkârasya pratîkatvena vâkakatvena lakshakatvena vâ brahmatvam uktam, om iti, kam sukham tasyârthendriyayogagatvam vârayitum kham iti, tasya bhûtâkasatvam vyâseddhum purânam ity uktam. Ân. Gi.
84:3 The doubt about the meaning of a word is preferably to be decided by means of a reference to preceding passages; where that is not possible (the doubtful word occurring at the beginning of some new chapter) complementary, i.e. subsequent passages have to be taken into consideration.