From regular monitoring of the Double Quasar QSO 0957+561 A,B there is now
general agreement on a time delay of about 416 days. This has made it possi
ble to determine the microlensing residual in the light-curve, see Pelt et
al. (1998). We have used two significant microlensing features: 1) A "quiet
" period with a variability less than 0.05 mag lasting about 8 years, and 2
) A change in the residual of 0.25 mag during a time interval of about 5 ye
ars. The first feature gives a lower limit for the lens mass, M, for a give
n normalized source radius, r, whereas the second feature gives an upper li
mit. We have considered the amount of mass in a continuum to be a free para
meter with possible values between 0% (all mass in lenses) and 90%. At a si
gnificance level of 1% the mass can only be constrained within a rather wid
e range (10(-6)M. to 5M.). For the radius R of the source an upper limit of
10(16) cm is found, whereas the normalized source radius r is restricted t
o be smaller than 30. At a level of 10%, however, the range of possible mas
ses is much narrower (2.10(-3) M. to 0.5M.), and the upper limit of R is ab
out 6 . 10(15) cm, whereas the value of r is restricted to be less than 2.
We have used an effective transverse velocity V equal to 600 km s(-1).