their imperfections, the similes of the reflected sun, of the ether
limited by jars, &c., are applicable to it (18).--Should it be said that
the illustration is not an appropriate one, because the sun is
apprehended in the water erroneously only while the antaryâmin really
abides within all things, and therefore must be viewed as sharing their
defects (19); we reply that what the simile means to negative is merely
that Brahman should, owing to its inherence in many places, participate
in the increase, decrease, and so on, of its abodes. On this view both
similes are appropriate (20).--Analogous similes we observe to be
employed in ordinary life, as when we compare a man to a lion (21).

Sûtras 22-30 constitute, according to /S/a@nkara, a new adhikara/n/a
(VI), whose object it is to show that the clause 'not so, not so' (neti
neti; B/ri/hadâr) negatives, not Brahman itself, but only the two forms
of Brahman described in the preceding part of the chapter. Sûtras 23-26
further dwell on Brahman being in reality devoid of all distinctive
attributes which are altogether due to the upâdhis. The last four Sûtras
return to the question how, Brahman being one only, the souls are in so
many places spoken of as different from it, and, two explanatory
hypotheses having been rejected, the conclusion is arrived at that all
difference is unreal, due to fictitious limiting adjuncts.

According to Rámânuja, Sûtras 22 ff. continue the discussion started in
Sûtra 11. How, the question is asked, can the ubhayali@ngatva of Brahman
be maintained considering that the 'not so, not so' of the
B/ri/hadâra/n/yaka denies of Brahman all the previously mentioned modes
(prakâra), so that it can only be called that which is (sanmâtra)?--The
reply given in Sûtra 22 is that 'not so, not so' does not deny of
Brahman the distinctive qualities or modes declared previously (for it
would be senseless at first to teach them, and finally to deny them