HIPPOCRATES’S QUADRATURE OF LUNES 199
this third hypothesis the fallacy would lie in the fact that
the sum of the circle and the lune is squared, and not the
circle alone.’
If, however, the reference of Aristotle was really to Hip
pocrates’s last quadrature alone, Hippocrates was obviously
misjudged; there is no fallacy in it, nor is Hippocrates likely
to have deceived himself as to what his proof actually
amounted to.
In the above reproduction of the extract from Eudemus
I have marked by italics the passages where the writer follows
the ancient fashion of describing points, lines, angles, &c., with
reference to the letters in the figure: the ancient practice was
to write to arj/ieiov k<p’ <p (or e<p’ ov) K, the (point) on which (is)
the letter K, instead of the shorter form to K arjpGov, the
point K, used by Euclid and later geometers; rj 1(f) rj AB
(evOeia), the straight line on which (are the letters AB, for
■}] AB {eiiOeia}, the straight line AB; to rpiycovov to k<f)’ ov
EZH, the triangle on which (are the letters) EFG, instead of
to EZH Tpiyoivov, the triangle EFG; and so on. Some have
assumed that, where the longer archaic form, instead of the
shorter Euclidean, is used, Eudemus must be quoting Hippocrates
verbatim; but this is not a safe criterion, because, e.g., Aristotle
himself uses both forms of expression, and there are, on the
other hand, some relics of the archaic form even in Archimedes.
Trigonometry enables us readily to find all the types of
Hippocratean lunes that can
be squared by means of the
straight line and circle. Let
ACB be the external circum
ference, ABB the internal cir
cumference of such a lune,
r, r' the radii, and 0, 0' the
centres of the two arcs, 6, 6'
the halves of the angles sub
tended by the arcs at the centres
respectively.
Now (area of lune)
= (difference of segments ACB, ABB)
= (sector OAGB-AAOB)-(sector 0'ABB —A AO'B)
— r 2 Q — r' 2 6' + \ (r' 2 sin 2 6' — r 2 sin 2 6).