First part. Assume that the equations (2) have real roots satisfying x1y1−x2y2=1. By the familiar formula, the roots of the quadratic equations are given by
x1,2=2−p±Kandy1,2=2p±L,
where the real numbers K,L satisfy K2=p2−4q and L2=p2−4r (we choose the signs of K,L in accordance with the labelling of the roots). Thus
1=x1y1−x2y2=4(−p+K)(p+L)−(−p−K)(p−L)=2p(K−L),
whence p=0 and K−L=2/p. Substituting this into the equality
(K+L)(K−L)=K2−L2=(p2−4q)−(p2−4r)=4(r−q)
yields K+L=2p(r−q). From these values of K+L and K−L we obtain K=1/p−p(q−r), so upon squaring, K2=1/p2−2(q−r)p2+(q−r)2. Comparing this with the equality K2=p2−4q, an easy manipulation leads to the desired equation (1).
Second part. Assume that (1) holds. Then clearly p=0. The equation (1) can be rewritten in either of the following two forms,
p4(q−r)2+2p2(q−r)+1=p4−4p2randp4(r−q)2+2p2(r−q)+1=p4−4p2q.
Upon dividing by p2 we find that the discriminants of the equations (2) are equal to
p2−4q=(pp2(r−q)+1)2andp2−4r=(pp2(q−r)+1)2;
hence, they are nonnegative and the (real) roots of (2) have the form (3), where
K=pp2(r−q)+1andL=−pp2(q−r)+1.
The signs of the numbers K and L have been chosen so that (see First Part)
x1y1−x2y2=2p(K−L)=2p⋅(pp2(r−q)+1+pp2(q−r)+1)=1.