of the discussion, i.e. that being which in the end is declared to be Brahman is referred to by means of a special word, in most cases a nominative form [12].  From this rule there is in the preceding part of the adhyaya only one real exception, viz. in I, 2, 1, which possibly may be due to the fact that there a new pada begins, and it therefore was considered superfluous to indicate the introduction of a new topic by a special word.  The exception supplied by I, 3, 19 is only an apparent one; for, as remarked above, Sutra 19 does not in reality begin a new adhikara/n/a.  A few exceptions occurring later on will be noticed in their places.—­Now neither Sutra 22 nor Sutra 23 contains any word intimating that a new Vedic passage is being taken into consideration, and hence it appears preferable to look upon them, with Ramanuja, as continuing the topic of the preceding adhikara/n/a.—­This conclusion receives an additional confirmation from the position of the next adhikara/n/a, which treats of the being ‘a span long’ mentioned in Ka/th/a Up.  II, 4, 12; for the reason of this latter passage being considered here is almost certainly the reference to the alpa/s/ruti in Sutra 21, and, if so, the a@ngush/th/amatra properly constitutes the subject of the adhikara/n/a immediately following on Adhik.  V, VI; which, in its turn, implies that Sutras 22, 23 do not form an independent adhikara/n/a.—­The two next adhikara/n/as are digressions, and do not refer to special Vedic passages.—­Sutra 39 forms a new adhikara/n/a, according to Sa@nkara, but not according to Ramanuja, whose opinion seems again to be countenanced by the fact that the Sutra does not exhibit any word indicative of a new topic.  The same difference of opinion prevails with regard to Sutra 40, and it appears from the translation of the Sutra given above, according to Ramanuja’s view, that ‘jyoti/h/’ need not be taken as a nominative.—­The last two adhikara/n/as finally refer, according to Ramanuja, to one Chandogya passage only, and here also we have to notice that Sutra 42 does not comprise any word intimating that a new passage is about to be discussed.

From all this we seem entitled to draw the following conclusions.  The Vedic passages discussed in the three first padas of the Vedanta-sutras comprise all the doubtful—­or at any rate all the more important doubtful—­passages from the Chandogya Upanishad.  These passages are arranged in the order in which the text of the Upanishad exhibits them.  Passages from other Upanishads are discussed as opportunities offer, there being always a special reason why a certain Chandogya passage is followed by a certain passage from some other Upanishad.  Those reasons can be assigned with sufficient certainty in a number of cases although not in all, and from among those passages whose introduction cannot be satisfactorily accounted for some are eliminated by our following the subdivision of the Sutras into adhikara/n/as adopted by Ramanuja, a subdivision countenanced by the external form of the Sutras.