g in the relation of identity, the assumption of
the samavaya relation has really no purport.

In what manner again do you--who maintain that the cause and the effect
are joined by the samavaya relation--assume a substance consisting of
parts which is an effect to abide in its causes, i.e. in the material
parts of which it consists? Does it abide in all the parts taken
together or in each particular part?--If you say that it abides in all
parts together, it follows that the whole as such cannot be perceived,
as it is impossible that all the parts should be in contact with the
organs of perception. (And let it not be objected that the whole may be
apprehended through some of the parts only), for manyness which abides
in all its substrates together (i.e. in all the many things), is not
apprehended so long as only some of those substrates are
apprehended.--Let it then be assumed that the whole abides in all the
parts by the mediation of intervening aggregates of parts[296].--In that
case, we reply, we should have to assume other parts in addition to the
primary originative parts of the whole, in order that by means of those
other parts the whole could abide in the primary parts in the manner
indicated by you. For we see (that one thing which abides in another
abides there by means of parts different from those of that other
thing), that the sword, for instance, pervades the sheath by means of
parts different from the parts of the sheath. But an assumption of that
kind would lead us into a regressus in infinitum, because in order to
explain how the whole abides in certain given parts we should always
have to assume further parts[297].--Well, then, let us maintain the
second alternative, viz. that the whole abides in each particular
part.--That also cannot be admitted; for if the whole is present in one
part it cannot be present in other parts also; not any more than
Devadatta can be present in /S/rughna and in Pa/t/aliputra on one and
the same day. If the whole were present in more than one part, several
wholes would result, comparable to Devadatta and Yaj/n/adatta, who, as
being two different persons, may live one of them at /S/rughna and the
other at Pa/t/aliputra.--If the opponent should rejoin that the whole
may be fully present in each part, just as the generic character of the
cow is fully present in each individual cow; we point out that the
generic attributes of the cow are visibly perceived in each individual
cow,