LOST GEOMETRICAL WORKS
431
We now come to the lost works belonging to higher
geometry. The most important was evidently
(J3) The For isms.
Our only source of information about the nature and con
tents of the Porisms is Pappus. In his general preface about
the books composing the Treasury of Analysis Pappus writes
as follows 1 (I put in square brackets the words bracketed by
Hultsch).
‘ After the Tangencies (of Apollonius) come, in three Books,
the Porisms of Euclid, a collection [in the view of many] most
ingeniously devised for the analysis of the more weighty
problems, [and] although nature presents an unlimited num
ber of such porisms, [they have added nothing to what was
originally written by Euclid, except that some before my time
have shown their want of taste by adding to a few (of the
propositions) second proofs, each (proposition) admitting of
a definite number of demonstrations, as we have shown, and
Euclid having given one for each, namely that which is the
most lucid. These porisms embody a theory subtle, natural,
necessary, and of considerable generality, which is fascinating
to those who can see and produce results].
‘ Now all the varieties of porisms belong, neither to theorems
nor problems, but to a species occupying a sort of intermediate
position [so that their enunciations can be formed like those of
either theorems or problems], the result being that, of the great
number of geometers, some regarded them as of the class of
theorems, and others of problems, looking only to the form of
the proposition. But that the ancients knew better the differ
ence between these three things is clear from the definitions.
For they said that a theorem is that which is proposed with a
view to the demonstration of the very thing proposed, a pro
blem that which is thrown out with a view to the construction
of the very thing proposed, and a porism that which is pro
posed with a view to the producing of the very thing proposed.
[But this definition of the porism was changed by the more
recent writers who could not produce everything, but used
these elements and proved only the fact that that which is
sought really exists, but did not produce it, and were accord
ingly confuted by the definition and the whole doctrine. They
based their definition on an incidental characteristic, thus :
A porism is that which falls short of a locus-theorem in
1 Pappus, vii, pp. 648-60.