merchant, and your cask; but the contents of the latter are not to be
wine, but assorted seeds. On the top is wheat, next beans, then barley,
below that lentils, then peas--and other kinds yet. You go to buy seeds,
and he takes some wheat out of that layer, and puts it in your hand as a
sample; now, could you tell by looking at that whether the peas were
Sound, the lentils tender, and the beans full?

_Her_. Impossible.

_Ly_. No more can you tell the quality of a philosophy from the first
statements of its professor; it is not uniform, like the wine to which you
compared it, claiming that it must resemble the sample glass; it is
heterogeneous, and it had better not be cursorily tested. If you buy bad
wine, the loss is limited to a few pence; but to rot with the common herd
(in your own words) is not so light a loss. Moreover, your man who wants
to drink up the cask as a preliminary to buying a pint will injure the
merchant, with his dubious sampling; but philosophy knows no such danger;
you may drink your fill, but this cask grows no emptier, and its owner
suffers no loss. It is cut and come again here; we have the converse of
the Danaids' cask; that would not hold what was put into it; it ran
straight through; but here, the more you take away, the more remains.

And I have another similar remark to make about these specimen drops of
philosophy. Do not fancy I am libelling it, if I say it is like hemlock,
aconite, or other deadly poison. Those too, though they have death in
them, will not kill if a man scrapes off the tiniest particle with the
edge of his nail and tastes it; if they are not taken in the right
quantity, the right manner, and the right vehicle, the taker will not
die; you were wrong in claiming that the least possible quantity is
enough to base a generalization on.