Measurements of the dynamic structure factor performed on the MARI spe
ctrometer at ISIS on the metallic glass Fe85B15 are briefly discussed.
The spectra were fitted by a sum of two Gaussian functions, one repre
senting the elastic peak and one the inelastic part of the spectrum, a
nd a linear background in order to obtain a 'dispersion curve' for the
collective excitations at small Q. Several interesting features are n
oticeable. (i) At 295 K, the measured data at small Q are very well de
scribed by the sound velocity calculated from the Young modulus while
at 20 K this is not the case. At this temperature, the data do not see
m to extrapolate to the origin but to a finite energy. (ii) The 'dispe
rsion curve' is asymmetric with regard to Q0/2.