178
APOLLONIUS OF PERGA
the value of A in order that the solution may be possible.
Apollonius begins by stating the limiting case, saying that we
obtain a solution in a special manner in the case where M is
the middle point of CD, so that the rectangle CM. MB or
GB'. AD has its maximum value.
The corresponding limiting value of A is determined by
finding the corresponding position of D or M.
We have B'G:MD = CM: AD, as before,
= B'M.MA ;
whence, since MD = CM,
B'G: B'M = GM-.MA
= B'M: B'A,
so that B'M 2 = B'G. B'A.
Thus M is found and therefore D also.
According, therefore, as A is less or greater than the par
ticular value of OG: AD thus determined, Apollonius finds no
solution or two solutions.
Further, we have
AD = B'A + B'G- (B'D + B'G)
= B'A+B'G-2B'M
= B'A + B'G— 2 VB'A . B'G.
If then we refer the various points to a system of co
ordinates in which B'A, B'JSG are the axes of x and y, and if
we denote 0 by (x, y) and the length B'A by h,
A = OG / AD = y/{h + x — 2 Vhx).
If we suppose Apollonius to have used these results for the
parabola, he cannot have failed to observe that the limiting
case described is that in which 0 is on the parabola, while
N'OM is the tangent at 0 ; for, as above,
B'M: B'A = B'G: B'M = N'O : N'M, by parallels,
so that B'A, N'M are divided at M, 0 respectively in the same
proportion.