The Vedanta Sutras of Badarayana, Commentary by Sankara (SBE38), tr. by George Thibaut  at sacred-texts.com
19. In the same (Sâkhâ also) it is thus (i.e. there is unity of vidyâ), on account of the non-difference (of the object of meditation).
In the Agnirahasya forming part of the Vâgasaneyi-sâkhâ there is a vidyâ called the Sândilya-vidyâ, in which we meet with the following statement of particulars, 'Let him meditate on the Self which consists of mind, which has the prâna for its body and light for its form,' &c.--In the Brihad-âranyaka again, which belongs to the same Sâkhâ, we read (V, 10,6), 'That person consisting of mind, whose being is light, is within the heart, small like a grain of rice or barley. He is the ruler of all, the Lord of all--he rules all this whatsoever exists.'--A doubt here presents itself whether these two passages are to be taken as one vidyâ in which the particulars mentioned in either text are to be combined or not.
The pûrvapakshin maintains that we have to do with two separate vidyâs whose particulars cannot be combined. For
otherwise, he argues, the text could not be cleared from the reproach of useless repetition. As long as we have to do with texts belonging to different Sâkhâs we can rebut the charge of useless repetition by pointing to the fact that the texts are read and known by separate classes of men; we can then ascertain the unity of the vidyâs and combine the particulars mentioned in one text only with those mentioned in the others; so e.g. in the colloquy of the prânas. On the other hand, texts belonging to one and the same Sâkhâ cannot be freed from the reproach of tautology as the same persons study and know them, and passages occurring in different places cannot therefore be combined into one vidyâ. Nor can we make out a separate position for each of the texts of the latter kind by saying that it is the task of one text to enjoin the vidyâ and that of the other to enjoin the particulars of the vidyâ. For in that case each of the two passages would mention only such particulars as are not mentioned in the other one; while as a matter of fact particulars common to both as well as not common to both are mentioned in each. Hence the particulars of the one passage are not to be combined with those of the other.
To this we make the following reply. Just as passages met with in different Sâkhâs form one vidyâ in which the different particulars are to be combined, so the two passages under discussion also, although belonging to one and the same Sâkhâ, constitute one vidyâ only, since the object of meditation is the same in both. For as such we recognise Brahman possessing certain qualities such as consisting of mind and so on. Now we know that the object constitutes the character of a meditation; as long as there is no difference of character we cannot determine difference of vidyâ; and if there is no difference of vidyâ the particulars mentioned in different places cannot be held apart.--But has it not been demonstrated above that the vidyâs have to be held apart, as otherwise tautology would arise?--Tautology does not result, we reply, because the two passages may be understood to have each its particular meaning, one of them enjoining the vidyâ, and the other the particulars of the vidyâ.--But in that case the Brihad-âranyaka ought to
mention only those points which are not mentioned in the Agnirahasya, as e.g. 'he is the Lord of all;' while it ought not to mention what is already mentioned in the Agnirahasya, as e.g. the Selfs consisting of mind!--Not so, we reply. Only the repetition, in one passage, of what is already mentioned in the other passage enables us to recognise the vidyâ. The Brihad-âranyaka-passage, by mentioning some common qualities, first enables us to recognise the Sândilya-vidyâ, and then teaches certain particulars with reference to the latter; how otherwise should we know that the Bri.--passage is meant to enjoin particulars for the Sândilya-vidyâ? Moreover, as in a passage which has a purpose of its own in so far as it teaches something not yet established, a reference to something already established is justified on the ground of its being a (so-called) nityânuvâda, we cannot overlook the recognition (of the identity of the passage with another one) which is rendered possible through that anuvâda. Hence, although the two passages belong to one and the same Sâkhâ, they yet constitute one vidyâ only, and their particulars have to be combined into one whole.