Hi again, Bartek, everyone, On Tue, 7 Jan 2003, Bartosz Lew wrote:

what is the true meaning of P(k) ???

It only has a true meaning if: (1) the curvature of comoving space is zero (it is perfectly flat) (2) the topology of comoving space is an equal-sided, right-angled, hypertorus If either (1) or (2) is false, then P(k) is only an *approximation*, and if you forget this you may make errors. (If everyone makes the same error, it is still an error. It's OK politically, but not OK scientifically.) I did a quick search using http://www.google.com "power spectrum density perturbations" and found a fairly standard definition by Pedro Viana: http://nedwww.ipac.caltech.edu/level5/Viana/Viana2_1.html What seems to be missing is the definition of delta_[k (scalar)] in terms of delta_[k (vector)]. My guess is that |delta_[k (scalar)]|^2 in the definition of P(k) should be replaced by < | delta_[k (vector)] | ^2 > where the average is taken over all [k (vector)] pointing in different directions, for the same fixed magnitude of k > 0 . But you should check, where you take an averages or absolute value can make a difference... boud