s
coming under it are not seen, either when these are only remembered,
or when some such attribute is seen which resembles some
other attribute seen before, or when a thing is seen in one way
but appears in another, or when what is seen is not definitely
grasped, whether rightly seen or not. He then discusses the question
whether sound is eternal or non-eternal and gives his reasons
to show that it is non-eternal, but concludes the discussion with
a number of other reasons proving that it is eternal.

The first chapter of the third book is entirely devoted to the
inference of the existence of soul from the fact that there must
be some substance in which knowledge produced by the contact
of the senses and their object inheres.

The knowledge of sense-objects (_indriyartha_) is the reason by

__________________________________________________________________

[Footnote 1: I have differed from _Upaskara_ in interpreting
"_sa@mjnakarma_" in II. i. 18, 19 as a genitive compound while
_Upaskara_ makes it a _dvandva_ compound. Upaskara's interpretation
seems to be far-fetched. He wants to twist it into an argument for
the existence of God.]

[Footnote 2: This interpretation is according to S'a@nkara Mis'ra's
_Upaskara._]

289

which we can infer the existence of something different from the
senses and the objects which appear in connection with them. The
types of inferences referred to are (1) inference of non-existence of
some things from the existence of some things, (2) of the existence
of some things from the non-existence of some things, (3) of the
existence of some things from the existence of others. In all
these cases inference is possible only when the two are known to
be connected with each other (_prasiddhipurvakatvat apades'asya_) [Footnote
ref 1]. When such a connection does not exist or is doubtful, we have
_anapades'a_ (fallacious middle) and _sandigdha_ (doubtful middle);
thus, it is a horse because it has a horn, or it is a cow because it
has a horn are examples of fallacious reason. The inference of
soul from the cognition produced by the contact of soul, senses
and objects is not fallacious in the above way. The inference of
the existence of the soul in others may be made in a similar way
in which the existence of one's own soul is inferred [Footnote ref 2], i.e.
by virtue of the existence of movement and cessation of movement. In
the second chapter it is said that the fact that there is cognition