observed. And in this our countryman, Servius, as it seems, thinks
that there is nothing to be observed except _post_, and he insists
upon it that _liminium_ is a mere extension of the word; as in
_finitimus, legitimus, ceditimus, timus_ has no more meaning than
_tullius_ has in _meditullius_.

But Scaevola, the son of Publius Scaeaevola, thinks the word is a
compound one, so that it is made up of _post_ and _limen_. So that
those things which have been alienated from us, when they have come
into the possession of our enemies, and, as it were, departed from
their own threshold, then when they have returned behind that same
threshold, appear to have returned _postliminio_. By which definition
even the cause of Mancinus may be defended by saying that he returned
_postliminio_,--that he was not surrendered, inasmuch as he was not
received. For that no surrender and no gift can be understood to have
taken place if there has been no reception of it.

IX. We next come to that topic which is derived from those things
which are disposed in some way or other to that thing which is the
subject of discussion. And I said just now that it was divided into
many parts. And the first topic is derived from combination, which the
Greeks call [Greek: sizugia], being a kindred thing to observation,
which we have just been discussing, as, if we were only to understand
that to be rain-water which we saw to have been collected from rain,
Mucius would come, who, because the words _pluna_ and _pluendo_ were
akin, would say that all water ought to be kept out which had been
increased by raining. But when an argument is derived from a genus,
then it will not be necessary to trace it back to its origin, we may
often stop on this side of that point, provided that which is deduced
is higher than that for which it is deduced, as, "Rain water in its
ultimate genus is that which descends from heaven and is increased by
showers," but in reference to its more proximate sense, under which
the right of keeping it off is comprised, the genus is, mischievous
rain water. The subordinate species of that genus are waters which
injure through a natural defect of the place, or those which are
injurious on account of the works of man: for one of these kinds may
be restrained by an arbitrator, but not the other.

Again, this argumentation is handled very advantageously, which is
derived from a species when you pursue all the separate parts by
tracing them back t